Volume 13 • Number 7 • 2010 VA L U E I N H E A LT H ISPOR 13th Annual European Congress Research Abstracts PODIUM SESSION I: HTA IN CANCER STUDIES AND METHODS CN1 EFFICACY OF THE COMBINATION BEVACIZUMAB PLUS CHEMOTHERAPY (BEV-CT) COMPARED TO CT ALONE IN PREVIOUSLY UNTREATED LOCALLY ADVANCED OR METASTATIC NON-SMALL CELL LUNG CANCER (NSCLC): SYSTEMATIC REVIEW (SR) AND META-ANALYSIS (MA) Botrel TEA, Clark O, Clark LGO, Paladini L, Faleiros E, Pegoretti B MedInsight-Evidências, Campinas, Brazil OBJECTIVES: To perform SR with MA of all randomized controlled trials (RCT) comparing the efficacy of BEV-CT versus CT alone in previously untreated locally advanced or metastatic NSCLC. METHODS: We searched MEDLINE, EMBASE, LILACS, and CENTRAL among others. Primary end points were overall survival (OS) and progression-free survival (PFS). Adverse events (AE) were analyzed. Extracted data were combined using hazard ratio (HR) or risk ratio (RR) with 95% confidence intervals (CI 95%). RESULTS: 544 references were identified and screened, three trials comprising 2020 patients were included. Overall response rate (RR = 0.53; CI 95% = 0.44 to 0.64; P < 0.00001) and PFS were higher in BEV-CT (HR = 0.71, CI 95% = 0.63 to 0.80; P < 0.00001), however with significant heterogeneity (χ2 = 4.9, df = 2 [P = 0.09]; I2 = 59%) and (χ2 = 4.33, df = 2 [P = 0.11]; I2 = 54%), respectively. Random-efffects model analysis favored BEV-CT. OS was higher in BEV-CT but with significant heterogeneity (χ2 = 5.92, df = 3 [P = 0.12]; I2 = 49%) and random-effects model analysis was not statistically significant (HR = 0.86, CI 95% = 0.71 to 1.05; P = 0.15). Neutropenia (RR = 0.77; CI 95% = 0.65 to 0.91; P = 0.002) and febrile neutropenia (RR = 0.42; CI 95% = 0.22 to 0.81; P = 0.009) were higher on BEV-CT. Rates of anemia (RR = 1.41; CI 95% = 0.93 to 2.13; P = 0.1) and thrombocytopenia (RR = 0.91; CI 95% = 0.69 to 1.20; P = 0.50) were similar. Non-hematologic toxicities were higher on BEV-CT: hemoptysis (RR = 0.28; CI 95% = 0.09 to 0.90; P = 0.03), hypertension (RR = 0.15; CI 95% = 0.07 to 0.30; P < 0.00001), proteinuria (RR = 0.05; CI 95% = 0.01 to 0.41; P = 0.005), venous thromboembolic events (RR = 0.87; CI 95% = 0.51 to 1.47; P = 0.6), vomiting (RR = 0.41; CI 95% = 0.22 to 0.77; P = 0.005), rash (RR = 0.19; CI 95% = 0.04 to 0.88; P = 0.03), epistaxis (RR = 0.32; CI 95% = 0.03 to 3.10; P = 0.33) and bleeding events (RR = 0.27; CI 95% = 0.13 to 0.56; P = 0.0004). CONCLUSIONS: The combination BEV-CT increased the response rate and PFS in patients with NSCLC. Benefits in overall survival remain uncertain, and toxicity rates were higher in the combination group. CN2 REAL-WORLD COST-EFFECTIVENESS OF OXALIPLATIN IN STAGE III COLON CANCER: A SYNTHESIS OF TRIAL DATA WITH DATA FROM DAILY CLINICAL PRACTICE van Gils C, de Groot S, Redekop W, Uyl-de Groot C Erasmus University, Rotterdam, The Netherlands OBJECTIVES: Previous cost-effectiveness analyses of oxaliplatin have been based on randomized trial settings which may not reflect actual daily practice. The objective of this study was to examine the real-world cost-effectiveness of oxaliplatin plus fluoropyrimidines versus fluoropyrimidines only, as adjuvant treatment of colon cancer. METHODS: A Markov model was developed to estimate lifetime costs and qualityadjusted life-years (QALYs) from a hospital perspective. Dutch real-world (RW) population-based data on use, costs, and disease-free survival after oxaliplatin use were combined with published efficacy data from the pivotal clinical registration trial (MOSAIC trial). Eighty-two percent of the patients in the RW study fulfilled the MOSAIC trial eligibility criteria (“eligibles”); the other 18% (“ineligibles”) had a poorer prognosis. The efficacy of the comparator was modelled using MOSAIC trial results. Cost-effectiveness analyses (CEAs) were performed for four different scenarios: 1) CEA based on MOSAIC trial patients; 2) CEA using eligible RW patients; 3) CEA using both eligibles and ineligibles, assuming that oxaliplatin had an equal effect in both groups; and 4) CEA using eligibles and ineligibles, assuming oxaliplatin had no effect among ineligibles. RESULTS: MOSAIC and eligible RW patients had similar 2-year disease-free survivals (79% vs. 78%). Oxaliplatin showed an incremental QALY gain of 0.86, 0.73, 0.81, and 0.60, and incremental costs of c13,105, c13,278, c13,225, and c13,456 in scenarios 1 to 4, respectively. The corresponding incremental cost-effectiveness ratios (ICERs) were c15,185, c18,115, c16,254, and c22,387 in scenarios 1 to 4, respectively. Sensitivity analyses of input parameters and model assumptions produced only minimal differences in the estimated ICERs showing the robustness of the model results. CONCLUSIONS: The real-world cost-effectiveness of oxaliplatin plus fluoropyrimidine versus fluoropyrimidine for the treatment of colon cancer can be estimated using different scenarios. We found that the various estimates were very similar, and all suggest that oxaliplatin is cost-effective. CN3 COST OF ABSENTEEISM DUE TO CANCER IN POLAND 1 1 2 Macioch T , Hermanowski T , Adamowicz O 1 Department of Pharmacoeconomics, Medical University of Warsaw, Warsaw, Poland; 2 HealthQuest sp.z o.o., Warsaw, Poland OBJECTIVES: Cancer is a leading cause of death worldwide and major health problem. It’s also a huge problem to worldwide economy due to high both direct and indirect costs. The aim of this study was to evaluate absenteeism costs of cancer in Poland in 2007. METHODS: Sickness absence data and data on disability were derived by ZUS (Social Insurance Institution). Costs of lost productivity due to premature death were estimated based on regional register data on cancer mortality (KRN—National Cancer Registry). Absenteeism costs were estimated based on gross value added per employee derived by Central Statistical Office (GUS). The analyses of indirect costs due to sickness absence were based on an assumption that number of missed days includes only working days (226 days per year). Human capital approach was used to estimate the absenteeism costs. Values are presented in Euro (exchange rate: 1 EUR = 4.10 PLN). RESULTS: Costs of lost productivity due to illness and disability were estimated to amount of c1081 million (c451 and c630 million, respectively). Costs of lost productivity due to premature death were estimated to amount of c4692 million and were twice as high in male population compare to female population (c3200 and c1492 million, respectively). The highest costs of lost productivity due to premature death were related to lung cancer (c1331 million) followed by brain (c356 million), stomach (c314 million), and breast cancer (c259 million). Lung cancer was the leading cost of lost productivity due to premature death in male population (c917 million), while breast cancer dominated in female population (c255 million). CONCLUSIONS: Indirect costs of lost productivity due to cancerrelated illness, disability, and premature death are substantial to polish economy and may account for near 0.6% of PKB loss in 2007 year. CN4 HPV-BASED PRIMARY CERVICAL CANCER SCREENING IN GERMANY. COST-EFFECTIVENESS RESULTS FROM A DECISION-ANALYTIC MODELING STUDY Sroczynski G1, Schnell-Inderst P1, Muehlberger N1, Lang K2, Aidelsburger P2, Wasem J3, Mittendorf T4, Engel J5, Hillemanns P6, Petry K-U6, Kraemer A7, Siebert U1 1 Department of Public Health, Information Systems and HTA, UMIT—University for Health Sciences, Medical Informatics and Technology, Hall i.T., Austria; 2CAREM GmbH, Sauerlach, Germany; 3Institute for Healthcare Management, University of Duisburg-Essen, Essen, Germany; 4Institute for Health Economics, University of Hanover, Hanover, Germany; 5 Munich Cancer Registry of the Munich Cancer Centre, Clinic Grosshadern, LudwigMaximilians-University, Munich, Germany; 6Department of Obstetrics and Gynecology, Hanover Medical School, Hanover, Germany; 7School of Public Health, University of Bielefeld, Bielefeld, Germany OBJECTIVES: The objective of this HTA commissioned by the German Agency for Health Technology Assessment (DAHTA@DIMDI) and the Federal Ministry of Health was to systematically evaluate the long-term effectiveness and cost-effectiveness of HPV-based primary cervical cancer screening for the German health care context using a decision analytic approach. METHODS: A Markov model simulating the natural history of cervical cancer was developed and validated for the German health care context. Different screening strategies were evaluated, including cytology alone, HPV testing alone or combined with cytology or with cytological triage for HPV-positive women, and different screening intervals. German clinical, epidemiological and economic data were used. Test accuracy data were retrieved from international metaanalyses. Predicted outcomes were reduction in cervical cancer cases and deaths, life expectancy, and discounted incremental cost-effectiveness ratios (ICER). a perspective of the health care system and 3% annual discount rate were adopted. Extensive sensitivity analyses were performed to evaluate robustness of results. RESULTS: HPVbased screening was more effective than cytology alone, with 71%–97% (depending on screening intervals) relative reduction in cervical cancer compared to 53%–80% for cytology. The ICER ranged between c2,600/LYG (cytology, 5-yr-interval) and c155,500/LYG (annual HPV-testing age 30 yrs, cytology 20–29 yrs). Annual cytology, the current recommended screening strategy in Germany, was dominated by other strategies. Increasing screening start-age to 25 yrs had no relevant loss in effectiveness but resulted in lower costs. CONCLUSIONS: Based on our analyses, HPV-based cervical cancer screening is more effective than cytology and could be cost-effective, when performed at intervals of 2 years or longer. For the German screening context, an optimal screening strategy could be biennial HPV testing starting at age 30 yrs and biennial cytology at the age 25–29 yrs. An extension to a 3-yearly screening interval may be considered for low-risk women with good screening adherence or in populations with low HPV incidence. © 2010, International Society for Pharmacoeconomics and Outcomes Research (ISPOR) 1098-3015/10/A239 A239–A480 A239 A240 PODIUM SESSION I: DEVELOPMENTS IN QUESTIONNAIRES EQ-5D AND SF-6D EQ1 PSYCHOMETRIC COMPARISON OF EQ-5D AND EQ-5D-5L IN STUDENT POPULATION Golicki D1, Zawodnik S1, Janssen MF2, Kiljan A1, Hermanowski T1 1 Department of Pharmacoeconomics, Medical University of Warsaw, Warsaw, Poland; 2 EuroQol Group, Rotterdam, The Netherlands OBJECTIVES: The new five-level version of the EQ-5D is now officially available. a pilot five-level version of EQ-5D questionnaire was successfully tested in patients with cancer. Other studies, using the official EQ-5D-5L, in moderately or severely disabled populations of patients with stroke, diabetes, injury, or psychiatric conditions are under way. Nothing is known about how EQ-5D-5L performs in relatively young and healthy populations. Aim of the study was to compare a polish version of the threelevel EQ-5D questionnaire (3L) with the five-level version (5L) in a student population. METHODS: In March 2010, all students from the Faculty of Pharmacy, Medical University of Warsaw were surveyed with a set of HRQoL questionnaires (5L, EQ-VAS, SF-36, and 3L). We examined percentage of reported problems, proportion and size of logical inconsistencies, ceiling effect, redistribution properties, convergent validity with SF-36 domains, and discriminatory power using Shannon’s indices. RESULTS: Four hundred eighty-five students were approached and 443 responded to all questionnaires (5L, SF-36, and 3L). The domains with the highest and the lowest proportion of reported problems by 5L were Anxiety/Depression (57.1%) and SelfCare (0.2%), respectively. 3L and 5L responses were highly correlated in Pain/Discomfort (Pearsons’s Rho = 0.71), Usual Activities (0.64), and Anxiety/Depression domains (0.64). Mean rate of logical inconsistencies was 3.02%, with 93% of them being level 1 (according to Janssen et al., 2008). The ceiling effect was indentified in 47% of 3L questionnaires and 34% of 5L questionnaires. Absolute informativity was higher for 5L, with similar relative informativity for both instruments. CONCLUSIONS: Results obtained in students support introduction of new 5-level version of EQ-5D in young and healthy populations. EQ-5D-5L appears to be more favorable in terms of ceiling effect and absolute informativity. EQ2 TESTING THE PERFORMANCE OF THE NEWLY DEVELOPED VERSION OF THE EQ-5D WITH 5 LEVELS OF SEVERITY: APPLICATION ON A COHORT OF PATIENTS WITH CHRONIC HEPATIC DISEASES Scalone L1, Ciampichini R2, Fagiuoli S3, Gardini I4, Del Prete A3, Gaeta L3, Fusco F2, Mantovani LG5 1 University of Milano—Bicocca, Monza, Monza e Brianza, Italy; 2Charta Foundation, Milan, Italy; 3Ospedali Riuniti Bergamo, Bergamo, Italy; 4EPAC—Hepatic Patient Association, Vimercate, Milan, Italy; 5Federico II University of Naples, Naples, Italy OBJECTIVES: To assess the performance of the EQ-5D-5L in a naturalistic context targeted to patients with hepatic diseases. Preliminary investigations suggest that the new version of EQ-5D, with five levels (5L) of severity per domain, is promising to assess individuals’ QoL more appropriately than the standard version with three levels. However, further research is encouraged to investigate the EQ-5D-5L properties in different subpopulations and contexts. METHODS: This study was conducted within a project aimed to estimate costs and QoL related to hepatic diseases. The participants self-completed a questionnaire including the 5L descriptive system, the 3L standard descriptive system, and the VAS. The following properties were tested: feasibility (amount of missing answers); amount of inconsistent and consistent responses between 3L and 5L versions; convergent validity with the 3L version and VAS (Spearman’s rank correlation coefficient); discriminatory power (informativity). RESULTS: Data from 426 patients were analyzed: 69% male, 19 to 84 (median = 57) years old. They were affected with: chronic hepatitis C (25.4%), chronic hepatitis B (22.5%), chronic hepatitis B and C (1.2%), cirrhosis (20.9%), liver transplantation (19.0%), hepatic carcinoma (4.5%), nonalcoholic steatohepatitis (1.9%), and other less frequent hepatic diseases. With the standard version (3L), patients reported some or severe problems with mobility (24.2%), self-care (10.4%), usual activities (29.0%), pain/ discomfort (35.5%), and anxiety/depression (37.4%). Median VAS was 70 (15–100). Missing answers were more frequent with the 3L version (6.6% vs. 4.0% of patients). In total, 97.5% of responses were consistent. Convergent validity of 5L-VAS (from −0.35 to −0.57) was similar to the one of 3L-VAS (from −0.41 to −0.56). 3L-5L correlation coefficient ranged from 0.83 to 0.91. Informativity was similar between the two versions. CONCLUSIONS: The EQ-5D-5L version holds promise as a valid extension of the standard 3L version for the assessment of QoL of individuals with hepatic diseases. EQ3 ESTIMATING THE SF-6D VALUE SET FOR A POPULATION BASED SAMPLE OF BRAZILIANS Cruz LN1, Camey SA1, Hoffmann JF1, Brazier J2, Rowen D2, Fleck MP1, Polanczyk CA1 1 Federal University of Rio Grande do Sul (UFRGS), Porto Alegre, Brazil; 2University of Sheffield, Sheffield, UK OBJECTIVES: The SF-6D is a preference-based measure of health developed to estimate utility values from the SF-36. The aim of this study was to estimate preference weights for SF-6D health states representing the preferences of a sample of Southern Brazilian general population. METHODS: A sample of 248 health states defined by the SF-6D has been valued by a sample of Southern Brazilian population using the 13th Euro Abstracts standard gamble (SG) method. SG responses were used to estimate regression models at the individual and mean levels to predict preference values for all SF-6D health states. The models were compared with those described in the UK study. RESULTS: Five hundred twenty-eight participants were interviewed, but 146 (28%) were excluded due to inconsistent SG responses. Data from 382 subjects were used to estimate the models, rendering 2224 health states valuations. All Brazilian models have a large number of significant coefficients and a mean absolute difference between observed and predicted values below 0.07. Inconsistent coefficients have been merged to produce the final recommended model. Compared to UK data, Brazilian health state values were lower, leading to a lower constant term in the models. The best model fitted to Brazilian data was a random effects model using only the main effects variables, different from the preferred British SF-6D mean model, highlighting the importance to adopt a country-specific algorithm in predicting SF-6D health states values. CONCLUSIONS: The results provide the first population-based value set for health states in Brazil, making it possible to generate QALYs for cost-utility studies using regional data. Utility scores based on Brazilian preferences values can be derived from existing SF-36 data sets. EQ4 DESCRIBING AND COMPARING HEALTH-RELATED QUALITY OF LIFE DERIVED FROM EQ-5D AND SF-6D IN A SWEDISH GENERAL POPULATION Andersson D, Brodtkorb TH, Tinghög G Linköping University, Linköping, Östergötland, Sweden OBJECTIVES: Measurement of health-related quality of life (HRQoL) is a crucial aspect of cost-effectiveness analysis. Two of the most widely used instruments for assessing HRQoL are EQ-5D and SF-6D. Previous studies have shown that the sensitivity of these instruments differ depending on the severity of the disease. However, few studies have systematically compared EQ-5D and SF-6D scores in a representative sample of the general population. The objective of this study is to describe and compare HRQoL derived from EQ-5D and SF-6D in a Swedish general population for age, gender, and seven disease groups (respiratory, tumor, endocrine, cardiovascular, orthopedic, mental, and rheumatic diseases). METHODS: The EQ-5D and SF-6D were included in a cross-sectional postal health survey in Östergötland County to a representative sample (n = 6628, age: 18–84). RESULTS: When comparing EQ-5D and SF-6D with regard to age, the mean HRQoL with EQ-5D across all ages was 0.81, ranging from 0.86 (18–29 years) to 0.71 (80–84 years). For SF-6D, mean HRQoL for the same population was 0.74, ranging from 0.75 (18–29 years) to 0.66 (80–84 years). With regard to gender, HRQoL was higher overall for men compared to women. Men had an EQ-5D score of 0.83 and a SF-6D score of 0.79; in contrast, women had an EQ-5D score of 0.79 and a SF-6D score of 0.73. Across disease groups, the mean EQ-5D scores ranged from 0.60 (rheumatic) to 0.75 (endocrine), mean SF-6D index values ranged from 0.64 (rheumatic) to 0.71 (endocrine). CONCLUSION: Although EQ-5D and SF-6D appear to measure similar constructs, our results show that HRQoL varies greatly among the general population depending on chosen instrument. However, when comparing our results across disease groups, differences between the instruments appear less substantial. Our findings could have serious implications for the transparency of cost-effectiveness analysis, if different studies use HRQoL measurements that are not equivalent. PODIUM SESSION I: DEVELOPMENTS IN MODELING METHODOLOGY MO1 STRUCTURAL UNCERTAINTY IN COST-EFFECTIVENESS MODELS OF TREATMENT FOR ALZHEIMER’S DISEASE Peters JL, Hoyle M, Anderson R Peninsula Medical School, Exeter, Devon, UK OBJECTIVES: To discuss and explore important sources of structural uncertainty in a decision model for drug treatment in Alzheimer’s disease. Progression in Alzheimer’s disease can be modeled on various domains (e.g., cognition, function, behavior), each of which can be measured by various scales. Defining disease progression and incorporating a treatment effect can therefore be difficult, requiring many assumptions based on little or inconsistent evidence, resulting in a great deal of structural uncertainty. METHODS: We explored the impact of structural assumptions in a published three-state Markov model of treatment for Alzheimer’s disease. This ultimately led to the development of a novel model. Elements of structural uncertainty included 1) defining the model states; 2) predicting occupancy of the states over time; and 3) allowance for disease progression within a state. RESULTS: Assessment of the stepby-step structural changes to the published model indicated a number of assumptions having a relatively large impact on the cost-effectiveness results, generally resulting in lower costs per quality-adjusted life-year (QALY) associated with the treatments. The cumulative impact of these assumptions was also considerable. However, there is little evidence to inform which assumptions are the most realistic. CONCLUSIONS: Investigation into sources of structural uncertainty has helped to identify which assumptions had the largest impact on the estimated costs per QALY. However, this does not help to reduce the uncertainty in the decision model, but does make the structural uncertainty explicit. Decision-makers are therefore forced to address this type of uncertainty as well as parameter uncertainty. Exploring structural uncertainty also helps to identify gaps in the current evidence base to help understand which assumptions may be the most credible. A241 13th Euro Abstracts MO2 HERD IMMUNITY AS A RESULT IN DYNAMIC AGENT-BASED EPIDEMIC MODELS Miksch F1, Popper N2, Zauner G2, Endel G3, Schiller-Frühwirth I3, Breitenecker F1 1 Vienna University of Technology, Vienna, Austria; 2Dwh Simulation Services, Vienna, Austria; 3 Main Association of Austrian Social Security Institutions, Vienna, Austria OBJECTIVES: Herd immunity describes a phenomenon in the area of communicable diseases. Pathogens are spread by infected persons. Protecting a part of the population—for example via vaccination—lowers the overall appearance of pathogens as these people cannot spread pathogens any more. Not protected people profit by fewer contacts with pathogens, and further, a lower number of infections for them can be expected. Classic Markov model cannot provide herd immunity as a result. In this work, we propose calculations of herd immunity, create a model that is able to simulate epidemics, and show herd immunity dynamically in different states of the model. Appearance of herd immunity is very disputed because it 1) cannot be measured directly in real life and 2) depends on several factors. METHODS: Classic Markov models require herd immunity as a static input parameter that cannot be provided. The developed agent-based model includes single persons with different infection states and a single pathogen. Every agent is part of a social contact model. It is possible to simulate scenarios without vaccinations and with different vaccination strategies. Herd immunity as a result of the dynamic model is calculated as the reduction of the carrier rate of nonvaccinated persons for a certain vaccination strategy compared with the scenario without vaccinations. RESULTS: Results show herd immunity as simulation result depending not only on vaccination strategies but also on other system parameters. Further work extends the social contact structure with places like households, schools, or workplaces that are expected to have an impact on herd immunity as well. CONCLUSIONS: Results can be implemented in systems for calculating new strategies for vaccination programs. Current work considers two or more concurrent serotypes where herd immunity and serotype replacement affects each other. In this case, different definitions of herd immunity are possible. MO3 THE DEVELOPMENT AND VALIDATION OF A DECISION MODEL REPRESENTING THE FULL DISEASE COURSE OF ACUTE MYELOID LEUKEMIA Leunis A1, van Beers EH2, Löwenberg B3, Redekop WK1, Uyl-De Groot CA1 1 Institute for Medical Techonology Assessment (iMTA), Rotterdam, The Netherlands; 2 Skyline Diagnostics BV, Rotterdam, The Netherlands; 3Erasmus University Medical Center, Rotterdam, The Netherlands OBJECTIVES: Acute myeloid leukemia (AML) is a heterogeneous disease, consisting of several subtypes with a variety in prognosis. a new genomics technology, the AML profiler, has been developed that identifies new genetic subtypes. Since no decision model exists that describes the full disease course of AML, the potential cost-effectiveness of this test cannot yet be determined. The aim of this study is to fill this gap and develop and validate a disease progression model for AML. METHODS: The structure of the model and the identification of relevant parameters were based on the literature and expert opinion. All input parameters were estimated from clinical trial data (HOVON data) for patients aged 18 to 60 years. The internal and external validity of the model was evaluated by comparing model-based survival results with the results from HOVON trials and the literature. RESULTS: Important prognostic factors for AML were derived from the literature and expert opinion a microsimulation model (i.e., individual patient sampling) was designed to incorporate all important prognostic factors in the model. The prognostic factors were included as covariates in parametric survival functions for two events: relapse and death. The model combined those survival functions with individual patient data to calculate life-years per patient. The average 5-year survival of the simulated patient cohort was 40%, which is similar to the survival found in HOVON trials and the literature. DISCUSSION: The content validity of the model was achieved by involving clinical experts in the construction of the model. The survival estimated using the model corresponds with those seen elsewhere, suggesting an acceptable level of internal and external validity. Therefore, the model can be used to assess the cost-effectiveness of AML genomics technologies such as the AML profiler. Moreover, the model can be used for other cost-effectiveness analyses in the field of AML. MO4 USING AHP WEIGHTS TO FILL MISSING GAPS IN MARKOV DECISION MODELS Steuten L, Hummel M, Wetering G, Groothuis-Oudshoorn K, Doggen C, Ijzerman M University of Twente, Enschede, The Netherlands OBJECTIVES: We propose to combine the versatility of the analytic hierarchy process (AHP) with the decision-analytic sophistication of health-economic modeling in a new methodology for early technology assessment. As an illustration, we apply this methodology to a new technology to diagnose breast cancer. METHODS: The AHP is a technique for multicriteria analysis, relatively new in the field of technology assessment. It can integrate both quantitative and qualitative criteria in the assessment of alternative technologies. We applied the AHP to prioritize a more versatile set of outcome measures than most Markov models do. These outcome measures include clinical effectiveness and costs, but also weighted estimates of patient comfort and safety. Furthermore, as no clinical data are available for this technology yet, the AHP is applied to predict the performance of the new technology with regard to all these outcome measures. Results of the AHP are subsequently integrated in a Markov model to make an early assessment of the expected incremental cost-effectiveness of alternative technologies. RESULTS: We systematically estimated priors on the clinical effectiveness and wider impacts of the new technology using AHP. In our illustration, AHP estimates for sensitivity and specificity of the new diagnostic technology were used as probability parameters in the Markov model. Moreover, the prioritized outcome measures including clinical effectiveness (weight = 0.61), patient comfort (weight = 0.09), and safety (weight = 0.30) were integrated into one outcome measure in the Markov model. CONCLUSIONS: Combining AHP and Markov modelling is particularly valuable in early technology assessment when evidence about the effectiveness of health care technology is still limited or missing. Moreover, combining these methods is valuable when decision makers are interested in other patient relevant outcomes measures besides the technology’s clinical effectiveness, and that may not (adequately or explicitly) be captured in mainstream utility measures. PODIUM SESSION I: RISK-SHARING SCHEMES RS1 A RISK FORECASTING MODEL TO HELP IN THE DESIGNING OF RISK-SHARING SCHEMES Ethgen O University of Liège, Liège, Belgium OBJECTIVES: To develop a financial risk forecasting model that could be used to negotiate risk-sharing schemes (RSS) conditions. RSS are increasingly being established as a market access strategy and have direct financial implications for both payer and manufacturer. However, underlying methodologies remain poorly researched. Additionally, HTA agencies may need analytical frameworks to evaluate the value of RSS. METHODS: We designed a financial-based agreement for a hypothetical technology. The financial risk to be shared is defined as N × p × r where N is the size of the target patients population, p is the price/dose of the technology, r is the proportion of patients for whom di > D, di being the number of doses/year given to an individual patient i and D being the maximum number of doses/year agreed in the schemes (the cap). a logistic growth curve is used to simulate the risk evolution as time progresses and patients progressively accrue in the RSS. Multiple risk evolution and “sharing” scenarios with their resulting financial implications for both parties are simulated. Finally, a Bayesian framework is introduced to allow both parties to make revisions as real-life information becomes available upon implementation of the RSS. RESULTS: For N = 1000 over a period of 3 years, D = 12 and a prior distribution for di centred on 12 doses but with 20% of patients receiving more than 12 doses, the model predicts that 23,092 doses will be delivered and that 542 doses (2.40%) will fall above the cap. These doses in excess determine the cost to be shared: total refund, partial refund, or price discount. CONCLUSIONS: Financial modelling and technological forecasting techniques can be combined to simulate different risk-sharing scenarios and their financial implications for payers and manufacturers. This provides both parties with an analytical framework to design win-win schemes and to make potential revisions as real-life information becomes available. RS2 PAYER ROADBLOCKS TO RISK-SHARING AGREEMENTS AROUND THE WORLD: WHERE, WHEN AND HOW? Ando G, Kowal S, Reinaud F IHS, London, UK OBJECTIVES: The increasing use of risk-sharing in reimbursement decisions across major markets necessitates that key stakeholders understand the role of this concept in shaping drug development and regulatory decision-making. The objective of this research was to examine global trends in risk-sharing agreements since 1990 to provide a comprehensive understanding of the current and future impact of this fastevolving concept. METHODS: Primary research was conducted through 50 in-depth 45-minute telephone interviews in native languages. Subjects were carefully selected and represented payers, government agencies, and HTA organizations in nine markets (Europe 5, Australia, New Zealand, United States, and Canada) to understand their assessment of the role which risk-sharing agreements have—or have not—played in their respective markets, and whether they will do so in the future. This was complemented with secondary research of reimbursement decisions around the world based on a newly created database of risk-sharing agreements around the world. RESULTS: In some countries such as the United Kingdom and Italy, for certain therapeutic areas such as oncology, these agreements almost act as a substitute for the normal reimbursement process, but primary research indicates that this practice faces significant resistance at many layers. Still, many other countries are seeking to understand the potential applicability of risk-shares to their own market. Also, risk-share agreements are being examined for their potential in several other therapeutic areas. While population- and patient-level agreements remain the most popular, we conclude that health outcomes-based arrangements are significantly on the rise, with 27 having been identified through the study in the markets that were studied, the majority of which were signed since 2007. Just over half were signed for oncology therapeutics. CONCLUSIONS: Outcomes-based agreements are becoming an increasingly important consideration to include in pricing models across the traditional development pathway for new molecules. A242 RS3 THE BAYESIAN APPROACH FOR PERFORMANCE-BASED PRICING: A CASE STUDY IN JAPAN Kamae I1, Araki D2, Kamae MS3 1 Keio University Graduate School of Health Management, Fujisawa, Kanagawa, Japan; 2Meiji University of Integrative Medicine, Nantan, Kyoto, Japan; 3Tufts Medical Center, Boston, MA, USA OBJECTIVES: To present a case study of Bayesian approach to performance-based pricing in Japan, and to discuss advantages and limitations of the proposed method. METHODS: The official price of any new drug listed on the National Reimbursement List in Japan is initially determined by Ministry of Health, Labor, and Welfare in political manners with some equations historically developed without considering the science of pharmacoeconomics. The initial price determined when approved is biennially re-priced downward regarding the average market price (AMP) in reality. At first, we reviewed the pricing and re-pricing methods of the government, then identified which portion of the current rules for pricing and re-pricing could be reworded with the quantitative methods of biostatistics and pharmacoeconomics, and finally tried to develop a Bayesian method for making the current re-pricing rule better quantified incorporating performance and value of a drug. RESULTS: We identified three portions such as premium rule in pricing, an adjustment in re-pricing, and the market extension premium in re-pricing could be objectively described by scientific methods. We developed in theory: 1) a pharmacoeconomic method associating the premium in pricing with ICER (incremental cost-effectiveness ratio); 2) a Bayesian method quantifying an adjustment in the current re-pricing rule with respect to estimating the AMP according to the degree of Bayesian belief; and 3) a mixture method of Bayes and ICER quantifying the market extension in re-pricing to incorporate the performancebased concept. CONCLUSIONS: The Japanese pricing and re-pricing rules, which employs the performance-based concept in a subjective manner, could be improved in terms of pharmacoeconomics and Bayesian statistics. Although the validation of the theory including data availability is left for further investigation, our approach would be thought-provoking for any country which considers a statistical approach to policymaking for performance-based pricing or reimbursement. RS4 THE INFLUENCE OF UK RISK-SHARING MODELS (PATIENT ACCESS SCHEMES) ON PAYERS IN GERMANY AND THE NETHERLANDS. A QUALITATIVE RESEARCH APPROACH Guhl N1, Kolominsky-Rabas P2, Toumi M1 1 University Claude Bernard Lyon 1, Lyon, France; 2Centre for Health Technology Assessment (HTA) and Public Health (IZPH), University of Erlangen-Nurnberg, Erlangen, Germany OBJECTIVES: The number of risk-sharing schemes in the UK increased significantly since the introduction of the new Pharma Pricing and Regulation Scheme (PPRS) in early 2009. Other health-care systems seem to follow. The objective is to investigate via qualitative research if payers in Germany and The Netherlands are influenced by UK decisions, if there are any preferred schemes, e.g., finance-based versus valuebased, and if the outcome for payers and manufacturers are comparable among the countries concerned. METHODS: A comprehensive questionnaire was designed, validated, and translated into German and Dutch. Decision-makers at various levels involved in the implementation of risk-sharing models in Germany and The Netherlands were interviewed face to face. The interviews lasted approximately 60 minutes and were held in their mother language. All interviews were recorded and a total of n = 10 interviews were conducted. RESULTS: German payers especially sick funds seem to be more open to adopt risk-sharing models than their colleagues in The Netherlands. There is a clear tendency in both countries that UK risk-sharing schemes have an influence on the decision-making process; however, most payers responded that there must be some adaptations to local needs. CONCLUSIONS: Risk-sharing models will become increasingly popular in the near future especially in Germany. Companies have to be prepared especially in therapeutic areas like oncology to follow this route as a way of market access. Although risk-sharing models have been introduced in other European countries before 2009, the UK models seem to represent a benchmark among payers especially with regard to their transparency. PODIUM SESSION II: ADHERENCE AND COMPLIANCE STUDIES AD1 SHORT MESSAGE SERVICE (SMS)-BASED STRATEGY TO IMPROVE ANTIPSYCHOTIC ADHERENCE AMONG PATIENTS WITH SCHIZOPHRENIA Maurino J1, Tesoro A2, Diez T3, Gomez-Beneyto M4 1 AstraZeneca, Madrid, Spain; 2Hospital Universitario Virgen Macarena, Sevilla, Spain; 3 AstraZeneca, Zaventem, Brussels, Belgium; 4University of Valencia, Valencia, Spain OBJECTIVES: Schizophrenia is a major psychotic disorder that has devastating effects on the lives of patients and their caregivers. Poor adherence to antipsychotic treatment is a well-recognized challenge among patients with schizophrenia and is associated with increased morbidity and consumption of health-care resources. Implementation of innovative strategies to enhance adherence is needed. To evaluate the effect of daily short message service (SMS) reminders on patients’ adherence to antipsychotic treat- 13th Euro Abstracts ment in schizophrenia. METHODS: A multicenter, randomized, open-label, controlled trial. The study included clinically stabilized schizophrenic outpatients with adherence problems (at least one affirmative answer of Morisky–Green questionnaire—MAQ). Participants were randomized to receive a daily SMS reminder on their cell phone to take their medication in the following 3 months or current standard of care. The SMS reminder was sent daily at 10 am or 2 pm according to participants’ preferences. The primary end point was the absolute difference in mean adherence rate between the two groups after 3 months using MAQ. RESULTS: A total of 254 patients were analyzed, 66.5% men. Mean age: 39.7 years (SD = 11.03). Baseline sociodemographic and clinical characteristics were similar between the two groups. No significant differences in baseline mean MAQ score were observed between groups (2.23, SD = 0.94 and 2.23, SD = 0.85, respectively). At month 3, mean MAQ score in the intervention group was 1.2 and 1.53 in the control group. The absolute difference in mean adherence rate between groups was −0.33 (95% CI −0.62, −0.03; P = 0.03). Largest change from baseline was observed in question 4 of the questionnaire “When you feel better, do you sometimes stop taking your medicine?” (percentage of affirmative answers in the SMS group vs. control was 45% and 59%, respectively, P = 0.02). CONCLUSIONS: SMS reminders may help many patients become sufficiently adherent to their antipsychotic medication in addition to standard interventions for schizophrenia control. AD2 THE SOCIETAL BURDEN OF POOR PERSISTENCE TO TREATMENT OF OSTEOPOROSIS IN SWEDEN Landfeldt E1, Lundkvist J2, Strom O1 1 i3 Innovus, Stockholm, Sweden; 2Amgen, Stockholm, Sweden OBJECTIVES: Poor persistence to prescribed treatment regimens is a well-documented health problem. The issue is of particular importance in treatment of chronic diseases, such as osteoporosis. The objective of this study was to estimate the annual societal burden of real-world persistence to treatment of osteoporosis in Sweden. a second aim was to estimate the monetary net benefit of improved persistence. METHODS: The annual societal burden was evaluated in relation to perfect persistence to a 5-year treatment duration and performed using a published Markov model by Ström and colleagues. The target population was extracted from the Swedish Prescription Register and based on all treatment-naïve patients who started therapy of primary osteoporosis in Sweden during 2009. Five hypothetical interventions were investigated, with improvements in the persistent proportion of between 10% and 50%. RESULTS: Annually, a total of 1018 fractures were estimated to be caused by nonpersistence to treatment of osteoporosis in Sweden. These fractures resulted in a substantial waste of health-care resources related to morbidity (c26 million annually) and a loss, in total, of 771 QALYs. Using a societal willingness-to-pay for a QALY of c60,000, the total annual societal burden, incorporating both monetary consequences and health effects, was estimated at c62.76 million. Given current Swedish cost-effectiveness guidelines, between approximately c225 and c1130 could be spent per patient to increase persistence, depending on the level of improvement (between 10% and 50%). CONCLUSIONS: The total annual societal burden of current, real-world persistence was estimated at c63 million. The estimated additional fracture-related costs associated with poor persistence were larger than the current total annual expenditure on all osteoporosis medications in Sweden. Poor persistence to treatment of osteoporosis should consequently be acknowledged as an important and costly health problem, and be taken into account when evaluating osteoporosis interventions. AD3 A CONCEPTUAL FRAMEWORK THAT CAPTURES THE MULTIDIMENSIONAL ASPECT OF ADHERENCE IN A HEALTHECONOMIC MODEL: AN APPLICATION IN CHRONIC NONMALIGNANT PAIN MANAGEMENT Treur M1, Houwing NS1, Heeg B1, Kristanto P2, Botteman M3, Vo P4, Marx S4, Lockhart E4, Van Hout B5 1 Pharmerit International, Rotterdam, The Netherlands; 2AARDEX Group Ltd., BE-4600, Belgium; 3Pharmerit North America, LLC, Bethesda, MD, USA; 4Abbott, Abbott Park, IL, USA; 5Pharmerit Ltd., York, UK Incorporating adherence in a health-economic evaluation is challenging. a novel conceptual framework was developed that captures the multidimensional aspect of adherence in a health-economic model, by considering the case of pain medication in chronic nonmalignant pain management. a 1-year microsimulation model was built. Realistic dosing profiles (combining dose-taking and dose-timing information) were obtained from electronically compiled dosing histories gathered in the Aardex/Pharmionic database. Probability and timing of dosing depended on time of intake during the day, day of the week, whether or not a previous dosage was taken, and number of days on treatment. The effect of dosages taken on pain was modeled via a pharmacokinetic–pharmacodynamic (PK-PD) model. The PK model translated dosages taken into drug concentration levels. Drug concentration levels influenced the amount of pain a patient experienced, quantified by the PD model. Quality of life (QoL) values were obtained by assigning a utility value to the pain intensity a patient experienced. Dosages taken by the patient contributed to medication costs. Work productivity losses due to pain exacerbations, based on a pain survey, were included. The model simulated 1000 unique patients to generate stable results, comparing realistic dosing profiles with an optimal dosing profile for a four times daily (QID) regimen. The realistic QID regimen had 14% lower average drug concentration levels and more than 3.5-fold higher standard deviation in concentration levels than the optimal regimen. Due to lower average concentration levels and higher fluctuation, this trans- 13th Euro Abstracts lated into patients experiencing 37% higher average pain intensity, an 8% decrease in QoL, and total additional costs of ~$1277 annually. This study provided a conceptual framework to model the health-economic impact of dosing variations based on actual dosing information in a realistic and intuitive way. This novel concept can also be applied to other disease areas to assess pharmaco-economic outcomes. AD4 THE IMPACT OF PHARMACIST INTERVENTION ON PATIENTS’ ADHERENCE TO ANTIDEPRESSANT MEDICATION AND PATIENTREPORTED OUTCOMES Aljumah K, Donyai P University of Reading, Reading, UK OBJECTIVES: Medications, regardless of effectiveness, do not work in patients who do not take them. Poor adherence to prescribed medication regimens has been a wellrecognized problem in all of medicine and patients with chronic conditions, such as depression, are less likely to follow prescription orders than those with acute conditions. In depression, several studies have reported a medication adherence rate of about 72% within the first month of treatment which drops sharply to about 43% after 6 months of treatment. Pharmacists may be able to help improve adherence rates, yet no single review has examined the impact of pharmacist interventions on adherence to antidepressants. The aim of this study was to summarize the literature and determine whether pharmacist intervention will 1) improve adherence in depression; and 2) improve patient-reported outcomes (PRO). METHODS: A systematic review of the literature was conducted using the PubMed database to retrieve studies examining the impact of pharmacist interventions on adherence to antidepressants and on patientreported outcomes, from 1990 to 2010. The following MESH terms were used: pharmacist intervention, medication intervention, depression, medication adherence, health-related quality of life, patient reported outcomes, and antidepressants. a total of 25 papers were retrieved with 11 excluded on the basis of abstract or full-text review resulting in 14 studies suitable for inclusion. RESULTS: The most common intervention strategy that pharmacists utilized was a combination of drug monitoring (baseline assessment and treatment follow-up), drug counseling by telephone and personal interviews, and patient education (about medication side effects).The results of these interventions were positive, improvement varying from 15% to 19%, and also, HRQL improved to varying degrees. CONCLUSIONS: The studies support the roles of pharmacists in providing interventions to improve medication adherence in depression. The results can provide a basis for future studies examining the costeffectiveness of pharmacist interventions in depression. PODIUM SESSION II: BIOLOGIC AGENT STUDIES AND METHODS BL1 COST-EFFECTIVENESS, VALUE OF INFORMATION, AND BUDGET IMPACT OF CERTOLIZUMAB PEGOL COMPARED TO SUBCUTANEOUS TUMOR NECROSIS FACTOR (TNF) INHIBITORS AND METHOTREXATE IN THE TREATMENT OF MODERATE-TO-SEVERE RHEUMATOID ARTHRITIS IN FINLAND Soini EJ1, Hallinen T1, Taiha M2, Honkanen V2 1 ESiOR Ltd, Kuopio, Finland; 2UCB Pharma Oy Finland, Espoo, Finland OBJECTIVES: To analyze the cost-effectiveness, multinomial expected value of perfect information (mEVPI), and budget impact of certolizumab pegol (CZP) compared to the used first-line subcutaneously administered tumor necrosis factor (TNF) − inhibitors + methotrexate (MTX) and MTX alone in the treatment of moderate-to-severe rheumatoid arthritis (RA) in the Finnish setting. METHODS: An Excel-based, probabilistic lifetime Markov cohort model was developed to assess the cost-effectiveness and mEVPI. Treatment efficacy was measured using the ACR-responses (no ACR20, ACR20, ACR50, or ACR70) at 3 months. ACR estimated response rates were based on adjusted indirect comparison (MTX as the common comparator) of published clinical trials. The health state utilities were estimated from the CZP-trials using regression models. The inpatient days were linked to patients’ HAQ-scores according to published literature. Cost estimates from Finnish sources were used and assessed from payer perspective. Cost and health outcomes were discounted with annual 3% discount rate. Undiscounted budget impact of CZP was estimated for years 2010– 2013 using Excel model. Equal inpatient costs and treatment efficacies were assumed for all TNF inhibitors. RESULTS: The lifetime costs for CZP + MTX, etanercept + MTX, and MTX alone were c179,986, c201,781, and c186,986, respectively. The corresponding QALYs (life-years) were 7.041 (15.697), 6.838 (15.646), and 6.336 (15.547). According to the cost-effectiveness acceptability frontier, CZP + MTX was the optimal treatment option with 58% to 73% probability of being cost-effective at willingness-to-pay values of c0–30,000 per QALY gained. The respective mEVPI was c5058–3701 per patient. Based on clinical responses, CZP + MTX dominated also adalimumab + MTX with 100–85% probability of cost-effectiveness (c0–30,000 per QALY gained; EVPI c0–218 per patient). The introduction of CZP produced annual budgetary net savings of c0.17–0.80 million during 2010–2013 in budget impact analysis. CONCLUSIONS: This analysis shows that, on average, CZP + MTX dominated the other subcutaneously injected TNF inhibitors considered or MTX alone in the Finish setting. The use of CZP resulted in budgetary net savings. A243 BL2 PATIENT PREFERENCES FOR BIOLOGIC AGENTS IN RHEUMATOID ARTHRITIS: A DISCRETE CHOICE EXPERIMENT Augustovski F1, Beratarrechea A2, Irazola V1, Rubinstein F1, Tesolin P3, Gonzalez JM4, Lencina V5, Scolnik M3, Wainmann C5, Navarta D3, Citera G5, Soriano E3 1 Institute for Clinical Effectiveness and Health Policy, Buenos Aires, Argentina; 2IECS & Hospital Italiano de Buenos Aires, Buenos Aires, Argentina; 3Hospital Italiano de Buenos Aires, Buenos Aires, Argentina; 4RTI Health Solutions, Research Triangle Park, NC, USA; 5 Instituto de Rehabilitación Psicofísica (I.R.E.P.), Buenos Aires, Argentina OBJECTIVES: To conduct a Discrete Choice Experiment (DCE) to elicit rheumatoid arthritis (RA) patients’ preferences regarding treatment with BIOLOGIC AGENTS (BA). METHODS: We designed a DCE with RA patients and seven treatment attributes: effectiveness, route of administration, frequency of administration, local and systemic adverse effects, severe infections, and out-of-pocket costs. RA patients who had never received BA from one private and one public hospital in Buenos Aires were included. a multinomial probit regression model (MNP) was used. RESULTS: A total of 240 RA participated (mean age 56.2 [SD 13.5], 87% women, median disease duration 9 years), all receiving conventional DMARDs (84.5% Metotrexate); median Clinical Disease Activity Index 7.5, and median HAQ 0.5. All the attributes showed to be significant factors affecting choice of treatment. Most attributes levels showed coefficients with the expected signs and were statistically significant. Attributes importance ranking was in the following order: cost, systemic adverse events, frequency of administration, efficacy, route of administration, local adverse events, and serious infection (table). Patients had relatively high monthly WTP for treatments that significantly reduced the risk of systemic adverse events: mean 331 (95% CI: 212–499) US dollars ($) for a reduction from 30% to 10%; of decreasing dose frequency: mean $302 (95% CI: 183–461) for going from weekly to monthly administration; increasing treatment efficacy: mean: $386 (95% CI: 285–532) for 40 versus 20 mm reduction in patient global assessment VAS), and also for switching from an intravenous to an oral therapy: mean $262 (95% CI: 262–555). CONCLUSIONS: Different treatment attributes had a significant and different influence in RA patients’ choice of BA. The results of the DCE indicated that most respondents would be willing to pay for treatments that importantly reduced the risk of systemic adverse effects, dose frequency, with increased treatment efficacy, and with an oral route of administration. BL3 A CONCEPTUAL MODEL FOR POMPE DISEASE: THE BACKBONE FOR AN ECONOMIC EVALUATION OF AN ORPHAN DRUG Kanters TA1, Redekop WK1, Hagemans MLC2, Van der Ploeg AT2, Hakkaart L1 1 Erasmus University Rotterdam, Rotterdam, The Netherlands; 2Erasmus University Medical Center, Rotterdam, The Netherlands OBJECTIVES: Studies of orphan drugs are, by nature, confronted with small patient populations, meaning that randomized controlled trials (RCTs) will have limited statistical power. Enzyme replacement therapy (ERT) is an orphan drug for Pompe disease, a metabolic orphan disease, with a prevalence in the The Netherlands of 1 per 40,000 births. In order to estimate the (cost-)effectiveness of ERT, we developed a disease model founded on all available clinical knowledge. METHODS: We developed a disease model with a strong clinical basis which linked disease-related factors with quality-adjusted life-years. The structure of the Wilson-Cleary health outcomes model was used as a blueprint. Based on literature and expert opinion, clinically relevant aspects of Pompe disease were applied to the distinct entities of the WilsonCleary model. Data from a Dutch cohort study (n = 94, mean follow-up = 2.7 years) were used to quantify the relationships between the different entities by means of regression analyses. RESULTS: A conceptual model for Pompe disease was developed by establishing a clinically plausible pathway from enzyme activity (“biological variables” in Wilson-Cleary model) via muscle strength and respiratory function (“symptom status”), and MCS and PCS from the SF-36 (“functional status”) to the final entity of health utility (“quality of life” using EQ-5D and SF-6D). The strengths of the relationships between these entities were based on the results of the regression analyses. Patient characteristics such as age, gender, and disease duration affected all entities in the model. Therapy was assumed to affect only enzyme activity in the model; all other health outcomes could only be affected via an impact on enzyme activity. CONCLUSION: We have developed a clinically based model to assess the long-term cost-effectiveness of ERT in Pompe disease. The approach used here is expected to be applicable in the assessment of other orphan drugs. BL4 COST-EFFECTIVENESS OF TREATMENT STRATEGIES FOR ANKYLOSING SPONDYLITIS: FIRST RESULTS OF A DISCRETE EVENT SIMULATION MODEL Tran Duy A1, Boonen A2, Arsenijevic J1, Severens JL3 1 Maastricht University, Maastricht, Limburg, The Netherlands; 2Maastricht University Hospital, Maastricht, Limburg, The Netherlands; 3Erasmus University Rotterdam, Rotterdam, South Holland, The Netherlands OBJECTIVES: Owing to rapid, persistent efficacy in reducing pain and preventing structural damage, tumor necrosis factor-alpha inhibitors (anti-TNFs) has revolutionized the treatment of ankylosing spondylitis (AS). However, the high price of antiTNFs has limited their use. Recent studies suggest that increased medication costs by adopting anti-TNFs may be offset by decreased values of other cost categories. In most cost-effectiveness studies on AS treatments, costs were estimated as a whole. Our A244 objectives were to 1) develop a simulation model which can simulate QoL and different cost categories from a societal perspective as affected by AS treatments with sequential drugs using the ASAS recommendations at group level; 2) parameterize the model using Dutch data; and 3) analyze simulated cost-effectiveness of treatment strategies with and without anti-TNFs. METHODS: Discrete-event paradigm was selected for model development. Pathways in AS treatments were simulated in conjunction with changes in AS-specific measures. Parameterization was realized using original data from an observational cohort of AS patients and literature where necessary. Frequencies of visits to rheumatologists and hospitals, and having paramedical treatments, informal/formal care, sick leave, and work disability were estimated using logit and Poisson regression models. Health utility was estimated using a proportional-odds model. Demographic variables were also considered in model selections. Two compared strategies in the simulation were 1) three Cox-1 plus two Cox-2 selective nonsteroidal anti-inflammatory drugs, and 2) same as strategy 1 plus Etanercept and Infliximab. The simulation was run for 1000 patients until death. RESULTS: Simulated values of QALYs and different cost categories were in reasonable ranges. Productivity and hospital costs constituted largest shares of the total costs, and were significantly lower in strategy 2 (P < 0.05). The incremental cost per QALY gained of strategy 2 against strategy 1 was c25,165. CONCLUSIONS: The model is flexible and promising for assessing actual costs-effectiveness of different treatment strategies for different societies. PODIUM SESSION II: IRT AND RACH MODELING IN QUALITY OF LIFE MEASUREMENT IR1 METHODS FOR PSYCHOMETRIC AND CLINICAL EVALUATIONS OF CAT-BASED MEASURES OF DISEASE IMPACT IN CHRONIC KIDNEY DISEASE (CKD) Lin P1, Ware JE1, Meyer K2, Richardson M3, Bjorner JB4 1 John Ware Research Group, Inc., Worcester, MA, USA; 2Tufts University School of Medicine, Boston, MA, USA; 3Tufts Medical Center, Boston, MA, USA; 4National Institute for the Working Environment, Copenhagen, Denmark OBJECTIVES: To evaluate a new standardized disease-specific PRO impact item bank, the Kidney Disease Impact Scale (KDIS), among adults with chronic kidney disease (CKD). METHODS: A 34-item bank of items measuring disease impact, with standardized content (differing only in disease attribution) and standardized IRT-based parameters from previous studies across therapeutic areas, was administered to 516 CKD patients, along with CKD-specific legacy (KDQOL Effects and Burden scales) and a generic (SF-12v2) health survey. All measures were administered in a clinic setting at three time points (baseline, 1 week, and 3 months) to evaluate: 1) reliability (internal consistency and 1-week test–retest); 2) validity in discriminating among CKD groups differing in clinically defined disease severity (nondialysis stages 3–5, dialysis patients, and transplant patients); and 3) responsiveness (better, same, or worse after 3 months). Real-data CAT simulations were performed, using only the responses to items that would have been selected and asked during a real CAT, to estimate KDIS scores based on the full bank, and 5-, 10-, and 15-item CATs. RESULTS: As hypothesized, CAT-based score distributions were less skewed and estimates were more efficient (covered a wider range and were as reliable with fewer items), in comparison with legacy measures. All KDIS forms (full bank, various lengths of CATs) discriminated among clinically defined groups as well or better than legacy CKD-specific measures and correlated substantially with generic HRQOL measures. Relative validity (RV) coefficients for discriminant validity and responsiveness tests for 1-minute (5-item) CAT-based administrations compared favorably with scores based on the entire bank (RV = 0.84), KDQOL scores (RV = 0.79 for Burden and 0.50 for Effects), and generic measures (RV = 0.00 to 0.59, median = 0.13). CONCLUSIONS: Diseasespecific PROs can be improved to be more practical and more valid, in comparison with legacy CKD-specific and generic PROs. CKD impact measures that perform better psychometrically tend to perform better in empirical tests of clinical validity. IR2 SCALING PROPERTIES OF TWO COMMONLY USED OUTCOME MEASURES IN DERMATOLOGY—THE DERMATOLOGY LIFE QUALITY INDEX (DLQI) AND THE PSORIASIS QUALITY OF LIFE SCALE (PSORIQOL) Twiss J, Mckenna S, Crawford S, Doward L Galen Research Limited, Manchester, UK OBJECTIVES: The DLQI is a 10-item generic dermatological HRQoL measure used widely with psoriasis patients. The PSORIQoL is a 25-item psoriasis-specific measure that employs the needs-based model of quality of life (QoL). Both measures have been shown to have good psychometric properties using Classical Test Theory. However, their scaling properties have not been fully assessed using Item Response Theory (IRT). This is the purpose of the study. METHODS: Psoriasis patients, recruited from an out-patients clinic, completed the DLQI and PSORIQoL. IRT (Rasch analysis) was used to determine each measure’s overall fit to the Rasch model, individual item fit, targeting of scales to severity of respondents, functioning of response categories, and the presence of Differential Item Functioning (DIF) by age or gender. RESULTS: 146 psoriasis patients were included in the study (male 50%, mean age = 44.2, range = 17–83 years). DLQI: There was overall misfit to the Rasch model (χ2 = 39.85, df = 20, P = 0.005). One item misfit the Rasch model and another showed borderline misfit. 13th Euro Abstracts Four of the 10 items had disordered response thresholds indicating that they did not work in a logical way. Too few DLQI items covered milder levels of HRQoL (meaning that the measure is not sensitive to change in patients with mild disease). DIF by age was found in one item and DIF by gender in another. PSORIQoL: Overall, the measure fit the Rasch model (χ2 = 56.45, df = 50, P = 0.247), although one item had a high fit residual suggesting misfit. The response options worked logically and items were well distributed across the QoL measurement range. No DIF by age or gender was found. CONCLUSIONS: Application of Rasch analysis indicated that there were several problems with the scaling properties of the DLQI. In contrast, the PSORIQoL fit the Rasch model and exhibited good measurement properties. IR3 THE USE OF IRT MODELING TO ACCOUNT FOR DIFFERENTIAL ITEM FUNCTIONING IN THE PROQOL-HIV QUESTIONNAIRE Lalanne C, Duracinsky M, Armstrong AR, Chassany O Assistance Publique-Hopitaux de Paris, Paris, France OBJECTIVES: This study aims to provide sample-free estimates of uniform DIF on a newly developed HRQL questionnaire, PROQOL-HIV, based on IRT. Patient characteristics considered here were gender, country of residence, and ethnicity, given the high prevalence of HIV in migrant populations. METHODS: We use a two-step ordinal logistic regression to test for the significance of person covariates, on two constructs extracted from factor analysis: Physical Health and Symptoms (PHS, 14 items), and Emotional distress and Health Concerns (EDHC, 10 items). a subset of N = 505 patients (71% males, median age 43 years) from five countries (Australia, United States, France, Brazil, and Thailand) was used. Person parameters were estimated using a Graded Response Model. Missing responses (<1%) were imputed using halfrule. RESULTS: The EDHC dimension was more prone to exhibit item biases compared to PHS items, especially when referring to ethnicity. Country-related effects were equally distributed on the two scales. In both cases, however, no DIF effects were found for gender. We reached similar conclusions when comparing the results to those obtained when using sum or rest scores as conditioning ability levels. Correlations between raw scores computed with or without flagged items were above 0.5. Restricting analyses to Western countries, however, yielded smaller hits. Altogether, these results suggest that DIF effects should be of limited impact when reporting scores for these two HRQL dimensions of the PROQOL-HIV questionnaire. Results are discussed in light of other latent variable models proposed to assess DIF, the MIMIC, and logistic mixed effects model. CONCLUSIONS: Using IRT instead of summated scale scores allows the incorporation of person covariates and the study of DIF effects in a probabilistic framework. The presence of DIF effects on biopsychological constructs raises important cross-cultural validity considerations for multi-country multiethnic studies, although assessment of DIF impact needs further investigations. IR4 HUNTINGTON QUALITY OF LIFE INTERVIEW (HQOLI): A NEW HUNTINGTON’S DISEASE-SPECIFIC QUALITY OF LIFE INSTRUMENT Clay E1, Guelfucci F1, Dorey J1, Tedroff J2, Verny C3, Toumi M4, Aballea S1 1 Creativ Ceutical, Paris, France; 2NEUROSEARCH, Ballerup, Denmark; 3Neurology Unit CHU, Angers, France; 4University Claude Bernard Lyon 1, Lyon, France OBJECTIVES: No specific health-related quality of life (HRQOL) instrument tool is available for Huntington’s disease (HD). In the context of a large European study on HD burden (Euro-HDB), a specific HRQoL tool (HQOLI) was created for patients in early and intermediate stages of HD. We describe the development and psychometric validation of HQOLI. METHODS: After semistructured interviews with patients, caregivers, and HD specialists, we ran a patient focus group. a self-reported questionnaire was then devised in French language. The questionnaire consists of 42 items, relating to 11 dimensions: Mobility, Motor Ability, Autonomy 1 and 2, Finance, Occupational Activities, Psychology, Feelings of Guilt, Vitality, Socializing, and Body Image. Face validity and internal and external validity were examined. RESULTS: Among 178 participating patients, 15 did not complete HQOLI. Item response rates range from 0.81 to 0.98, excluding the occupational activities domain. As expected, there is a ceiling effect on several items, suggesting low sensitivity at advanced HD stage. Cronbach’s alphas are greater than 0.76. HQOLI total score is highly correlated with generic HR-QOL (SF36 and EQ5D) as well as with the specific caregiver HRQoL instrument (HDQoL-C). Correlations were −0.73 (P < 0.01) with EQ5D utility, −0.76 (P < 0.01) with “Physical Functioning,” and −0.74 (P < 0.01) with “Mental Health.” Voluntary movement, fall/balance, and dystonia symptoms contribute significantly to HQOLI (P = 0.01), but chorea does not (P = 0.21). CONCLUSIONS: These data support the validity of the HQOLI in patients with HD. The ceiling effect limits the use of HQOLI in late stages but appears necessary to achieve good discrimination at early and intermediate stages. Correlations with other instruments suggest good external validity. High correlations of several items within dimensions suggest item number can be reduced. a new, shorter version of the questionnaire is in development. Assessments of responsiveness to change, test–retest reliability, and cross-cultural validation will be performed/are underway. 13th Euro Abstracts PODIUM SESSION II: MIXED TREATMENT COMPARISON METHODOLOGY AND APPLICATIONS MT1 DOES MODEL CHOICE IN MIXED TREATMENT COMPARISONS AFFECT THE OUTCOME FOR DECISION-MAKERS? Adams RC1, Schmitz S2, Fitzgerald O3, Walsh C2, Barry M1 1 National Centre for Pharmacoeconomics, Dublin, Ireland; 2Trinity College Dublin, Dublin, Ireland; 3St. Vincents University Hospital, Dublin Academic Healthcare, Dublin, Ireland BACKGROUND: An incremental cost-effectiveness ratio is calculated by dividing the difference in cost of an intervention by the difference in effect. In the absence of head to head trials, it is often necessary to combine evidence from placebo-controlled trials of different treatments and thereby derive an estimate of effect of one treatment against another. There are formal methods to do this and various models can be used. We present four different models and examine whether using different models can change the conclusion of the economic evaluation in rheumatoid arthritis (RA). METHODS: A literature review was carried out and the relevant data were extracted according to defined inclusion criteria. The outcome measures chosen were American College of rheumatology (ACR) 20 and 50 at 6 months. The MTC was carried out in WinBUGS and the method used was that by Nixon et al.[1]. This model allows inclusion of methotrexate as a parameter. We then used the four different estimates from each of the models in an RA model. RESULTS: The four different models differed in whether random or fixed effect was applied to the treatment or to methotrexate. Pairwise risk ratios (RR) are presented. The RR differ slightly for the same drug using the different models. Both the positioning of the ICERs on the CE plane and the probability of cost-effectiveness change from model to model. CONCLUSIONS: We see a difference in the results of using different models for MTC. However, the relative positioning of the estimates remains the same. Therefore, in deciding which method to choose in the case of biological agents, it is unlikely that the choice of model will impact greatly on the overall decision. MT2 CONDUCTING A MIXED TREATMENT COMPARISON (MTC) IN A SPECIFIC POPULATION: POSTMENOPAUSAL WOMEN WITH METASTATIC HORMONE RECEPTOR POSITIVE BREAST CANCER WHICH OVEREXPRESSES HER2 Pacou M1, Gauthier A1, Abrams K2, Moore L3, Jiang Y1, McNamara S3 1 Amaris Consulting UK, London, UK; 2University of Leicester, Leicester, UK; 3Roche Products Limited, Welwyn Garden City, UK OBJECTIVES: To estimate the relative treatment effects on progression free survival (PFS) and overall survival (OS) of trastuzumab plus an aromatase inhibitor (AI), lapatinib plus an AI, and AI monotherapy (anastrozole, letrozole, or exemestane) in postmenopausal women with untreated metastatic hormone receptor positive breast cancer which overexpresses HER2. METHODS: A systematic literature review was conducted in MEDLINE, MEDLINE-IN-PROCESS, EMBASE, Cochrane CENTRAL, Web of Science Proceedings, and BIOSIS to retrieve relevant randomized clinical trials. Heterogeneity and inconsistency were discussed with clinical experts and assessed statistically. An MTC was developed within the Bayesian framework using noninformative priors. It was decided a priori to investigate whether AIs could be considered as a class, through the evaluation of hazard ratios between single agents. As few trials focused on the exact target population, extensive sensitivity analyses were planned. RESULTS: The base-case analysis indicated no significant difference between AIs on PFS. Combined therapies were significantly more effective than AIs: the hazard ratios were estimated at 0.55 (95% CrI: [0.41; 0.74]) for trastuzumab + AI and 0.71 (95% CrI: [0.53; 0.96]) for lapatinib + AI. Although the probability of being best varied in the sensitivity analyses, trastuzumab + AI was always associated with the highest probability of best efficacy. The effect on OS was similar across AIs. No significant difference was found between the two combination therapies, which were associated with high probabilities of best efficacy (48% for trastuzumab + AI and 50% for lapatinib + AI). The sensitivity analyses indicated that adjustment for cross-over had an important impact on the OS results. CONCLUSIONS: Our study confirmed that anastrozole, letrozole, and exemestane had similar efficacy with respect to PFS and OS and that combined therapies were associated with significantly improved PFS compared to AIs. Conducting an MTC in this specific population was addressed through clinical and statistical assessments of heterogeneity and inconsistency, and extensive sensitivity analyses. MT3 COMPARATIVE EFFECTIVENESS OF ANTIDEPRESSANTS (AD) IN THE TREATMENT OF THE ELDERLY WITH MAJOR DEPRESSIVE DISORDER (MDD): A MIXED TREATMENT COMPARISON AND META-REGRESSION ANALYSIS Benedict A1, Naci H2, Fahrbach K3, Bozkaya D3, Happich M4, Deberdt W5, Raskin J6 1 United BioSource Corporation, Budapest, Hungary; 2United BioSource Corporation, London, UK; 3United BioSource Corporation, Lexington, MA, USA; 4Eli Lilly & Company, Bad Homburg, Hessia, Germany; 5S.A. Eli Lilly Benelux N.V., Bruxelles, Belgium; 6Eli Lilly & Company, Toronto, ON, Canada OBJECTIVES: To estimate the probability of clinical response, remission, and treatment withdrawal of antidepressants including TCAs, SSRIs, mirtazapine, buproprion, and SNRIs in elderly patients with MDD using indirect comparison (MTC) and metaregression (MR) methods. METHODS: Published systematic reviews of AD in elderly patients with MDD were used to extract data from RCTs of comparator treatments A245 reporting response or remission or treatment withdrawal. Studies with less than 6 weeks duration, open label extension studies, and studies in patients with significant dementia were excluded. Bayesian MTC and MR analyses, enabling indirect comparisons of interventions while respecting randomization, were performed using OpenBugs 3.0.2, adjusting for placebo control, baseline disease severity, treatment duration, age, patients with late onset depression as study, or arm level covariates as appropriate. Assuming similar class efficacy TCAs and SSRIs were grouped. RESULTS: The final data set included 40 studies, with 87 comparator arms. Placebo control was associated with a reduction in the log odds ratio (LOR) of response versus placebo for all treatments. Severity and publication year was marginally correlated with response. Posterior median predicted response and remission rates in non-placebo-controlled studies ranged between 71.9% (buproprione) and 77.1% (duloxetine), and remission between 29.4% (SSRIs), 42.7% (duloxetine), and 58.1% (TCAs). Predicted withdrawal rates for non-placebo-controlled studies vary between 25.7% (duloxetine) and 40% (venlafaxine). Remission results were sensitive to how the mean placebo incidence across studies was modeled. CONCLUSIONS: In the absence of head-to-head clinical trials, this analysis provides comparative estimates of relative treatment effect of AD among elderly with MDD while statistically taking into account heterogeneity. Although statistically not different, results suggest that among the elderly, duloxetine and TCAs produce the largest clinical response and remission rates, with relatively favorable withdrawal rates for duloxetine. MT4 MIXED TREATMENT COMPARISON (MTC) OF ANTIFUNGAL DRUGS FOR PROPHYLAXIS TREATMENT AGAINST INVASIVE FUNGAL INFECTIONS (IFIS) IN PATIENTS RECEIVING CHEMOTHERAPY FOR HEMATOLOGICAL MALIGNANCIES OR ALLOGENEIC HEMATOPOIETIC STEM CELLS TRANSPLANTATION Pechlivanoglou P, Le HH, De Vries R, Postma MJ University of Groningen, Groningen, The Netherlands OBJECTIVES: Neutropenic patients treated with chemotherapy for hematological malignancies and recipients of allogeneic hematopoietic stem cells transplantation (HSCT) are at high risk of acquiring invasive fungal infections (IFIs). Prognoses for IFIs are poor in terms of mortality and morbidity. An important factor in improving outcome is early treatment with antifungal agents. Early treatment, however, is complicated by a number of factors, including poor diagnostic measures. As a consequence, antifungal prophylaxis may be an effective strategy. Use of antifungal prophylaxis remains a matter of controversy, without a clear consensus on the choice of drug treatment. In this study, we used mixed treatment comparison (MTC) methodologies to compare the effectiveness of different antifungal drugs used in prophylaxis treatments against IFIs in high-risk patients. METHODS: Through systematic and transparent methodologies, MTC techniques allow for the combination of evidences from different sources and as a result may provide information where none exists or where direct comparisons are incomplete. For the estimation of effectiveness, we collected evidence on proven or probable IFIs from 13 different randomized clinical trials (RCTs). Comparisons of interest included placebo, fluconazole, itraconazole, voriconazole, and micafungin. The MTC analyses were carried out through the application of Bayesian hierarchical models. RESULTS: Significant evidence was found on the superiority of all study drugs versus placebo for preventing IFIs. Furthermore, MTC analyses indicated that the probability of acquiring an IFI after voriconzole prophylaxis is significantly lower than after fluconazole prophylaxis (dOR = 0,17 95% CI = [0.03, 0.94]). Comparisons between all other antifungals yielded no significant differences in incidence of IFIs. CONCLUSIONS: After the application of MTC methodologies on RCT evidence, the authors found evidence suggesting that prophylaxis with antifungals is superior to placebo and prophylaxis with voriconazole is superior to fluconazole regarding reduction in incidence of IFIs. PODIUM SESSION II: DEVELOPMENTS IN HTA AGENCIES LIKE NICE NI1 RETROSPECTIVE ANALYSIS OF TECHNOLOGY APPRAISALS CONDUCTED BY UK’S NATIONAL INSTITUTE FOR HEALTH AND CLINICAL EXCELLENCE (NICE) IN SEVERAL CANCER INDICATIONS: HIGHER EVIDENCE BARRIERS FOR TARGETED THERAPIES? Banerji L, Das S J&D Associates, London, London, UK OBJECTIVES: We undertook a retrospective analysis of NICE technology appraisals in NSCLC, hematological malignancies, gastric and breast cancer, and compared submission evidence with final NICE recommendation. METHODS: The analysis was based on a review of published appraisals on the NICE website from 2007 to 2010. RESULTS: Over the past 3 years, NICE completed 15 appraisals in the indications of NSCLC, hematological malignancies, and gastric and breast cancer. Ten of these appraisals were of targeted therapies. A total of five out of 15 appraisals were not recommended mainly because of weak clinical evidence surrounding efficacy. One of these technologies, Lapatinib, was not recommended for breast cancer despite the manufacturer offering a patient-access scheme. Focusing on the 10 appraisals that received positive recommendation, four appraisals received unconditional reimbursement. Rituximab is the only targeted therapy to receive unconditional reimbursement because it was cost-effective (<£30, 000 per QALY) due to the strong clinical evidence that demonstrated PFS >10 months compared to chemotherapy in CLL. The remaining A246 six assessments that resulted in conditional reimbursement were targeted therapies. Typically, with these targeted therapies, PFS or OS ranged from ~3 months to 9 months with the cost-per-QALY > £45.000. Five of the six manufactures participated in a patient-access scheme which consisted of fixed-price discounts such as Cetuximab (CRC) and Gefitinib (NSCLC) or performance schemes like sunitinib (GIST), Bortezomib (myeloma), and Lenalidome (myeloma). CONCLUSIONS: Based on the retrospective analysis, it is clear that the biggest challenge for targeted-cancer therapies is affordability with only one of the targeted therapies receiving unconditional reimbursement. However, nearly all the other targeted therapies evaluated that offered >3 months OS or PFS were recommended by NICE with a proviso to bring down the cost of treatment. Therefore, when companies develop their market access strategy, they should include a patient-access scheme in order to enter the UK market. NI2 PATIENT ACCESS SCHEMES IN UK ARE DRIVEN BY HEALTH TECHNOLOGY ASSESMENT Toumi M1, Jaroslawski S2 1 University of Lyon, Lyon, France; 2Creativ Ceutical, Paris, Ile de France, France OBJECTIVES: Achieving market access for new products has become complex for pharmaceutical companies. Faced with growing expenditure, health-care authorities accept or propose various schemes (risk sharing/payment for performance/commercial): UK’s Department of Health coined a term Patient Access Scheme (PAS) and published specific guidance for the industry. We performed an in-depth analysis of design of PASs in UK to revisit their typology and rationale. METHODS: We reviewed official and grey literature on the Web sites of UK’s HTA Agency—NICE, the Department of Health (DoH), the industry, and in the Internet. We searched for documents containing all synonimes of PAS and different scheme types. We selected PASs launched after 2006. RESULTS: We identified 13 PASs, all of which were designed/ implemented in consultation with NICE. Drug’s comparative effectiveness was central to the rationale behind the design of PASs. If effectiveness was acknowledged in the HTA, PAS was based on cost-containment (rituximab, erlotinib). If it was not recognized, this was for one of the two reasons: (1) the uncertainity about the long-term effect of the drug, or (2) the value of ICER was questioned in the HTA. In case of (1), the PAS consisted in free provision of the drug by manufacturer after a predefined period (lenalidomide, ranibizumab). In the case of (2), the PAS aimed at lowering the ICER either through cost containment (sunitinib, cetuximab, pemetrexed), through linking payment to outcomes (bortezomib, omalizumab), or by a mix of the two (certolizumab, ustekinumab). CONCLUSIONS: Formalized Health Technology Assessment is both a prerequisite and reason for implementing Patient Access Schemes in the UK. If the comparative effectiveness of a drug is acknowledged, the agreement is based on cost containment. On the other hand, if it is questioned, the PAS may have a form of a risk-sharing scheme and may be linking the payment to health outcomes (performance-based scheme). NI3 NICE’S COST-EFFECTIVENESS THRESHOLD REVISITED: NEW EVIDENCE ON THE INFLUENCE OF COST-EFFECTIVENESS AND OTHER FACTORS ON NICE DECISIONS Devlin N1, Dakin H2, Rice N3, Parkin D4, O’Neill P1 1 Office of Health Economics, London, UK; 2University of Oxford, Oxford, UK; 3University of York, York, UK; 4NHS South East Coast, Horley, Surrey, UK OBJECTIVES: Since its establishment, NICE has become increasingly explicit about the way it uses evidence on cost-effectiveness in decision-making—and, more recently, about the other factors it considers. This, together with other ways in which decisionmaking has evolved, suggests a number of testable hypotheses. We propose and empirically test alternative ways that NICE decision-making might be modeled, building on and extending Devlin and Parkin (2004) and Dakin et al. (2006). The large number of NICE decisions now observable facilitates the use of more sophisticated modeling techniques. METHODS: NICE’s decisions are characterized as binary choices: yes or no to a technology in a specifically defined patient group or indication. NICE Guidance often contains multiple such decisions. The probability of NICE recommending a technology is modeled as depending on evidence on effectiveness and cost-effectiveness; characteristics of the patients, disease, or treatment; and contextual factors. Data were obtained from HTAinSite (http://www.htainsite.com) on November 2009. RESULTS: Initial results, drawing on data for 262 decisions, suggest costeffectiveness alone explains the vast majority of NICE’s decisions, correctly classifying 85%, with high sensitivity and specificity. The estimated threshold, around £40k, is higher than NICE’s stated threshold (20k–£30k) but similar to that estimated by Devlin and Parkin (2004). Results across alternative model specifications showed that almost none of the other variables exert a statistically significant effect on decisions, with two exceptions. First, technologies for the treatment of cancer have a significantly higher probability of being accepted, ceteris paribus, implying a willingness to pay an additional >£10k per QALY gained by cancer patients. Second, analysis of the subset of decisions made after NICE’s second “social value judgement” document suggest an increased probability of rejection. CONCLUSIONS: This is work in progress; further results will be available to report from additional data extraction and modeling. 13th Euro Abstracts NI4 DO PATIENT ACCESS SCHEMES RESULT IN AN ACCEPTABLE ADMINISTRATIVE BURDEN? Haynes S, Costello S, Kusel J, Hamer N, Brooks-Rooney C Costello Medical Consulting Ltd., Cambridge, UK OBJECTIVES: In the UK, Patient Access Schemes (PAS) have become more common in submissions to the National Institute for Health and Clinical Excellence (NICE). The increase in PAS is a result of the essential role such schemes play in enhancing the availability of high-cost treatments to payers. In published appraisals, minimal emphasis has been placed upon the administrative burden of PAS, which is typically described as “acceptable.” The aim of this study was to assess the impact of administering PAS in the UK, using both primary research and existing literature to identify key administrative challenges. METHODS: A literature search was conducted using PubMed and Google Scholar. Freedom of information requests were sent to NICE for data on PAS administration. a pilot questionnaire was distributed to all 19 contacts listed on the directory of NHS Chief Pharmacists in Wales, to assess the real-life burden of PAS administration. RESULTS: Limited literature is available on the administration of PAS. However, the literature search uncovered evidence that the administrative impact of PAS is being recognized. The creation of the Patient Access Scheme Liaison Unit (PASLU) in October 2009 and the publication of the Pharmaceutical Price Regulation Scheme (PPRS) are two such developments, both of which are steps toward a system that more accurately reflects the needs of NHS administrators. The return rate for the questionnaire was low; however, responders showed dissatisfaction with multiple aspects of PAS management. Responders emphasized the need to address NHS requirements (both financial and temporal) in order to facilitate accurate PAS administration. CONCLUSIONS: Encouraging steps have been taken to recognize the burden of PAS on the NHS; however, further research is required to assess whether these recent developments are meaningful in everyday practice. Additional support for appropriate PAS implementation must also be provided if these important schemes are to continue effectively. PODIUM SESSION III: BIASES, METHODOLOGICAL CHALLENGES, AND SOLUTIONS BI1 ANALYZING OVERALL SURVIVAL IN RANDOMIZED CONTROLLED TRIALS WITH CROSS-OVER Jonsson L1, Sandin R2, Ekman M1, Ramsberg J1, Charbonneau C3, Huang X4, Jonsson B5, Weinstein MC6, Drummond M7 1 i3 Innovus, Stockholm, Sweden; 2Pfizer Oncology, Sollentuna, Stockholm, Sweden; 3Pfizer, Inc, New York, NY, USA; 4Pfizer Oncology, La Jolla, CA, USA; 5Stockholm School of Economics, Stockholm, Sweden; 6Harvard School of Public Health, Boston, MA, USA; 7 University of York, York, UK BACKGROUND: Offering patients in oncology trials the opportunity to cross over to active treatment at disease progression is a commonly used strategy to address ethical issues associated with the use of placebo controls, but could lead to statistical challenges for the analysis of key end points such as overall survival. While an advantage from the perspective of the treated patient enrolled in the trial, cross-over leads to loss of information and dilution of the comparative clinical efficacy and costeffectiveness results. OBJECTIVES: The purpose of the study is to compare alternative methods for analyzing overall survival data in the presence of cross-over, thus illustrating differences between methods, and providing guidance on choice of methodology. METHODS: Two promising methods for dealing with cross-over are inverse probability of censoring weighting and the rank-preserving structural failure time model. The methods are compared with naïve censoring of data at cross-over and intention-totreat analysis ignoring cross-over using two recent examples of trials in oncology: the receptor tyrosine kinase inhibitor sunitinib in renal cell carcinoma (RCC) and in gastrointestinal stromal tumor (GIST). RESULTS: The analyses showed that for a trial with a low proportion of cross-over from placebo to active treatment (RCC), the choice of statistical method did not affect the results to a great extent; the range of relative mortality risk for active treatment versus control was narrow. With a high proportion of cross-over (GIST), the range of relative mortality risks was broader. CONCLUSIONS: Naïve censoring at cross-over can lead to bias and should be avoided. If cross-over occurs frequently, the inverse probability of censoring weighting method or the rank-preserving structural failure time model are recommended depending on the characteristics of cross over in the trial, trial size, and available data. BI2 MULTIPLE COHORT MODELING OF LONG-DURATION INTERVENTIONS: QUESTIONING TIME HORIZONS AND AGGREGATION ACROSS COHORTS O’Mahony J Erasmus University Medical Center, Rotterdam, The Netherlands BACKGROUND: Models are widely used in the cost-effectiveness analysis of healthcare interventions. Most models only simulate one patient cohort, but some use multiple cohorts. Advocates of multicohort modeling contend it better represents actual health-care implementation, especially where interventions are applied over specific age ranges, as in cancer screening. When such an intervention is introduced, cohorts already older than the starting age only receive a partial intervention, possibly A247 13th Euro Abstracts resulting in cost-effectiveness different to that of younger cohorts that receive the complete intervention: multi-cohort models can include both these “complete” and “partial” cohorts. Some multi-cohort models described as population models impose finite time horizons at which the intervention is assumed to cease, although health effects are typically assessed until death. ANALYSIS: If cost-effectiveness differs between partial and complete cohorts, then the overall cost-effectiveness estimate from a multi-cohort model will depend on the relative numbers of partial and complete cohorts. The total number of complete cohorts depends on how long the intervention is used, which is uncertain. Therefore, the overall estimate may depend, in part, on the number of future cohorts assumed. The appropriateness of time horizons depends on whether a cross-sectional or a longitudinal cohort approach is used. Assuming an intervention ceases at a time horizon is unrepresentative of actual implementation and may result in biased cost-effectiveness estimates for curtailed cohorts. CONCLUSION: Multi-cohort modeling is advocated as being more representative of actual implementation. However, a single cost-effectiveness estimate for multiple cohorts necessarily implies an aggregation of estimates. Such aggregation leaves estimates sensitive to assumptions of the number of cohorts included, can hide useful information, and lead to nonoptimal policy choices. We suggest cost-effectiveness estimates for the complete and incomplete cohorts should not be aggregated, but reported separately. Implementation time horizons should not be used in longitudinal cohort-based modeling in cost-effectiveness analysis. BI3 COMMON AND AVOIDABLE ERRORS IN ECONOMIC MODELING: A REVIEW OF THE FREQUENCY AND IMPACT OF MODELING MISTAKES Taylor M, Kenworthy J, Lewis L York Health Economics Consortium, York, North Yorkshire, UK BACKGROUND: Cost-effectiveness models are often used to predict the costs and health outcomes that are likely to be associated with various different interventions. Models are a useful tool for representing the detailed and complex “real world” in a more simple and understandable structure. While models do not claim to necessarily create an exact replica of the real world, they can be useful in demonstrating the relationships and interactions between various different factors. However, developers of models often consciously, and unconsciously, make assumptions that are avoidable and may bias the results of a model. METHODS: A review was undertaken on a random selection of published models in different disease areas to aim to identify the frequency of typical “errors” in economic models. In addition, a simple model was developed and used to explore the relative impact of different types of errors in models. Each type of error was examined for its likely impact on the model’s overall findings and conclusions. This helped to gain a greater understanding of both the frequency of different errors and their magnitude of effect. RESULTS: Mistakes are commonly observed in economic models. These were often due to limitations in scope of the model, but all were found to be avoidable given unlimited time and data availability. As well as identifying “major” errors in models, the review also identified many common errors, such as excluding “half cycle correction,” that often have very little impact on a model’s results, relative to other common errors. CONCLUSIONS: While many errors in economic models are frequent, many errors often go unnoticed and have significant impact upon a model’s results. This analysis has highlighted the relative importance of each type of error and has provided suggestions as to how these might be avoided. BI4 ARE SECOND OPINIONS OBJECTIVE? BIASES IN SECOND-OPINION CONSULTATIONS Vashitz G1, Pliskin JS2, Parmet Y2, Kosashvili Y3, Ifergane G4, Wientroub S5, Davidovitch N2 1 Technion Institute of Technology, Haifa, Israel; 2Ben Gurion University of the Negev, Beer-Sheva, Israel; 3Assaf Harofeh Medical Center, Zerrifin, Israel; 4Soroka University Medical Center, Beer-Sheva, Israel; 5Dana Children’s Hospital, Tel-Aviv Sourasky Medical Center, Sackler Faculty of Medicine, Tel-Aviv University, Tel-Aviv, Israel OBJECTIVES: Discrepancies in diagnosis, treatment, or prognosis may emerge among physicians. a known decision-making bias is the tendency to shift personal opinion either toward or away from a previous opinion. We sought to evaluate such biases in the context of second-opinion medical consultations. METHODS: We distributed a survey questionnaire to a nationwide sample of orthopedic surgeons and neurologists. The questionnaires presented eight scenarios, each with conventional treatment options with no clear-cut preference. In four scenarios, the physicians were told that a previous opinion had already been given by another physician, or that a second opinion will be given, and the other four scenarios were used as controls. The physicians’ responses were coded according to the level of intervention (conservative to interventional). RESULTS: 172 orthopedic surgeons and 160 neurologists filled out the questionnaires, which represent about 50% of these specialties in Israel. In the orthopedic questionnaire, when a first opinion had already been given, there was a shift toward a more interventionist treatment (P < 0.05). This was especially prominent when the first opinion was known to the second physician. When the patient intended to seek a second opinion, there was a shift toward a more conservative treatment. No such effect was found among neurologists. CONCLUSIONS: Physicians’ judgment may be affected by another physician’s opinion (compared to their choices without a first opinion). This bias mainly tends toward a more interventionist treatment. Due to the immense impact of any decision on patient health and resource use, further research should address such biases and develop tools to address them. PODIUM SESSION III: HTA POLICY APPROACHES HT1 REAL-WORLD DATA—IMPROVING APPROACHES FOR DEMONSTRATING BENEFITS AND RISKS OF NEW DRUGS Lucas F1, Arenson SM2 1 Pope Woodhead and Associates, St. Ives, UK; 2Cambridge University, Cambridge, UK OBJECTIVES: Regulators and payers view randomized controlled trials (RCT) as the gold standard for establishing the benefit/risk of new drugs. However, they are increasingly interested in real-world data (RWD) due to their external validity. This survey explored stakeholders’ perceptions and emerging trends in the area of RWD. METHODS: We identified relevant literature since 2006 via Google Scholar and manual search, and reviewed it based on several topics: types of RWD, pros and cons of different approaches, and impact of new statistical techniques and technology on availability and quality of RWD. We then conducted 45–60 min in-depth, semistructured discussions with 17 experts from Academia, HTA bodies, health insurance, research organizations, and pharmaceutical industry—from the UK, France, Germany, the The Netherlands, and the United States. Their views about value and future directions of RWD approaches were elicited. RESULTS: Experts unanimously thought that RCTs would remain a mandatory approach for the foreseeable future due to the limitations of RWD, mainly potential for confounding. New study designs (e.g., randomized database studies) and statistical techniques (e.g., high-dimensional propensity scoring) remove confounding only partially and need to gain credibility. There was a strong view that, while registries have been the reference source of observational data, there is an opportunity for (claims) database and electronic medical records to form an efficient platform for automatic, real-time analysis of naturalistic data. Despite a few good examples, it will, however, require time to resolve technical difficulty of linking databases and, crucially, the challenges of data ownership and privacy issues. Several experts predict the short-term rise of at-home monitors, “smart pills,” and “smart phones” that automatically feed into databases, and the increased use of data from Google Health and Microsoft HealthVault. CONCLUSIONS: RWD may eventually become the new gold standard in drug development, but this will occur only through incremental progress. HT2 INFLUENCE OF HTA AND HOSPITAL FUNDING SYSTEMS ON PATIENT ACCESS TO INNOVATIVE MEDICINES: EXAMPLE OF ADVANCED RENAL CELL CANCER IN WESTERN EUROPE Toumi M1, Misset J2, Jaroslawski S3, Aballea S3, Baynton E4, Lamure M1 1 University Claude Bernard Lyon 1, Lyon cedex 08, Rhone Alpes, France; 2Hôpital Saint-Louis, Paris, Ile de France, France; 3Creativ Ceutical, Paris, Ile de France, France; 4 European Oncology Monitor Synovate Healthcare, London, London, UK OBJECTIVES: Prices of recently launched targeted therapies are relatively high and patient access differs between European countries. Advanced renal cell cancer (aRCC) is a rare malignancy with poor prognosis for which four such drugs are available: sunitinib, sorafenib, bevacizumab, and temsirolimus. We investigated relations between funding decisions of national health authorities and patient access to these treatments. METHODS: We reviewed Web sites of health authorities in France, Germany, Italy, and UK. Data on drug utilization from June 2006 to July 2009 were extracted from the Synovate Oncology Monitor, an ongoing prescription database based on doctors’ diaries. Total sample size varied between countries, from 7766 to 9463 patients within the year ending 2009 Q2. RESULTS: Sunitinib was granted restricted recommendation by NICE (UK) in March 2009, but other treatments were not assessed during study period. All drugs were financed through payment-for-performance schemes (P4PS) with a registry in Italy. In France, bevacizumab and temsirolimus were reimbursed on top of DRGs and in Germany only bevacizumab. Sunitinib was the first line treatment in 62%, 50%, 47%, and 31% of drug-treated patients in France 2006Q3–2009Q2), Germany, Italy, and UK, respectively, followed by temsirolimus in France and Germany but sorafenib in Italy. Sorafenib was the most widely used second line treatment in Germany and Italy. In France, temsirolimus was used off-label in first line and bevacizumab before funding decision in aRCC was granted. In UK, many patients remained untreated. CONCLUSIONS: Funding on top of DRGs in France contributed to early uptake and off-label usage. P4PS were associated with enhanced drug uptake in Italy. In UK, the lack of assessment by NICE prevented patient access, which raises concern about the current financing system. The large differences in patient access to recently approved cancer treatments raise the issue of equity and health outcomes associated with innovative drugs. HT4 EVIDENCE EXPECTATIONS FROM PAYERS ACROSS THE EU: DOES THE DISEASE BURDEN ON HEALTH-CARE BUDGETS HAVE AN EFFECT? Kirpekar S, Mukku SR Double Helix Consulting Group, London, UK OBJECTIVES: Payer expectations for reimbursement from novel drugs are constantly increasing. Understanding these is crucial during evidence generation. Expectations in terms of clinical outcomes vary across EU countries. This study compared payer awareness and expectations in two disease areas with contrasting prevalence and subsequent impact on health-care budgets, and thus their impact on reimbursement. METHODS: The study was conducted in EU5 markets. Opinions of 36 stakeholders were collected via telephone interviews. Value drivers for new drugs were tested including unmet need, clinical data—safety and efficacy, cost-effectiveness, budget A248 impact, price, price against comparator, and innovation. The results in the two disease areas were separately analyzed and then compared with each other qualitatively on the basis depth of the responses. RESULTS: Awareness about the evidence expectations were seen to vary substantially between the disease areas. Payers were seen to be far more conscious in the high-burden disease area. Of the 12 stakeholders interviewed for the high-burden disease area (cardiovascular disease used as an example), all of them could provide an in-depth understanding of the value drivers for a new drug. On the other hand, of the stakeholders interviewed from a low-burden disease area within urology, a very small part of the sample (n = 2) was able to provide details about necessary expectations for the indication. CONCLUSIONS: The variation in awareness among payers in disease areas tends to affect the clinical outcomes that are presented by the drug manufacturer for a novel drug for the indication. The low awareness regarding evidence expectations in disease areas with low budgetary impact may lead to lower quality of evidence being accepted for reimbursement. This needs to be further investigated. PHP22 THE USE OF ECONOMIC EVALUATIONS IN DECISION-MAKING AT MACRO LEVEL: A LITERATURE REVIEW Silva ML1, Späth HM2, Moumjid NF1 1 Research Group in Health Economics, University Lyon 1; University Lyon 2, Lyon, France; 2 University Lyon 1, Lyon, France OBJECTIVES: The number of economic-evaluations (EE) increases but their role in decision-making remains unclear. Our literature review of the use of EE in decisionmaking at macro-level aimed to analyze the methods presented in the articles and to investigate differences within and between countries regarding the use of EE, its facilitators and barriers. METHODS: We searched major databases (PubMed, Web-ofScience, Euronheed, EconLit) in seven languages, without time limit, using keywords related to economics, health care policy decision-making, and outcomes research. Abstracts were screened according to four criteria reflecting the objectives of our review. Selected articles were analyzed and compared using a checklist of items related to study context (period, location, domain . . .), methods (population, design . . .), and outcomes (use of EE, facilitators, barriers . . .). RESULTS: Twenty-nine articles were selected. Most reported single-country- (24), mainly developed-country-studies (20). Five multi-country-studies compared countries of Europe, the USA and LatinAmerica. The studies’ population generally included public or private health authorities and used questionnaires (10), interviews (6), focus-groups (1), observations (4) or a combination of interviews and other methods (8). The main facilitator to using EE in decision-making was governmental/institutional incentives (UK and Australia). Although the use of EE has increased since the late 1990s, barriers remain, without apparent variation, overtime and between countries. Most relate to the accessibility and acceptability of EE. CONCLUSIONS: Decision-makers are increasingly aware of the importance of using EE in their practice. Despite large differences in the level of EE use between countries, barriers are very similar. Studying these barriers could narrow gaps between researchers and decision-makers and encourage governmental incentives to using EE. Consequently, we decided to develop a methodology for assessing the use of EE in radiotherapy decision-making, a domain where many EE have been undertaken, without actually investigating their use. This study will be conducted in France, and in European or developing-countries. PODIUM SESSION III: MAPPING SCORE FROM ONE QOL MEASUREMENT INTO ANOTHER MA1 HOW TO HANDLE THE MAPPING PROBLEM IN COST-UTILITY ANALYSES? Siani C1, de Peretti C1, Castelli C2, Phung T3, Duru G4, Daurès JP2 1 University Claude Bernard Lyon 1, Lyon, France; 2Centre Hospitalier Universitaire de Nimes, Nimes, France; 3Institut Universitaire de Recherche Clinique, Montpellier, France; 4 Cyklad Group, Rilleux la Pape, France OBJECTIVES: In cost-utility analyses (CUA), utility values are rarely available for the entire patients sample and they are generally predicted using a “mapping” interpolation from a functional status questionnaire, known for the entire sample. This mapping method is not accounted for in pharmaceutical industry and in literature studies, when building confidence regions around the utility and the incremental costutility ratio, leading to a wrong confidence region and consequently, to a wrong decision-making. The purpose of this paper is to build a confidence interval (CI) around the mean utility, accounting for the uncertainty coming from the “mapping” interpolation. METHODS: Analytical and bootstrap methods are developed to handle the fact that values are interpolated. Linear, multilinear, and nonlinear mapping are considered. Monte Carlo experiments are carried out to compare the performance of these methods. These methodologies are applied on data issued from an observational study dealing with prostate cancer treatment. Utility is assessed with Standard Gamble method and some of these values are interpolated from the questionnaires: EORTC QLQC-30; IPSS and IIEF-5; SF-36 and Visual Analogic Scales. RESULTS: Monte Carlo experiments show that the analytic and bootstrap 95% CI display coverage between 94% and 96% for various sample sizes. If mapping is not accounted for (“naive method”), the coverage is between 20% and 40%. The cross validation shows similar results. From prostatectomy data, the utility is explained by SF-36, role functioning, diarrhea, and age. For instance, mean utility equals 0.94. The analytic and 13th Euro Abstracts bootstrap CIs equal [0.59, 1.51] and [0.51, 1.63] respectively. The naive interval equals [0.95, 1.15]. CONCLUSIONS: In CUA, decision-making based on utility values interpolated from mapping is not reliable: a naive interval would lead to a serious mistake. The uncertainty due to mapping has to be accounted for. Our analytic and bootstrap procedures, integrating the mapping, provide very accurate results. MA2 ESTIMATING THE ASSOCIATION BETWEEN SF-12 RESPONSES TO EQ-5D UTILITY VALUES USING A RESPONSE MAPPING AND A PARAMETER UNCERTAINTY APPROACH Rivero-Arias O1, Gray A1, Ramos-Goñi JM2 1 University of Oxford, Oxford, Oxfordshire, UK; 2Servicio de Evaluación del Servicio Canario de Salud, Santa Cruz de Tenerife, Tenerife, Spain BACKGROUND: Two types of mapping exercises have been suggested to translate data from the short-form health instrument SF-12 into measurements for the EQ-5D instrument. One of the trends associates SF-12 data directly to the EQ-5D index, whereas a second approach suggests a response mapping method where SF-12 data have been mapped to the five domains of the EQ-5D questionnaire. OBJECTIVES: To improve the current response mapping approach and provide a comparison between the direct and the response mapping approaches to the EQ-5D. METHODS: Five multinomial logit regressions were implemented to estimate the association between SF-12 variables and each of the EQ-5D domains. Predicted EQ-5D responses were estimated using a Monte Carlo method. a parameter uncertainty approach was introduced to calculate confidence intervals for the predicted EQ-5D index. The direct mapping approach for the comparison was also conducted. Several large data sets were used for internal and external validation. Actual versus predicted EQ-5D index were compared using mean square error (MSE), mean absolute error (MAE), and confidence intervals. RESULTS: In the internal validation data set, both the response and direct mapping predicted a similar mean EQ-5D index, but the response mapping yielded a smaller MSE of 0.018 compared to 0.020 in the direct mapping, and a smaller MAE of 0.091 in the response mapping compared with 0.105 in the direct mapping method. Using the external validation data set, the MSE and MAE were one decimal point less in the response compared to the direct mapping. CONCLUSIONS: The revised response approach provided marginally better results than the direct mapping method. The response method can be implemented to country-specific EQ-5D data with available value sets. To facilitate the implementation of the revised response mapping algorithm, a Stata command has been programmed. MA3 GLAUCOMA DISEASE PROGRESSION STATES: ESTABLISHING A DIRECT, QUANTITATIVE LINK BETWEEN VISUAL FIELD DEFECTS AND UTILITY LOSS Gerlier L1, Lamotte M1, Verboven Y2 1 IMS Health Consulting, Brussels, Belgium; 2ALCON Couvreur NV, Puurs, Belgium OBJECTIVES: Treatments of ocular hypertension (OHT) and glaucoma aim mainly at maintaining the patient’s visual function and related quality of life (QoL). However, it is not known to what extent the disease progression impacts the health-related QoL. Using recent publications, the objectives were to 1) derive a range of EQ-5D utility scores in OHT/glaucoma patients; (2) establish a direct, quantitative link between the mean defect in perimetry (MD) measured in decibels (db) and the utility loss; and (3) estimate the utility level of patients reaching a glaucoma stage. METHODS: The glaucoma stages were characterized using MD thresholds of 0 db (OHT), 0–6 db (early glaucoma), 6–12 db (moderate), 12–22 db (advanced), and >22 db (legal blindness), as per the Hodapp classification. An algorithm allowing to derive EQ-5D utility scores from the eight mean subscale scores to the SF-36 (Ara and Brazier 2008) was applied to the range of SF-36 scores published in a systematic review of QoL in glaucoma (Mills 2009). Besides, the same algorithm was used to estimate the utility loss corresponding to a MD of 1 db, based on a regression model assessing the correlation between visual function and QoL in glaucoma patients (Lin and Yang 2010). RESULTS: Over eight studies in OHT/glaucoma patients, utility scores ranged from 0.65 to 0.89. Based on a significant regression coefficient for the SF-36 Role Physical subscale, a utility loss of 0.0295/db was calculated. Applying this utility loss/db to the predefined thresholds, the ranges of unadjusted utility scores expected per glaucoma stage equal: 0.72–0.90 (early), 0.55–72 (moderate), 0.25–0.55 (advanced), and <0.25 (blindness), assuming a baseline utility of 0.90 for OHT. CONCLUSIONS: These utility scores per disease stage are consistent with published direct measures of HUI3. The estimated utility loss per db is sizeable and could be implemented in cost-utility models where disease progression is tracked. MA4 ANCHOR-BASED DETERMINATION OF THE MINIMAL IMPORTANT DIFFERENCE OF A PRO SCALE—A CRITICAL LOOK ON A WIDELY USED METHOD BY MEANS OF A SIMULATION STUDY Kemmler G, Giesinger J, Holzner B Innsbruck Medical University, Innsbruck, Tyrol, Austria OBJECTIVES: Anchor-based methods are frequently used for determining the minimal important difference (MID) of scales employed to measure patient-reported outcomes (PRO). The anchor may, e.g., consist of a global rating by the patient or the doctor or of a clinical measure closely related to the issue to be measured. The role of the psychometrical properties of the anchor has been rarely studied in this context. Aim of this contribution is to shed more light on the relationship between A249 13th Euro Abstracts the reliability of the anchor and the estimated MID. METHODS: We performed a simulation study in which the reliability of the anchor used for MID estimation was varied systematically. Features of real-life data (e.g., skewed distribution, discreteness of PRO scale) and anchors were used to generate simulated PRO scales and anchors. MIDs were then estimated on the basis of the simulated data. RESULTS: Compared to the MID value obtained with an anchor with perfect reliability (r = 1), a marked attenuation of the MID was observed when reducing the reliability of the anchor. Thus, an anchor with reliability 0.7 gave rise to a 24% to 35% decrease of the MID estimate and an anchor with reliability 0.5 led to a 45% to 55% reduction. Based on the findings and on theoretical considerations, we suggest a method for bias correction. CONCLUSIONS: When determining the MID of a PRO scale by an anchor-based method, the reliability of the anchor plays a crucial role. Anchors with poor to moderate reliability may lead to considerable underestimation of the MID. Bias correction is possible provided the reliability of the anchor is known. PODIUM SESSION III: PRICING AND MARKET ACCESS PR1 THE APPLICATION OF PHARMACOECONOMIC MODELING TO ESTIMATE A VALUE-BASED PRICE FOR NEW CANCER DRUGS IN A PUBLICLY FUNDED HEALTH-CARE SYSTEM Dranitsaris G1, Truter I1, Lubbe M2, Spirovski B3, Cottrell W4, Edwards J5 1 Nelson Mandela University, Port Elizabeth, Eastern Cape, South Africa; 2North-West University, Potchefstroom, North-West Provi, South Africa; 3Humber River Regional Cancer Centre, Weston, ON, Canada; 4Princess Margaret Hospital, Toronto, ON, Canada; 5Dr. H. Bliss Murphy Cancer Center, St Johns’s, NF, Canada OBJECTIVES: Value-based pricing has recently been discussed by international organizations as a means to estimate a drug price that is linked to the benefits it offers patients and society. However, one of the challenges with value-based pricing is determining the optimal threshold for health policy decision-making. The World Health Organization (WHO) has recommended using multiples of a country’s per capita GDP as the value threshold. In this study, pharmacoeconomic modeling was used to estimate a value-based monthly price for a hypothetical new cancer drug that provides a 3-month survival to patients with metastatic colorectal cancer (mCRC). METHODS: A decision model was developed to simulate progression free and overall survival in mCRC patients receiving standard chemotherapy ± the new drug. Outcomes for cancer control and side effects were abstracted from randomized trials in mCRC. Costs for chemotherapy were obtained from Canadian cancer centers. Utility estimates measured as quality-adjusted life-years (QALYs) were determined by interviewing 24 oncology nurses and pharmacists using the Time Trade-Off technique. The monthly price of the new drug was then modeled using a threshold of $117,000 per QALY gained, which is three times the Canadian per capita GDP, as recommended by the WHO. RESULTS: The analysis suggested that a monthly price of $2180 would be considered cost-effective from the Canadian public health perspective. If the drug were able to improve patient quality of life or survival from 3 to 6 months, the monthly price could increase to $4100 and $3430 and offer the same value. CONCLUSIONS: The use of the WHO criteria for estimating a value-based price is feasible. However, one of the challenges would be to identify an appropriate threshold that would provide a balance between what governments can afford to pay and the commercial viability of the product in the reference country. PR2 DECIDING ON VALUE FOR MONEY: A COMPARISON OF THE DUTCH, BELGIAN, SWEDISH, AND FRENCH DRUG REIMBURSEMENT SYSTEMS Franken M1, le Polain M2, Senn A2, Koopmanschap M1, Cleemput I2 1 Erasmus University, Rotterdam, The Netherlands; 2Belgian Health Care Knowledge Centre, Brussel, Belgium OBJECTIVES: Many countries are adapting their pharmaceutical reimbursement system, increasingly emphasizing the role of pharmacoeconomics in decision-making. The aim of our study is to analyze European regulatory systems to obtain insight into best practice systems that deliver value for money. METHODS: The analytical Hutton Framework was used for comparing and assessing “fourth hurdle” drug reimbursement systems in the The Netherlands, Belgium, Sweden, and France. We investigated policy documents, explored literature, and conducted interviews with policymakers and representatives of the pharmaceutical industry. RESULTS: All systems have a centralized decision body, even though the financial responsibility may be regional (Sweden). Only in Sweden, the minister has no role at the individual reimbursement decision level. None of the systems has a fully independent evaluation process and the impact of the systems is mainly assessed on drug expenditure. All countries make efforts to increase transparency. However, in Sweden manufacturers may withdraw their application before the final reimbursement decision, guaranteeing confidentiality at the cost of less transparency. Policies to deal with uncertainty vary per country: financial risk-sharing agreements by price/volume contracts—France—versus outcomes-based agreements for expensive inpatient drugs—the The Netherlands. The actual value of a drug and disease severity is reflected in the level of reimbursement in France and Belgium, whereas in the The Netherlands and Sweden, enlisted drugs are fully reimbursed. All countries attempt to increase the importance of pharmacoeconomics in decision-making. However, no country expresses the relative importance of cost-effectiveness compared to other criteria nor applies a strictly defined threshold. CONCLUSIONS: This study reveals that while there is a convergence in scientific evaluation processes, important differences remain between the Dutch, Belgian, Swedish, and French regulatory frameworks. All countries recognize that pharmacoeconomics has a place in decision-making on value for money, but for the time being, pharmacoeconomics seems to play a rather undefined role. PR3 THE IMPACT OF FINNISH PHARMACEUTICAL PRICING SCHEME IN COST-EFFECTIVENESS ANALYSES Hallinen T, Soini EJ ESiOR Ltd, Kuopio, Finland OBJECTIVES: Finnish retail prices for drugs are determined with a pricing scheme (PS). The PS is of the form “multiplier × wholesale price + fixed sum.” The multiplier ranges from 1.125 to 1.5 (being smaller for higher wholesale prices), whereas the fixed sum ranges from 0.5 to 47.68 euros (being larger for higher wholesale prices). Although PS is regressive, it nevertheless provides higher absolute pharmacy margins for drugs with higher wholesale prices. At the lower end of wholesale prices, PS results in retail prices that do not cover dispensing costs. Despite this, the retail prices (excluding VAT 8%) are used to represent all drug and drug delivery costs in economic evaluations. This study assesses the impact of this Finnish system-derived “distortion” in cost-effectiveness analyses. METHODS: The cost utilities of new hypothetical treatments were assessed in a setting where the new and old treatments produce different amounts of quality-adjusted life-years (QALYs) and the only cost difference comes from the pharmaceutical prices. The treatments are assumed not to differ regarding the real costs of drug delivery and patient survival. The PS-induced computational cost difference was deducted from the retail price differences of new and old treatments to estimate the impact of PS on the incremental cost-effectiveness ratios (ICER). RESULTS: The computational cost differences due to PS ranged from 7.3 to 1 951 euros and the QALYs gained ranged from 0.004 to 0.070 in estimated scenarios. The respective ICERs increased by 104 to 487 840 euros/QALY due to the PS. CONCLUSIONS: The PS significantly worsens the ICERs obtained for more expensive and often innovative pharmaceuticals. The Finnish PS is problematic when the aim is to provide optimal, cost-effective treatments to Finnish patients. In the current form, the PS discourages innovation and may prevent reimbursement of otherwise cost-effective treatments. PR4 GLOBAL MARKET ACCESS STRATEGY: AN INTEGRATED APPROACH 1 2 Pang F , Mukku S 1 Shire Human Genetic Therapies, Inc, Basingstoke, UK, 2Double Helix Consulting Group, London, UK OBJECTIVES: To develop a framework for integrating pricing and reimbursement with health economics and outcomes research and health policy to achieve commercially desirable prices and levels of access in 2010 and beyond. METHODS: A review of recent pricing policy and regulatory changes of countries, especially those in the financially troubled Eurozone, was conducted. This was supplemented by a review of P&R decisions for a selection of drug launches between 2005 and 2009 and categorized according to the level of therapeutic innovation and disease type (conventional, rare diseases, oncology). a search was performed on the OHE and NHS EED databases and HTA reports to establish the level of published value evidence in support of these launches, and finally, the components of most importance to a market access strategy were identified and validated through interviews across different stakeholder types. RESULTS: The review identified since January 2010, there have been 11 pricing policy and regulatory changes. From the review of recent P&R decisions and stakeholder interviews, the main components identified were: competitive and environmental analysis (market assessment, reimbursement, revenue forecasts, policy trends); analysis of payer’s decision drivers (payer, physician, and other stakeholder qualitative research); value demonstration (value hypotheses, economic modeling, patientreported outcomes, scientific advice); pricing strategy (price targets, cross-market revenue optimization modeling, country launch sequencing, scenario planning); and local market access tactics (HTA, risk sharing, contracting negotiations with payers). The review of the P&R decisions also demonstrated an increasing trend toward deployment of risk-sharing schemes since 2008. CONCLUSIONS: Development of a successful market access strategy requires an understanding of pricing, health economics and outcomes research, health technology assessment (HTA), and health policy, and continually keeping vigilant and adapting to rapid changes in the policy environment. This research gives direction to health economics, P&R, and government affairs professionals for the development of an integrated framework for the design and implementation of a global market access strategy. PODIUM SESSION III: HTA IN VACCINE AND EPIDEMICS VA1 ARE THE BENEFITS OF FLU VACCINATION IN THE ELDERLY CORRECTLY SIMULATED IN ECONOMIC ASSESSMENT MODELS? Chen YC1, Van Bellinghen LA2, Van Vlaenderen I2, Standaert B1 1 GlaxoSmithKline Biologicals, Wavre, Belgium; 2Deloitte, Diegem, Belgium BACKGROUND: In literature, economic models of flu vaccination in elderly (65+) most often consider the target population as one homogeneous age group evaluated during a 1-year time period (= 1-year 65+ group cohort model). Because the mortality A250 rates in elderly steeply increase with age and the transition probabilities for specific health states are age specific, this population should be evaluated as non-homogeneous and over a lifetime with annual vaccinations. This study compares the cost results of flu vaccination between these two different modeling approaches. METHODS: Two models were developed to estimate the direct costs of annual flu vaccination compared with no vaccination: 1) a 1-year 65+ group cohort model; and 2) a lifetime multi-age cohort model with target population and clinical pathways stratified in five age cohorts (65–69 years; 70–74 years; 75–79 years; 80–84 years; 85+ years) eligible for annual vaccination. Both models were populated with US specific data. Vaccination coverage and disease management were identical in both models. The decision tree included the following states: natural deaths, infected, and symptomatic states followed by GP visits, hospitalizations (pneumonia, influenza, stroke, myocardial infarction, and congestive heart failure), disease-specific death rates, and recovery in nursing homes. Undiscounted costs per individual per year are compared for vaccinated and unvaccinated groups, using both approaches. RESULTS: The cost per individual per year is higher in the 1-year 65+ group cohort model versus the lifetime multi-age cohort model (no vaccination: $205 vs. $139; vaccination: $185 vs. $113) as expected: considering additional age cohorts with decreasing life expectancies in the multi-age cohort lowers the average cost per individual per year. Meanwhile, the selection of model type impacts the estimated incremental cost of vaccinated versus unvaccinated groups ($−20 vs. $−26). CONCLUSIONS: In economic assessments, a 1-year 65+ group cohort approach undervalues the impact of heterogeneity in elderly on the benefit of flu vaccination, and therefore, a lifetime multi-age cohort is preferred. VA2 GATHERING INFORMATION BY COMPARISON OF DIFFERENT DYNAMIC MODELING APPROACHES FOR EPIDEMIC MODELS Zauner G1, Popper N1, Miksch F2, Urach C2, Einzinger P2, Endel G3, Schiller-Frühwirth I3, Breitenecker F2 1 Dwh Simulation Services, Vienna, Austria; 2Vienna University of Technology, Vienna, Austria, 3 Main Association of Austrian Social Security Institutions, Vienna, Austria OBJECTIVES: Several dynamic approaches can simulate epidemics and vaccination strategies. Generally, the models can be divided into top-down approaches like Markov models and differential equations and bottom-up approaches like cellular automata and agent-based models. Top-down approaches are characterized by cumulative values that are representing groups of people. Bottom-up approaches, in contrast, consider individuals. Both approaches have advantages and disadvantages. Top-down approaches can be analyzed very well with mathematical methods, while bottom-up approaches require comparison of the outcome of simulation runs with different parameter sets. To improve validity of model structures, a method that compares different approaches for epidemic models is introduced. METHODS: Statistical calculations and Markov models are static, while other approaches like differential equations or individual-based models are dynamic. In this context, dynamic does not only stand for simulation over time but also for models where the calculation of the next time step or period depends on the current state of the model. Since the transition matrices in Markov models are calculated before execution time, it is not considered to be dynamic. The advantage of dynamic models is that they can produce highly nonlinear behavior that cannot be reached with static calculations. To validate the structure of such nonlinear models, different model types are implemented and compared. Results are compared; sensitivity analysis is done separately. RESULTS: Outcome of vaccination against streptococcus pneumoniae was tested. a differential equations model and an agent-based model could reproduce results of published Markov models. As soon as we consider population dynamics, herd immunity, and serotype replacement, the Markov model was not able to fulfill the structural requirements anymore, while dynamic approaches still work. CONCLUSIONS: Dynamic models offer more information and opportunities for epidemic simulation. Usage of different approaches provides at least comparable reliability. PIN61 INFLUENZA RISK AND VACCINATION RATES IN EUROPE: A NATIONWIDE SURVEY OF ADULTS DiBonaventura M1, Goren A1, Gupta S2, Wagner JS1, Freedman D2 1 Kantar Health, New York, NY, USA; 2Kantar Health, Princeton, NJ, USA OBJECTIVES: The aim of the current study was to determine influenza vaccination rates among high- and non-high-risk adults across Europe (UK, France, Germany, Italy, and Spain). METHODS: Data from the 2008 EU National Health and Wellness Survey (NHWS) were used. Demographics, comorbidities, and vaccination behavior in the past year were assessed for all respondents. Health-related quality of life (SF12v2) and resource use (number of emergency room visits, hospitalizations, and physician visits) in the past 6 months were also measured. RESULTS: Only 23.7% of respondents received an influenza vaccine in the past year (UK: 25.3%, Germany: 25.2%, France: 20.2%, Italy: 24.8%, Spain: 24.1%). a total of 28,158 respondents (52.6%) were at high risk for influenza complications (i.e., over age 50, had chronic conditions such as asthma, diabetes, COPD, cardiovascular conditions, or HIV/AIDS). Those at high risk reported significantly lower levels of both physical quality of life (mean = 45.88 vs. mean = 52.10) and health utilities (mean = 0.72 vs. mean = 0.73), and significantly higher levels of emergency room visits (mean = 0.21 vs. mean = 0.17), hospitalizations (mean = 0.22 vs. mean = 0.12), and provider visits (mean = 5.96 vs. mean = 4.14) in the past 6 months relative to those not at high risk, all P < 0.0001. Despite the significantly worse health profile, only 35.9% of high-risk respondents received the vaccine. High-risk status was the strongest driver of vaccination in the 13th Euro Abstracts UK (high risk: 42.2% vaccinated vs. non-high risk: 5.4% vaccinated, Φ = 0.42) and the weakest in Germany (high risk: 31.8% vaccinated vs. non-high risk: 16.2% vaccinated, Φ = 0.18). The most common reason for nonvaccination was a belief that the vaccine was unimportant (35.9%). CONCLUSIONS: Despite influenza vaccine recommendation guidelines, only a modest percentage of respondents in Europe were vaccinated. Even those at high risk for influenza complications, who reported significantly worse health outcomes than non-high-risk respondents, were vaccinated at less than a 40% rate. VA4 COST-EFFECTIVENESS OF UNIVERSAL HEPATITIS B IMMUNIZATION IN VIETNAM: APPLICATION OF COST-EFFECTIVENESS AFFORDABILITY CURVES IN HEALTH DECISION-MAKING Tu HAT, de Vries R, Woerdenbag HJ, van Hulst M, Postma MJ University of Groningen, Groningen, The Netherlands OBJECTIVES: To perform a cost-effectiveness analysis of newborn universal vaccination against hepatitis B virus (HBV) and to identify the cost-effective affordability levels of the vaccination program in Vietnam. METHODS: We simulated a birth cohort using 1,693,000 newborns in 2002. Incremental cost-effective ratios (ICERs) per quality-adjusted-life-year (QALY) gained with universal newborn vaccination against HBV was calculated using a Markov model. Two types of analyses (including and excluding expenditure on the treatment of chronic hepatitis B and its complications) were performed. We used 5000 Monte Carlo simulations to examine the costeffectiveness acceptability and affordability of the vaccination program from the payer’s perspective and to derive a cost-effective affordability curve to assess the program’s cost and health effects. All costs were expressed in 2002 US dollars. RESULTS: In the base-case scenario, newborn universal vaccination against HBV reduced the carrier rate by 58% at a cost of US$42 per carrier averted. From the payer’s perspective, marginal cost per life-year and per QALY gained were US$4.76, much lower than GDP per capita of ~US$440 in 2002. The vaccination could be potentially affordable starting at a relatively low budget of US$1.7 million. Newborn universal vaccination would save US$ 1 billion from the treatment cost of complications due to chronic HBV infections. The probability of vaccination being both costeffective and affordable is 27% at an annual budget of US$4.1 million at the cost-effectiveness threshold of US$3.9 per QALY. CONCLUSIONS: Universal newborn vaccination against HBV is highly cost-effective in Vietnam. In low-income, high-endemic countries, where funds are limited and economic results of vaccination are uncertain, our findings on the cost-effectiveness affordability options would assist decision-makers in making proper health investments in vaccination strategies against HBV. POSTER SESSION I CANCER – Clinical Outcomes Studies PCN1 TOLERABILITY OF FIRST-LINE TREATMENTS OF LOCALLY ADVANCED OR METASTATIC NON-SMALL-CELL LUNG CANCER (NSCLC): A SYSTEMATIC REVIEW AND ADJUSTED INDIRECT COMPARISON Edwards SJ1, Welton N2, Borrill J1 1 AstraZeneca UK Ltd, Luton, Bedfordshire, UK; 2University of Bristol, Bristol, UK OBJECTIVES: Platinum-based chemotherapy is a common first-line treatment of NSCLC; tolerability impacts on choice of regimen. This research compared the tolerability of gefitinib and doublet chemotherapy in this setting in patients with activating epidermal growth factor receptor tyrosine kinase (EGFR-TK) mutations (M+). METHODS: Systematic searching of CENTRAL, EMBASE, and MEDLINE for randomized controlled trials (RCTs) comparing ≥2 doublet chemotherapies (carboplatin or cisplatin in combination with either docetaxel, gemcitabine, paclitaxel, pemetrexed, or vinorelbine) for the first-line treatment of advanced NSCLC was completed in May 2009. Data were extracted on the following grades 3/4/5 adverse events (AEs) most commonly reported with doublet chemotherapy or EGFR-TK inhibitors: anemia, diarrhea, fatigue, febrile neutropenia, nausea/vomiting, neutropenia, and rash. We performed a meta-analysis of the available gefitinib versus paclitaxel/carboplatin RCTs in EGFR-TK M+ patients. We then carried out a mixed treatment comparison (MTC) of doublet chemotherapies in unselected advanced NSCLC patients using paclitaxel/ carboplatin as a baseline. Treatment effect for the risk of AE occurrence was estimated as an odds ratio (OR > 1.0 favors paclitaxel/carboplatin). RESULTS: Three RCTs were identified for gefitinib, of which two were comparisons with paclitaxel/carboplatin. Metaanalysis of these two trials gave the following statistically significant results: anemia—OR 0.12, 95% confidence interval: 0.03–0.47; diarrhea—OR 5.78, 95% CI: 1.01–33.11; neutropenia—OR 0.01, 95% CI: 0.00–0.03. Twenty-nine trials were appropriate for inclusion in the MTC. The alternative doublet chemotherapy regimens did not demonstrate a statistically significant reduction in risk of any of the AEs assessed versus paclitaxel/ carboplatin, with the exception of gemcitabine/cisplatin, which had a lower risk of febrile neutropenia (OR 0.39, 95% credible interval: 0.12–0.96). CONCLUSIONS: In the absence of RCTs comparing all doublet chemotherapies with gefitinib in EGFR-TK M+ patients with advanced NSCLC, this adjusted indirect comparison suggests that gefitinib may have important tolerability advantages over other first-line treatments in this targeted population. 13th Euro Abstracts PCN2 PALONOSETRON VERSUS OTHER 5-HT3-RECEPTOR ANTAGONISTS IN PREVENTING CHEMOTHERAPY-INDUCED NAUSEA AND VOMITING IN PATIENTS WITH CANCER TREATED IN A HOSPITAL OUTPATIENT SETTING Craver C1, Gayle J1, Balu S2, Buchner D2 1 Premier, Inc, Charlotte, NC, USA; 2Eisai, Inc., Woodcliff Lake, NJ, USA OBJECTIVES: To explore the risk of uncontrolled chemotherapy-induced nausea and vomiting (CINV) associated with palonosetron (a 5-hydroxy tryptamine3-receptor antagonist [5-HT3-RA]) initiation versus other 5-HT3-RAs among patients with cancer on chemotherapy (CT) treatment in a hospital outpatient setting. METHODS: Patients with a cancer diagnosis initiating CT and anti-emetic prophylaxis with palonosetron (Group 1) and other 5-HT3-RAs (Group 2) for the first time (index date) between April 1, 2007 and March 31, 2009 were identified from the Premier Perspective (TM) Database. Inclusion criteria were patients aged ≥18 years, no evidence of nausea and vomiting or a hospital charge for a CT or anti-emetic medication in the 6-month pre-index date period, and 36 consecutive months of hospital data submission. Patients were followed through eight CT cycles or 6 months post-index date, whichever occurred first. a negative binomial distribution-generalized linear multivariate regression model estimating the number of CINV events on CT-matched groups in the follow-up period was developed after adjusting for several demographic and clinical variables. RESULTS: Of 9144 identified patients, 1775 initiated palonosetron (Group 1; 19.4%). Group 1 patients were significantly younger (61.2 [SD: 13.0] vs. 62.8 [13.1] years; P < 0.0001), comprised more females (52.5% vs. 41.1%; P < 0.0001), less Blacks (8.6% vs. 13.2%; P < 0.0001), more highly emetogenic CT (43.3% vs. 28.5%; P < 0.0001), and more lung (26.1% vs. 22.4%; P < 0.0001) and breast cancer patients (19.3% vs. 15.3%; P < 0.0001). In the follow-up period, the number of unadjusted CINV events was statistically lower for Group 1 (8336 vs. 9400; P = 0.0007) patients. The regression model predicted a 13.7% decrease in the total CINV events per patient per cycle for Group 1 patients versus Group 2 patients; P = 0.0006. CONCLUSIONS: In this retrospective hospital outpatient study, patients with cancer initiated on palonosetron were more likely to experience a significantly lower rate of CINV events versus those initiating other 5-HT3-RAs. PCN3 CHEMOTHERAPY-INDUCED NAUSEA AND VOMITING EVENTS BY CHEMOTHERAPY EMETOGENICITY IN PATIENTS WITH CANCER TREATED IN A HOSPITAL OUTPATIENT SETTING Craver C1, Gayle J1, Balu S2, Buchner D2 1 Premier, Inc, Charlotte, NC, USA; 2Eisai, Inc., Woodcliff Lake, NJ, USA OBJECTIVES: To characterize the changes in chemotherapy-induced nausea and vomiting (CINV) events by chemotherapy (CT) emetogenicity among patients with cancer initiating CT treatment in a hospital outpatient setting. METHODS: Patients with cancer initiating CT for the first time (index date) between April 1, 2007 and March 31, 2009 were extracted from the Premier Perspective comparative database. Patients aged ≥18 years, no evidence of nausea and vomiting or a hospital charge for a CT agent or antiemetic medication in the 6-month pre-index period, and 36 consecutive months of hospital data submission were included. Patients were followed through eight CT cycles or 6 months post-index date, whichever occurred first. CT was categorized as highly emetogenic (HEC), moderately emetogenic (MEC), low emetogenic (LEC), or minimal emetogenic (MinEC) per National Comprehensive Cancer Network guidelines. a descriptive analysis of changes in CINV events (either a diagnosis of nausea and/or vomiting or evidence of CINV-related medications) per CT cycle per CT emetogenicity was performed in the follow-up period. RESULTS: The overall study population (N = 11,495) had an average age of 63.3 years (SD 13.4), was 50.7% female, and 86% white. Most common tumor types were lung (19.8%), breast (15.9%), and urinary tract (13.8%). Use of HEC (cycle 1: 26.0% vs. study end: 9.8%; P = < 0.0001) decreased over the follow-up period. Change in CINV events from cycle 1 to study end was statistically significant for HEC (32.1% vs. 29.9%; P < 0.0001), MEC (38.4% vs. 39.9%; P < 0.0001), LEC (26.4% vs. 27.8%; P < 0.0001), and MinEC (3.2% vs. 2.4%; P < 0.0001). The % point change from cycle 1 to study end was also statistically significant between groups (HEC [−2.2%] vs. MEC [+1.5%] vs. LEC [+1.4%] vs. MinEC [−0.8%]; P < 0.0001). CONCLUSIONS: In this retrospective hospital outpatient study, patients with cancer initiated on MEC and LEC experienced significantly higher CINV events over time versus HEC-initiated patients. PCN4 LIKELIHOOD OF A SUBSEQUENT CHEMOTHERAPY-INDUCED NAUSEA AND VOMITING (CINV) EVENT IN PATIENTS RECEIVING MODERATELY OR HIGHLY EMETOGENIC CHEMOTHERAPY (MEC/HEC) Feinberg B1, Gilmore J1, Haislip S1, Jackson J2, Jain G2, Balu S3, Buchner D3 1 Georgia Cancer Specialists, Atlanta, GA, USA; 2Xcenda, LLC., Palm Harbor, FL, USA; 3Eisai, Inc., Woodcliff Lake, NJ, USA OBJECTIVES: CINV with MEC and HEC therapy is well studied, but the association of prior history of CINV with the future risk of CINV is not well quantified. This study assessed the increased likelihood of a subsequent CINV following a first administration CINV in patients on single-day MEC/HEC therapy. METHODS: A retrospective analysis was conducted utilizing Georgia Cancer Specialists electronic medical records database (October 2006–August 2009). Patients who received >1 single-day MEC/HEC administration with no chemotherapy 3 months prior were included. Patients who received multiday chemotherapy, started with low emetogenic chemotherapy, or had no dosing information were excluded. Two cohorts, a first administra- A251 tion CINV (Group 1) and no first administration CINV (Group 2), were created and followed for 6 months. a multivariate logistic regression assessed the likelihood of subsequent CINV during the 6-month follow-up, controlling for age, gender, Charlson comorbidity index, cancer type, gap between administrations, and chemotherapy emetogenicity. Sub-analyses were performed for patients initiated on MEC and HEC. RESULTS: A total of 3721 patients met inclusion criteria; 423 (11.37%) experienced a first administration CINV. These patients were younger (56.6 vs. 59.4; P < 0.0001), had lower comorbidity index (2.1 vs. 2.2; P = 0.0154), and had more gaps between administrations (18.5 vs. 17.4; P = 0.0204). Unadjusted subsequent CINV rate was higher in the Group 1 cohort (52.3% vs. 24.2%; P < 0.0001). After controlling for covariates, Group 1 patients were 3.5 times more likely to have a subsequent CINV compared to Group 2 patients (odds ratio [OR]: 3.48 [95% CI: 2.81–4.30]; P < 0.0001). Sub-analyses by MEC/HEC supported overall analysis (HEC OR: 2.9 [95% CI: 2.1–3.9; P < 0.0001]) and (MEC OR: 4.1 [95% CI: 3.1–5.5; P < 0.0001]). CONCLUSIONS: In this retrospective analysis, patients receiving single-day MEC/HEC who had a prior CINV were at increased risk of subsequent CINV. Further research on the clinical and economic impact of early and appropriate anti-emetic prophylaxis is required. PCN5 EFFICACY OF PALONOSETRON (PAL) COMPARED TO OTHER SEROTONIN INHIBITORS (5-HT3R) IN PREVENTING CHEMOTHERAPYINDUCED NAUSEA AND VOMITING (CINV) IN PATIENTS RECEIVING MODERATELY OR HIGHLY EMETOGENIC (MOHE) TREATMENT: AN UPDATE OF THE PREVIOUSLY PUBLISHED SYSTEMATIC REVIEW AND META-ANALYSIS Botrel TEA, Clark O, Clark LGO, Paladini L, Faleiros E, Pegoretti B MedInsight-Evidências, Campinas, Brazil OBJECTIVES: To perform an update of the previously published systematic review and meta-analysis (Engel T. et al. ISPOR 2009) of all randomized controlled trials comparing a single intravenous dose of PAL 0.25 mg with other 5-HT3R in patients receiving MoHE chemotherapy. METHODS: Several databases were searched, including MEDLINE, EMBASE, LILACS, and CENTRAL. The primary end points were the incidence of acute and delayed nausea and vomiting. The adverse events of each treatment were analyzed. a subgroup analysis was performed to evaluate the impact of the use of concomitant corticosteroids. The results are expressed as risk ratio (RR) and the correspondent 95% confidence interval (CI). RESULTS: Six studies were included, comprising 2201 patients. PAL was compared to Ondansetron, Granisetron, and Dolasetron. Patients in PAL group had less nausea, both acute (RR = 0.86; CI 95% = 0.76 to 0.96; P = 0.007) and delayed (RR = 0.82; CI 95% = 0.75 to 0.89; P < 0.00001). They also had less acute vomiting (RR = 0.77; CI 95% = 0.67 to 0.88; P = 0.0001) and delayed vomiting (RR = 0.78; CI 95% = 0.70 to 0.86; P < 0.00001). There were no statistical differences in adverse events like headache (RR = 0.84; CI 95% = 0.61 to 1.17; P = 0.30), dizziness (RR = 0.40; CI 95% = 0.13 to 1.27; P = 0.12), constipation (RR = 1.29; CI 95% = 0.77 to 2.17; P = 0.33), or diarrhea (RR = 0.67; CI 95% = 0.24 to 1.85; P = 0.44). Patients receiving PAL presented less nausea and vomiting regardless of the use of corticosteroids. CONCLUSIONS: PAL was more effective than the other 5-HT3R in preventing acute and delayed CINV in patients receiving MoHE treatments, regardless of the use of concomitant corticosteroids. PCN6 PRESERVED ANTITUMOR ACTIVITY AND REDUCED CARDIOTOXICITY OF FIRST-LINE PEGYLATED LIPOSOMAL DOXORUBICIN COMPARED WITH CONVENTIONAL DOXORUBICIN IN PATIENTS WITH METASTATIC BREAST CANCER Mann K, Rana C, Wadhwa A Heron Health Private Ltd., Chandigarh, India OBJECTIVES: Doxorubicin is an anthracycline used in the treatment of breast cancer, but its use is limited by cumulative dose-dependent cardiotoxicity. Pegylated liposomal doxorubicin is a novel drug delivery system that alters biodistribution of doxorubicin resulting in reduced cardiotoxicity. Our aim is to compare the efficacy and cardiotoxicity of pegylated liposomal doxorubicin and conventional doxorubicin in first-line treatment of metastatic breast cancer (MBC). METHODS: The literature databases (Cochrane Library, EMBASE, and PUBMED) were searched from inception to May 2010. Randomized controlled trials assessing the efficacy and cardiotoxicity of pegylated liposomal doxorubicin compared to conventional doxorubicin in first-line treatment of MBC were included. Two reviewers independently selected trials, assessed quality, and extracted data, and a third reviewer resolved discrepancies. The fixed effects meta-analysis was performed in STATA 9.0 using a standard meta-analysis approach. RESULTS: Two studies assessing pegylated liposomal doxorubicin with conventional doxorubicin for first-line treatment of MBC met the study criteria. Overall survival (HR = 0.869 [95% CI; 0.720, 1.049]) and overall response rate (RR = 0.908 [95% CI; 0.725, 1.139]) were comparable between the two arms. Treatment with pegylated liposomal doxorubicin resulted in significant reduction in cardiotoxicity as compared to conventional doxorubicin (RR = 0.312 [95% CI; 0.198, 0.490]). The results of random effect analysis were similar (data not shown). CONCLUSIONS: Pegylated liposomal doxorubicin provides comparable antitumor activity in terms of efficacy and is better tolerated as compared to conventional doxorubicin in the treatment of MBC. Thus, the liposomal drug delivery system can play a significant role in the use of doxorubicin in MBC treatment which is otherwise limited by its cardiotoxicity and supports its use in first-line treatment of MBC. A252 PCN7 ASSESSING INTERFERON-ALPHA MONOTHERAPY IN PATIENTS WITH ADVANCE OR METASTATIC RENAL CELL CARCINOMA Rai MK1, Nair SR1, McEwan P2 1 CRC, Capita India Pvt. Ltd, Mumbai, Maharashtra, India; 2CRC, Cardiff, UK OBJECTIVES: The objective was to evaluate the clinical efficacy and safety of interferon-α 2a (IFN) in the treatment of advanced/metastatic renal cell carcinoma in treatment-naïve patients. METHODS: Studies were retrieved from Embase, Pubmed, Cochrane, and DARE databases using relevant search strategies. Randomized controlled trials, which compared IFN with other pharmacological interventions/best supportive care (BSC), were included according to prespecified inclusion/exclusion criteria. The outcomes of interest were overall survival (OS), progression free survival (PFS), response rate (RR), and adverse events (AEs). Two reviewers independently extracted data from the included studies. Data were analyzed using RevMan (5). RESULTS: Of the 736 studies identified, seven studies met the inclusion criteria. In total, 1147 patients were randomized to IFN, and 1150 were randomized to comparator interventions. Two studies reported comparison with interleukin-2 (IL-2), two with BSC and one each with sorafenib, sunitinib, and temsirolimus. Median OS ranged from 9 to 21.8 months with IFN. Progression-free survival ranged from 1.9 to 5.6 months and overall RR ranged from 4.83% to 12.27% with IFN. Sunitinib had significantly better overall RR (P < 0.001), PFS (P < 0.001), and OS (P < 0.01) compared to IFN. Sorafenib and temsirolimus had better overall RR than IFN (P < 0.01). Results of meta- analysis demonstrate that IFN has better overall RR than BSC (OR: 2.51 [95% CI: 0.87, 7.27], P = 0.089) and similar RR as IL-2 (OR: 1.09 [95% CI: 0.48, 2.45], P = 0.836). The AE profile (gastrointestinal, vascular, infectious, and blood disorders) was similar with IFN and comparators. CONCLUSIONS: Survival benefit with IFN-α was lower than the newer therapeutic agents. Anti-angiogenic agents targeting through multiple receptor kinases, such as sunitinib and sorafenib have significantly improved response rates and survival. These agents would be preferred for treatment naïve patients with advanced/metastatic renal cancer. PCN8 CLINICAL AND ECONOMIC BURDEN OF TOXICITIES ASSOCIATED WITH MONOCLONAL ANTIBODIES FOR METASTATIC COLORECTAL CANCER (MCRC) Burudpakdee C1, Zhao Z2, Trochlil K1, Gao SK3, Munakata J4, Barber B3 1 IMS Consulting, Falls Church, VA, USA; 2Amgen, Newbury Park, CA, USA; 3Amgen Inc., Thousand Oaks, CA, USA; 4IMS Health, Redwood City, CA, USA OBJECTIVES: As overall survival improves with newer therapies for mCRC, treatmentlimiting toxicities and related costs will be important when evaluating treatment decisions. Little is known about toxicity-related cost of currently available monoclonal antibody treatments. This study was designed to identify cetuximab-, bevacizumab-, and panitumumab-related toxicities and estimate direct costs of treating these toxicities. METHODS: A comprehensive literature search was performed to identify English language phase II/III studies of monoclonal antibodies for mCRC. The search utilized PubMed, conference abstracts, treatment guidelines, and product labels. Commonly reported grade 3 and 4 toxicities were identified, and outpatient and inpatient costs were estimated for all toxicities. Outpatient costs were estimated by applying 2010 Medicare reimbursement rates to resource use assumptions (assessed based on indepth clinical interviews). Inpatient costs were estimated using ICD-9 codes and 2007 Medicare payments from the HCUP database; then were converted to 2010 values using the Consumer Price Index for medical care services. RESULTS: Clinically significant toxicities associated with bevacizumab include hypertension, arterial thrombosis, hemorrhage, gastrointestinal (GI) perforation, fistula, and wound healing complication; while treatment-related toxicities associated with cetuximab and panitumumab include skin rash, hypomagnesemia, and infusion reactions, although the incidence of these toxicities differ between the two drugs. Cost of toxicities treated in outpatient setting ranged from $185 (hypertension and skin rash) to $585 (woundhealing complications). Inpatient cost per event for GI perforation is the highest at $32,443, followed by fistula $29,062, arterial thrombosis $20,346, wound healing complication $13,240, hemorrhage $12,956, infusion reaction $10,326, and hypertension $8453, while inpatient cost per event for skin rash and hypomagnesemia is among the lowest at $4424 and $6174, respectively. CONCLUSIONS: Monoclonal antibodies have different toxicity profiles and the costs associated with managing these toxicities vary greatly. PCN9 OBSERVATIONAL STUDY OF PATIENTS WITH NON SMALL CELL LUNG CANCER (NSCLC) TREATED BY ERLOTINIB: CLINICAL PRACTICES AND MAIN OUTCOMES IN FRANCE Vergnenegre A1, Monnet I2, Chouaid C3, Hureaux J4, Mazières J5, Quéré G6, Lombard JN7, Cumin I8, Abdiche S9, Nocent Ejnaini C10, Decroisette C11 1 Hôpital du Cluzeau, Limoges, France; 2CHI Créteil, Créteil, France; 3Hôpital St Antoine, Paris, France; 4CHU Angers, Angers, France; 5CHU Toulouse, Toulouse, France; 6CHU Morvan, Brest, France; 7Cabinet de Pneumologie, Dijon, France; 8CH Bretagne Sud Site Lorient, Lorient, France; 9Hôpital Robert Boulin, Libourne, France; 10CHI Côte Basque, Bayonne, France; 11Centre Hospitalier de la Région d’Annecy, Pringy, France OBJECTIVES: Few data are available about the use of erlotinib in real-life in France for patients with non small cell lung cancer (NSCLC) in a selected population. METHODS: An epidemiological multicenter observational study was built in 35 french centers. The study was retrospective (2006 to 2008) and a cohort was created 13th Euro Abstracts with a follow-up period of 1 year. The main objective was to describe practices, use of erlotinib, response, and adverse events. RESULTS: A total of 533 patients (333 males, 200 females) have been included. The histological types were as follows: 330 (62.5%) adenocarcinoma, 107 (20.2%) squamous cell carcinoma, 60 (11.3%) large cell carcinoma, 36 (3.8%) undifferentiated carcinoma. In terms of practice, 502 patients had a first line chemotherapy (81% a doublet, 11% three drugs, 8.2% one drug). Among them, 61.2% received a second line of treatment (83.4% one drug, 15.7% two drugs and 0.9% three drugs), 17.6% received a third line (91% one drug). Erlotinib was prescribed a first line treatment (n = 30; 5.6%), second line treatment (n = 190; 35.6%), third line (255; 47.8%), fourth line, and more (n = 50; 9.3%) and as a maintenance therapy (n = 9; 1.7%); the stage at treatment initiation were stage I–II (1.2%), IIIA (3.8%), IIIB (5.3%), and IV (89.7%). For the first line, the median duration of erlotinib treatment was 123 days (d) in second line 98 d, in third line 77 d, in maintenance 127 d. Global response rate was 20% with a maximum of 32% in first line and 33% in maintenance. Grade III adverse events occurred in 11.5% of patients and grade IV in 3.4%. CONCLUSIONS: Erlotinib was widely used in France in 2d and 3d line treatment with a good response rate and tolerance. Adenocarcinoma is the main indication. PCN10 EXPLORATIVE ANALYSIS ABOUT THE POTENTIAL OF A LARGE GPS LONGITUDINAL DATABASE ON SEARCHING CAUSAL ASSOCIATIONS AMONG PATHOLOGIES Katz PM, Heiman F BKL Consulting SRL, Milan, Italy OBJECTIVES: The main objective of this study was to analyze different approaches and methods to explore potential causal associations among prevalent pathologies. We have focused on diabetes mellitus (DM) and its well-known association with incident neoplasia. METHODS: For this retrospective cohort study, data were obtained from CSD LPD, a General Practitioner’s longitudinal database. We have evaluated the risk of neoplasia incidence among people with diabetes mellitus compared with those without this pathology, in patients who had no reported history of benign or malign neoplasia at the start of the follow-up on January 2006. For the DM group, patients with at least one diagnosis of DM from January 2005 to December 2005 have been selected, while for the DM-free group, patients without a diagnosis of DM and a date of registration in the GPs office before January 2006 have been selected. Both groups have been followed up for 48 months. RESULTS: During the selection period, 45.121 (3.4%) patients with a diagnosis of DM (females: 22.330, males: 22.791) and 1.290.597 (96, 6%) patients without a diagnosis of DM (females: 690.462, males: 600.135) have been selected. During the follow-up 6.648 and 80.880 incident cases of neoplasms have been documented from the DM and DM free groups respectively. The mean follow-up duration was 43 and 45 months for the DM and the DM-free groups respectively. CONCLUSIONS: The selected cohort has shown to match quite well with general population in terms of gender and age. The estimated prevalence of diabetes also matches with the one of the general population. Statistical analysis has shown an adjusted (for age and sex) hazard ratio of 1,88 (95% Cl 1,83–1,93) suggesting an association between DM and incident neoplasms, evidencing that GPs longitudinal databases could be a valid instrument for evaluating causal association among prevalent pathologies. PCN11 GEFITINIB COMPARED WITH DOUBLET CHEMOTHERAPY FOR FIRST-LINE TREATMENT OF NON-SMALL-CELL LUNG CANCER (NSCLC): A SYSTEMATIC REVIEW AND ADJUSTED INDIRECT COMPARISON Edwards SJ1, Welton N2, Borrill J1 1 AstraZeneca UK Ltd, Luton, Bedfordshire, UK; 2University of Bristol, Bristol, UK OBJECTIVES: Objective response rate (ORR) is an early indicator of successful treatment in patients with NSCLC. This research compared gefitinib with platinum-based doublet chemotherapies for first-line treatment of advanced NSCLC in patients harboring activating epidermal growth factor receptor tyrosine kinase (EGFR-TK) mutations (M+). METHODS: Systematic searching of CENTRAL, EMBASE, and MEDLINE for randomized controlled trials (RCTs) comparing ≥2 doublet chemotherapies (carboplatin or cisplatin in combination with either docetaxel, gemcitabine, paclitaxel, pemetrexed, or vinorelbine) for the first-line treatment of advanced NSCLC was completed in May 2009. a meta-analysis was performed on ORR using data from published RCTs of gefitinib versus paclitaxel/carboplatin in EGFR-TK M+ patients. a mixed treatment comparison (MTC) was carried out with doublet chemotherapies in unselected advanced NSCLC patients using paclitaxel/carboplatin as a baseline. Treatment effect was calculated as an odds ratio (OR) with 95% credible interval (95% CrI). a sensitivity analysis was conducted on the inclusion of the gefitinib trials within the MTC. For this analysis, it was assumed that the efficacy of doublet chemotherapy is consistently affected by EGFR-TK mutation status. RESULTS: Three RCTs were identified for gefitinib, of which two were comparisons with paclitaxel/carboplatin. Meta-analysis of these two trials gave an estimated ORR favoring gefitinib: OR 4.04, 95% confidence interval: 2.73–5.98. Twenty-nine trials were appropriate for inclusion in the MTC, of which 25 reported ORR. The MTC found no significant difference in ORR among other doublet chemotherapies versus paclitaxel/carboplatin, with the exception of pemetrexed/cisplatin, in patients with predominantly non-squamous tumor cell histology, which was associated with a significantly higher ORR (OR 1.64, 95% CrI: 1.15–2.27). In the sensitivity analysis, ORR was significantly higher with 13th Euro Abstracts gefitinib versus all doublet chemotherapies (gefitinib vs. pemetrexed/cisplatin OR 3.05, 95% CrI: 1.58–5.51). CONCLUSIONS: This adjusted indirect comparison suggests that gefitinib may have important ORR advantages over other first-line treatments in EGFR-TK M+ patients. PCN12 EFFICACY OF SECOND LINE TREATMENTS IN PATIENTS WITH METASTATIC HORMONE REFRACTORY PROSTATE CANCER (MHRPC) IS NOT DEMONSTRATED BY PUBLISHED EVIDENCE FROM NONRANDOMIZED TRIALS Freemantle N1, Mason M2, Jasso Mosqueda JG3, Nixon F4, Budhia S4 1 University of Birmingham, Birmingham, UK; 2Cardiff Medical School, Cardiff, UK; 3 Sanofi-Aventis, Massy, France; 4Heron Evidence Development Ltd, Luton, UK OBJECTIVES: Standard first-line treatment for patients with mHRPC is Docetaxel(D)based chemotherapy. Published results from randomized clinical trials of second-line treatments after D failed to provide definitive conclusions about clinical efficacy largely due to paucity of data. This study sought to identify nonrandomized trials of secondline chemotherapy in mHRPC patients pretreated with D and present related survival and clinical benefits. METHODS: Pubmed and Embase were used to perform a systematic literature review (SLR) (2000–2010). Both comparative and noncomparative nonrandomized evidence were extracted from prospective and retrospective studies. Targeted population was patients with mHRPC failing previous D-based regimens. End points included overall-survival (OS), progression-free-survival (PFS), and PSAresponse rate. RESULTS: Among the 825 records screened, 30 studies met the inclusion criteria, two of which were comparative. Of these, 10 addressed rechallenge with D and seven addressed mitoxantrone (MTX); the remaining 18 studies considered various other regimens. Treatment was with either single-agent or combination regimens. Ninety-three percent of studies included <50 patients. PFS and PSA response definitions varied between trials. For studies evaluating rechallenge with D, the median OS and PFS varied from 41 to 76 weeks and from 15 to 39 weeks respectively. For MTX, the median OS and PFS varied from 39 to 48 weeks and 13 to 16 weeks, respectively. For other chemotherapy regimens, the median OS and PFS varied from 51 to 104 weeks and 9 to 17 weeks, respectively. PSA response rates varied from 24% to 70% to D rechallenge, from 4% to 33% to MTX-based regimens and from 0% to 60% to other regimens. CONCLUSIONS: The SLR showed a lack of available nonrandomized evidence, and among the selected studies, evidence was not strong enough due to small sample sizes, noncomparative nature and variable PFS and PSA response definitions. This literature review demonstrates that it is difficult to infer the clinical efficacy of mHRPC 2nd line chemotherapy. PCN13 EFFECT OF ZOLEDRONIC ACID AND PAMIDRONATE ON SKELETALRELATED EVENTS AND MORTALITY IN WOMEN WITH BONE METASTASES FROM BREAST CANCER IN A MANAGED CARE PLAN: A RETROSPECTIVE DATABASE ANALYSIS Henk HJ1, Kaura S2 1 i3 Innovus, Eden Prairie, MN, USA; 2Novartis Pharmaceuticals Corporation, Florham Park, NJ, USA OBJECTIVES: Patients with breast cancer (BC) and bone metastases are at risk for skeletal-related events (SREs) that are associated with significant morbidity, mortality, and reduced quality of life. The intravenous bisphosphonates (IVBPs) zoledronic acid (ZOL) and pamidronate (PAM) are approved for treating patients with bone metastases from BC. We compared incidence of SREs and mortality in women with BC who received ZOL or PAM, and assessed the long-term benefit of ZOL in a real-world setting. METHODS: A claims-based analysis of commercial and Medicare Advantage data from >45 US managed care plans was used to evaluate SRE rates and mortality in patients treated with ZOL or PAM. Inclusion criteria were age >18 years, BC with bone metastasis diagnosis between 01/01/01 and 12/31/06, continuous enrollment in the health plan, no evidence of bone metastasis or IVBP for 6 months before an index date of first receipt of ZOL or PAM. Patients were followed until disenrollment (including mortality) or study completion (12/31/07). Persistency was defined as the absence of a >45-day gap between treatments. SREs were defined as evidence of pathologic fracture, spinal cord compression, and radiotherapy or surgery to bone. RESULTS: Among 8757 patients (mean age, 58.1 ± 12.4 years) approximately 30% received ZOL, 15% received PAM, and 55% received no IVBP. Longer persistency with ZOL was associated with lower risk of fracture and of all SREs versus shorter persistency (trend test, P = 0.0026 and P = 0.0216, respectively). ZOL-treated patients had a moderately lower SRE incidence (36.2 vs. 40.0 per 100 person-years; P = 0.0707) and significantly fewer deaths (6.2 vs. 8.9 per 100 person-years; P = 0.0130) versus PAM-treated patients. CONCLUSIONS: In a real-world assessment of women with bone metastases from BC, ZOL reduced SRE incidence and significantly improved survival versus PAM. Longer ZOL persistency was associated with lower SRE risk, reinforcing the importance of regular monthly ZOL dosing. A253 PCN14 CLINICAL CONSEQUENCES OF PRIMARY PROPHYLAXIS WITH PEGFILGRASTIM VERSUS FILGRASTIM FOR THE PREVENTION OF FEBRILE NEUTROPENIA IN NON-HODGKIN LYMPHOMA AND STAGE II BREAST CANCER PATIENTS IN GERMANY Ozer-Deniz S1, Taylor DC1, Hill G1, Skornicki M1, Danel A2, Kunz E3 1 i3 Innovus, Medford, MA, USA; 2Amgen Inc., Zug, Switzerland; 3Amgen (Europe) GmbH, München, Germany OBJECTIVES: To assess the clinical consequences of primary prophylaxis (PP) with pegfilgrastim versus 6- or 11-day filgrastim (F6, F11) in the prevention of febrile neutropenia (FN) in non-Hodgkin lymphoma (NHL) patients receiving CHOP-14 chemotherapy and in breast cancer (BC) patients receiving TAC chemotherapy in Germany. METHODS: A lifetime Markov model was developed, consisting of two phases: 1) on-chemotherapy phase (OCP), where model cycle length equals chemotherapy cycle length (CHOP-14:14 days, TAC: 21 days), and 2) post-chemotherapy phase (PCP) with annual model cycles. PP is defined as prophylaxis initiated with the first chemotherapy cycle. Cycle 1 FN risk with no prophylaxis (NP) was estimated to be 21% for NHL CHOP-14 and 14% for BC TAC. All cycle relative risk of FN using PP with pegfilgrastim versus no PP, F6, and F11 was 0.25, 0.87, and 0.61. FN case fatality was estimated (NHL: 8.9%; BC: 3.6%). In PCP, all-cause mortality was estimated from German life-tables; NHL and BC mortality from US data; patients experiencing FN were assumed to have higher mortality due to reduced chemotherapy dose intensity. All inputs were estimated from clinical trials and published literature. The model estimates life-years, number of FNs, and number needed to treat (NNT) to prevent an FN. RESULTS: NNT to prevent an FN were 1.3, 6.2, 2.2 in NHL; 2.3, 11.1, 4.0 in BC for Pegfilgrastim, F6, and F11 compare to NP. Overall, FN episodes per patient were 0.15, 0.76, and 0.47 in NHL; 0.09, 0.43, and 0.27 in BC. Per-patient life-months gained using PP with Pegfilgrastim were 3.4 and 1.8 versus F6 and F11, respectively in NHL, and 2.2 and 1.2 in BC. CONCLUSIONS: Primary prophylaxis with pegfilgrastim results in a lower NNT, fewer FN events, and more life-years than with 6-day filgrastim or 11-day filgrastim in both NHL and BC. PCN15 SYSTEMATIC REVIEW OF LAPATINIB PLUS LETROZOLE WITH OTHER FIRST LINE TREATMENTS FOR HORMONE POSITIVE (HR+) HER2+ ADVANCED OR METASTATIC BREAST CANCER (MBC) Riemsma R1, Forbes CA1, Amonkar M2, Lykopoulos K3, Diaz JR4, Kleijnen J1, Rea DW5 1 Kleijnen Systematic Reviews Ltd., York, North Yorkshire, UK; 2GlaxoSmithKline, Philadelphia, PA, USA; 3GlaxoSmithKline, London, UK; 4GlaxoSmithKline, Uxbridge, Middlesex, UK; 5 Institute for Cancer Studies, Birmingham, UK OBJECTIVES: To undertake a systematic review of lapatinib plus letrozole (LAP + LET) with other first-line treatments for HR+ HER2+ advanced or MBC in postmenopausal women who have not received prior therapy for advanced or metastatic disease. METHODS: Seven databases were searched through January 2009 for randomized controlled trials. Relevant interventions were lapatinib (alone/in combination), aromatase inhibitors (letrozole (LET), anastrozole (ANA), exemestane (EXE)), tamoxifen (TAM), and trastuzumab (TRAS) (alone/in combination). Outcomes included overall survival (OS), progression-free survival (PFS), time to progression (TTP), and objective response rate (ORR). From the available evidence, it was possible to directly compare LAP + LET with LET. Using a network meta-analysis, LAP + LET could be indirectly compared with the four other interventions. RESULTS: Eighteen studies (62 papers) met the inclusion criteria. LAP + LET was significantly superior to LET based on a direct head-to-head study in terms of PFS/TTP and ORR. In the indirect comparison with LAP + LET, TAM (hazard ratio [HR] = 0.45 [95% CI: 0.32, 0.65]), EXE (HR = 0.52 [0.34, 0.79]), and ANA (HR = 0.53 [0.36, 0.80]) scored significantly worse in terms of PFS/TTP and ORR (TAM: odds ratio [OR] = 0.25 [0.12, 0.53], ANA: OR = 0.27 [0.12, 0.58], EXE: OR = 0.47 [0.20, 1.09]). LAP + LET also seemed better, although not significantly, in terms of OS versus TAM: HR = 0.74 (0.49, 1.12), EXE: HR = 0.65 (0.39, 1.11), and ANA: HR = 0.71 (0.45, 1.14). LAP + LET when indirectly compared with TRAS + ANA, seemed to be better in terms of OS (HR = 0.85 [0.47, 1.54]), PFS/TTP (HR = 0.89 [0.54, 1.47]) and ORR (OR = 0.92 [0.24, 3.48]), although, none of these results were significant. CONCLUSIONS: Using indirect methods, LAP + LET appeared to be the best treatment in this HR+ HER2+ patient population. However, the results are based on a network analysis for which the basic assumptions of homogeneity, similarity, and consistency were not fulfilled. Therefore, despite the fact that these are the best available data, the results need to be interpreted with caution. PCN16 MIXED TREATMENT COMPARISON OF BEVACIZUMAB-BASED THERAPIES RELATIVE TO DOUBLET-CHEMOTHERAPY COMBINATIONS TO ESTIMATE THE RELATIVE EFFICACY IN PROGRESSION-FREE SURVIVAL FOR TREATMENT OF FIRST-LINE ADVANCED OR METASTATIC NON-SMALL CELL LUNG CANCER (NSCLC) Yi Y1, Chouaid C2, Vergnenegre A3, Bischoff HG4, Bergman G5, Walzer S6, Philips Z1 1 Mapi Values, Bollington, UK; 2Hôpital Saint-Antoine, Paris, France; 3SIME, Limoges, France, 4 Thoraxklinik Heidelberg GmbH, Hedelberg, Germany; 5Mapi Values, Houten, The Netherlands; 6F. Hoffmann-La Roche Pharmaceuticals AG, Basel, Switzerland OBJECTIVES: To compare the efficacy in progression-free survival (PFS) of bevacizumab plus cisplatin and gemcitabine (BCG) and bevacizumab plus carboplatin and paclitaxel (BCP), relative to doublet-chemotherapy combinations for the treatment of A254 first-line advanced or metastatic NSCLC. METHODS: A systematic literature review identified randomized controlled trials (RCTs) reporting PFS for bevacizumab-based and doublet-chemotherapy combinations. Studies were evaluated for comparability of design and patient population. Reported PFS hazard ratios (HR) were analyzed simultaneously with a Bayesian mixed treatment comparison. The base-case analysis compared BCG and BCP with grouped platinum-based doublets (PLD) and grouped nonplatinum-based doublets (NPLD). Scenario analyses explored BCP and BCG versus different combinations of doublet treatments. RESULTS: Eight identified RCTs, considered comparable in design and patient characteristics, allowed for a comparison between bevacizumab-based therapies and grouped doublet-chemotherapy combinations. The expected PFS HRs relative to PLD, for BCP, BCG, and NPLD were 0.66 (95% interval: 0.57; 0.77), 0.80 (0.71; 0.89), and 1.05 (0.92; 1.19), respectively. BCP and BCG were ranked as the top two most efficacious treatments in terms of PFS across all included regimens. Scenario analyses confirmed the top ranking for BCP and BCG. When BCP and BCG were compared to individual doublet chemotherapies, BCP showed the greatest benefit (HR of 0.63 [0.45; 0.88]), followed by BCG 7.5 mg/kg (0.75 [0.64; 0.87]) and BCG 15 mg/kg (0.85 [0.73; 0.99]). Further analyses confirmed the robustness of the findings. CONCLUSIONS: Compared to all available doubletchemotherapy combinations, bevacizumab-based therapy is expected to be more efficacious in terms of PFS, and could therefore be considered as the first treatment option in advanced or metastatic NSCLC. PCN17 NO CONCLUSIVE EVIDENCE FROM RANDOMIZED CONTROLLED TRIALS (RCTS) FOR IMPROVED SURVIVAL WITH SECOND-LINE TREATMENT OPTIONS, IN PATIENTS WITH METASTATIC HORMONEREFRACTORY PROSTATE CANCER (MHRPC) PREVIOUSLY TREATED WITH DOCETAXEL Mason M1, Freemantle N2, Parnaby A3, Högberg D4 1 Cardiff Medical School, Cardiff, UK; 2University of Birmingham, Birmingham, UK; 3 Sanofi-Aventis, Vitry-sur-Seine, France; 4i3 Innovus, Stockholm, Sweden OBJECTIVES: A docetaxel (D)-based regimen is recommended first-line treatment for mHRPC patients. Currently, there are no recommended second-line treatments for D pretreated patients. This study sought to identify phase II and III RCTs of second-line treatments for mHRPC in D pretreated patients to provide information regarding survival. METHODS: PubMed and Embase were used to perform a systematic literature review (2000–2009). Primary and secondary efficacy end points were extracted. Safety outcomes were reviewed according to grade. RESULTS: Among 52 records screened, three trials were included and 47 were excluded (35 not clinical trials; four not second line to D; eight not comparative or randomized). Primary end points included overall survival (OS), progression-free-survival (PFS), PSA response rate, and objective tumor response (OTR). a phase III study comparing satraplatin plus prednisone (SP) to prednisone (P) alone (n = 950, 51% post-D) was identified. Two phase II trials compared ixabepilone (ixa) with mitoxantrone plus prednisone (MP) (n = 82), and custirsen in combination with prednisone plus D (DPC) versus curtisen plus MP (MPC) (n = 42). SP demonstrated significant improvements compared to P in PSA response (25% vs. 12%, P < 0.001), OTR (7% vs. 1%, P < 0.002), and pain response (24% vs. 14%, P < 0.005). Median PFS (11 weeks vs. 9.7 weeks), but median OS (66.1 weeks vs. 62.9 weeks) were similar. In the second trial (Ixa vs. MP), there was no significant improvement in either PSA response (17% vs. 20%) or OS. In the third trial, PSA response was better for DPC than MPC (40% vs. 27%); no OS data reported. Grade 3 or 4 neutropenia occurred in 54% and 63% with Ixa and MP respectively. CONCLUSIONS: This review found a limited number of published phase II and III RCTs second-line treatments for mHRPC in D pretreated patients. None demonstrated a survival benefit. Results should be interpreted with caution in terms of clinical benefits. PCN18 RETROSPECTIVE DATABASE ANALYSIS OF THE EFFECT OF ZOLEDRONIC ACID ON SKELETAL-RELATED EVENTS IN MEN WITH PROSTATE CANCER AND BONE METASTASES Henk HJ1, Kaura S2 1 i3 Innovus, Eden Prairie, MN, USA; 2Novartis Pharmaceuticals Corporation, Florham Park, NJ, USA OBJECTIVES: Patients with bone metastases from prostate cancer (PC) are at risk for skeletal-related events (SREs) including pathologic fracture, spinal cord compression, the need for radiotherapy or surgery to bone, and hypercalcemia of malignancy. Zoledronic acid (ZOL), an intravenous bisphosphonate (IVBP), has proven efficacy for reducing the incidence and delaying the onset of SREs in multiple tumor types. This retrospective study was designed to assess the fracture risk in patients receiving ZOL or no treatment, and to examine the benefit of long-term ZOL use in a real-world setting among men with PC and bone metastases. METHODS: Commercial and Medicare Advantage databases were used to evaluate fracture rates and medication persistency. Patients included in this analysis were ≥18 years old, had PC and bone metastasis diagnosed between January 1, 2001 and December 31, 2006, were continuously enrolled in the health plan, and had no evidence of bone metastasis or IVBP for 6 months before first infusion of ZOL. Patients were followed until discontinuation (including mortality) or study completion. Fractures were categorized as vertebral, hip, or other nonvertebral fractures. Persistency was defined as the absence of a >45-day gap between ZOL treatments. RESULTS: Among 4976 men (mean age, 70.9 ± 9.7 years), approximately 26% received ZOL and 74% received no IVBP. Regardless 13th Euro Abstracts of fracture site, ZOL reduced the fracture rate compared with no IVBP (5.9 vs. 8.5 per 100 person-years; P = 0.0003). Longer persistency with ZOL was associated with a reduced fracture rate (trend test, P = 0.0179). The mortality rate was also significantly lower in ZOL patients versus patients receiving no IVBP (6.2 vs. 9.4 per 100 person-years; P = 0.0018). CONCLUSIONS: In men with bone metastases from PC, ZOL was associated with a significantly lower fracture rate and mortality compared with no IVBP. Furthermore, longer persistency with ZOL was associated with a lower fracture rate. PCN19 SYSTEMATIC REVIEW OF ENDOSCOPIC SUBMUCOSAL DISSECTION VERSUS ENDOSCOPIC MUCOSAL RESECTION FOR EARLY GASTRIC CANCER Park YM1, Cho E2, Kang HY2, Kim JM1, Kim SY3 1 Health Insurance Review & Assessment Agency, Seoul, South Korea; 2Yonsei University, Seoul, South Korea; 3Korea University, Seoul, South Korea OBJECTIVES: Endoscopic submucosal dissection (ESD) allows en bloc resection of the entire lesion which permits higher curative resection rate, lower local recurrence, and consequently, increases quality of life by minimizing the resection size compared to endoscopic mucosal resection (EMR). While ESD has been implemented in most university hospitals in Korea currently, potential complications of ESD like hemorrhage and perforation waver over the therapeutic decision on the ESD for early gastric cancer patients as well as the reimbursement decision-making. The study aims to address both effectiveness and safety outcomes of ESD versus EMR in early gastric cancer by systematic review. METHODS: MEDLINE, EMBASE, Cochrane Central Register of Controlled Trials (CENTRAL), and Koreamed were searched using primary key words: “stomach neoplasm” and “endoscopic submucosal dissection” and “endoscopic mucosal resection.” To assess the quality of selected studies, the methodological approach of Scottish Intercollegiate Guidelines Network were used. Five effectiveness-relevant and three safety-relevant outcome measures were extracted. Bibliography management and meta-analysis for each outcome were conducted using Review Manager 5.0. RESULTS: Three nonconcurrent cohort studies and nine retrospective cohort studies were identified. Meta-analyses showed significantly greater effectiveness of ESD as compared to EMR for en bloc resection (OR = 8.43, 95% CI: 5.20–13.67), complete resection (OR = 8.54, 95% CI: 4.44–16.45), curative resection (OR = 2.56, 95% CI: 1.68–3.91), local recurrence (RR = 0.13, 95% CI: 0.04–0.40), and all-cause mortality (RR = 0.65, 95% CI: 0.08–5.38). While intraoperative bleeding (RR = 2.16, 95% CI: 1.14–4.09) and perforation risk (RR = 3.58, 95% CI: 1.95–6.55) were significantly greater for ESD, overall bleeding risk (RR = 1.22, 95% CI: 0.76–1.98) and longer resection time (RR = 1.55, 95% CI: 0.74–2.37) were not significantly different between ESD and EMR. CONCLUSIONS: Considering bleeding risk was not significantly different between ESD and EMR, and the perforation risk usually does not lead to life-threatening disease, the effectiveness benefit of ESD can overweigh the overall harm compared to EMR on condition that ESD was performed by surgeons with certain experiences. PCN20 MAINTENANCE ERLOTINIB VERSUS PEMETREXED FOR THE TREATMENT OF NON-SMALL CELL LUNG CANCER: INDIRECT COMPARISON APPLYING REAL-LIFE OUTCOMES Casciano R1, Bischoff H2, Nuijten M3, Malangone E1, Ray J4 1 Analytica International Inc., New York, NY, USA; 2Thoraxklinik Heidelberg GmbH, Heidelberg, Germany; 3Ars Accessus Medica, Rotterdam, The Netherlands; 4F. Hoffmann-La Roche Ltd., Basel, Switzerland OBJECTIVES: Recent clinical trials have established superior efficacy of both erlotinib and pemetrexed as first-line maintenance therapies for metastatic non-small cell lung cancer (mNSCLC) over placebo. Results indicated that erlotinib improved survival for all histology types and pemetrexed improved survival in nonsquamous patients. To date, there have been no head-to-head trials directly comparing the two agents. An indirect comparison analysis was performed to examine the relative efficacy of these two treatment regimens as maintenance treatment options following platinum-based first-line therapy. METHODS: An adjusted-matched indirect analysis approach was used to compare overall survival (OS) estimates in mNSCLC patients treated with erlotinib from SATURN versus pemetrexed patients from JMEN. Patient distributions of key characteristics between the two studies were unbalanced; JMEN trial patients had a better prognosis at baseline. Patient distributions observed in the pemetrexed study for race and smoking status were used to match erlotinib-treated patients using patient-level data from the SATURN trial, employing an adjusted matching approach to make the populations more comparable. a distribution of survival outcomes was derived from each of 1000 repeated random matching samples of the SATURN data, with 95% confidence intervals (CI) around the mean of the aggregate of all observed median OS survival estimates generated by ordering the outcome measures and identifying the 2.5 percentile observations. To indirectly compare treatments, the median ratio (MR) for OS was calculated to approximate the hazard ratio. RESULTS: The estimated median OS after adjusted-matching was 13.9 months (95% CI 10.9–16.8) for erlotinib, compared with the published median OS reported for pemetrexed of 13.4 months (95% CI: 11.9–15.9). Erlotinib patients had similar median OS compared to pemetrexed patients with an MR of 0.96 (0.95, 1.09). CONCLUSIONS: Erlotinib and pemetrexed are similarly efficacious in first-line maintenance NSCLC differing in other parameters than efficacy such as tolerability, administration, and patient convenience. 13th Euro Abstracts PCN21 FIRST-LINE BEVACIZUMAB-BASED THERAPY VERSUS PEMETREXED + CISPLATIN FOR THE TREATMENT OF ADVANCED ADENOCARCINOMA NONSQUAMOUS NON-SMALL CELL LUNG CANCER: INDIRECT COMPARISON APPLYING REAL-LIFE OUTCOMES Berenson K1, Chouaid C2, Vergnenegre A3, Sherman S1, Walzer S4 1 Analytica International Inc., New York, NY, USA; 2Hôpital St Antoine, Paris, France; 3SIME, Limoges, France; 4F. Hoffmann-La Roche Pharmaceuticals, Basel, Switzerland OBJECTIVES: In the absence of head-to-head clinical trial data, an indirect comparison of bevacizumab (BEV) versus pemetrexed (PMX) was conducted to compare survival outcomes among adenocarcinoma nonsquamous metastatic Non-Small Cell Lung Cancer (mNSCLC) patients. METHODS: An adjusted matched indirect analysis was conducted to estimate overall survival (OS) in adenocarcinoma mNSCLC patients treated with BEV + cisplatin doublet therapy using patient-level data from SAiL (ECCO/ESMO 2009). These estimates were indirectly compared to previously published survival outcomes for PMX + cisplatin-treated patients (Oncologist 2009;14:253–263) by calculating the median ratio (MR) for OS. a subset of the SAiL population was selected to more closely approximate the PMX population by excluding patients who did not have cisplatin doublet as their baseline treatment, those with a baseline Eastern Cooperative Oncology Group (ECOG) performance status (PS) of > = 2, and those with non-adenocarcinoma histology. This sample of BEV patients was matched to the adenocarcinoma subgroup from the PMX trial on stage of disease and ECOG PS. One thousand repeated random matched samples of the SAiL data were produced to generate a distribution of survival outcomes and infer a 95% confidence interval (CI) around the mean of all sampled median survival estimates. RESULTS: After adjusted matching, the estimated median OS benefit for BEV patients was 15.6 months (95% CI: 15.0, 16.5) compared to the published median OS of 12.6 months (95% CI: 10.7, 13.6) for PMX patients. BEV patients had longer median OS with an MR of 0.81 (95% CI: 0.71, 0.82). CONCLUSIONS: Results from this indirect comparison show that BEV-based therapy provides superior overall survival outcomes when compared to PMX in adenocarcinoma mNSCLC patients. PCN22 EFFICACY AND CARDIAC SAFETY OF TRASTUZUMAB (T) IN THE ADJUVANT TREATMENT OF HER2-POSITIVE EARLY-STAGE BREAST CANCER: A SYSTEMATIC REVIEW (SR) AND META-ANALYSIS (MA) Botrel TEA, Clark O, Clark LGO, Paladini L, Faleiros E, Pegoretti B MedInsight-Evidências, Campinas, Brazil OBJECTIVES: Trastuzumab (T) is now part of the standard adjuvant treatment for HER2 positive, breast cancer patients. However, the results of the studies are not uniform and there are still doubts about the ideal indication and schedule for its use. Our objective was to perform a systematic review (SR) and meta-analysis (MA) of all randomized controlled trials (RCT) comparing efficacy of Chemotherapy (CHEM) plus T versus CHEM plus observation in the adjuvant treatment of HER2-positive early-stage breast cancer. METHODS: Several databases were searched, including MEDLINE, EMBASE, LILACS, and CENTRAL. The primary end points were progression-free survival (PFS) and overall survival (OS). a subgroup analysis was performed to evaluate the influence of the use of T concurrent or sequential to CHEM. RESULTS: Overall, 730 references were identified and screened. The final analysis included six trials comprising 13,940 patients. The PFS was higher in the group of patients treated with CHEM plus T (fixed effect: HR = 0.61, CI 95% = 0.56 to 0.66; P < 0.00001) with significant heterogeneity (χ2 = 13,33 df = 5 (P = 0.02); I2 = 62%). This result remained favorable to the use of T after the random-effects model analysis was performed (HR = 0.63, CI 95% = 0.54 to 0.73; P < 0.00001). OS was better for patients who received T (fixed effect: HR = 0.71, CI 95% = 0.62 to 0.81; P < 0.00001 and random-effects: HR = 0.71, CI 95% = 0.60 to 0.84; P < 0.0001) with moderatelevel heterogeneity (χ2 = 7,12 df = 5 (P = 0.21); I2 = 30%). There was a significant interaction between the concurrent and sequential use of T and CHEM, suggesting that the concurrent use may be more effective. CONCLUSIONS: Trastuzumab increased progression-free survival and overall survival of patients especially when administered concomitantly to chemotherapy. PCN23 SIMULATION AND COMPARISON OF PROGRESSION-FREE-SURVIVAL OUTCOMES OF SEQUENTIAL TARGETED THERAPY IN METASTATIC RENAL-CELL CARCINOMA Mickisch GH1, Schwander B2, Cassinello JG3, Carles J4, Walzer S5, Nuijten M6 1 Center of Operative Urology Bremen, Bremen, Germany; 2AiM GmbH—Assessment in Medicine, Research and Consulting, Lörrach, Germany; 3University Hospital Guadalajara, Guadalajara, Spain; 4Vall d’Hebron University Hospital, Barcelona, Spain; 5F. Hoffmann-La Roche Pharmaceuticals AG, Basel, Switzerland; 6Ars Accessus Medica, Jisp, The Netherlands OBJECTIVES: The novel targeted agents bevacizumab (BEV), sorafenib (SOR), sunitinib (SUN), everolimus (EVR), and pazopanib (PAZ) have each demonstrated activity in metastatic renal-cell carcinoma patients. One of the remaining key questions is which therapy sequence provides the most valuable outcome in terms of progressionfree-survival (PFS). METHODS: A Markov disease model was developed using pivotal trial evidence. All patients were assumed to be treatment-naïve, with a good or intermediate prognosis, and to enter the model in “PFS,” receiving first-line treatment with either BEV + interferon-alpha-2a (IFN), SUN, PAZ, or IFN alone, taking into account current EMA licenses (e.g., BEV + IFN indicated only for first-line therapy). After initial treatment, patients were assumed to either die, progress to a subsequent line of A255 therapy, or remain in PFS. Hence, in case of p. RESULTS: The most valuable therapy sequence in terms of total PFS was BEV + IFN → PAZ → SUN → SOR → EVR resulting in a mean PFS time of 33.2 months (95% confidence interval [CI]: 31.3– 35.2). The sequence BEV + IFN → PAZ → SOR → SUN → EVR obtained comparable results (mean PFS 33.2; 95% CI: 31.3–35.1). The first-line PAZ sequences PAZ → SUN → SOR → EVR (mean PFS 28.6; 95% CI 26.6–31.2) and PAZ → SOR → SUN → EVR (mean PFS 28.6; 95%CI 26.6–31.2) were the second-best alternatives, followed by the most valuable IFN first-line sequence (IFN → PAZ → SUN → SOR → EVR; mean PFS 28.5; 95% CI 26.8–30.4) and the most valuable SUN first-line sequence (SUN → PAZ → SOR → EVR; mean PFS 26.7; 95% CI 24.4–29.1). The incremental PFS difference between the best therapy sequences (first-line BEV + IFN; mean PFS 33.2 months) and the second-best therapy sequences (PAZ first-line; mean PFS 28.6 months) of 4.6 months reached statistical significance (P < 0.004 for each possible comparison). Additional overall survival simulations have confirmed these findings. CONCLUSIONS: Modeling simulation indicates that patients’ PFS outcomes could be improved significantly, if therapy started with first-line BEV + IFN compared to other first-line agents (PAZ, SUN, or IFN alone). PCN24 EFFICACY OF ADJUVANT CHEMOTHERAPY WITH GEMCITABINE (GEM) COMPARED TO SURGERY-ONLY IN PATIENTS WITH RESECTED PANCREATIC CANCER: SYSTEMATIC REVIEW (SR) AND METAANALYSIS (MA) Botrel TEA, Clark O, Clark LGO, Paladini L, Faleiros E, Pegoretti B MedInsight-Evidências, Campinas, Brazil OBJECTIVES: We aimed to perform a systematic review (SR) with meta-analysis (MA) of all randomized controlled trials (RCT) comparing the efficacy of adjuvant chemotherapy with gemcitabine (GEM) versus observation in patients with resected pancreatic cancer. METHODS: Several databases were searched, including MEDLINE, EMBASE, LILACS, and CENTRAL. The primary end points were progression-free survival (PFS) and overall survival (OS). The data extracted from the studies were combined by using hazard ratio (HR) with their corresponding confidence intervals of 95% (CI 95%). RESULTS: Overall, 233 references were identified and screened. The final analysis included two trials comprising 472 patients evaluated. The proportion of patients that underwent surgery with curative intent (R0 or R1 resection) was similar between the studies as well as their ages and nodal status. The progression-free survival was higher in the group of patients who were treated with adjuvant chemotherapy including GEM (fixed effect: HR = 0.59, CI 95% = 0.50 to 0.70; P < 0.00001) and no heterogeneity was found (χ2 = 0.01, df = 1 (P = 0.94); I2 = 0%). Overall survival was also higher in patients treated with GEM (fixed effect: HR = 0.81, CI 95% = 0.67 to 0.98; P = 0.03) yet again no heterogeneity was detected (χ2 = 0.07, df = 1 [P = 0.79]; I2 = 0%). CONCLUSIONS: Adjuvant chemotherapy with gemcitabine increased progression-free survival and overall survival of patients with resected pancreatic cancer. PCN25 DID THE UPTAKE OF NEW TREATMENT OPTIONS CHANGE THE TREATMENT STRATEGY IN PATIENTS WITH COLORECTAL CANCER AND PRIMARY NONRESECTABLE METASTASES? THE RESULTS OF LARGE POPULATION-BASED SURVEY IN GERMANY 2006–2007 Kellermann L1, Arnold D2 1 Oncology InformationService, Freiburg, Germany; 2Hematology & Oncology, Martin Luther University Halle-Wittenberg, Halle/Saale, Germany OBJECTIVES: The survey was initiated to gain insights into the changes of treatment patterns in treatment of metastatic colorectal cancer and the implementation of the results of clinical trials in daily practice. METHODS: A representative sample of centers (82) was selected with regard to the distribution of treated prevalence in colorectal cancer in institutions (university hospitals, community hospitals, officebased oncologists) and regional population density. The physicians reported all pts. with a treatment decision in colorectal cancer in the respective reporting period May 2006 to April 2007. The database contains 3254 pts. with a retrospective record of their entire treatment history. The treatment patterns were analyzed in the whole patient group and in subgroups according to resectability of metastases, the treatment objectives (especially secondary resectability of metastases), used systemic treatment regimen, age, concomitant diseases, and performing institution type. The statistics were performed in SPSS by bivariate analyses with two-sided chi-square test. In the next step, the decisive parameters for treatment choice were defined by logistic regression in multivariate analysis. RESULTS: The clinical trial data were taken up very soon in clinical reality. The correlation of drug efficacy and resectability of metastases was transferred into the disease management of colorectal cancer. The patient share with treatment objective “secondary resection of metastases” increased significantly (18% 2004 vs. 27% 2006–2007, P = 0.000%). In this subgroup, the patient share treated with targeted therapy was significantly higher than in patients with other treatment objectives (34% vs. 19%, P = 0.000%). CONCLUSIONS: The method used for creation of the database and for the statistic analyses has been proven as appropriate for the objectives of this survey. The resectability of metastases is recognized as an important treatment objective. Therefore, targeted therapy was implemented more frequently in treatment regiments for patients deemed secondary resectable, compared to other treatment aims. A256 PCN26 ESTIMATION OF THE EPIDEMIOLOGICAL EFFECT OF RITUXIMAB FOR THE TREATMENT OF AGGRESSIVE NON-HODGKIN LYMPHOMA Barreto WG1, Oliveira JSR1, Borges LG2, Santos E2 1 Hospital Santa Marcelina, São Paulo, SP, Brazil; 2Roche Brazil, São Paulo, SP, Brazil BACKGROUND: Non-Hodgkin lymphoma (NHL) are a heterogeneous group of malignancies of the lymphoid system and represents approximately 80% of all lymphomas. Mortality associated with NHL in Brazil has increased in the last few years, from 2772 cases in 2000 to 3194 in 2004. Aggressive NHL accounts for approximately 60% of all cases and cure can be achieved in more than 70% of cases with rates ranging 50% to 90% when under immunochemotherapy depending on the clinical stage and international prognostic index (IPI). Currently, only CHOP therapy is available to patients presenting with aggressive NHL in Brazilian Public Healthcare System. OBJECTIVES: To evaluate the epidemiological impact of adding rituximabe (R) to CHOP first-line therapy (cyclophosphamide, doxorubicin, vincristine, and prednisone) on the annual number of patients experiencing recurrence and deaths between 2010 and 2020 in Brazil. METHODS: Annual aggressive NHL incidence for 2010 to 2020 was estimated by applying published age-specific NHL incidence to Brazilian population projections for the mentioned period. Projections were made for aggressive NHL and population above 20-year-olds only. Therapy effect for R-CHOP and CHOP was derived from Gao (2009) and 2-year overall survival and disease control data were adjusted for a 1-year rate. Considering the subgroups analysis, annual overall survival rate were 54.5% and 45.7% and disease control rate were 43.7% and 31.1% for R-CHOP and CHOP, respectively. RESULTS: Between 2010 and 2020, it is estimated that 8886 more patients would be saved if R-CHOP was given (54,917 vs. 46,031 for CHOP). Regarding disease progression, R-CHOP has the potential to control the disease in 12,703 more patients than CHOP (44,048 vs. 31,345). CONCLUSIONS: Findings suggest that R-CHOP has the potential to save about 16% more lives annually when compared to CHOP in the management of aggressive NHL, particularly in Brazilian patients who have late diagnosis which implies a worse prognosis. PCN27 ESTIMATION OF THE EPIDEMIOLOGICAL EFFECT OF TRASTUZUMAB OVER THE 200–2015 PERIOD IN PORTUGAL Calado F1, Monteiro I1, McNiven PM2, Weisgerber-Kriegl UM3 1 Roche Farmacêutica Química, Amadora, Portugal; 2Strategyst Consulting, Austin, TX, USA; 3 F. Hoffmann-La Roche Ltd., Basel, Switzerland OBJECTIVES: Trastuzumab was approved for the treatment of HER2-positive early breast cancer (EBC) in 2006 in several European countries, Portugal included. The model assesses the impact of trastuzumab treatment in EBC on the annual number of patients entering the metastatic setting between 2005 and 2015 in Portugal. METHODS: Annual EBC incidence for 2005 to 2015 was projected by applying stage-specific proportions for stages I–III to female breast cancer incidence rates estimated for 2008 (GLOBOCAN 2010), using a conjoint estimate based on English cancer registries, where overall breast cancer survival is comparable to that seen in Portugal. Age-specific rates were applied to UN population projections for 2000 to 2025. The annual number of patients with HER2-positive metastatic breast cancer (MBC) includes de novo and recurrent MBC patients. The baseline 10-year recurrence rate was estimated as 37%, based on 4-year follow-up in the control arm of a combined trial analysis in patients with HER2-positive breast cancer and the long-term timing of recurrence in all patients with breast cancer. To model recurrence in trastuzumab-treated EBC patients, the hazard ratio (HR) from the HERA trial (0.49; 95% CI: 0.38, 0.63) was applied. RESULTS: In 2004, prior to approval of trastuzumab in EBC, the pool of de novo and relapsed MBC patients was estimated to be 650. Between 2005 and 2015, the model predicts that the use of trastuzumab will result in an average annual decline in recurrence of MBC of 5.2% (95% CI: 3.0, 6.7). Cumulative recurrence prevented by trastuzumab treatment between 2005 and 2014 (projected through full 10-year window for recurrence) is projected to be 884 (95% CI: 537; 1075). CONCLUSIONS: Trastuzumab treatment of HER2-positive EBC over a 10-year period is expected to prevent nearly 900 women from developing metastases in Portugal, which may result in a similar number of breast cancer deaths being avoided. PCN28 UTILIZATION OF ANTIEMETIC PROPHYLAXIS ALONGSIDE CHEMOTHERAPY IN FOUR EUROPEAN COUNTRIES Cristeau O1, Toumi M2 1 Creativ Ceutical, Paris, France; 2University Claude Bernard Lyon 1, Lyon, France OBJECTIVES: To investigate prescription patterns for setron antiemetics used alongside chemotherapy in France, Germany, Italy, and the UK. METHODS: Prescription data from June 2008 to July 2009 were extracted from the Synovate Oncology Monitor, an ongoing database tracking prescriptions of anticancer therapies and concomitant medications. Patients receiving setron prophylaxis (dolasetron, granisetron, ondansetron, palonosetron, or tropisetron) with Moderately emetic chemotherapy (MEC) or highly emetic chemotherapy (HEC) were selected. RESULTS: The sample included 1715, 1658, 1830, and 1651 patients with MEC regimens in France, Germany, Italy, and UK, respectively, and 557, 761, 903, and 627 patients with HEC regimens. Ondansetron was used in 85.1%, 45.8%, 52.1%, and 83.3% of patients with MEC regimens in France, Germany, Italy, and UK respectively, and 92.5%, 46.0%, 43.8%, and 62.2% of patients with HEC regimens. The shares of different setrons were relatively similar between MEC and HEC regimens in Gemany. a stronger 13th Euro Abstracts association between setron and regimen category was found in other countries, with different trends according to country. Doses of setrons were slightly higher for HEC regimens than for MEC regimens in Italy, but substantially higher in other countries. The average daily dose was found to increase with treatment duration in all countries. Neurokinin 1 antagonists (NK1) were used alongside setron in up to 46% of patients in Germany, depending on setron, but were very rarely used in Italy. CONCLUSIONS: Utilization of setrons and concomitant medications (NK1, antipsychotics) varied widely between European countries. It appears that German prescribers adapted the dose of setron rather than the drug according to regimen category, contrary to Italian prescribers. Such international comparisons may provide useful insight for improving practice in different countries. PCN29 EPIDEMIOLOGY AND HEALTH-CARE UTILIZATION FOR RESECTED SQUAMOUS CELL CARCINOMA OF HEAD AND NECK (SCCHN) IN SOUTH KOREA Kim K1, Amonkar M2, Lykopoulos K3, Kasteng F1, Högberg D1 1 i3 Innovus, Stockholm, Sweden; 2GlaxoSmithKline, Philadelphia, PA, USA; 3GlaxoSmithKline, London, UK OBJECTIVES: In South Korea, 2617 new cases of head and neck cancer (HNC) were diagnosed and 1473 deaths were reported in 2005. In Western countries, SCCHN represents more than 90% of HNCs and surgery is the primary treatment modality for SCCHN. Chemotherapy is often an important treatment component combined with radiotherapy for patients diagnosed with locally advanced disease. This study aimed to estimate the number of resected SCCHN patients and describe their treatment and health-care utilization in South Korea. METHODS: The study included the following cancer sites: oral cavity, pharynx, and larynx. The main source of data for the probabilities of having different types of SCCHN treatments was results from retrospective analyses found through literature review using MEDLINE, EMBASE, Cochrane, and the Korean Medical Database. Cancer incidence was based on data from the National Cancer Registry. In addition, expert opinion of a HNC surgeon in South Korea complemented the literature review. RESULTS: We estimated that SCCHN constituted 91% of HNC cases; 69% of the SCCHN cases were surgically resected. The number of new cases of resected SCCHN in South Korea was 1642 (oral cavity, 501; pharynx, 397; and larynx, 744). Fifity-two percent of resected SCCHN patients had neck dissections (45% of them were radical neck dissections) and 6% of resected SCCHN patients had total laryngectomy. Ten percent of patients had secondary surgery and 75% had reconstructive surgery. Sixty-four percent of patients received radiotherapy, whereas 33% received chemotherapy combined with surgery or as concomitant therapy with radiotherapy. CONCLUSIONS: In South Korea, resected SCCHN is estimated to constitute 1% of all cancers. However, 29% of patients with resected SCCHN had major surgeries (radical neck dissection or total laryngectomy), causing a significant negative impact on the quality of life of patients and contributing to a high economic burden to the patients and society. PCN31 THE PREDICTIVE AND PROGNOSTIC ROLE OF ERCC1 IN ADVANCED NON-SMALL CELL LUNG CANCER: A SYSTEMATIC REVIEW AND META-ANALYSIS Roth J, Carlson JJ University of Washington, Pharmaceutical Outcomes Research and Policy Program, Seattle, WA, USA OBJECTIVES: ERCC1 mRNA expression has the potential to be a useful tool in the clinical treatment of advanced NSCLC; however, considerable uncertainty remains as to the biomarker’s magnitude of effect on tumor response and survival. The objective of this study was to evaluate the association between ERCC1 mRNA expression level and treatment outcomes in advanced NSCLC patients treated with platinum-based chemotherapy using a systematic review and meta-analytic techniques. METHODS: A systematic search of the literature was conducted using MEDLINE and EMBASE databases from June 1995 to May 2010. Additionally, a search of ASCO Annual Meeting abstracts was conducted. Eligible studies were evaluated for clinical, methodological, and statistical heterogeneity. Pooled RESULTS: The pooled analysis demonstrated a statistically significant higher probability of response (according to the RECIST criteria) to platinum-based chemotherapy for patients with low ERCC1 mRNA expression levels relative to patients with high ERCC1 mRNA expression levels (OR: 0.64, 95% CI 0.42–0.95). Low ERCC1 mRNA expression level was also associated with significantly improved overall survival relative to high mRNA expression levels (HR: 0.44, 95% CI 0.31–0.62). CONCLUSIONS: In conclusion, ERCC1 mRNA expression level may provide useful clinical information in terms of the likelihood of platinum-based chemotherapy response and overall survival prognosis. Ongoing analyses are focused on evaluating the predictive role of ERCC1 in overall survival. The findings of this analysis also reveal a need for standardization in characterizing “high” and “low” ERCC1 mRNA expression, as the studies evaluated herein varied considerably in their methods. 13th Euro Abstracts CANCER – Cost Studies PCN32 COMPARING FORECASTED WITH ACTUAL BUDGET IMPACT: REIMBURSED CANCER DRUGS IN THE NETHERLANDS Thuresson PO1, Heeg B1, Botteman M2 1 Pharmerit, Rotterdam, The Netherlands; 2Pharmerit North America, LLC, Bethesda, MD, USA OBJECTIVES: In the The Netherlands, a new drug is priced to the level of an existing drug cluster, if these are considered similar. However, when a drug is not considered similar to a cluster, a separate price can be requested. Before this separate price is approved, sufficient cost-effectiveness and budget impact (BI) data have to be provided to the Dutch Health Care Insurance Board (CVZ). The present analysis investigated how close the submitted BI-forecasts were to the actual BIs for cancer treatments in the The Netherlands. METHODS: The publicly available evaluations by the CVZ were assessed for cancer agents. From these, forecasted number of users, drug costs, treatment lengths, and the BIs were derived. The predicted BIs were compared with actual BIs from the GIP database (Dutch Drug Information System). The predicted BI was standardized with respect to time to ensure that actual and forecasted BIs shared the same starting point. Broadening of the indication was taken into account and substitution was considered. To further explain any difference in forecasted and actual BI, various variables such as drug cost, number of users, market share (when applicable), and treatment lengths were investigated. RESULTS: The search provided five relevant cases. The forecasted budgets were lower than the actual ones in four out of the five cases. The forecasted and actual BI differed up to 250%. Data on drug substitution was insufficient and therefore not considered. The differences between predicted and observed BIs were explained primarily by an underestimation of the number of patients eventually receiving the evaluated treatments. The published BI forecasts contained no/very limited sensitivity analyses. CONCLUSIONS: Most BI forecasts underestimated the actual BI for new cancer drugs. One explanation was that the growth of the patient population was often underestimated. Improvements in predicting total market size and penetration should be considered, as well as more elaborate sensitivity analysis. PCN33 IMPACT BUDGETING IN CROATIA: FULVESTRANT EXAMPLE 1 1 2 2 3 4 Culig J , Leppee M , Reic I , Serdar S , Boskovic J , Colak N 1 Andrija Stampar Institute of Public Health, Zagreb, Croatia; 2AstraZeneca d.o.o, Zagreb, Croatia; 3University of Zagreb, Zagreb, Croatia; 4Colpharm d.o.o, Siroki Brijeg, Bosnia OBJECTIVES: Chemotherapy is standard choice as third-line treatment in patients with hormone-dependent metastatic breast cancer. The field research has shown that there is a need for additional hormonal drug in order to delay expensive and harmful chemotherapy. The aim is to analyze the impact on the budget Croatian Health Insurance Institute (CHII) budget including the drug fulvestrant 250 mg on the basic drug list as a third-line treatment in patients with hormone-dependent metastatic breast cancer. METHODS: Markov model was used to develop a new treatment scenario. Program to calculate the financial impact on the budget was developed in Microsoft Excel. RESULTS: The scenario with fulvestrant given as a third line of treatment in a period of 4 months, followed by 4 months of paclitaxel, 3 months kapecitabine, and, finally, 4 months with only supportive care was designed. This would delay the application of chemotherapy for an average of 4 months. Therefore, significantly fewer patients will be treated with chemotherapy. Additionally to that, a small part of patients would receive chemotherapy over a shorter period. The quality of life of metastatic breast cancer patients during their life expectancy would be significantly better, with the same expected survival rate and potentially lower cost of treatment. The impact budget analysis has shown reduce of therapy cost (4,202,922 Croatian Kuna or c579,713). CONCLUSIONS: A need for additional line of hormonal drug as a third-line treatment that can delay the implementation of poorly tolerated and expensive chemotherapy was established. The effectiveness of fulvestrant qualified it as a good candidate after the failure of the current endocrine therapy. The economic evaluation using Markov models and analysis of the budget impact in Croatia has been proven that adding fulvestrant in sequential treatment of metastatic breast cancer HR+ patients is cost-effective as third-line hormonal therapy in advanced breast cancer. PCN34 BUDGET IMPACT ANALYSIS OF CAPECITABINE IN ADJUVANT TREATMENT OF PATIENTS WITH RESECTED DUKES’ C COLON CANCER (CC) FROM POLISH PUBLIC PAYER’S AND PATIENT’S PERSPECTIVES Oleszko R1, Kawalec P2, Drobniak A3, Stec R4, Streb J5, Russel-Szymczyk M6, Szkultecka-Debek M6 1 Centrum HTA, Kraków, Poland; 2Jagiellonian University, Kraków, Poland; 3Holycross Cancer Center, Kielce, Poland; 4Military Institute of Medicine, Warsaw, Poland; 5The University Hospital in Krakow, Krakow, Poland; 6Roche Polska Sp. z o.o., Warsaw, Poland OBJECTIVES: The aim of the analysis is to determine budget impact of reimbursement of capecitabine monotherapy used in the adjuvant therapy of patients after resection of stage III CC (Dukes’ C). METHODS: Cost data were collected from Polish public payer’s (National Health Fund) and patient’s perspectives and calculated for a 3-year time horizon. The population was estimated based on the Polish National Cancer Register. The following direct medical costs were included: cost of drugs used in A257 adjuvant, I and II line chemotherapy, drug administration costs, adverse events, and monitoring costs. In “scenario A,” capecitabine was not reimbursed by the public payer, while the “scenario B” was developed under assumption of 100% capecitabine reimbursement. The proportion of patients treated with capecitabine and other drugs used in CC treatment was assessed based on the results of a questionnaire study conducted among Polish clinical experts. a cohort Markov model was used for simulation of long-term health outcomes and costs. a range of variables was tested in oneand multi-way sensitivity analyses. RESULTS: “Scenario B” introduction led to savings equal to 24,730,000 PLN in the first year, 25,379,000 PLN in the second year, and 26,712,000 PLN in the third year from public payer’s perspective (1 EURO = 4.1 PLN). Similarly, savings were observed when patient’s perspective was assumed. Results of sensitivity analysis confirmed conclusions from base-case analysis. CONCLUSIONS: One hundred percent reimbursement of capecitabine used in adjuvant therapy of patients after the resection for Dukes’ C CC leads to savings from both a public payer’s and patient’s perspectives in Poland. PCN35 CLINICAL AND ECONOMIC IMPLICATIONS OF SCREENING FOR KRAS MUTATIONS IN METASTATIC COLORECTAL CANCER PATIENTS IN SPAIN: A COST-EFFECTIVENESS AND BUDGET IMPACT MODEL Saz-Parkinson Z, Amate JM Instituto de Salud Carlos III, Madrid, Spain OBJECTIVES: Colorectal cancer is the second most common cancer in Spain (25,600 incident cases and over 13,500 deaths yearly). Mutations in the KRAS oncogene are associated with a poor response to epidermal growth factor receptor (EGFR) inhibitor therapy in metastatic colorectal cancer (mCRC). In addition to avoiding unnecessary toxicity, implementing routine KRAS screening and limiting the use of EGFR inhibitors to patients with wild-type KRAS may lead to cost savings. The objective is to estimate health and economic outcomes derived from carrying out KRAS determination in mCRC patients requiring an aggressive treatment in Spain. METHODS: Cost-effectiveness and budget impact analyses from national, regional, and hospital perspectives were developed in an interactive mode. Information was obtained from literature review of clinical trials (treatment response rates and prevalence of KRAS mutations), official statistics (epidemiological data), and local databases (therapy and KRAS test costs). Multivariate sensitivity analysis can be performed by changing the technology for KRAS testing (PCR, real-time PCR), the chemotherapy regime (FOLFIRI, FOLFOX-4), and the proportion of wild-type patients to be treated with each therapy (bevacizumab, cetuximab). RESULTS: About 7000 mCRC patients will require an aggressive treatment in Spain in 2010. a strategy based on treating wild-type patients with cetuximab would raise first-line response rate from 44.80% to 53.84% (20.18%, 633 patients more) when combined with FOLFIRI (19.05%, 620 patients with FOLFOX-4) with an additional cost estimate of c22,117 per response (c17,201 with FOLFOX-4). Budget impact of this approach amounts to a 9.06% (5.01% with FOLFOX-4) increase in resources devoted to mCRC treatment. Univariate and multivariate sensitivity analyses yield ratios between c19,101 and c27,531 and budget impact estimates of 0.16–9.16%. CONCLUSIONS: Treating wild-type KRAS mCRC patients with cetuximab would lead to over 600 additional patients (19–20%) with first-line response every year in Spain, with an associated budget impact of 9–11% of therapy cost. PCN36 COST IMPLICATIONS OF RESTRICTING PALONOSETRON UTILIZATION FOR THE TREATMENT OF CINV: RESULTS FROM AN ANALYTIC MODEL FROM A PAYER PERSPECTIVE Powers A1, Sarnes E2, Knoth RL1, Balu S1, Buchner D1 1 Eisai, Inc., Woodcliff Lake, NJ, USA; 2Xcenda, Palm Harbor, FL, USA OBJECTIVES: Palonosetron is a 5-HT3 receptor antagonist (5-HT3-RA) indicated for the prevention and treatment of chemotherapy induced nausea and vomiting (CINV). Using an analytic model, this study explored the consequences of the restricted use of palonosetron for the treatment of CINV from a payer perspective. METHODS: The analytic model utilized rates of uncontrolled CINV among patients with cancer and initiated on palonosetron or another 5-HT3-RA for antiemetic prophylaxis. Other inputs included direct medical cost of a CINV event defined by inpatient and outpatient visits, fixed Medicare reimbursement fees for medications (average sale price + 6%), national prescription utilization data, and the costs associated with executing a prior authorization. Total costs associated with restricting palonosetron usage by 10% with a corresponding 10% increase in generic 5-HT3-RA utilization were modeled for six cycles of moderately (MEC) and highly emetogenic chemotherapy (HEC). RESULTS: In a 1 million member health plan, 15,000 patients (1.5%) were estimated to require chemotherapy treatment annually. Of these, 10,800 patients (72%) were estimated to receive 5-HT3-RA prophylaxis on the first day of chemotherapy, with 7587 (51%) on either a MEC or HEC regimen. Result of the model demonstrated that a 10% reduction in palonosetron led to an overall savings in antiemetic costs equivalent to $762,616. However, total direct medical costs to the health plan (including the cost to treat uncontrolled CINV and the cost of the prior authorization program) increased by an estimated $865,457. Total cost of care, therefore, increased by an estimated $0.02 PMPM. CONCLUSIONS: Results from an analytic model showed that restricting palonosetron utilization by 10% through a prior authorization, while decreasing medication costs, produced an overall increase in total direct costs to the MCO. These results suggest that further studies on the clinical and economic impact of appropriate utilization of antiemetic prophylaxis for CINV are warranted. A258 PCN37 BI OF IRESSA IN NSCLC IN THE NETHERLANDS: A HOSPITAL PERSPECTIVE Langenfeld M, Scheer F AstraZeneca BV, Zoetermeer, The Netherlands OBJECTIVES: To gain insight in the total Costs of first-line treatment of NSCLC phase III/IV in the Dutch hospital setting. To calculate the budget impact of adding Iressa (gefitinib), including EGFR mutation testing, to the treatment sequence. To identify costs of EGFR mutation testing. METHODS: A budget impact model was constructed by Medaxial and adapted by AstraZeneca the The Netherlands to better reflect the Dutch situation. The model calculates the budget impact of EGFR testing and the resulting therapy change for NSCLC patients in Dutch hospitals. The model covers the first-line treatment of NSCLC fase IIIb/IV patients and calculates total costs in a hospital setting and budget impact, taken into account the following costs—Medication—EGFR mutation testing—Administration and monitoring—Treating grade III/ IV adverse events. The model also calculates—Costs of delivery of oral therapies (outside hospital costs). RESULTS: Before introduction of Iressa, total costs of firstline NSCLC treatment in Dutch hospitals was c41,982.936. After introduction of Iressa, total costs in hospital setting slightly decreased to c41.939,604 (year 1). Cost of EGFR mutation testing is c198,432, but medication costs of the hospital will decrease with c248,193. CONCLUSIONS: Adding extra diagnosis will introduce new costs to the NSCLC treatment. However, for an individual hospital, the extra costs are limited. And the total costs of first-line NSCLC treatment even decrease, since the reimbursement of Iressa is outside of the hospital budget. As a result, the hospital has less costs if patients are tested and the appropriate patients then treated with Iressa. At the moment, for some hospitals, EGFR mutation testing is still seen as a hurdle. Apart from the ethical point of view (making sure patients will receive the medication they will benefit most from), also from costs perspective, it is worth to test. PCN38 INFLUENCE OF ME-TOOS TO POSSIBLE SAVINGS DUE TO BIOSIMILARS Fuezi J, Reichardt B Sickness Fund Burgenland, Eisenstadt, Burgenland, Austria OBJECTIVES: To avoid the waste of market shares, me-toos of established top-selling biopharmaceuticals enter the market before their patent expiry. After switching the sales to the me-too, called “step-innovation” by the provider, the possible savings of biosimilars diminish as the sales are parked at the me-toos. The objective of the study is to analyze this hypothesis for the two2 substance classes erythropoiesis-stimulating agents (ESA) and granulocyte colony-stimulating factors (G-CSF). METHODS: By analyzing the account data of regional Austrian sickness funds, the share of prescriptions of filgrastim in GCSF (filgrastim plus lenograstim plus pegfilgrastim) and of epoetin in ESA (epoetin plus darbepoetin) are correlated to the costs per package. These data are shown for several periods, 6 months before the availability of Biosimilars, 6 montha after their availability, and 2 further half-years for ESA. To take regional influence into consideration, the data are shown for all nine regional sickness funds of Austria. RESULTS: The shares of prescription for epoetin in ESA show a huge regional dispersion from 42% to 88% and those of filgrastim in GCSF from 23% to 66%. The average costs per package have an inverse relation to their market share. These data are reconfirmed through data of the other periods. The average costs per package have been declined with increased market share of the biosimilars. The influence of the market share of the me-toos is though overwhelming. CONCLUSIONS: For pharmaceutical companies, it is a profit-maximizing strategy to substitute their products in time before patent expiry. The less the substitution by patent-protected alternatives succeeds, the higher the achievable cost saving potential for the sickness funds is. PCN39 COST UTILITY OF HUMAN PAPILOMA VIRUS VACCINE IN SPAIN Callejo D, Lopez-Polin A, Blasco JA Agencia Lain Entralgo, Madrid, Spain OBJECTIVES: It is well known that a persistent infection by human papillomavirus (HPV) is an essential cause of cervical cancer. Prophylactic HPV vaccines aimed at preventing precancerous cervical lesions and cervical cancer are currently available and may be used for primary prevention of cervical cancer. To define the efficiency of using HPV vaccine within a cervical cancer screening program with cytology compared with the strategy of only screening with cytology. METHODS: A comprehensive search of studies was developed in the main electronic databases including primary studies assessing HPV vaccine efficacy and/or safety or HPV. Meta-analysis was conducted when the studies were homogeneous. a decision analysis was performed using a Markov model based on the results of the systematic review and information about the natural history of the disease (HPV infection, cervical cancer). The model allowed to estimate cancer incidence and associated mortality, life expectancy, and associated costs, with the objective of performing an economic evaluation. RESULTS: A total of 11 studies were included to assess HPV vaccines: seven for efficacy and 11 for safety assessment. The efficacy of vaccine in preventing CIN2+ was 96% (95% CI: 91% to 98%). Modified intention-to-treat and intention-to-treat analyses were also undertaken and the results supported this trend, although efficacy was lower. The economic evaluation showed that the vaccination strategy would imply a cost-effectiveness ratio of around c10,000 per quality-life adjusted-years (QALY). CONCLUSIONS: There is evidence from RCTs that the HPV vaccines are safe and effective in 13th Euro Abstracts the prevention of cervical cancer precursor lesions. On top of this, this efficacy is reached with a reasonable cost-effectiveness ratio and within the accepable limits of the Spanish National Health System. PCN40 COST ANALYSIS OF MANAGING ADVERSE EVENTS IN THE TREATMENT OF METASTATIC RENAL CELL CARCINOMA IN PORTUGAL: A COMPARISON OF BEVACIZUMAB IN COMBINATION WITH INTERFERON ALFA-2A AND SUNITINIB Silva CI1, Monteiro I2, Schwander B3 1 Eurotrials Scientific Consultants, Lisbon, Portugal; 2Roche Farmacêutica Química, Amadora, Portugal; 3AiM GmbH—Assessment in Medicine, Research and Consulting, Lörrach, Germany OBJECTIVES: The burden of metastatic renal cell carcinoma (mRCC) is substantial for patients and society. Bevacizumab (BEV) combination with interferon alfa-2a (IFN) has demonstrated to prolong mRCC patients’ progression-free survival and to have comparable efficacy to sunitinib (SUN). However, tolerability differs between these treatment alternatives and it is therefore of importance to evaluate the economic impact of adverse events (AEs) management, for each alternative, in the daily clinical practice in Portugal. METHODS: A linear decision analytical model was applied considering direct medical costs only in the Portuguese Health System perspective. AEs incidences associated with each of the two alternatives were extracted from published literature. Health resource consumption was estimated based on an expert panel of Portuguese oncologists and urologists. Corresponding unitary costs were obtained through national official sources. The considered time horizon was 1 year. The basecase analysis includes all grades AEs and a normal dose (nine MUI, three times weekly) of IFN scenario. Deterministic univariate sensitivity analyses were performed to test the robustness of the model including grade 3–4 AEs analysis only and IFN low-dose (six MUI or three MUI, three times weekly) scenarios. RESULTS: The associated costs of managing AEs were considerably lower with BEV + IFN than with SUN in Portugal. The average treatment costs for all grade AEs per patient was c1472 for SUN and c1093 for BEV + IFN resulting in a difference of −379 c (a cost saving of 26% for BEV + IFN vs. SUN). Sensitivity analyses showed that BEV + IFN remains the less costly option when alternative scenarios are considered and that a low-dose IFN would lead even to further cost savings. CONCLUSIONS: BEV + IFN is a more tolerable and hence cost-saving alternative for the Portuguese Health System considering the AE management costs of mRCC treatment when compared to SUN. These results are consistent with previous evidence for other countries. PCN41 ECONOMIC OUTCOMES AMONG 2ND LINE NON-SMALL CELL LUNG CANCER PATIENTS IN THE OUTPATIENT COMMUNITY SETTING Gruschkus S1, Reyes C2, Forsyth M1, Ravelo A2, Nadler E1 1 US Oncology, The Woodlands, TX, USA; 2Genentech, Inc., South San Francisco, CA, USA OBJECTIVES: Second-line monotherapy options for advanced NSCLC include erlotinib (E), docetaxel (D), or pemetrexed (P). The purpose of this retrospective study was to compare economic outcomes among patients (pts) receiving 2nd line monotherapy in the outpatient community setting. METHODS: Using US Oncology’s iKnowMed EMR data, we identified advanced NSCLC patients who received 2nd line monotherapy from July 1, 2006 to June 30, 2008. Economic outcomes were derived using outpatient claims and pharmacy data and included total outpatient, chemotherapy, supportive care costs, and frequency of outpatient physician visits, lab procedures, and acute care (ER/inpatient) visits. All economic outcomes were calculated as per-patient month (PPM) metrics over a 12-month follow-up period. Multiple regression analyses were used to estimate the independent association between treatment (E, D, or P) on outcomes after controlling for age, gender, stage at diagnosis, baseline hemoglobin, and performance status. RESULTS: We identified 610 pts—73 received E, 87 received D, and 450 received P. Total cost, chemotherapy costs, and supportive care costs differed significantly by treatment, as did frequency of outpatient visits and lab procedures. Relative to P, total adjusted costs PPM is $1579 lower for D and $1584 lower for E (P < 0.05). Majority of the cost savings are due to decreased chemo-related costs. Outpatient visits, lab procedures, and acute care visits are also less frequent with E relative to P (−2.6 PPM, P < 0.05). CONCLUSIONS: In US outpatient setting, pts receiving E and D have statistically significant lower costs and resource use relative to pts receiving P. PCN42 COSTS OF MANAGING ADVERSE EVENTS OF FIRST-LINE THERAPY FOR METASTATIC RENAL CELL CARCINOMA IN MEXICO: BEVACIZUMAB IN COMBINATION WITH INTERFERON-ALPHA-2A COMPARED WITH SUNITINIB Carlos F, Ramirez J, Aguirre A R a C Salud Consultores S.A. de C.V., México, D.F., Mexico OBJECTIVES: Bevacizumab plus interferon-α2a (BEV+IFN) prolongs progressionfree survival (PFS) to >10 months, providing comparable efficacy to sunitinib in patients with metastatic renal cell carcinoma (mRCC). However, the type and frequency of adverse events (AE) differ between these two regimens. We aimed to assess the costs of managing AE of grade 3/4 of these regimens from the perspective of public health-care system in Mexico. METHODS: A linear decision analytic model was developed to compare the direct medical costs of managing AE of grade 3/4 of BEV + IFN and sunitinib in patients with mRCC. AE of grade 1/2 are assumed to entail 13th Euro Abstracts very low costs or even no medical attention, and therefore were not taken into account. Data sources included the published incidence rates for the 25 most frequent AE in controlled clinical trials with BEV + IFN or sunitinib. a panel integrated by 10 local experts from different specialties was constituted to estimate medical and nonmedical resource use for diagnosis and treatment of each AE grade 3/4. Cost of medications involved in treating AE were taken from public bids and unit cost of medical services (outpatient medical consultations, laboratory and image tests, hospitalization at general ward and at intensive care unit, surgical and nonsurgical procedures, etc.) was gathered from official tariff lists. All costs are expressed in 2009 Mexican pesos (MXN). RESULTS: The average cost per patient for the management of grade 3/4 AE were 76.5% higher for sunitinib ($17,577) than those for BEV + IFN ($9959). The main cost drivers for sunitinib were hypertension, heart failure, and non-febrile neutropenia; for BEV + IFN, main cost drivers included proteinuria and arterial and venous thromboembolic events. CONCLUSIONS: BEV + IFN has a more tolerable AE profile when compared to sunitinib, which is also reflected in the nearly double cost for managing AE with sunitinib in patients with mRCC. PCN43 COST COMPARISON OF ERLOTINIB VERSUS PEMETREXED FOR THE FIRST-LINE MAINTENANCE TREATMENT OF PATIENTS WITH LOCALLY ADVANCED OR METASTATIC NON-SMALL CELL LUNG CANCER IN ITALY Ravera S1, Walzer S2, Ray J2 1 Roche S.p.A., Milano, Italy; 2F. Hoffmann-La Roche Pharmaceuticals AG, Basel, Switzerland OBJECTIVES: First-line chemotherapy for locally advanced or metastatic non-small cell lung cancer (mNSCLC) is usually limited to four to six cycles, as prolonged exposure leads to cumulative toxicity without additional survival benefit. Maintenance therapy represents a new treatment option which can delay disease progression and extend survival in patients with mNSCLC. Erlotinib and pemetrexed are currently the only treatments specifically approved for this indication by the European Medicines Agency and US Food and Drug Administration; therefore, it is important to compare the monthly treatment costs of using erlotinib or pemetrexed for the maintenance therapy of patients with mNSCLC. METHODS: Italian monthly treatment costs were calculated as the sum of the ex-factory costs for the average dose (erlotinib = 150 mg/ day, pemetrexed = 500 mg/m2) over a 30-day treatment duration plus administration costs. Monthly administration costs were derived from regional tariffs for oncology drugs. RESULTS: Monthly drug costs for erlotinib maintenance therapy are lower than for pemetrexed (c1517 vs. c2770, respectively). In addition, as an intravenous treatment, pemetrexed is associated with additional costs related to administration (estimated at c140 per month), whereas orally administered erlotinib is not associated with any administration costs. Pemetrexed total monthly treatment costs are therefore c2910, c1393 higher than erlotinib total monthly treatment costs. The cost saving associated with erlotinib would allow approximately 92% more patients to be treated with erlotinib maintenance therapy, based on a fixed health-care budget. Furthermore, it is anticipated that the management of pemetrexed-related adverse events (e.g., neutropenia, anaemia) would be more costly than those related to erlotinib use (e.g., rash, pruritus). Therefore, the cost saving when using erlotinib versus pemetrexed for first-line maintenance therapy may be greater in a real-world setting. CONCLUSIONS: Based on Italian costs, erlotinib is a cost-saving treatment option compared with pemetrexed, for the first-line maintenance therapy of patients with locally advanced or mNSCLC. PCN44 COST COMPARISON OF ERLOTINIB VERSUS PEMETREXED FOR THE FIRST-LINE MAINTENANCE TREATMENT OF PATIENTS WITH LOCALLY ADVANCED OR METASTATIC NON-SMALL CELL LUNG CANCER IN SPAIN Castro de Carpeño J1, Castro-Gómez AJ2, Walzer S3, Ray J3 1 La Paz University Hospital, Madrid, Spain; 2Roche Farma, Madrid, Spain; 3F. Hoffmann-La Roche Pharmaceuticals AG, Basel, Switzerland OBJECTIVES: First-line chemotherapy for locally advanced or metastatic non-small cell lung cancer (mNSCLC) is usually limited to four to six cycles, as prolonged exposure leads to cumulative toxicity without additional survival benefit. Maintenance therapy represents a new treatment option which can delay disease progression and extend survival in patients with mNSCLC. Erlotinib and pemetrexed are currently the only treatments specifically approved for this indication by the European Medicines Agency and US Food and Drug Administration; therefore, it is important to compare the monthly treatment costs of using erlotinib or pemetrexed for the maintenance therapy of patients with mNSCLC. METHODS: Spanish monthly treatment costs were calculated as the sum of the ex-factory costs for the average dose (erlotinib = 150 mg/day, pemetrexed = 500 mg/m2) over a 30-day treatment duration plus administration costs. Monthly administration costs were obtained from regional tariffs (Galician Health Service). RESULTS: Monthly drug costs for erlotinib maintenance therapy are lower than for pemetrexed (c2045 vs. c2914, respectively). In addition, as an intravenous treatment, pemetrexed is associated with additional costs related to administration (estimated at c235 per month), whereas orally administered erlotinib is not associated with any administration costs. Pemetrexed total monthly treatment costs are therefore c3149, c1104 higher than erlotinib total monthly treatment costs. The cost saving associated with erlotinib would allow approximately 54% more patients to be treated with erlotinib maintenance therapy, based on a fixed health-care budget. Furthermore, it is anticipated that the management of pemetrexed-related adverse events (e.g., neutropenia, anaemia) would be more costly than those related A259 to erlotinib use (e.g., rash, pruritus). Therefore, the cost saving when using erlotinib versus pemetrexed for first-line maintenance therapy may be greater in a real-world setting. CONCLUSIONS: Based on Spanish costs, erlotinib is a cost-saving treatment option compared with pemetrexed, for the first-line maintenance therapy of patients with locally advanced or mNSCLC. PCN45 DIFFERENCES IN HEALTH-CARE COSTS FOR PATIENTS WITH CASTRATION-RESISTANT PROSTATE CANCER (CRPC) TREATED BY ONCOLOGISTS OR UROLOGISTS Engel-Nitz NM1, Alemayehu B2, Nathan F2, Parry D3, Kulakodlu M1 1 i3 Innovus, Eden Prairie, MN, USA; 2AstraZeneca, Wilmington, DE, USA; 3AstraZeneca, Macclesfield, Cheshire, UK OBJECTIVES: Patients with CRPC may be treated by urologists or oncologists. This study examined differences in total health-care costs and prostate cancer-specific costs in patients treated by oncologists or urologists. METHODS: A retrospective study design used medical and pharmacy claims (2001–2007) to identify patients with CRPC from a large US-managed care health plan. Patients were stratified based on the specialist providing treatment following castration; an oncologist (with/without a urologist, ONC), and a urologist without an oncologist (URO). A 6-month baseline period was used to assess patient characteristics and initial clinical status; a variable follow-up period (until disenrollment or December 31, 2008) was used to assess total health-care costs. Lin’s regression was used to assess costs adjusting for the variable follow-up and patient and treatment characteristics. RESULTS: A total of 995 URO and 1590 ONC patients with CRPC were identified. Mean age was higher in URO patients than in ONC patients (75.5 vs. 71.1 years, P < 0.001). The URO cohort had a lower average Charlson comorbidity score (3.7 vs. 4.9, P < 0.001), fewer comorbid illnesses (10.1 vs. 11.1, P < 0.001), and were less likely to have other cancers (17.7% vs. 27.4%, P < 0.001) or to have had hormones, chemotherapy, and radiation treatment during the baseline period. After multivariate adjustment, mean total health-care costs during the first year were $31,792 (URO), $54,306 (ONC with chemotherapy, P < 0.05), and $30,894 (ONC without chemotherapy); during 6 years of follow-up, cumulative costs rose to $86,706 (URO), $168,794 (ONC with chemotherapy), and $114,180 (ONC without chemotherapy), P < 0.05 for all. a similar pattern was observed for prostate cancer-specific cumulative costs. CONCLUSIONS: CRPC patients treated by oncologists, particularly patients with chemotherapy, had higher total and prostate cancerrelated health-care costs than patients treated by urologists. PCN46 ECONOMIC EVALUATION OF ONCOTYPE DX® TO TARGET CHEMOTHERAPY USE IN LYMPH-NODE–NEGATIVE, OESTROGENRECEPTOR–POSITIVE, EARLY-STAGE BREAST CANCER IN IRELAND Lacey L1, Hornberger J2 1 Lacey Solutions Ltd., Skerries, Ireland; 2Stanford University & Cedar Associates, Menlo Park, CA, USA OBJECTIVES: Oncotype DX® is a clinically validated assay used to guide chemotherapy decision-making for patients with early-stage breast cancer. Patients classified as low risk by Oncotype DX® have low likelihood of benefitting from chemotherapy. By foregoing chemotherapy, patients avoid the risk of chemotherapy-related toxicities. For those patients reclassified by Oncotype DX® as high risk, the assay identifies patients who are likely to gain a large benefit from chemotherapy. The study objective was to estimate the health-care costs of using Oncotype DX® testing in early-stage, lymph node-negative breast cancer in Ireland. METHODS: A cost-analysis estimated the health-care costs (chemotherapy, administration, adverse events [AEs], and G-CSF costs) in patients whose treatment decisions are informed by Oncotype DX® testing. The perspective was that of the Irish health-care system. The chemotherapy regimen was docetaxel and cyclophosphamide (4 × 21-day cycles), costing approximately c9200. Univariate sensitivity analysis was performed, together with a probabilistic sensitivity analysis (PSA) of the net reduction in chemotherapy usage from Oncotype DX® testing. In a meta-analysis of seven published studies, there was an estimated 30% (95% CI −40%, −21%; P = 0.0003) absolute reduction in chemotherapy usage after Oncotype DX® testing (ratio 0.49 [95% CI 0.41, 0.58]; P < 0.00001). RESULTS: Adoption of Oncotype DX® testing resulted in approximate cost-neutrality (0.4% increase in cost) to the Irish health-care system, under the above conditions. The main cost drivers were: net reduction in chemotherapy usage from Oncotype DX® testing and the rate of G-CSF usage. From the PSA, the probability of Oncotype DX® being cost-saving is approximately 47%. CONCLUSIONS: Using Oncotype DX® to inform chemotherapy decisions in early-stage breast cancer has the potential to reduce the incidence of chemotherapy-induced AEs, while being approximately cost-neutral to the Irish health-care system. a cost-effectiveness analysis would be expected to result in a low incremental cost-effectiveness ratio. PCN47 CHANGE OF ANTIFUNGAL TREATMENT PATTERNS AND ASSOCIATED COSTS IN PATIENTS WITH ACUTE MYELOGENOUS LEUKEMIA (AML) AFTER CHEMOTHERAPY IN A GERMAN HOSPITAL FROM 2004 TO 2006 Boehme A1, Atta J1, Mousset S1, Steffen B1, Serve H1, Hoelzer D1, Shlaen R2, Ehlken B2, Bug G1 1 Med. Clinic II, J.W. Goethe-University, Frankfurt, Germany; 2IMS Health, Munich, Germany OBJECTIVES: To describe changes in outcomes, treatment patterns and costs of the management of hospitalized patients with acute AML after chemotherapy in Germany A260 over a 3-year time horizon, with a special focus on prophylaxis and treatment of invasive fungal infections (IFI). METHODS: This was a retrospective, single-center study on AML patients hospitalized for chemotherapy, neutropenia, and infections after myelosuppressive chemotherapy from January 2004 to December 2006. Data on occurrence of IFI, treatment patterns, and resource utilization were collected by chart review. Direct medical costs were calculated from hospital provider perspective. RESULTS: In total, 471 hospitalization episodes in 212 patients were eligible for analysis. Occurrence of IFI decreased from 5.9% in 2004 to 1.9% in 2006. Mean hospital stay decreased from 28.7 ± 17.9 days (2004) to 22.4 ± 11.8 days (2006) (P < 0.05). From 2004 to 2006, use of a single antifungal drug increased from 30.4% to 46.9% of episodes, whereas use of multiple antifungal drugs decreased from 24.4% to 13.1%. Single antifungal drug use was dominated by azoles and increased from 23.7% (2004) to 43.4% of episodes (2006). Posaconzole monotherapy was applied in 26.7% of episodes. Use of liposomal amphotericin B declined from 21.4% to 3.8%, caspofungin from 19.3% to 8.1%, fluconazole from 25.2% to 11.9%, and voriconazole from 31.9% to 15.0%. Total costs per episode declined from c19,051 ± 19,024 (2004) to c13,531 ± 9,260 (2006) (P < 0.05); main reduction was observed for antimycotics, blood products, and hospital stay. CONCLUSIONS: These real-life data from a university hospital in Germany indicate that the antifungal management of AML patient hospitalized for chemotherapy, neutropenia, and infections after chemotherapy changed between 2004/5 and 2006. This change was accompanied by a decline in treatment costs. Results suggest that the introduction of posaconazole prophylaxis in 2006 has not only reduced the use of antifungal therapies but also the need for treatment with multiple antifungal drugs. PCN48 COST-EFFECTIVENESS COMPARISON OF APPROACHES FOR LOW-RISK PROSTATE CANCER CARE: THE NEED TO BALANCE COST AND UTILITY Olaye A1, New M2, Abenhaim L3 1 London School of Economics, London, UK; 2LA-SER Europe Ltd., London, UK; 3London School of Hygiene and Tropical Medicine, London, UK Prostate cancer is the most common form of cancer in UK men and incidence is on the rise, primarily driven by increased screening. Low-risk prostate cancer has diseasespecific survival consistently over 95%. Consequently, the UK has deemed radical treatments (RTs) comprising radical prostatectomy, radiotherapy, and brachytherapy as unnecessary for low-risk prostate cancer and recommends active surveillance (AS). AS does not address the disutility for patients living with cancer and in real-life over 50% of patients proceed to RT. Alternative “focal therapies” for this patient population are therefore generating interest. OBJECTIVES: Determine the cost-effectiveness of AS in the UK as practiced and contrast with focal therapy. METHODS: A Markov model was used to evaluate the cost-utility of treatments for low-risk prostate cancer. Input parameters for progression rates, efficacy, and side effects were derived from the literature. Biochemical and histological progression to RT plus patient choice were modeled. Hormonal treatment was included as salvage therapy. Age-related mortality rates were applied throughout the model. For focal therapy, all low-risk patients received treatment. Failure of treatment led to patients commencing AS, while biochemical progression at any time led to RT. Costs were taken from the UK NHS perspective. RESULTS: Over 25 years, AS delivered 13.3 QALYs (3.5% discount rate applied). The QALY value was heavily weighted by the utility of 0.84 associated with AS. Focal therapy delivered improved QALYs (14.6) over the same period and could be delivered for cost parity with AS. Cost-effectiveness is discussed. CONCLUSIONS: UK guidelines recommend AS for low-risk prostate cancer but do not meet the needs of patients, evinced by low QoL and by many patients choosing RT. Hence, AS has “hidden costs” in the UK system and there is a place for alternative treatment approaches such as focal therapies. PCN49 PREDICTORS OF THE DIRECT COSTS OF BREAST CANCER IN THE UNITED STATES ELDERLY POPULATION Davis KL1, Iyer S2, Candrilli S1 1 RTI Health Solutions, Research Triangle Park, NC, USA; 2Pfizer, Inc., New York, NY, USA OBJECTIVES: To assess cancer-related costs and related predictors among elderly breast cancer patients in the United States. METHODS: A retrospective study was conducted in subjects aged ≥65 years and diagnosed with breast cancer between January 1, 2000 and December 31, 2005. Patients were identified from the SEERMedicare linked database that combines clinical information on cancer cases with longitudinal (1991–2006) Medicare claims. An index date was defined as the date of the first observed breast cancer diagnosis. Costs (2009 US$) were aggregated from subjects’ index date until death, Medicare disenrollment, or database end (December 31, 2006) and included breast cancer-related surgery, radiotherapy, chemotherapy, and other medical encounters carrying a breast cancer diagnosis. Generalized linear models with a log link function and gamma distribution were used to assess predictors of costs. Age, race, stage at diagnosis, hormone receptor status (ER/PR), nodal status, Charlson comorbidity score, and chemotherapy use were key explanatory variables. RESULTS: The majority of the 66,217 breast cancer subjects selected were aged 70 to 79 years (47%), Caucasian (88%), and in localized stage (67%) at diagnosis. Median follow-up was 50 onths. Approximately 5% of cases were diagnosed in the metastatic stage with a median follow-up of 26 months. Approximately 73% of subjects with metastatic disease died during follow-up compared to 21% of localized cases. Cancer-related adjusted costs per patient were $55,120 (median $25,991) for 13th Euro Abstracts all cases and $153,421, $82,789, and $38,099, for metastatic, regional, and local stage cases, respectively. Regional and metastatic stage at diagnosis, increased age, increased number of cancer-positive lymph nodes, negative estrogen/progesterone receptor status, and chemotherapy use were found to be significant (P < 0.001) predictors of higher costs. CONCLUSIONS: Clinical characteristics indicating poorer prognosis are associated with significantly higher breast cancer costs. Patients with metastatic disease carry the highest cost of care despite having shorter follow-up and poorer survival. PCN50 BURDEN OF HOSPITALIZATIONS FOR HEPATOCELLULAR CARCINOMA PATIENTS IN A US POPULATION Tsong W1, Singer ME2, Ray S1 1 Abbott Laboratories, Abbott Park, IL, USA; 2Case Western Reserve University School of Medicine, Cleveland, OH, USA OBJECTIVES: Hepatocellular carcinoma (HCC) is a rapidly progressing, fatal disease. However, little is known regarding the hospitalization burden for these patients. This study compares the burden of hospitalizations due to HCC and hepatobiliary conditions relative to other nonhepatobiliary comorbid conditions. METHODS: Insurance claims (Janaury 1, 2000—December 31, 2008) from a geographically diverse, commercially insured US population were used to identify a cohort of patients with > = 1 HCC claim (index = 1st claim), age > = 18, and no other cancer diagnoses in the year prior to index. Hospitalizations were grouped by primary diagnosis (ICD-9) codes into the following categories: 1) HCC and hepatobiliary, and 2) nonhepatobiliary. Hospitalization burden was compared between these categories based on: number of hospitalizations, time to occurrence since HCC diagnosis, length of stay, and cost (2009 USD). The number of hospitalizations was compared using a sign test for patient-level differences. The remaining parameters were compared using general estimating equations to adjust for patients with multiple hospitalizations. RESULTS: This study identified 2927 HCC patients (mean age 50.4 years, 57% male) and 2192 hospitalizations. The subset of patients with > = 1 hospitalization (n = 1083, 37%) had an average of 2.02 admissions per patient with a median of 11.8 follow-up months. Compared to the nonhepatobiliary hospitalizations, the HCC and hepatobiliary hospitalizations had: a higher number of admissions (0.40 vs. 0.35 per patient; P < 0.001), a shorter time to occurrence (112 vs. 330 days; P < 0.001), a longer length of stay (6.4 vs. 5.4 days; P < 0.007), and a higher average cost per admission ($48,539 vs. 23,221; P < 0.001). The total cost of HCC and hepatobiliary hospitalizations was 2.3 times higher than nonhepatobiliary hospitalizations ($56,451,173 vs. $23,894,444). CONCLUSIONS: HCC and hepatobiliary conditions accounted for the majority of the hospitalization burden in liver cancer patients. Future HCC therapies demonstrating reduced symptom progression may reduce the prevalence, duration, and cost of HCC and hepatobiliary-related hospitalizations. PCN51 HOSPITALIZATION COSTS FOR HPV-RELATED CANCERS IN MALES AND FEMALES IN FRANCE Borget I1, Mathevet P2, Abramowitz L3, Largeron N4 1 Institut Gustave Roussy, Villejuif, France; 2Hopital Edouard Herriot, Lyon, France; 3Hopital Bichat Claude Bernard, Paris, France; 4Sanofi Pasteur MSD, Lyon, France OBJECTIVES: Human papillomavirus (HPV) infection is a necessary cause of cervical cancer and is associated with a subset of other anogenital (anal, vulvar, vaginal, penile) and head/neck cancers (oral cavity, oropharynx/pharynx, larynx). The overall HPVrelated disease burden is considerable in Europe. This study aimed to assess the hospital costs of HPV-related cancers for both genders in France. METHODS: Hospitalization costs were collected from a recent publication for cervical cancer and from the French national hospital database (PMSI) for other cancers. Costs included hospital stays, radiotherapy, and chemotherapy sessions. Annual costs for each cancer were estimated from the health-care payers’ perspective. RESULTS: In 2006, the annual costs of cervical cancer were estimated at c43.9 million. In addition, vulvar and vaginal cancers were associated with c9.7 million annually. For anal cancer, the annual costs were c20.3 million (male: c6.3 million; female: c14 million). Penile cancer hospitalizations represented c2.6 million annually. The estimated annual costs for head/neck cancers in 2007 for males and females, respectively, were c54.4 and c17.1 million for oral cavity, c128.2 and c21.7 million for oropharynx/pharynx, and c47.8 and c5.5 million for larynx. Considering the assumed proportion of cancers attributable to HPV (cervical: 100%; vulvar: 34.7%; vaginal 76.8%; anal: 84.2%; penile: 46.7%; oral cavity: 16.0%; oropharynx: 28.2%; larynx: 21.3%, in Europe), the overall hospitalization costs due to HPV-related cancers were estimated at c61.6 million in males and c70.8 million in females. CONCLUSIONS: The hospital burden of HPV-associated cancers in males is almost similar as in females, in France. This burden is probably underestimated since outpatient and indirect costs were not included. PCN53 ANALYSIS OF SOCIOECONOMIC BUREDN OF HEPATOCELLULAR CARCINOMA IN RUSSIA Omelyanovsky VV, Avksentieva MV, Krysanov I, Ivakhnenko O Research Center for Clinical and Economic Evaluation and Pharmacoeconomics, Moscow, Russia OBJECTIVES: In our study, we estimated social and economic burden of hepatocellular carcinoma (HCC) in Russian Federation resulting from HCC from the position of public health-care system. METHODS: All types of costs were calculated—direct 13th Euro Abstracts and indirect (the “cost of illness” analysis), connected with patient treatment at different stage of disease treatment. Epidemiology of ÍÑÑ in Russia was investigated: prevalence, incidence, structure of disease depending on age, disease progression, and death rate indicators. The expert opinion of real practice of HCC treatment was collected. All these methods allowed to estimate the direct and indirect costs of HCC. RESULTS: HCC incidence rate was 2.4 patients per 100,000 population (85% in the structure of primary liver cancer). HCC incidence rate index was 8658 patients as of 2008. In the HCC, structure intermediate stage prevails—61%, the terminal and local stages—30% and 9% correspondingly. HCC treatment costs were 2370 bln RUB (c67.7 mln) (2008). Direct medical costs were 2208 bln RUB (c63.1 mln) including inpatient care (90%), outpatient care (6.6%), and diagnostics (3.4%). Indirect costs were 0.161 bln RUB (c4.6 mln) including GDP loses (26%) and payment related with temporary disability (74%) The current HCC treatment standards do not correspond to international approaches. Very few patients with primary liver cancer get target pharmacotherapy. The analysis of actual practice of managing patients with HCC shows prevalence of drugs with no indications for usage from the point of view of the existing standards and recommendations (form 33% according to regional reimbursement to 58% according to Federal reimbursement). CONCLUSIONS: Developing of new standards of HCC treatment including target therapy can reduce the cost of illness by reducing off-label use and optimizing the treatment strategy. PCN54 HOSPITAL UNDERTAKING OF PATIENTS WITH A RESECTION OF LUNG TUMOR IN FRANCE Cancalon C, Schmidt A, Bénard S St[è]Ve Consultants, Lyon, France OBJECTIVES: With 30,651 new cases diagnosed in 2005, lung cancer is the fourth most frequent cancer in France and the first in terms of mortality, with 26,624 deaths per year. The survival rate at 5 years is less than 15%. The purpose of this study was to describe the 2 years hospital undertaking of patients with a resection of lung tumor and to estimate associated hospital costs. METHODS: The 2006 to 2008 PMSI French hospital databases were used. Patients with a resection linked to a lung cancer (ICD10 diagnoses: C33* and C34*) in 2006 were identified and followed up during 2 years. Hospital stays, chemotherapy, and radiotherapy sessions were extracted and associated costs (excluding expensive drugs) were assessed using DRG. Kaplan Meier method was applied to estimate associated costs over time, by taking into account survival probabilities. RESULTS: In 2006, 8798 patients were hospitalized for a resection of lung tumor of which 75.8% were men. Mean age at inclusion was 62.4 years; 2343 patients (26.6%) died in hospital during follow-up. The mean number of hospitalizations for repeated surgery was 0.11 per patient, 2.00 for radiotherapy sessions, 2.99 for chemotherapy sessions, and 2.57 for other hospitalizations. Total mean hospital cost per patient was estimated at c16,169.89 for the 2 years follow-up. First surgery account for 48%, repeated surgeries for 6%, radiotherapy session for 2%, chemotherapy sessions for 8%, and other hospitalizations for 37%. First month of first year supported half of the total cost, and first year bore 86% of it. CONCLUSIONS: In France, lung resections for cancer represent a heavy charge for hospitals. During the follow-up period, major burden is dedicated to hospital stays especially for the first months. These results could be relevant to estimate the impact of coming drugs which will be associated to resection of lung tumor. PCN55 COST OF TREATMENT OF BREAST CANCER IN RUSSIA Yagudina R, Kulikov AU, Nguyen TTT Moscow Medical Academy named after I.M.Sechenov, Moscow, Russia OBJECTIVES: Estimating the amount and the structure of annual expenses for treatment of breast cancer (BC) at different stages in Russia. METHODS: Indentifying the annual direct costs of treatment BC based on the Russian standard of treatment confirmed by Minister of Health of Russian Federation. Direct costs of BC include cost of medical services (MS) (hospitalizations, physician services, diagnosis, surgery, and other supporting services), and cost of pharmacotherapy (PT) (cytostatics, hormones and antihormones, accompanying drugs, and other drugs). RESULTS: Annual direct costs of treatment BC totaled 138,680, 1,379,980, and 1,923,050 roubles at stages I–II, III, IV accordingly. With increase in weight of disease, the relative share of expenses for MS decreases and for PT increases. Analyzing the structure of costs at different stages showed that the cost of PT is substantial, especially at metastatic stage. The presence and relative share of components vary at different stages in structure of expenses for MS, in which expenses for radiotherapy share about 60% and 68% at stage III and IV, accordingly. In structure of expenses for PT, the expenses for cytostatics are substantial and take 85% at both stages III and IV. Analyzing the gain of expenses for separate groups of drugs from stage III to IV showed that expenses for cytostatics and accompanying drugs have grown by 43% and 128% accordingly, while expenses for other groups have almost not changed. CONCLUSIONS: Expenses for treatment BC at different stages vary in size and structure, and rises with increase in weight of disease. With the amount of 51,865 patients with BC in 2007 (62.3%, 26.3%, 10.5% at stages I–II, III, IV accordingly), we received total expenses for treatment of 33,777.217.655 roubles ($1.125.907.255), in which expenses for treatment of stage I–II, III, IV are 4.481.031.599 roubles ($149,367,120), 18,823,610,290 roubles ($627,453,676), and 10,472,593,166 roubles ($349,860,459), accordingly. A261 PCN56 CLINICO-ECONOMIC ANALYSIS OF TREATMENT OF CHILDREN WITH RHABDOMYOSARCOMAS ON PROTOCOLS CWS Teplykh E1, Zhukovskay E2 1 Children’s Oncohematology Center, Chelyabinsk, Russia; 2State Medical Academy of Chelyabinsk, Chelyabinsk, Russia OBJECTIVES: To carry out clinico-economic analysis of treatment of children with rhabdomyosarcomas on protocols CWS in Chelyabinsk Region, to estimate medical, social, and economic efficiency of technology. METHODS: We reviewed 21 patients <16 years old with rhabdomyosarcomas. All patients received therapy on protocols CWS. We used following methods of clinico-economic analysis: cost of illness, costeffectiveness analysis, and analysis of the kept years of a life. RESULTS: The sum of direct medical expenses for one patient has made US$16,904 dollars, and the total sum of direct medical expenses for treatment of 21 patients included in research, has made US$354,984. The overall survival rate was 52% that testifies to high medical efficiency of applied technology. For all groups of patients, the treatment keeps 694.4 years of a life that makes 33 years of a life on one patient.The parity of a cost/efficiency for one patient has made US$512 dollars for 1 year of the kept life. Considering that in 2008 gross national product has made US$9.5 thousand per capita, an expense for 1 year of the kept life in 18.5 times there is less than parameter of gross national product, means that expenses are justified. Before achievement of a pension age, the survived patients will work 422 years and will make gross national product for the sum US$4009 thousand. The survived patients provide the state with the income 11.2 times the society for their treatment. CONCLUSIONS: The clinico-economic analysis has shown high medical, economic, and social efficiency of technology of treatment of children with rhabdomyosarcomas on protocols CWS. PCN57 EXAMINING PATIENT-BASED COSTS FOR IRINOTECAN CHEMOTHERAPY: UK PRACTICE-BASED MICROCOSTING STUDY Shabaruddin FH1, Elliott RA2, Payne K1 1 Health Sciences—Economics, The University of Manchester, Manchester, UK; 2University of Nottingham, Nottingham, UK OBJECTIVES: To conduct a robust economic evaluation, it is necessary to describe current practice and associated costs. Available data on clinical pathways and cost of chemotherapy are clinical trial-based, which may not reflect UK National Health Service (NHS) practice. Practice-relevant costs of drug administration, patient monitoring and management of adverse events, required for a practice-relevant economic model, are not available. This study aimed to inform an economic evaluation by describing patient-based cost of NHS patients with advanced colorectal cancer (CRC) undergoing irinotecan-based chemotherapy. METHODS: Resource use data were collected from the medical records of 48 patients prescribed irinotecan-based (IrMdG) chemotherapy at a UK tertiary care center. Using the hospital perspective, data were collected from starting chemotherapy until treatment ended. Unit costs were assigned, based primarily on NHS Reference Costs 2008/09. Data were analyzed using descriptive statistics and variations around the costs were obtained. Predictors of cost were identified from a stepwise multiple regression analysis (ordinary least squares). RESULTS: Total cost for 48 patients was £598,765.54 (UK £ 2008/09). Mean cost per patient was £12,474.28 (95% CI: £11,233.24–£13,715.32, median £13,307.82, range £3,024.48–£21,276.18). Cost components comprised: chemotherapy drugs (36.9%), chemotherapy delivery (21.4%), pharmacy cost (15.0%), oncology appointments (9.5%), central line insertion (5.0%), management of complications and comorbidities (5.1%), management of adverse events (4.9%), and imaging (2.2%). Significant predictors of increased cost (P < 0.05) identified from the stepwise regression were: number of chemotherapy cycles received (adjusted R2 0.81), neutropaenia (adjusted R2 0.83), no prior chemotherapy (adjusted R2 0.85), and full-dose chemotherapy (adjusted R2 0.86). CONCLUSIONS: This study provides the first data describing patient-based costs associated with current NHS practice in this patient group, derived from a pragmatic observational study with no trial protocol dictating practice. These data should be used in ensuing economic evaluations to ensure relevance to current clinical practice. PCN58 SURVEY AND ANALYSIS OF THE COSTS OF METASTATIC COLORECTAL CANCER TREATMENT IN BULGARIA Rutkowski J1, Haldas M1, Jedynasty K2 1 HTA Consulting, Krakow, Poland; 2Amgen GmbH, Headquarters Office for CEE, Vienna, Austria OBJECTIVES: To describe chemotherapy regimens used in the first, second, third, and fourth line of treatment in patients with metastatic colorectal cancer. Costs of chemotherapy regimens used as well as supportive care and medical procedures in Bulgaria will be estimated (as part of a multinational central European study). METHODS: This opinion-based study collected required data by online questionnaire. All information concerning treatment of colorectal cancer was based on experts opinion from four oncology centers in Bulgaria. Oncologists had access to medical records of approximately 1220 patients treated in year 2008. RESULTS: The leading first line regimen (60% of patients) was FOLFOX 4 (oxaliplatin, calcium folinate, and fluorouracil). The most commonly prescribed second-line regimen (50%) was FOLFIRI (irinotecan, calcium folinate, and fluorouracil). Capecitabine was the most popular in both the third- (26%) and fourth-line (4%) settings. The percentage of patients receiving supportive care increased with disease stage, from 1% in the first, A262 21% in the second, 53% in the third, and 94% in the fourth line. The most common treatment algorithm (18%) was FOLFOX, FOLFIRI and supportive care in the first-, second-, and third-lines, respectively. Mean regimen costs per patient were estimated from a public payer perspective. The most expensive first, second, third and fourth-line regimens were FOLFOX 4 (c14,200), FOLFIRI + bevacizumab (c7912c), cetuximab + irinotecan (c7237) and capecitabine (c2609), respectively. CONCLUSIONS: The most common regimen in the first line was also the most expensive one. New chemotherapeutic agents are associated with improvements in survival time but also with substantial costs. Factors influencing the selection of chemotherapy included: previous therapies, course of the disease, the patient’s performance status, adverse events after previous chemotherapies, and concomitant diseases. However, open-ended coverage policies for new chemotherapeutic agents may prove difficult to sustain as costs continue to rise. PCN59 COST OF MANAGEMENT OF BREAST CANCER WITH BRAIN METASTASES USING FRENCH HOSPITAL PATIENT CHAINING SYSTEM Benjamin L1, Cotté FE1, Vainchtock A2, Mercier F3, Vidal-Trécan G4, Durand-Zaleski I5 1 GlaxoSmithKline, Health Outcome Studies, Marly le Roi, France; 2HEVA, Lyon, France; 3 StatProcess, Port-Mort, France; 4Hôpital Cochin, Unité de Santé publique, Paris, France; 5 Hôpital Henri Mondor, Service de Santé publique, Créteil, France OBJECTIVES: Breast cancer (BC) is the second neoplasm which disseminates brain metastases (BM). We estimated the incidence of patients with BCBM, those overexpressing HER2 (HER2+), and the costs related to their management using the new chaining system of the French hospital information program (PMSI). This was to identify predictors of costs. METHODS: A retrospective analysis using the PMSI database (2006–2008) was conducted to estimate the number of public and private stays related to the diagnosis of BCBM. Stays were extracted and chained to patients’ identification number to calculate the number of patients concerned by BCBM. The administration of trastuzumab was used as a surrogate for the HER2 status. Costs were estimated from the health insurance perspective including health-related group tariff, supplements for intensive care, number and length of stays, and expensive-drugs status. Spearman’s rank correlation coefficient and nonparametric test (Kruskal– Wallis) were used for univariate analyses. RESULTS: In 2008, 3610 women were hospitalized for BCBM (vs. 3273 and 3523 in 2006 and 2007, respectively) of whom average age was 56.1 years (SD: 13.3). Patients had an average of 4.8 hospitalizations (SD: 6.4) mostly for palliative care (42%), chemotherapy (39%), and radiotherapy (14%). Twenty-one percent of patients suffered from BM only and 79% had multiple metastases; 16% were identified as HER2+. Annual mean cost of care was c8049 per patient with BCBM compared with c19,412 specifically for HER2+ patients (respectively, 11% and 59% were dedicated to expensive drugs). Age (P < 0.001), patients with newly diagnosed BCBM (P < 0.001), and the number of metastases (P < 0.001) were associated with the cost of BCBM. CONCLUSIONS: Incidence of BCBM seemed to increase during 2006–2008. BCBM management appeared resource-consuming especially for HER2+ patients. The development of chaining system in PMSI database is an opportunity to estimate economic data accurately as well as to generate epidemiological data from an exhaustive database. PCN60 CLINICAL AND ECONOMIC OUTCOMES ASSOCIATED WITH ADJUVANT CHEMOTHERAPY IN ELDERLY PATIENTS WITH EARLY STAGE OPERABLE BREAST CANCER Sail K, Franzini L, Du XL, Lairson D University of Texas School of Public Health, Houston, TX, USA OBJECTIVES: Chemotherapy is a major factor contributing to the economic burden associated with breast cancer in the elderly. However, there are no clear recommendations for adjuvant chemotherapy use in elderly women aged 70 and above due to lack of efficacy data in that age group. The study objective was to examine the clinical and economic outcomes associated with adjuvant chemotherapy in elderly patients aged 65 and above with early stage operable breast cancer. METHODS: We studied a cohort of 23,110 node positive and 31,572 node negative women aged 65 and over diagnosed with incident American Joint Committee on Cancer (AJCC) stage I, II, or IIIa breast cancer between January 1, 1991 and December 31, 2002 using SEERMedicare data. Total treatment and chemotherapy costs were estimated from the Medicare payments using the phase of care approach. Cox proportional hazard ratio of mortality was used to determine the effectiveness of adjuvant chemotherapy after adjusting for selected patient and tumor characteristics. a propensity score analysis was also employed to minimize the bias associated with the receipt of adjuvant chemotherapy. RESULTS: The difference in the total unadjusted costs for patients who received chemotherapy in contrast with patients not receiving any chemotherapy was $16,795 in node positive patients and $11,882 in node negative patients. Regression adjusted cost estimates for all node positive patients receiving chemotherapy was approximately $6500 and was significantly higher (P < 0.05) than for patients not receiving chemotherapy. Mortality was significantly reduced in node positive women aged 65–69 who received adjuvant chemotherapy compared to those who did not receive chemotherapy (HR, 0.66; CI, 0.58–0.74) and in patients aged 70–74 (HR,0.66; CI, 0.59–0.74), after adjustment for factors that may affect survival. CONCLUSIONS: Decision makers can use cost and effectiveness estimates from this study to assess relative value of chemotherapy in different age groups. 13th Euro Abstracts PCN61 COST IMPACT OF ORAL CAPECITABINE COMPARED TO 5-FLUOROURACIL FOR TREATMENT OF PATIENTS WITH METASTATIC COLORECTAL CANCER Citarella A, Cammarota S, Riegler S, Putignano D, Menditto E CIRFF, Federico II University, Naples, Italy OBJECTIVES: To evaluate the cost of biweekly oxaliplatin plus oral capecitabine (OXXEL) versus oxaliplatin combined with leucovorin-modulated 5-fluorouracil (5-FU) given as i.v. bolus every 2 weeks (OXAFAFU) in patients with metastatic colorectal cancer (MCRC) in Italy. METHODS: We conducted a multicenter, retrospective longitudinal treatment-cost analysis. Direct medical costs attributable to MCRC were quantified using 2008 prices and tariffs. The analysis was applied to a time horizon of 6 months. The study was conducted from the perspective of the National Healthcare Service (NHS). RESULTS: A total of 322 patients (59.9% males; mean age 65.2 ± 9.4 years) were analyzed. Mean total cost per patient over follow-up period was estimated at c5242.18 ± 2542.06 and c6732.80 ± 3423.72 in the Capecitabine and 5-FU arms respectively (P < 0.0001). CONCLUSIONS: The study estimated that oral capecitabine administration would produce a saving of c1490.62 to the NHS. The differences in cost between the two arms are determined by the administration route (i.v. vs. oral administration). Therefore, the important economic and practical advantage of capecitabine oral home-based therapy is the reduced number of hospital visit and the relative costs. Avoiding the hospital access fees reduces the impact of higher acquisition cost of capecitabine. Moreover, capecitabine in comparison to the 5-FU regimen was associated with lower complication. Therefore, oral capecitabine may represent a valid alternative in the management of metastatic colorectal cancer. PCN62 SURVEY AND ANALYSIS OF THE COSTS OF METASTATIC COLORECTAL CANCER TREATMENT IN SLOVAKIA Rutkowski J1, Haldas M1, Salek T2, Jedynasty K3 1 HTA Consulting, Krakow, Poland; 2National Cancer Institute, Bratislava, Slovak Republic; 3 Amgen GmbH, Headquarters Office for CEE, Vienna, Austria OBJECTIVES: To describe chemotherapy regimens used in the first-, second-, and third-line treatment of patients with metastatic colorectal cancer and to estimate costs of regimens, supportive care, and medical procedures in Slovakia (part of a multinational study in central Europe). METHODS: In this opinion-based study, data were collected by online questionnaire. All information concerning treatment of colorectal cancer was based on expert opinion at three oncology centers in Slovakia. Oncologists had access to medical records of approximately 1600 patients treated in year 2008. RESULTS: The most commonly used first-line regimen (27% of patients) was IFL (irinotecan, 5-FU, leucovorin) + bevacizumab. The most commonly prescribed secondline regimen (27%) was cetuximab + irinotecan and capecitabine was the most common third-line regimen (15%). None of patients received supportive care in the first line. The percentages of patients receiving supportive care in the second and third lines were 27% and 45%, respectively. The most common treatment strategy (8%) was first-line capecitabine and supportive care in the second line. Mean regimen costs per patient were estimated from a public payer perspective. The most expensive firstline regimen was IFL + bevacizumab (c36,369). In the second and third lines, modified FOLFOX 4 (oxaliplatin, 5-FU, leucovorin) was the most expensive regimen at c31,318 and c31,572, respectively. CONCLUSIONS: More than 50% of patients received an active treatment until the second line. The most common regimen in the first line was also the most expensive one. New chemotherapeutic agents are associated with improvements in survival time but also with substantial costs. Factors influencing the selection of chemotherapy included: previous therapies, course of the disease, the patient’s performance status, adverse events after previous chemotherapies and concomitant diseases. However, open-ended coverage policies for new chemotherapeutic agents may prove difficult to sustain as costs continue to rise. PCN63 SURVEY AND ANALYSIS OF THE COSTS OF METASTATIC COLORECTAL CANCER TREATMENT IN SERBIA Rutkowski J1, Haldas M1, Jedynasty K2 1 HTA Consulting, Krakow, Poland; 2Amgen GmbH, Headquarters Office for CEE, Vienna, Austria OBJECTIVES: To describe chemotherapy regimens used and to estimate costs of chemotherapy regimens, supportive care, and medical procedures in the first-, second-, third-, and fourth-line treatment of patients with metastatic colorectal cancer in Serbia (part of a multinational study in central Europe). METHODS: An online questionnaire was used to collect necessary information in this opinion-based study. All information concerning treatment of colorectal cancer was based on experts opinion from four oncology centers in Serbia. Oncologists had access to medical records of approximately 1760 patients treated in year 2008. RESULTS: The leading first-line regimen (38% of patients) was Mayo (Folinic acid, 5-FU). The most commonly prescribed second-line regimen (46%%) was FOLFOX 4. Modified FOLFIRI (irinotecan, folinic acid, 5-FU) was the most popular regimen in the third line (35%), while FOLFIRI/ cetuximab (35%) was the most commonly used fourth-line regimen. The percentage of patients receiving supportive care was 7%, 5%, 10%, and 56% in the first, second, third, and fourth lines, respectively. The most common treatment path (8%) was FOLFOX 4 B in the first line, FOLFIRI B in the second, and Mitomicin mono in the third. Mean regimen costs per patient were estimated from a public payer perspective. The most expensive regimen in the first line was bevacizumab + capecitabine + oxali- 13th Euro Abstracts platin (c21,126), the same regimen was also the most expensive in the second line (c13,366), irinotecan/cetuximab in the third line (c25,633), and FOLFIRI/cetuximab in the fourth line (c6479). CONCLUSIONS: New chemotherapeutic agents are associated with improvements in survival time but also with substantial costs. Factors influencing the selection of chemotherapy included: previous therapies, course of the disease, the patient’s performance status, adverse events after previous chemotherapies, and concomitant diseases. However, open-ended coverage policies for new chemotherapeutic agents may prove difficult to sustain as costs continue to rise. PCN64 SURVEY AND ANALYSIS OF THE COST OF METASTATIC COLORECTAL CANCER TREATMENT IN SLOVENIA Rutkowski J1, Haldas M1, Jedynasty K2, Ocvirk J3 1 HTA Consulting, Krakow, Poland; 2Amgen GmbH, Headquarters Office for CEE, Vienna, Austria; 3Onkoloski Institut Ljubljana, Ljubljana, Slovenia OBJECTIVES: To describe chemotherapy regimens used in the first-, second-, third-, and fourth-line treatment of patients with metastatic colorectal cancer and to estimate costs of chemotherapy regimens, supportive care, and medical procedures in Slovenia (part of a multinational study in central Europe). METHODS: In this opinion-based study, necessary data were collected by online questionnaire. All information concerning treatment of colorectal cancer was based on experts opinion from an oncology center in Slovenia. Oncologist had access to medical records of approximately 800 patients treated in year 2008. RESULTS: The most commonly used regimen in the first line (30% of patients) was XELIRI (capecitabine, irinotecan) + bevacizumab. The most commonly prescribed regimen in the second (28%) and third lines (12%) was XELOX (capecitabine, oxaliplatin) + bevacizumab. XELIRI + bevacizumab was most commonly used in the fourth line (6%). Supportive care was not used in the first line with 9%, 55%, and 76% receiving it in the second, third, and fourth lines, respectively. The most common treatment algorithm (21% of patients) was first-line XELIRI + bevacizumab and second-line XELOX + bevacizumab. Mean regimen costs per patient were estimated from a public payer perspective. FOLFOX + cetuximab was the most expensive regimen in all lines. Costs of this regimen were c35,896 in the first line and c36,179 in the second, third, and fourth lines. CONCLUSIONS: More than 50% of patients received an active treatment until the second line. Costs of treatment vary between lines. New chemotherapeutic agents are associated with improvements in survival time but also with substantial costs. Factors influencing the selection of chemotherapy included: previous therapies, course of the disease, the patient’s performance status, adverse events after previous chemotherapies, and concomitant diseases. However, open-ended coverage policies for new chemotherapeutic agents may prove difficult to sustain as costs continue to rise. PCN65 COST-EFFECTIVENESS ANALYSIS OF A FOBT-BASED COLORECTAL CANCER SCREENING PROGRAMME Pizzo E1, Bracci E2, Vagnoni E2, Wilschut J3, van Ballegooijen M3 1 Imperial College London, London, UK; 2University of Ferrara, Ferrara, Italy; 3Erasmus University Medical Center, Rotterdam, The Netherlands OBJECTIVES: Colorectal cancer (CRC) is one of the most common forms of cancer in western countries and represents the second leading cause of cancer mortality in Europe (AIRTUM 2009). Early detection and removal of cancerous lesions can reduce CRC and mortality and improve patients’ quality of life (Taupin et al. 2006). The main literature on this topic refers to the United States and few studies have been conducted in Italy to date (Zappa et al. 1997; Tappenden 2007). Aim of the paper is to shed some light on the effectiveness and costs of screening programs in the Italian health-care system, presenting the results of a cost-effectiveness analysis of a CRC screening program in Italy. METHODS: We use as case study a Regional CRC screening program to determine the full costs and the effectiveness of the adopted techniques, FOBT combined with colonoscopy.The costs involved in each phase of the program are evaluated using a microcosting analysis. Effectiveness is valued in terms of early detected lesions and years of life gained. Cost and effectiveness data are used to estimate the costs for year of life gained, using a MISCAN-COLON Model© to simulate and compare two alternative scenarios, with or without the screening program. RESULTS: The preliminary results show that the screening will prevent almost 2.0 deaths (11.2%) per 1000 screened individuals, corresponding to 19.4 years of life gained in 30 years with an incremental cost-effectiveness ratio of c2.400 for life-year gained. CONCLUSIONS: The results outpace those of previous studies (Sonnenberg 2000), signaling an increasing effectiveness of CRC screening program. Besides, the paper highlights the importance of implementing a screening not only for the effects that prevention can have in clinical terms, but also for the economic impact of such a policy in relation to the long-term sustainability of health-care systems. PCN66 ECONOMIC ANALYSIS OF CAPECITABINE PLUS OXALIPLATIN (XELOX) VERSUS FLUOROURACIL/LEUCOVORIN PLUS OXALIPLATIN (FOLFOX) IN THE TREATMENT OF ADVANCED COLON-RECTUM CANCER IN CHINA Chen W Fudan University, Shanghai, China OBJECTIVES: The objective of the study was to examine the direct medical cost of XELOX (capecitabine plus oxaliplatin) compared to FOLFOX (fluorouracil/leucovorin plus oxaliplatin) for the treatment of advanced colon-rectum cancer in China. A263 METHODS: Since the equal efficacy was already demonstrated by the published literature and local clinical guideline, cost minimization analysis was performed to compare the direct medical costs of XELOX and FOLFOX for the treatment of advanced colon-rectum cancer. The direct medical costs were associated with the drug costs, drug administration costs, hospitalization costs, and adverse events management costs. The costs were calculated based on a questionnaire survey from an expert panel of 23 pharmacists and 10 gastrointestinal surgeons and medical oncologists. RESULTS: According to the recommendation of expert panel, the standard treatment duration of XELOX and FOLFOX was eight cycles and 12 cycles, respectively. The drug cost of XELOX regimen was CNY 47,306 (US$6926), higher than FOLFOX by CNY 22,118 (US$3238). However, the cost increment of XELOX regimen was offset by the higher drug administration cost (deviation CNY 6,820), hospitalization cost (deviation CNY 10,200), and adverse events management cost (deviation CNY 7,710) of FOLFOX regimen. As a result, XELOX showed a significant overall cost savings of CNY 2612 (US$382) compared with FOLFOX. CONCLUSIONS: According to the study, XELOX is cost saving in comparison with FOLFOX for the treatment of advanced colon-rectum cancer in China, especially in the chemotherapy administration and hospitalization utilization. PCN67 MANAGEMENT OF MALIGNANT ASCITES IN GERMANY—TREATMENT PATTERNS, RESOURCE CONSUMPTION, AND COSTS Ehlken B1, Berger K1, Shlaen R1, Gonschior AK2, Lordick F3 1 IMS Health, Munich, Germany; 2Fresenius Biotech GmbH, Munich, Germany; 3Medizinische Klinik III, Klinikum Braunschweig, Braunschweig, Germany OBJECTIVES: To describe treatment patterns, resource use, and associated costs for cancer patients with malignant ascites (MA) receiving paracentesis in Germany. METHODS: The study was conducted as an observational, multicenter, prevalencebased cohort study. Inclusion criteria were: age ≥18 years, diagnosis of ovarian or gastrointestinal carcinoma (CA) with MA, paracentesis as treatment option for MA at the time of enrolment. Resource consumption data were collected by chart review and patient questionnaire covering the time period from the first paracentesis documented in the study until ascites diagnosis retrospectively and subsequent paracenteses prospectively. Direct medical costs were analyzed from third-party payers’ (TPP) and patients’ perspective. RESULTS: A total of 29 patients (38% male) with a mean age of 65 ± 9 years were enrolled at 11 centers (six hospitals, five office-based practices) between July 2008 and August 2009. Seven patients had ovarian CA (24%), 5 gastric CA (17%), and 17 other gastrointestinal CAs (59%). a total of 101 paracenteses were documented for all patients. From ascites diagnosis to death patients received on average 4.6 paracenteses. Mean time between two paracenteses was 11.9 ± 12.3 days. Data from 42 paracenteses were eligible for resource and cost analysis. Diuretics were applied in 57% of paracentesis units and human albumin in 29%. Intraperitoneal chemotherapy was applied rarely (5%). From TPPs’ perspective, mean total costs per paracentesis unit amounted to c1064 ± 1453, from patients’ perspective c17 ± 46. Direct medical costs per paracentesis unit varied from c671 ± 1070 at office-based practices to c2.742 ± 1.535 at hospitals (inpatient treatment). CONCLUSIONS: This is the first comprehensive study evaluating the burden of MA in cancer patients undergoing paracentesis in Germany. Our findings indicate that the costs for paracentesis range around c670 to c2700 depending on health-care setting. Our results might serve as a basis for further research on the economic implication of malignant ascites. PCN68 COST-EFFECTIVENESS OF PRIMARY PROPHYLAXIS WITH PEGFILGRASTIM VERSUS FILGRASTIM FOR THE PREVENTION OF FEBRILE NEUTROPENIA IN NON-HODGKIN LYMPHOMA AND STAGE II BREAST CANCER PATIENTS IN GERMANY Taylor DC1, Ozer-Deniz S1, Hill G1, Skornicki M1, Danel A2, Kunz E3 1 i3 Innovus, Medford, MA, USA; 2Amgen (Europe) GmbH, Zug, Switzerland; 3Amgen (Europe) GmbH, München, Germany OBJECTIVES: To assess the cost-effectiveness in Germany of primary prophylaxis (PP) with pegfilgrastim versus 6- or 11-day filgrastim (F6, F11) in the prevention of febrile neutropenia (FN) in non-Hodgkin lymphoma (NHL) patients receiving CHOP-14 chemotherapy and in breast cancer (BC) patients receiving TAC chemotherapy. METHODS: A payer perspective Markov model of febrile neutropenia prophylaxis in chemotherapy patients was developed. PP was defined as initiating prophylaxis with the first chemotherapy cycle. Model cycle length matches chemotherapy cycle length (CHOP-14:14 days, TAC: 21 days); model time horizon is the duration of chemotherapy (6 cycles). Cycle 1 FN risk with no prophylaxis was estimated to be 21% for NHL CHOP-14 and 14% for BC TAC; all cycle relative risks of FN versus no prophylaxis for PP using Pegfilgrastim, F6, and F11 were 0.25, 0.87 and 0.61, respectively, based on published literature and meta-analyses. Pegfilgrastim cost was estimated as c1686 per chemotherapy cycle; corresponding costs for F6 and F11 were c1347 and c2469 based on German national pricing. Incremental costeffectiveness ratios (ICERs) were calculated per FN events avoided. Costs and outcomes were discounted (3%/year). Sensitivity analyses were performed. RESULTS: For NHL FN events per patient were 0.15, 0.76, and 0.47 for Pegfilgrastim, F6, and F11, respectively. ICER for Pegfilgrastim versus F6 was c1386 per FN avoided. For BC, corresponding FN events per patient were 0.09, 0.43, and 0.27. The ICER for Pegfilgrastim versus F6 was c6651 per FN avoided. Pegfilgrastim was dominant (less costly, more effective) compared with F11 in both populations. Results were most sensitive to baseline risk of FN, cost of prophylaxis and cost of FN events. CONCLUSIONS: Primary prophylaxis with pegfilgrastim costs <c1400 per additional FN A264 avoided compared with 6-day filgrastim for NHL patients treated with CHOP-14, and <c7000 for BC patients treated with TAC. Pegfilgrastim dominated 11-day filgrastim. PCN69 COST-EFFECTIVENESS ANALYSIS OF DASATINIB 100 MG VS. IMATINIB 800 MG IN PATIENTS WITH IMATINIB-RESISTANT CHRONIC MYELOID LEUKEMIA IN SPAIN Ramirez de Arellano A1, Sbarigia U2, Taylor MJ3, Martin P1, Restovic G4 1 Bristol-Myers Squibb Iberia, Madrid, Spain; 2Bristol-Myers Squibb, Braine-l’Alleud, Belgium; 3 University of York, York, UK; 4BCN Health, Barcelona, Spain OBJECTIVES: To asses the cost-effectiveness relationship of dasatinib in comparison to high dose of imatinib in the treatment of CML in patients with imatinib-resistant chronic myeloid leukemia in Spain. METHODS: A Markov model was developed to estimate, in the long term, the costs and clinical outcomes (life-years gained and quality-adjusted life-years gained) of dasatinib compared to imatinib in imatinibresistant patients. Four health states were considered in the analysis: Chronic phase; Accelerated phase; Blast phase; and Death. Cycle length is on a monthly basis and health effects and costs were counted until all patients reached the “death” health state. The efficacy outcomes are estimated from a direct comparison derived from the clinical trial BMS 017. The health-care resource use has been set up by a Spanish clinical expert and direct costs are in euros (2009). The perspective used is the Spanish Health System. Both costs and effects were discounted annually at 3.5%. The robustness of the results was tested in deterministic sensitivity analyses. RESULTS: In the base-case scenario, treatment with dasatinib proves to be a dominant option with a lower total cost and a higher level of effectiveness (potential cost saving of c56,995 and 0.19 QALY gained). The sensitivity analysis indicates that dasatinib remains as a dominant alternative in front of changes in the most relevant variables: costs, utility values, age at the start of the treatment, time horizon, and discount rate. CONCLUSIONS: Compared to imatinib, dasatinib shows a slower disease progression with relatively lower direct medical costs. Dasatinib can be regarded as a dominant treatment option in patients with imatinib-resistant CML in the Spanish Health System. PCN70 COST-EFFECTIVENESS ANALYSIS OF RITUXIMAB THERAPY IN PATIENTS WITH PREVIOUSLY UNTREATED CHRONIC LYMPHOCYTIC LEUKEMIA (CLL) IN BRAZIL Chiattone CS1, Borges LG2, Santos E2 1 Faculdade de Ciências Médicas da Santa Casa de São Paulo, São Paulo, SP, Brazil; 2Roche Brazil, São Paulo, SP, Brazil BACKGROUND: Chronic lymphocytic leukemia (CLL) is the most common of adult leukemias, comprising about 30–40% of all cases (Watson 2008). Incidence of CLL varies worldwide, being 4.3/100,000 in Caucasian men and increases with age, with median age at diagnosis of 64–70 years (Yamamoto 2005). In a phase III trial (CLL-8), the combined immunochemotherapy R-FC (rituximab plus fludarabine and cyclophosphamide) showed longer progression-free and overall survival, higher complete response rate, and longer duration of response than FC alone in previously untreated CLL patients. OBJECTIVES: To assess the incremental cost-utility ratio for R-FC versus FC alone in untreated CLL patients under the public payer perspective in Brazil. METHODS: The cost-effectiveness analysis was based on the pivotal study ML17102 (CLL-8). a Markov model was developed consisting of three health states: “Progression-Free Survival” (PFS), “Progression,” and “Death”. The model cycle length is monthly and the time horizon of the analysis is 15 years. Costing was based on public sources. Only direct costs were considered in the calculation, including costs for treating severe adverse events and further treatment patterns. Costs were reported in 2010 (US$1.00~$Brz1.8) Brazilian Reais and discounted at a 5% rate according to local guidelines for economic evaluation (Vianna 2007). RESULTS: R-FC combined therapy resulted in a gain of 1.031 life-years (Lys) (5.611 vs. 4.579) at an incremental cost of $Brz44,780. The ICER of R-FC versus FC is, therefore, estimated to be $Brz43,414 per LY gained. The probability of R-FC being cost-effective is 98.84%, considering a willingness-to-pay of R$100,000. CONCLUSIONS: In untreated CLL patients, R-FC therapy improves overall survival and progression-free survival compared with FC alone. Results suggest that F-CR combined therapy is a cost-effective intervention for the Brazilian Public Healthcare System. PCN71 A COMPARISON OF THE COST-EFFECTIVENESS OF ZOLEDRONIC ACID FOR PREVENTING SKELETAL-RELATED EVENTS IN PATIENTS WITH BONE METASTASES FROM PROSTATE CANCER IN 4 EUROPEAN COUNTRIES Botteman M1, Carter J1, Kaura S2 1 Pharmerit North America, LLC, Bethesda, MD, USA; 2Novartis Pharmaceuticals Corporation, Florham Park, NJ, USA OBJECTIVES: Zoledronic acid (ZOL) is the only bisphosphonate indicated for preventing skeletal-related events (SREs) in patients with bone metastases from prostate cancer (PC). We estimated and compared the cost-effectiveness of ZOL versus placebo for this indication in France, Germany, Portugal, and the The Netherlands. METHODS: Incremental costs and quality-adjusted life-years (QALYs) associated with ZOL and placebo were estimated using a literature-based decision analytic model using data from a 15-month randomized trial comparing ZOL (4 mg monthly; n = 214) with placebo (n = 208). The model included assumptions about SREs, mortality, drug and administration 13th Euro Abstracts costs, SRE costs, quality of life, and therapy duration. SRE costs were estimated using Diagnosis Related Group tariff information (supplemented with published literature) in France and Germany, and published retrospective medical record review cost analyses in Portugal and the The Netherlands. RESULTS: Over 15 months, the cumulative projected SREs were 0.83 for ZOL and 1.66 for placebo. ZOL reduced SRE costs by c2659 to c4005, depending on the country. SRE cost savings were greatest in the The Netherlands, followed by Portugal, Germany, and France. ZOL reduced total costs (including drug costs) by c62 in Portugal and c301 in the The Netherlands, but increased costs by c562 in Germany and c1022 in France versus placebo. ZOL increased qualityadjusted survival by 0.03566 QALY per patient, with an incremental cost per QALY gained versus placebo of c15,770 in Germany and c28,648 in France. In all countries, the cost-effectiveness ratio for ZOL was favorable and substantially below the internationally accepted c50,000/QALY threshold. Costs and QALYs were saved with ZOL in Portugal and the The Netherlands. CONCLUSIONS: In patients with bone metastasis from PC, ZOL is economically attractive. The cost-effectiveness ratio for ZOL is below standard cost-effectiveness thresholds used by most health-care systems. In Portugal and the The Netherlands, ZOL is cost-saving versus placebo. PCN72 COST-EFFECTIVENESS OF LETROZOLE AND OF ANASTROZOLE VERSUS TAMOXIFEN AS ADJUVANT THERAPY IN POSTMENOPAUSAL WOMEN WITH EARLY BREAST CANCER USING UPDATED SURVIVAL DATA FROM THE BIG 1–98 AND ATAC TRIALS: A UK PERSPECTIVE Karnon J1, Kaura S2 1 University of Sheffield, Sheffield, UK; 2Novartis Pharmaceuticals Corporation, Florham Park, NJ, USA OBJECTIVES: The 74-month follow-up of the BIG 1–98 trial reported improved overall survival (OS) for 5 years letrozole (LET) versus tamoxifen (TAM; hazard ratio 0.83; 95% CI 0.71, 0.97; P < 0.05). The 100-month follow-up of the ATAC trial did not show a significant difference in OS for anastrozole (ANA) versus TAM. Using reported differences in OS, we estimated the incremental cost per quality-adjusted life-year (QALY) gained for 5 years LET or ANA versus 5 years TAM in postmenopausal women with endocrine-responsive breast cancer (ERBC), from a UK NHS perspective. METHODS: Annual survival probabilities postoperatively were extracted from BIG 1–98 and ATAC results. Survival was extrapolated to 20 years using data reported by the EBCTG for women receiving 5 years TAM. Conservatively, equivalent annual survival probabilities were assumed for TAM, LET, and ANA groups after follow-up. Published adverse event (AE) costs and 5-year costs for locoregional recurrence (LR) and metastases were applied. Published utility weights for disease-free survival with AEs, LR, and metastases were used. All costs and health benefits were discounted at 3.5% annually. RESULTS: Over a 20-year period, the discounted additional treatment costs are £3618 for LET and £3736 for ANA. When accounting for AEs and reduced BC recurrence, the total cost difference between LET and TAM is £2964, and between ANA and TAM is £2929. The model estimated a difference in discounted QALYs of 0.297 for LET versus TAM, with an incremental cost per QALY gained of £9999. The incremental cost per QALY gained for ANA versus TAM is £46,829. CONCLUSIONS: Using updated OS data, economic analysis of 5 years LET or ANA versus 5 years TAM in postmenopausal women with ERBC suggests that LET is substantially more cost-effective than ANA. The current analysis is consistent with earlier extrapolations based on differences in time to recurrence. PCN73 COST-EFFECTIVENESS ANALYSIS OF SUPERIOR HYPOGASTRIC PLEXUS INHIBITION IN CANCER PATIENTS WITH VISCERAL PAIN IN LOWER ABDOMEN Contreras-Bello R, González-Buendía NI, Guajardo-Rosas J, Cerezo O, Plancarte-Sánchez R Oncology National Institute, Mexico City, D.F., Mexico OBJECTIVES: The aim of a neurolytic sympathetic blockade is to reduce consumption or the side effects of opioids, to improve and enhance the analgesic response, and get efficiency of costs related to treatment. We assessed the cost-effectiveness (CEA) of superior hypogastric plexus inhibition (SHPI) in patients with cancer and visceral pain in the lower abdomen. METHODS: We conducted a CEA within a retrospective follow-up clinical study at the National Cancer Institute in Mexico City in patients >18 years with cancer and visceral pain. We assessed patients that underwent SHPI between March 2005 and June 2009. We evaluated the visual analog pain scale (VAS), drugs resource consumption, and medical direct costs. The measures were evaluated before and after (1 day, 1 week, 1, 2, 3, and 6 months) the procedure. Incremental cost-effectiveness ratio (ICER) was calculated. RESULTS: Twenty-six patients underwent SHPI. They were matched with 26 patients with cancer and visceral pain managed with standard treatment (WHO analgesic ladder steps). The average cost per patient was not significantly different between treatment groups ($7372 vs. $6768 MXP, P = 0.54); however, the effectiveness (treatment success: 50% decrease in drug consumption within 30 days) was much higher for SHPI (65% vs. 19%, P < 0.001). The ICER was $1308 (IC 95% 1104–1485) MXP per patient. CONCLUSIONS: SHPI was effective for treating visceral pain in cancer patients. The ICER ($1313 MXP) shows that SHPI is a cost-effectiveness alternative in Mexico (threshold of 1 GDP per capita). 13th Euro Abstracts PCN74 ADDING ZOLEDRONIC ACID TO ENDOCRINE THERAPY IN PREMENOPAUSAL WOMEN WITH HORMONE-RESPONSIVE EARLY BREAST CANCER CAN BE COST-EFFECTIVE FROM ITALIAN, SPANISH, AND PORTUGUESE HEALTH-CARE PERSPECTIVES, BASED ON THE ABCSG-12 TRIAL Taneja C1, Delea TE2, Kaura S3, Sternini P4, Gerzeli S5, Gnant M6 1 PAI, Brookline, MA, USA; 2Policy Analysis Inc., Brookline, MA, USA; 3Novartis Pharmaceuticals Corporation, Florham Park, NJ, USA; 4Novartis Oncology, Origgio, Italy; 5 University of Pavia, Pavia, Italy; 6Medical University of Vienna, Vienna, Austria OBJECTIVES: To retrospectively estimate the cost-effectiveness of adding zoledronic acid (ZOL; 4 mg intravenously q6m) to adjuvant endocrine therapy (ET; goserelin plus tamoxifen or anastrozole) in premenopausal women with endocrine-responsive early breast cancer (ERBC) from Italian, Spanish, and Portuguese health-care perspectives. METHODS: A Markov model projected lifetime outcomes and costs of care for ERBC patients receiving 3 years’ adjuvant ET or adjuvant ET plus ZOL. Cost-effectiveness was measured as the incremental cost per quality-adjusted life-year (QALY) gained. Probabilities of BC recurrence were from the ABCSG-12 trial. Other probabilities and country-specific costs were from published literature. Results were generated under two scenarios: 1) benefits of ZOL persist to the 7-year maximum follow-up (trial benefit); 2) benefits persist until recurrence or death (lifetime benefit). RESULTS: Expected ZOL costs (medication and administration) were c1500 (Italy), c2100 (Spain), and c2300 (Portugal). Under the trial benefit scenario, resulting savings from reduced BC recurrence partially offset costs by c900 (both Spain and Italy) and c200 (Portugal). Therefore, projected total ZOL costs were c600 (Italy), c1300 (Spain), and c2100 (Portugal). Projected QALY gains with ZOL were 0.46 (Italy), 0.47 (Spain), and 0.33 (Portugal). Costs per QALY gained were c1304 (Italy), c2766 (Spain), and c6364 (Portugal) (all favorable). Under the lifetime benefit scenario, savings from reduced BC recurrences completely offset ZOL costs and yielded net savings of c2900 (Italy) and c2100 (Spain). Incremental total costs were c1400 for Portugal. Projected QALY gains with ZOL were 1.57 (Italy), 1.59 (Spain), and 0.96 (Portugal). The cost per QALY gained for Portugal was highly favorable (c1458). CONCLUSIONS: Adding ZOL to ET in premenopausal women with ERBC can be highly cost-effective (<c50,000 per QALY gained) in Italy, Spain, and Portugal. Additionally, ZOL would be considered cost saving to patients in Italy and Spain if these benefits persist >7 years. PCN75 COMPARISON OF ADVANCED NONINVASIVE TECHNIQUES TO SCREEN COLORECTAL CANCER: FECAL IMMUNOCHEMICAL TEST VS. FECAL DNA; A COS-EFFECTIVENESS STUDY Bhattacharya R, Shah J, Sansgiry S University of Houston, Houston, TX, USA OBJECTIVES: This study aims to compare the guaiac-based fecal immunochemical test (FIT), the primary colorectal cancer (CRC) detection technique, with the fecal DNA (F-DNA) test which has been recommended as an alternative to FIT as the standard of care. METHODS: A hybrid decision tree-Markov model was created to estimate the CRC screening cost per quality-adjusted life-year (QALYs) of using the FIT annually, or the F-DNA every 3, or the F-DNA every 5 years in individuals at average CRC risk from a third-party payer’s perspective. a hypothetical cohort of 10,000, 50-year-old individuals transitioning between the health states: healthy, polyps <10 mm, polyps >10 mm, local cancer, regional cancer, advanced cancer, and dead, were followed until they were 75 years. Colonoscopy followed every positive test result. Sensitivity, specificity, transition probabilities, and costs (in 2010 US Dollars) were obtained from clinical trials and published peer-reviewed articles. The costs and QALYs were discounted at 3% and sensitivity analyses were conducted. RESULTS: Using FIT annually would result in an average cost of $56,716.94/QALY for each individual with an incremental cost-effectiveness ratio (ICER) of $76,181/ QALY when compared to F-DNA used every 5 years. In the ICER plane of 1000 Monte Carlo simulations, FIT was more costly but more effective technique compared to F-DNA used every 5 years, in 77% of the samples. FIT was the most cost-effective screening strategy at willingness to pay (WTP) of $100,000/QALY. However, at a lower WTP of $50,000/QALY, F-DNA every 5 years was cost-effective until a threshold of $71,000/QALY. F-DNA conducted every 3 years was completely dominated by FIT. CONCLUSIONS: Further research is needed, and third-party payers may need to assess variables such as compliance and patient characteristics, before considering the F-DNA as a standard of care for screening CRC. PCN76 A SIMULATION MODELLING APPROACH TO QUANTIFY THE COST-EFFECTIVENESS OF EXTRA-CORPOREAL PHOTOPHERESIS IN CHRONIC GRAFT VERSUS HOST DISEASE (CGVHD) IN SPAIN Perez-Simon J1, Sierra J2, Crespo C3, Rodríguez J4 Brosa M3 1 University Hospital of Salamanca Cancer Research Center (CIC/CSIC) of Salamanca and Center of Regenerative Medicine and Cellular, Salamanca, Spain; 2Hospital de la Santa Creu i Sant Pau, Barcelona, Spain; 3Oblikue Consulting, Barcelona, Spain; 4Johnson & Johnson, Madrid, Spain OBJECTIVES: Chronic graft-versus-host disease (cGVHD) is a major complication after allogeneic hematopoietic stem cell transplantation (HSCT) that impairs quality of life, functional status, and long-term survival. There is no standard therapy for patients whose cGVHD does not resolve with immunosuppressors and corticosteroid A265 treatment. The aim of this study was to compare the cost-effectiveness of extra-corporeal photopheresis (ECP) with Rituximab (Rmb) or Imantinib (IMT) or pooled comparators (pooled) in addition to the usual care of cGvHD after standard treatment failure in Spain. METHODS: The model assessed the incremental cost-effectiveness ratio (ICER) of ECP versus Rmb or IMT or pooled comparator. The incremental cost and quality-adjusted life-year (QALY) gained were estimated using a short-term decision analysis and a long-term Markov cohort modeling approach. Model probabilities were obtained from literature, while treatment pathways and adverse event where derived from expert opinion. Local data on health resources use and costs were used and validated by clinical experts. The time horizon of the study was 5 years and only direct local medical costs (euros 2010) were considered. a probabilistic sensitivity analysis was performed. RESULTS: Preliminary results show that the higher efficacy of ECP leads to a gain of 0.19–0.20 QALY at first year and 0.15–0.19 at year 5 when compared to Rmb or IMT or pooled. The short-term cost of ECP is higher than Rmb (c2.900), IMT (c800) and pooled (c1.800). The ICER results for ECP for the first year were c15,340 versus RMB, c3.663 verus IMT and c8.977 versus pooled. At 3 years, ECP was dominant versus IMT and pooled, and showed ICER less than c3.000 vs. Rmb. The results of the evaluation were sensitive to limited data available. CONCLUSIONS: Preliminary results of this study indicate ECP is a cost-effective, below the Spanish threshold, or dominant option with respect altenatives. PCN77 COST-EFFECTIVENESS OF OCTREOTIDE LAR IN PATIENTS WITH METASTATIC NEUROENDOCRINE MIDGUT TUMORS FROM THE PRIVATE PAYER PERSPECTIVE IN BRAZIL Takemoto ML1, Fernandes RA1, Chinen R2, Alves MR2 1 ANOVA—Knowledge Translation, Rio de Janeiro, RJ, Brazil; 2Novartis Pharmaceuticals, São Paulo, SP, Brazil OBJECTIVES: Octreotide LAR has shown antiproliferative activity in advanced midgut neuroendocrine tumors (NET) increasing time to tumor progression (TTP) compared to placebo. This study aims to assess the costs and consequences of OCT-LA versus best supportive care (BSC) in patients with metastatic midgut NET from the private payer perspective. METHODS: A three health state (progression-free survival, progression, and death) Markov model with a 10-year time horizon was developed with data from the phase III PROMID trial. Within the trial, subjects remained on treatment until progression. Resource use was estimated through published data and input from clinical experts to reflect clinical practice in the Brazilian private setting. Unit costs were obtained from Brazilian official sources. Costs and outcomes were discounted 5% per annum. RESULTS: The model estimated 14 months PFS with OCT-LA versus 6 months with BSC. Estimated PFS gain was 0.60 years (1.07 vs. 0.46). Total cost of treatment was 275,497 BRL for BSC and 303,111 BRL for OCT-LA. The incremental cost per progression-free year gained was 28,706 BRL in the OCT-LA arm versus BSC due to treatment until progression. The mean cost of supportive care for progressive disease represented 87.3% (239,883 BRL) and 76.9% (224,388 BRL) of the final cost of treatment for BSC and OCT-LA, respectively. Results remained consistent when univariate sensitivity analyses were run. CONCLUSIONS: OCT-LA is a clinically effective option to control tumor growth in patients with metastatic midgut NET. OCT-LA provides longer TTP compared to BSC for those patients. Although there is ecological evidence to suggest improvement in OS after introduction of OCT LA, the ICER for an additional life-year gained is not currently calculable as the PROMID trial was not designed to evaluate OS. Further areas of research to elucidate the association between PFS and OS in NET are needed. PCN78 OST-EFFECTIVENESS ANALYSIS OF ADJUVANT THERAPY WITH IMATINIB MESYLATE IN PATIENTS AFTER RESECTION OF LOCALIZED PRIMARY GASTROINTESTINAL STROMAL TUMOR Krysanov I, Zorin N, Pyadushkina E, Koval DA Institute of Clinico-Economic Expertise and Pharmacoeconomics, RSMU, Moscow, Russia OBJECTIVES: Imatinib is a low molecular tyrosine kinase inhibitor that blocks the kinase activity KIT and PDGFRα, and a first-line drug in the treatment of unresectable and metastatic gastrointestinal stromal tumor (GIST). The standard treatment of patients with localized primary GIST is a complete surgical resection of the tumor. Several studies have shown that target therapy improves survival of patients after GIST resection. The purpose of this study was to estimate the costs and effectiveness of adjuvant imatinib therapy versus no treatment in patients who have undergone GIST resection. METHODS: A Markov model was used to estimate costs and effectiveness of adjuvant imatinib therapy in the long-term follow-up period. Data on overall and recurrence-free survival were taken from the phase III clinical trial ACOSOG Z9001 and were used to assess efficacy. Measures of effectiveness include such indicators as life-years saved and quality-adjusted life-years (QALYs) gained for adjuvant imatinib following surgical resection and surgical resection only. Data on the common practice of GIST treatment in the Russian oncology centers were used in the model. Costs, life-years, and QALYs gained were calculated over the 50-year time horizon and discounted at an annual rate of 5%. RESULTS: The number of life-years saved was 10.01 for imatinib treatment against 8.67 for no treatment. The number of QALYs was 7.97 and 6.82, respectively. The costs of 1-year patient management with adjuvant imatinib therapy were c44,348 per person; a patient who had not received imatinib in adjuvant mode required c32,102 per person. CONCLUSIONS: The analysis showed that adjuvant imatinib therapy is more costly compared with no treatment. However, it is more effective and can increase the life expectancy of patients. In this A266 case, the costs of therapy are partly offset by reducing the frequency of relapses and the progression of the disease. PCN79 ECONOMIC MODELING FOR TREATMENT FAILURE PATIENTS USING MULTIPLE ROUNDS OF THERAPY AS COMPARATOR Aggarwal S1, Stevens CA2 1 PAREXEL Consulting, Bethesda, MD, USA; 2PAREXEL Consulting, Waltham, MA, USA OBJECTIVES: Treatment failure patients in various disease areas are often treated by multiple rounds of therapy. However, new treatment options are emerging that have potential to replace that treatment with single-agent or single round of combination treatment. It is challenging to demonstrate cost-effectiveness of these new agents, especially when comparator is not one single regimen but sequential treatment. We present here our results from a study where we developed a model that can incorporate multiple rounds of treatment or relapses to estimate cost-effectiveness of new emerging therapies. METHODS: Intervention was chosen as an emerging T-cell lymphoma drug candidate. Comparator was chosen as sequential treatment with 1–5 chemo regimens (called DHAP, ESHAP, ICE, HyperCVAD, and EPOCH). All comparator chemo regimens are generics and their prices were obtained from Medispan’s PriceRx. Intervention’s price was assumed as median price of branded chemotherapy agents. Cost, efficacy, adverse events, and utilities were sourced and estimated from published studies for T and B-cell lymphoma. Relapses and number of chemo regimens for comparators were varied from 1–5. Sensitivity analyses were performed for all base calculations. RESULTS: Model results show that a new agent that can replace multiple rounds of treatment is relatively more cost-effective than another agent that replaces relatively fewer rounds of treatments. Our base-case incremental cost-effectiveness with one chemo regimen as comparator was $262,908. However, if there are 2,3,4, or 5 sequential rounds, the ICER values change to $223,078, $183,249, $143,420, and $103,591, respectively. CONCLUSIONS: For newer agents that are indicated for treatment failure patients, the use of sequential treatments as comparator can significantly improve their cost-effectiveness. The model approach described here can be used for arthritis, hepatitis C, and diabetes and oncology TF patients. PCN80 COST-EFFECTIVENESS ANALYSIS OF OXLIPLATIN IN ADJUVANT THERAPY FOR STAGE 3 COLON CANCER PATIENTS IN JAPAN Fukuda T1, Shiroiwa T2, Takeuchi T1, Shimozuma K2, Ohashi Y1 1 Tokyo University, Tokyo, Japan; 2Ritsumeikan University, Kusatsu, Shiga, Japan OBJECTIVES: FOLFOX (folic acid [l-LV], 5-FU, and oxaliplatin) is a standard therapy for metastatic colorectal cancer. FOLFOX in adjuvant therapy was approved in 2009. However, cost-effectiveness of FOLFOX, which can prolong DFS (diseasefree survival), is not known. METHODS: We performed cost-effectiveness analysis of FOLFOX in adjuvant therapy for stage3 colon cancer compared with FU/LV as a standard regimen. Our analysis is based on the patient-level data of MOSAIC (the Multicenter International Study of Oxaliplatin/5-Fluorouracil/Leucovorin in the Adjuvant Treatment of Colon Cancer) trial. Survival curve of DFS and OS (overall survival) was extrapolated by cure model, which uses parametric regression considering some patients can cure without recurrence. Death of any other causes was treated as competing risk. Expected value of mean survival year was gained by calculating area under the estimated survival curve during 15 years. QALY (quality-adjusted life-year) was calculated weighting survival time by utility scores. Since our analysis is from the perspective of health-care payer, only direct medical costs were included. Three percent annual discount rate was used for both costs and outcome. RESULTS: Adjuvant FOLFOX therapy for stage 3 colon cancer patients can gain more QALY than standard FU/LV therapy. The difference of both therapies is about 0.5 QALY. ICER (incremental cost-effectiveness ratio) of FOLFOX compared with FU/LV estimated to be less than JPY 2.5 million (US$28,000, US$ 1 = JPY 90) per QALY. This value is thought to be a little conservative because time horizon of our analysis is 15 years, not lifetime to avoid uncertainty of long-term future. CONCLUSIONS: FOLFOX therapy in adjuvant of stage 3 colon cancer is cost-effective. The ICER compared with FU/LV is acceptable from the Japanese health-care payer. PCN81 COST-EFFECTIVENESS ANALYSIS OF THREE STRATEGIES OF ERLOTINIB TREATMENT IN NON0-SMALL-CELL LUNG CANCER: A PROSPECTIVE MULTICENTRIC FRENCH STUDY (ERMETIC) Borget I1, Cadranel J2, Mauguen A1, Coudert B3, Dansin E4, Friard S5, Daniel C6, Quoix E7, Madelaine J8, Madroszyk A9, Morin F10, Pignon JP1, Chouaid C11 1 Institut Gustave Roussy, Villejuif, France; 2Hospital Tenon, PARIS, France; 3CLCC Dijon, Dijon, France; 4CHU Lille, Lille, France; 5Hospital Foch, Suresnes, France; 6Institut Curie, Paris, France; 7CHU Strasbourg, Strasbourg, France; 8CHU Caen, Caen, FRANCE, France; 9CLCC Marseille, Marseille, France; 10IFCT, Paris, France; 11Hôpital St Antoine, Paris, France OBJECTIVES: Although several clinical and biological parameters are prognostic factors of NSCLC patients outcome, their medico-economic impact in the prescription of erlotinib has never been evaluated. a French NCI prospective study aimed to compare cost and effectiveness of three strategies of erlotinib initiation in second line or more treatment of advanced NSCLC patients: initiation in all patients, patients selected on clinical-guided strategy, and patients selected on biological-guided strategy. METHODS: A Markov model compared the outcomes and costs (limited to direct medical costs from the third-party payer perspective) of a prospective multicentric cohort of consecutive advanced NSCLC patients newly treated by erlotinib, to a cohort 13th Euro Abstracts of clinical-selected patients (non/ex-smoking women with adenocarcinoma histology) and a cohort of biomarker-selected patients (EGFR mutation). Utility data were extracted from literature. Sensitivity analyses were performed. RESULTS: A total of 522 patients were enrolled between March 2007 and March 2008. Median age was 63 years; 32% were females; 65% had adenocarcinoma; and 8% had EGFR mutation. The strategy which consists to treat all patients was dominated, as it was both the less effective and the most expensive strategy (0.495 QALY/c22,396). The clinical-guided strategy was slightly more effective than the biological-guided strategy (respectively 0.568 and 0.563 QALY), but it was also more expensive (respectively c16,299 and c15,187). The dominant strategy was then the biological-guided strategy (c26,975/ QALY). The model was robust to variations of biological exam costs, palliative costs, and utility data. Biological-guided strategy appears the most effective and the less expensive strategy when the prevalence of EGFR mutation exceeds 10%. CONCLUSIONS: Biological-guided strategy appears the dominant strategy if the prevalence of EGFR mutation was >10%. This suggests determining EGFR mutation status in priority to non/former smokers, females with adenocarcinoma. PCN82 RESEARCH PRIORITIZATION FOR PROSPECTIVE COMPARATIVE EFFECTIVENESS RESEARCH (CER) IN CANCER GENOMICS Veenstra D1, Thariani R1, Carlson JJ1, Garrison L2, Mohr P3, Deverka P4, Tunis SR3, Hoban C5, Baker LH5, Ramsey S6 1 University of Washington, Pharmaceutical Outcomes Research and Policy Program, Seattle, WA, USA; 2University of Washington, Department of Pharmacy, Seattle, WA, USA; 3Center for Medical Technology Policy, Baltimore, MD, USA; 4University of North Carolina, Chapel Hill, Phoenix, AZ, USA; 5University of Michigan, Ann Arbor, MI, USA; 6Fred Hutchinson Cancer Research Center, University of Washington, Seattle, WA, USA OBJECTIVES: The Center for Comparative Effectiveness Research in Cancer Genomics (CANCERGEN) is a multidisciplinary, national consortium established to conduct CER in Genomics and Personalized Medicine (GPM). The objective was the evaluation and prioritization of GPM applications for study in a prospective, randomized CER trial. METHODS: Candidate GPMs were identified through a landscape analysis of recent literature. Initial candidates were examined by cancer genomics experts and study investigators to identify 5–7 GPMs for evaluation by a diverse group of external stakeholders, including representatives from patients groups, payers, test developers, state-funded public HTA programs, and practicing oncologists. We developed Topic Briefs and Test Target Profiles assessing the following domains: population impact, current standard of care, clinical validity, potential benefits, potential harms, economic impact, evidence of need, trial feasibility, and current payer status. RESULTS: We identified 43 studies from 183 GPMs based on our landscape analysis, which were narrowed to four GPMs through feedback from cancer genomics experts; two additional GPMs were identified by investigators from our clinical trials consortium (SWOG). The six GPMs included: ERCC1 testing for platinum chemotherapy in NSCLC, EGFR mutation testing for TKIs in NSCLC maintenance, tumor markers for breast cancer recurrence, EGFR FISH testing for first-line cetuximab in NSCLC, BRAF testing in colorectal cancer, and gene expression profiling in multiple myeloma. External stakeholders preliminarily identified the first three of these as most likely providing the greatest value of research. CONCLUSIONS: A rapid process for research prioritization involving literature evaluation, expert input, and stakeholder feedback is feasible with adequate resources and processes. Elements include an organization facilitating collaboration between investigators with CER/clinical trials experience, and utilizing systematic and timely evidence-assessment accessible to stakeholders. Final selection of a GPM for study in a prospective CER trial will be based on quantitative value-of-information analyses, implementation feasibility, and funding mechanisms, including coverage with evidence development. PCN83 A COST-EFFECTIVENESS ANALYSIS (CEA) FOR DENOSUMAB, A FULLY HUMAN MONOCLONAL ANTIBODY FOR CANCER TREATMENTINDUCED BONE LOSS (CTIBL) IN NON-METASTATIC PROSTATE CANCER (PRCA): A SWEDISH PERSPECTIVE Shroff S1, Martin M1, Kearney M2, Lothgren M2, Bracco A2 1 i3 Innovus, Uxbridge, Middlesex, UK; 2Amgen (Europe) GmbH, Zug, Switzerland BACKGROUND: Androgen deprivation therapy (ADT) decreases bone mineral density (BMD), increasing risk of fragility fractures and decreasing quality of life over time. Until recently, there were no licensed treatments despite high unmet medical need. In a randomized, double-blind, placebo-controlled trial, denosumab increased BMD and reduced the incidence of vertebral fractures in nonmetastatic PrCa patients receiving ADT. OBJECTIVES: To assess the cost-effectiveness (CE) of denosumab versus no treatment in nonmetastatic PrCa CTIBL patients in Sweden. METHODS: A Markov model was adapted from previously developed models in osteoporosis. This includes six fracture states: hip, vertebral, wrist, other, post-hip, and post-vertebral fracture. Model inputs were based on a literature review conducted in PUBMED. The target population reflected patient characteristics of the trial. The model horizon was 5 years, reflecting progression to metastatic disease. General population fracture risks from Swedish males were adjusted by the relative risk of fracture due to ADT. Only the efficacy of denosumab on vertebral fractures was included in the base case. a societal perspective was used. Published trial data were used to validate the model in terms of fractures. RESULTS: The cost per QALY (ICER) for denosumab versus no treatment ranged from c46,683 to c58,282. Multiple deterministic sensitivity analyses (SA) were performed. The main driver of CE was the efficacy of denosumab; when it 13th Euro Abstracts provides benefits at other fracture sites as shown in the postmenopausal pivotal trial FREEDOM, the ICER reduces to c19,726. a probabilistic SA showed that denosumab was a cost-effective option for a willingness to pay >c60,000. CONCLUSIONS: Denosumab prevents vertebral fractures in patients with PrCa receiving ADT and is cost-effective versus no treatment. Vertebral fractures significantly reduce quality of life and since there is no other licensed treatment in Sweden, denosumab represents an important option in PrCa at commonly accepted CE thresholds in Sweden. PCN84 COST-EFFECTIVENESS OF ERLOTINIB IN FIRST-LINE TREATMENT OF ADVANCED NON-SMALL CELL LUNG CANCER (NSCLC) IN FIT ELDERLY PATIENTS: AN ECONOMICAL ANALYSIS OF A PROSPECTIVE PHASE 2 STUDY (GFPC 0504) Chouaid C1, Le caer H2, Crequit J3, Monnet I4, Chouabe S5, Locher C6, Paillotin D7, Auliac JB8, Thomas P9, Vergnenegre A10 1 Hôpital St Antoine, Paris, France; 2Hôpital de draguignan, Draguignan, France; 3Centre Hospitalier de Beauvais, Beauvais, France; 4Centre Intercommunal de Creteil, Creteil, France; 5 Centre hospitalier de Charleville Meziere, Charleville Mézières, France; 6Hôpital Saint Faron, Meaux, France; 7CHU de Rouen, Rouen, France; 8Service De Pneumologie, mantes la jolie, France; 9Centre Hospitalier, GAP, France; 10Hôpital du Cluzeau Service de Pathologie Respiratoire, Limoges, France OBJECTIVES: Median age of newly diagnosed non-small cell lung cance (NSCLC) is 70 years (with 1/3 older than 75 years) and elderly are more vulnerable to chemotherapy. In this population, weekly gemcitabine and docetaxel or erlotinib are both active in advanced NSCLC treatment. The GFPC0504 randomized prospective phase 2 study assess in fit elderly patients with advanced NSCLC, efficacy of weekly chemotherapy followed by erlotinib if progression (Arm A) versus erlotinib followed by chemotherapy if progression (arm B). The main objective of this study was time before second progression, secondary objective was overall survival. The objective of this study is to assess the cost-effectiveness of erlotinib in first-line treatment of NSCLC in fit elderly patients. METHODS: Outcomes (PFS and overall survival) and direct medical costs until second progression (from the third-party payer perspective) were prospectively collected. Costs after second progression and health utilities (based on disease states and grade 3–4 toxicities) were derived from the literature. RESULTS: For respectively 48 and 51 patients randomized respectively in arm a and B, PFS were 6.4 and 5.2 months, overall survival were 9.2 and 7.9 months; mean Qualy and mean direct costs (euros value 2010) were respectively c0.434 ± c0.394 and c26,297 ± c25,297 and c0.471 ± c0.451 and c25,948 ± c18,206. Acceptability curve will be presented at the meeting. CONCLUSIONS: In this population of fit ederly patients, erlotinb in first line, followed by chemotherapy if progression appears as dominant compare to chemotherapy followed by erlotinib if progression. PCN85 SORAFENIB AND SUNITINIB IN METASTATIC RENAL CELL CARCINOMA: COST-EFFECTIVENESS ANALYSIS IN REIMBURSEMENT PROCEEDINGS VS. DATA FROM CLINICAL PRACTICE Ondrackova B1, Demlova R2 1 Masaryk University, Faculty of Medicine, Brno, Czech Republic; 2Masaryk Memorial Cancer Institute, Brno, Czech Republic OBJECTIVES: Sorafenib and sunitinib are approved for patients with advanced or metastatic renal cell carcinoma after INF-a or IL-2 therapy failure or intolerance, with PS 0-1 and without CNS metastasis in defined cancer centers in the Czech Republic; sunitinib is reimbursed for first-line therapy in mRCC patients of good or intermediate risk. METHODS: We assessed the cost of sunitinib and sorafenib in patients treated in comprehensive cancer center and prepared cost-effectiveness analysis (CEA) to compare our data to CEA submitted by manufacturers to Czech authority (SUKL = State Institute for Drug Control) in reimbursement proceedings between 2008 and 2010. (1c = 26CZK). RESULTS: CEA of sunitinib submitted to SUKL was based on cost of pharmacotherapy and clinical data of Motzer et al. study (NEJM 2007; time to PD: sunitinib 11 months, INF-a 5 months; duration of PD to death 6 months). Cost per progression-free year (PFY) was 324144CZK/12467c in manufacturer’s analysis, CZK867,946CZK/c33,383 in SUKL analysis (after INF-a cost reduction and costs after PD removal) and CZK2,304,914/c88,651 in our analysis (cost and effects of sunitinib based on our results; INF-a data were assumed identically). CEA of sorafenib was performed for patients after cytokine intolerance or failure (Escudier et al.; NEJM 2007) in comparison with sunitinib (70% pts) or BSC (30% pts). The cost per PFY was CZK965,726/c37,143 in manufacturer’s analysis. Although sorafenib was cheaper alternative according to our results, time to progression was shortened by 18 days (ICER CZK516,820/c1,9878 per PFY). CONCLUSIONS: The cost per PFY in sunitinib was seven times lower in manufacturer’s analysis than in CEA based on real data from cancer center. We assume that this was mainly caused by shorter time of pharmacotherapy in original study (6 vs. 11 months in our data). CEA of sorafenib demonstrated lower costs and effects in our analysis, because the significance of comparator (70% pts sunitinib) was underestimated in manufacturer’s analysis. A267 PCN86 COST-EFFECTIVENESS ANALYSIS OF SPLANCHNIC NERVE BLOCKADE IN PATIENTS WITH CANCER AND VISCERAL PAIN IN THE UPPER ABDOMEN Domínguez-Ocadio G, Cerezo O, González-Buendía NI, Guajardo-Rosas J, Plancarte-Sánchez R Oncology National Institute, Mexico City, D.F., Mexico OBJECTIVES: The aim of a sympathetic blockade is to improve the analgesic response, diminish the opioid consumption, reduce the adverse effects from opioides, and get efficiency of costs related to treatment. We analyzed the cost-effectiveness of Splanchnic Nerves Blockade (SNB) versus drug therapy in patients with cancer and visceral pain at the upper abdomen. METHODS: A cost-effectiveness analysis was conducted within a retrospective, follow-up study in patients >18 years with cancer and visceral pain. Using medical records, we assessed patients that underwent a SNB between March 2005 and December 2009. We evaluated the visual analog pain scale (VAS), Karnofsky performance scale (KPS), and medical direct costs. The measures were evaluated before and after (1, 2, 3, 6, 9, and 12 months) the procedure. Cost methodology was calculated trough cost of illness and microcosting technique, to get the incremental cost-effectiveness ratio (ICER). RESULTS: Sixty-five patients were treated with SNB and 19 with drug treatment-WHO analgesic ladder steps (mean age 52.7 ± 12.9 and 54 ± 12.9, respectively). Basal characteristics were not different between them. VAS scores diminished in both arms, but at repeated measures ANOVA patients on SNB had better pain control (P < 0.05) and higher KPS (P < 0.05). The mean cost per patient in 1-year follow-up for the drug treatment group was $7512 MXP (CI 95% $1587–$13,436 MXP) and $5433 (CI 95% $5114–$5752) for SNB. The effectiveness measure was 80% for SNB versus 20% for the drug treatment group, respectively. The ICER obtained was negative (−$3526 MXP, IC 95% −5860 to −1191), favoring the SNB as a cost-saving alternative. CONCLUSIONS: SNB showed to be less costly and more effective than drug treatment alone. However, when a sensitive analysis (bootstrap methodology) was conducted, the sample size was not powerful enough for a precise CE estimate. PCN87 PHARMACOECONOMIC ANALYSIS OF DIRECT MEDICAL COSTS OF METASTATIC COLORECTAL CANCER THERAPY WITH XELOX OR FOLFOX4 WITH OR WITHOUT BEVACIZUMAB AS THE FIRST-LINE TREATMENT Tikhomirova A1, Kulikov A2, Yagudina R2 1 FGU NCESMP Roszdravnadzora, Moscow, Russia; 2Moscow Medical Academy, Moscow, Russia OBJECTIVES: Pharmacoeconomic analysis of direct medical costs of mCRC therapy using XELOX/FOLFOX4, XELOX + BV/FOLFOX4 + BV. METHODS: Costs of diagnosis, medical services, and hospitalization were based on the price list for diagnostic and therapeutic procedures of Cancer Research Center n.a. N.N.Blokhin RAMS. The medical services patient should receive during the treatment and the frequency of their appointments were taken from the standards of medical care for patients with colon and rectum cancer. Cost analysis of anticancer drugs (16 courses of XELOX/XELOX + BV, 24 courses of FOLFOX4/FOLFOX4 + BV) and related drugs were based on the information about maximum selling import prices, registered, and entered into the State Register of prices of vitally essential drugs. The cost of other drugs was based on a database of retail prices for drugs in pharmacies, which was subsequently reduced by trade discount. RESULTS: In was calculated that the cost of diagnosis was 16,757 rubles and the medical services—379,815 rubles. The mCRC therapy as a first line by XELOX was 1,172,731 rubles and by XELOX + BV—2,526,110 rubles; by FOLFOX4—1,487,627 rubles and by FOLFOX4 + BV—2,843,558 rubles. The cost saving in applying the regime XELOX compared to FOLFOX4 regime amounted to 314,896 rubles. In applying the regime of XELOX in combination with BV in comparison with the regime of FOLFOX4 in combination with BV amounted to 317,448 rubles. Sensitivity analysis showed that the decrease and increase of the cost of capecitabine and bevacizumab in 20% for XELOX/XELOX + BV does not exceed the cost of regimes FOLFOX/FOLFOX4 + BV. CONCLUSIONS: From the pharmacoeconomic point of view, the most optimal is the use of XELOX and XELOX + BV regimes because of lower costs for neutropenia treatment, associated with an increased risk of infectious complications, as well as with a large number of hospitalization days. PCN88 COST-EFFECTIVENESS ANALYSIS OF CANCER TREATMENTS IN SOUTH OF IRAN Ahmad Kiadaliri A1, Bastani P2, Hatam N3, Ahmadloo N3 1 Lund University, Malmo, Sweden; 2Iran University of Medical Sciences, Tehran, Iran; 3Shiraz University of Medical Sciences, Shiraz, Iran OBJECTIVES: To calculate the incremental cost-effectiveness of docetaxel-adriamicine-cyclophosphamide (TAC) against adriamicine- cyclophosphamide-5 flourouracil (FAC) in treatment of breast cancer in south of Iran. METHODS: A double blind study was applied on a cohort of 100 patients suffering from breast cancer with nodepositive in the radiotherapy center of Namazi Hospital, Shiraz, Iran. The European organization for research and treatment of cancer questionnaire (EORTC QLQ-C30) was used for the measuring of quality of life at the first and last session of chemotherapy cycle. Third-party payer perspective was applied for costing side of evaluation. At last, two-way sensitivity analysis was used for ensuring the robustness of the results. A268 RESULTS: In spite of the same quality of life score at the first session of chemotherapy (74.5 out of 100), after finishing the chemotherapy cycle, patients in TAC arm had the lower score of QOL (64 in TAC vs. 68 in FAC) and higher range of toxicity and their medical costs were higher as well (the average costs in TAC was 391,176,968.2 Rials vs. 2,427,775.2 in FAC). ICER was negative that showed the dominant result for FAC comparing with TAC. CONCLUSIONS: It seems that because of the short horizon of the study, TAC regimen had the worse impact on the patient’s quality of life during the chemotherapy cycle because of more side effects than FAC. It is believed that there is need for other studies with longer time horizons and specific attention to the effects of these treatments on survival and quality of life. PCN89 PROJECTING THE POTENTIAL COST-EFFECTIVENESS OF A BREAST CANCER VACCINE IN COMPARISON TO OTHER STANDARD TREATMENTS: A DECISION ANALYTIC MODEL Patel TB, Zaveri VB, Gohil NS, McGhan WF University of the Sciences in Philadelphia, Philadelphia, PA, USA OBJECTIVES: Breast cancer is known to be one of the leading causes of death among the female population. Preventive measures may provide an economic and outcome advantage by reducing treatment costs and increasing survival. The objective of this study was to evaluate the cost-effectiveness of a breast cancer vaccine versus current standard treatments. METHODS: TreeAge software was used to calculate the costeffectiveness. a decision tree was constructed for different probabilities of success and failure for the vaccine versus standard treatment. Costs and outcomes (life-years saved) ranges were obtained from published clinical trials. The vaccine effectiveness was projected from animal studies, with human clinical trials expected within a year. The range of effectiveness of the vaccine was considered between 30% and 90% with a baseline at 80%. The costs included for standard treatments ranged from $20,000 to $45,000 and the cost of the vaccine was assumed at $450 for three doses; therefore, the cost for vaccine ranged from $300 to $2000 depending on the number of doses. The incremental cost-effectiveness ratios were calculated from the range of costs and outcomes. Sensitivity analyses were performed to determine the robustness of the findings. RESULTS: Vaccination was found to be a potentially cost-effectiveness option with an ICER of 2384.146 relative to standard treatment. The incremental effectiveness was 8.2 life-years saved. The highest cost-effectiveness of the vaccine was at 90% success and a cost of not more than $1000 per individual. Sensitivity analyses indicated that the vaccine remained cost-effective over the range of model parameters. CONCLUSIONS: The breast cancer vaccine was projected to be the most costeffective treatment option in this analysis. It is expected that better screening for breast cancer vaccine patient candidates will be available in the future. PCN90 COMPARATIVE RETROSPECTIVE NON-RANDOMIZED PHARMACOECONOMIC TRIAL OF EFFICIENCY AND SAFETY OF USE OF PACLITAXELS (PACLITAXEL-LENS OR TAXOL) IN A MONOMODE FOR 2ND LINE OF TREATMENT OF METASTATIC BREAST CANCER PATIENTS Pavlysh A1, Kolbin A2, Livshits R2, Koroleva O2, Manikhas A1, Tkachenko E1, Atrashevskaya N1, Demicheva N1 1 Saint Petersburg City Clinical Oncology Dispensary, Saint Petersburg, Russi; 2Saint Petersburg State University, Saint Petersburg, Russia OBJECTIVES: For the first time in a modern Russian economic conditions, it has been made pharmacoeconomics trial (PE) uses Russian generic of paclitaxel (Paclitaxel-Lens [PL]) in comparison with original drug (Taxol (T)) at chemotherapy (ChT) in a monomode for 2nd line of metastatic breast cancer (MBC) in real clinical practice. METHODS: It has been provided retrospective comparative nonrandomized clinical trial which have been included 70 patients for 35 patients of each group (PL or T) after analysis of 148 case records. RESULTS: At the analysis of effectively treatment MBC in group of the patients who have received T, the partial remission (PR, 28.5% against 10%) statistically significantly has been more often reached. At the analysis of safety, it has been shown that in group of the patients who have received PL, statistically significantly has been more often fixed hepatotoxicity (23.3% against 3.8%) and an anemia (19.2% against 3.5%). In group of the patients who have received T, statistically significantly has been more often fixed arthralgia/ myalgia (29.8% against 0%). Total direct costs (DC) in group of patients with T also there were above, than in group of PL, namely $10,727 and $9765 accordingly. Calculation of efficiency of expenses has shown that treatment of MBC by T more expensive and more effective, than treatment by PL. CONCLUSIONS: Thus, as a result of research, it has been established that: 1) Applying of T was more (from 7% to 11%) expensive, than PL, but gave the PR is much more often; 2) The alternative scenario and the sensitivity analysis shown to choose conditions when application of compared drugs will be economically more expedient; and 3) Thus, it is necessary to take into consideration, what application of PL was more often accompanied by hepatotoxicity and anemia, like arthralgia/ myalgia after using of T. 13th Euro Abstracts PCN91 BEVACIZUMAB + PACLITAXEL + CARBOPLATIN (BEV + PAC + CAR) VS. PEMETREXED + CISPLATIN (PEM + CIS) IN ADENOCARCINOMA NON-SQUAMOUS NON-SMALL CELL LUNG CANCER (NSCLC): A COST-EFFECTIVENESS ANALYSIS FROM A POLISH PUBLIC PAYER’S PERSPECTIVE Kawalec P1, Badurak P2, Denisso T2, Jastrzebski D3, Marek M4, Pluzanski A2, Szczesna A5, Szkultecka-Debek M6, Russel-Szymczyk M6 1 Jagiellonian University, Kraków, Poland; 2Maria Sklodowska-Curie Memorial Cancer Center and Institute, Warsaw, Poland; 3Medical University of Silesia, Zabrze, Poland; 4Leszczynski Memorial Hospital, Katowice, Poland; 5Mazovian Center of Lung Diseases and Tuberculosis, Otwock, Poland; 6Roche Polska Sp. z o.o., Warsaw, Poland OBJECTIVES: To determine and compare the cost-effectiveness of Bev + Pac + Car versus Pem + Cis regimens in the treatment of patients with adenocarcinoma nonsquamous NSCLC from a Polish Public Payer’s perspective. METHODS: Efficacy and safety of 15 mg of bevacizumab + 200 mg/m2 of paclitaxel + 6 mg/mL/min of carboplatin versus 500 mg/m2 of pemetrexed and 75 mg/m2 of cisplatin was assessed based on a systematic review performed for both therapies according to evidence-based medicine principles. A cost-effectiveness analysis was performed with a lifetime (5 years) horizon and the National Health Fund perspective. a three state (progressionfree, progression, death) Markov model was developed. Costs of 1st and 2nd line therapy, administration and monitoring, adverse events treatment, and palliative care were included. Sensitivity analyses testing the influence of length of time horizon, probability of progression, utilities, discounting rates, cisplatin dose, and the length and costs of 2nd line therapy were performed. RESULTS: Bev + Pac + Car results in 0.21 life-years gained per patient when compared to Pem + Cis in the treatment of patients with adenocarcinoma non-squamous NSCLC. The additional cost per patient was 18,840 pln (1 EURO = 4.1PLN) over patient’s lifetime when Bev + Pac + Car was used instead of Pem + Cis regimen. The incremental cost-effectiveness ratio (ICER) was at an acceptable 91,216 pln. The sensitivity analyses demonstrated that the duration of 2nd line treatment (assumption of 2nd line treatment continuation for more than six cycles) considerably influenced the ICER (1,198 pln). Other sensitivity analyses confirmed the base-case results, proving conclusions’ robustness. CONCLUSIONS: Based on this modeling analysis, 1st line Bev + Pac + Car therapy is a clinically superior and cost-effective treatment for patients with adenocarcinoma non-squamous NSCLC when compared to chemotherapies such as Pem + Cis. PCN92 PHARMACOEPIDEMIOLOGICAL AND PHARMACOECONOMIC EVALUATION OF OXALIPLATIN IN PALLIATIVE CHEMOTHERAPY OF METASTATIC COLORECTAL CANCER (MCCR) Kolbin A1, Orlova R2, Pavlysh A3, Llivshits M1 1 Saint Petersburg State University, Saint Petersburg, Russia; 2Saint Petersburg Medical Academy of Postgraduate Education, Saint Petersburg, Russia; 3Saint Petersburg City Clinical Oncology Dispensary, Saint Petersburg, Russia The problem of original drugs substitution on generics presents in the Russian clinical practice due to rational expenditures allocation. Pharmaceutical bioequivalence of generic should be confirmed by therapeutic one. Only after such kind of confirmation, the mentioned substitution could be made in different segments of doctors’ practice especially in anticancer chemotherapy. OBJECTIVES: To evaluate the clinical-economic interchangeability of the original oxaliplatin Eloxatine (EL) and local generic Exorum (EX) in the chemotherapy of mCCR. METHODS: The retrospective clinicaleconomic analysis of FOLFOX scheme for chemotherapy of mCCR with EL and EX in the real practice has been performed. Fifty case histories (23 with using of EL, 27—EX, was used nomogram of Altman’s) were studied. The calculation of direct cost and cost-effectiveness ratio (CER) based on “partial regress + stabilization” parameter no less than 80% has been performed. RESULTS: For achievement of equal efficacy EL had less number of chemotherapy cycles and total dosage compared with EX (5,0 and 7,3; 670 mg and 900 mg, respectively). Adverse effects were more frequent in EX versus EL (59 and 38, respectively) and caused additional costs and prolonged hospitalization (9 days/patient compared to EL group). The utilitarian EX program cost per patient was less compared to EL by 7,7%. In the same time, CER calculated with total costs due to side effects treatment was practically equal (difference is 1,6% only). Cost prognosis for equal efficacy results with EL using is less by 28,6% versus EX. The alternative scenario has confirmed the cinical-economic added value of EL. CONCLUSIONS: The change of original EL for generic EX in FOLFOX scheme for mCCR has no economic advantages. EL substitution leads to increased number of chemotherapy cycles, higher dose of oxaliplatin, higher rate of adverse effects, and higher costs. PCN93 COST-MINIMIZATION ANALYSIS OF XELOX (CAPECITABINE + OXALIPLATIN) VERSUS FOLFOX-4 (5-FU/LV + OXALIPLATIN) AS ADJUVANT TREATMENT IN STAGE III COLON CANCER UNDER THE BRAZILIAN PRIVATE PAYER PERSPECTIVE Prolla G1, Borges LG2, Santos E2 1 Hospital Mãe de Deus, Porto Alegre, RS, Brazil; 2Roche Brazil, São Paulo, SP, Brazil BACKGROUND: Colorectal cancer is the third leading cancer worldwide (INCA) with nearly 1.2 million cases and about 630,000 deaths expected in 2007 (ACS 2007). In Brazil, it is estimated 28,110 new cases in 2010 (INCA 2010). For patients with stage III colon cancer, the benefits from fluorouracil (5-FU)-based adjuvant chemo- 13th Euro Abstracts therapy are well established and the combination regimens including a fluoropyrimidine + oxaliplatin are the current standard of care. OBJECTIVES: To compare costs of XELOX with FOLFOX-4 as adjuvant treatment for stage III colon cancer under Brazilian private payer perspective. METHODS: Both regimens demonstrated to significantly improve disease-free survival when compared to 5-FU/LV for adjuvant treatment of stage III colon cancer (MOSAIC and XELOXA trials). In the absence of head-to-head trials comparing both regimens, an indirect comparison using Butcher approach (Butcher 1997) was conducted. No difference was found regarding efficacy of regimens (XELOX vs. FOLFOX-4 in disease-free survival: HR 1.03, 95% CI 0.81, 1.29); therefore, a cost-minimization analysis was used. a modified Delphi panel identified local practices to manage severe adverse events (SAEs) of each scheme. Only direct costs were considered for a patient with 1.7 m2. Drug prices were obtained from official public sources (Kairos Magazine, April 2010) and administration costs from medical society physicians fee list (CBHPM2008, v.5). Time horizon was 6 months according to clinical recommendations: eight cycles for XELOX and 12 for FOLFOX4. Discounting was not applied. RESULTS: XELOX is less costly than FOLFOX-4 ($Brz49,862 vs. $Brz57,846). XELOX has higher acquisition costs which is offset by savings in medical resource utilization. Mean acquisition costs for XELOX were R$4185 higher than with FOLFOX-4, but costs to treat SAEs and administration costs were $Brz12,169 higher for FOLFOX-4. One-way sensitivity analysis confirmed the robustness of results. CONCLUSIONS: Findings suggest XELOX as a cost-saving therapy for the adjuvant setting under the private payer perspective in Brazil when compared to FOLFOX-4. PCN94 CAPECITABINE + OXALIPLATIN (XELOX) VS. 5-FU/LV + OXALIPLATIN (FOLFOX4) IN THE ADJUVANT TREATMENT OF PATIENTS WITH COLON CANCER (ACC): COMPARISON OF DIRECT MEDICAL AND SOCIETAL (INDIRECT) COSTS Winterhalder R1, Delmore G2, van Lier M3, Urspruch A4, Hieke K5 1 Luzerner Kantonsspital, Luzern, Switzerland; 2Kantonsspital Thurgau, Frauenfeld, Switzerland; 3 Roche Pharma (Schweiz), Reinach, Switzerland; 4F. Hoffmann-La Roche Ltd., Basel, Switzerland; 5NEOS Health AG, Binningen, Switzerland OBJECTIVES: FOLFOX4 has been the chemotherapy of choice for patients with stage III colon cancer. Recently, the international NO16968 study reported results confirming the efficacy of XELOX in this setting, and evidence suggests that both regimens have at least equivalent efficacy. Therefore, medical and societal resource utilization are important factors for providers, patients, and payers. The objective of this analysis was to compare total costs required to treat an average aCC patient with either XELOX or FOLFOX4 in Switzerland. METHODS: In the absence of a direct comparison, detailed medical resource utilization (MRU) data collected for XELOX from study NO16,968 (aCC) and for FOLFOX4 from study NO16,966 (metastatic colorectal cancer) were analyzed. The FOLFOX4 regimens are identical in both indications; therefore, MRU data from NO16,966 were considered valid proxies. In addition to direct MRU (chemotherapy, hospitalizations due to adverse events [AEs], ambulatory encounters, AE medication, and central venous access [CVA] placements), patient time and travel costs for hospitalizations, ambulatory encounters, and drug administration were estimated. Unit costs were derived from official tariffs (Spezialitätenliste, Tarmed 2010 for drug costs and physician services), official statistics (hospital cost, mean hourly salary) and tax guidelines (travel costs). Total costs while on treatment (24 weeks) for an average patient with aCC were compared. RESULTS: On average, XELOX saved CHF 11,471 per patient versus FOLFOX4. CHF 8883 resulted from savings in direct costs, mainly driven by savings in drug administration (CHF 9312) and CVA placements (CHF 1730). Savings in patient time and travel costs amounted to CHF 2588. CONCLUSIONS: XELOX appears to be cost-saving versus FOLFOX4 in aCC from both a Swiss health-care system and the societal perspective, assuming equivalent efficacy for the two regimens. Considering the high incidence of colon cancer in Switzerland, substantial overall savings may be realized by routine use of XELOX in this indication. PCN95 A MARKOV MODEL TO ESTIMATE THE COST-EFFECTIVENESS OF OMACETAXINE IN CHRONIC MYELOID LEUKEMIA Brereton NJ1, Batty AJ1, Foy CF2, McCormick AL2 1 BresMed Health Solutions, Sheffield, UK; 2Complete Market Access, Macclesfield, UK OBJECTIVES: In patients with chronic myeloid leukemia (CML), first-line treatment with imatinib therapy is beneficial. In cases of imatinib failure, second-generation tyrosine kinase inhibitors (TKIs) are recommended. Omacetaxine has a novel mode of action and acts independently of TKIs; thus, it may have therapeutic advantages for patients who have developed resistance to TKI therapy and have no available treatment options. The objective was to develop a health economic model to estimate the cost-effectiveness of omacetaxine in the treatment of CML. METHODS: A costutility Markov model was developed to capture the progression of CML and treatment effects. The model was developed from the perspective of the French health-care system. Patients entered the model treated either with omacetaxine or standard care, in one of three phases: chronic, accelerated, or blast phase, having failed on imatinib therapy (through resistance or intolerance). Patients then moved to states of response, no response, or death. Survival estimates for nonresponding and responding patients were taken from studies 202 and 203. These were extrapolated using parametric curve fits to estimate survival beyond the end of the trial. Resource use was based on the trial and from the expert opinion of a panel of French clinicians. Unit costs and utilities A269 were elicited from the literature. One-way and probabilistic sensitivity analyses (PSA) were performed. RESULTS: The deterministic results demonstrated that treatment with omacetaxine is cost-effective at a threshold of c30,000. Sensitivity analysis showed that results were most sensitive to cost of omacetaxine, utility score, and survival benefit. PSA results showed that the model was sufficiently robust to parameter uncertainty. CONCLUSIONS: The analysis demonstrated that omacetaxine is cost-effective in the treatment of CML patients who are resistant to TKI therapy and have no available treatment options. PCN96 A UK COST-UTILITY ANALYSIS OF PACLITAXEL ALBUMIN COMPARED TO SOLVENT-BASED PACLITAXEL MONOTHERAPY AND DOCETAXEL MONOTHERAPY FOR PRETREATED METASTATIC BREAST CANCER (MBC) McLeod EJ1, Lloyd A1, Samyshkin Y1, Prunièras F2, Canney P3 1 IMS Health, London, UK; 2ABRAXIS BIOSCIENCE, Paris, France; 3Beatson West of Scotland Cancer Centre, Glasgow, UK OBJECTIVES: Paclitaxel albumin (P-A, Abraxane®) is nanoparticle albumin-bound paclitaxel formulated without use of irritant solvents that are responsible for many of the hypersensitivity and dose-limiting adverse events (AEs). Previous research has compared its cost-effectiveness to solvent-based paclitaxel (S-P) and docetaxel (DOC) in a cohort of patients with mixed treatment history. This study examined P-A’s costeffectiveness for pretreated MBC, the population specified in the European license. METHODS: A Markov model with progression-free, progressed, and mortality states was developed to estimate costs and outcomes over 5 years from a UK NHS perspective. Included from published sources were the costs at 2009 prices of drugs, administration, AEs, and supportive care. Published utility weights were applied to health states to estimate the impact of response, disease progression, and AEs on qualityadjusted life-years (QALYs). Clinical data for pretreated patients receiving P-A 260 mg/m2 3-weekly (q3w) and S-P 175 mg/m2 q3w were from Gradishar (2005). Using Bucher’s methods, an indirect comparison with Jones (2005) provided estimates of clinical parameters for DOC 100 mg/m2 q3w. Weibull extrapolations of survival data generated transition probabilities. RESULTS: Compared to S-P, P-A achieved an extra 0.164 QALYs, 0.263 life-years, and incurred additional costs of £4,137 per patient treated. This translated to an incremental cost-effectiveness ratio of £25,209/ QALY. P-A saved £697 when compared to DOC, with a marginal QALY gain of 0.0037 and no life-expectancy divergence. Probabilistic sensitivity analysis versus DOC indicated a 61% likelihood of P-A satisfying a willingness-to-pay threshold of £30,000/QALY. Both comparisons were sensitive to drug costs and survival estimates. Accounting for potential drug wastage did not influence interpretation of results from either comparison. CONCLUSIONS: The model found that P-A gave better outcomes than S-P or DOC and was cost-effective compared to both interventions. This depended upon greater efficacy than S-P and a more favorable safety profile than DOC. PCN97 A COST-UTILITY ANALYSIS ON THE USE OF TRASTUZUMAB + ANASTROZOLE COMPARED TO LAPATINIB + LETROZOLE, LETROZOLE MONOTHERAPY OR ANASTROZOLE MONOTHERAPY IN THE TREATMENT OF HER2+/HORMONE RECEPTOR POSITIVE (HR+) METASTATIC BREAST CANCER (MBC) FROM THE PERSPECTIVE OF THE UK NATIONAL HEALTH SERVICE (NHS) McNamara S1, Moore L1, Ray J2 1 Roche Products Limited, Welwyn Garden City, Hertfordshire, UK; 2F. Hoffmann-La Roche Ltd., Basel, Switzerland OBJECTIVES: To assess the cost-effectiveness of trastuzumab/anastrozole compared to lapatinib/letrozole, anastrazole, and letrozole for the treatment of HER2+/HR+ mBC patients in whom treatment with an aromatase inhibitor is suitable from a UK NHS perspective. METHODS: An area under the curve model based on the TAnDEM (trastuzumab/anastrozole vs. anastrozole) and EGF30008 (lapatinib/letrozole vs. letrozole) RCTs and the findings of a mixed treatment comparison (MTC) conducted on endocrine treatments in HR+ mBC was developed in Excel. a rank preserving structural failure time (RPSFT) model was utilized to account for the 70% crossover in TAnDEM. In the base-case, no attempt to account for the sizeable additional imbalance in 2nd line chemotherapy was made. The anastrozole PFS and RPSFT-adjusted OS curves from TAnDEM were utilized as a baseline from which to implement the required indirect comparisons under the assumption of an AI “class effect” (as suggested by expert clinical opinion and confirmed by the MTC). The present value of all costs and health outcomes attributable to each treatment option were calculated and the efficiency frontier defined. Extensive deterministic and probabilistic sensitivity analyses were conducted. RESULTS: Anastrozole is dominated by letrozole. Lapatinib/letrozole is extendedly dominated by a combination of letrozole monotherapy and trastuzumab/anastrozole. Trastuzumab/ anastrozole produced the most QALYs of all regimens. Trastuzumab/anastrozole and letrozole define the efficiency frontier with a base-case ICER of £54,336/QALY. The use of the utility values derived from EGF30008 caused this ICER to fall to £44,497/ QALY. CONCLUSIONS: Lapatinib/letrozole is not a cost-effective use of finite NHS resources at any cost-effectiveness threshold. As no attempt was made to account for the imbalance of 2nd line chemotherapy in TAnDEM (31% in anastrozole vs. 8% for trastuzumab/anastrozole) and relatively conservative utility values were used within the model the base-case ICER of trastuzumab/anastrozole vs. letrozole (£54,336/ QALY) should be regarded as conservative and the true ICER likely lies below £50,000/QALY gained. A270 PCN98 COST-EFFECTIVENESS OF A EUROPEAN COMMUNITY-BASED INTERVENTION: “10,000 STEPS GHENT” De Smedt D1, De Cocker K1, Cardon G1, De Bourdeaudhuij I1, Annemans L2 1 Ghent University, Gent, Belgium; 2Ghent University—Brussels University, Ghent, Belgium OBJECTIVES: Physical inactivity is linked with inverse health effects and chronic disease. The aim of this study was to evaluate the cost-effectiveness of the European community-based project “10,000 Steps Ghent.” a published comparative controlled trial showed that the intervention resulted in a significant decrease in sedentary time and a significant increase in step counts (896 steps/day) and self-reported walking time (66 minutes/week). METHODS: A Markov model, with a time horizon of 20 years and a cycle length of 1 year was designed in Excel to estimate the development of diabetes, cardiovascular events, and colorectal cancer. All individuals start in a healthstate free of events. The model transitions were age dependent and based on epidemiological data. The effect of the intervention was based on published relative risk reductions (RRR) related to increased walking time. Costs (from a public payer perspective) and utility decrements related to events were obtained from published literature. To assess the impact of the uncertainty of the parameters on incremental costs and QALYs one-way sensitivity analyses and a Monte Carlo analysis were performed. RESULTS: Implementing the community-based program increased average QALYs with 0.14 to 12.50 QALY and decreased the total costs with approximately c490 to c2749. Hence, the intervention program was dominant. One-way sensitivity analyses indicated that relative risk reductions had the most pronounced effect on the incremental QALYs and costs, however without changing the conclusion of dominance. The results of the Monte Carlo analysis were favorable as well and the intervention, based on 5000 simulations, remained dominant. CONCLUSIONS: The communitybased “10,000 Steps Ghent” campaign is a dominant intervention. Sensitivity analyses have proved the robustness of the results; hence, implementing this intervention on a population-based level could lead to improved health outcomes and reduced costs. PCN99 COST-EFFECTIVENESS OF IMATINIB AS ADJUVANT TREATMENT FOR RESECTED GASTROINTESTINAL STROMAL TUMORS (GIST) VERSUS BEST SUPPORTIVE CARE: CANADIAN PERSPECTIVE El Ouagari K1, Pawar V2, Coombs J3, Rubin J2 1 Novartis Pharmaceuticals Canada, Dorval, QC, Canada; 2i3 Innovus, Medford, MA, USA; 3 Novartis, Florham Park, NJ, USA OBJECTIVES: Clinical studies have highlighted the high risk of recurrence following complete resection of primary GIST. Published data from the phase III (ACOSOG Z9001) trial have demonstrated significant clinical benefit with adjuvant imatinib versus placebo with respect to recurrence-free survival (RFS 98% vs. 83% at 1 year). We conducted a health economic evaluation for imatinib as adjuvant therapy for GIST that can be used to support this indication. METHODS: A Markov model was used to project lifetime outcomes and costs for patients who undergo complete gross resection of primary GIST. Cost-effectiveness was measured in terms of the incremental cost per quality-adjusted life-year (QALY) gained with the addition of imatinib. Probabilities of disease recurrence, resource use, utilities, and costs were derived from ACOSOG Z9001 trial and other secondary sources. Results were generated under three scenarios regarding the treatment duration with imatinib: 1-year, 3-year, and continuous treatment with imatinib. RESULTS: Adding imatinib was projected to result in a gain of 0.745, 1.538, and 5.180 QALYs assuming 1-year, 3-year, and continuous treatment scenarios, respectively. These clinical benefits of imatinib are obtained at an additional expected per-patient lifetime cost of $30,042, $81,125, and $345,360 assuming 1-year, 3-year, and continuous treatment scenarios, respectively. The incremental cost per QALY gained with imatinib was therefore $40,328, $52,760, and $66,669 assuming 1-year, 3-year, and continuous treatment scenarios, respectively. Deterministic sensitivity analyses showed the results to be robust with respect to variations in assumptions and estimates. The probability that imatinib is costeffective given a threshold value of $100,000 per QALY was over 98% in all scenarios. CONCLUSIONS: Results of this evaluation suggest that, from a Canadian health-care system perspective, imatinib is cost-effective and represents good value for the money according to currently accepted standards of cost-effectiveness. PCN100 COST-EFFECTIVENESS OF GEFITINIB VERSUS DOUBLET CHEMOTHERAPY IN FIRST-LINE TREATMENT OF NON-SMALL CELL LUNG CANCER (NSCLC) IN SWEDEN Jacob J, Henriksson M, Brattström D AstraZeneca Nordic MC, Södertälje, Sweden OBJECTIVES: The IPASS study (NCT00322452) showed that in patients with EGFR mutation-positive tumors (EGFRm+), gefitinib significantly increased progression-free survival (PFS) compared with doublet chemotherapy, reducing the risk of progression by 52% (HR 0.48, 95% CI 0.36 to 0.64, P < 0.001) and increasing median PFS by 3.2 months (9.5 months vs. 6.3 months) for the first-line treatment of advanced nonsmall cell lung cancer (NSCLC). The aim of the study reported here was to evaluate the cost-effectiveness of a clinically relevant treatment strategy with gefitinib based on data from IPASS. The strategy with gefitinib involves EGFR mutation testing prior to treatment, followed by selective gefitinib treatment of EGFRm+ patients and doublet chemotherapy for EGFRm- patients and patients with unknown mutation status, and is compared to treating all patients with doublet chemotherapy without mutation testing. METHODS: A Markov model was developed to integrate IPASS study data 13th Euro Abstracts with external data on costs and quality of life. The model estimated costs and QALYs from a lifetime horizon for each treatment strategy. The key clinical data inputs were event rates of PFS and overall survival data. Other important parameters, e.g., prevalence of EGFRm+, cost of EGFR-diagnostics, resource utilization, and utility estimates were retrieved from the literature. RESULTS: The test and treat strategy, including gefitinib, was associated with a QALY gain of 0.0116 at an incremental cost of c300 yielding a cost per QALY gained of c25,900. CONCLUSIONS: This cost-effectiveness analysis of the IPASS study demonstrates that testing patients for EGFR status, followed by gefitinib treatment for EGFR m+ patients is a cost-effective option compared to treating all patients with doublet chemotherapy in a Swedish setting. PCN101 COST-UTILITY ANALYSIS OF DASATINIB AS A SECOND-LINE TREATMENT IN THE CHRONIC PHASE OF CHRONIC MYELOID LEUKAEMIA IN RUSSIA Kuznetzov S1, Mungapen LJ2, Samyshkin Y2, Jakouloff DE3, Sbarigia U4, van Baardewijk M4 1 Haematology Research Centre, Moscow, Russia; 2IMS Health, London, UK; 3BMS, Moscow, Russia; 4BMS, Braine l’Alleud, Belgium OBJECTIVES: To evaluate the cost-effectiveness of dasatinib 100 mg once daily in second-line therapy for chronic myeloid leukemia (CML) patients in the chronic phase (CP) resistant to imatinib 400 mg compared with imatinib 800 mg and nilotinib 800 mg in Russia. METHODS: A Markov cost-utility model was developed to estimate lifetime outcomes and resource use reflecting treatment practice for CML patients in Russia. Treatment efficacy, disease progression, and rates of adverse events in the model were based on published multicenter randomized controlled trials. RESULTS: Dasatinib appeared to be dominant over imatinib and nilotinib in CP-CML in the Russian setting. Incremental life expectancies were 0.17 years and 0.26 years when comparing dasatinib with imatinib and nilotinib, respectively; quality-adjusted lifeyears (QALYs) gains were of 0.18 and 0.22 QALY versus imatinib and nilotinib, respectively. The life-years and QALY gains on dasatinib treatment were due to a larger proportion of patients who achieved complete cytogenic response (CCyR). Mean cost saving per patient over a lifetime horizon with dasatinib were Rubles (RUB) 1,364,220 versus imatinib and RUB 778,621 versus nilotinib. Limitations of the model include a lack of direct comparative efficacy data at licensed doses, which precluded formal indirect comparison. CONCLUSIONS: Dasatinib was a dominating strategy, resulting in outcome gains (greater life expectancy and greater quality-adjusted life expectancy) and cost saving compared both to nilotinib and high-dose imatinib in CP-CML patients. Expanding access to new tyrosine kinase inhibitors for the treatment of CP-CML in Russia would ensure a greater choice of modern and effective therapies. PCN102 COST-EFFECTIVENESS OF DIGITAL MAMMOGRAPHY IN A BREAST CANCER POPULATION-BASED SCREENING PROGRAM Comas M1, Arrospide A2, Mar J2, Roman R1, Sala M1, Hernandez C1, Castells X 1 Hospital del Mar-IMIM, CIBER de Epidemiología y Salud Pública (CIBERESP), Barcelona, Spain; 2Hospital Alto Deba, Mondragon, Spain OBJECTIVES: The introduction of digital mammography presents benefits at the technological. However, there are doubts about its impact on the effectiveness of breast cancer screening. The aim of this study was to analyze the cost-effectiveness of the introduction of digital mammography in a population-based breast cancer screening program. METHODS: A discrete-event simulation model was implemented including the processes under a breast cancer screening program and the natural history of breast cancer. The screening events included: invitation (biennial) of the target population (women aged 50–69 years), participation, screening test, confirmatory tests after a positive mammography result, cancer diagnosis, and treatment. Natural history of breast cancer included the following health states: no cancer, preclinical (nonsymptomatic) cancer, clinical (or symptomatic) cancer, and death. Natural history was modeled as time until a change of health state, and health states were managed using attributes in order to condition the sensitivities and specificities of the tests to the current health state of the woman. Interval cancers were also detected according to the health state. Digital and analogical mammography had the same sensitivity, but different specificities were applied according to type of mammography and also initial or successive screening. Cost-effectiveness was calculated under a 20-year screening scenario and five simulations. RESULTS: Simulation started with a target population of 28,020 women. Other 29,552 women were enrolled in the target population during the simulated 20 years. This population resulted in 56,136 screening mammograms. The number confirmatory tests needed was 1864 under analogical mammography and 1724 under digital screening. Screen-detected cancers were 344 with analogical screening and 312 with digital screening. The overall ICER was c349.14. CONCLUSIONS: Results suggest that population-based breast cancer screening with digital mammography is cost-effective. It does not improve the results of conventional analogical mammography, but it reduces the cost in confirmatory tests. PCN103 COST-EFFECTIVENESS ANALYSIS OF ADDING HPV VACCINATION TO CERVICAL CANCER SCREENING PROGRAM IN HUNGARY Vokó Z1, Nagyjanosi L2, Kalo Z1 1 Eötvös Loránd University, Budapest, Hungary; 2Syreon Research Institute, Budapest, Hungary OBJECTIVES: Despite opportunistic and organized screening, mortality of cervical cancer is still high in Hungary in international comparison. The study aimed to 13th Euro Abstracts investigate the cost-effectiveness of including HPV vaccination in the preventive strategy of cervical cancer. METHODS: We developed an interactive Markov model based on modeling disease progression. Data on prevalences, incidences, test characteristics, efficacy of vaccination and of cancer treatments, and quality of life of cervical cancer patients were taken from the scientific literature. Data on the costs of screening and of cancer treatments were provided by the Hungarian National Health Fund. We applied 5% discount rate for both cost and utility values. The effect of herd immunity was taken into account in the model. GDP PPP exchange rate in 2009 was employed to convert local currency to EUR (1 EUR = 158 HUF). RESULTS: We analyzed the cost-effectiveness of different scenarios. In case of the most likely scenario to come through—vaccination at age 12, coverage 80%, vaccine effectiveness against HPV16/18 (responsible for 72% of cervical cancers caused by HPV) being 95%, vaccine is efficient for at least 20 years, the effectiveness of the screening program is not improved radically—the ICER for vaccination was c49,587/QALY compared to no screening. CONCLUSIONS: As disease burden of cervical cancer is high in Hungary, including HPV vaccination in the preventive strategy of cervical cancer would result in a cost-effectiveness ratio below the implicit upper threshold (3× GDP/ capita = c49,986 in PPP). PCN104 SOCIETAL COST SAVINGS IN METASTATIC NON-SMALL CELL LUNG CANCER (NSCLC) IN FRANCE Stanisic S1, Vergnenegre A2, Chouaid C3, Mueller E1, Walzer S4 1 Analytica International Inc., Loerrach, Germany; 2SIME, Limoges, France; 3Hôpital St. Antoine, Paris, France; 4F. Hoffmann-La Roche Pharmaceuticals AG, Basel, Switzerland OBJECTIVES: Bevacizumab in combination with platinum-based chemotherapy improves progression-free survival (PFS) over chemotherapy alone in advanced or metastatic NSCLC. The aim of this analysis was to assess potential cost savings for society derived from the clinical benefits of bevacizumab-based therapy in metastatic NSCLC patients returning to work in France. METHODS: Indirect cost calculation was performed using human capital approach in the population of progression free metastatic NSCLC patients. Clinical outcomes in terms of PFS from the two bevacizumab phase III trials (Reck et al. 2009, Sandler et al. 2006) were applied to the incidence cases of nonsquamous metastatic NSCLC patients in France, taking into account only age groups eligible to work. Clinical experts were consulted to estimate the percentage of PFS patients who return to work after the induction phase. It was estimated that 20% of PFS patients, below 55 years of age and with performance status 0–1 would maintain their prior employment status (60% part time, 40% full time). Cost savings were analyzed for 1 year and 1.5 year time horizons applying French labor costs. RESULTS: Indirect cost savings per patient returning to work were c21,667 at year 1 and c39,001 at year 1.5. a change in employment pattern (40% part time, 60% full time) led to an approximate 14% increase of savings for both time horizons (c24,763 at year 1 and c44,573 at year 1.5). a 10% alteration of labor cost in either direction resulted in cost savings that were 10% higher/lower. CONCLUSIONS: Bevacizumab-based therapy could result in considerable cost savings in progression-free NSCLC patients. This societal economic benefit adds to the patients’ clinical benefit of increased survival outcomes. PCN105 COST SAVINGS ASSOCIATED WITH BEVACIZUMAB-BASED THERAPY IN PATIENTS WITH METASTATIC NON-SMALL CELL LUNG CANCER (NSCLC) IN GERMANY Schmidt E1, Bischoff HG2, Heigener DF3, Stanisic S1, Walzer S4 1 Analytica International Inc., Loerrach, Germany; 2Thoraxklinik Heidelberg GmbH, Hedelberg, Germany; 3Krankenhaus Grosshansdorf, Grosshansdorf, Germany; 4F. Hoffmann-La Roche Pharmaceuticals AG, Basel, Switzerland OBJECTIVES: First-line treatment of advanced or metastatic NSCLC patients with bevacizumab demonstrated significantly improved progression-free survival (PFS) when added to platinum-based chemotherapy (Reck 2009; Sandler 2006). The aim of this analysis was to investigate whether this clinical benefit would translate into cost savings due to reduced productivity losses in metastatic NSCLC patients who return to work while treated with bevacizumab-based therapy in Germany. METHODS: Potential cost savings were calculated on the basis of reduced productivity losses for patients returning back to work during PFS periods of 1 and/or 1.5 years, applying the human capital approach. The percentage of progression-free patients with ECOG performance status 0 and 1 was derived from the two bevacizumab phase III trials. Epidemiologic data as well as country-specific employment rates were applied stepwise to estimate the metastatic NSCLC patient population eligible to work and employed at diagnosis. It was assumed that 20% of progression-free patients would return to work after the induction phase while maintaining their prior employment status (40% full time). Cost savings related to productivity losses were calculated by applying German labor cost data to the derived population of interest. Results were weighted according to prescribing patterns of bevacizumab combination regimens (bevacizumab/cisplatin/gemcitabine or bevacizumab/carboplatin/paclitaxel). Sensitivity analyses were performed for employment patterns and labor cost. RESULTS: Mean cost savings were c21,171 at year 1 and c38,107 at year 1.5 per progression-free patient returning to work in Germany. a change in employment pattern (60% full time) led to an approximate 14% increase of savings for both time horizons (c24,195 at year 1 and c43,551 at year 1.5). a 10% alteration of labor cost in either direction resulted in cost savings that were 10% higher/lower. CONCLUSIONS: In addition to its PFS benefits bevacizumab-based therapy could result in considerable cost savings in progression free metastatic NSCLC patients. A271 PCN106 INDIRECT COST IN NON-SMALL CELL LUNG CANCER (NSCLC): A SYSTEMATIC LITERATURE REVIEW Schmidt E1, Stanisic S1, Neumann M1, Walzer S2 1 Analytica International Inc., Loerrach, Germany; 2F. Hoffmann-La Roche Pharmaceuticals AG, Basel, Switzerland OBJECTIVES: Novel treatment options for NSCLC patients were shown to improve survival outcomes. Thus, it is of interest whether published evidence can be found on work productivity and employment consequences for this patient group and their informal caregivers, and on respective indirect cost data. METHODS: A systematic literature search from 1990 onward was performed using 19 bibliographic databases available via DIMDI (German Institute of Medical Documentation and Information), supplemented by searches for other publicly available sources such as abstracts at major health economic and clinical conferences, HTAs, and further gray literature. RESULTS: So far, published indirect cost data are elusive and rather limited to overall lung cancer. In lung cancer, indirect cost accounted for the major portion of total estimated cost, e.g., for Germany 78% (Ruff 2000) to 89% (Weißflog 2001). Indirect costs per case seem to increase with stage. In an Australian study, overall stage IV lung cancer patients had direct and indirect costs that were 19% higher than the average (Bishop 2009). The only indirect cost information on advanced NSCLC was reported by Romanus (2008) for the United States (mean wage loss per newly diagnosed end-stage IV NSCLC patient was $1697 [$126–$3371] in the first few months after diagnosis) and by Perrone (2004) for Italy (mean productivity loss per patient of c60,263). Employment rates of NSCLC patients were reported to be about 13% of NSCLC stage IIIb or IV, treated with 2nd line chemotherapy in Italy (Gridelli 2007). In the same population, the informal care given by the principal caregiver was the main assistance cost item with average 3-month cost of c2368 representing 74% of total assistance cost. CONCLUSIONS: The indirect cost burden of NSCLC has scarcely been assessed so far and needs further quantitative investigation, particularly in view of the use of new treatment options. PCN107 FIRST-LINE TREATMENT OF METASTATIC NON-SMALL CELL LUNG CANCER (NSCLC) WITH BEVACIZUMAB AND SOCIETAL COST SAVINGS IN SPAIN Stanisic S1, Castro Carpeño JD2, Walzer S3 1 Analytica International Inc., Loerrach, Germany; 2La Paz University Hospital, Madrid, Spain; 3 F. Hoffmann-La Roche Pharmaceuticals AG, Basel, Switzerland OBJECTIVES: Bevacizumab-based therapy as first-line treatment of advanced or metastatic NSCLC is associated with longer progression-free survival (PFS) when compared with chemotherapy alone (Sandler et al. 2006; Reck et al. 2009). This study analyzes the implications of the clinical benefits of bevacizumab-based treatment for the societal costs in Spain. METHODS: Clinical inputs (PFS) were obtained from two clinical trials (E4599 and AVAiL) comparing a bevacizumab regimen (combined with carboplatin/paclitaxel or cisplatin/gemcitabine) with chemotherapy alone. Percentage of patients returning to work was estimated as 20% of all progression-free patients with ECOG performance status 0–1 in the age of ≤55 years. PFS percentage was applied to the incidence estimation for nonsquamous metastatic NSCLC in Spain after applying specific employment rate. Indirect costs were calculated using the human capital approach. Spanish labor costs available at EUROSTAT were applied after inflating to 2009 values. Employment patterns were suggested by the clinical experts (full time/part time), and were applied to the estimation of patients returning to work. Time frames for the calculations were 12 months and 18 months. RESULTS: Average cost saving per progression-free patient was c12,401 at 12 months and c22,322 at 18 months. Sensitivity analysis on changes in employment patterns from 40% full time/60% part time to 60% full time/40% part time increased cost savings per patient to c14,173 and c25,511 at 12 and 18 months, respectively. CONCLUSIONS: Treatment with bevacizumab-based therapy results in the reduction of productivity losses due to improved clinical benefits and hereby leads to considerable societal cost savings. PCN108 SOCIETAL COST SAVINGS ASSOCIATED WITH BEVACIZUMAB-BASED TREATMENT IN NON-SMALL CELL LUNG CANCER (NSCLC) PATIENTS IN ITALY Mueller E1, Nuijten M2, Ravera S3, Stanisic S1, Walzer S4 1 Analytica International Inc., Loerrach, Germany; 2Ars Accessus Medica, LG Jisp, Amsterdam, The Netherlands; 3Roche S.p.A., Milano, Italy; 4F. Hoffmann-La Roche Pharmaceuticals AG, Basel, Switzerland OBJECTIVES: The high societal cost of lung cancer derives mostly from productivity losses associated with premature retirement and premature mortality (Weißflog et al. 2001). Bevacizumab-based therapy is associated with significantly improved progression-free survival (PFS) over chemotherapy alone in patients with advanced or metastatic NSCLC (Reck et al. 2009; Sandler et al. 2006), the most common form of lung cancer. The aim of this analysis was to assess the potential cost savings for Italy resulting from a higher number of progression-free metastatic NSCLC patients returning to work while treated with bevacizumab-based therapy. METHODS: Productivity losses were calculated on the basis of number of days lost due to illness. Country labor costs and employment rates were taken from EUROSTAT. It was assumed, based on expert opinion, that 20% of progression-free patients with ECOG performance status 0–1 and eligible to work would go back to their original employment status after the induction therapy. The percentage of PFS patients was calculated from two clinical A272 trials (Reck et al. 2009; Sandler et al. 2006) at two time points from baseline: 12 and 18 months. PFS percentages were applied to the incidence estimations for metastatic NSCLC in Italy. Incidence of nonsquamous metastatic NSCLC was estimated as 37% of lung cancer incidence retrieved from GLOBOCAN. Sensitivity analysis was conducted varying the percentage of employment pattern (part time/full time) and the labor cost. RESULTS: Bevacizumab-based therapy resulted in mean savings due to reduced productivity losses per progression-free patient of c20,089 at 12 months and c36,160 at 18 months. Changes in employment patterns, from 60% full time/40% part time to 80% full time/20% part time, increased cost savings per patient to c22,600 and c40,680 12 and 18 months, respectively. CONCLUSIONS: Bevacizumab-based therapy results in societal cost savings due to improved PFS in metastatic NSCLC patients. PCN109 COMPARISON OF HOSPITALIZATION BURDEN BETWEEN METASTATIC AND NON-METASTATIC MELANOMA PATIENTS IN A US POPULATION Ray S1, Tunceli O2, Kamat SA2, Ganguli A1 1 Abbott Laboratories, Abbott Park, IL, USA; 2HealthCore, Wilmington, DE, USA OBJECTIVES: The clinical burden of melanoma patients worsens significantly upon metastasis. However, little is known regarding how hospitalization burden changes upon metastasis. This study compares the hospitalization burden between metastatic and nonmetastatic melanoma patients in a geographically diverse commercially insured US population. METHODS: Insurance claims (January 1, 2004–June 30, 2009) from the HealthCore Integrated Research Database was used to identify patients aged ≥18 years with ≥2 melanoma claims (ICD-9-CM 172.xx, V10.82). Two mutually exclusive cohorts were formed: 1) metastatic melanoma (MM) cohort with ≥1 claim for metastasis (ICD9-CM 196.x-198.x), and 2) non-MM cohort comprising the remaining patients. The first claims for MM and melanoma were the index claims for the respective cohorts. Hospitalization burden was compared between the two cohorts based on four outcomes. The rate of inpatient visits (per-patient-per-month [PPPM]), and the risk of first hospitalization from index claim, were compared using Poisson and Cox-proportional hazards regression, respectively. The mean days to first hospitalization and the mean length of stay (per-patient-per-month [PPPV] ) were compared using generalized linear models. RESULTS: The study included 17,756 (mean age 54.4 years, 49.69% female) non-MM patients and 636 (mean age 56.3 years, 41.04% female) MM patients. The proportion of patients with ≥1 hospitalization was 2.4 times higher in the MM compared to the non-MM cohort (45% vs. 19%, P < 0.01). Compared to the non-MM cohort, the MM cohort had: four times higher PPPM inpatient visit rate (0.04 vs. 0.01, P < 0.001), threefold greater risk of first hospitalization (HR: 3.01; 95% CI: 2.61–3.46), and shorter mean days to first hospitalization (35 days vs. 309 days, P < 0.001). The mean days of inpatient visit (PPPV) were comparable (3.93 vs. 3.54, P = 0.32). CONCLUSIONS: The hospitalization burden in metastatic melanoma patients is significantly higher than patients whose melanoma has not progressed to metastasis. Treatments that improve progression-free survival may reduce this burden. PCN110 RESOURCE UTILIZATION AND DIAGNOSTIC APPROACHES IN NON-SMALL-CELL LUNG CANCER (NSCLC) ACROSS EUROPE: EPICLIN-LUNG STUDY Vergnenegre A1, Cruciani G2, Thomas M3, Guallar E4, Medina E5, Carrato A6 1 SIME, Limoges, France; 2Ravenna Hospital, Ravenna, Italy; 3Clinic for Thoracic Diseases, University of Heidelberg, Heidelberg, Germany; 4Welch Center for Prevention, Epidemiology, and Clinical Research, Baltimore, MD, USA; 5AstraZeneca Pharmaceuticals, Madrid, Spain; 6Hospital Universitario Ramon y Cajal, Madrid, Spain OBJECTIVES: The EPICLIN-Lung study aims to provide information on NSCLC clinical management across European countries. As part of this study, information on resources associated with the diagnosis of NSCLC from first visit to final diagnosis across Europe was collected and described. METHODS: The EPICLIN-Lung study (NCT00831909) is a noninterventional prospective cohort study conducted in Belgium, France, Germany, Greece, Italy, Portugal, Spain, and Turkey. Patients with confirmed NSCLC attending a participating clinical department for the first time between January and March 2009 were included. Information on the approach to disease diagnosis was recorded at the first visit. Descriptive analyses were performed. RESULTS: A total of 3500 patients were included in the analysis and 2691 (76.9%) presented with symptoms at first visit. The mean age (SD) of patients was 64.5 (± 10.5) years. From first visit to final diagnosis, 3124 (89.3%) patients utilized healthcare resources including radiology (51.5%), general practitioner office (37.7%), other resources (19.3%), oncology (19.0%), surgery (17.9%), emergency room (12.8%), drug administration facilities (1.6%), and blood-specific resources (0.3%). The application of tests for diagnosis occurred in 96.0% of patients and included laboratory tests (77.5%), bronchoscopy (73.9%), fine needle aspiration biopsy (26.5%), other tests (20.5%), and biomarker determination (11.3%). Imaging technology was also used in disease assessment with scanning the most frequently used technology (88.7% of patients), followed by x-ray (65.9%), MRI/CT brain scan (39.5%), PET imaging (34.1%), and other imaging tests (21.5%). CONCLUSIONS: The resource used from first visit to final diagnosis of NSCLC is not uniform within the patient demographic across Europe. In particular, biomarker determination was low but bronchoscopy and laboratory tests were relatively high as was the use of imaging technology. a more uniform approach in the use of these resources may improve the impact of current treatments. More specific results will be presented at the meeting. 13th Euro Abstracts PCN111 HEALTH-CARE RESOURCE UTILIZATION IN ADVANCED MELANOMA: AN ANALYSIS FROM THE MELODY OBSERVATIONAL STUDY Lorigan P1, Maio M2, Middleton M3, Testori A4, Bédane C5, van Baardewijk M6, Konto C7, Dueymes A8, Lebbe C9 1 The Christie NHS Foundation Trust, Manchester, Cheshire, UK; 2University Hospital of Siena, Siena, Italy; 3Churchill Hospital, Oxford, UK; 4Istituto Europeo di Oncologia, Milan, Italy; 5Hopital Dupuytren, Limoges, France; 6Bristol-Myers Squibb, Brussels, Belgium; 7 Bristol-Myers Squibb, Paris, France; 8PAREXEL International, Semoy, France; 9Hopital Saint Louis, Paris, France OBJECTIVES: We conducted this study to document the health-care resource utilization associated with treatment of patients with advanced melanoma. METHODS: MELODY (Melanoma treatment patterns and outcomes among patients with unresectable stage III or stage IV disease: a retrospective longitudinal survey) is an observational study managed at 31 centers in France, Italy, and the UK. Eligible patients had attended at one of the sites with a diagnosis of unresectable stage III or IV melanoma between July 1, 2005 and June 30, 2006; data were retrieved from diagnosis (no limit date) until 2008. The primary objective was to document the first-line treatments received by patients. Secondary objectives included ascertaining health-care resource utilization related to up to three lines of treatment (anticancer and supportive care). Data were collected from patients (n = 776) using a case-report form that included information on hospitalizations, outpatient visits, hospice care, and adverseevent management (transfusions and concomitant medications including antiemetics and growth factors). Resource use data were collected from patients (n = 606) that received systemic treatment outside a clinical trial and/or supportive care. RESULTS: Twenty-nine percent (176/606) of patients required medical management for treatment-related adverse events and 32% (195/606) were hospitalized while receiving systemic treatment and/or supportive care with 25% of these having at least four hospitalizations. The median duration of hospitalization was 17 days, with 25% spending at least 29 days in hospital (may comprise multiple stays). The hospitalization rate was higher in patients receiving supportive care than those receiving anticancer treatment (86/170; 51% vs. 140/553; 25%), but median duration of hospitalization was similar (15.0 vs. 14.5 days). Results by line of anticancer therapy and supportive care will be presented. CONCLUSIONS: These results from MELODY suggest that the systemic and palliative treatments used to manage advanced melanoma are associated with considerable resource utilization, highlighting the need for more effective treatment options. PCN112 PATIENT CHARACTERISTICS, MEDICATION USE, RESOURCE USE AND MEDICAL COSTS AMONG PATIENTS WITH METASTATIC PANCREATIC CANCER (MPC) Oglesby A1, Lage M2 1 Amgen, Thousand Oaks, CA, USA; 2HealthMetrics Outcomes Research, Groton, CT, USA OBJECTIVES: To examine the patient characteristics, cancer therapy use, and costs among patients with mPC. METHODS: Subjects in the Marketscan Commerical Claims and Encounters and Supplemental Medicare databases (July 1, 2004–June 30, 2008) were included if they received a diagnosis of metastastasis (ICD-9 codes: 197.x—199.x) on or after the first occurrence of PC (ICD-9 code: 157.x), had no claims for other secondary metastases within 6 months of the initial mPC claim (e.g. index date), and had continuous insurance coverage from 6 months prior through at least 1 month post index. Outpatient cancer therapies and costs are described from mPC index until loss to follow-up or end of the data collection period. RESULTS: There were 4938 subjects with incident mPC included in the analysis. The mean age at index date was 64.8 years (SD = 11.8) and 53.8% were male. Subjects were followed an average of 8.3 months post index. Among patients with incident mPC, 53.3% received outpatient chemotherapy with gemcitabine (76.4%) being the most commonly received therapy. 20.1% of mPC subjects received the targeted agent, erlotinib, post-index, with 63.6% of its use occurring in combination with gemcitabine. Subjects who received gemcitabine, received an average of 9.5 doses (SD: 8.8) over 108.9 days (SD: 129.5). Subjects who received erlotinib, filled an average of 3.9 prescriptions (SD: 4.1) following mPC index. Post-metastases, 84.1% and 73.3% of subjects were hospitalized or visited the ER respectively. The mean cost per month of treating mPC subjects was $16,192 (SD = $21,639), with the majority of these costs attributed to inpatient stays (57.8%) and outpatient visits (35.0%). Outpatient cancer therapies contributed 5.3% of the mean monthly cost post index. CONCLUSIONS: Gemcitabine was the predominant cancer therapy among subjects with incident mPC. Outpatient cancer drugs constituted a small portion of the total mean monthly cost of mPC. 13th Euro Abstracts PCN113 TREATMENT PATTERNS AND OUTCOMES IN THE FRENCH COHORT OF PATIENTS WITH UNRESECTABLE STAGE III OR STAGE IV MELANOMA (MELODY STUDY): A RETROSPECTIVE LONGITUDINAL SURVEY Bédane C1, Leccia MT2, Sassolas B3, Mansard S4, Guillot B5, Mortier L6, Robert C7, Saiag P8, Truchetet F9, Oukessou A10, Bregman B10, Lebbé C11 1 Hopital Dupuytren, Limoges, France; 2Hôpital Michallon, La Tronche, France; 3Hôpital Morvan, Brest, France; 4Hôtel-Dieu, Clermont-Ferrand, France; 5Hôpital St-Eloi, Montpellier, France; 6Hôpital Claude Huriez, Lille, France; 7Institut Gustave Roussy, Villejuif, France; 8 Hôpital Ambroise Paré, Boulogne-Billancourt, France; 9Hôpital Beauregard, Thionville, France; 10Bristol-Myers Squibb, Rueil-Malmaison, France; 11Hôpital St-Louis, Paris, France OBJECTIVES: Melanoma is the first mortality cause by skin cancer. The more advanced stages prognosis is remaining poor. The MELODY study had as primary objectives to describe the disease characteristics, treatment modalities/outcomes, and resources use in patients with unresectable Stage III-Stage IV (UNRSIII-SIV) melanoma. METHODS: MELODY was a three-country (France, Italy, UK) longitudinal, retrospective, observational survey. In the French cohort, 10 expert dermatology departments had to register all melanomas seen between July 2005 and June 2006. The UNRSIII-SIV cases, with minimum 2-month follow-up, were extracted for detailed description of the disease characteristics, treatments (systemic, local, supportive care), and outcomes and resource use, until May 1, 2008 or death. RESULTS: In France, 1224 patients (pts) were registered, in which 278 UNRSIII-SIV cases were extracted. At initial diagnosis, 34/1224 pts (2.8%) were UNRSIII-SIV; 253 pts (91%) received systemic treatment, while 230 (91%) had a chemotherapy of any line. In first-line systemic (n = 249), 198 pts (80%) were treated outside clinical trials, in whom 119 (60%) were on dacarbazine (D) and 35 (18%) on fotemustine (F) monotherapy; complete/partial response was noted for 21 pts (11%), with 2.9-month median response duration, median of progression-free survival (PFS) was 2.8 months (2.6; 3.5). In second line, 159 pts (57%) were treated, in whom 139 (87%) outside of clinical trials, with 75 (54%) on F, and 16 (12%) on D; median PFS was 2.5 months (2.1; 3.9). In case of systemic treatment/supportive care, there was a median of 20 (11; 30) hospitalization days (13 [6; 22] days for first line systemic) and 3 [1; 6] outpatient visits (1.5 [1.0; 5.0] for first-line systemic). CONCLUSIONS: In the French cohort of MELODY, chemotherapy treatments gave modest results in terms of complete/partial response and PFS, at a high cost of resource use, highlighting the unmet medical need. CANCER – Patient-Reported Outcomes Studies PCN114 VISITING DOCTORS WITH MALIGNANT MELANOMA 1 1 1 1 2 Kívés Z , Boncz I , Kriszbacher I , Németh T , Sándor J 1 University of Pécs, Pécs, Hungary; 2University of Debrecen, Debrecen, Hungary OBJECTIVES: The incidence of malignant melanoma has been continuously growing; however, early identification contributes significantly to its effective treatment. As a prerequisite, the individual needs to see a specialist soon after the recognition of the first symptoms. Therefore, the aim of our study is to disclose the factors and patient attitude behind seeing a specialist in case of skin deformation in order to establish the method for effective treatment. METHODS: Retrospective cross-sectional study was carried out among patients in four cities (n = 280) who had been diagnosed with malignant melanoma within 5 years. Besides sociodemographic data, the questionnaire focused on the awareness and knowledge of the patient regarding melanoma, on risk factors, the general attitude of seeing doctors, and that of delaying factors in respect of melanoma. Single and multiple logistic regression were used in statistical analysis. RESULTS: After the recognition of skin deformation, only half of the patients visit a doctor within 1 year. The increase of latency (seeing a specialist more than 3 months after the incidence of the first symptoms) was four to seven times more frequent among those who demonstrate anxiety and distress, feel exposed to doctors, lie in the sun without any control, lack awareness of the disease, and have a lower rate of social support (P < 0.01). The study also revealed the shortages of primary care (e.g., insufficient examination, lack of information supply). Seeing a specialist in most of the cases was motivated by external factors. CONCLUSIONS: To estimate the benefit of developing programs focusing on health behavior and awareness building of the prevalence and incidence of malignant melanoma instead of facilitating several preventive interventional projects based on the screening of the population. PCN115 EVALUATION OF SUPPORT SERVICES TO COLORECTAL CANCER PATIENTS IN MEXICO Kuo KL1, Encarnacion V2, Monzalvo B2, Hernández-Cadena L3, Oderda G1, Brixner D1, Zapata L4 1 University of Utah, Salt Lake City, UT, USA; 2Centro Integral Farmacéutico, Col. Cubitos, Pachuca, Hidalgo, Mexico; 3Instituto Nacional de Salud Pública, Mexico, Cerrada Los Pinos y Caminera, Cuernavaca Mor., Mexico; 4Guia Mark, Mexico City, D.F., Mexico OBJECTIVES: To identify the association of sociodemographic, resource use, and adverse events variables with adherence to measure the return on investment of a patient education service for colorectal cancer patients taking capecitabine. A273 METHODS: This was a prospective intervention of government patients in the country of Mexico of any age, with colorectal cancer and capecitabine treatment. The intervention group had access to a hot line, booklets, nutritional and psychological support; control group patients did not. a structured ad-hoc questionnaire applied the Morisky–Green Test to determine adherence to therapy via follow-up phone calls. Associations between variables and adherence between intervention group and controls were assessed using chi-square test. Emergency rooms (ER) visits were the economic indicator to establish cost differences between groups. Power analysis was conducted using PS Power and Sample Size Calculations version 3.0, January 2009. RESULTS: A total of 220 patients were included in this study (mean age 58.4 [26 to 89] years), 121 intervention patients (48 responded to first call; 70% adherence) and 99 control patients (30 responded to first call; 57% adherence). If the true odds ratio for adherence in intervention relative to control is 1.86, a power analysis projects statistical significance with 290 intervention patients and 181 controls to reject the null hypothesis with a probability (power) of 0.8. The reason for nonadherence “Does not know administration regime” did show statistical significance (P = 0.019) between the intervention (10% n = 2) and control group (47% n = 7). For ER visits related to adverse events, the control group rate was 13.33% (n = 4) versus 2.32% (n = 1) in the intervention group (P = 0.067). The cost for ER/day is $Mex 1,144.44 (USD 86.82). CONCLUSIONS: These preliminary results indicate that patients who receive intervention services may have improved adherence to medication and therefore fewer ER visits and costs. Continued enrollment will be able to determine true significance. PCN116 ANALYSIS OF THE EQ-5D QUESTIONNAIRE FOR PATIENTS WITH ADVANCED HER2-POSITIVE GASTRIC CANCER (HER2+ AGC) BASED ON THE TRASTUZUMAB FOR GASTRIC CANCER (TOGA) STUDY Aultman R, Urspruch A F. Hoffmann-La Roche Ltd., Basel, Switzerland OBJECTIVES: In the ToGA study, patients with HER2+ aGC (n = 584) who received trastuzumab + chemotherapy achieved significantly improved overall survival compared to patients receiving chemotherapy alone (HR = 0.74 (95% CI 0.60; 0.91, P = 0.0046; van Cutsem et al. 2009). Since published utilities in gastric cancer are rare, we report results of the EQ-5D analysis from ToGA here, particularly the obtained utility for progression-free-survival (PFS). METHODS: The EQ-5D was provided to patients at baseline and every 3 weeks until disease progression. The questionnaire was completed by the patient prior to any study specific assessment or discussion with their study care provider. Patients’ EQ-5D responses were scored using UK-tariffs, according to the guidelines of the EuroQol group. Only fully completed questionnaires were included. Mixed effects random slope models in SAS v8.2 were developed to estimate the health state utility. The intercept and day of assessment were incorporated as random effects. Covariance structures were assessed for goodness of fit using Akaike Information Criteria. Secondary analyses that included a treatment indicator were performed to explore a possible treatment effect. EQ-5D analysis of a patient subgroup (with tumors highly expressing the HER2-protein, i.e., IHC3+ or IHC2+ and FISH+, metastatic at screening, n = 431), was also conducted. RESULTS: The unstructured covariance matrix was found to be the best fit and resulted into a PFS-utility of 0.73 (95% CI: 0.71; 0.75). The treatment effect was not found to be significant (P = 0.1542); i.e., no difference between the two treatment regimens was identified. Results were confirmed for the subgroup. CONCLUSIONS: Although only HER2+ patients were included in ToGA, the obtained utility value is in line with other published PFSutilities in aGC. Given the large patient numbers, 0.73 can be regarded as a robust estimate of a patient-reported outcome. Further studies in aGC should attempt to assess the utility in patients following disease progression. PCN117 UTILITY ELICITATION STUDY IN THE UK GENERAL PUBLIC FOR LATE STAGE CHRONIC LYMPHOCYTIC LEUKEMIA Tolley K1, Goad C2, Yi Y2, Maroudas PA3, Thompson G3 1 Tolley Health Economics, Buxton, UK; 2Mapi Values, Bollington, UK; 3GlaxoSmithKline, Uxbridge, UK OBJECTIVES: In the UK, chronic lymphocytic leukemia (CLL) makes up 40% of all leukemias in patients over 65 years. The objective of this UK-based study was to obtain societal preferences for “progression-free” and “progressive” stages of late-stage CLL, and selected treatment-related adverse events. METHODS: A utility study, using the time-trade off (TTO) method, was conducted in the UK with 110 members of the general public for a baseline disease state (before treatment), three primary disease states (progression-free survival [PFS] response, PFS non-response, and disease progression) and four adverse event (AE) substates (PFS response with thrombocytopenia, neutropenia or severe infection, and PFS non-response with severe infection). Disease state vignettes were developed using literature and validated by expert CLL clinicians. Face-to-face interviews were conducted by trained interviewers. The TTO scores were converted into a utility value for each disease state and disutilities were calculated for AEs. Visual analogue scale (VAS) scores were also obtained. RESULTS: All participants were included in the analysis. The mean utility scores from the TTO for the primary disease states were: baseline: 0.549; PFS response: 0.671; PFS nonresponse: 0.394; and progression: 0.294. The mean TTO utility (disutility) scores for the AEs were: PFS response with thrombocytopenia, neutropenia, or infection, 0.563 (−0.108), 0.508 (−0.163), 0.476 (−0.195), respectively; PFS nonresponse with infection, 0.333 (−0.061). The VAS results were in line with the TTO results. CONCLUSIONS: A274 Overall, the utility was higher for the PFS state than for baseline, but decreased below baseline in nonresponse and disease progression states. AEs had an important impact on utility within the PFS response state. The severe infection AE appeared to have a greater impact on patients responding to treatment compared to nonresponders, which may be related to the quality of life which is already low for the latter. PCN118 UTILITY VALUES FOR CHRONIC MYELOID LEUKAEMIA-CHRONIC PHASE (CML-CP) HEALTH STATES FROM THE GENERAL PUBLIC IN THE UNITED KINGDOM Guest J1, Naik N1, Coombs J2, Gray L3, Jenkins A3 1 Catalyst Health Economics Consultants, Northwood, Middlsex, UK; 2Novartis, Florham Park, NJ, USA; 3Novartis Pharmaceuticals UK Limited, Frimley, Surrey, UK OBJECTIVES: To estimate utility values associated with CML-CP health states among members of the general public in the UK. METHODS: Interviewer-administered time trade-off utilities were elicited for four CML-CP health states related to risk of progression, from a random sample of 241 members of the general public from eight cities across the UK, using health-state descriptions validated by clinicians and members of the general public. Mean utility values with 95% confidence intervals (CI) were calculated for each health state. RESULTS: The respondents’ mean age was 45 years and 51% were female. Seven percent (n = 18) of respondents had a cancer at the time of the interview which had been diagnosed for a mean 7.0 ± 6.5 years. The mean utilities with 95% CI were: 0.72 (0.69; 0.75) for untreated chronic phase CML, 0.80 (0.79; 0.82) for hematologic response, 0.89 (0.87; 0.90) for cytogenetic response, and 0.94 (0.94; 0.95) for molecular response. The utility values for each state are significantly different from one another (P < 0.001). The respondents’ preference values for any of the states were not significantly affected by their demographics or whether they had cancer. Nevertheless, the values elicited from respondents with cancer were lower than those elicited from respondents who did not have cancer: 0.65 versus 0.73 for chronic phase CML; 0.72 versus 0.81 for hematologic response; 0.83 versus 0.89 for cytogenetic response; and 0.89 versus 0.95 for molecular response. CONCLUSIONS: The health states with poorer outcome (e.g., hematologic response) were associated with a lower preference value than the state with the best outcome (i.e., molecular response). The data demonstrate the impact that different treatment responses may have on the health-related quality of life of patients with chronic phase CML and can be used to estimate the outcomes of interventions in terms of quality-adjusted life-years. PCN119 COMPARISON OF EQ-5D SCORE BETWEEN TREATMENT WITH 4 CYCLES OF ANTHRACYCLINE FOLLOWED BY 4 CYCLES OF TAXANE AND 8 CYCLES OF TAXANE FOR NODE POSITIVE BREAST CANCER PATIENTS AFTER SURGERY: N-SAS BC 02 TRIAL Shimozuma K1, Shiroiwa T1, Fukuda T2, Mori M3, Ohashi Y2, Watanabe T4 1 Ritsumeikan University, Kusatsu, Shiga, Japan; 2The University of Tokyo, Tokyo, Japan; 3Japan Clinical Research Support Unit, Tokyo, Japan; 4Hamamatsu Oncology Center, Hamamatsu, Shizuoka, Japan OBJECTIVES: We investigated the effect of adjuvant chemotherapy regimens on utility scores assessed by the EQ-5D instrument in a randomized controlled trial for breast cancer patients after surgery. METHODS: In the National Surgical Adjuvant Study of Breast Cancer-02 (N-SAS BC 02), 1060 patients were randomly assigned to the following four chemotherapy groups: 1) four cycles of anthracycline (ADM 60 mg/m2 or EPR 75 mg/m2 + CPM 600 mg/m2, q3 wks x 4) followed by paclitaxel (175 mg/m2, q3 wks x 4) (ACP); 2) four cycles of anthracycline followed by docetaxel (75 mg/m2 q3, wks x 4) (ACD); 3) eight cycles of paclitaxel (175 mg/m2, q3 wks x 8) (PTX); and 4) eight cycles of docetaxel (75 mg/m2, q3 wks x 8). The first consecutively registered 300 women were the subjects of the present utility study. Utility scores were assessed using the EQ-5D instrument at baseline, 3rd cycle, 5th cycle, 7th cycle, 7 months, and 1 year. The obtained data were analyzed using a linear mixed model with baseline, time, group, and interaction between time and group as explanatory variables. RESULTS: Missing data was observed between 1.9 and 6.1% of cases depending on the time of measurement. The utility score was significantly lower in the DTX group than in the ACP and ACD groups. In the DTX group, the mean utility score was lowest at 7 months, and it tended to remain low for a long time. In a comparison of the anthracycline and taxane groups, the anthracycline group had significantly higher utility scores. There were no significant differences depending on the type of taxane. The estimated mean utility scores were 0.81, 0.83, 0.79, and 0.76 (ACP, ACD, PTX, and DTX group). CONCLUSIONS: The results of this study will be beneficial not only for clinical decision-making but also for appropriate allocation of medical resources. PCN120 UTILITY AND WORK PRODUCTIVITY DATA FOR ECONOMIC EVALUATION OF BREAST CANCER THERAPIES IN THE NETHERLANDS AND SWEDEN Lloyd A1, Quadri N1, Tamminga H2, Hövels A2 1 Oxford Outcomes Ltd., Oxford, UK; 2GlaxoSmithKline B.V, Zeist, The Netherlands OBJECTIVES: Survival and quality of life (utility) are often the main measure of benefit used in an economic evaluation. Additionally, some decision-makers will consider benefits in terms of work productivity. The present study was designed to estimate utilities and productivity loss for women with metastatic breast cancer (MBC) 13th Euro Abstracts which is Human Epidermal Growth Factor Receptor 2 positive (HER 2+). METHODS: Health-state vignettes describing MBC progressive disease, stable disease, and seven grade 3/4 adverse events (diarrhea, fatigue, anemia, leukopenia, anorexia, decreases in left ventricular ejection fraction [LVEF], and skin rash) were developed based on interviews with women with MBC in the The Netherlands and Sweden and clinicians. a general public sample rated the states (100 men and women in NL; 100 women aged 50+ in Sweden) using the time trade off method. Women (161 The Netherlands, 52 Sweden) who were currently or recently treated for MBC were surveyed using the Work Productivity and Activity Impairment scale regarding the impact of disease on their ability to work. RESULTS: MBC progressive disease and stable disease were rated more highly in Sweden (0.61, 0.81) than the The Netherlands (0.50, 0.69). Utilities for toxicities ranged from 0.52 to 0.69 (Sweden), and 0.47 to 0.66 (NL). The productivity survey identified that women currently receiving treatment reported that their overall productivity was reduced by 69% (NL) and 72% (Sweden); while those who had recently completed therapy reported reductions of 41% (NL) and 40% (Sweden). CONCLUSIONS: This study captured utility and productivity data for the The Netherlands and Sweden regarding the impact of HER 2+ MBC. Important differences in utilities emerged in the study which could impact cost-effectiveness estimates. The productivity survey demonstrated how the negative impact of breast cancer on productivity persists after women have completed their treatment. PCN121 CONFIRMATION OF BRIEF PAIN INVENTORY SHORT FORM (BPI-SF) “WORST PAIN” ITEM CUT-POINT FOR THE ASSESSMENT OF PAIN PROGRESSION IN CASTRATION-RESISTANT PROSTATE CANCER (CRPC) Regnault A1, Gater A2, Battersby C3, Meunier J1, Abetz L2 1 Mapi Values, Lyon, France; 2Mapi Values Ltd, Bollington, Cheshire, UK; 3AstraZeneca R&D Alderley Park, Macclesfield, England, UK OBJECTIVES: Previous studies in cancer patients have found scores of ≥5 on 11-point pain scales to indicate pain that has a significant impact on patients’ lives. This study sought to confirm the adequacy of a ≥5 cutpoint on the BPI-SF “worst pain” item for defining pain progression in CRPC patients using data collected as part of a multinational phase III clinical trial. METHODS: Patients with a BPI-SF worst pain score ≥5 were compared with patients with a score <5 in terms of Functional Assessment of Cancer Therapy—Prostate (FACT-P) subscales and total score and EQ-5D item scores. Exploratory analyses were also conducted to investigate any potential differences within designated regional subgroups of patients. All analyses were performed using treatment-blinded data collected at the first post-baseline trial visit including the above assessments (Week 12). RESULTS: A total of 464 patients completed the BPI-SF at W12 (<5 n = 411, ≥5 n = 53). Mean FACT-P total scores for patients with a BPI-SF worst pain score ≥5 were 24.5 points lower than for patients with a score <5 (91.1 vs. 115.6, P < 0.0001), indicating poorer well-being. Patients with BPI-SF worst pain scores ≥5 consistently had lower scores for all FACT-P subscales (P < 0.0001) except for social well-being. The magnitude of these differences, for all scales, was considerably greater than reported thresholds for meaningful difference. Results for EQ-5D item scores were in a similar direction with significantly greater impairment reported in patients with a BPI-SF worst pain score ≥5 compared with patients with a score <5 (P < 0.0001). Exploratory analyses also revealed similar results across all regional subgroups of patients. CONCLUSIONS: Patient scores ≥5 on the BPI-SF “worst pain” item are associated with significant and meaningful impairments in CRPC patients, thus supporting the adequacy of this cutpoint as an appropriate definition of pain progression in this population. PCN122 DEVELOPMENT OF THE PATIENT-REPORTED VERSION OF THE COMMON TERMINOLOGY CRITERIA FOR ADVERSE EVENTS (PRO-CTCAE) Basch E1, Reeve B2, Cleeland C3, Sloan J4, Schrag D5, Atkinson TM1, Mendoza T3, Hay J1, Abernethy A6, Minasian L7, Kwitkowski V8, Trentacosti AM8, Burke L9, Sit L1, Bruner D10 1 Memorial Sloan-Kettering Cancer Center, New York, NY, USA; 2National Cancer Institute, Bethesda, MD, USA; 3U.T.M.D. Anderson Cancer Center, Houston, TX, USA; 4Mayo Clinic, Rochester, MN, USA; 5Dana-Farber Cancer Institute, Boston, MA, USA; 6Duke University, Durham, NC, USA; 7National Cancer Institute, Rockville, MD, USA; 8U.S. Food and Drug Administration, Silver Spring, MD, USA; 9Center for Drug Evaluation and Research, U.S. Food and Drug Administration, Silver Spring, MD, USA; 10University of Pennsylvania, Philadelphia, PA, USA OBJECTIVES: The standard lexicon for reporting adverse events in National Cancer Institute (NCI) sponsored clinical trials is the Common Terminology Criteria for Adverse Events (CTCAE), which consists of 790 individual items. Currently, all adverse events are reported by clinicians. However, multiple studies have found that clinicians tend to underreport symptom severity and onset compared with patient self-reports. In 2008, the NCI contracted a multi-institution consortium to develop patient versions of CTCAE items and an electronic platform for capturing symptoms from patients and reporting data to health care providers and researchers. METHODS: A committee including clinical investigators, methodologists, patients, and representatives of NCI and FDA systematically identified CTCAE items with a subjective component amenable to patient reporting. Systematic review and analyses of publications and existing symptom survey data sets and questionnaires were conducted to determine optimal formats for questions, response options, and terms for new PRO-CTCAE items. RESULTS: 81 symptoms were identified in the CTCAE to be amenable to patient reporting. The format and content of these items were found to be inappropri- 13th Euro Abstracts ate for patient reporting in their current form, due to mixing of symptom attributes within items and technical jargon. The committee determined that measured attributes for each symptom should include frequency, severity, and activity interference. Questions and response options were standardized, along with plain language terms for each symptom. a web-based platform was developed for administering the new PROCTCAE items. CONCLUSIONS: In response to a charge from the NCI, the PROCTCAE, a patient version of the CTCAE system, has been developed. The prototype is undergoing testing to assess validity, usability, and logistical feasibility in a variety of cancer care settings. The PRO-CTCAE system has the potential to enhance adverse event reporting by integrating patient experiences and can foster consistency of data collection methods across studies. PCN123 DEVELOPMENT OF A COMPUTER-ADAPTIVE PATIENT REPORTED OUTCOME TOOL FOR THE SCREENING FOR PSYCHO-ONCOLOGICAL TREATMENT NEEDS Gamper E, Meraner V, Giesinger J, Kemmler G, Oberguggenberger A, Sperner-Unterweger B, Holzner B Innsbruck Medical University, Innsbruck, Austria OBJECTIVES: Diagnosis and treatment of cancer entail a considerable amount of distress which in a large percentage of patients would require psychooncological treatment (POT). There is though a lack of precise and economic measures for assessing the need for POT. Its detection in clinical routine often depends on procedural and personnel conditions. The aim of the study was the development and implementation of a patient-reported outcome tool for the screening of distress. a further aim is the construction of a computer-adpative testing (CAT) version. METHODS: 115 breast cancer outpatients attending the Department of Gynaecology at Innsbruck Medical University were consecutively included in a pilot study. Logistic regression analysis and ROC analyses identified the most predictive items from a set of questionnaires (EORTC-QLQ-C30, HADS, and Hornheide Screening Instrument) and other additional questions. The development of the CAT screening tool comprises generating an item bank, developing the CAT-algorithm and applying and implementing the CAT in clinical practice. RESULTS: Data from 105 breast cancer patients (mean age 58.8, SD 12.3) were analyzed. The EORTC QLQ-C30 scales Emotional Functioning and Role Functioning as well as the yes–no question after psychiatric/psychological/psychotherapeutic treatment at any point in lifetime showed high predictive power with regard to need for POT (AUC = 0.88; CI 95% 0.82–0.95). The generation of the item bank resulted in 52 prelinimary items which will be subjected to further analyses. CONCLUSIONS: The implementation of precise and effective measures for POT needs is essential for providing comprehensive and high-quality cancer care. CAT methodology contributes to the reduction of patient burden and assessment duration, and increases efficiency as well as measurement precision. PCN124 COMPUTER-ADAPTIVE TESTING OF FATIGUE IN ONCOLOGICAL PATIENTS Giesinger J1, Petersen M2, Groenvold M2, Gamper E1, Conroy T3, King M4, Luckett T4, Arraras J5, Young T6, Verdonck-de Leeuw I7, Kemmler G1, Holzner B1 1 Innsbruck Medical University, Innsbruck, Austria; 2Bispebjerg Hospital, Copenhagen, Denmark; 3Centre Alexis Vautrin, Vandoeuvre les Nancy, France; 4University of Sydney, Sydney, Australia; 5Hospital of Navarre, Pamplona, Spain; 6Mount Vernon Cancer Centre, Northwood, UK; 7VU University, Amsterdam, The Netherlands OBJECTIVES: Computer-adaptive testing (CAT) is an advanced method for measuring patient-reported outcomes. With help of an algorithm, CAT selects the items most relevant for an individual patient from an item bank. Our study aim was the development of a CAT version of the EORTC QLQ-C30 Fatigue scale, its implementation in a software package, and its use in daily clinical routine. METHODS: Our project is part of a large project on CAT development conducted by the EORTC Quality of Life Group. To set up an initial English fatigue item list, an extensive literature research was performed. These items were refined through multistage expert reviews, translated to German, Danish, Spanish, French, and Dutch, and filled in by a pilot patient sample to collect feedback. In a next step, a large patient sample is recruited for all language versions to gain data for development of the item bank and the CAT algorithm. RESULTS: Literature research resulted in 588 fatigue items, from which 44 were selected after comprehensive expert reviews. Based on feedback from 52 oncological patients, wording and translation of several items were revised. Data have been collected from about 1200 patients. Preliminary results of the CAT-analysis will be presented. CONCLUSIONS: By generating individually tailored item sets, CAT reduces patient burden and assessment duration, and increases measurement precision. In addition, electronic data capture increases data quality and reduces the amount of human resources required for data collection. PCN125 ECONOMIC AND PATIENT-REPORTED OUTCOMES OF OUTPATIENT HOME-BASED VERSUS INPATIENT HOSPITAL-BASED CHEMOTHERAPY FOR PATIENTS WITH COLORECTAL CANCER Kang HY1, Joo EH1, Cho E1, Kim HS1, Kim SY2 1 Yonsei University, Seoul, South Korea; 2Korea University, Seoul, South Korea OBJECTIVES: To compare the economic and patient-reported outcomes between outpatient home-based and inpatient hospital-based chemotherapy in advanced colorectal cancer patients. METHODS: A total of 80 patients from Severance Hospital A275 in Seoul, Korea, who had stage III colorectal cancer and underwent home-based (n = 40) or hospital-based chemotherapy (n = 40) with a FOLFOX regimen between January 2007 and April 2008 were enrolled. Patient satisfaction data were collected by a self-administered questionnaire survey. Based on hospital charge records, average cost (in 2008 Korean won) per chemotherapy session was estimated and compared between home- and hospital-based chemotherapy from a societal perspective. RESULTS: Patients receiving chemotherapy at home showed higher satisfaction with their treatment (mean satisfaction score: 3.58 ± 0.15, 5-point Likert-type scale, with a higher score indicating higher satisfaction) than did those treated at the hospital (3.23 ± 0.21; P < 0.01). After adjusting for differences in baseline characteristics between the two groups using multivariate analysis, those receiving home-based chemotherapy still showed significantly higher satisfaction than those undergoing hospital-based therapy (β = 0.271, P < 0.001). Additionally, home-based therapy reduced the cost per chemotherapy session by 16.6%, compared with hospital-based treatment (1,694,216 vs. 2,030,383 Korean won [KW], 1200 KW ≈ 1 US dollar). The largest cost reduction was attributable to medical costs (−201,122 KW), followed by caregiver’s opportunity costs (−135,000 KW). CONCLUSIONS: Higher satisfaction and lower economic cost for home-based chemotherapy suggests that home-based chemotherapy could be a popular and cost-effective treatment option for colorectal cancer patients who are eligible for home-based chemotherapy. PCN126 REVIEW OF PATIENT-REPORTED OUTCOMES IN PHASE II ONCOLOGY CLINICAL TRIALS Danheiser S1, Barkalow F1, Finnern HW2 1 Citeline, New York, NY, USA; 2Boehringer Ingelheim GmbH, Ingelheim, Germany OBJECTIVES: Patients’ own assessment of the impact of anticancer therapy has been considered important from a patient, physician, payor, and regulatory perspective. Approximately 12% of all oncology clinical trials list at least one patient-reported outcome (PRO) measure; 50% of these are phase II trials. Use and publication of PROs results were systematically reviewed for breast, colorectal, ovarian, and non-small cell lung cancer (NSCLC). METHODS: Citeline’s TrialTrove database was searched for oncology trials that included PROs and were planned, ongoing or completed as of October 2009. Specific trial factors examined include phase, disease type, location, sponsorship, type of PRO instrument(s), and publication of PRO endpoint results. RESULTS: Of the 5483 phase II trials retrieved for breast, colorectal, ovarian, and NSCLC, 9% listed PROs. PRO measures were more frequently included in NSCLC (13%) and ovarian cancer (10%) than in breast (7%) or colorectal (8%) cancer phase II trials. Fifty-two percent of these trials were sponsored by industry, with 25% and 19% being sponsored by academic and cooperative groups, respectively. One-third of the trials were conducted in the United States only. PROs were most frequently measured with the EORTC and FACT series of questionnaires. Phase II PRO results were published 24%, 36% and 39% of the time for breast, ovarian, and NSCLC, respectively, and only 18% of the time for colorectal cancer. This compares with an overall publication frequency for PRO trials (any end points) of 58% to 67% which was highest in NSCLC. PRO results were published more frequently for trials conducted in Asia or Europe. However, trial sponsorship did not impact publication frequency. CONCLUSIONS: PROs are infrequently assessed in phase II cancer trials. PRO measures were most frequently included in NSCLC trials and sponsored by industry. PRO results were published less frequently than other end points in these trials. Publication frequency differed by tumor type and trial location. PCN127 MINIMAL CLINICALLY MEANINGFUL DIFFERENCES FOR THE EORTC QLQ-C30 AND EORTC QLQ-BN20 SCALES IN BRAIN CANCER PATIENTS Maringwa J1, Quinten C1, King M2, Ringash J3, Osoba D4, Coens C1, Martinelli F1, Cleeland C5, Flechtner H6, Gotay C7, Greimel E8, Taphoorn M9, Reeve B10, Schmucker-Von Koch J11, Weis J12, Van Den Bent MJ13, Stupp R14, Bottomley A1 1 European Organisation for Research and Treatment of Cancer, Brussels, Belgium; 2 University of Sydney, Sydney, Australia; 3The Princess Margaret Hospital, Toronto, QC, Canada; 4Quality of Life Consulting, West Vancouver, BC, Canada; 5U.T.M.D. Anderson Cancer Center, Houston, TX, USA; 6University of Magdeburg, Magdeburg, Germany; 7 University of British Columbia, Vancouver, BC, Canada; 8Medical University of Graz, Graz, Austria; 9Medisch Centrum Haaglanden—Westeinde, Den Haag, The Netherlands; 10 National Cancer Institute, Bethesda, MD, USA; 11University of Regensburg, Regensburg, Germany; 12University of Freiburg, Freiburg, Germany; 13AZ Rotterdam-Daniel Den Hoed Kliniek, Rotterdam, The Netherlands; 14University Hospital CHUV, Lausanne, Switzerland OBJECTIVES: The aim of this study was to determine the smallest changes in healthrelated quality-of-life (HRQOL) scores in the European Organization for Research and Treatment of Cancer quality of life questionnaire (EORTC QLQ-C30) and the EORTC Brain Cancer Module (QLQ-BN20), which could be considered as clinically meaningful in brain cancer patients. METHODS: World Health Organization (WHO) performance status (PS) and the Mini Mental State Examination (MMSE) were used as clinical anchors appropriate to related subscales to determine minimal clinically important differences (MCID) in HRQOL change scores (range 0–100) in the EORTC QLQ-C30 and QLQ-BN20. a threshold of 0.2SD (small effect) was used to exclude anchor-based MCID estimates considered too small to inform interpretation. RESULTS: Based on WHO PS, our findings support the following integer estimates of the MCID for improvement and deterioration, respectively: physical functioning (6, 9), role functioning (14, 12), cognitive functioning (8, 8), global health status (7, 4*), fatigue (12, 9), and motor dysfunction (4*, 5). Anchoring with MMSE, cognitive A276 functioning MCID estimates for improvement and deterioration were (11, 2*) and those for communication deficit were (9, 7). The estimates with asterisks were less than the 0.2 SD threshold and were therefore excluded from our MCID ranges. Our MCID estimates therefore range from 5 to 14. CONCLUSIONS: These estimates can help clinicians to interpret the clinical relevance of changes in HRQOL over time and, in conjunction with other measures of efficacy, help to assess the value of a health-care intervention. The findings also indicate that more sensitive measures may be needed to detect certain changes. Furthermore, the estimates can be useful in determining sample sizes in the design of future clinical trials. PCN128 EFFECTS AND MEDICAL COSTS OF A STRUCTURED PSYCHOSOCIAL GROUP INTERVENTION FOR BREAST CANCER PATIENTS AFTER SURGERY Shimozuma K1, Shiroiwa T1, Sagara Y2, Tobata R2, Ueo H3, Kubota Y3, Hori T4, Amano K4, Terada S5, Mori M6, Yajima T6, Kurahashi I7, Saito S8, Hosaka T9 1 Ritsumeikan University, Kusatsu, Shiga, Japan; 2Sagara Hospital, Kagoshima, Japan; 3Ueo Breast Clinic, Oita, Japan; 4Shiga Medical Center for Adults, Moriyama, Shiga, Japan; 5Wakaba Association, Kariya, Aichi, Japan; 6Japan Clinical Research Support Unit, Tokyo, Japan; 7The University of Tokyo, Tokyo, Japan; 8Okayama University, Okayama, Japan; 9Tokai University, Tokyo, Japan OBJECTIVES: To clarify the details of psychosocial health status, the effects and related costs of psychosocial group intervention in post-breast cancer surgery patients. METHODS: Structured psychosocial group intervention (90 min/week × five sessions) was conducted for 66 patients aged 20–79 years who were receiving breast cancer treatment at three hospitals in Japan (intervention group). The intervention was conducted 2 weeks to 3 months after radical surgery. HRQOL and psychosocial function were investigated prospectively for outcome. The survey included 1) EORTC QLQ-C30 for HRQOL; and 2) POMS, MAC scale, and a characteristic self-efficacy scale for psychosocial function. It was conducted at registration, week 4 (group therapy conclusion), and month 6. Prior to the intervention study, a group that fulfilled the same eligibility criteria but did not receive psychosocial group intervention was studied (n = 116; nonintervention group). The same outcomes were measured and compared using multivariate analysis adjusted for disease stage, chemotherapy, and hospital. Direct medical costs were obtained from receipt data, and information on travel expenses, direct nonmedical costs, and indirect costs (productivity costs) were collected simultaneously with the HRQOL survey. RESULTS: In HRQOL and psychosocial function, an intervention effect was seen in fatalism (MAC_F) only. In a subgroup analysis by hospital, effects of cognitive function were seen in HRQOL (EORTC_CF), tension-anxiety (POMS_TA), helplessness/hopelessness (MAC_H), and fatalism (MAC_F). However, these effects were not seen at 6 months. Total medical costs including indirect costs during the 6 months were ¥840,000 for the intervention group and ¥750,000 for the nonintervention group. CONCLUSIONS: Structured psychosocial group intervention for post-breast cancer surgery patients had a uniform effect in improving HRQOL and psychosocial function. It is significant that a benefit was obtained in the early postoperative period when HRQOL is most easily damaged. Medical costs did not differ significantly between the groups. PCN129 A DATABASE REVIEW OF PATIENT-REPORTED OUTCOME STUDIES IN EORTC CANCER CLINICAL TRIALS Bottomley A1, Quinten C1, Mauer M1, Taphoorn M2, Flechtner HH3, Koller M4, Ghislain I1, Coens C1 1 European Organisation for Research and Treatment of Cancer, Brussels, Belgium; 2Medisch Centrum Haaglanden—Westeinde, Den Haag, The Netherlands; 3Otto-von-Guericke University Magdeburg, Magdeburg, Germany; 4University Hospital Regensburg, Regensburg, Germany OBJECTIVES: For the last two decades, QL has been increasingly assessed in the EORTC. a detailed database review of all completed and ongoing EORTC QL studies has been undertaken. METHODS: A database of all EORTC QL clinical trial protocols was established in 2000. This regularly updated database comprises all trials where QL has been included as an end point. It summarizes details of closed and ongoing trials, QL tools used, assessment timing, disease and treatment types, compliance, and published outcomes. These were examined and reported. RESULTS: The database contains over 128 EORTC clinical trial protocols over a 15-year period between 1995 and 2010 which included a QL component. This is an average of eight new clinical trials with QL per year with only two having QL as a primary end point. The QLQ-C30 was the instrument of choice in 85% of trials. In the last decade, 15 EORTC clinical groups have been recruiting patients. The majority of trials were conducted by the radiotherapy, brain and breast groups. Over 20,000 patients, mostly from Belgium, France, Germany, and the The Netherlands, have been entered into closed and ongoing multinational trials. Typically, EORTC trials are conducted across an average of 12 countries which include over 85 centers and often involve only a single trial. Several early EORTC trials experienced lower than ideal compliance. This presents a major challenge for the EORTC, but trials in the last 6 years have demonstrated significantly improved compliance. Most QL studies were published as separate papers from the main clinical paper, and over 80% were published in high-impact factor journals (IF > 12). CONCLUSIONS: QL is now a major component of cancer clinical trials, an almost standard secondary end point. This suggests that researchers in the European context increasingly acknowledge the importance of patient-reported outcome assessments for evaluating cancer therapeutic modalities. 13th Euro Abstracts PCN130 QUALITY OF LIFE, OUTCOMES, AND COSTS IN THE BELGIAN POPULATION RECEIVING 90Y-ZEVALIN FOR NON-HODGKIN LYMPHOMA: A STUDY FOR REIMBURSEMENT REVISION Caekelbergh K, Moeremans K IMS Health Consulting, Brussels, Belgium OBJECTIVES: To compare hematological toxicity, costs, health-related quality of life (HR-QOL), and outcomes observed in real life in the Belgian non-Hodgkin lymphoma (NHL) population receiving 90Y-Zevalin, with model-predicted data at reimbursement on the basis of a clinical trial in heavily pretreated NHL. METHODS: Twelve of the 13 centers trained for 90Y-Zevalin administration at the time of reimbursement approval participated. All consecutive patients receiving 90Y-Zevalin for Rituximabrelapsed/refractory NHL were included in this 1-year, multicenter, prospective, observational study. QOL assessments, based on a generic (EQ-5D) and a disease-specific (FACT-LYM) questionnaire, were performed at baseline (date of first administration Rituximab), day 8, month 1, 2, and 3 after baseline. Costing (c, 2008) was based on official tariffs and observed resource use. RESULTS: Over the 2-year recruitment period, 30 patients were included. Only one1 patient received hematopoietic growth factor prophylaxis. Hematological toxicity occurred in 79% of patients, but was generally mild. Grade IV neutropenia was reported in 16% of patients (35% predicted), grade IV thrombocytopenia in 23% (9% predicted), and grade IV anemia was not reported (4% predicted). EQ-5D as well as FACT-LYM assessments suggested no significant changes in HR-Qol over the study period and no impact of hematological toxicity on HR-Qol. The average total cost from the health-care payer perspective, covering costs of drugs, drug administration setting, and prophylaxis, mounted to c16,886 (Standard error 79 c). Hematological toxicity added c1191 (Standard error: c528), constituting 7% of total costs (18,076 c). Observed costs were within 2% from predicted costs. Median time to progression (TTP) was 11.7 months (6.8 months predicted). CONCLUSIONS: This observational study confirmed the predicted tolerability profile, costs and outcomes associated with 90Y-zevalin in a real-life patient population. Furthermore, the patient numbers treated with 90Y-Zevalin in real life were lower than anticipated (maximum 100 predicted). PCN131 HEALTH-RELATED QUALITY OF LIFE (HRQOL) IN CERVICAL CANCER: CLINICAL SIGNIFICANCE AND METHODOLOGICAL ISSUES Siddiqui MK, Gupta J, Sehgal M Heron Health Private Ltd., Chandigarh, India OBJECTIVES: To assess the clinical use of HRQoL instruments in RCTs in cervical carcinoma and evaluate pertinent methodological issues. METHODS: RCTs investigating HRQoL benefit of therapies in cervical carcinoma in adult females, published in English language were included. Cochrane Central Register of Controlled Trials, MEDLINE and EMBASE were searched from inception to 2010. Two reviewers independently assessed the trials for inclusion and performed extraction. RESULTS: Of the 96 RCTs identified, 13 RCTs met the inclusion criteria. Ten trials compared different chemotherapy regimens in first- to third-line treatment. The most commonly used QoL instrument was Functional Assessment of Cancer Therapy (FACT-G) (46% trials). Alternatively, cancer-specific European Organization for Research and Treatment of Cancer-Quality of Life Questionnaire-Core 30 (EORTC QLQ-30), SF-36, and Linear Analogue Scale Assessment (LASA 100 mm) were used either alone or in conjunction with the Brief pain Inventory (BPI). Only one trial reported HRQoL as a primary objective. HRQoL results were presented adequately in 61% trials which were analyzable. Statistically significant differences in HRQoL between treatment groups were studied in 69% studies, but the clinically meaningful difference was examined in only one study. Prognostic value of QoL was assessed in only two trials (15%). Most widely observed methodological issues included lack of priori hypothesis (92% studies) and lack of methods to deal with missing data (85%). Other issues included lack of details on domain, variability in time points of administration of a tool, and patient’s noncompliance, reported in ~60% studies. CONCLUSIONS: HRQoL is included as an outcome in the studies investigating therapies in cervical carcinoma but clinically meaningful interpretations of HRQoL results are rarely considered. The methodological shortcomings in the assessment and analysis of HRQoL outcomes should be examined further to derive clinically meaningful and correlating evidence. PCN132 ERLOTINIB MAINTENANCE THERAPY FOR NON-SMALL CELL LUNG CANCER PRESERVES QUALITY OF LIFE Juhász E1, Kim JH2, Stelmakh L3, Cicenas S4, Klingelschmitt G5 1 Országos Korányi TBC, Budapest, Hungary; 2St. Vincent’s Hospital, Seoul, South Korea; 3I.P. Pavlov State Medical University, St. Petersburg, Russia; 4Vilnius University, Vilnius, Lithuania; 5 F. Hoffmann-La Roche Pharmaceuticals AG, Basel, Switzerland OBJECTIVES: Maintenance therapy can delay progression and prolong survival in metastatic non-small cell lung cancer (mNSCLC). The impact of treatments for mNSCLC on patient quality-of-life (QoL) is an important consideration, as treatment is noncurative and QoL in this population is already compromised. The SATURN study demonstrated that, compared with placebo, erlotinib maintenance therapy improved progression-free survival and overall survival by 41% and 23%, respectively. The impact of erlotinib maintenance therapy on QoL was also evaluated as a secondary end point. METHODS: Patient QoL was assessed until disease progression or withdrawal using the Functional Assessment of Cancer Therapy-Lung (FACT-L) questionnaire. Disease progression was assessed radiographically at regular intervals, often diagnosed prior to symptomatic progression. Patient QoL was analyzed in terms 13th Euro Abstracts of the time to symptom progression (TSP), time to deterioration (TTD) in trial outcome index (TOI), and time to deterioration (TTD) in QoL. An exploratory analysis based on the time to analgesia and appearance of key symptoms (pain, cough, and dyspnea) was also performed. RESULTS: FACT-L completion rates were above 90% at almost all study visits. At baseline, QoL measures were similar between the two treatment groups. Maintenance therapy with erlotinib did not negatively impact on QoL, compared with placebo, as illustrated by comparable TSP (HR = 0.91 [0.74– 1.12], n = 785), TTD in TOI (HR = 1.06 [0.87–1.31], n = 781), or TTD in QoL (HR = 0.96 [0.79–1.16], n = 776). Exploratory analysis of NSCLC-related symptomatology showed that time to pain and time to analgesic use were significantly delayed in patients receiving erlotinib compared with placebo (HR = 0.61 [0.42–0.88]; P = 0.0080 and HR = 0.66 [0.46–0.94]; P = 0.0199, respectively). There was also a nonsignificant trend toward delayed time to cough and time to dyspnea (HR = 0.77 [0.49–1.21] and HR = 0.75 [0.48–1.17], respectively). CONCLUSIONS: Erlotinib maintenance therapy significantly extends progression-free survival, without compromising patient QoL and with some improvement in symptoms. PCN133 EPICLIN-LUNG STUDY: NON-SMALL-CELL LUNG CANCER (NSCLC) PATIENT QUALITY OF LIFE AND HEALTH-STATE ASSESSMENT Thomas M1, Cruciani G2, Vergnenegre A3, Guallar E4, Medina E5, Carrato A6 1 Clinic for Thoracic Diseases, University of Heidelberg, Heidelberg, Germany; 2Ravenna Hospital, Ravenna, Italy; 3SIME, Limoges, France; 4Welch Center for Prevention, Epidemiology, and Clinical Research, Baltimore, MD, USA; 5AstraZeneca Pharmaceuticals, Madrid, Spain; 6Hospital Universitario Ramon y Cajal, Madrid, Spain OBJECTIVES: The aim of the EPICLIN-Lung study is to provide information on the impact and overall resource burden of the diverse strategies used across Europe for the management of NSCLC. Quality of life (QoL) and health state (HS) were determined by patient-reported outcome (PRO) questionnaires. METHODS: The EPICLIN-Lung study (NCT00831909) is a noninterventional prospective cohort study conducted in Belgium, France, Germany, Greece, Italy, Portugal, Spain, and Turkey. Patients with confirmed NSCLC attending a participating clinical department for the first time between January and March 2009 were included. QoL and HS were assessed at baseline (Visit 1) by the responses to the PRO questionnaires FACT-L (Functional Assessment of Cancer Therapy-Lung) and EQ5D (EuroQoL-5D), respectively. RESULTS: Patients (N = 1500) received the questionnaires at Visit 1. QoL data werer available for 1500 patients and overall FACT-L mean score (±SD) was 93.4 (21.3). HS data were available for a total of 1402 patients, overall EQ5D mean score (±SD) was 63.3 (19.9). Mean scores (±SD) split by disease stage IIIb (FACT-L, n = 276; EQ5D, n = 258) and IV (FACT-L, n = 694; EQ5D, n = 653) were 93.5 (20.1) and 91.3 (21.6) for FACT-L, respectively, and 64.2 (18.7) and 61.0 (20.5) for EQ5D, respectively. Mean scores (±SD) split by age <80 (FACT-L, n = 1,428; EQ5D, n = 1,348) and age ≥80 (FACT-L, n = 61; EQ5D, n = 54) were 93.5 (21.4) and 92.9 (18.7) for FACT-L, respectively, and 63.5 (19.9) and 60.4 (19.9) for EQ5D, respectively. Mean scores (±SD) for patients with performance status 0, 1, 2, 3, and 4 were 100.9 (18.7), 94.3 (21.3), 84.9 (20.9), 84.4 (21.0), and 85.0 (20.4), respectively, for FACT-L and 71.6 (17.6), 63.4 (19.3), 57.3 (17.9), 54.5 (21.3), and 55.0 (15.3), respectively, for EQ5D. CONCLUSIONS: This study provides a robust utility that could be used in the final analysis to assess the QoL and HS of NSCLC patients across Europe. More mature results and analysis will be provided at the meeting. PCN134 DEALING WITH CULTURALLY SENSITIVE QUESTIONS IN THE COURSE OF TRANSLATING EORTC QUALITY-OF-LIFE GROUP QUESTIONNAIRES Kulis D1, Jeglikova P1, Greimel E2, Bottomley A1, Koller M3 1 EORTC, Brussels, Belgium; 2Medical University of Graz, Graz, Austria; 3University Hospital Regensburg, Regensburg, Germany OBJECTIVES: The Translation Unit of the EORTC is responsible for coordinating new translations of HRQOL tools. According to the Translation Procedure described in the Translation Manual, one step is pilot-testing, involving 10 to 15 patients who comment on the new translation. This abstract’s aim is to review the difficulties in translating sensitive issues (body image, death, etc.). METHODS: During the pilot testing, patients are interviewed about the translation, filling in answer sheets. a report with their comments is reviewed by the Translation Unit. All questions with comments from at least two patients must be analyzed. The three possible reactions are to accept new translations suggested by the patients, to reword the English item, and provide a new translation or to provide no change (e.g., because the only solution would be to delete the item). Fifteen reports of QLQ-MY20 translations were analyzed. RESULTS: In the most recent 15 translations of QLQ-MY20, pilot-tested on 85 patients in 16 countries, there were five language versions that caused no problem and 10 that received comments about offensiveness or disturbing nature of questions about body image and future perspectives (especially in countries such as China, Hong Kong, Thailand, Lebanon, where body image and death are taboos). All together, there were 82 comments about four items (giving 20 translated questions causing difficulties) which required thorough analysis and discussion. Results of the analysis included changing six translations (rewording, accepting patients’ suggestions), refusing five suggestions (they deviated from the source too much) and leaving nine translations without changes (since there were no suggestions and rewording was impossible). CONCLUSIONS: Scales concerning sexual functions, body image, and future perspectives tend to raise concerns, especially in Asian and Arabic countries. However, such A277 issues are resolved in the pilot-testing stage of the Translation Procedure through discussions, and linguistic and medical analyses of both source and target items. PCN135 IMPROVEMENT IN QUALITY OF LIFE OUTCOMES IN INTERFERONALPHA TREATED PATIENTS COMPARED TO SUNITINIB IN ADVANCE OR METASTATIC RENAL CELL CARCINOMA Goyal R1, Anand R1, Rai MK2 1 Value Edge Research Pvt Ltd., New Delhi, India; 2Cardiff Research Consortium, Capita India Pvt. Ltd, Mumbai, Maharashtra, India OBJECTIVES: The objective was to assess the improvement in quality-of-life outcomes with interferon-alpha compared to sunitinib in treatment-naïve advanced/metastatic renal cell carcinoma in patients. METHODS: Studies were retrieved from Embase, Pubmed, Cochrane, and DARE databases using relevant search strategies. Randomized controlled trials which compared IFN with sunitinib were included according to prespecified inclusion/exclusion criteria. The quality-of-life (QoL) data presented in the studies was independently extracted by two reviewers and differences were reconciled by a third reviewer. All studies were critically appraised and data was analyzed using STATA version 9.2. RESULTS: Of the 463 studies identified, three studies met the inclusion criteria. FKSI-DRS index was reported in one study with baseline and endpoint values as 29.55 and 27.4 with IFN-α and 29.74 and 29.4 with sunitinib (P < 0.0001). FKSI-15 index was reported in two studies with mean baseline and endpoint values as 46.1 and 42.1 with IFN-α and 46.45 and 45.3 with sunitinib (P < 0.0001). FACT-G score was reported in one study. The baseline FACT-G score was 81.25 with IFN-α and 82.3 with sunitinib. The endpoint FACT-G score was 76.8 with IFN-α and 82.3 with sunitinib (P < 0.0001). EQ-5D score was reported in two of the included studies and was reported to be 0.76 and 0.73 as mean baseline and endpoint score for IFN-α group and 0.76 and 0.76 for sunitinib group. EQ-VAS score was reported in one study and was 71.43 and 68.7 as baseline and end point for IFN-α group and 73.8 and 73.4 for sunitinib group. The overall survival rate and response rate was better with sunitinib as compared to IFN-α. CONCLUSIONS: Improvement in quality-of-life outcomes was better in patients treated with IFN-α as compared to sunitinib at the end point. IFN-α continues to remain a treatment of choice despite of limited efficacy and tolerability. PCN136 QUALITY OF LIFE AFTER CHEMOTHERAPY IN BREAST CANCER: A STUDY IN SOUTH OF IRAN Ahmad Kiadaliri A1, Bastani P2 1 Lund University, Malmo, Sweden; 2Iran University of Medical Sciences, Tehran, Iran OBJECTIVES: The aim of this study was to compare the impact of two common treatment of breast cancer on quality of life in women with early stage of breast cancer. METHODS: A double-blind cohort study was done in 100 breast cancer patients with node positive that used 5-fluorouracil, doxorubicin, cyclophosphamide (FAC) or docetaxel, doxorubicin, and cyclophosphamide (TAC) regimen as adjuvant therapies. Patients were followed for 4 months since end of chemotherapy. Health-related quality of life was assessed using questionnaire from European organization for research and treatment of cancer (EORTC) QLQ-C30. Independent t-test analysis was used at the significant level of 0.05 for analyzing the results. RESULTS: The mean of age was 49.2911.59 and 46.718.23 years old in TAC and FAC groups, respectively. In the end of chemotherapy, QoL score were 64 and 68 in TAC and FAC groups, respectively (P < 0.005). After 4 months, patients in TAC and FAC groups experienced 11.45 and 7.14 units improvement in QoL scores, respectively (P = 0.02). CONCLUSIONS: Although, TAC had a more negative impact on QoL during chemotherapy, it created a higher improvement than FAC during 4 months since end of treatment. These effects on quality of life should be considered in making decision for providing and financing cancer treatments in Iran. PCN137 EMPLOYMENT STATUS AND WORK-RELATED DIFFICULTIES IN LUNG CANCER SURVIVORS COMPARED WITH GENERAL POPULATION Kim YA1, Yun Y1, Zo JI1, Shim YM2 1 National Cancer Center, Goyang-si, Gyeonggi-do, South Kore; 2Samsung Medical Center, Seoul, South Korea OBJECTIVES: Although improved lung cancer survival is likely to result in increased lung cancer survivors, little was known about work situation among lung cancer survivors. The purpose of this study was to investigate employment status and workrelated difficulties of lung cancer survivors compared with the general population. METHODS: We enrolled 917 lung cancer survivors from two hospitals 12 months after lung cancer surgery and 1000 volunteers from the general population. Multivariate logistic regression was used to identify the factors associated with work situation. RESULTS: Employment decreased from 69.6% to 38.7% after cancer treatment. The proportion of lung cancer survivors who remained working was significantly smaller relative to that of the general population (63.5%) (adjusted odds ratio [aOR] = 2.59; 95% confidence interval [CI]: 1.91 to 3.51). In subgroup analyses, female survivors over 65 years had unemployed after treatment (aOR = 89.24; 95%CI = 10.52 to 756.91) than at the diagnosis of cancer. Among cancer survivor who remained employed after treatment (n = 284), 71.8% found no meaning of working and 15% experienced decrease wages. CONCLUSIONS: This is the first study with the largest number of patients investigating employment situation among lung cancer survivors reported poorer employment status than the general population. Among cancer sur- A278 vivors, specifically female survivors over 65 years had greater risk of work loss than general population. These findings should help working people concerned about employment after lung cancer. CANCER – Health-Care Use and Policy Studies PCN138 HEALTH-CARE COSTS ASSOCIATED WITH BREAST CANCER MANAGEMENT Cammarota S1, Citarella A1, Menditto E1, Putignano D1, Riegler S1, De Luca L1, Malorni L2, De Placido S2, Arpino G2 1 CIRFF, Federico II University, Naples, Italy; 2Federico II University Hospital School of Medicine, Naples, Italy OBJECTIVES: To assess the outpatient direct costs related to early and metastatic breast cancer (BC) management in Campania, a Southern Italy region. METHODS: This is a retrospective cohort study based on clinical records from 457 general practitioners who managed an average of 630,000 inhabitants in Campania. Incident early BC cases from 2005 to 2007 were identified and costs related to outpatients management were calculated until evidence of local recurrence or metastases (BC Event), death, revocation or the end of the database (December 31, 2009). For those patients who developed a BC event, costs for their disease management were further analyzed from the time of the event until death, revocation, or the end of the database. Monthly cost per patient was expressed in Euros. RESULTS: A total of 1529 patients with early BC were identified in the database. Of these, 112 women developed a BC event during the study period. At a median follow-up of 34 months, adjusted monthly primary care cost per patient was c151.87 in the subset of women with early BC. For those who experienced a BC event, at 24 months of follow-up, adjusted monthly primary care cost per patient was almost doubled: c289.15 (P < 0.0001). Main causes for this cost difference were related to increased number of specialists’visits, diagnostic procedures, and laboratory test once a BC event developed. CONCLUSIONS: Outpatient’s management in women with metastatic BC is twice more expansive compared to management of women with early BC. Reasons for this increase are mainly due to increased frequency of imaging and diagnostic procedures in the metastatic BC subset. However, our study underestimates the total costs for metastatic BC patients’ management because hospitalization and chemoptherapy costs are not included in our analyses. Based on our data, secondary and tertiary prevention strategies must be significantly implemented in order to rationalize resource allocation. PCN140 IMPROVING PATIENT ACCESS TO CANCER DRUGS IN INDIA: USING ECONOMIC MODELING TO ESTIMATE A DRUG COST BASED ON MEASURES OF SOCIETAL VALUE Dranitsaris G1, Truter I1, Lubbe M2, Sriramanakoppa N3, Mendonca V3, Mahagaonkar S3 1 Nelson Mandela Metropolitan University (NMMU), Port Elizabeth, Eastern Cape, South Africa; 2North-West University, Potchefstroom, South Africa; 3PharmARC Analytic Solutions, Bangalore, India OBJECTIVES: Cancer patients from lower-income countries such as India often have limited access to modern medicines because of high costs. Using multiples of India’s per capita GDP as the threshold for economic value as suggested by the World Health Organization (WHO), decision analysis modeling was used to estimate a monthly cost for a hypothetical new cancer drug that provides a 3-month survival benefit to patients with metastatic colorectal cancer (mCRC). METHODS: A decision model was developed to simulate progression-free and overall survival in mCRC patients receiving chemotherapy ± the new drug. Outcomes for cancer control and side effects were obtained from randomized trials evaluating 1st and 2nd line chemotherapy in mCRC. Costs for chemotherapy were obtained from both public and private hospitals in India. Utility estimates measured as quality-adjusted life-years (QALY) were determined from 24 oncology nurses using the Time Trade-Off technique. These data were then used to estimate the monthly cost of the new drug using a threshold of $9300 per QALY gained, which is three times the Indian per capita GDP, as recommended by the WHO. RESULTS: The base-case analysis suggested that a monthly cost of $U.S.98 would be considered cost-effective from the Indian public health-care perspective. If the drug were able to improve patient quality of life above the standard of care or survival from 3 to 6 months, the monthly drug cost could increase to $U.S.170 and $U.S.253 and offer the same value. CONCLUSIONS: The use of the WHO criteria for estimating a country-specific drug price based on economic value for a developing country is feasible. However, the challenge would be to identify an appropriate threshold that would provide a balance between what patients/governments can afford to pay and the commercial viability of the product in the reference country. PCN141 TRANSPARENCY OF DRUG REIMBURSEMENT IN POLAND—ONCOLOGY Kujawska A, Nogas G Association CASPolska, Myslenice, Poland OBJECTIVES: Increasing the level of transparency of decision-making process of reimbursement for drugs used in oncology by facilitating online access to public information and other information regarding the refund of cancer drugs in Poland compared to solutions used in the world. METHODS: The project will develop a comprehensive system to monitor the transparency of reimbursement decision-making process in Poland consisting of: Guidelines and Clinical and Reimbursement Recom- 13th Euro Abstracts mendations Database (WiRKliR database) whose purpose is to collect documents from the Polish and selected countries of the clinical guidelines, registration, and decisions about the recommendations and decisions regarding reimbursement for cancer drugs and present information contained in them in a comprehensible and transparent. Reimbursement monitoring which aims to carry out continuous monitoring of the implementation of law and regulation in the field of oncology drug reimbursement and the acquisition of complementary information about the cancer drugs in the database WiRKliR. Reimbursement reports whose purpose is to discuss the issue of transparency of reimbursement through the analysis of procedures drawn up and applied by public authorities in dealing with citizens, with special emphasis on patients and pharmaceutical companies in the reimbursement decision-making process and to present the results of the various stages of our work. RESULTS: The most important results of the project include: increased transparency of institutions involved in the drug reimbursement decision-making process, the democratization of medical information, provide decision-makers Polish health-care system expertise, to increase patient awareness about their rights and proposals for corrective actions (conclusions and recommendations) for public administration in the area of the refund. CONCLUSIONS: A comprehensive monitoring system for civil funding allocation process in the health-care system to counteract the imbalance of information. PCN142 WHAT IS THE CURRENT R&D LANDSCAPE FOR METASTATIC BREAST CANCER? AN INVESTIGATION INTO RECENT CLINICAL TRIAL ACTIVITY Scrutton HC, Samuels ER Heron Evidence Development Ltd, London, UK OBJECTIVES: A number of systemic therapies are available for the treatment of metastatic breast cancer (MBC)—including hormonal therapies, chemotherapeutic agents, and biologics—but long-term prognosis remains poor. There is a need for new treatments that improve survival and are effective in a greater proportion of patients. This research evaluates the recent clinical trial activity directed toward the treatment of MBC. METHODS: MBC trials were identified through a systematic search of the records within http://www.clinicaltrials.gov, using the search term “metastatic breast cancer OR stage IV breast cancer OR advanced breast cancer.” Trials with a start date from January 2008 onward were included and categorized by nature of intervention (new investigational agent, drug launched for indications other than breast cancer, or approved breast cancer product); all trials that did not include a pharmaceutical (biological or drug) were for an unsuitable indication, or that were suspended, terminated, or withdrawn were excluded. The remaining trials were evaluated for a number of variables, including phase of development and product type. The overall number of pharmaceuticals under investigation was also explored. RESULTS: The original search term identified 2014 trials, 365 of which met the inclusion criteria. a large proportion of these trials (41%) were evaluating products already approved for the treatment of breast cancer. Of the remaining trials, 68% were investigating new agents; the remainders were evaluating a product launched for an indication other than breast cancer. There were a total of 116 new products under investigation, although these were mainly at an early stage of development with just 8% of trials at Phase III/IV. CONCLUSIONS: MBC is a key area of research, with a large number of products in development. However, since the majority of clinical trials are at an early stage, it will be some years before these products impact patient treatment. PCN143 PRIMARY PREVENTION AS EXPANDED INDICATION TO MITIGATE COMPETITION: A CASE STUDY Nichols E, Wiederkehr DP, Doyle J Quintiles Global Consulting, Hawthorne, NY, USA OBJECTIVES: While there are several instances of products gaining expanded indications from secondary treatment to primary treatment/prevention, in some cases, expanded indications appear to buffer the drug utilization from competitors. However, the extent to which the later expanded indications impact the utilization of the product across the lifecycle, including following loss of exclusivity, is not well understood. The objective was to examine the prescribing volume of a drug throughout its lifecycle in conjunction with the uptake of a novel class of drugs launched for similar indications to examine the impact of the expanded primary treatment/prevention indications on competition. METHODS: Tamoxifen and the aromatase inhibitors letrozole and anastrozole were selected as case products. All three drugs are FDA indicated for the treatment of advanced breast cancer, yet tamoxifen is the only drug of the group FDA indicated for the primary prevention of breast cancer. From January 1992 to April 2010, the volume of prescriptions (TRx) was collected monthly using SDI’s VONA database and grouped according to class sales by active molecule. RESULTS: Generic sales of tamoxifen maintained a high level after the 1998 approval for the primary prevention indication of “reduction in breast cancer incidence in high-risk women” despite competition within active breast cancer treatment indications from the aromatase inhibitors. While the aromatase inhibitors launched in the mid-1990s, their utilization did not begin to encroach on total tamoxifen prescriptions until the expiration of exclusivity for tamoxifen’s prevention indication in 2003. CONCLUSIONS: This preliminary analysis shows that the strategy of obtaining a later primary prevention indication may help maintain utilization of a compound across the lifecycle. The hypothesis that a prevention indication expansion could mitigate competition should be further examined among products where primary prevention and primary treatment indications have distinctly different dosages and/or branding and where the additional indication is protected by extended exclusivity. 13th Euro Abstracts PCN144 CURRENT CHEMOTHERAPY AND MONOCLONAL ANTIBODY USE PATTERNS IN METASTATIC COLORECTAL CANCER IN WESTERN EUROPE Zhao Z1, Pelletier E2, Barber B3, Bhosle M4, Wang S1, Klingman D5, Gao SK3 1 Amgen, Newbury Park, CA, USA; 2IMS Health Incorporated, Watertown, MA, USA; 3 Amgen Inc., Thousand Oaks, CA, USA; 4IMS Health, Falls Church, VA, USA; 5IMS Consulting, Falls Church, VA, USA OBJECTIVES: Treatment outcomes improved in metastatic colorectal cancer (mCRC) due to the introduction of the monoclonal antibodies (mAb) in combination with chemotherapy. This study described current treatment patterns of chemotherapy and mAbs in clinical practice in 4 EU countries. METHODS: This cohort study used physician-surveyed data from the LifeLinkTM Oncology Analyzer Database for mCRC patients in four EU countries (France, Germany, Italy, and Spain). All patients aged ≥21 years at mCRC diagnosis were included. Treatment patterns in 2009 were examined descriptively by lines of therapy. RESULTS: The study sample includes 2734 mCRC patients (61% male, median age category 61–70 years) with 862, 656, 567, and 649 from France, Germany, Italy, and Spain, respectively. In 1st-line, more patients received FOLFOX-containing regimens than FOLFIRI-containing regimens in Germany (42% vs. 30%) and Spain (25% vs. 16%), while in Italy and France, the reverse was true (Italy: 34% FOLFIRI vs. 29% FOLFOX: France: 26% vs. 19%). In 2nd-line, more patients received FOLFIRI-containing regimens than FOLFOX-containing regimens in Germany (36% vs. 18%), Italy (29% vs. 14%), and Spain (34% vs. 6%), while similar proportions of FOLFOX and FOLFIRI were used in France (18% vs. 15%). In 1st line, Bevacizumab (Bmab) was administered to 44% of patients in Italy, 42% in France, 37% in Germany, and 30% in Spain, while Cetuximab (Cmab) use ranged from 14% in Spain to 7% in Italy. In 2nd-line, Bmab was used in 37% of the patients in Germany, 38% in France, 33% in Italy, and 30% in Spain, while Cmab was used in 30% of the patients in Spain, followed by 26% in Italy, 20% in Germany and 17% in France. CONCLUSIONS: FOLFOX- and FOLFIRI-based regimens are common standard of care chemotherapies, and monoclonal antibodies are routinely combined with these chemotherapies. PCN145 INEQUALITIES IN GEOGRAPHICAL ACCESS TO ONCOLOGY SERVICES IN GREECE AND THEIR IMPACT ON PATIENTS AND CARERS Souliotis K1, Athanasakis K2, Palaka E3, Kyriopoulos II2, Golna C4, Kyriopoulos J2 1 University of Peloponnese, Corinth, Corinth, Greece; 2National School of Public Health, Athens, Greece; 3Foundation of Economic and Industrial Research, Athens, Attica, Greece; 4 Roche Hellas SA, Athens, Greece OBJECTIVES: Previous studies (NSPH 2008, 2009) demonstrated that clustering of oncology resources exceeds the spatial concentration pattern of health-care services in Greece, thus resulting to substantial cross-regional flows of cancer patients. The objective of this study was to assess the impact of geographic accessibility on patients when selecting care and during treatment for cancer. METHODS: Face-to-face interviews with 106 patients diagnosed with cancer from three specialized anticancer hospitals (two Athens and one Thesaloniki). Questionnaire was designed by a specialized Delphi panel of the NSPH to capture patient preferences. Median patient age was 54.5 years. Data were analyzed using SPSS v.15.0. RESULTS: Patients across the board choose their hospital on the basis of specialization (50%), physician reference (41.5%), and hospital reputation (35.8%). Sixty-three percent of patients face access barriers, most commonly cost of health-care services (44.3%), distance from place of residence (37%), and demand on time (33%). 76.4% of patients return to place of residence at treatment intervals and 43.4% immediately after treatment. To receive treatment, 23.6% stay at homes of relatives, 14.2% at hotels, and 1% in hospital-owned hospices. Fifty percent of patients undergo treatment cycles repeating every 15–30 days. Patients travel predominately by own car (48.1%). 4.7% travel by taxi reimbursed by their insurance fund and 1% by hospital ambulance. 84% of patients travel accompanied by one or more carers. At treatment intervals, only 8% of patients are supported by physicians at place of residence. CONCLUSIONS: Significant cross-regional flows of cancer patients to access adequate treatment lead to substantial direct and indirect costs for patients and their carers in a strained financial environment. Patients also face significant gaps in integrated support during treatment intervals. a shift in the organization of cancer services is essential for the system to be responsive to expressed patient needs especially during treatment. PCN146 HEALTH RESOURCE UTILIZATION OF SUBJECTS RECEIVING DENOSUMAB AND ZOLEDRONIC ACID IN A RANDOMIZED PHASE 3 TRIAL OF ADVANCED BREAST CANCER PATIENTS WITH BONE METASTASES Body JJ1, von Moos R2, Stopeck A3, Qian Y4, Braun A4, Chung K4 1 Centre Hospitalier Universitaire Brugmann, Brussels, Belgium; 2Kantonsspital Graubünden, Chur, Switzerland; 3University of Arizona, Tucson, AZ, USA; 4Amgen Inc., Thousand Oaks, CA, USA OBJECTIVES: Bone metastases occur in up to 75% of patients with advanced breast cancer. Skeletal complications of bone metastases include pathologic fracture, spinal cord compression, and surgery and radiation to bone. In a recently completed trial of breast cancer patients with bone metastases, denosumab, a fully human monoclonal anti-RANKL antibody, was superior to zoledronic acid (ZA) in delaying/preventing A279 skeletal related events (SRE). METHODS: Eligible patients (N = 1026, denosumab; N = 1020, ZA) received monthly subcutaneous denosumab 120 mg or intravenous ZA 4 mg in a double-bind, double-dummy design. Health resource utilization (HRU) data, excluding protocol-specified visits, were collected from all randomized patients monthly through end of study (median duration 17 months, both groups). RESULTS: Comparing patients who experienced any on-study SRE (n = 315, denosumab; n = 372, ZA) with patients without an on-study SRE, a greater proportion reported at least one radiation oncology unit visit (22.9% vs. 8.7%), clinic visit (75.5% vs. 71.1%), and inpatient hospitalization (62.9% vs. 48.6%). The mean number of clinic visits and inpatient hospitalizations was also greater for patients who experienced an on-study SRE. A between-treatment group comparison of denosumab versus ZA demonstrated that for patients with one or more clinic visits who experienced any on-study SRE, the denosumab group had 10% fewer mean clinic visits compared with the ZA group (15.5 vs. 17.2 visits). Additionally, mean radiation oncology clinic visits were decreased by 32% in the denosumab compared with the ZA group (4.2 vs. 6.2 visits) and mean inpatient hospitalizations were decreased by 12% (3.7 vs. 4.2 hospitalizations). CONCLUSIONS: Regardless of treatment group, patients who had an on-study SRE had greater HRU compared with patients not experiencing an on-study SRE. In this study, among those with any on-study SRE, patients treated with denosumab had fewer clinic visits, radiation oncology unit visits, and inpatient hospitalizations compared to patients treated with zoledronic acid. PCN147 MEDICAL RESOURCES UTILIZATION OF FIVE MOST PREVALENT CANCERS IN TAIWAN (LUNG CANCER, LIVER CANCER, COLORECTAL CANCER, GASTRIC CANCER, AND BREST CANCER): 2001~2007 Liu JS, Chen LT, Hsiao CF National Health Research Institutes, Miaoli County, Taiwan OBJECTIVES: We want to study the real cost structure of cancer treatment using the National Health Insurance Research Database (NHIRD) from Taiwan. METHODS: This study adopted a retrospective observational design and the data were retrieved from the NHIRD, which is managed by the National Health Research Institutes (NHRI). We used descriptive statistical methods to display data and time trend. RESULTS: The number of patients with these five cancers showed an increasing trend in these 7 years. Breast cancer had the highest annual increase (7.34%), followed by colorectal cancer (3.75%). Lung cancer and breast cancer had the highest average outpatient cost for each patient per year, while lung cancer, colorectal cancer, and gastric cancer had the highest average inpatient cost for each patient per year. The annual increase rates in the total cancer treatment cost for liver cancer, lung cancer, colorectal cancer, gastric cancer, and breast cancer were 10.70%, 11.61%, 16.73%, 18.38%, and 17.21% respectively. We divided the cost structure of cancer treatments into chemotherapy, hormone therapy, and supportive treatment. The highest percentages of chemotherapy and hormone therapy were for colorectal cancer (more than 95%) and for breast cancer (more than 11%). On average, the annual increase rates in the cancer chemotherapy cost per patient for liver cancer, lung cancer, colorectal cancer, gastric cancer, and breast cancer were respectively 3.59%, 3.66%, 9.86%, 16.78%, and 11.72%. CONCLUSIONS: This study showed a positive correlation between annual cancer patient number and NHI reimbursement, and an increasing trend for chemotherapy, hormone therapy, and supportive treatment. The results also showed that the number of the total amount of NHI reimbursement and the drug costs had increasing trends in these 7 years. This increased the financial burden of cancer patients and may alter the allocation of NHI resources. PCN148 IMAGING TESTS IN STAGING AND SURVEILLANCE OF EARLY BREAST CANCER (EBC) IN ITALY—CHANGES IN ROUTINE CLINICAL PRACTICE AND COSTS IMPLICATIONS Cammarota S1, Menditto E1, Citarella A1, Putignano D1, Riegler S1, Arpino G2 1 CIRFF, Federico II University, Naples, Italy;, 2Federico II University Hospital School of Medicine, Naples, Italy OBJECTIVES: New imaging tests such as computed tomography (CT), [18F]fluorodeoxyglucose-positron emission tomography (FDG-PET) scanning, and magnetic resonance imaging (MRI) are not recommended for staging or follow-up of asymptomatic patients with EBC according to current guidelines. However, frequently these tests are requested even in the absence of a clinical indication. The purpose of this study was to evaluate how the availability of new imaging techniques has changed staging and follow-up modalities in EBC patients and to estimate its cost implications. METHODS: We analyzed clinical computerized information from 457 general practitioners assisting an average of 630,000 inhabitants of the Campania region in the south of Italy. Incident EBC cases were selected and divided into calendar years from 2005 to 2008. The mean number of diagnostic tests prescribed per patient (N/Pt) and the mean costs per patient were evaluated during the first post-diagnosis year. The mean was compared between each year using one-way analysis of variance. Costs were expressed in Euros (mean ± SD). RESULTS: We identified 576, 489, 474, and 497 newly diagnosed cases of EBC in 2005, 2006, 2007, and 2008, respectively. Overall, there was a significant increase of the mean number of imaging tests prescribed per patient from 2005 to 2008 (P < 0.0001). No change of the mean number of mammograms, bone scan, and chest x-ray requested per patient was observed. The mean costs per patient also significantly increased from c354.96 ± 581.32 in 2005 to c546.78 ± 837.36 in 2008 (P = 0.004). CONCLUSIONS: CT, FDG-PET, and MRI employment for EBC patients in daily clinical practice has been steadily growing over the past 4 years with cost repercussions. However, there are no data to support their role in routine breast A280 cancer staging or surveillance in asymptomatic patients. Further studies are needed to characterize patients’ typology who deserve intensive staging and follow-up procedures. PCN149 DECIDING UPON NEW AND EXPENSIVE TECHNOLOGIES IN HEALTH CARE: REAL OPTIONS ANALYSIS IN PROTON THERAPY Grutters J1, Abrams K2, Deruysscher D3, Lambin P3, Pijls-Johannesma M3, Beutner E1, Peters H1, Joore MA4 1 Maastricht University, Maastricht, Limburg, The Netherlands; 2University of Leicester, Leicester—UK; 3MAASTRO Clinic, Maastricht, Limburg, The Netherlands; 4Maastricht University Medical Center, Maastricht, Limburg, The Netherlands OBJECTIVES: Radiotherapy with protons is a promising new treatment modality, for which adoption decisions are being made worldwide. However, the investment costs of proton therapy (PT) are high (roughly c90 million) and limited clinical evidence is available. Also, previous studies have indicated that PT may be cost-effective, but show considerable decision uncertainty. Consequently, it is unclear whether we should adopt PT now, or wait for more information. Adoption involves a risk of facing high sunk costs, while delay may impose opportunity losses because patients receive suboptimal treatment. Real options analysis (ROA), a technique originating from financial economics, assists in making this trade-off. METHODS: We examined whether to adopt PT, as compared to stereotactic body radiotherapy, in the treatment of stage I non-small cell lung cancer (NSCLC). Three options are available: adopt without further research (AN); adopt and undertake a trial (AT); or delay and undertake a trial (DT). The decision depends on the expected net gain of each option, which is calculated by subtracting its total costs from its expected benefits. RESULTS: The expected net gain of at and DT were positive, indicating that we should not decide to adopt without further research (AN). Up to a sample size of 1000 patients, the expected net gain of at was higher than DT, indicating that the best option was to adopt and trial. The expected net gain of at was highest for a sample size of 450 patients, which is thus considered the optimal sample size. CONCLUSIONS: Based on these results, we recommend to adopt PT in the treatment of stage I NSCLC, and to perform a trial with 450 patients. We have shown that ROA provides a transparent method of weighing the costs and benefits of all available options, to assist in decisionmaking upon new and expensive technologies. PCN150 SYSTEMATIC LITERATURE REVIEW ON THE INTER AND INTRA LABORATORY VARIABILITY OF MOLECULAR TESTING OF RESPONSE TO TREATMENT OF CHRONIC MYELOID LEUKEMIA (CML) PATIENTS AND THE ASSOCIATED COSTS AND COST-EFFECTIVENESS Ratcliffe M1, Hudson PE1, Ossa D2 1 Phmr Consulting, London, UK; 2Novartis Pharma AG, Basel, Switzerland OBJECTIVES: During disease monitoring of patients with CML, for patients with a complete response, residual leukemia can be assessed by real-time quantitative polymerase chain reaction (RQ-PCR). There are several “home-brew” and commercially available BCR-ABL gene transcript detection methodologies in use, each requiring internal validation for the specific laboratory and giving rise to laboratoryspecific data. Harmonization of results according to an international scale is underway, but use is limited for several technical reasons. Information is required for decision makers on the accuracy and reproducibility of the tests and their costs and cost-effectiveness. The objective of this study was to assess the quantity and quality of such information. METHODS: English language systematic literature review on the intra- and inter-laboratory variability for BCR-ABL molecular monitoring testing, inter-rater reliability across manual assays and the costs and cost-effectiveness of molecular testing in CML. RESULTS: From 88 papers retrieved for detailed analysis, we found no studies which conducted a repeated test procedure on the same patient sample using the same technical approach in the same laboratory. There are a large number of studies which have compared alternative approaches using the same patient sample in molecular monitoring in the same laboratory. Several well-conducted studies have examined the variability of results from different laboratories in controlled environments. We found no studies which compared inter-rater reliability or examined the costs or cost-effectiveness of molecular testing in CML. CONCLUSIONS: There is a reasonable body of evidence on certain aspects of analytical validity for CML molecular testing, but other aspects of analytical validity and the costs and economics of molecular diagnostics in CML appear to be an unresearched area. Testing variability has potentially serious implications for patient outcomes and more information for decision-makers to assess relative costs and cost-effectiveness is required. PCN151 TARGETED CANCER THERAPIES: PRICING, ACCESS, AND UPTAKE 1 2 Aggarwal S , Stevens CA 1 PAREXEL Consulting, Bethesda, MD, USA; 2PAREXEL Consulting, Waltham, MA, USA OBJECTIVES: The oncology market has become one of the major focus areas for pharmaceutical and biotech firms. As of March 2009, 15,752 of 39,747 Phase I, II, and III trials listed on clinicaltrials.gov were related to cancer (approximately 40%). This large interest in oncology stems from market success of cancer therapies launched in the past decade and the existence of high unmet need to treat different types of cancers. As the number of FDA approved cancer therapies increases, there is a need to understand treatment patterns of these cancer drugs. METHODS: To understand 13th Euro Abstracts the trends in usage and sales of cancer therapies, we analyzed the US market (sales and prescription) 2005–2008 data for all FDA-approved cancer drugs. Drugs were categorized as targeted cancer therapies, chemotherapies, monoclonal antibodies, small molecules, branded, and generics. RESULTS: During the past 5 years, the usage of both targeted cancer therapies and chemotherapy drugs has increased by high double digit rates. From 2005 to 2008, the total prescriptions for targeted cancer therapies and chemotherapies increased by 66% and 30%, respectively. While the sales of both types of these drugs are expanding, the majority of sales growth is attributed to an increasing uptake of targeted cancer drugs. The sales share of targeted cancer therapies in the US oncology market increased from 36% in 2004 to 56% in 2008. Among targeted cancer therapies, majority (more than 75%) of uptake belongs to monoclonal antibodies. CONCLUSIONS: The usage and sales trends show a significant increase in the use of cancer drugs. The high usage of targeted cancer therapies versus chemotherapies shows the rapidly changing nature of cancer treatment regimen. PCN153 HOW DO HTA AGENCIES RECOGNIZE AND REWARD INNOVATION? CASE STUDIES IN BREAST CANCER AND COLORECTAL CANCER Shah K1, Mestre-Ferrandiz J1, Towse A1, Nash-Smyth E2, Ball D2, Grainger D3 1 Office of Health Economics, London, UK; 2Eli Lilly & Company, Indianapolis, IN, USA; 3Eli Lilly & Company, Sydney, Australia OBJECTIVES: This paper examines how different payers and health technology assessment (HTA) agencies recognize and reward innovation, using treatments for breast and colorectal cancer as case studies. METHODS: Breast and colorectal cancer were chosen given the extent of clinical advancements to date and supporting publicly available data. Sixteen cancer medicines across these two tumor types were considered. For each medicine/indication, the reimbursement decision and the reasons behind it were obtained from assessment reports published by the respective agencies in Australia, Canada, England and Wales, France, and Scotland. RESULTS: Seventyseven decisions were reviewed (39 and 38 for breast and colorectal cancer, respectively). Twenty-four (62%) and 16 (42%) were positive for breast and colorectal cancer, respectively, while 21% and 45% were negative. In general, HTA agencies appear to consider advancements in breast cancer treatments as representing good uses of health-care resources with some assessments yielding less positive or more restrictive results. The majority of appraisals for newer colorectal cancer treatments have failed to receive positive recommendations. CONCLUSIONS: We identified some broad areas where differences of approach have led to different decisions. These relate to the: 1) acceptability of surrogate end points (e.g., progression-free survival) in cases where improvement in overall survival has not been established; 2) extent to which agencies formally consider input from clinical and patient representative organizations as part of their decision-making process; 3) methods used to assess medicines where pivotal trial did not use a comparator reflecting standard therapy; and 4) mechanisms for re-review or adopting performance-based risk-sharing arrangements following rejection due to uncertain clinical and/or cost-effectiveness. Addressing these issues may improve the likelihood of innovative medicines meeting reimbursement requirements, for breast and colorectal cancer as well as other therapeutic areas, thereby increasing the overall health benefit from pharmaceutical development that accrue to patients. PCN154 A COMPARISON OF HTA RECOMMENDATIONS FOR CANCER TREATMENT TECHNOLOGIES PUBLISHED BY AGENCY FOR HEALTH TECHNOLOGY ASSESSMENT (AHTAPOL) AND NATIONAL INSTITUTE FOR HEALTH AND CLINICAL EXCELLENCE (NICE) Kiljan A, Kolasa K, Hermanowski T Medical University of Warsaw, Warsaw, Poland OBJECTIVES: The objective of this study was to compare HTA recommendations for cancer drug technologies issued by both Agency for Health Technology Assessment (AHTAPol) in Poland and National Institute for Health and Clinical Excellence (NICE) in the UK. METHODS: The review of HTA recommendations concerning cancer technologies published online in the period August 2007–June 2010 (AHTAPol) and March 2000–June 2010 (NICE) was performed. The classification of HTA recommendations based on Raftery’s approach labeling them as positive, positive with major or minor restriction, and negative was conducted. Negative guidance was categorized as clinical or nonclinical. Reasons for HTA recommendations for drug technologies appraised by both AHTAPoL and NICE were compared. Contradictory and noncontradictory recommendations were identified as well. RESULTS: A total of 149 drug technologies were appraised by AHTAPol, of which 39 concerned cancer technologies (seven resubmissions). NICE published 136 drug appraisals, of which 59 related to cancer technologies (12 resubmissions). In total, 12 cancer drug technologies were appraised by both AHTAPoL and NICE. Among them, there were nine contradictory and three noncontradictory pairs of guidance (two negative and one positive). In the group of drug technologies appraised by both agencies, there were 42% and 67% positive HTA recommendations issued by AHTAPoL and NICE, respectively. Negative recommendations based on nonclinical reasons prevailed in Poland (58%). At the same time, there were as many positive recommendations with major restrictions (33%) as negative recommendations based on nonclinical reason (33%) in the UK. The positive guidance without restriction constituted 8% and 17% of all appraisals for cancer drug technologies published by AHTAPol and NICE, respectively. A281 13th Euro Abstracts CONCLUSIONS: In a studied period, NICE published more positive recommendations for cancer drug technologies than AHTAPoL. The contradictory recommendations prevailed in the group of drug technologies appraised by both jurisdictions. PCN156 KNOWLEDGE TRANSFER REGARDING CANCER SCREENING BASED ON INTERESTS OF DIFFERENT TARGET POPULATIONS Hamashima C National Cancer Center of Japan, Tokyo, Japan OBJECTIVES: To develop targeted leaflets regarding cancer screening guidelines that take account the knowledge required by different target groups. METHODS: Before developing leaflets, public interest data regarding cancer screening were collected by the problem solving model (the so-called KJ method developed for field surveys in Japan). The KJ method includes two steps: label making of information corresponded to specific issues and grouping similar concepts. Contents of leaflets were edited based on the results of the KJ methods but changed in the final version according to discussions at a committee meeting involving public members. We compared the results of the KJ method and contents of the final version of the leaflets for the following groups: cervical cancer screening targeted at 20-year-old subjects (first group), cervical cancer screening targeted at subjects aged 30 years and older (second group), and colorectal cancer screening targeted at subjects aged 40 years and older (third group). RESULTS: Common interests among the three groups included targeting cancer and screening methods. Although the KJ method revealed that the first group expected broad information, in the final version of the leaflet, only basic information regarding participation in cancer screening programs remained. The final versions in group 2 and 3 almost corresponded to the first interest of an actual plan to participate in a screening program including details of the screening methods as well as physical and financial burden expected. Although information regarding the harm of cancer screening was initially included according to the results of the KJ methods in the second and third groups, there was no expectation in the first group. CONCLUSIONS: A targeted leaflet is a powerful tool to share knowledge regarding cancer screening. We must understand the expectations of different target groups and prepare appropriate leaflets that support the decision to take part in cancer screening. PCN157 REASONS FOR DISCONTINUATION OF HORMONAL THERAPY IN BREAST CANCER PATIENTS ACROSS FIVE EUROPEAN COUNTRIES Petrilla AA, Ferrufino CP IMS Health, Falls Church, VA, USA OBJECTIVES: Recent evidence suggests that many breast cancer patients discontinue their hormonal therapy treatment regimen before the end of the recommended treatment period, but reasons for treatment discontinuation are not well understood. Therefore, we examined reasons for hormonal therapy discontinuation in a European treatment cohort of female breast cancer (BC) patients. METHODS: Female patients ages 21+ with a diagnosis of BC between January 2006 and December 2008 were identified within IMS Oncology AnalyzerTM which provides retrospective, cross-sectional, cancer-treatment data from 14 countries including France, Germany, Italy, Spain, and the United Kingdom and encompasses over 60,000 de-identified BC patient records from a physician panel of nearly 800 doctors in the EU5. Patients within the EU5 with early-stage BC (stages I-IIIa) who received hormonal therapy and for whom their physician provided a reason for discontinuing first course of therapy were included in the study. Patients enrolled in clinical trials were excluded. Reasons for discontinuation of therapy include course completion, progression of disease, stabilization of disease, adverse events, terminal outlook, poor performance, and patient’s choice. RESULTS: 10,949 patients were identified. Approximately 57.2% discontinued their first course of hormonal therapy prior to course completion. Of these patients, the top reasons for early discontinuation included progression of the disease (59.5%), adverse events (21.7%) and patient choice (10.1%). Mean duration of therapy for all patients was 95.4 months (range 45.5–152.1). Time to discontinuation was significantly shorter for those who quit due to Aes than those who quit by choice (36.2 months vs. 57.8 months, P < 0.0001). Of patients that discontinued due to AEs, hot flushes (46.1%), pain (19.1%), and nausea and vomiting (14.0%) were the most commonly reported events. CONCLUSIONS: This analysis in a real world setting provides new insight into reasons for early discontinuation of hormonal therapy in the EU5. PCN158 TREATMENT DECISION CHANGE WITH 21-GENE RECURRENCE SCORE IN PATIENTS WITH EARLY STAGE BREAST CANCER (ESBC): A META-ANALYSIS Chien R1, Hornberger J2 1 Cedar Associates LLC, Menlo Park, CA, USA; 2Cedar Associates LLC and Stanford University, Menlo Park, CA, USA OBJECTIVES: The 21-gene Recurrence Score® (RS) is a validated assay for estimating a woman’s recurrence risk and chemotherapy benefit after the diagnosis of estrogen receptor positive (ER+), ESBC. We conducted a meta-analysis of RS’s influence on treatment recommendations/decisions in lymph node negative (LN-) disease. METHODS: Literature was abstracted on cohort studies that reported the change in the recommendation or actual use of adjuvant chemotherapy (CT) for patients with ER+, LN-, ESBC before and after the RS. Outcomes evaluated were: treatment decision change from 1) CT plus hormone therapy (HT) to HT-only or 2) HT-only to CT plus HT. Actual treatment change was used when available. Reductions in the relative and absolute CT use associated with the RS were computed with Review Manager (5 Version 5.0.24; Copenhagen: The Nordic Cochrane Centre, the Cochrane Collaboration 2008). The relative change in CT was calculated as a before-to-after ratio of proportion of patients recommended or received CT, weighted by study sample size. RESULTS: Seven published studies (n = 912 patients) were eligible for the metaanalysis. One was a prospective physician survey; six were retrospective chart reviews. a total of 569 (62%) patients were recommended CT prior to RS testing. After RS testing, 272 (30%) patients were recommended or received CT. In six of seven studies (n = 652 patients) that reported a treatment change, 212 (32%) patients switched from CT plus HT before RS to HT-only after RS. The absolute reduction in CT before and after RS testing was 30% (95% CI [−39%, −21%]). The relative reduction in CT before and after RS was 49% (95% CI [42%, 58%]). Estimates varied little when analysis omitted a single study. Limitations include heterogeneity in study designs. CONCLUSIONS: The meta-analysis shows consistent overall reduction in CT with the use of RS. CANCER – Conceptual Papers and Research on Methods PCN160 THE ROLE OF PATIENT SELECTION CRITERIA IN IDENTIFYING OVARIAN CANCER PATIENTS Meyers J1, Mitra D1, Doan J2, Leeflang C3 1 RTI Health Solutions, Research Triangle Park, NC, USA; 2Genentech, Inc, South San Francisco, CA, USA; 3IMS Health, Watertown, MA, USA BACKGROUND: Retrospective claims databases are commonly used in outcomes research. Since physician charts are rarely available to confirm diagnoses, care must be taken when choosing patient populations. OBJECTIVES: To show how patient selection criteria affects sample size and chemotherapy treatment rates using an ovarian cancer (OC) population. METHODS: Patients were initially selected if they met the following inclusion criteria: at least one diagnosis of OC (ICD-9-CM codes 183.0x) between 1/1/2002 and 12/31/2007 (first OC diagnosis date termed index), 6 months pre-index and 12 months post-index eligibility, and no OC diagnosis in the 6 months pre-index. Additional criteria were imposed to further refine the sample and assess variations in chemotherapy treatment rates. First, patients were required to have at least two diagnoses of OC at least 14 days apart. Next, patients were required to have both OC diagnoses on a record labeled as medical, surgical, facility, or pharmacy (i.e., ancillary records were excluded). RESULTS: A total of 37,172 patients had at least one diagnosis of OC. Of those, 16,418 had 6 months pre-index and 12 months post-index eligibility with no pre-index OC diagnoses. In this population, 26% of patients received chemotherapy. When patients were also required to have one additional OC diagnosis at least 14 days from index, the sample size dropped to 7431 patients, of whom 47% received chemotherapy. When OC diagnoses on ancillary records were excluded, a total of 6213 patients were identified, of whom 52% received chemotherapy. CONCLUSIONS: Chemotherapy rates among OC patients varied significantly by the sample selection criteria used. Care must be taken to identify the correct patient sample in any retrospective database analysis since selection criteria affect the appropriateness of the sample, and thus the study results. PCN161 DECISION-ANALYTIC MODELING IN CHRONIC MYELOID LEUKEMIA—A SYSTEMATIC OVERVIEW Rochau U1, Schwarzer R1, Sroczynski G1, Jahn B1, Wolf D2, Gastl G2, Siebert U1 1 UMIT—University for Health Sciences, Medical Informatics and Technology; Oncotyrol— Center for Personalized Cancer Medicine, Hall, Tyrol, Austria; 2Medical University Innsbruck, Innsbruck, Austria OBJECTIVES: To provide an overview on published decision-analytic models evaluating various treatment strategies in chronic myeloid leukemia (CML). We sought to describe and analyze the structural and methodological approaches used and to derive recommendations for future CML models. METHODS: We performed a systematic literature review in electronic databases (Medline/PreMedline, EconLit, EMBASE, and others) to identify published studies evaluating CML treatment strategies using mathematical decision models. The models were required to compare different treatment strategies and to comprise relevant clinical health outcomes such as life-years gained or QALYs over a defined time horizon and population. We used standardized forms for data extraction, description of study design, methodological framework, and data sources. RESULTS: We identified 14 different decision-analytic modeling studies, among which, 13 included economic evaluations. The modeling approaches varied substantially and comprised decision trees, Markov models, Monte Carlo simulations, and mathematical equations. Time horizons ranged from 2 years to lifetime. Health outcomes included survival, life expectancy, and QALYs. Compared treatment strategies comprised bone or blood marrow transplantation, conventional chemotherapy, interferon-alpha, and first-generation tyrosine kinase inhibitor (TKI) imatinib. None of the models evaluated comprehensive personalized medicine strategies or second generation TKI (e.g., nilotinib, dasatinib). Only few models were validated using A282 independent data. CONCLUSIONS: We found several well-designed models for different CML treatment strategies. However, the quality of reporting varied substantially. We recommend that future models should include novel treatment options, subgroup evaluations for a more personalized decision making, and validation using independent data. Already available models with a short time horizon could be updated with new survival data. PCN162 EXTRAPOLATION IN TRIAL-BASED COST-EFFECTIVENESS MODELING: IN SEARCH OF A STANDARD Ekman M1, Charbonneau C2, Ramsberg J1, Jonsson L1, Sandin R3, Jonsson B4, Drummond M5, Weinstein MC6 1 i3 Innovus, Stockholm, Sweden; 2Pfizer, Inc, New York, NY, USA; 3Pfizer Oncology, Sollentuna, Stockholm, Sweden; 4Stockholm School of Economics, Stockholm, Sweden; 5 University of York, York, UK; 6Harvard School of Public Health, Boston, MA, USA BACKGROUND: Extrapolation is often a key element in health economic modeling. Although any model should use empirical data if possible, the effects of treatments on long-term health outcomes are seldom observed within the follow-up time of a clinical study. Extrapolation over a lifetime horizon will generally be required in economic models where treatments have different cumulative survival at the end of the clinical trial. Typically, a within-trial analysis of costs and health effects, in which outcomes are truncated at the conclusion of the trial, will be overly conservative. OBJECTIVES: The purpose of this study is to compare different methods of extrapolation in the context of examples concerning oncology, although the principles apply across all therapeutic areas. METHODS: There is a set of standard assumptions regarding extrapolation of survival data from clinical studies, ranging from very cautious (“stop-and drop”) to very optimistic (“continued benefit”). The impact of different assumptions regarding extrapolation is explored, and the implications are discussed. CONCLUSIONS: The choice of extrapolation method has significant impact on comparative clinical effects, costs, and cost-effectiveness. Based on our findings and supporting examples, we propose the following: 1) Analysts should perform and report results under a range of specific standard extrapolation assumptions to increase comparability across studies. 2) The choice of a base-case approach in any particular study should be guided by knowledge about the biology of the indication under evaluation and the mechanism of action of the treatment. A case could be made for a reference case method of extrapolation, but we believe that sensitivity analysis across a standard set of possibilities is sufficient. Adherence to these modeling practices will contribute to increased transparency in modeling and hence potentially to a greater confidence among health-care decision-makers in the results from costeffectiveness analyses building on modeling and extrapolation. PCN163 EXTENDING FIXED EFFECT MODELS TO CENSORED COST DATA 1 2 Baser O , Yuce H 1 STATinMED Research/University of Michigan, Ann Arbor, MI, USA; 2New York City College of Technology-CUNY/STATinMED Research, New York, NY, USA OBJECTIVES: Challenges in analyzing cost data include addressing skewness in cost distributions, observed and unobserved heterogeneity across samples, and even more challenging complexities due to censoring. We combined generalized random effect models with inverse probability weighted (IPW) estimation techniques to address those challenges in a single model. METHODS: Generalized fixed effect models have been used with weights that are calculated as inverse due to probability being uncensored. The Gaussian family and log link function was chosen and we applied a test to see if possibly censoring bias exists. We also calculated the deviation from the consistent value if standard pooled ordinary least squares were used. RESULTS: A total of 4824 observations were used in the analysis. We obtained Medicare claim files for the 2 years following patients’ lung cancer diagnosis. Costs had high kurtosis and skewness. Moreover, 30% of the cases were censored, and therefore, their annual costs were not observed. The total cost of all care was $63,000 for the 2 years following a lung cancer diagnosis and $57,000 for incomplete cases. Results significantly diverged from the standard regression model (P = 0.000). CONCLUSIONS: This paper applied inverse probability weighted estimation and fixed effect panel data models to an inception cohort of patients newly diagnosed with lung cancer. Our findings suggest that standard regression models yield inconsistent estimators due to censoring bias. The IPW least square estimation method removes that bias. PCN164 THE ECONOMICS OF CHRONIC MYELOGENOUS LEUKEMIA: A COMPARISON OF MODELING APPROACHES Taylor MJ University of York, York, UK OBJECTIVES: Chronic myelogenous leukemia (CML) is a progressive disease which arises from damage to the DNA of a stem cell in the bone marrow. This results in the uncontrolled growth of white blood cells which, in turn, can lead to severe impairment of an individual’s functioning. The National Institute for Health and Clinical Excellence (NICE) models the costs and benefits of medicines. The structure of these models is not prespecified and wide variations are often observed, both in the model’s choice of input parameters and in the structure of the modeling approach. While there is no such thing as a “correct” model, it is important that different models are compared and critiqued in order to identify any particular strengths and weaknesses of differing approaches. METHODS: A review was undertaken, identifying existing published 13th Euro Abstracts models for CML. The data sources and choice of inputs were compared across each model and presented in a comparative table. Furthermore, the different approaches to model structure were examined, and attempts were made to explore the consequences of each approach on the models, costs, effectiveness, and cost-effectiveness findings. RESULTS: The approaches to modeling CML vary significantly between different studies. While different data sources are utilitized in each model, this can usually be explained by emerging data which were not available to other researchers. However, the overall approach to modeling the disease varied considerably across each study. Model structures and assumptions for long-term outcomes were key drivers of the cost-effectiveness results in each model, but were often based on contrasting and contradicting approaches. CONCLUSIONS: This review has highlighted significant variation in approaches to modelling CML. It is recommended that longterm follow-up from previously published trials should be used to predict the likely outcomes associated with shorter-term outcomes, such as treatment response. PCN165 INFLATION ANALYSIS AS A TOOL TO ASSESS COST-EFFECTIVENESS OF CANCER TREATMENT Eliseeva E, Apanasevich V, Solodyankina T Vladivostok State Medical University, Vladivostok, Russia OBJECTIVES: Cost-effectiveness of chronic diseases drug-therapy is estimated with the help of various pharmacoeconomic analysis methods. Cost-effectiveness analysis, cost minimization analysis, cost-utility analysis, and cost-effectiveness analysis are used most often. We have proposed a modification of the cost-effectiveness analysis, which makes possible to record patient’s losses during treatment. Inflation coefficient K and its average value—Kmed may be more reliable tool in assessing the costeffectiveness of treatment of recurrent disease. METHODS: Formula evaluation of the effectiveness can be expressed as Kt &die;T DC(M/Nt), where k—coefficient of effectiveness of treatment at time t DC—amount invested in the treatment of one person M—number of patients in the beginning of treatment (original group) N—number of patients at the time T t—regular interval, running on the account (month, day, year, etc.) RESULTS: Analysis of the coefficient K shows a tendency to increase with an increase in DC price for the treatment or—on decreasing the number of patients N, remaining in the group. As the period t becomes longer, entire formula is decreasing (inflation). The average coefficient k meaning is calculated according the formula Kmed &die;T ∑ Kt/t. Coefficient shows the average cost of one free of recurrence month in one patient from the group. The number of patients in group N decreases as a disease recurs, which leads to Kmed increase because of reallocating funds spent on chronic patients for the rest of patients. CONCLUSIONS: The study of economic efficiency through inflationary coefficient K, we have proposed, is a sensitive method for estimating treatment costs, and may prove to be a reliable tool for cost-effectiveness analysis of chronic patient drug therapy. PCN166 DEVELOPMENT OF A FLAG SYSTEM FOR THE COMPUTERIZED DETECTION OF CANCER PATIENTS WITH ADDITIONAL TREATMENT NEEDS BY MEANS OF THE “COMPUTER BASED HEALTH EVALUATION SYSTEM” (CHES) Holzner B1, Kemmler G1, Giesinger J1, Gamper E1, Oberguggenberger A1, Zabernigg A2, Sperner-Unterweger B1 1 Innsbruck Medical University, Innsbruck, Tyrol, Austria; 2County Hospital Kufstein, Kufstein, Tyrol, Austria OBJECTIVES: Studies evaluating computer-aided routinely assessment of patientreported outcomes (PRO) suggest important benefits for physicians, patients, and their treatment. This project’s aim was developing a flag system for computerized detection of cancer patients with additional treatment needs. METHODS: The detection system is part of the “Computer based Health Evaluation System (CHES),” a software for the computerized collection, processing, and presentation of PRO data. The flag system helps identify cancer patients with conspicuous subscale patterns. This includes cutoff values (patients marked with orange or red buttons) and significant changes over time (yellow deltas). The testing data set was derived from a longitudinal study on QOL (assessed using the EORTC QLQ-C30) in chemotherapy patients at an oncological outpatient unit. Several criteria of “relevant” changes were compared regarding the prevalence of such changes in the study population and their statistical significance on an individual patient level. RESULTS: QOL data of 167 cancer patients were analyzed (on average 5.3 assessments per patient). The 75th and 90th percentile showed to be useful cutoff values. Recommended thresholds of relevant QOL changes appear to be unduly low when considering changes in the individual patient. Based on empirical data, we suggest a modified criterion of relevant change for the EORTC QLQ-C30 which appears clinically and statistically more meaningful. CONCLUSIONS: The developed flag system enables physicians to detect patient-reported health deficits (e.g., fatigue) at one glance. However, more research involving various diagnostic groups is needed for a more profound empirical basis for developing refined criteria. 13th Euro Abstracts PCN167 INFORMING THE DECISION: IMPROVING GENERALIZABILITY BY PARAMETERIZING RELEVANT INFORMATION IN THE ECONOMIC EVALUATION OF TRASTUZUMAB IN EARLY BREAST CANCER Joore MA1, Hoomans T2, Tjan-Heijnen V1, Seferina S1, Goettsch W3, Grutters J2 1 Maastricht University Medical Center, Maastricht, Limburg, The Netherlands; 2Maastricht University, Maastricht, Limburg, The Netherlands; 3Health Insurance Council, Diemen, Noord-Holland, The Netherlands OBJECTIVES: To better inform decision-makers on how practice may deviate from scientific evidence, and for which factors further research is particularly worthwhile, through parameterizing factors that may impact the generalizability of economic evaluations. METHODS: We modeled the long-term cost-effectiveness of trastuzumab in early breast cancer. Based on a review, five real-world factors that could influence the cost-effectiveness of trastuzumab in daily practice were identified: comparators, patient consequences, clinical practice, patient mix, and professional use. To explore the impact of these factors on (cost-)effectiveness, parameters were added to the model. Parameter values, ranges, and distributions were based on literature and expert opinion. Analyses were performed without (trial scope) and with real-world factors included (decision scope). Cost-effectiveness acceptability curves were drawn and value of information was analyzed. RESULTS: In the trial-scope analysis, the FinHer regimen dominated the other two trastuzumab regimens, and was cost-effective compared to usual care (c1.845/QALY). In the decision-scope analysis, the HERA regimen was most effective, amounting to c50.746/QALY gained compared to FinHer. At a ceiling ratio of c30.000, the FinHer regimen was most likely to be cost-effective in both analyses. However, this probability was higher in the trial-scope analysis (98%) than in the decision-scope analysis (54%). Hence, the probability that in the real world a wrong decision is made is considerably higher than suggested in the trial-scope analysis. The parameter for which in the real world further research was most valuable was the credibility of the FinHer regimen. CONCLUSIONS: Incorporating parameters that improve generalizability is feasible, even in absence of evidence. Parameterizing the uncertainty of these parameters provides an estimation of how likely it is for each comparator to be cost-effective in the real world. Also, it informs decision-makers and researchers for which parameters further research is most valuable, for instance, in patient access schemes. DIABETES/ENDOCRINE DISORDERS – Clinical Outcomes Studies PDB1 HYPOGLYCEMIA-RELATED HEALTH-CARE UTILIZATION FOLLOWING INITIATION OF INSULIN ASPART IN A VIAL/SYRINGE OR IN A PREFILLED DISPOSABLE PEN (FLEXPEN®). ANALYSIS OF REAL-WORLD UTILIZATION IN THE UNITED STATES Aagren M1, Luo W2 1 Novo Nordisk Inc, Princeton, NJ, USA; 2Novo Nordisk Inc, Yardley, PA, USA OBJECTIVES: Diabetes is a chronic and progressive disease that for many patients requires initiation and intensification of insulin in order to sustain glycemic control and thereby lower risk of long-term complications. Insulin intensification, however, can have some negative consequences for patients in the form of increased risk of hypoglycemic events. This analysis asks the question to what extent the administration vehicle, prefilled pen versus vial and syringe, has an impact on the risk of hypoglycemic events. METHODS: This is a retrospective database analysis comparing the incidence of hypoglycemia-related health-care utilization between patients that were previously naïve to short-acting insulin initiating the short-acting insulin analogue, insulin aspart (IAsp), in either a vial or in a prefilled disposable pen (FlexPen®). Data were derived from health-care claims for individuals with employer-sponsored primary or Medicare supplemental insurance between 2004 and 2007. The analysis was conducted using logistic regression technique controlling for age, gender, diabetes type, pre-index hypoglycemic incidence, pre-index diabetes treatments, and daily consumption of IAsp. Evaluation period was 12 months. RESULTS: The cohorts consisted of 5523 vial patients and 6065 FlexPen® patients. Mean age of the two cohorts were 53.6 and 56.2 years (P < 0.0001) in vial and FlexPen® groups, respectively. Fifty-one percent and 54% (P = 0.05) of the populations were males. Generally, a larger proportion the FlexPen® versus the vial cohort were already on an insulin regimen (basal or premix). Mean number of annual hypoglycemia-related health-care utilization incidence in the period following initiation of IAsp were 1.27 and 0.87 (P < 0.0001) for vials and FlexPen®, respectively. After multivariate adjustment, the odds ratio of hypoglycemic incidence with vials versus FlexPen® was 1.36 (CI: 1.20–1.54). CONCLUSIONS: After controlling for potential confounders, the incidence of hypoglycemia-related health-care utilization were 36% higher when initiating IAsp in a vial compared to initiation with FlexPen®. PDB2 HOSPITALIZATIONS FOR SEVERE HYPOGLYCEMIAS IN PATIENTS WITH DIABETES MELLITUS IN SPAIN Ramirez de Arellano A1, Lindner L2, Alvarez C3, Sanchez-Zamorano M1, De Rivas B3, Badia X2 1 Bristol-Myers Squibb Iberia, Madrid, Spain; 2IMS Health, Barcelona, Spain; 3AstraZeneca, Madrid, Spain OBJECTIVES: To estimate the number of hospitalizations for severe hypoglycemia (SH) in patients with diabetes mellitus in Spain and assess the differences amongst A283 regions. METHODS: A retrospective study was conducted using the hospital Minimum Basic Data Set (CMBD) which collects up to 73% of all hospitalizations in Spain. All hospitalization episodes of diabetic patients in which SH was the principal or secondary diagnosis for hospitalization during 2007 were accounted for. a SH was defined as a hypoglycemic episode which requires hospitalization. RESULTS: The number of SH episodes as primary diagnosis was estimated at 8242 in relation to total discharged patients with a diagnosis of DM2 (primary and secondary) at 496,660 patients (1.66% of total). When the number of SH as primary diagnosis is related to the total number of discharged patients with a diagnosis of DM2 as the primary diagnosis (23,343 patients), the percentage of episodes become 35.31%. The number of SH as a secondary diagnosis for hospitalization was estimated at 17,302 episodes: 16,649 in patients with DM2 (96.2%) and 653 in patients with DM1 (3.8%). For DM2 when SH as secondary diagnosis are added to primary diagnosis, the number of episodes is 24,891, which means that hypoglycemia diagnosis accounts for 5.01% of hospitalizations for DM2 patients (496,660 patients). Catalonia shows the highest number of SH, with 1477 episodes as a primary diagnosis in DM2 patients (17.9% of total) and 3836 episodes of SH as a secondary diagnosis in DM2 patients (23.0% of total). CONCLUSIONS: Severe hypoglycemia represents a significant cause for hospitalization for DM2 patients in Spain: incidence of 1.66% estimated in the overall DM2 patients increases to 35.3% in the subpopulation of patients discharged for DM2 as primary diagnosis. There is a great variation in the incidence of severe hypoglycemia between regions. PDB3 EFFICACY AND SAFETY OF DIPEPTIDYL-PEPTIDASE 4 [DPP 4] INHIBITORS IN TYPE 2 DIABETES: META-ANALYSIS Park H, Kim YA, Rascati KL The University of Texas at Austin, Austin, TX, USA OBJECTIVES: To assess the efficacy and safety of dipeptidyl peptidase-4 (DPP-4) inhibitors, including saxagliptin, sitagliptin, and vildagliptin, in type 2 diabetic patients. METHODS: Meta-analysis was conducted for efficacy (Hemoglobin A1c and weight) and safety outcomes (hypoglycemia and other adverse events). Both published and unpublished randomized controlled trials (RCTs) for type 2 diabetic patients using DPP-4 inhibitors were included. RCTs were selected if criteria included a duration of at least 12 weeks and they compared DPP-4 inhibitors with placebo or other hypoglycemic medications. Two reviewers independently assessed trials for inclusion and extracted data. Differences were resolved by consensus. RESULTS: Of 579 potentially relevant articles identified, 118 were retrieved for detailed evaluation, and 52 met inclusion criteria. DPP-4 inhibitors significantly reduced hemoglobin A1c compared with placebo (weighted mean difference, −0.851% [95% confidence interval {CI}, −1.057% to −0.645%]) and were noninferior to other hypoglycemic agents (0.117% [95% CI, −0.042% to 0.276%]). DPP-4 inhibitor increased weight compared with placebo (0.186 kg [95% CI, 0.128 kg to 0.244 kg]), but the increase was not significantly different from other hypoglycemic agents. Also compared with placebo, DPP-4 inhibitors had an increased risk of hypoglycemia (odd ratio [OR], 1.430 [95% CI, 1.198–1.708]) and any adverse event (OR, 1.072 [95% CI, 1.003–1.145]). However, the risk of experiencing hypoglycemia or any adverse event was significantly lower in patients using DPP-4 inhibitors compared to other hypoglycemic agents (OR, 0.307 [95% CI, 0.247–0.381]; OR, 0.828 [95% CI, 0.765–0.896], respectively). a similar result was also seen with serious adverse events (OR, 0.834 [95% CI, 0.699–0.996]) and GI adverse events (OR, 0.810 [95% CI, 0.715–0.917]) for DPP-4 inhibitors versus other oral antidiabetic medications. CONCLUSIONS: DPP-4 inhibitors had similar efficacy for hemoglobin A1c reduction and weight loss, and had a decreased risk of hypoglycemia and other adverse events compared with other hypoglycemic agents. PDB4 A CROSS-SECTIONAL STUDY ON GLYCEMIC CONTROL AND ADVERSE EVENTS IN TYPE 2 DIABETES MELLITUS PATIENTS TREATED WITH ORAL ANTIDIABETIC DRUGS IN CHINA Hu SL1, Sun F2, Xu ZR3 1 Fudan University, Shanghai, China; 2Novo Nordisk (China) Pharmaceuticals Co., Ltd., Beijing, China; 3The 306th Hospital of PLA, Beijing, China OBJECTIVES: The study reported the utilization of oral antidiabetic drugs (OADs), glycemic control, hypoglycemia, and adverse events of type 2 diabetes mellitus (T2DM) patients with OADs therapy during their clinical visits in urban hospitals in China. METHODS: The study was a cross-sectional survey conducted at 75 hospitals in nine cities in China. There were 9577 T2DM patients with OADs therapy completed the questionnaires. The survey period was from December 3, 2008 to July 31, 2009. Patients’ self-report on their diabetes condition and the latest health examination was used to evaluate the diabetes management situation. Descriptive statistics method was used for the analyses. RESULTS: A total of 51.1% of the patients were male, the mean age (±SD) was 59.5 ± 12.7 years, and the mean duration of disease (±SD) was 7.9 ± 6.3 years. The percentage of patients treated with biguanides, sulfonylurea, meglitinides, glitazones, α-glucosidase inhibitors, and others (including traditional Chinese medicine) were 78.4%, 65.1%, 14.0%, 12.6%, 31.1%, and 18.1% respectively. Blood glucose control was inadequate, the mean HbA1c and fasting blood glucose were 9.1 ± 2.4% and 9.4 ± 3.4 mmol/L, respectively. Only 7.9% of the patients achieved the treatment target (HbA1c < 6.5%). There were 15.3% patients who experienced hypoglycemia episodes in the last 4 weeks, the mean hypoglycemia rates were 26.0/patient/year, of which major hypoglycemia rates were 7.8/patient/year, and nocturnal hypoglycemia rates were 6.5/patient/year. There were 60.3% OADs A284 patients who reported adverse events in the last 4 weeks, of which, 14.8% experienced digestive symptoms, 8.9% with skin allergies, 22.2% with blurred vision, 4.0% with liver and kidney damages. CONCLUSIONS: The majority of T2DM patients treated with OADs in China had suboptimal glycemic control. The adverse events of OADs treatment impose a high disease burden on patients. It underscores the urgent need for new therapies in the prevention and management of diabetes. PDB5 IMPROVING PATIENT KNOWLEDGE AND CLINICAL OUTCOMES THROUGH A MEDICATION THERAPY MANAGEMENT (MTM) PROGRAM Pinto SL, Howe JL, Partha G University of Toledo, Toledo, OH, USA OBJECTIVES: To examine the effectiveness of an employer-sponsored, pharmacistprovided medication therapy management (MTM) program by evaluating changes in patient knowledge and clinical outcomes over 1 year in patients with diabetes. METHODS: A prospective pre–post longitudinal study. Three 10-question knowledge tests were developed collaboratively by an expert panel of clinical pharmacists and researchers, as part of the MTM program. Throughout the duration of the program, pharmacists administered the tests to City of Toledo employees and their dependents with diabetes, at baseline, 6 months and 12 months. The tests were designed to assess patients’ understanding of the causes, symptoms, and clinical goals associated with diabetes, hypertension, and hyperlipidemia. Based on the test results, pharmacists were able to tailor counseling sessions by educating patients based on the questions answered incorrectly. Clinical data were simultaneously recorded at 3-month intervals over 1 year, and included the following variables: A1c, systolic blood pressure (SBP), and diastolic blood pressure (DBP). Data were analyzed with SPSS v16.0 using descriptive statistics, and Wilcoxin tests. RESULTS: One hundred and one patients enrolled in the MTM program. Diabetes knowledge improved for 29 patients (58.00%; P < 0.001) after only 6 months. Hypertension test scores improved for 15 patients (65.22%; P < 0.01) after 1 year. Of those patients, reductions in SBP and DBP toward goal were seen in 11 (73.33%, P = 0.023) and 15 patients (100%, P = 0.001), respectively. Overall, 41 patients (56.16%; P = 0.042) reduced their A1c toward goal within 3 months; 22 patients (59.46%; P = 0.050) were able to decrease their SBP toward goal, whereas 24 patients (64.86%; P = 0.018) reduced their DPB toward goal after 12 months in the program. CONCLUSIONS: This community pharmacy MTM program has effectively shown that consistent patient education delivered by a clinical pharmacist can have a positive impact on knowledge of disease for patients with diabetes while improving clinical outcomes concomitantly. PDB6 PERFORMANCES OF COMORBIDITY MEASURES IN HEALTH CARERELATED BEHAVIORS AND OUTCOMES IN TYPE 2 DIABETES Ou HT, Balkrishnan R, Erickson SR, Bagozzi RP, Mukherjee B, Piette JD University of Michigan, Ann Arbor, MI, USA OBJECTIVES: To assess and compare the predictive and discriminative performances of comorbidity indexes for health-care outcomes and evaluate comorbidity dimensionality using psychometric techniques. METHODS: The sample was type 2 diabetes in the Medicaid setting from 2003 to 2007. The conceptual framework was based on the Aday-Anderson’s Healthcare Utilization model. Four comorbidity indexes targeted were the Charlson Comorbidity Index, Elixhauser Index (EI), Chronic Disease Score (CDS), and Health-related Quality-of-Life Comorbidity Index (HRQL-CI). Three types of outcomes were health-care behaviors, including physician treatment adherence and patient medication adherence, utilization and expenditures. Multiple regression analyses assessed the predictive performance of comorbidity index. The c statistic (the area under the receiver operator curve) evaluated discriminative validity of the comorbidity index. Confirmatory factor analysis identified comorbidity dimensionality. The SASTM, STATATM, and LISRELTM statistical software were utilized. RESULTS: A total of 9832 patients were finally included, with mean age of approximate 45 years and the majority of them was female (73%) and White (52%). The CDS demonstrated the best performance in predicting physician treatment adherence and discriminating medication adherence behavior. The CDS and HRQL-CI mental aspect index had better predictive validity for medication adherence and similar discrimination for physician treatment adherence. Diagnosis-driven indexes (e.g., EI) had better performances for health-care utilization and expenditures outcomes compared to medication-based index (CDS). A 7-factor pattern/dimensionality was noticed and it provided best model fit and predictive performance across different health-care outcomes. Individual comorbidity dimensions demonstrated differential impacts for a given outcome. CONCLUSIONS: The CDS and HRQL-CI mental aspect index served as better risk adjustment tools for studying health-care behaviors. Diagnosis-driven indexes remained the first choice for health-care utilization and expenditures data. Comorbidity index which accounts for comorbidity dimensionality provided better risk adjustment and insightful knowledge regarding the impacts of different features of comorbidities in predicting patient outcomes. 13th Euro Abstracts PDB7 PREDICTORS FOR THE INITIATION OF A BASAL SUPPORTED ORAL THERAPY (BOT) IN TYPE 2 DIABETIC PATIENTS UNDER REAL-LIFE CONDITIONS IN GERMANY Hammer H1, Dippel FW2, Kostev K3, Fuchs S4, Kotowa W4 1 Dr. Med. Harm Hammer, Bremen, Germany; 2Sanofi-Aventis Deutschland GmbH, Berlin, Germany; 3IMS Health GmbH & Co. OHG, Frankfurt am Main, Germany; 4IMS Health GmbH & Co. OHG, Nürnberg, Germany OBJECTIVES: To assess the predictors for the initiation of a basal-supported oral therapy (BOT) in type 2 diabetics (T2D) under real-life conditions in Germany. METHODS: This historical cohort study included T2D who started an oral antidiabetic drug (OAD) treatment between January 1995 and June 2006 and whose records were eligible at least 12 months before and 36 months after OAD initiation. Data were extracted from a representative real-life database (IMS® Disease Analyzer). BOT initiation was defined according to the ATC code of the European Pharmaceutical Market Research Association, with A10C2 (NPH insulin) and A10C5 (long-acting insulin analogues) prescribed additionally to OADs. The time-dependent rate of T2D starting BOT was calculated by use of the Kaplan–Meier method. Univariate and multivariate Cox regression analyses were applied to identify predictive associated factors. RESULTS: A total of 9028 T2D on initial OAD therapy were included, of whom 1450 patients have been switched to BOT during the observational period. The probability of BOT initiation was associated with poor metabolic control, midlife age, and OAD therapy before insulinization. The combined Cox regression analysis identified three groups of particular importance: group I (hazard ratio [HR] = 2.72; P < 0.001): HbA1c > 8%, age 51–60 years, and sulfonylurea (SU), alpha-glucosidase inhibitor (AGI), or glinide (GLI) as last OAD prescribed, group II (HR = 2.62; P = 0.032): HbA1c > 8%, age ≤ 50 and pretreatment with at least three OADs, and group III (HR = 2.10; P = 0.019): HbA1c > 8%, pretreatment with a combination of at least three OADs and SU, AGI, or GLI as last OAD prescribed. The HbA1c threshold of 7.5% led to comparable results, although only group I reached significance. CONCLUSIONS: The probability of BOT initiation for T2D under real-life conditions in Germany was associated with poor metabolic control, midlife age, and pretreatment with SU, AGI, or GLI. This knowledge may help to identify patients who might benefit from an early initiation of BOT. PDB8 FACTORS ASSOCIATED WITH THE CHOICE OF FIRST INJETABLE THERAPY AND 6-MONTH TREATMENT OUTCOMES IN PATIENTS WITH TYPE 2 DIABETES: DATA FROM THE CHOICE STUDY IN GERMANY Reaney M1, Matthaei S2, Kiljanski J3, Arellano J1, Nicolay C4 1 Eli Lilly & Company Ltd., Surrey, UK; 2Diabetes Centre Quakenbrück, Quakenbrück, Germany; 3Eli Lilly & Company Ltd., Warsaw, Poland; 4Eli Lilly & Company Ltd., Bad Homburg, Germany OBJECTIVES: CHOICE is an ongoing prospective, multicountry observational study designed to evaluate time to significant treatment change among adults with type 2 diabetes (T2D) initiating injectable therapy in routine practice, and describing baseline characteristics, clinical outcomes, and common adverse events over 24 months. Results presented are from 6-month data analyses of the German sample. METHODS: Data were collected at initiation of exenatide or insulin, and 3 and 6 months thereafter. RESULTS: In Germany, 848 patients (394 exenatide, 454 insulin) were enrolled. Multivariate logistic regression indicated that high BMI, hypoglycemia, and high triglyceride levels were associated with exenatide initiation, while increased age, high blood glucose levels, and increased frequency of blood glucose monitoring were associated with insulin initiation. Significant treatment changes were made in 20.7% of exenatide-treated and 29.7% of insulin-treated patients; corresponding discontinuation rates of initiated injectable therapy were 15.5% and 4.1%. At 6 months, mean (SD) HbA1c change from baseline was −0.8% (1.4%) in the exenatide cohort (baseline 8.2%) and −1.6% (1.7%) in the insulin cohort (baseline 8.8%). 21.5% of patients initiating exenatide achieved an HbA1c of <6.5% and 42.5% of <7.0%. Corresponding values for patients initiating insulin were 17.7% and 41.7%, respectively. Mean body weight changes were −3.7 kg in the exenatide cohort and +0.9 kg in the insulin cohort. Gastrointestinal symptoms and hypoglycemic episodes were reported by 19.7% and 2.3% of patients in the exenatide cohort, respectively, and 2.5% and 11.3% of patients in the insulin cohort. CONCLUSIONS: Differences between German patients who initiated insulin and exenatide appear to reflect recommendations of the German Diabetes Association and the German health-care authorities, with exenatide favored when weight is high and hypoglycemia is experienced on oral therapy and HbA1c being only modestly raised. Treatment outcomes were consistent with results of clinical trials and both are important components of T2D care. PDB9 LONG-TERM CLINICAL OUTCOMES OF EXENATIDE ONCE-WEEKLY VERSUS INSULIN GLARGINE FOR THE TREATMENT OF TYPE 2 DIABETES PROJECTED USING THE CORE DIABETES MODEL Timlin L1, Beaudet A2, Wilson B3, Bruhn D4, Boye KS5, Palmer JL2, Lloyd A6 1 Lilly, Windlesham, Surrey, UK; 2IMS Health, Allschwil, Basel-Landschaft, Switzerland; 3Lilly, Basingstoke, Hampshire, UK; 4Eli Lilly & Company, Indianapolis, IN, USA; 5Lilly, Indianapolis, IN, USA; 6IMS Health, London, United Kingdom OBJECTIVES: This analysis aimed to determine the long-term incremental difference in clinical outcomes for exenatide once-weekly (EQW) compared with insulin glargine. 13th Euro Abstracts a 26-week, randomized, clinical trial in 456 patients with type 2 diabetes failing treatment with oral antidiabetic agents (OAD) was performed to compare EQW to insulin glargine. EQW and insulin glargine were associated with (LS) mean decreases in HbA1c (−1.47% and −1.31%), decreases in systolic blood pressure (−3.03 mmHg and −0.63 mmHg), and changes in body weight (−2.6 kg and +1.4 kg) from baseline, respectively. METHODS: The published and validated CORE Diabetes Model (CDM) was used to project clinical outcomes over patient lifetimes using the trial data. The model simulates disease progression by combining 15 Markov-based submodels to determine the occurrence and time to onset of diabetes-related complications, life-years gained, and qualityadjusted life-years (QALYs). Standard CDM utility values for diabetes-related complications and weight were included in the model. a discount rate of 3.5% was applied. RESULTS: Treatment with EQW compared to insulin glargine was associated with a higher life expectancy (11.91 vs. 11.81 years) and quality-adjusted life expectancy (8.019 vs. 7.846 QALYs). The cumulative incidence of all diabetes-related complications was lower for EQW compared to glargine except for stroke and amputation due to recurrent foot ulcer: congestive heart failure (31.25% vs. 32.44%), stroke (29.61% vs. 29.50%), myocardial infarction (20.50% vs. 21.16%), angina (15.73% vs. 16.06%), peripheral vascular disease (12.61% vs. 12.73%), amputation recurrent ulcer (1.22% vs. 1.19%). Moreover, EQW was associated with a longer mean time to onset of first complication versus glargine (4.79 vs. 4.59 years). CONCLUSIONS: Long-term projections based on the findings of a recent randomized controlled trial indicate that EQW is likely to improve life expectancy and quality-adjusted life expectancy, reduce complication rates, and delay the time to onset of diabetes-related complications compared with insulin glargine. PDB10 PREVALENCE AND OUTCOME OF INSULIN ANALOGUE TREATMENT COMPARING GERMANY, UK, AND FRANCE Holz B, Schröder-Bernhardi D, Kostev K IMS Health, Frankfurt, Germany OBJECTIVES: New classes of antidiabetic medications have been introduced, but details of their use are not well known. In Germany, several measures have been undertaken to limit the usage of analogues in T1D and T2D patients. The aim was to assess the usage patterns of short acting (SA) and basal analogue (BA) insulin in primary care in Germany compared to UK and France. METHODS: Computerized patient data from general medicine practices (IMS Disease Analyzer, 01/2007 to 12/2009) throughout Germany, UK, and France have been analyzed on treatment rates and outcome. RESULTS: In 2009 10,792 insulin patients in France, 20,244 in UK, and 103,238 in Germany were analyzed. The dominance of BA and SA insulins is differing by country and diabetes type (Germany: SA insulins in T2D 40.8% vs. UK 95.8% and France 92.3%). Treatment of T1D with analogous insulins is in general higher compared to T2D (Germany: 56.1% BA in T2D vs. 66.3% in T1D; 40.8% SA in T2D vs. 74.3% in T1D). Usage share of analogous insulins increased over the 3-year period in all three countries (T2D patients, Germany: BA usage 47.6% in 2007 vs. 56.1% in 2009; UK: 78.0% in 2007 vs. 83.0 % in 2009, France 82.2% in 2007 vs. 88.5% in 2009). The analysis of HbA1c values in UK and Germany revealed a higher share of patients with uncontrolled HbA1c values (>7.0) for population treated with analogues (70.8% in Germany) vs. human insulin (68.4%) vs. OAD (42.9%). Increase hints to high unmet medical need in human insulin usage and shift to analogues. CONCLUSIONS: Despite several health political measures in Germany favoring human insulins over analogues, the prevalence of both SA and BA insulin usage for diabetes patients continuously increased in Germany in line with UK and France from 2007 to 2009 to meet medical need by higher analogue usage. PDB11 DRUG-USE PATTERNS OF INITIALLY PRESCRIBED INSULIN DETEMIR AND INSULIN GLARGINE IN THE NETHERLANDS; A COMPARATIVE ANALYSIS USING PHARMACY DATA FROM IADB.NL Visser ST1, Vegter S1, Boersma C1, De Grooth R2, Postma MJ1 1 University of Groningen, Groningen, The Netherlands; 2sanofi-aventis The Netherlands B.V., Gouda, Zuid Holland, The Netherlands OBJECTIVES: Newer long-acting insulin analogs have shown to result in several treatment improvements if compared with NPH insulins. Promising results from clinical trials require confirmation from observational settings reflecting potential “real-life” benefits. Therefore, the current study aimed to evaluate trends in prescribing (uptake), treatment adherence, and costs of (newer) insulin analogs (e.g., insulin detemir and insulin glargine). METHODS: A drug-utilization analysis was conducted based on using dailypractice prescription data from IADB.nl (one of the databases available in the Mondriaan infrastructure), a dispensing database of 50 pharmacies, covering a Dutch population of 500,000. Drug-utilization data were collected for a 7-year period (2000–2007). All patients that received at least one prescription of an insulin analog—defined following the “Anatomical Therapeutical Classification” (ATC) as “A10A”—were included for this study. The prevalence and incidence of insulin use were both presented per 1000 population. Survival analysis was conducted to calculate the persistence on different insulin analogs. RESULTS: Preliminary results show a fast increase in the proportion of patients using the newer long-acting insulin analogs with 7% in 2003 to over 30% in 2006. Among patients initiating these newer insulin analogs, use of insulin glargine is more than two times higher compared to use of insulin detemir (507 vs. 190 starters in 2006). Patients that initiated treatment with long-acting insulin analogs were significantly more adherent to insulin glargine compared to insulin detemir (exact figures will be presented during the presentation). CONCLUSIONS: Current preliminary results A285 show an increase in use of long-acting insulin analogs. Of these analogs, insulin glargine is most often initiated and seems to be well-received by patients. Further results will be presented based on a longer follow-up period including differences in HbA1c level dependent dosing and trends in cost development for insulin analogs over time. Also, the potential feature of “channelling” will be investigated. PDB12 TIME TO OPIOID USE AMONG COMMERCIALLY-INSURED PATIENTS WITH DIABETIC PERIPHERAL NEUROPATHIC PAIN WHO INITIATED DULOXETINE VERSUS OTHER TREATMENTS—A PROPENSITY SCORE APPROACH Zhao Y1, Wu N2, Chen SY2, Fraser K2, Boulanger L2 1 Eli Lilly & Company, Indianapolis, IN, USA; 2United BioSource Corporation, Lexington, MA, USA OBJECTIVES: Diabetic peripheral neuropathic pain (DPNP) is often managed using opioids which are subject to tolerance and dependence issues. Newer classes of medications have been approved for DPNP; however, published findings assessing utilization of opioids following treatments with these agents are limited. This retrospective cohort study compared the subsequent use of opioids between commercially insured DPNP patients initiating duloxetine treatment versus other standard of care (SOC) therapies. METHODS: DPNP patients aged 18 to 64 who initiated duloxetine or other SOC treatments (tricyclic antidepressants, venlafaxine, gabapentin, pregabalin) between March 1, 2005 and December 31, 2005 were selected. Initiation was defined as a 90-day period without available study medication. The initial dispense date for the study medication was defined as the index date. Selected patients had no opioid pill coverage during the 90 days prior to the index date, and no depression, neuralgia, or epilepsy diagnosis in the 12-month pre-index period. Duloxetine and SOC patients were matched via propensity scoring (1:1 ratio), controlling for demographics, comorbidities, prior health-care utilization and costs, and prior medication history. Opioid utilization and health-care costs over the 12-month post-index period were compared between study cohorts. RESULTS: A total of 113 patients in each of the duloxetine and SOC cohorts were matched. Duloxetine-treated patients were less likely to use any opioids versus SOC-treated patients (52.2% vs. 79.6%, P < 0.01) over the 12-month post-index period. Compared with SOC patients, duloxetine patients had, on average, 110 days delay in opioid use, two fewer prescriptions dispensed, 32 fewer days on opioids, and 2064 mg lower morphine equivalent dosage (all P < 0.01) over the 12-month post-index period. Duloxetine patients also had significantly lower total ($19,408 vs. $30,470, P < 0.01) and outpatient costs ($7,606 vs. $15,272, P < 0.01). CONCLUSIONS: Among commercially insured DPNP patients, duloxetine treatment appears to be associated with delayed and reduced opioid utilization and lower healthcare costs than SOC therapies. PDB13 DIAPS 79, PRELIMINARY REPORT OF AN OBSERVATIONAL STUDY OF COSTS OF TYPE 2 DIABETES MELLITUS (T2DM) TREATMENT IN THE BRAZILIAN PRIVATE HEALTH-CARE SYSTEM (PHS) DIAPS79 study group1, Bahia L2, Eliaschewitz FG3, Franco L4, Hayashida CY5, Schaan BD6, Forti A7, Cintra M2, Andrade PC8, Barbosa E8, Nita ME8, Rached R8 1 Diabetes Mellitus Type 2 in the Brazilian Private Health Care System DIAPS79, São Paulo, SP, Brazil; 2MedInsight, Rio de Janeiro, RJ, Brazil; 3CPCLIN—Centro de Pesquisas Clínicas, São Paulo, SP, Brazil; 4FMUSP, Ribeirão Preto, SP, Brazil; 5Hospital Santa Cruz, São Paulo, SP, Brazil; 6Clínica Pró Gastro & UFRGS, Porto Alegre, RS, Brazil; 7Centro Integrado de Diabetes e Hipertensão & UFCE, Fortaleza, CE, Brazil; 8Bristol-Myers Squibb Farmaceutica S/A, São Paulo, SP, Brazil BACKGROUND: Despite the progress of clinical trials, clinicians and decision-markers often lack knowledge on how to better promote health-care delivery to patients with T2DM in clinical practice. Outcome research studies can provide effectiveness and economic analysis in actual practice in order to improve clinical performance. OBJECTIVES: To describe the resources used and the costs associated with outpatient treatment of T2DM in the Brazilian PHS. METHODS: This is a cross-sectional analysis of 383 T2DM outpatients treated in five cities by various health plans. Data were collected using info from the previous year by interviewing patients using a validated questionnaire complemented by medical chart review. Direct costs included expenses associated with medications, diagnostic tests, procedures, blood glucose test strips, and health professional visits. Nonmedical direct costs included expenses with diet products. Indirect costs were not assessed. RESULTS: The group had a mean age of 60.5 ± 9.6 years and a mean duration of diabetes of 12.2 ± 8.75 years. Annual outpatient cost of T2DM care was R$10,645.78 per patient, being R$9,534.58 for direct medical costs and R$1,111.20 for nonmedical costs. Costs escalated as duration of diabetes increased (R$8796.68 per patient <9 years vs. R$13,231.74 >19 years of disease; P = 0.05). Those patients with associated microvascular and macrovascular complications (n = 18%) have higher costs (R$15,755.12 per patient) compared to those with microvascular (R$9178.38 per patient; n = 32.4%), macrovascular (R$8726.13 per patient; n = 8.1%), or no complications (R$8246.53 per patient; n = 41.5%). Annual per patient expenses included medications (R$5542.94), blood glucose home monitoring (R$1,014.62), private health professional visits (R$4,264.04), and diet products (R$1,111.20). Health plans expenses per year were due to exams/ procedures (R$588.02 per patient), medications co-payment (R$1,029.00 per patient), and health professional visits (R$734.12 per patient). CONCLUSIONS: The outpatient care costs assessed reveal that diabetes poses a serious social and economic burden to patients and to Brazilian PHS. A286 PDB14 HOSPITALIZATION ATTRIBUTED TO DIABETES MELLITUS AND ITS COMPLICATIONS IN THE PRIVATE HEALTH-CARE SYSTEM IN RIBEIRÃO PRETO REGION, BRAZIL Rosa R1, Bahia L2, Monteiro R3, Asano E4, Nita ME4, Donato BM5, DIAPS79 study grou P6 1 UFRGS, Porto Alegre, RS, Brazil; 2MedInsight, Rio de Janeiro, RJ, Brazil; 3USP, Ribeirão Preto, SP, Brazil; 4Bristol-Myers Squibb S/A, São Paulo, SP, Brazil; 5Bristol-Myers Squibb, Wallingford, CT, USA; 6Diabetes Mellitus Type 2 in the Brazilian Private Health Care System DIAPS79, São Paulo, SP, Brazil INTRODUCTION: Studies about the impact of hospitalizations for diabetes in the Private Health System are unknown in Brazil. OBJECTIVES: To estimate the dimension of hospitalization for diabetes mellitus (DM) and its complication in the private health-care system in Ribeirão Preto (RP) region, São Paulo. METHODS: Hospitalization data from 26 private hospitals of RP region (26 cities) were collected from a database of Social Medicine Department from São Paulo University (CPDH-Centro de Processamento de Dados). These data comprised an average of 35,964 admissions per year (2006–2007) of patients from 30 to 74 years old. Combinations of DM prevalence and hospitalization relative risk for chronic complications and general medical conditions (GMC) were added to DM first-listed hospitalizations (attributable risk methodology). The chronic complications diagnoses were divided into five groups: neurological disease, peripheral vascular disease (PVD), cardiovascular disease (CVD), renal disease, and other diabete-related complications. RESULTS: From the whole admissions, 6037 (17%) hospitalizations were attributed to DM (first-listed, chronic complications, and GMC), 3520 (58.3%) women and 2517 (41.7%) men, 1145 (19%) were from 30 to 44 years old patients, 2284 (38%) from 45 to 59, and 2607 (43%) from 60 to 74. DM as first-listed diagnosis accounted for 313 (5.2%) hospitalizations, chronic complications for 2394 (39.6%), and GMC for 3330 (55.2%). CVD (5.3%) and PVD (8.7%) represented 14% of all hospitalizations in the whole population, contrasting to 34.3% (20.9% and 13.4% respectively) in hospitalizations attributed to DM. Hospitalization rates were very similar between sex, except for cardiovascular disease (25% men vs. 18% women). Admissions due to DM represent 37% out of all hospitalizations for chronic diseases (6478). They represent 28% for neurological diseases, 42% for PVD, 41% CVD, and 20% for renal diseases. CONCLUSIONS: DM and its complications impose a relevant burden to private health-care system in Ribeirão Preto region, especially for PVD and DCV hospitalizations. PDB15 WEIGHT CHANGE FOR PATIENTS WITH TYPE 2 DIABETES: SECULAR TRENDS AND IMPACT OF DIFFERENT COMBINATIONS OF THERAPY Morgan CL1, Poole CD1, Jenkins-Jones S1, Evans LM2, Currie CJ 1 Pharmatelligence, Cardiff, UK; 2University Hospital of Wales, Cardiff, Vale of Glamorga, UK; 3 Cardiff University, Cardiff, Wales, UK OBJECTIVES: To describe the secular trend of weight gain for patients with type 2 diabetes and patterns of weight change associated with different diabetes therapy regimens using data from general clinical practice. METHODS: Data were extracted from the THIN database, where patients with type 2 diabetes were identified. The secular trend in weight for patients with type 2 diabetes was plotted for each year from 1995 to 2008 for both prevalent and for incident cases. Baseline weight at the start of the treatment period was compared to absolute and relative weight at circa 6-, 12-, and 24-month periods and compared using the Wilcoxon signed rank test for different treatment regimens. RESULTS: Mean, standardized prevalent weight increased from 84.9 to 93.6 kg for males and from 75.2 kg to 82.1 kg for females between 1995 and 2008. For incident cases, the respective figures were 89.6 kg to 96.6 kg for males and 78.4 kg to 85.7 kg for females. The total number of valid therapy periods was 105,991, accounting for over 70 different regimen combinations. There were 28 therapy combinations with a minimum of 50 observations which were selected for analysis. Between baseline and 6, 12, and 24 months, there were significant changes in weight for the majority of treatment regimens. The largest weight increase at 12 months was for patients prescribed metformin, insulin, sulfonylurea, and thiazolidinedione (median increase of 4.9 kg [95% CI 1.5 to 7.4, P < 0.001]). The largest weight decrease at 12 months patients prescribed metformin and exenatide with a median decrease of −8.1 kg (95% CI −11.0 to −5.0, P < 0.001). CONCLUSIONS: There was evidence of a secular trend of weight increase for patients with diabetes. As the impact of weight increase upon vascular outcomes is clear, consideration should be given to the affect that different treatment regimens have upon weight change. PDB16 INSULIN GLARGINE IS ASSOCIATED WITH A LOWER INCIDENCE OF DIABETIC FOOT SYNDROME AND MACROVASCULAR COMPLICATIONS COMPARED TO NPH INSULIN IN TYPE 2 DIABETICS UNDER GERMAN REAL-LIFE CONDITIONS Siegmund T1, Dippel FW2, Kostev K3, Lauterbach S4, Fuchs S5, Kotowa W5 1 Städt. Klinikum München GmbH, Munich, Germany; 2Sanofi-Aventis Deutschland GmbH, Berlin, Germany; 3IMS Health GmbH & Co. OHG, Frankfurt am Main, Germany; 4Apotheke Rotes-Kreuz-Krankenhaus, Kassel, Germany; 5IMS Health GmbH & Co. OHG, Nürnberg, Germany OBJECTIVES: The purpose of this study was to evaluate the relationship between the long-acting insulin analogue glargine (GLA) versus Neutral Protamine Hagedorn insulin (NPH) regarding the incidence of diabetic foot syndrome (DFS), myocardial infarction (MI), and ischemic stroke (IS) in type 2 diabetics (T2D) under real-life conditions in Germany. METHODS: A historic cohort study based on a representative German database (IMS® Disease Analyzer) included T2D who started a basal sup- 13th Euro Abstracts ported oral therapy (BOT) with either GLA or NPH between July 2000 and September 2007 (index date) and who provided continuously documented data between 12 months before and 24 months after initiation of BOT. Data were collected from index date until the occurrence of an event (DFS or MI/IS) or until August 2009. Duration of therapy was ≥24 months to demonstrate a potential effect of insulins considered. Therefore, the end points were measured as from the second year after index date. The identification of the end points was according to ICD-10 codes (MI: I22 and I23; IS: I63 and I64) or the original doctors texts (DFS). Kaplan–Meier curves were generated and compared by log-rank tests. Cox proportional hazard models were used to estimate the adjusted hazard ratio (HR) for the incidence of DFS and MI/IS. RESULTS: A total of 23,395 T2D fulfilled the inclusion criteria and started a BOT either with GLA (n = 9,638) or with NPH (n = 13,757). After adjustment for demographic and clinical variables, it could be demonstrated that GLA reduced the relative DFS risk by 64 % in T2D when compared to NPH (HR = 0.611; P = 0.0405). Furthermore, GLA decreased the relative MI/IS-risk by 49 % (HR = 0.671; P = 0.0562). CONCLUSIONS: When compared to NPH, GLA significantly reduces the risk of DFS in T2D under real-life conditions. Additionally, a reduction of the macrovascular events MI/IS was shown with GLA versus NPH. Prospective trials should be conducted to confirm these results. PDB17 GLYCATED HEMOGLOBIN AS A SURROGATE MARKER FOR THE APPEARANCE AND PROGRESSION OF MACRO VASCULAR COMPLICATIONS IN TYPE 2 DIABETES MELLITUS: SYSTEMATIC REVIEW AND META-ANALYSIS Weczorek A1, Marcisz A1, Rys P2, Skrzekowska-Baran I1, Plisko R1, Wladysiuk M1 1 HTA Consulting, Krakow, Poland; 2NovoNordisk, Warsaw, Poland OBJECTIVES: We performed a systematic review and meta-analysis to examine the association between HbA1c and the appearance and progression of macrovascular complications (MVC) in T2DM. METHODS: The two electronic medical databases (MEDLINE, CENTRAL) were searched to identify all papers reporting HbA1c level and macrovascular complications in T2DM. Observational and randomized, controlled trials (RCTs) with at least 1 year of follow-up were included. Estimates were made of the adjusted relative risk (or odds ratio) of complications for an increase in HbA1c of 1%. If data were insufficient to calculate RR, the odds ratio (OR) was estimated. Weighted mean differences (WMD) in HbA1c level between the case group and the control group were also calculated. RR (OR) was estimated for HbA1c increase of 1%. RESULTS: We identified 11 trials that fulfilled the inclusion criteria. Pooled data from two randomized studies showed that RR of the incidence of stroke was 3.40 (95% CI: 1,77; 6.56), myocardial infarction (two studies): 3.19, (95% CI: 1.41; 7.24) and the risk of extremity amputation or vascular surgery caused by peripheral vessels (two studies): 2.12 (95% CI: 1.08; 4.16). The results were confirmed in observational studies for stroke and myocardial infarction. No correlation was found between deaths from cardiovascular disease and HbA1c increase: RR = 1.21, (95% CI: 0.64; 2.28), but it was shown in observational studies: RR = 5.72 (95% CI: 0.76; 43.02). CONCLUSIONS: Relatively low number of studies and the number of observed macrovascular complications impede unequivocal conclusions. However, the results of our systematic review indicate a correlation between HbA1c level and appearance and progression of MVC in T2DM, especially stroke, heart infarction, and peripheral vessel disease. Thus, HbA1c might be considered a surrogate end point for MVC in T2DM, although influence of other factors (e.g., blood pressure, LDL cholesterol) should be considered. DIABETES/ENDOCRINE DISORDERS – Cost Studies PDB18 EVALUATION OF CLINICAL INERTIA IN DIABETES CARE OF NON-INSULIN-DEPENDENT TYPE 2 DIABETIC PATIENTS Font B1, Galera J2, Lahoz R3, Gambús G2, González Clemente JM4 1 Novartis Pharma, Barcelona, Spain; 2Novartis Pharma, Barcelona, Barcelona, Spain; 3Novartis Farmacéutica S.A., Barcelona, Barcelona, Spain; 4Corporació Sanitària Parc Taulí, Sabadell, Barcelona, Spain OBJECTIVES: Clinical inertia (CI) is usually defined as the absence of drug therapy intensification in patients who fail to achieve the objectives of standard treatment. This study aims to describe the CI associated with a noninsulinized patient population with type 2 diabetes (DM2) in Spain. METHODS: Epidemiologic, retrospective, and multicenter study conducted in primary and specialist care consultations from Spain. Inclusion criteria were: adults with non-gestational DM2 and uncontrolled status (HbA1c ≥ 7%), oral hypoglycemic therapy initiated at least 2 years before, and HbA1c values available for a minimum period of three visits. Complete CI was defined as the absence of hypoglycemic treatment intensification in all uncontrolled HbA1c values. Partial CI was defined as the absence of intervention for at least one uncontrolled HbA1c value. RESULTS: 1555 patients who met all inclusion criteria have been included. Mean (SD) patients’ age was 64.7 ± 10.3 years and time from diagnosis was 9.2 ± 5.7 years. Seventy-four percent of patients presented abdominal obesity. Complete CI was estimated as 13.5%, while partial CI was observed in 31.4% of controlled patients and 71.8% of uncontrolled patients. The mean changes of treatment were 1.0 ± 1.02 in controlled patients vs. 1.8 ± 1.2 in uncontrolled patients (P < 0.0001). Significant factors statistically associated (P < 0.05) with current lack of glycemic control were: obesity, sedentary lifestyle, hypertension, hyperlipidemia, total number 13th Euro Abstracts of risk factors, and smoking status. Patients with ≥4 risk factors account for 71.7% of current glycemic control group, while this value reached 84.3% in the uncontrolled group (P < 0.0001). CONCLUSIONS: In one out of seven patients with DM2 and poor glycemic control, none action to intensify treatment has been taken during the past 2 years. Patients without current glycemic control have more than two times higher clinical inertia than the controlled ones. Intensification of treatment is twice as common in patients currently uncontrolled (85.1% vs. 44.9%). PDB19 BUDGET IMPACT ANALYSIS OF SAXAGLIPTIN FOR THE TREATMENT OF TYPE 2 DIABETES IN MEXICAN POPULATION AT PEMEX Juarez-Garcia A1, Martinez-Rivera G1, Anaya P2, Donato BM3 1 Bristol-Myers Squibb, Mexico City, Mexico; 2AstraZeneca, Edo. de Méx., Mexico; 3 Bristol-Myers Squibb, Wallingford, CT, USA OBJECTIVES: Diabetes affects approximately 8 million people in Mexico and is the first cause of death in the country. Ninety percent of all diabetes is classified as type II diabetes (T2DM). Saxagliptin, a DPP4 inhibitor, is one of a class of drugs orally administered for treatment of T2DM. Petroleos Mexicanos (PEMEX) health-care system covers approximately 43,977 patients diagnosed with T2DM. The objectives of this study are to: 1) analyze the current utilization and expenditure for oral antidiabetics (OADs) by PEMEX; and 2) evaluate the budget impact of saxagliptin for treatment of T2DM population. METHODS: An MS Excel-based budget impact model of the total population diagnosed with T2DM in PEMEX was used. OAD usage was based on the total amount purchased, by the Institution in 2009. The prices of medications were taken from the published price listing by PEMEX (2009). The following OAD medications were included in the analysis: pioglitazone, rosiglitazone, vildagliptin, and saxagliptin. Pharmaceutical expenses of OADs were considered excluding other medical costs. The time horizon was 3 years and the assumptions of the model including market dynamics were estimated by Bristol Myers Squibb. The budget impact is reported in terms of additional annual total costs. Results are presented in US dollars with an exchange rate of $13.4 MXN. RESULTS: The usage of saxagliptin in PEMEX represents savings to the institution of US$56,132 for the first year of use, with increases in savings for year two and year three US$102,910 and US$154,441 respectively. The impact of saxagliption on the budget was primarily driven by the gradual substitution of pioglitazone, rosiglitazone, and vildagliptin with saxagliptin over the 3-year of analysis. CONCLUSIONS: The budget impact of adding saxagliptin as a treatment option for T2DM patients reveals that the accumulated savings for PEMEX for the estimated timeframe is around US$ 313,485. PDB20 A PHARMACOECONOMIC MODEL FOR ADJUVANT TREATMENT IN A SMOKING CESSATION INTERVENTION IN PATIENTS WITH TYPE 2 DIABETES MELLITUS AND CHRONIC OBSTRUCTIVE PULMONARY DISEASE Kayadibinli M1, Durguner B1, Ulus P1, Ozdemir B2 1 Pfizer Pharmaceuticals, Istanbul, Turkey; 2Omega Contract Research Organization, Ankara, Turkey OBJECTIVES: The present model was aimed to demonstrate the annual cost of the smoking-related diseases in patients with type 2 diabetes mellitus (DM) and chronic obstructive pulmonary disease (COPD), and to establish the budget that would be saved with varenicline (nicotinic receptor partial agonist) reimbursement by the government. METHODS: The model was carried out in two contexts; the costs of the smoking-related diseases and exacerbations in patients with type 2 DM and COPD were estimated in smoking conditions, and varenicline use. The model was constructed on a total of 900,000 type 2 DM patients and 106,410 COPD patients by assuming approximately 3.8 million type 2 DM patients and 2.0 million COPD patients in Turkey of which 23% and 48% were considered to be smokers, respectively. RESULTS: According to the model, the ratio of patients willing to quit smoking was estimated as 35% in type 2 DM group and 54% in COPD group. Of those, 20% (n = 63,000) type 2 DM patients and 30% (n = 17,238) COPD patients were assessed to use varenicline. The annual cost of the smoking-related diseases and exacerbations was calculated as 72.40 million USD according to the 43,341 events; the unit direct costs for myocardial infarcts, stroke, and congestive heart failure were calculated as US$2,523.55, US$1,930.70, and US$1,412.33, respectively, in type 2 DM patients, whereas it was US$1,567.55 in COPD patients. After varenicline use, the government would save US$9.47 million per year by 5608 preventable events. Moreover, the annual cost of varenicline was estimated to be US$23.46 million for 80,238 patients. Accordingly, the total cost of the smoking-related diseases and exacerbations would be US$13.99 million for the first year. CONCLUSIONS: Varenicline reimbursement decreases the annual cost of the smoking-related diseases and exacerbations in patients with type 2 DM and COPD. PDB21 COMPARISON OF COSTS OF THE INSULIN TREATMENT OF TYPE 2 DIABETES MELLITUS WITH INSULIN GLARGINE AND INSULIN DETEMIR Álvarez Guisasola F1, Mauricio Puente D2, Garcia Coscolin T3, Rubio Terres C4 1 Center of Health La Calzada, Gijón, Asturias, Spain; 2Vilanovaxs Arnau Hospital, Lleida, Spain; 3Sanofi-Aventis, Madrid, Spain; 4Health Value, Madrid, Spain OBJECTIVES: Large published data suggested that some patients initiating with the recommended once-daily detemir administration require twice-daily dosing to opti- A287 mize blood glucose control; therefore, the clinical outcome in this selected population was tested in a randomized controlled trial. The objective of this study is to compare the costs of two treatments of type 2 diabetes mellitus (DM2): Insulin glargine (glargine) and insulin detemir (detemir) in patients with DM2 not controlled with OADs. METHODS: A costs analysis of the insulin treatment was carried out with National Health System perspective. Costs with glargine or detemir linked to DM2 patients were calculated according to the administered doses in a clinical trial of 24 weeks of duration, which included 964 insulin-naive patients, with a DM2 average duration of 10 years and average HbA1c = 8.7%. For both insulin, started dose was 0, 2 U/kg. The administration of a daily dose of glargine and two daily doses of detemir led to similar average levels of glycosylated hemoglobin (−1.46 ± 1.09 and −1.54 ± 1.11 %; P = 0.149). However, patients treated with glargine needed significantly minor insulin than those treated with detemir (43.5 vs. 76.5 U, P < 0.001). RESULTS: The application of this model would be translated in minor daily costs for glargine and concretely its use would suppose an annual cost of c814.52 opposite to the c1461.5 of detemir. In consequence, utilization of glargine instead of detemir would be associated with an annual saving of c647.13 for patient with DM2, which supposes a saving of 44.2% with glargine opposite to detemir. CONCLUSIONS: According to the present model, in the above mentioned population, insulin glargine is a treatment of the DM2 associated with minor costs than insulin detemir. PDB22 COST STUDY OF SELF-MONITORING BLOOD GLUCOSE THROUGH GLUCOMETERS IN PATIENTS WITH DIABETES MELLITUS II IN SPAIN Franch J1, Orozco D2, García A3, Coiduras A4, Buendía C5, Febrer L6, Lizan L7 1 CAP Raval Sud, Barcelona, Catalunya, Spain; 2Research Unit Alicante—San Juan, Alacant, Spain; 3Primary Care Ciudad Real, Ciudad Real, Spain; 4SAP Esquerre, Barcelona, Catalunya, Spain; 5EAP Sants, Barcelona, Catalunya, Spain; 6Bayer Healthcare, Barcelona, Spain; 7 Outcomes’10, Castellon, Spain OBJECTIVES: The objective of this study is to determine the potential differences in the economical impact for the National Health System (NHS) in Spain of using different glucometric systems for self-monitoring blood glucose (SMBG) in patients with type II diabetes mellitus (DM II). METHODS: An economic model was built based on three information sources: 1) literature review; 2) costs databases; and 3) expert opinion. Six different glucometers were analyzed according to their codification characteristics and their corresponding strips package characteristics, leading to four different system possibilities: 1) autocoded/individual strip package; 2) manually coded/ individual strip package; and 3) autocoded/ collective strip package; and 4) manually coded/collective strip package. The perspective was that of the NHS. RESULTS: With more than 2 million people suffering from DM II in Spain, we calculated the impact of glucometer miscoding relating it to cardiovascular episodes and glucose disorders occurred. Also, collective strip package implied, due to shorter time of strip stability after its opening, considerable strip waste especially in those patients with low frequency of SMBG recommendation. Results show that autocoded glucometers which have individual package for strips safe c5 million (10%) versus those manually coded and having individual strip package in diabetic patients’ management to the NHS; versus those autocoded but with collective strip package, savings rise to c15 million (24%) and versus those manually coded and with collective package the amount of saving is of c22 million (31%). The one-way sensitivity analysis performed with the most relevant variables confirmed this tendency. CONCLUSIONS: Glucometric systems not requiring patient intervention for coding and with individual strip package minimize the total cost of SMBG of type II diabetic patients in Spain. PDB23 COST ANALYSIS FOR THE TREATMENT OF DIABETES MELLITUS IN DIFFERENT HEALTH SECTORS OF HEALTH-CARE SYSTEM OF PAKISTAN Malik M, Hussain A, Khan J Hamdard University, Islamabad, Punjab, Pakistan OBJECTIVES: Despite the efforts made worldwide, data regarding the cost of care in Pakistan are scarce. This study was aimed at assessing the direct costs incurred in the treatment of diabetes and its variation among government, semi-government, and private sectors of Islamabad, Pakistan. METHODS: It was a comparative cross-sectional study in which data was collected from patients by conveniently sampling. The direct cost of treatment of diabetes was determined in terms of variables consultation fee, cost of medicine, travelling cost, fees for laboratory test for glucose monitoring, cost of home blood glucose monitoring device, and cost of strips used for home blood glucose monitoring. RESULTS: The annual mean direct cost for government, semigovernment, and private sectors were Rs. 6481.73, Rs. 9785.25 and Rs. 27790.31, respectively, while the monthly mean directs costs were Rs. 540.14, Rs. 815.43, and Rs. 2315.85, respectively. The total direct cost of treatment of diabetes per month for all health facilities was Rs. 986.61. It was found from the analysis that the mean costs for individual variables were highest in the private sector as compared to semi-government and government sector. The consultation fee charged by the private sector is much higher than the other two sectors. CONCLUSIONS: The private sector of Pakistan is incurring more cost for the treatment of diabetes mellitus. This difference in cost among health sectors is a burden on the economic status of the country. Moreover, untreated diabetes or comorbidities increase the overall treatment cost and this can affect the affordability of the patient. A288 PDB24 DIRECT HEALTH CARE COSTS OF DIABETES MELLITUS IN HUNGARY Vokó Z1, Nagyjanosi L2, Kalo Z1 1 Eötvös Loránd University, Budapest, Hungary; 2Syreon Research Institute, Budapest, Hungary OBJECTIVES: Diabetes mellitus is responsible for a huge burden of disease. Our objective was to estimate the direct health-care costs of patients with diabetes in Hungary. METHODS: Real-world data were retreived from the National Health Insurance Fund database. Diabetic patients were defined as persons who filled in a prescription of oral antidiabetics (OAD) or insulin in Q3-Q4 2007. Study population was divided into two groups depending on whether they were hospitalized for major complications of diabetes in 2007–2008. Patients without hospitalization were further divided into three subgroups according to the use of drugs (only OAD, only insulin, OAD and insulin). In all subgroups, we estimated health-care costs for each cost item by age group in the whole study group and among those who actually used a particular service. Additionally, we took samples of patients who were hospitalized for specific complications, and estimated health-care costs for the first and second year after the occurrence of the complication. Hungarian Forint values were converted to Euros by employing the 2008 GDP specific PPP exchange rate (1c = 157.64HUF). RESULTS: Mean health-care cost of 521,545 diabetic patients was c2125 in 2008. It was c4016 for those with hospitalization for complications, c1533 for OAD users without complications, and c2847 for insulin users without complications. Fifty-three percent of the total cost covered drug treatment and 27% acute hospital treatment; 26% of the total drug cost was spent on OADs and on insulin. CONCLUSIONS: Health-care cost of diabetes is already high in Hungary, especially care for its complications. Public health-care cost of diabetes exceeds 0.65% of GDP and 13% of total direct public health-care expenditure. Considering the burden of disease that manifests in premature mortality, reduction in QoL, and high cost, and the epidemiological trends, diabetes mellitus should be a public health priority in Hungary. PDB25 RELATIONSHIPS OF QUALITY OF LIFE AND COSTS WITH CLINICAL CHARACTERISTICS OF DIABETES PATIENTS Cristiani M1, Scalone L2, Morsanutto A3, Moneghini M4, Cortesi PA2, Mantovani LG5 1 Charta Fondation, Milano, Italy; 2University of Milano—Bicocca, Monza, Italy; 3Friuli Venezia Giulia Regional Health Autority, Trieste, Italy; 4University of Trieste, Italy; 5CIRFF, Federico II University, Naples, Italy BACKGROUND: Recent research suggests that direct medical costs and quality of life in diabetes depends on number of diabetes-related complications. OBJECTIVES: To analyze relationships of quality of life and medical costs with clinical characteristics of diabetes mellitus patients. METHODS: A retrospective longitudinal cost of care study was conducted; type 1 and 2 diabetic patients accessing at two hospitals in the north-east area of Italy were recruited between October 2008 and March 2009. At enrollment data on demographic, clinical status and QoL (EQ-5D) were collected. Information on costs occurring during the previous 2 years was obtained from a chart review: hospitalizations, specialist medical visits, diagnostic examinations, drugs, and the main clinical parameters. Costs were quantified from the National Health Service (NHS), by applying tariffs and prices valid in 2009. Data were analyzed with a multivariable linear regression model. RESULTS: A total of 411 valid patients (mean + SD age = 64.1 + 12.7, 56.5% male) were enrolled: 15.9% had type 1, 83.4% type 2 diabetes, and 0.7% had other type of diabetes. Costs were on average c234.36/patientmonth; hospitalization accounted for the greatest proportion of costs (58.5%), followed by pharmacological therapies (32.6%) and diagnostic exams (8.9%). With EQ-5D: VAS was on average + SD = 67.74 ± 16.71. Both Costs and HRQoL showed a linear-positive (costs) and -negative (HRQoL) relationship with number of diabetesrelated complications (diabetic retinopathy, diabetic nephropathy, diabetic neuropathy, ischemic cardiopathy, vascular diseases, and diabetic foot), adjusting for age and gender and type of diabetes. On the contrast, no relationship was found with type of complications. CONCLUSIONS: Long-term complications carry a considerable impact on medical cost and HRQoL. Although apparently costly, strategies aimed to optimize the prevention of the onset of diabetic complications should be considered as a potential investment to gain health and reduce costs in the long run. PDB26 ECONOMIC BURDEN OF PAINFUL DIABETIC PERIPHERAL NEUROPATHY IN KOREA Ko KS1, Cha BY2, Kim CH3, Kwon HS2, Lee JH4, Park TS5, Won JC1, Park HJ6, Ko SK6 1 Inje University Sanggye Paik-Hospital, Seoul, South Korea; 2The Catholic University of Korea, St. Mary’s Hospital, Seoul, South Korea; 3Sejong General Hospital, Bucheon-Si, Gyeonggi-do, South Korea; 4Daegu Catholic University Medical Center, Daegu-City, South Korea; 5 Chonbuk National University Hospital, Jeonju Si, Jeollabuk-Do, South Korea; 6Pfizer Pharmaceuticals Korea Ltd., Seoul, South Korea OBJECTIVES: The painful diabetic peripheral neuropathy (DPN) is the most common complication of diabetes. Despite the prevalence of painful DPN and its potential risk of foot ulcer and amputation, there has been no study about painful DPN on economic burden in Korea. This study was conducted to assess the patient-level economic burden among subjects with painful DPN. METHODS: A cross-sectional multicenter study was performed using a standardized questionnaire, to estimate recent 3-month healthcare and non-health-care cost, and productivity loss of diabetic patients. a total of 4000 patients were recruited from 40 hospitals between December 2009 and May 2010. Cost items mainly included health-care cost such as outpatient, pharmacy, inpatient, and oriental medicine; non-health-care cost such as traffic expenses, nursing cost, complementary, and alternative medicine. Cost included insurance-covered cost 13th Euro Abstracts as well as patient’s out-of-pocket expenses during 3 months. To estimate productivity loss due to morbidity, days away from work due to painful DPN were also investigated. RESULTS: Among 2681 diabetic patients completed questionnaire (response rate = 67.0 %), 26.3% (n = 706) had painful DPN. Numbers of outpatient visit within 3 months were higher in patients with painful DPN compared to those in patients without painful DPN, 3.79 ± 2.83 and 3.25 ± 2.36, respectively (P < 0.01). Total costs over 3 months were also higher in patients with painful DPN than in those without painful DPN (1,049,477 ± 1,549,446 and 721,933 ± 1,394,970 KRW, respectively, P < 0.01); Median costs were higher among patients with painful DPN (656,585 vs. 421,668 KRW). Within 3 months, 8.2% and 43.5% of patients with painful DPN had been away from work and reported the decreased work productivity, respectively. CONCLUSIONS: Painful DPN increased health-care cost and decreased work productivity of diabetic patients in Korea. PDB27 INDIRECT COSTS OF ILLNESS FOR DIABETES IN PORTUGAL 1 2 2 Gouveia M , Borges M , Costa J 1 Catholic University of Portugal, Lisbon, Portugal; 2University of Lisbon, Lisbon, Portugal OBJECTIVES: As in so many other countries, diabetes is one of the largest health problems faced by Portugal. Up to now, there have been no “cost of illness” studies for diabetes in Portugal. This paper provides a contribution to fill that gap by estimating the indirect costs of illness, more specifically the output loss due to short- and long-term disability attributable to diabetes in Portugal. METHODS: The estimates are based on the microdata of the 4th National Health Survey conducted in 2005/2006. An employment logit is estimated with covariates including age, gender, education levels, and regional dummies as well as a dummy for diabetes and dummies for other relevant health conditions for all people in the survey with ages between 20 and 74. a comparison of the baseline labor market participation/employment estimates and model predictions assuming zero diabetes prevalence provides the estimates for the labor market impact of diabetes. The estimates are specified by age groups and gender. At this point, the analysis uses microdata from the Labor Ministry, covering about 3 million workers, to estimate gross wages and employer Social Security contributions by age and gender, allowing us to use the human capital approach to put a value on the diabetes-induced labor market nonparticipation. RESULTS: The nonemployment estimates generated by the logit-based methodology are that diabetes reduces employment by 22,150 in a 4.6 million demographic group. The corresponding output loss is estimated to have been c324 million. CONCLUSIONS: The output loss is one of the main costs of diabetes in Portugal. Its amount is four times larger than the available estimates for diabetes’ attributable inpatient care in National Health Service hospitals. PDB28 HEALTH INSURANCE COST OF DIABETES MELLITUS IN HUNGARY: A COST OF ILLNESS STUDY Gresz M1, Varga S2, Kriszbacher I2, Sebestyén A3, Boncz I2 1 National Health Insurance Fund Administration, Budapest, Hungary; 2University of Pécs, Pécs, Hungary; 3National Health Insurance Fund Administration, Pécs, Hungary OBJECTIVES: Diabetes is one of that chronic diseases in which the increase of the costs cause financial problems for the National Health Insurance Fund in Hungary. Aim of the study is the examination of it in a 5-year period from 2003 to 2008. METHODS: The number of diabetic patients, the type of treatment services, the financial support, and the costs were analyzed. Data were derived from the National Health Insurance Fund Administration. RESULTS: Measuring by the medical ID, the numbers of patient increased by 34% during 5 years. The numbers of case increased in the outpatient services by 14% (from 213,790 to 243,960). The numbers of the hospital treatment day decreased by 5% (3,342,857; 3,168,263) The day off work due to sickness and the cost of it increased by 345% (from 8275c to 36,860c). While financial support of devices for the measurement of blood sugar increased by 93% c4900 to c9400 and support of insulin treatment increased 111%, than the increase of the number of insulin-treated patients was only 46%, 91,920 (2003) and 134,617 (2008). Similarities were seen in numbers (40%) and in costs (94%) in the noninsulin-treated patients from 195,662 to 274,886 and from c7900 to c15,300. The total change in financial support was 72%. CONCLUSIONS: Increase in the number of diabetic patients was seen during this 5-year period. The increase of the number in outpatient services was higher than in inpatient services. The decrease in numbers of hospital bed could explain it. The economic crisis could cause the growing in the numbers of day off work. The increase in the medical financial support compared to the number of patients had to be mentioned, because it could be caused by enhancement in the administration of the innovative medical products. Basic strategy should be found against health cost explosion. PDB29 COST OF DIABETES MELLITUS TYPE 1 AND 2 STUDIES IN COUNTRIES OF CENTRAL AND EASTERN EUROPE—A SYSTEMATIC REVIEW OF THE LITARATURE Kawalec P1, Czech M2 1 Jagiellonian University, Kraków, Poland; 2Novo Nordisk Pharma Sp z.o.o., Warsaw, Poland OBJECTIVES: Despite the rapid development of pharmacoeconomics and outcomes research in new countries joining European Union (EU), there is still scarcity of cost of illness studies compared to old EU members. The aim of the study was to review all studies concerning costs of diabetes type 1 and 2 and its complications in old and new EU members. The following countries were taken into account: Slovenia, Poland, Czech Republic, Hungary, Slovakia, Bulgaria, Romania, Lithuania, Latvia, and 13th Euro Abstracts Estonia—as new EU members and remaining EU countries as old members. METHODS: A systematic review of the following databases was conducted: Medline (PubMed), Embase, Cochrane. In addition, NICE, INAHTA, HTAI, and ISPOR Internet pages were searched. The following key words were used: cost of illness diabetes with name of a country or burden of illness diabetes with name of a country. Studies’ design, methods, scope, and results were compared. Direct and indirect costs were considered separately, detailed costs’ categories and costs related to complications were distinguished. Perspectives, time frames were taken into consideration and current exchange rates for communication purposes were used. Studies published in English language were included. RESULTS: Only 12 studies concerning cost of diabetes type 1 and 2 and their complications in new EU countries were found as compared to 67 from old EU members. On the side of new EU members, one study was found in Lithuania and Bulgaria, three in Czech Republic, and seven in Poland. In old EU countries, one study was found in Luxembourg and Greece, two in Austria, Belgium, Finland, and Ireland, three in France and The Netherlands, seven in Spain and Italy, eight in Great Britain, 13 in Sweden, and 16 in Germany. CONCLUSIONS: Results of this review revealed the necessity of carrying out more studies concerning cost of diabetes and its complications in new EU members. PDB30 USE OF MEDICAL INFORMATION SYSTEM FOR ASSESSMENT OF THE COST OF THERAPY OF DIABETIC FOOT PATIENTS IN RUSSIA Matveev N1, Galstyan G2, Sergeeva S2, Dolotova D3 1 Nycomed Russia-CIS, Moscow, Russia; 2Scientific Center for Endocrinology, Moscow, Russia; 3 Russian State Medical University, Moscow, Russia OBJECTIVES: Diabetic foot (DF) is one of the most important complications of diabetes mellitus. If not treated properly, diabetic foot may often lead to infections, amputations, and finally to disability. Unfortunately, only few studies of the cost of diabetic foot treatment were available in Russia; most of the studies were completed about 10 years ago. We suggested that the data of earlier studies could not reflect present level of expenditures. Therefore, we launched a study of the cost of diabetic foot treatment in present conditions. METHODS: We have analyzed the data of 146 inpatients treated in diabetic foot department at Scientific Centre for Endocrinology (Moscow, Russia) in 2008–2009. The patients’ data were input into a specially designed medical information system. The patients were sorted into three main groups: 1) diabetic polyneuropathy without diabetic foot (N = 37); 2) diabetic foot without amputations (N = 58); and 3) patients with amputations due to diabetic foot (N = 51). To calculate the total cost of DF treatment, costs of diagnostic procedures, medicines, bandaging, surgical operations, and staying in the hospital were summed. RESULTS: Mean cost of treatment of one DF patient was equal to 81,700 rubles (US$2645), which is about 30% higher than previously reported figures. It is mainly due to larger introduction of recombinant insulins into routine treatment of diabetic patients. Mean cost of treatment of DF patients with amputations was significantly higher than in those without amputations, mostly due to additional costs of surgical treatment and longer stay in a hospital. Moreover, the mean cost of medicines for DF patients with amputations was almost twice higher than for DF patients without amputations. CONCLUSIONS: The cost of diabetic foot treatment in Russia increased approximately 30% during last 10 years. The results will be used to assess costeffectiveness of various drug treatments of diabetic foot. PDB31 THE IMPACT OF NEUROMONITORING ON THYROID SURGERY COSTS 1 1 2 3 Beccagutti G , Grifi M , Pantaleoni M , Dionigi G 1 Medtronic Italia, Sesto San Giovanni, Italy; 2IMS Health S.p.A, Milano, Italy; 3Università dell’Insubria, Varese, Italy OBJECTIVES: Damage to the recurrent laryngeal nerve (RLN) is one of the principal reasons for malpractice claims against surgeons in otorhinolaryngology. Intraoperative neuromonitoring (IONM), facilitating the identification of RLN, has been demonstrated to be a valid technical support to reduce intraoperative risks. Aim of this study was to evaluate the additional hospitalization costs for thyroidectomy due to IONM, against the mentioned clinical and administrative advantages. METHODS: The study was performed in an experienced Italian University Hospital, in which the learning curve for this technology is considered completed. Through a microcosting approach, the thyroidectomy patient care process (with and without IONM) was costed considering direct costs only (staff time, consumables, equipment, drugs, operating room, and general expenses) and according to the hospital perspective. Unit costs were collected from hospital accounting and standard tariff lists. a differential analysis was performed to highlight additional resource consumption (time effort, consumables, technology equipment) due to IONM usage. To assess the impact of the technology on hospital management, three scenarios were considered: 1) traditional thyroidectomy; 2) thyroidectomy with IONM in a high-volume setting (five procedures per week); and 3) thyroidectomy with IONM in a low-volume setting (one procedure per week). RESULTS: The cost for hospitalization for a traditional thyroidectomy was c3471. If IONM is used, costs increase as follows: +7% in a high-volume setting (c3713) and +9% in a low-volume setting (c3770). IONM therefore represents only the 6% to 7% of the total hospitalization costs. CONCLUSIONS: IONM for thyroid surgery could reduce the risk of RLN damages and of the consequent malpractice claims against a very low impact on hospital budget, accounting only for the 7% of the hospitalization costs for a thyroidectomy. Considering that IONM could be useful in all surgical procedures where nerves are at risk, the economic impact could be even lower due to a higher level of usage of the equipment. A289 PDB32 COST-EFFECTIVENESS CONSEQUENCES OF OBESITY IN T2DM BY INSULIN ANALOGUE THERAPY Józsa Z1, Tóth E2, Nagy B2 1 Novo Nordisk Hungary Ltd., Budapest, Hungary; 2Healthware Ltd., Budapest, Hungary OBJECTIVES: The available T2DM (type 2 diabetes mellitus) therapies attend with weight gain, which affects for the disease outcomes in a very extent way. Weight gain could worsen the long-term effect of insulin therapy (increased risk of insulin resistance, hypertension, dyslipidemia), which influences the total treatment cost. Longacting insulin analogue therapies (insulin detemir and glargine) offer improved pharmacokinetic and pharmacodynamic properties compared to regular human insulin. In a previous meta-analysis, detemir caused significant lower weight gain than glargine (−2.04 kg after 6 months therapy). Our research aimed to determine the effects of weight gain and calculate the total treatment cost. METHODS: On the basis of a systematic literature review, we found many consequences, which could be obesity related. Because of its important effect and bride prevalence, we analyzed the CHD (coronary heart disease) risk changes related to the weight gain. We modified the 10 years risk—calculated with the UKPDS risk engine—with the effects of the weight gain published in a meta-analysis. We aggregated the differences in natural outcomes (number of avoided fatal and nonfatal events) and in monetary unit, based on realworld data. RESULTS: In our analysis, we assessed that insulin detemir-treated patients—due to the favorable features regarding weight gain compared to glargine— experience smaller (cca 2%) CHD risks. It means, for the current treated Hungarian population, that there are 223 coronary (158 fatal of this) events avoidable in 10 years horizon using detemir insulin instead glargine. The estimated savings of social insurance would be around 224 million HUF on the basis of a Hungarian burden of diabetes complication study. CONCLUSIONS: Obesity and weight gain-related aspects should be priorized as the main international tendencies showed. It is not only necessary on the policy level, but also in “individual” level in the cost-effectiveness analysis as well. PDB33 EVALUATION OF COST AND CLINICAL OUTCOMES BY HBA1C AT DIAGNOSIS USING VARIOUS DIABETES TREATMENT STRATEGIES Lee LJ1, Klein RW2, Klein TM2, Furiak N2, Peltz G1, Bansal M2, Jackson JA3, Juneja R4 1 Eli Lilly & Company, Indianapolis, IN, USA; 2Medical Decision Modeling Inc., Indianapolis, IN, USA; 3Lilly USA, LLC, Indianapolis, IN, USA; 4Indiana University, Indianapolis, IN, USA OBJECTIVES: To examine the association of HbA1c at diagnosis with cost and clinical outcomes in patients with type 2 diabetes (T2D) using three different diabetes treatment strategies. METHODS: An existing Monte Carlo diabetes model was used to generate demographic and clinical characteristics for 10,000 random patients simulated from the time of T2D diagnosis. Fixed initial HbA1c values were assigned (range: 7–12%). Input distributions were derived from the literature, with diabetes complications occurring in: retinopathy, nephropathy, neuropathy, coronary heart disease, and stroke. The diabetes-related costs, percentage of patients reaching HbA1c target (<7%), complication events, and mortality over 10 years were evaluated using the following treatment strategies: (S1) addition of oral antihyperglycemic agents (OHAs) at 3-month intervals then starting insulin after 9 months; (S2) addition of OHAs at 6-month intervals then starting insulin after 2 years; and (S3) addition of OHAs only. All treatment strategies began with metformin then sulphonylurea and/or thiazolidinedione were added if target was not reached. RESULTS: Diabetes-related costs increased as initial HbA1c increased for all three strategies. S1 had the greatest increase ($5300–$7750) followed by S3 ($4200–$5250) per 1% HbA1c increase from 7% to 10%. S2 had the smallest rate of increase ($3350–$4950). S3 was the least costly until HbA1c exceeded 10%; however, even at HbA1c of 8%, S3 had fewer patients ever reaching target (S3 = 79% vs. S1 and S2 = 95% [standard errors, SE < 0.41%]). For 10,000 patients with initial HbA1c of 9%, the total counts of complications were: S1 = 4360; S2 = 4126; S3 = 5009 (SE < 67) with mortality rates of 42.5%, 41.9%, and 45.2% (SE < 0.56%), respectively. CONCLUSIONS: In this model, S2 had the lowest complication rates and mortality in patients with T2D. Starting HbA1c affected S1 cost more than other strategies. Strategies with other treatments or alternative timing strategies can be specified and analyzed using this model. PDB34 THE ECONOMIC VALUE OF THE EASYPOD® ELECTRONIC AUTOINJECTOR IN IMPROVING THE RESPONSE TO GROWTH HORMONE (GH) IN CHILDREN WITH IDIOPATHIC GROWTH HORMONE DEFICIENCY (IGHD): A COST-CONSEQUENCE ANALYSIS Chatelain P1, Latour S2, Maetzel A3 1 Université Claude Bernard Lyon—Hôpital Femme Mère Enfant, Bron, France; 2Merck Serono S.A., Geneva, Switzerland; 3Stratas Partners, Basel, Switzerland OBJECTIVES: Response to GH therapy in children with IGHD can be further optimized. To evaluate the economic benefit of injecting GH with easypod®, an electronic autoinjector that objectively monitors drug administration, enabling differentiation of poor adherers from low responders. METHODS: A discrete event simulation model was developed to model continuous, intermittent (four injections/week) and discontinued GH usage in children with IGHD until final height. a cohort of children (age: 4–12 years, growth delay: −4.0– −2.5 standard deviation scores [SDS] at baseline) was modeled to initiate GH (0.03 mg/kg/day). Annual height gains of 1.2 to 0.8 SDS in year 1 were assumed to be 30% and 60% lower in each subsequent year of continuous and intermittent use, respectively. Baseline nonadherence was 9.3 persons per 100 person- A290 years, of which 20% (1.86%) discontinued and the remainder used GH intermittently. a one-time identification of nonresponders with GH dose increase of 50% was assumed after month 6 for children with low growth gains. The model was validated against final height gains observed in large international registries of children with IGHD. The rate of poor adherence was reduced by 10% and 30% in sensitivity analyses. Cost of GH was c32.80 per mg, and discounted at 3.5% per annum. RESULTS: The baseline cohort was modeled to achieve near final height SDS of −1.0 (± 1.23) as observed in real-life registries, effectively gaining 16.0 cm at a cost per cm gained of c3929. Identification of low responders increased height gain to 17.3 cm at a cost per cm gained of c3987. Growth gains were maximized at 18.2 cm at a cost per cm gained of c4073 when identifying low responders and decreasing poor adherence by 30%. CONCLUSIONS: Easypod® features can identify adherence profiles, allowing early recognition of low responders, giving the potential to optimize final height gain and investment in GH therapy in IGHD. PDB35 COST-CONSEQUENCE ANALYSIS OF CSII VS. MDI: A CANADIAN PERSPECTIVE Sadri H1, Bereza BG2, Longo CJ3 1 Medtronic of Canada, Brampton, ON, Canada; 2University of Toronto, Toronto, ON, Canada; 3McMaster University, Hamilton, ON, Canada OBJECTIVES: Several studies have evaluated the ICER of CSII versus MDI over a patients’ lifetime. However, decision-makers are also concerned with shorter time horizons. As such, the annualized incremental cost and consequences of CSII versus MDI may be informative to this audience. METHODS: An economic model was developed using Excel spreadsheet. Deterministic data derived from various sources including ICES Diabetes Atlas, Statistics Canada Population, and published literature was used to determine the annualized costs and consequences of CSII versus MDI in the Province of Ontario. RESULTS: Overall, the use of CSII for all type 1 diabetic patients (T1DM) age 18–64 in the base-case analysis would result in savings associated with the event reductions of $77,885,000 per year (95% CI: 55,920,148–123,835,211). Assuming all T1DM patients switched to CSII, annual resource saving include: 63 fewer AMI (95% CI: 20.8–233.0), 23 fewer major amputations (95% CI: 6.4–133.9), 14 fewer blindness cases (95% CI: 3.6–224.5), 411 fewer dialysis years (range 329.2– 493.8), and 140 fewer hypoglycemia-related hospitalizations (range 112.3–168.5). Additional investments in CSII would result in a net expenditure of $4,724,000. However, extensive sensitivity analysis shows that there is a potential cost saving of $26,689,000–$41,225,000 for the province in direct medical cost assuming all 48,000 T1DM patients adopted this technology. CONCLUSIONS: Use of CSII has demonstrated improved glycemic control compared to MDI. Additional health benefits including reductions in AMI, amputations, cataract surgery, dialysis rates, and hypoglycemia-related hospitalization could potentially save the health-care system both costs and precious resources. Sensitivity analyses suggest that under the most favorable conditions, adoption of CSII would be a dominant strategy (saving $41,000,000/ annum in the steady state). This analysis underestimates the CSII benefits by ignoring smaller benefits including minor amputations, costs associated with gangrene. Also, the effects on patient’s HR-QoL and other socioeconomic benefits related to lost productivity has not been calculated. PDB36 LONG-TERM COST-CONSEQUENCE ANALYSIS OF EXENATIDE ONCE WEEKLY VERSUS SITAGLIPTIN OR PIOGLITAZONE IN TYPE 2 DIABETES PATIENTS IN THE UNITED STATES Aledort Gaebler J1, Samyshkin Y2, Guillermin ALG2, Lloyd A2 1 Amylin Pharmaceuticals, Inc., San Diego, CA, USA; 2IMS Health, London, UK OBJECTIVES: Exenatide once-weekly (EQW) is a GLP-1 receptor agonist that controls glucose in patients with type 2 diabetes (T2DM) and is associated with weight loss and improvement in cardiovascular risk factors such as blood pressure and lipids. We estimated the long-term clinical and financial benefits of EQW compared to sitagliptin or pioglitazone in the United States. Pending US approval and EQW price, we conducted a cost-consequence analysis that excluded pharmacy costs. METHODS: We used the CORE Diabetes Model, a validated computer simulation model of diabetes progression and treatment, to project lifetime clinical outcomes and complication costs. Patient characteristics (mean age 52.5 years; diabetes duration 6 years; HbA1c 8.51%; body mass index 32.12 kg·m−2) and clinical data were derived from a phase 3 clinical trial comparing EQW to sitagliptin or pioglitazone against metformin background. Compared to sitagliptin and pioglitazone patients, trial patients on EQW had HbA1c improvement (0.63% and 0.32% points, respectively) and BMI reduction (0.51 kg·m−2 and 1.77 kg·m−2). Health outcomes and complication costs were discounted at 3% per year. Sensitivity analyses were performed to determine how changes in estimated values affected results. RESULTS: Compared to sitagliptin and pioglitazone, EQW increased life expectancy by 0.31 years (13.80 ± 0.18 vs. 13.50 ± 0.16) and 0.14 years (13.80 ± 0.18 vs. 13.66 ± 0.18), respectively, and quality-adjusted life-years (QALY) by 0.30 (9.49 ± 0.12 vs. 9.20 ± 0.11) and 0.21 (9.49 ± 0.12 vs. 9.28 ± 0.12). EQW was also associated with lower complication costs: Compared to sitagliptin and pioglitazone, EQW saved US$1654 (US$55,385 ± 1781 vs. US$57,039 ± 1848) and US$573 (US$55,385 ± 1781 vs. US$55,958 ± 1838) lifetime direct cost per patient. In both cases, cost savings resulted mainly from lower cumulative incidence of cardiovascular diseases and neuropathic complications (e.g., ulcer, amputation). CONCLUSIONS: Over a patient’s lifetime, EQW is projected to improve health and decrease complication costs compared with sitagliptin or pioglitazone. Incremental cost-effectiveness ratios will depend upon EQW price. 13th Euro Abstracts PDB37 THE COST-EFFECTIVENESS OF SAXAGLIPTIN VERSUS NPH INSULIN IN THE TREATMENT OF TYPE 2 DIABETES MELLITUS (T2DM) IN POLAND Kolasa K1, Niewada M2, Puelles Fernandez de Troconiz J3, Townsend R4, McEwan P4 1 Bristol-Myers Squibb, Warsaw, Poland; 2HealthQuest sp z o.o., Warszawa, Poland; 3 Bristol-Myers Squibb, Braine l’Alleud, Belgium; 4CRC, Cardiff, UK OBJECTIVES: To compare the cost-effectiveness of saxagliptin versus NPH insulin as second-line therapy in combination with either metformin (MET) or sulphonylurea (SU) after failure of monotherapy with MET or SU in a Polish setting. METHODS: The perspective was the Polish National Health Fund using a published fixed time increment, stochastic simulation model set to a 40-year time horizon. Disease progression was taken from the UKPDS 68 outcomes study. Relative effectiveness for change in Hba1c, weight, and hypoglycemia was determined from published trials. Costs were from published estimates and local data. Utilities were derived from UKPDS outcomes supplemented with information from published sources for hypoglycemia-, weight-, and injection-associated disutility. Costs and health-related effects were discounted annually at 5% and 3.5%, respectively. RESULTS: When compared with insulin + MET, saxagliptin + MET was associated with reduced severe hypoglycemia and less weight gain, resulting in an incremental benefit of 0.13 quality-adjusted life-years (QALYs) and an incremental cost-effectiveness ratio (ICER) of 27,454 zloty (PLN) per QALY. Treatment with saxagliptin + SU was associated with a reduction in symptomatic and severe hypoglycemia when compared with insulin + SU, leading to an incremental benefit of 0.14 QALYs and an ICER of 24,663 PLN per QALY gained. Key model drivers were: baseline HbA1c, treatment-associated weight gain, thresholds for switching treatment, age, and disutility associated with weight gain, injection fear, and hypoglycemia. The results were robust to various assumptions concerning inputs and modeling parameters, with all ICERs < 50,000 PLN per QALY gained (GDP per capita for 2009 was 30,000 PLN or c7300, based on 1 PLN = EUROS 0.243, June 2010). CONCLUSIONS: Saxagliptin is cost-effective as a second-line therapy in combination with MET or SU in T2DM in the Polish setting. The availability of saxagliptin will provide T2DM patients with an additional treatment option to insulin. PDB38 ADDITION OF INCRETIN-THERAPY TO METFORMIN IN TYPE-2DIABETES MELLITUS (T2DM): COST-EFFECTIVENESS OF LIRAGLUTIDE VERSUS SITAGLIPTIN FROM THE PERSPECTIVE OF THE GERMAN STATUTORY HEALTH INSURANCE (SHI) Schlander M1, Mentrup S2, Lund N3 1 Institute for Innovation & Valuation in Health Care, Wiesbaden, Germany; 2Novo Nordisk Pharma GmbH, Mainz, Germany; 3Novo Nordisk A/S, Soeborg, Denmark BACKGROUND: The novel incretin analogue, once-daily liraglutide, mimics the effect of endogenous glucagons-like peptide 1 (GLP-1). Liraglutide was found to be significantly more effective than the DPP-4 inhibitor, sitagliptin (increasing the half-life of GLP-1 by inhibiting its degradation by DPP-4), in lowering HbA1c and weight in patients with T2DM (Pratley et al. 2010). OBJECTIVES: To compare, from a German payer’s (SHI) perspective, the long-term clinical and cost-effectiveness of liraglutide (1.2 mg or 1.8 mg OD) versus sitagliptin (100 mg OD), in combination with metformin, in T2DM patients, based on data from the randomized clinical trial (RCT) by Pratley et al. 2010. METHODS: RCT data were used to populate the CORE Diabetes Model (CDM), calibrated to clinical study baseline characteristics and background mortality in a German diabetes population. Costing (direct costs only) was done from the SHI perspective for year 2009. a discounting rate of 3% was used for costs and clinical effects. For hypothetical cohorts of 1000 patients, each followed for 20 years, liraglutide or sitagliptin were assumed to be maintained for 5 years and patients to be subsequently switched to insulin. Consequences (costs and effects) were projected over a 20-year time horizon. RESULTS: Estimated 20-year survival rates were higher for liraglutide 1.8 mg (41.1%) and 1.2 mg (40.3%) compared to sitagliptin 100 mg (39.5%), and cumulative costs were c47,436, c45,627, and c43,298, respectively. Base-case ICERs were c37,163 per life-year gained for liraglutide 1.8 mg (or c48703 for 1.2 mg) versus sitagliptin, and c20,702 (or c20,870 for liraglutide 1.2 mg) per QALY gained. Sensitivity analyses, including longer time horizons, different risk (CDM base case, Framingham study, and UKPDS) cohorts, indicated robustness of findings. CONCLUSIONS: Long-term projections combining RCT data with the CDM indicate an acceptable to attractive cost-effectiveness of liraglutide compared to sitagliptin (both plus metformin) according to currently prevailing standards. PDB39 EFFECT OF VARYING MODELLING ASSUMPTIONS ON THE COSTEFFECTIVENESS OF LIRAGLUTIDE 1.2 MG (AS A COMPONENT OF DUAL THERAPY) IN TYPE 2 DIABETES Valentine WJ1, Chubb BD2, Smith IC2 1 Ossian Health Economics and Communications, Basel, Switzerland; 2Novo Nordisk, Crawley, West Sussex, UK OBJECTIVES: To compare the effect of different sets of modeling assumptions on the cost-effectiveness of liraglutide 1.2 mg, as a component of dual therapy in type 2 diabetes. METHODS: Data from three published clinical trials were used as the basis of the modeling analysis, comparing: liraglutide 1.2 mg versus rosiglitazone 4 mg, both added to glimepiride (Marre 2009); liraglutide 1.2 mg versus glimepiride 4 mg, both added to metformin (Nauck 2009); and liraglutide 1.2 mg versus sitagliptin 100 mg, both added to metformin (Pratley 2010). The published and validated CORE Diabetes Model was used to simulate the progression of type 2 diabetes and predict costs (in 2009 Pounds Sterling [£]) and quality-adjusted life-years (QALYs) over 13th Euro Abstracts patients’ lifetimes. Duration of treatment was assumed to be 5 years, before switching to a basal insulin regimen. Changes were made to the base-case assumptions. First, the utility change per BMI unit gained (in patients with a BMI > 25 kg/m2) was decreased from −0.01 to −0.0061. Second, the treatment duration was increased from 5 to 10 years. All other variables were kept constant. RESULTS: In the comparison with rosiglitazone, liraglutide was associated with a base-case incremental cost-effectiveness ratio (ICER) of £6226 per QALY gained, which increased to £7545 with BMI utility changes and to £16,477 when the treatment duration was increased. Similar increases were seen for comparisons with glimepiride (£13,257 (base case) to £25,343 (BMI utility change) and £38,368 (10-year treatment)) and sitagliptin (£9,851 [base case] to £14,616 [BMI changes] and £17,089 [10-year treatment]). CONCLUSIONS: Increasing the treatment duration and decreasing the impact of BMI on quality of life increased the ICER of liraglutide versus comparators. Liraglutide was shown to be cost-effective in dual therapy (assuming a threshold of £20,000 per QALY gained) versus rosiglitazone and sitagliptin in all three scenarios (base case, BMI utility changes, and 10-year treatment). PDB40 THE ECONOMIC IMPORTANCE OF “METABOLIC MEMORY” IN THE TREATMENT OF TYPE 2 DIABETES MELLITUS (T2DM) IN SWEDEN Willis M1, He J2, Neslusan C2, Chicevic E3 1 IHE, Lund, Swedenl; 2Johnson & Johnson Pharmaceutical Services, LLC, Raritan, NJ, USA; 3 Janssen-Cilag EMEA, Prague, Czech Republic OBJECTIVES: Analysis of the post-interventional follow-up of the UKPDS found that the benefits of intensive therapy persisted even 10 years after the trial, a finding consistent with “metabolic memory” (i.e., early metabolic status influences long-term outcomes). We assessed the potential impact of “metabolic memory” on the costeffectiveness of intensive versus conventional care in Sweden. METHODS: We used the Economic and Health Outcomes (ECHO)-T2DM model to simulate lifetime health outcomes (including QALYs) and medical costs for 500 cohorts of 2000 newly diagnosed patients. In each cohort, patients were randomized to intensive or conventional care and HbA1c treatment effects corresponding to the UKPDS study were applied for the first 10 years. Consistent with findings from the follow-up study, HbA1c values were assumed to converge by year 10. Subsequently, in the intensive care arm, “metabolic memory” effects (reduced risk for certain microvascular and macrovascular events and mortality) were applied. Two sets of “metabolic memory” effects from the UKPDS follow-up study were used: those observed in the sulfonylurea/insulin subsample and those observed in the metformin subsample. a scenario assuming no “metabolic memory” effects was simulated for comparison. Unit costs were derived from the Swedish literature (modeling studies and a regression analysis of inpatient care costs based on administrative hospital data linked to the Swedish National Diabetes Register). RESULTS: Including “metabolic memory” had a large effect on the cost-effectiveness estimates. The incremental cost-effectiveness ratio declined from SEK2,387,292 (~c250,000) without “metabolic memory” to SEK731,308 (~c75,000) assuming effects from the sulfonylurea/insulin sample and to SEK445,425 (~c45,000) assuming effects from the metformin sample. CONCLUSIONS: These results suggest that good glycemic control early in the disease continuum may confer significant medical cost savings over the long term. Evaluations of the cost-effectiveness of intensive glycemic control in newly diagnosed patients should potentially consider the health and cost consequences of “metabolic memory.” PDB41 COST-EFFECTIVENESS OF LIRAGLUTIDE IN PEOPLE WITH TYPE 2 DIABETES IN THE SLOVAK REPUBLIC Ilavska A1, Uliciansky V2, Wrona W3, Lacka J4, Czech M5 1 ISPOR Slovakia Regional Chapter, Bratislava, Slovak Republic; 2Via medica, s.r.o., Košice, Slovak Republic; 3HealthQuest sp z o.o., Warsaw, Poland; 4Novo Nordisk Slovakia s.r.o., Bratislava, Slovak Republic; 5Novo Nordisk Pharma Sp z.o.o., Warsaw, Poland OBJECTIVES: This study aimed to assess cost-effectiveness of treatment with liraglutide added to a standard therapy with metformin (MET) or/and sulphonylurea (SU) compared with rosiglitazone and exenatide. METHODS: Our study used a health economic model (the CORE Diabetes Model) to project the long-term costs and clinical outcomes of liraglutide based on clinical data from LEAD-1 trial: liraglutide + SU versus rosiglitazone + SU; and LEAD-6 trial: liraglutide + MET+/−SU versus exenatide + MET+/−SU. The analysis was performed from the Slovak health-care services payer’s perspective in a 20-year time horizon. The analysis used health-state utility values from published sources to assess the effect of treatment on QALYs. The unit costs of treatment and complications were derived from published sources and official tariff lists for health-care services paid by public payer. All figures are shown in EUR. RESULTS: QALYs increased with liraglutide 1.2 mg + SU versus SU + rosiglitazone 4 mg by 0.203. Total costs increased by c2561 resulting in incremental costs per QALY of c12,615. The incremental cost-effectiveness ratio for liraglutide 1.8 mg + MET+/−SU versus exenatide 10 µg + MET+/−SU was estimated at c24,013 per QALY gained (QALYs increased by 0.112). CONCLUSIONS: Using the CORE Diabetes Model and data from the LEAD 1 and LEAD 6 trials treatment with liraglutide is a cost-effective intervention compared with both rosiglitazone and exenatide. Sensitivity analysis showed the results are only moderately changing when altering the key parameters and assumptions. A291 PDB42 INCRETIN-THERAPY IN ADDITION TO METFORMIN AND/OR SULFONYLUREA FOR TYPE 2 DIABETES MELLITUS (T2DM): COSTEFFECTIVENESS OF LIRAGLUTIDE VERSUS EXENATIDE FROM THE PERSPECTIVE OF THE GERMAN STATUTORY HEALTH INSURANCE (SHI) Schlander M1, Mentrup S2, Lund N3 1 Institute for Innovation & Valuation in Health Care, Wiesbaden, Germany; 2Novo Nordisk Pharma GmbH, Mainz, Germany; 3Novo Nordisk A/S, Soeborg, Denmark BACKGROUND: Unlike most other antihyperglycemic drugs, glucagon-like peptide 1 (GLP-1) receptor agonists have a glucose-dependent action and promote weight loss. In a randomized clinical trial (RCT) over 26 weeks reported by Buse et al. (2009), the novel GLP-1 agonist liraglutide (1.8 mg once daily, OD) was found to be significantly more effective than exenatide (10 µg twice daily) in terms of HBA1c reduction. OBJECTIVES: To compare, from a German payer’s (SHI) perspective, the long-term clinical and cost-effectiveness of liraglutide (1.8 mg OD) versus exenatide (10 µg BID) in T2D patients, based on data from the randomized clinical trial (RCT) by Buse et al. 2009. METHODS: The CORE Diabetes Model (CDM) was applied using RCT data. The model was calibrated to RCT baseline characteristics and background mortality in a German diabetes population, applying epidemiological data from longterm studies including UKPDS and Framingham to project morbidity and mortality of T2DM. Unit costs for direct costs were applied from an SHI perspective for year 2009. a discounting rate of 3% was used for costs and clinical effects. For hypothetical cohorts of 1000 patients, each followed for 20 years, patients were assumed to be maintained on liraglutide and exenatide, respectively, for 5 years and subsequently switched to insulin. Costs and effects were projected over a 20-year time horizon. RESULTS: Estimated 20-year survival rates were higher for liraglutide 1.8 mg once daily (36.8%) compared to exenatide 10 µg administered twice daily (35.6%), and cumulative costs were c46,308 (liraglutide) and c45,025 (exenatide), respectively. Base-case ICERs were c16,632 per life-year gained for liraglutide versus exenatide, and c11,606 for liraglutide versus exenatide per QALY gained. Sensitivity analyses indicated robustness of findings. CONCLUSIONS: Long-term projections combining RCT data with the CDM strongly suggest an acceptable to attractive cost-effectiveness of liraglutide compared to exenatide according to currently prevailing standards. PDB43 LONG-TERM COST-EFFECTIVENESS OF LIRAGLUTIDE VS. SULPHONYLUREA IN POLAND Niewada M1, Wrona W1, Czech M2, Schubert A2, Skrzekowska-Baran I2 1 HealthQuest sp z o.o., Warsaw, Poland; 2Novo Nordisk Pharma Sp z.o.o., Warsaw, Poland OBJECTIVES: To assess the long-term cost-effectiveness of treatment with liraglutide on top of standard therapy with metformin (MET) compared with sulphonylurea (SU) in people with type 2 diabetes. METHODS: The extensively published and validated CORE Diabetes Model was populated with the clinical data from LEAD 2: liraglutide + MET versus SU + MET. The analysis was performed from the polish health-care services payer’s perspective. a 20-year time horizon was chosen to reflect the costs and outcomes of diabetes as these are often only seen in the later stages of the disease. The analysis used health-state utility values from published sources to assess the effect of treatment on QALYs. The unit costs of treatment and complications were derived from published sources or based on expert opinion survey and Polish official tariff lists for health-care services paid by public payer. All figures are shown in EURO (1 EURO = 3.9 PLN). RESULTS: QALYs increased with liraglutide 1.2 mg + MET versus SU + MET by 0.191. Total costs increased by c3,349 resulting in incremental costs per QALY of c17,565. The incremental cost-effectiveness ratio for liraglutide 1.8 mg + MET versus SU + MET was estimated at c24,842 per QALY gained (QALYs increased by 0.207). Sensitivity analysis showed the results to be moderately changing when altering the key parameters and assumptions (for liraglutide 1.2 mg range from 12,944 to c30,275/QALY). CONCLUSIONS: Treatment with liraglutide is a cost-effective intervention compared with sulphonylurea and is likely to represent good value for money in Polish setting. PDB44 THE COST-EFFECTIVENESS OF SAXAGLIPTIN VERSUS SULFONYLUREA (SU) IN THE TREATMENT OF TYPE 2 DIABETES MELLITUS (T2DM) IN GERMANY Erhardt W1, Bergenheim K2, Townsend R3, Puelles Fernandez de Troconiz J4, McEwan P3 1 Bristol-Myers Squibb GmbH & Co. KGaA, München, Germany; 2AstraZeneca, Mölndal, Sweden; 3CRC, Cardiff, UK, 4Bristol-Myers Squibb, Braine l’Alleud, Belgium OBJECTIVES: This study evaluates the long-term economic consequences of saxagliptin versus sulfonylurea (SU, glipizide) as second-line therapy when used in combination with metformin (MET) after failure of monotherapy treatment with MET, in patients with type 2 diabetes mellitus (T2DM) in Germany. METHODS: A published discrete event simulation model with a fixed time increment was used and set to a 40-year (life-) time horizon. Disease progression was modeled using evidence from the UK Prospective Diabetes Study (UKPDS 68). The treatment sequence matched that of published guidelines, and efficacy and safety data were derived from published sources. The model assumes that quality-adjusted life-years (QALYs) are affected by complications and hypoglycaemic events over a lifetime. As such costs and utility decrements for macroand micro-vascular complications, and adverse events such as severe hypoglycaemia are included. Costs were specific to the German setting where SUs are generic. Costs and effects were discounted annually at 3%. The perspective of the national sick funds was taken, and recommendations from the Institute for Quality and Efficiency in Health Care (IQWiG) were considered. RESULTS: Treatment with saxagliptin + MET was A292 associated with lower incidence of both symptomatic and severe hypoglycaemic events, resulting in an incremental benefit of 0.12 QALYs and an incremental cost-effectiveness ratio (ICER) of c13,931 per QALY gained. Modest reductions in all macro-vascular and micro-vascular complications were seen in those receiving saxagliptin + MET compared with SU + MET. Sensitivity analysis showed that treatment-related weight changes, as a risk factor for complications, represent the most influential driver of costeffectiveness. CONCLUSIONS: Saxagliptin is associated with improved outcomes, a lower incidence of hypoglycaemic events, and weight neutrality, when compared with generic SU, at a cost that would likely be considered acceptable in the German setting. PDB45 COST-EFFECTIVENESS ANALYSIS OF SAXAGLIPTIN IN THE TREATMENT OF DIABETES MELLITUS TYPE 2 IN SPAIN Ramirez de Arellano A1, Brosa M2, Franch J3, Mauricio D4, Alvarez C5, Sanchez-Zamorano M1, De Rivas B5 1 Bristol-Myers Squibb Iberia, Madrid, Spain; 2Oblikue Consulting, Barcelona, Spain; 3Red GEDAPS, Barcelona, Spain; 4Hospital Universitario Arnau de Vilanova, Lerida, Spain, 5 AstraZeneca, Madrid, Spain OBJECTIVES: Saxagliptin is indicated as an add-on combination therapy for adult patients with diabetes mellitus type 2 (T2D) to improve glycaemic control in combination with metformin, a sulphonylurea (SU) or a thiazolidinedione (TZD). The objective of this study was to evaluate the cost-effectiveness in Spain of saxagliptin when added to metformin in comparison with SU plus metformin or TZD plus metformin. METHODS: The analysis uses the Cardiff Long Term Model which simulates treatment pathways of pharmacological treatment of T2D from initiation of treatment until death. Efficacy and safety data are based on an indirect comparison of saxagliptin and SU and TZD based on the similarity in safety and efficacy between saxagliptin and sitagliptin demonstrated in a 18-week non-inferiority study (CV181,056). Available direct comparative trials between sitagliptin and TZDs and SUs are used to serve as data sources to assign efficacy and safety parameters. Spanish costs are used as for macro and microvascular complications as well as adverse events such as severe hypoglycemia. Utility decrements for ischemic heart disease, myocardial infarction, congestive heart failure, stroke, blindness, end-stage renal disease, transplant, amputation and body mass index are also accounted in the model. Health outcomes are measured in terms of QALYs, assuming that the lifetime QALY is affected by complications, occurrence of hypoglycaemic episodes and weight changes. The perspective used is that of the Spanish Health System. RESULTS: Saxagliptin as add on to metformin is cost-effective compared with SU and TZD (ICER < c10,000). Extensive univariate sensitivity analysis shows that the most influential factor is the weight variation, which increases with treatment with SU and TZD whereas DPP-4 inhibitors have shown to be weight neutral. CONCLUSIONS: The cost-effectiveness analysis shows that saxagliptin is cost-effective compared with both SU and TZD in combination with metformin for the treatment of T2DM in Spain. PDB46 COST-EFFECTIVENESS OF NEW ANTIDIABETICS IN TYPE 2 DIABETES: A REVIEW Baytar S, Malhan S Baskent University, Ankara, Turkey OBJECTIVES: Although having substantial costs, the new antidiabetics for Type 2 diabetes treatment present more alternatives for glycemic control of the disease. To evaluate their cost-effectiveness, the New Antidiabetics for Type 2 diabetes indexed on PubMed, EMBASE databases and American Diabetes Association abstracts were evaluated. METHODS: The cost-effectiveness of Type 2 diabetes based on the new antidiabetics was analyzed through literature review. Searches were carried out in PubMed, EMBASE and ADA abstracts to identify the articles published from 2008 to 2010, keying in the terms “cost-effectiveness” and “type 2 diabetes” with language filtering, “English.” The language filter for “Turkish” was also used but no result was achieved. Upon this filtration, the abstracts were reviewed to determine whether they included antidiabetics. RESULTS: Ten full texts, seven abstracts and three peer reviews were identified. In all studies, the cost-effectiveness of antidiabetics was assessed using the validated CORE Diabetes Model, except for two based on the Discrete Event Simulation Model. The outcomes from IMPROVE, PROactive, PRESENT, UKPDS and PREDICTIVE trials were used. In these studies, biphasic insulin aspart, exenatide, pioglitazone, insulin detemir, insulin glargine and sitagliptin were all studied under different settings and against various comparators. Biphasic insulin aspart versus human premix insulin was mostly found to be cost-effective in certain studies. Exenatide versus insulin glargine was established to be likely costeffective in two studies. Pioglitazone was found to be dominant compared to rosiglitazone. Insulin detemir was established to be cost-saving in comparison with OAD or NPH insulin, or insulin glargine. Sitagliptin was regarded as either cost-effective or cost-saving compared to rosiglitazone. CONCLUSIONS: Different results in terms of cost-effectiveness of various antidiabetics under certain settings and against varied comparators were achieved. In future, demonstrated gains in QALYs will be found to be essential for these antidiabetics to be regarded as cost-effective. PDB47 THE COST-EFFECTIVENESS OF CONTINUOUS GLUCOSE MONITORING IN TYPE 1 DIABETES PATIENTS IN THE NETHERLANDS vanGenugten ML Medtronic Trading NL BV, Heerlen, The Netherlands OBJECTIVES: Continuous glucose monitoring (CGM) has been found to improve glycemic control in type 1 diabetes in recent trials. The objective of this study is to 13th Euro Abstracts assess the cost-effectiveness of CGM compared with self-monitoring of blood glucose (SMBG) in type 1 diabetes patients, from a societal perspective in the The Netherlands. METHODS: The analysis was performed using the CORE Diabetes model with inputs for the The Netherlands. Clinical effectiveness of CGM in terms of lowering HbA1c was taken from the GuardControl study. Costs of complications and complicationrelated days off work were based on data from Erasmus University Rotterdam. Annual treatment costs were estimated using cost information from the Healthcare Insurance Board. Costs and effects were discounted at 4% and 1.5%, respectively. a lifetime horizon (60 years) was adopted for the analysis. a range of sensitivity analyses were conducted. RESULTS: The total costs per patient over 60 years for the CGM arm were c10,069 (73,348) higher than costs for the SMBG arm. Complication costs were lower, and treatment costs were higher for the CGM compared to the SMBG arm. Compared to the SMBG arm, 0.463 (3.063) QALYs were gained in the CGM group. The cost per QALY gained was c21,731. Probabilistic sensitivity analyses showed that this result is robust. CONCLUSIONS: At a willingness to pay threshold of c40,000 per QALY, CGM appears to be a cost-effective treatment option compared to SMBG in patients with type 1 diabetes. PDB48 COST-EFFECTIVENESS OF 2 DIABETES HEALTH CARE PROGRAMMES IN BELGIUM Benoit K1, Borgermans L2, Annemans L3 1 Ghent University, Ghent, Belgium; 2Catholic University Leuven, Leuven, Belgium; 3Ghent University—Brussels University, Ghent, Belgium OBJECTIVES: A multifaceted quality improvement programme for the care of diabetes was implemented in two different Flemish regions (Aalst and Leuven) and with slightly different modalities. The overall objective of the programmes was to improve adherence to evidence-based guidelines on diabetes in primary care physicians. The aim of this study was to assess the cost-effectiveness of the programs compared to regular care in both regions. METHODS: Short-term effects of the programmes were extrapolated to long-term hard endpoints, using the validated UKPDS Outcomes Model, supplemented by two self-developed Markov models to account for benefits on process parameters (screening for retinopathy and nephropathy). a simulation of the evolution of patients was made over a time horizon of 25 years with cycles of 1 year, from a public health care payer perspective. Cost data of the different disease states and extra direct medical costs due to intensified patient management were collected from literature and from the National Institute for Health and Disability Insurance. For the self developed Markov models, utility data for all states were obtained from published studies and transition probabilities were obtained from local epidemiological studies and published trials. In the simulation, the intervention was assumed to be implemented three times over the predicted life expectancy. Deterministic sensitivity analyses were performed on the combined results of outcome and process parameters. RESULTS: At a cost of intervention of c185 (Aalst) and c284 (Leuven) per patient, the analyses show ICER’S for Aalst and Leuven of respectively c15,206.70/QALY and c10,397.96/QALY. Sensitivity analyses show few influence of changed input variables. CONCLUSIONS: When using a ratio of c30,000/QALY as threshold of willingness to pay for health gain, the diabetes health care programmes have an acceptable ICER in both regions. PDB49 PHARMACOECONOMICS MODELING OF LONG TERM RESULTS OF TYPE 2 DIABETES MELLITUS TREATMENT IN PATIENTS USING MODERN INSULIN ANALOGUES IN CONTRAST TO ORAL ANTIDIABETIC DRUGS OR DIET Yagudina R, Kulikov AU, Arinina EE Moscow Medical Academy, Moscow, Russia OBJECTIVES: To analyze the effectiveness and safety of three treatment regimens, and to project and compare long-term outcomes and costs of complications of type 2 diabetes mellitus (T2DM). METHODS: Data of 3678 patients (mean age 49.6 ± 0.19 years; 47% men; mean diabetes’s duration—6 years) included in “evaluation of diabetes mellitus complications” program were taken for mathematic simulation using the validated Center for Outcomes Research (CORE) Diabetes Model. Patients were randomly divided into three groups: insulin therapy (NovoMix® 30), oral antidiabetic drugs (OAD; combination of glibenclamide + metformin in most cases) and diet. Life expectancy, cumulative incidence of cardiovascular, ocular events and health care costs were estimated over period duration—50 years. RESULTS: Estimated life expectancy was higher in insulin group compared with OAD and diet groups (17.2, 16.5, and 16.0 years). The best QALY’s results were also achieved in insulin group (10.7, 10.2 and 9.9 years) due to decreased rates of myocardial infarction (morbidity/mortality), decompensated heart failure, ocular complications and diabetic retinopathy. Higher direct costs for years of life gained in insulin group (1287, 1203, and 1180 thousands of rub.) were associated with concomitant decreasing of indirect costs (362.7, 382.5, and 381.3 thousands of rub.). The costs for one patient with T2DM per 1 year were nearly the same in three groups (95.9, 95.9, and 97.6 thousands of rub.). CONCLUSIONS: use of the CORE Diabetes Model life expectancy, diabetes complications, and costs favored insulin NovoMix® 30 therapies compared with OAD or diet in treatment of T2DM. 13th Euro Abstracts PDB50 ASSESSING THE INFLUENCE OF INCORPORATING SECONDARY CARDIOVASCULAR EVENTS INTO A TYPE 2 DIABETES MELLITUS (T2DM) COST-EFFECTIVENESS MODEL McEwan P1, Evans LM2, Bergenheim K3 1 CRC, Cardiff, UK; 2University Hospital of Wales, Cardiff, Vale of Glamorga, UK; 3 AstraZeneca, Mölndal, Sweden OBJECTIVES: Cost-effectiveness assessments in T2DM are commonly based on models that predict only primary cardiovascular events. This study aimed to assess the implications of incorporating secondary cardiovascular events on predicted cost-effectiveness. METHODS: Routine UK hospital data, between 2000 and 2005, were analyzed to quantify the cumulative incidence of first, second and third myocardial infarction (MI) or stroke events in T2DM subjects. Adjustments were made for out of hospital mortality and under-diagnosis of T2DM. Cardiovascular risk equations, used in a previously published cost-utility model, were re-calibrated, using the ratio of primary plus subsequent event to primary event, to predict subsequent MIs and strokes consistent with the observed UK data. The cost-effectiveness analysis compared two treatment strategies: A: 1st line metformin; 2nd line DPP-4 inhibitor add-on; 3rd line sulphonylurea add-on. B: 1st line metformin; 2nd line sulphonylurea add-on; 3rd line thiazolidinedione add-on. RESULTS: Of the 1,124,846 T2DM patients identified, 55,868 and 65,436 experienced primary MI and stroke events, respectively. There were 2159 (3.86%) and 185 (0.003%) second and third MI admissions, and 5808 (8.88%) and 755 (0.012%) second and third stroke admissions, respectively. Modelled risk multipliers of 1.04 for MI and 1.1 for stroke were required to predict cumulative incidence consistent with the UK data. Incorporating subsequent events had little impact on the cost-utility analysis with the ICER decreasing from £3129 to £3105 per quality adjusted life-year. More noteworthy, was the impact on cost per life-year gained, which decreased from £257,902 to £90,055, with subsequent events included. CONCLUSIONS: The inclusion of subsequent cardiovascular events in models of T2DM provides greater face validity but has little impact upon cost-effectiveness. Thus, economic assessments of therapies that modify glycaemic control, using models that do not incorporate subsequent MI and stroke events, are not significantly biased. PDB51 COST-EFFECTIVENESS OF EXENATIDE VERSUS INSULIN GLARGINE AND VERSUS BIPHASIC INSULIN ASPART FOR THE TREATMENT OF TYPE 2 DIABETES IN PORTUGAL: A LONG-TERM HEALTH ECONOMIC ANALYSIS Palmer JL1, Pinto CG2, Duarte R3, Miguel L4, Gregor Z5 1 IMS Health, Allschwil, Basel-Land, Switzerland; 2Instituto Superior de Economia e Gestao, Lisboa, Portugal; 3Portuguese Diabetic Association, Lisbon, Portuga; 4CISEP—ISEG/UTL, Lisbon, Portugal; 5Eli Lilly & Company, Prague, Czech Republic OBJECTIVES: Two recent multicenter, comparator-controlled, open-label, randomized, parallel group clinical trials comparing exenatide with insulin glargine and with biphasic insulin aspart provided evidence of the short-term clinical profile of exenatide. The objective of this cost-effectiveness analysis was to use these results as the basis for longterm projections to estimate the clinical and economic outcomes associated with exenatide treatment versus insulin glargine and versus biphasic insulin aspart in Portuguese health care setting. METHODS: The previously published and validated IMS Core Diabetes Model was used to project the long-term clinical and cost outcomes for a cohort defined as the intention-to-treat (ITT) population of patients in the H8O-MC-GWAA and H8O-MC-GWAD clinical trials having a baseline BMI ≥ 35 kg/m2. Portuguesespecific direct medical costs data were used in the analysis to model outcomes over a 35-year time horizon from the National Health Service perspective. RESULTS: Exenatide was associated with ICERs of c61,637 per life-year gained and c17,222 per QALY gained versus biphasic insulin aspart. Exenatide was also associated with ICERs of c53,275 per life-year gained and c14,697 per QALY gained versus insulin glargine from the National Health Service perspective. Results from 18 sensitivity analyses and two BMI subgroup analyses indicated a limited impact of baseline BMI on the final results. Results were sensitive to disutilities applied for excess BMI and nausea. Results were also sensitive to assumed insulin daily doses (IU) for insulin glargine and biphasic insulin aspart after the first year. CONCLUSIONS: The outcomes of this CEA and CUA were that exenatide has been projected to improve life expectancy and quality-adjusted life expectancy compared to both insulin glargine and to biphasic insulin aspart in patients with type 2 diabetes failing OADs. Based upon these results exenatide could be considered good value for money in Portugal regardless of baseline BMI levels. PDB52 LONG-TERM COST-EFFECTIVENESS OF LIRAGLUTIDE VS. ROSIGLITAZONE IN THE CZECH REPUBLIC Dolezal T1, Niewada M2, Rychna K3, Czech M4 1 Institute for Health Economics and Technology Assessment, Prague, Czech Republic; 2 HealthQuest sp z o.o., Warsaw, Poland; 3Novo Nordisk, Prague 6, Czech Republic; 4Novo Nordisk Pharma Sp z.o.o., Warsaw, Poland OBJECTIVES: To assess the long-term cost-effectiveness of treatment with liraglutide on top of standard therapy with glimepiride (SU) compared with rosiglitazone in people with type 2 diabetes. METHODS: The extensively published and validated CORE Diabetes Model was populated with the clinical data from LEAD-1 trial: liraglutide + SU vs. rosiglitazone + SU. The analysis was performed from the Czech health care services payer’s perspective. a 20-year time horizon was chosen to reflect the costs and outcomes of diabetes as these are often only seen in the later stages of the disease. The analysis used health state utility values from published sources to assess the effect of treatment on QALYs. The unit costs of treatment and complications were derived from published sources or A293 based on expert opinion survey and official tariff lists for health care services paid by public payer (insurance company). All figures are shown in CZK and EUR (100 CZK = 3.94 EUR). RESULTS: QALYs increased with liraglutide 1,2 mg + SU vs. SU + rosiglitazone 4 mg by 0.236. Total direct costs increased by CZK 45,679 (c1800) resulting in incremental costs per QALY of CZK 193,468 (c7623). The incremental cost-effectiveness ratio for liraglutide 1.8 mg + SU vs. SU + rosiglitazone 4 mg was estimated at CZK 378,762 (c14,923) per QALY gained (QALYs increased by 0.270). Total costs (including indirect costs) increased by CZK 44,028 (c1735) and CZK 100,301 (c3952) resulting in an incremental cost per QALY gained of CZK 186,475 (c7347) and CZK 371,188 (c14,624), respectively. CONCLUSIONS: Treatment with liraglutide added to a sulphonylurea is a cost-effective intervention compared with adding rosiglitazone and is likely to represent good value for money in the Czech Republic setting. PDB53 COST-EFFECTIVENESS OF PREGABALIN VERSUS USUAL CARE IN REFRACTORY OUT-PATIENTS WITH PAINFUL DIABETIC PERIPHERAL NEUROPATHY (PDPN) FOLLOWED IN PRIMARY CARE SETTINGS de Salas-Cansado M1, Pérez C2, Navarro A3, Saldaña MT4, Rejas J5 1 TFS, Madrid, Spain; 2Pain Clinic, Hospital de la Princesa, Madrid, Spain; 34Primary Care Health Centre Puerta del Ángel, Madrid, Spain; 4Primary Care Health Centre Raíces, Castrillón, Asturias, Spain; 5Pfizer España, Alcobendas/Madrid, Spain OBJECTIVES: Estimate the cost-effectiveness (CE) of Pregabalin (PGB) and Usual Care (UC) in refractory outpatients with pDPN treated in usual medical practice in Primary Care settings in Spain. METHODS: Data extracted from a 12-week noninterventional study were used in the CE analysis. Previously, PGB naïve patients treated with UC or PGB, matched by age (+5 years), sex and pain intensity (+5 pts), refractory (≥40 VAS-MPQ) to previous treatment were selected. Patients could switch to PGB (monotherapy/add-on) or to UC other than PGB. Time horizon was 12 weeks. Effectiveness was expressed as quality-adjusted life-years (QALY) gain. The CEA included the perspectives of the NHS and society (2006), with results expressed as incremental cost-effectiveness ratio (ICER). Bootstrapping techniques (10,000 resamples) were used to obtain the probabilistic ICER, its 95% percentile confidence interval (CI) and the CE acceptability curve. Univariate probabilistic sensitivity analysis was also performed. RESULTS: A total of 189 patients, 112 in PGB group and 77 in UC were identified. Compared with UC, PGB was associated with higher QALY gain; 0.0406 ± 0.0343 versus 0.0285 ± 0.0350 (P = 0.598). Although drug costs were higher for PGB (c262 ± 132 vs. c66 ± 66, P < 0.001), overall total costs (c1368 ± 1229 vs. c1258 ± 1474; P = 0.587), or health care costs (c628 ± 590 vs. c469 ± 420; P = 0.134) were similar, although due its observational design and small sample size, ICERs varied extensively from c5302 (95% CI: dominant; c144,105) for total costs to c14,381 (dominant; c115,648) for health care costs and c39,592 (dominant; c131,754) for drug costs. However, probabilistic analyses showed 79% to 84% of ICERs were below the threshold of c30,000/QALY. CONCLUSIONS: This study suggests that using PGB to treat refractory out-patients with pDPN in community medical practice in Spain is cost-effective compared to UC in majority of patients. It also highlights the burden of the disease and supports the availability of effective treatments available for patients not achieving pain relief from older therapies. PDB54 COMPARISON OF TREATMENT COSTS BETWEEN BASAL-SUPPORTED ORAL THERAPY (BOT) WITH INSULIN GLARGINE (GLA) AND BOT WITH INSULIN DETEMIR (DET) IN PATIENTS WITH TYPE 2 DIABETES (T2D): ECONOMIC EVALUATION BASED ON THE RESULTS OF THE INSULIN GLARGINE (LANTUS®) VERSUS INSULIN DETEMIR (LEVEMIR®) TREAT-TO-TARGET (L2T3) STUDY Schädlich PK1, Koltermann KC1, Dippel FW2, Hagenmeyer EG1, Häussler B1 1 IGES Institut GmbH, Berlin, Germany; 2Sanofi-Aventis Deutschland GmbH, Berlin, Germany OBJECTIVES: To compare, from the perspective of Statutory Health Insurance (SHI) in Germany, direct diabetes-related treatment costs (DTC) in T2D patients during the first year after initiation of a BOT with either of the long-acting insulin analogues GLA or DET, based on the results of a randomized controlled trial (RCT), the L2T3 study [1]. METHODS: According to the study protocol of the 24-week RCT, GLA was administered once daily, DET [2] twice daily. The respective insulin consumption was extrapolated to 52 weeks via logarithmic regression. Due to proof of non-inferiority in the L2T3 study, a cost-minimization analysis was conducted. DTC from the SHI perspective comprised insulin consumption, test strips, needles and lancets. In the base-case analysis, average values of all model parameters were applied. Taking a conservative approach, it was assumed that needles were changed daily (disadvantage for GLA). a new test strip and lancet were assumed for each blood glucose measurement. In comprehensive sensitivity analyses (impact analysis, analysis of extremes, Monte Carlo simulation), the robustness of the base-case results was tested. RESULTS: The base-case analysis revealed savings of c767 in annual DTC per patient when using BOT with GLA (c1141) compared to DET (c1908). Of these savings, c517 (67%) fell upon insulin, c214 (28%) upon test strips and c36 (5%) upon lancets. Savings in favour of GLA turned out to be robust in the sensitivity analyses. Price and insulin consumption of DET had the highest impact on these savings. CONCLUSIONS: Initiation of a BOT with GLA in T2D patients after failure of oral antidiabetic therapy alone may lead to substantial savings for SHI compared to BOT with DET. ACKNOWLEDGMENT: This study was supported by Sanofi-Aventis Deutschland GmbH, Berlin, Germany. [1] Swinnen et al. Diabetes Care 2010: doi:10.237/doc09– 2294 [2] In the treatment of T2D, in combination with oral medications, it is recommended to use DET once daily. A294 PDB55 COMPARISON OF RESOURCE USE AND COSTS IN TYPE 1 DIABETES PATIENTS TREATED WITH DIFFERENT LONG ACTING INSULINS IN A BASAL-BOLUS REGIMEN IN GERMANY Bierwirth RA1, Kostev K2, Dippel FW3, Fuchs S4, Kotowa W4 1 Ambulantes Diabteszentrum am Elisabeth-Krankenhaus, Essen, Germany; 2IMS Health GmbH & Co. OHG, Frankfurt am Main, Germany; 3Sanofi-Aventis Deutschland GmbH, Berlin, Germany; 4IMS Health GmbH & Co. OHG, Nürnberg, Germany OBJECTIVES: Compare resource utilization and treatment costs with three different basal insulins in type-1-diabetics (T1D). METHODS: A cohort study based on a representative database (IMS® Disease Analyzer) included T1D who had started an intensified conventional therapy (ICT) with NPH-insulin (NPH), insulin glargine (GLA) or insulin detemir (DET) between July 2000 and February 2008 and whose data were continuously documented at least 12 months before and 18 months after ICT initiation. The variables age, gender, diabetes duration, HbA1c, Body Mass Index (BMI), insurance status, geographical region and specification of the practice were collected. Diabetesrelated resource utilization (insulin, test strips, lancets, pens, needles, glucose i.v., glucagon, physician visits and hospitalization) and associated direct treatment costs (excluding physicians visits and hospitalization) were determined for a time period of 12 months for patients receiving NPH, GLA and DET, respectively. The results were adjusted applying a multivariate regression model. RESULTS: A total of 1218 T1D received an ICT with NPH, 1079 with GLA and 443 with DET, respectively. The unadjusted annual direct treatment costs were c1308 for NPH, c1512 for GLA and c1729 for DET. After adjusting ICT with GLA showed economic advantages compared to NPH (−c234/year; P < 0.0001) or DET (−c425/year; P = 0.2800). The consumption of basal insulin and test strips was lower in patients treated with GLA compared to NPH (−6.00 U/day; P = 0.3514 and −0.31 strips/day; P = 0.8291) or DET (−3.23 U/ day; P < 0.0001 and −0.59 strips/day; P = 0.0235). CONCLUSIONS: After adjustment this analysis of German real-life data showed that ICT with GLA is related to lower annual treatment costs than ICT with NPH or DET. In view of the equal clinical efficacy as reported in several randomized clinical trials [1, 2] and the economic advantages, GLA should be regarded as the favored therapeutic option in ICT for T1D in Germany. PDB56 PHARMACOECONOMIC CONSEQUENCES OF LOSARTAN THERAPY IN PATIENTS UNDERGOING DIABETIC END-STAGE RENAL DISEASE IN EU AND USA Citarella A, de Portu S, Cammarota S, Menditto E, Mantovani LG CIRFF, Federico II University, Naples, Italy OBJECTIVES: Diabetic nephropathy is the most frequent cause of End Stage Renal Disease (ESRD). As ESRD incidence continuously increases, more resources are needed for treatment. The objective was to evaluate the economic impact of losartan added to the standard care administered to diabetic subjects with ESRD. The analysis has involved more than 500 million inhabitants. METHODS: We used standard methods to conduct an economic evaluation comparing the economic outcomes deriving from the administration of losartan added to standard care versus standard care alone in patients with type 2 diabetes mellitus (DM) and nephropathy over 3.4 years. The study was conducted from the perspective of the third-party payer hence. The clinical outcome data were based on the results from the RENAAL trial. Direct medical costs are referred to the purchase costs of losartan and the cost of hospitalizations. The costs were discounted back at an annual rate of 3%. Also sensitivity analysis was performed. RESULTS: RENAAL study established that losartan confers strong renal protection in patients with DM and nephropathy. Losartan results into a cost saving in all countries considered: 3 c602.98/Italy, c4531.35/France, c3019.66/Germany, c3949.50/Switzerland and c3855.50/USA per patient. Results are not sensitive to both clinical and economic variables. CONCLUSIONS: In addition to the medical benefit, this analysis demonstrates the economic relevance of treatment with losartan in DM patients with nephropathy. PDB57 LOWER TREATMENT COSTS WITH INSULIN GLARGINE COMPARED TO INSULIN DETEMIR IN TYPE 1 DIABETES PATIENTS TREATED WITH A BASAL-BOLUS REGIMEN IN GERMANY Wiesner T1, Schädlich PK2, Dippel FW3, Koltermann K2, Hagenmeyer EG2 1 MVZ Stoffwechselmedizin Leipzig, Leipzig, Germany; 2IGES Institut GmbH, Berlin, Germany; 3 Sanofi-Aventis Deutschland GmbH, Berlin, Germany OBJECTIVES: To compare, from the perspective of Statutory Health Insurance (SHI) in Germany, direct diabetes-related treatment costs in patients with type 1 diabetes mellitus (T1DM) during the first year after the switch from Neutral Protamin Hagedorn insulin (NPH) to the respective long acting insulin in the course of a basal-bolus insulin regimen (ICT) with insulin glargine or insulin detemir as the basal insulin component. METHODS: Natural units of resource consumption incurred by basal and bolus insulin, needles, lancets, and test strips were modelled over a period of 1 year in each of the two cost-minimization analyses, based on the results of two controlled clinical trials [1, 2]. Resources were valued in prices of 15 January 2010 relevant to SHI in the outpatient sector. In the base-case analyses, average values of all the model parameters were applied. In comprehensive sensitivity analyses (impact analysis, analysis of extremes, Monte Carlo simulation), the robustness of the basecase results was tested. RESULTS: In the base-case analyses, there were savings of c378 or 15% and c311 or 14%, respectively, per patient and year obtained by insulin glargine compared to insulin detemir. Savings in favour of GLA turned out to be 13th Euro Abstracts robust in the sensitivity analyses. Even in the analyses of extremes, there were always savings obtained by insulin glargine, irrespective of insulin detemir being given once, once to twice, or twice daily. When simulating real-life conditions the savings obtained by insulin glargine instead of insulin detemir were maintained. CONCLUSIONS: Treatment of T1DM patients with insulin glargine as the basal insulin component of an ICT may lead to substantial savings from the German SHI perspective as compared to insulin detemir. [1] Pieber et al. Diabet Med 2007;24:635–42; [2] Heller et al. Clin Ther 2009;31:2086–97. PDB58 PHARMACOEPIDEMIOLOGICAL ASSAY AND COST-MINIMIZATION ANALYSIS OF ORAL ANTIDIABETIC MEDICATIONS AND INSULINS IN LITHUANIA Kildonaviciute G, Stankunaite E, Kadusevicius E, Petraityte A Kaunas Medical University, Kaunas, Lithuania OBJECTIVES: To conduct pharmacoepidemiological research and cost-minimization analysis of oral antidiabetic medications and insulins in Lithuania. METHODS: Medications were grouped according to the ATC classification system. Our research results were reported in DDDs per 1000 inhabitants per day (DDD/TID). Calculations of drug prices and total expenditures on antidiabetic medications were made by using data from National Patient Funds Price List in 2006–2009 years. Reference pricing methodology was used to accomplish our cost-minimization analysis. RESULTS: Total consumption of hypoglycaemic agents increased by 33% from 21.54 DDD/TID in 2006 to 28.72 DDD/TID in 2009. Utilization of insulin increased by 30% reaching the value 9.43 DDD/TID in 2009 and oral antidiabetic medications increased by 35%—19.29 DDD/TID in 2009. Total expenditures on hypoglycaemic agents increased by 23% from LTL 57.138 mlln in 2006 to LTL 70.531 mlln in 2009 (1EUR = 3.4528LTL). Single DDD prices ranged from 0.70 LTL/DDD to 5.01 LTL/DDD of oral antidiabetics, and from 5.67 LTL/DDD to 2.97 LTL/DDD for insulins. With reference to meta-analysis and NICE recommendations, considering the similar efficacy and safety within drug classes, cost-minimization analysis using the reference-based pricing could be implemented and total expenditures could be decreased by 27% (saving LTL 19 mlln. yearly). CONCLUSIONS: Our findings suggest that implementation of reference-based pricing could be a strong fiscal measure helping to rationalize increasing direct health care expenditures by 27%. PDB59 HEALTH ECONOMIC EVALUATIONS COMPARING THE BASAL INSULIN ANALOGUE GLARGINE (GLA) WITH NEUTRAL PROTAMINE HAGEDORN (NPH) INSULIN IN INTENSIFIED INSULIN THERAPY (ICT) IN PATIENTS WITH TYPE 1 DIABETES: A SYSTEMATIC REVIEW Hagenmeyer EG, Koltermann KC, Schädlich PK, Häussler B IGES Institut GmbH, Berlin, Germany OBJECTIVES: To perform a systematic literature review of health economic evaluations comparing GLA with NPH as the basal component of an ICT in patients with type 1 diabetes. METHODS: The search was performed between January 1, 2000 and December 1, 2009 via Embase, Medline, the Cochrane Library, the databases of German Medical Science and of DAHTA (Deutsche Agentur für Health Technology Assessment), and abstract books of relevant scientific congresses. The inclusion of retrieved studies was based on predefined criteria. The included studies were assessed according to established methodological and quality aspects. RESULTS: A total of seven health-economic evaluations from four different countries were included: six modeling studies, all of them cost-utility analyses (CUA), and one cost-minimization analysis (CMA) based on a claims data analysis. One CUA showed dominance of GLA because of higher utilities and lower costs. The other five CUAs varied in their additional costs per quality adjusted life-year (QALY) gained for treatment with GLA between c3.859 and c57.002 (incremental cost-effectiveness ratio, ICER). The CMA revealed about c160 higher diabetes-specific costs per year for GLA in the German Statutory Health Insurance (SHI) setting. All the included studies showed good quality despite a few constraints. Nevertheless, they all contained enough explanatory power to evaluate the effectiveness of GLA in comparison to NPH. CONCLUSIONS: Despite some differences concerning evaluation methods (CUA or CMA), data sources (randomized controlled trial, claims data) and country specific conditions (pricing and reimbursement situation) the identified health economic analyses showed high conformity concerning the main target parameter. Most of the studies (5 of 7) showed a good to very good cost-effectiveness in favour of GLA compared to NPH depending on the respective design of the health economic analysis chosen. ACKNOWLEDGMENT: This study was supported by Sanofi-Aventis Deutschland GmbH, Berlin, Germany. PDB60 HEALTH ECONOMIC EVALUATIONS COMPARING THE BASAL INSULIN ANALOGUES INSULIN GLARGINE (GLA) AND INSULIN DETEMIR (DET) IN INTENSIFIED INSULIN THERAPY (ICT) IN PATIENTS WITH TYPE 1 DIABETES: A SYSTEMATIC REVIEW Hagenmeyer EG, Koltermann KC, Schädlich PK, Häussler B IGES Institut GmbH, Berlin, Germany OBJECTIVES: Due to limited health care resources economic evaluations of alternative drug treatment methods become more important, especially in chronic diseases like diabetes mellitus. Therefore, a systematic literature review of health economic evaluations comparing GLA with DET as the basal component of an ICT in patients 13th Euro Abstracts with type 1 diabetes was performed. METHODS: The search was performed between January 1, 2000 and December 1, 2009 via Embase, Medline, the Cochrane Library, the databases of German Medical Science and of DAHTA (Deutsche Agentur für Health Technology Assessment), and abstract books of relevant scientific congresses. The inclusion of retrieved studies was based on predefined criteria. The included studies were assessed according to established methodological and quality aspects. RESULTS: A total of four health-economic evaluations from four different countries were included: two modeling studies, comprising one cost-utility analysis (CUA) and one cost-minimization analysis (CMA), as well as two claims data analyses, both CMAs. Two of the CMAs show an economic advantage in favour of GLA vs. DET and the third CMA showed cost neutrality between the basal insulin analogues. The CUA showed an economic advantage for DET. CONCLUSIONS: Despite some differences concerning evaluation methods (CUA or CMA), data sources (randomized controlled trial, claims data) and country specific conditions (pricing and reimbursement situation) the identified health economic analyses showed high conformity concerning the target parameters. Two of the studies showed an advantage in favour of GLA compared to DET, the extent of which depended on the respective design of the health economic analysis chosen. a systematic review to compare the health economic outcomes of GLA and NPH-insulin was done separately. ACKNOWLEDGMENT: This study was supported by Sanofi-Aventis Deutschland GmbH, Berlin, Germany. PDB61 ECONOMIC EVALUATION OF LIRAGLUTIDE VS. ROSIGLITAZONE OR EXENATIDE FOR TYPE 2 DIABETES MELLITUS IN BULGARIA Ivanova A1, Petrova G1, Wrona W2, Valov V3, Czech M4 1 Medical University, Faculty of Pharmacy, Sofia, Bulgaria; 2HealthQuest sp z o.o., Warsaw, Poland; 3Novo Nordisk Pharma EAD, Sofia, Bulgaria; 4Novo Nordisk Pharma Sp z.o.o., Warsaw, Poland OBJECTIVES: In these cost-utility analyses CORE Diabetes Model and LEAD-1 and LEAD-6 trials data were used to compare liraglutide (a glucagon-like peptide-1 receptor agonist) to rosiglitazone, both in combination with glimepiride, and to exenatide, both with metformin and/or sulfonylurea. METHODS: The analyses were performed from the health care services payer’s perspective. In the base-case analysis a time horizon of 20 years has been chosen. The analysis compared patients treated with liraglutide 1.2 mg or rosiglitazone 4 mg (LEAD-1) or liraglutide 1.8 mg or exenatide 10 µg b.i.d. (LEAD-6). The analysis used health state utility values derived from literature. The cost of treatment and complications were based on officially published sources for medicines prices (www.mh.government.bg), for hospital charges (www. nhif.bg) and verified by expert opinion survey (1 BGN = 0.51 EUR). RESULTS: QALYs increased with liraglutide 1.2 mg by 0.252 (SD 0.129) years (LEAD-1). Total costs increased by BGN7722 (c3948) resulting in an incremental cost per QALY gained of BGN30,674 (c15,684). Based on LEAD-6 trial data liraglutide 1.8 mg resulted in increase of 0.151 (SD 0.124) QALYs. Total costs increased by BGN4151 (c2122) with incremental cost BGN27,404 (c14,012) per QALY. CONCLUSIONS: In Bulgarian health care system settings liraglutide added to standard treatment have been shown to be cost-effective in comparison with rosiglitazone and exenatide for type 2 diabetes. PDB62 COST-UTILITY ANALYSIS OF SAXAGLIPTIN AS AN ADD-ON THERAPY TO METFORMIN IN TYPE 2 DIABETES PATIENTS FROM THE BRAZILIAN PRIVATE HEALTH SYSTEM Eliaschewitz FG1, Asano E2, Nita ME2, Rached R2, Donato BM3 1 CPCLIN—Centro de Pesquisas Clínicas, São Paulo, SP, Brazil; 2Bristol-Myers Squibb S/A, São Paulo, SP, Brazil; 3Bristol-Myers Squibb, Wallingford, CT, USA OBJECTIVES: This is a cost-utility analysis of saxagliptin (treatment group) vs. thiazolidinediones (control group) as add-on therapy in type 2 diabeties (T2D) patients not achieving appropriate glycaemic control with metformin, from the Brazilian private health system (PHS) perspective. METHODS: A discrete event simulation model based on UKPDS68 study was developed in order to simulate 40 years for a cohort of 1000 patients. Safety and efficacy data were obtained from a systematic review and meta-analysis of published literature. Epidemiological and costing data were obtained from DIAPS79, an outcome study of the treatment patterns and costs of T2D patients in the Brazilian PHS. Pharmaceutical costs were based on Brazilian official factory price. Insulin plus metformin was defined as rescue therapy. An annual discount rate of 5% was applied to both costs and benefits. Deterministic and probabilistic sensitivity analyses were conducted to assess the robustness of the results. RESULTS: According to the model, the lipid profile benefits from thiazolidinediones did not translate into long-term vascular benefits when compared to saxagliptin (vascular fatal events risk reduction of −0.0034 vs. pioglitazone and −0.0053 vs. rosiglitazone). Saxagliptin was dominant when compared to both pioglitazone and rosiglitazone as the add-on therapy of choice to metformin (costs savings per patient of R$3.874 vs. rosiglitazone and R$3.996 vs. pioglitazone; incremental 0.13 QALY per patient vs. pioglitazone and 0.14 QALY per patient vs. rosiglitazone). In the deterministic sensitivity analysis, HbA1c level was the most impactful parameter in the model, but saxagliptin remained the dominant option in all cases. In the probabilisitc sensitivity analysis, saxagliptin had a greater than 90% probability of being cost-effective for a willingness-to-pay of zero. CONCLUSIONS: Saxagliptin is associated with lower costs and increased quality-adjusted life expectancy compared to thiazolidinediones as add-on therapy in T2D patients failing to achieve adequate glycaemic control on metformin monotherapy. A295 DIABETES/ENDOCRINE DISORDERS – Patient-Reported Outcomes Studies PDB63 LIRAGLUTIDE IN THE TREATMENT OF TYPE 2 DIABETES MIELLITUS— ECONOMIC EVALUATION IN ROMANIAN SETTING Ionescu D1, Niewada M2, Czech M3 1 Carol Davila University of Medicine and Pharmacy, Bucharest, Romania, 2HealthQuest sp z.o.o., Warsaw, Poland, 3Novo Nordisk Pharma Sp z.o.o., Warsaw, Poland OBJECTIVES: Our aim was to assess cost-utility ratios of liraglutide, a GLP-1 receptor agonist, in comparison with rosiglitazone, glimepiride and exenatide for type 2 diabetes in Romanian health care system settings. METHODS: The CORE Diabetes Model and clinical data from LEAD-1 (liraglutide vs. rosiglitazone, both with glimepiride, 1st case), LEAD-2 (liraglutide vs. glimepiride, both with metformin, 2nd case) and LEAD-6 trial (liraglutide vs. exenatide, both with metformin and/or sulfonylurea, 3rd case) trials were used. The health care services payer’s perspective and 20-year time horizon have been chosen. Health state utility values and cost date were based DRG Data from “Center for Research and Evaluation of Healthcare Services,” Romania, CaNaMed National Catalog of Medicines Prices (official tariff lists) and expert opinion. RESULTS: In the 1st case QALYs increased by 0.201 (SD 0.105) years with liraglutide 1.2 mg and by 0.231 (SD 0.107) years with liraglutide 1.8 mg. Total costs increased by c3266 and c5378 resulting in an incremental cost per QALY gained of c16,209 and c23,282, respectively. In the 2nd case, liraglutide 1.2 mg resulted in increase of 0.155 (SD 0.099) QALYs and liraglutide 1.8 mg 0.170 (SD 0.146) QALYs with incremental cost per QALY gained of EUR 29,909 and EUR 38,830, respectively. In the 3rd case QALYs increased by 0.125 (SD 0.102) years with liraglutide 1.8 mg, total costs increased by c1898, an incremental cost was c15,123 per QALY. CONCLUSIONS: Based on efficacy data from clinical trials and validated model liraglutide has been shown to be cost-effective when compared with rosiglitazone, glimepiride and exenatide for the treatment of type 2 diabetes. PDB64 TREATMENT PERSISTENCE AMONG PATIENTS INITIATING INSULIN THERAPY WITH INSULIN DETEMIR IN A FLEXPEN® VERSUS NPH INSULIN IN A VIAL. RETROSPECTIVE DATABASE ANALYSIS BASED ON A LARGE US MANAGED CARE ORGANIZATION Conner C1, Buysman E2, Liu F2, Aagren M3, Bouchard J4 1 Novo Nordisk, Seattle, WA, USA; 2i3 Innovus, Eden Prairie, MN, USA; 3Novo Nordisk Inc, Princeton, NJ, USA; 4Novo Nordisk, Plaistow, NH, USA OBJECTIVES: Persistence with respect to injectable therapy in type 2 diabetes is known to impose a challenge to patients. One potential explanation is aversion to injection and risk of hypoglycemic events. Insulin analogues have shown in clinical trials to reduce the risk of hypoglycemic events versus human insulins, and insulin injection using a pen device is generally perceived as less burdensome than administration via vial and syringe. METHODS: This retrospective data analysis compared persistence between two types of basal insulin and administration: Neutral protamine hagedorn, NPH, administered by vial and syringe and insulin detemir, IDet, administered by FlexPen®. Data were derived from health care claims between 2004 and 2009 from a large national US payer and included type 2 diabetes patients that initiated either IDet FlexPen® or NPH in vials without any prescription fills for any insulin in the previous 12 months. Patients were defined as being persistent to therapy as long as they filled their prescription within the 80th percentile of days between fills adjusted to reflect differences in pack sizes. RESULTS: The IDet FlexPen® cohort (n = 1082) and the NPH vial cohort (n = 794) were of similar age (54.06 vs. 53.13, P = 0.134), but IDet FlexPen® had a lower proportion of female patients than NPH vials (44% vs. 55%, P < 0.001) and fewer treatment naïve patients (no pre-index OADs) (9% vs. 45%, P < 0.001). Persistency to therapy at 12 months after initiation was 23% and 13% (P < 0.001) for IDet FlexPen® and NPH vials, respectively. Average days of persistence was 167 days for the IDet FlexPen® cohort and 123 days for NPH vials (P < 0.001). CONCLUSIONS: Persistence to insulin therapy among type 2 diabetes patients could be improved. This study suggests that insulin persistency may be improved by initiating insulin using therapies associated with lower risk of hypoglycemic events and administering it in a pen. PDB65 THE METHODOLOGICAL QUALITY AND EFFECTIVENESS OF ADHERENCE INTERVENTIONS: A REVIEW OF DIABETES TYPE II INTERVENTIONS Wilke T1, Mueller S2, Groth A1 1 Institute for Pharmacoeconomics and Drug Logistics, University of Wismar, Wismar, Germany; 2University of Wismar, Wismar, Germany OBJECTIVES: Adherence interventions (AI) are an important part of the health care provision situation on the ground. For ethical, clinical and health economic reasons, it is vital to identify methodological characteristics of successful AI. The aim of this review is to do this for AI focused on oral anti-diabetics (Diabetes type II). METHODS: A comprehensive review of Diabetes type II AI effectiveness studies was conducted [Strings: (oral hypoglycemic agents; oral anti-diabetic medications; diabetes; hyperglycemia; Biguanide; Metformin; Potassium channel inhibitors; Dipeptidyl peptidase-4 inhibitors) and (improvement; enhancement; pharmacy, pharmacist; doctors; interventions; programs; reminder; prevention; patient education)]. Only interventions aiming to improve medication adherence/persistence were included. RESULTS: A total of 6977 contributions were identified; after detailed examination by two reviewers 15 publications evaluating 19 different AI were included. 10 AI were able to improve the A296 adherence/persistence and eight were able to improve the blood glucose levels of patients (double counting in three cases); five had no effect at all. Four dimensions of the methodological quality of AI programs were identified: 1) measurement of adherence/persistence/clinical outcomes, 2) measurement of NA/NP causes, 3) use of effective/validated intervention measures; and 4) effective program evaluation. The authors defined 5 detailed methodological requirements per dimension and, based on this, developed a corresponding scoring model (MIN Score 0, MAX score 20). All 19 AI programs were evaluated in the scoring model (average score 8.05): • Score <5: 3 AI—no adherence/blood glucose level improvement; • Score 5–9: 8 AI—6 with improvement in both adherence and/or blood glucose levels; • Score >9: 8 AI—all improved adherence and/or blood glucose levels. CONCLUSIONS: The scoring model provides a starting point for the methodical evaluation of AI. However, further development and testing of both the elements and construction is needed for medical indications other than diabetes type II. PDB66 THE 8-ITEM MORISKY MEDICATION ADHERENCE SCALE MMAS: TRANSLATION AND VALIDATION STUDY OF THE MALAYSIAN VERSION Al-Qazaz HK1, Hassali MA1, Shafie AA2, Sulaiman SAS1, Sundram S3, Morisky DE4 1 Universiti Sains Malaysia, Minden, Penang, Malaysia; 2Universiti Sains Malaysia, Penang, Penang, Malaysia; 3Hospital Balik Pulau, Balik Pulau, P.Penang, Malaysia; 4UCLA School of Public Health, Los Angeles, CA, USA OBJECTIVES: To translate the MMAS into the Malaysian language, and to examine the psychometric properties of the Malaysian version of the MMAS among patients with type 2 diabetes, including its validity and reliability. METHODS: After obtaining permission, a standard “forward-backward” translation procedure was used to create the Malaysian version of the MMAS from the original English version. a convenience sample of 223 outpatients with type 2 diabetes was identified between May and September, 2009. All data were collected from the Penang General Hospital, Penang, Malaysia. Instruments consisted of the Malaysian version of MMAS, the Malaysian version of the old four-item Morisky scale and a sociodemographic questionnaire. Medical records were reviewed for hemoglobin A1C (HbA1C) levels and other clinical data. Reliability was tested for internal consistency using Cronbach’s α coefficient. Validity was confirmed using convergent and known group validity. RESULTS: Employing the recommended scoring method, the mean ± SD of MMAS scores was 6.13 ± 1.72. Moderate internal consistency was found, (Cronbach’s α = 0.675), the test-retest reliability value by using Spearman’s rank correlation was 0.816 (P < 0.001). a positive correlation between the eight- and four-item MMAS was found (r = 0.792; P < 0.01). For known group validity, a significant relationship between MMAS categories and HbA1c categories (χ2 = 20.261; P ≥ 0.001) was found. The MMAS sensitivity and specificity, with positive and negative predictive values were 77.61%, 45.37%, 46.84% and 76.56%, respectively. CONCLUSIONS: The MMAS can be used for medication adherence measurement in diabetes. The findings of this validation study indicate that the Malaysian version of the MMAS is a reliable and valid measure of medication adherence which can now be used in clinical practice. PDB67 FACTORS INFLUENCING VALUATION OF- AND WILLINGNESS TO PARTICIPATE IN- A LIFESTYLE INTERVENTION: AN EXPLORATORY CONJOINT ANALYSIS WITH DIABETES TYPE 2 PATIENTS Van Gils PF1, Lambooij MS1, Struijs JN1, Flanderijn MH2, van den Berg M3, van den Berg B4 1 National Institute of Public Health and the Environment, Bilthoven, The Netherlands; 2Leids UniversityMedical Centre, Leiden, The Netherlands; 3National Institute for Public Health and the Environment (RIVM), Bilthoven, The Netherlands; 4University of York, York, UK OBJECTIVES: The last decade several studies have proven that lifestyle interventions can be effective for people with or at risk for diabetes. Because effectiveness of preventive interventions is affected by non-participation (adherence), it is important to understand factors influencing people’s willingness to participate in a lifestyle intervention. Therefore, the aim of this exploratory analysis is to examine which factors of a lifestyle program influence its valuation and willingness to participate. METHODS: We used conjoint analysis to empirically examine associations between the factors that influence participation and participants’ valuation of an intervention and participants’ willingness to participate in a lifestyle intervention. For this purpose participants received a questionnaire with four hypothetical lifestyle interventions. They were asked to value the hypothetical scenarios with a grade from “1” to “10” and furthermore they were asked if they would be willing to participate in these hypothetical programs. Linear and logistic regression techniques were used for the analyses. RESULTS: The factors “group activity,” “counselling,” and “receiving money” were positively associated with the scores of the valuation of the programmes. Logistic regression analysis showed that money was the only factor that was independently associated with respondents’ willingness to participate in a lifestyle intervention. Subgroup analysis showed that receiving an amount of money was not associated with willingness to participate, but having to pay is negatively associated with participation in the lifestyle intervention. CONCLUSIONS: It appeared that only financial disincentives were independently associated with willingness to participate in a lifestyle intervention. Our conjoint analysis results suggest that financial incentives, in the form of bonuses, cannot be used to encourage people to participate in lifestyle interventions. Financial incentives, in the form of payments might however discourage participation, regardless of the content of the program. 13th Euro Abstracts PDB68 HEALTH RELATED QUALITY OF LIFE (HRQL) AND EQ-5D UTILITIES IN A TYPE 2 DIABETES (T2D) POPULATION: RESULTS FROM A SWEDISH SURVEY Sörngård H1, Lagging E2, Lindh A3, Jörgensen L1 1 AstraZeneca Nordic MC, Södertälje, Sweden; 2Stockholm Diabetic Patient Association, Stockholm, Sweden, 3Husläkarna i Österåker, Åkersberga, Sweden OBJECTIVES: To present HRQL data from a previously presented survey on T2D patients’ perceived quality of care in Stockholm, Sweden. METHODS: A postal survey including EQ-5D was distributed to 1000 members of the Stockholm Diabetic Patient Association and 1000 patients from five primary health care centers. Patients were randomly selected, >18 years, having a diabetes diagnosis. Analysis of variance was used to test the statistical hypothesis that patients within each subgroup had equal mean utility. RESULTS: Response rate was 65% (1319/2000 questionnaires). T2D diagnose was reported for 961 respondents of which 858 completed the EQ-5D questionnaire. Mean age 69 years, 48.5 % female, BMI 28.4 kg/m2, mean duration of T2D 11.3 years. Overall, the mean (SD) utility was 0.765 (0.260) and the current health status reported on the VAS scale was 0.727 (0.189). Patients without any hypoglycemic episode the previous month had a utility of 0.799 while those with 1, 2–5 or >5 episodes reported 0.774, 0.687 and 0.633, respectively. More than one hypoglycemic episode resulted in significantly lower utility compared to none or only one episode (P < 0.0001). The utility of obese patients, 0.704, was significantly lower than for patients of normal weight, 0.806, or overweight, 0.790 (P < 0.0001). The utility of patients not considering themselves responsible for the management of their T2D was significantly lower, 0.608, compared to patients taking a limited, 0.774, or full, 0.759, responsibility (P = 0.0005). Patients considering themselves having insufficient knowledge to cope with their T2D reported a lower utility, 0.689, compared to patients with sufficient knowledge, 0.789 (P < 0.0001). Male respondents had a higher utility, 0.796, compared to females, 0.731 (P = 0.0002). CONCLUSIONS: Experience of hypoglycemic episodes, obesity, gender, patients’ perception of personal responsibility and perceived knowledge about type 2 diabetes has significant impact on health related quality of life in patients with type 2 diabetes. PDB69 UTILITY VALUES FOR DIABETES COMPLICATIONS Donatti C, Lloyd A, Henry N, Grant D IMS Health, London, UK OBJECTIVES: Cost-utility analysis in diabetes requires utility estimates for diabetic complications. Models frequently rely on UKPDS data. UK requirements for preference elicitation have changed since UKPDS publication. We conducted a systematic review of the literature to identify utility values for diabetes complications suitable for use in Health Technology Assessment (HTA). METHODS: A systematic search of online databases was conducted using key words relating to diabetes, major complications, utility assessment and quality of life. Reference lists of identified citations were reviewed. Studies reporting utility single-index measures in patients with any of 33 pre-specified diabetes related states were included: states considered were diabetic complications and adverse events associated with anti-diabetic therapies (AEs). Papers were qualitatively assessed: criteria included relevance of studied population to Type 1 or Type 2 diabetes, sample size, methodological quality and consistency with current UK HTA guidelines. Comorbidity is common in diabetes: methodology papers addressing combination of utility values were identified in a structured search and reviewed. RESULTS: The search returned 3024 hits, 169 articles were reviewed and 32 publications were identified as suitable for review. Utility or disutility values suitable for UK HTA were obtained for 23 diabetes states. For 10 complications, including late stage renal disease and some AEs, no utility value were identified that met UK HTA criteria. There is no consensus in the literature on how utility measures should be combined in patients with more than one complication. CONCLUSIONS: We identified a set of utility values suitable for economic analysis for HTA in diabetes. To further inform UK HTA, additional research should create robust utility values for diabetic renal disease, and evaluate the empirical accuracy of alternative methods of combining utility values in patients with multiple complications. PDB70 MEASUREMENT OF HRQOL USING EQ-5D IN TYPE 2 DIABETES MELLITUS PATIENTS TREATED WITH ORAL ANTI-DIABETIC DRUGS IN CHINA Li HC1, Chang JH2, Liu GG3 1 China Pharmaceutical University, Nanjing, China; 2Novo Nordisk (China) Pharmaceuticals Co., Ltd., Beijing, China, 3Peking University, Beijing, China OBJECTIVES: The study is to measure the health-related quality of life (HRQOL) in type 2 diabetes mellitus (T2DM) patients with oral anti-diabetic drugs (OADs) therapy using the Chinese version of EQ-5D, and examine their health status. METHODS: The study was a cross-sectional survey conducted at 75 hospitals in nine cities in China. There were 9577 T2DM patients administered with OADs therapy completed the questionnaires. The survey period was from December 3rd, 2008 to July 31st, 2009. Patients evaluated their health status using five dimensions (5D) and a visual analog scale (VAS). Descriptive statistics was used to describe patients’ demographic characteristics, duration of the disease, the frequency of 5D responses and VAS score. STATA 9.2 was used for the analyses. RESULTS: The mean age of patients (±SD) was 59.5 ± 12.7 years. 51.1% were male. The mean body mass index (±SD) was 24.3 ± 3.4 kg/ m2. The mean duration of disease (±SD) was 7.9 ± 6.3 years. For the five dimensions 13th Euro Abstracts of EQ-5D, the frequency of T2DM patients responding as having “some problem” and “extreme problem” was 4.6% and 1.6% for mobility, 7.9% and 1.4% for selfcare, 13.9% and 1.7% for usual activities, 27.5% and 1.1% for pain/discomfort, and 26.6% and 1.3% for anxiety/depression, respectively. The mean VAS score (±SD) of patients was 70.4 ± 15.1. CONCLUSIONS: The rates of T2DM patients with “some problem” in pain/discomfort, and anxiety/depression were relatively high; rate with “extreme problem” for usual activities was also higher than other dimensions. The results of 5D were consistent with the low VAS score. These findings imply that there was a significant impact on the health status of T2DM patients with OADs therapy due to pain/discomfort, and anxiety/depression. It underscores the urgent need to adopt effective measures toward prevention and control of diabetes to improve patients’ quality of life. PDB71 QUALITY-ADJUSTED LIFE-YEAR (QALY) WEIGHTS ASSOCIATED WITH DIFFERENT SEVERITY LEVELS OF DIABETIC RETINOPATHY Heintz E1, Bourghardt Peebo B2, Wiréhn AB3, Rosenqvist U4, Levin LÅ1 1 Linköping University, Linköping, Sweden; 2Ryhov County Hospital, Jönköping, Sweden; 3 Östergötland County Council, Linköping, Sweden; 4Motala Hospital, Motala, Sweden OBJECTIVES: The objective of this study was to elicit quality-adjusted life-year (QALY) weights for the different severity levels of diabetic retinopathy (DR) and to evaluate the adequacy of using certain health-related quality of life (HRQoL) instruments for this purpose. METHODS: The study population comprises 151 patients with diabetes (type 1 and 2) that either attended the eye clinic at Linköping University Hospital or were registered at any of the two vision centrals in Östergötland County, Sweden. Participants were interviewed over the phone using time trade-off (TTO) questions, the EuroQol Health Questionnaire (EQ-5D), the Health Utilities Index Mark III (HUI-3) and the National Eye Institute Visual Function Questionnaire (NEIVFQ-25). The effect of other variables than DR on QALY weights was investigated using ANCOVA and the generic instruments were tested for correlation with NEIVFQ-25. RESULTS: The ranges of the QALY weights estimated with the three generic instruments were for no DR, BR, PDR, maculopathy and legal blindness 0.81–0.88, 0.72–0.78, 0.75–0.82, 0.74–0.81 and 0.39–0.68 respectively. In general, the differences in QALY weights between the different severity levels were reduced when adjusted for clinical characteristics and co-morbidities. The difference in QALY weights between the patients with no DR and legal blindness was significant for all instruments. The correlations between the results from NEI-VFQ-25 and TTO, EQ-5D score, EQ-5D VAS and HUI-3 were 0.27, 0.31, 0.38 and 0.68 respectively. CONCLUSIONS: This study presents QALY weights for different severity levels of DR, which can be used in cost-effectiveness analyses of interventions directed to DR. Of the instruments we used, HUI-3 seems to be the most sensitive to changes in HRQoL due to progression of DR. PDB72 FURTHER DEVELOPMENTS OF THE QUALITY OF LIFE ASSESSMENT OF GROWTH HORMONE DEFICIENCY IN ADULTS (QOL-AGHDA) Twiss J1, Mckenna S1, Doward L1, Hána V2, Karbownik-Lewinska M3, Popovic V4, Pura M5, Ribeiro-Oliveira A6, Koltowska-Häggström M7 1 Galen Research Limited, Manchester, UK; 2Charles University, Prague, Czech Republic; 3 Medical University of Lodz, Lodz, Poland; 4University Clinical Centre, Belgrade, Serbia; 5 National Institute of Endocrinology & Diabetology, Lubochna, Slovak Republic; 6 Universidade Federal de Minas Gerais, Belo Horizonte, Brazil; 7Pfizer Endocrine Care, Sollentuna, Sweden OBJECTIVES: The QoL-AGHDA is the first true quality of life (QoL) measure for adult growth hormone deficiency and is widely used in clinical practice and trials. In the UK NICE advises that scores on the QoL-AGHDA should be used to guide treatment selection. The scale has good psychometric properties and has been shown to be responsive to changes in disease severity. The objective of the study was to adapt the QoL-AGHDA for the Czech Republic, Poland, Serbia, Slovakia and Brazil. METHODS: The adaptation in each country required three stages: Translation, cognitive debriefing and a validation survey. The dual panel translation method was used to ensure the items were translated accurately and expressed in everyday language. Cognitive-debriefing interviews with local patients assessed face and content validity. The validation survey tested the psychometric properties of the new scales and included the Nottingham Health Profile (NHP) as a comparator measure. RESULTS: Validation data are not available for Slovakia. Mean scores on the new versions of the QoL-AGHDA ranged from 6.2 to 11.8 (maximum possible = 25). Internal consistency ranged from 0.89–0.91 and test-retest reliability from 0.88–0.93. QoLAGHDA scores were statistically significantly related to; perceived general health and level of fatigue in the Czech Republic, perceived physical activity and level of fatigue in Poland and Serbia and to perceived general health and rated QoL in Brazil. Across the countries mean correlations with NHP sections were (as expected) highest with energy level and emotional reactions (correlations 0.68–0.83) and lowest with sleep disturbance and pain (correlations 0.38–0.46). CONCLUSIONS: This study indicates that (with the exception of Slovakia which requires further validation) the new language versions of the QoL-AGHDA meet the standards of the original UK version and the other 9 existing versions. The new adaptations represent valid and reliable tools for measuring QoL in international clinical trials. A297 PDB73 THE PANORAMA PAN-EUROPEAN SURVEY: HYPOGLYCAEMIA ASSOCIATED WITH DIFFERENT PHARMACOLOGICAL TREATMENTS FOR TYPE 2 DIABETES Simon D1, Bradley C2, Gönder-Frederick L3, Eschwège EM4, de Pablos-Velasco P5, Vandenberghe H6, Bouzamondo H7, Parhofer KG8 1 Hôpital de la Pitié/INSERM U-1018, Paris, France; 2Royal Holloway, University of London, Surrey, UK; 3University of Virginia Health System, Charlottesville, VA, USA; 4INSERM, Villejuif, France; 5Dr Negrin Hospital, Las Palmas University, Las Palmas de Gran Canaria, Spain; 6 AstraZeneca, Zaventem, Belgium; 7Bristol-Myers Squibb, Paris, France; 8Klinikum der Universität München, Munich, Germany OBJECTIVES: Hypoglycaemia can be a side effect of glucose-lowering treatment in patients with type 2 diabetes (T2D) that may counterbalance the beneficial effects of diabetes control. PANORAMA is a large (n = 5156) pan-European cross-sectional survey (NCT00916513) of patients assessing patient reported outcomes and glycaemic control. This subgroup analysis compared rates of severe and non-severe hypoglycaemic events in patients taking different pharmacological treatment regimens. METHODS: Patients with T2D were randomly or consecutively selected from medical practices in eight countries. Patients were aged ≥40 years, with T2D diagnosed >1 year and a clinic medical record available >1 year. All patients received dietary/exercise advice and most were also taking either oral antidiabetic drugs (OADs) and/or injectables (insulin and/or GLP-1 receptor agonists). Patients included in this subgroup analysis had been taking the same pharmacological treatment regimen for ≥12 months. Patient-reported frequency of severe (symptomatic episodes requiring external assistance) and non-severe hypoglycaemic episodes in the past year were examined. RESULTS: In this subgroup analysis 3106 patients were evaluated including: 1346 taking only OADs without secretagogues; 1452 taking only OADs including secretagogues (sulphonylurea/glinides) and 308 on insulin alone. The percentages of patients experiencing >1 non-severe hypoglycaemic episode in each treatment group were: 8.9% for patients taking OADs without secretagogues; 17.5% for patients taking OADs including secretagogues and 47.4% for patients using insulin alone. The differences between these three treatment categories (pair-wise comparisons) were highly significant (P < 0.001). The percentage of patients reporting ≥1 severe hypoglycaemic episode was greater for OADs including secretagogues versus no secretagogues (3.0% versus 1.3%; P = 0.011) and for insulin alone versus OADs including secretagogues (13.7% versus 3.0%; P < 0.001). CONCLUSIONS: Among patients with T2D on glucose-lowering medication, rates of non-severe and severe hypoglycaemic episodes were lowest amongst patients treated with OADs not including secretagogues and highest among patients treated with insulin alone. PDB74 TYPE 2 DIABETES PATIENT PERSPECTIVES ON HYPOGLYCEMIA 1 2 3 Gwaltney C , Martin ML , Gruenberger JB 1 Brown University, Providence, RI, USA; 2Health Research Associates, Inc, Seattle, WA, USA; 3 Novartis, Basel, Switzerland OBJECTIVES: Understanding the perspectives of Type 2 Diabetes (T2D) patients on hypoglycemia is important, in order to understand the burden of the disease and its treatment, and to develop improved outcome measures. The goal of this project was to identify key hypoglycemia-related concepts among T2D patients using qualitative interviews. METHODS: Participants were 19 T2D patients who were prescribed: 1) an oral anti-diabetic (OAD) and had been diagnosed with T2D in the past 3 years (n = 11); 2) a sulfonylurea (SU) or thiazolidinedione (TZD) that caused weight gain, following the failure of an OAD in the last year (n = 5); or 3) an incretin mimetic that caused weight loss, following the failure of an OAD in the last year (n = 3). One-hour in-person interviews were conducted using a semi-structured interview guide. Interviews were coded using ATLAS.ti software and code frequency was used to identify key experiences. RESULTS: Patients were 54.4 years of age on average and 63% were female. Tremor, sweating, and dizziness were the most commonly noted symptoms of hypoglycemia; all five of the patients on an SU or TZD reported experiencing tremor. Patients reported concerns about rare, severe events, such as passing out. Patients generally did not report that their lives were substantially impacted by hypoglycemia, but completing compensatory behaviors to prevent hypoglycemia emerged as an important theme (keeping glucotabs in multiple locations, eating large meals). CONCLUSIONS: Hypoglycemia has a negative impact on patients with T2D. Many report experiencing hypoglycemia symptoms, and, although T2D patients do not often report anxiety about hypoglycemia, several reported concerns about the consequences of severe events. Patients also engage in compensatory behaviors, which suggests that avoiding hypoglycemia is important. Further refinement of the concepts associated with the patient’s experience of hypoglycemia may yield new PRO instruments that can be implemented in clinical trials with T2D patients. PDB75 THE IMPACT OF PERCEPTIONS OF WEIGHT ON OVERALL HEALTHRELATED WELL-BEING IN EUROPEAN PATIENTS WITH TYPE 2 DIABETES MELLITUS (T2DM) Traina SB1, Jamieson C2, Neslusan C1, Li R3, Dennis M3, Ho KF4 1 Johnson & Johnson, Raritan, NJ, USA; 2Johnson & Johnson, Fremont, CA, USA; 3Knowledge Networks, Menlo Park, CA, USA; 4STAT-TU, Inc., Toronto, ON, Canada OBJECTIVES: Maintaining a healthy weight is important in the management of T2DM. We investigated the relationship between weight and patient concerns in relation to overall health-related well-being and function among those with T2DM. A298 METHODS: This cross-sectional study was conducted in the UK, France (FR), and Germany (DE), with data collected via a representative European online volunteer/ opt-in panel (called “MRops”). A 34-minute survey was fielded to 500 subjects with T2DM who were currently receiving pharmacotherapy in each country (n = 1500) from December 2007 to January 2008. Survey items asked about weight, concerns about weight and health, and health-related well-being and function as assessed by the SF-12® Health Survey. RESULTS: Based on self-reported height and weight, 85% (UK), 71% (FR), and 86% (DE) were overweight or obese. Weight was noted as the most important current health concern by 19% (UK), 19% (FR), and 21% (DE). Not all respondents (88% (UK), 79% (FR), 82% (DE)) indicated a desire to lose weight. More subjects with T2DM in DE were “very much” distressed (27%) by their weight than subjects in UK and FR (both 18%). However, the percentage who were “very much” worried about their health because of their weight (UK [20%], FR [14%], DE [21%]) and “very much” embarrassed by their weight (UK [12%] FR [16%] DE [11%]) was similar across countries. Health-related well-being and function, as assessed by the SF-12®, was worse across all domains in those with higher body mass index (BMI) and those with greater weight-related distress, worry, and embarrassment. CONCLUSIONS: Results indicate that elevated BMI and perceptions about weight affect health-related well-being and function as assessed by the SF-12®. Therapeutic options that help people with T2DM to lose weight may have a positive impact on patient lives. PDB76 HEALTH RELATED QUALITY OF LIFE OF TYPE 2 DIABETES MELLITUS PATIENTS IN THE BRAZILIAN PRIVATE HEALTH CARE SYSTEM: FIRST RESULTS OF DIAPS79 STUDY DIAPS79 study group1, Cintra M2, Bahia L2, Franco L3, Nita ME4, Rached R4 1 Diabetes Mellitus Type 2 in the Brazilian Private Health Care System DIAPS79, São Paulo, SP, Brazil; 2MedInsight, Rio de Janeiro, RJ, Brazil; 3FMUSP/Ribeirão Preto, São Paulo, SP, Brazil; 4 Bristol-Myers Squibb S/A, São Paulo, SP, Brazil BACKGROUND: Health-related quality of life (HRQoL) is a multidimensional concept that provides insight regarding the impact of a disease from the patient’s perspective and yield an important health outcome to evaluate treatment interventions and quality of health care. OBJECTIVES: The goal of this study is to reveal HRQoL of type 2 diabetes mellitus (T2DM) patients treated in the Brazilian private health care system. METHODS: A generic instrument to assess health status (EuroQol—EQ-5D), commonly used in diabetic patients worldwide, was applied in 383 outpatients with T2DM from five cities (Fortaleza, Porto Alegre, Rio de Janeiro, Ribeirão Preto e São Paulo). Euroqol uses five dimensions to describe HRQoL states (mobility, self-care, usual activities, pain/discomfort, anxiety/depression) and a visual analogue scale (VAS) recorded by the patient and was measured during only one time during the visit to the physician. RESULTS: The group was composed of 201 women and 182 men, aged 60.5 ± 9.6 years, mean BMI 29.1 kg/m2 and mean duration of diabetes of 12.2 ± 8.75 years. For the entire group, mean EQ VAS score was 75.15 ± 16.72, 24.5% reported some problems with mobility, 4.2% with self-care, 17.5% problems in their usual activities, 44.9% any pain/discomfort and 45.2% anxiety/depression. Mean EQ VAS score decreased as diabetes duration increased (78.19 < 9 years vs. 73.26 > 19 years of disease; P = 0.05). The patients without micro and macrovascular complications (41.5%) had a higher VAS compared to those with these complications (18%) (VAS 77.5 vs. 70.19; P = 0.002). Retinopathy (VAS = 68.13) and heart failure (VAS = 67.05) were conditions associated with lower EQ VAS scores. Regarding the type of treatment, VAS was lower in insulin-treated patients compared to oral agents users (VAS = 76.59 vs. 71.54; P = 0,00004). CONCLUSIONS: A longer duration of diabetes, the presence of chronic complications and the use of insulin showed a negative impact on T2DM patient’s HRQoL. PDB77 QUALITY OF TREATMENT OF DIABETES MELLITUS TYPE 2 IN THE CZECH REPUBLIC Dolezal T1, Kvapil M2, Petrikova A3 1 Institute for Health Economics and Technology Assessment, Prague, Czech Republic; 2 Czech Diabetes Society, Praha, Czech Republic; 3VFU Brno, Brno, Czech Republic OBJECTIVES: Aim of this study was to evaluate the quality of medical treatment of diabetes patients and its trends in the Czech Republic and to compare the findings with the international research. Based on IDF prevalence of diabetes was 9,7% in 2007 and is expected to rise to 11,7% in 2025. METHODS: Data were extracted from Czech cross-sectional studies from 2002 and 2007 and from European-based studies. Follow-up of short term parameters of quality of health care (HbA1c, BMI, blood pressure, lipids, treatment algorithms) which help to predict long-term incidence of complications. Frequency of microvascular and macrovascular complications was also assessed and the data were compared to statistics of the Czech Institute for Health Information and Statistics. RESULTS: Concerning short-term parameters there is wide variation accross European countries. We have not found significant differences between CR and selected European countries although there is insufficient evidence in revealing end-point values, e.g. HbA1c (7,7% in CR vs. 7,8% in Great Britain), BMI (29,9 in CR vs. 28,7 CODE-2 study) and reaching of target HbA1c values (36–42% in CR vs. 36% in CODE-2). 74% of Czech patients compared with 50% European patients are treated with metformin. Percentage of patients using antihypertention drugs (83%) and hypolipidemics (63%) is similar in CR and the European average. The diabetic patients are reaching the tagret therapeutic values only in minority of cases (31% in systolic blood pressure, 27% in total cholesterol and in 36% of HbA1c). There is high prevalence of diabetic macrovascular and microvascular com- 13th Euro Abstracts plications (CHD = 49%, stroke = 9,3%, nephropathy = 28,3%, retinopathy = 25%, diabetic foot = 4,6%). CONCLUSIONS: This comparative analysis is the first example of systematic evaluation of the quality of health care concerning diabetes patients in CR. The future objective is to set proper quality indicators and follow them on continuous basis. DIABETES/ENDOCRINE DISORDERS – Health Care Use & Policy Studies PDB78 WILLINGNESS-TO-PAY FOR DIABETES DRUG THERAPY BASED ON META-ANALYSIS RESULTS Jendle J1, Torffvit O2, Ridderstråle M3, Ericsson Å4, Nilsen B4, Bøgelund M5 1 Faculty of Health Sciences, Örebro, Sweden; 2Department of Nephrology, Inst Clinical Sciences, Malmö, Sweden, Sweden; 3Department of Clinical Sciences, Malmö, Sweden, Sweden; 4Novo Nordisk Scandinavia AB, Malmö, Sweden; 5Incentive Partners, Birkerød, Denmark OBJECTIVES: This study aimed to investigate in people with type 2 diabetes (T2D) the relative willingness-to-pay (WTP) for different diabetes drug therapies based on outcomes of clinical trials with liraglutide. METHODS: WTP for diabetes drug therapy of people with T2D was assessed by combining results from a meta-analysis of liraglutide compared with other diabetes drug therapies in the liraglutide clinical development program (LEAD) and a survey on WTP for important aspects of diabetes medication in people with T2D. a meta-analysis of six randomised trials with 3967 subjects in the LEAD program compared efficacy and safety of liraglutide, a once-daily human glucagon-like peptide-1 (GLP-1) analogue, vs. rosiglitazone, glimepiride, insulin glargine, and exenatide. The WTP survey had 461 participants with T2D from Sweden and used a discrete choice experiment methodology to evaluate convenience and clinical effects of treatments in T2D. Results were converted from SEK to c (10.14 SEK/c). RESULTS: Combining meta-analysis and WTP results revealed that people with T2D preferred liraglutide to all comparators. They were willing to pay an extra c2.49/day for liraglutide 1.2 mg compared with rosiglitazone, c1.82/day compared to glimepiride, c3.17/day compared to insulin glargine, and c0.74/day compared to exenatide. For the comparisons with rosiglitazone, glimepiride and insulin glargine, the largest component was based on the additional weight improvements with liraglutide. Compared to exenatide, the largest component of preference was administration of the drug. CONCLUSIONS: WTP for liraglutide by people with T2D was noticeably higher compared to other standard therapies based on the clinical results from the meta-analysis. Primary drivers were weight decrease (compared to rosiglitazone, glimepiride and insulin glargine) and administration (compared to exenatide). In total, people were willing to pay up to c3.17/day more to use liraglutide 1.2 mg rather than use other glucose lowering treatments. PDB79 DURATION OF FIRST INSULIN THERAPY IN PREVIOUSLY UNCONTROLLED TYPE 2 DIABETES: COMPARISON OF INSULINS GLARGINE, DETEMIR AND NPH AND ASSOCIATION WITH GLYCEMIC CONTROL Hall GC1, McMahon AD2, Dain MP3, Home PD4 1 Grimsdyke House, London, UK; 2Glasgow University Dental School, Glasgow, UK; 3 Sanofi-Aventis, Paris, France; 4University of Newcastle upon Tyne, Newcastle upon Tyne, UK OBJECTIVES: The duration of treatment with NPH, as a first insulin medication in type 2 diabetes is reported to be shorter than that of insulin glargine, with comparisons of glargine and insulin detemir inconsistent. We investigated whether differences remained after adjustment for potential confounders, including baseline HbA1c, and which factors are associated with longer duration of use of first insulin. METHODS: People on two or three oral glucose-lowering agents (OGLA) with poor glycaemic control who started long-term insulin treatment (2000–2007) were grouped by insulin type, NPH, glargine or detemir, and followed until a prescription for a different insulin or a GLP-1 mimetic. Time to treatment change was compared between groups in a Cox model adjusting for baseline characteristics: HbA1c, age, sex, year, time from diagnosis, concomitant OGLA, glomerular filtration rate, cardio- and micro-vascular medical history and cardiovascular risk factors. The association of this duration with baseline characteristics and mean treatment HbA1c was investigated by univariate analyses. All data came from THIN database of electronic UK primary care records. RESULTS: The analysis included 1044 people started on glargine therapy; 187 on detemir and 453 on NPH. Compared to glargine the adjusted hazard ratios for time to treatment change were 1.78 (95%CI 1.43, 2.20) for detemir and 1.52 (1.29, 1.80) for NPH. Lower mean HbA1c with any insulin type was associated with longer time to change of regimen (Spearman rank correlation coefficient −0.30, P < 0.001). No concomitant OGLA use, increasing age and time from diagnosis, decreasing BMI and baseline HbA1c and heart failure were also associated with longer first insulin duration. CONCLUSIONS: People who commenced insulin treatment with glargine remained on their initial insulin for longer than those who initiated detemir or NPH after adjustment for potential confounders. Better glycaemic control resulted in a longer time to change of insulin. 13th Euro Abstracts PDB80 A CONTROLLED, PROSPECTIVE LONGITUDINAL, INTERVENTIONAL, NATURALISTIC STUDY TO EVALUATE THE UNIQUE HEALTH CARE PROGRAM “CONVERSATION MAP©” FOR TURKISH PATIENTS WITH TYPE-2 DIABETES IN ONE GERMAN MAJOR CITY Korolewa V, Kretschmer B, Clouth J Lilly Deutschland GmbH, Bad Homburg, Germany OBJECTIVES: Considering a cultural and epidemiological disparity between Turkish and German type-2 diabetes patients in Germany, resulting in a health care gap, a customised diabetes program for Turkish migrants is required. The objective was to evaluate “Conversation Map©”, a unique health care education program for Turkish migrants with type-2 diabetes in Germany. METHODS: Ninety-two Turkish patients with poor German language skills were divided into interventional (IV, n = 44) and control (C, n = 48) groups, 42 German (G) patients served as additional control group. The IV group attended a structured once weekly, 2 hr “Conversation Map©” training over 4 weeks in Turkish language. The two control groups had no intervention. Sociodemographic data were recorded at baseline, and standardized questionnaires assessing treatment, diabetes-knowledge (0–6; 6 = everything right), and therapy adherence (“Essential Training Content (ETC)” 0–10; 10 = best) were applied at baseline and after 4 weeks after training in the IV group. Baseline differences between the groups and changes of data after 4 weeks compared to baseline were tested (paired t-test, α = 0.05, two-sided). RESULTS: Ethnic groups were comparable on demographic data, whereas Turkish patients had lower education level (P < 0.001). 55% of all Turkish patients (IV and C) were insulin-dependent and 63% were previously not attending any diabetes-training. Baseline diabetes knowledge of Turkish patients was significantly lower than in the German control group (IV 2.38, T 2.77, G: 4.10, P < 0.001), as well as the ETC score (IV 6.37, C 5.92, G 7.98, P < 0.001). After “Conversation Map©” training the Turkish group achieved significantly higher mean knowledge- and ETC-score values compared to baseline (difference +70%, ETC: +25%, P < 0.001) and were comparable to the German group. CONCLUSIONS: The study demonstrated that the cultural sensitive training “Conversation Map©”, led to a significant increase of disease-knowledge and therapy adherence, helping to improve the health care situation of Turkish patients, with cost savings potential in the long term. PDB81 TIME TO ADD-ON MEDICATION USE FOR PATIENTS WITH TYPE 2 DIABETES MELLITUS (T2DM) WHO FAILED METFORMIN MONOTHERAPY Qiu Y1, Fu AZ2, Radican L1 1 Merck & Co., Inc., Whitehouse Station, NJ, USA; 2Cleveland Clinic, Cleveland, OH, USA OBJECTIVES: Add-on medications are recommended if target glycemic goals for patients with T2DM are not achieved or sustained after initial metformin monotherapy. This study examined the time to add-on medication use after metformin monotherapy failure in clinical practice. METHODS: Selected from a large US EMR database between Janujary 1, 1997 and December 31, 2008, included patients had to be aged ≥18 years with a diagnosis of T2DM who had HbA1c ≥ 7.0% or ≥ two fasting blood glucose levels of 126 mg/dL or greater. Treatment failure was defined as HbA1c ≥ 7% (index date) after metformin monotherapy for at least 6 months. Baseline data were extracted during 1 year prior to the index date. Time to add-on medication use was time between index date to the first add-on medication use during follow-up and was evaluated for the overall cohort and for three index HbA1c subgroups: <8%, 8–9%, and >9%. a Cox proportional hazard model was employed to determine baseline clinical and demographic characteristics associated with shorter time to add-on medication use. RESULTS: There were 12,566 patients meeting the inclusion criteria; 8656, 2175 and 1735 had index HbA1c < 8%, 8–9% and >9%, respectively. The overall mean (SD) age was 63 (12) years and 51% were female. The median time to add-on medication use was 15.7 months overall and 17.0, 13.9 and 11.3 months for patients with index HbA1c < 8%, 8–9% and >9%, respectively. Higher index A1c, greater body mass index, higher Charlson comorbidity index, younger age, males, lower LDL were significantly associated with shorter time to add-on medication use (all P < 0.05). CONCLUSIONS: This indicates, in US clinical practice, it takes over a year for a diabetic patient who has suboptimal glycemic level after initial metformin monotherapy to receive add-on medications. There is room through disease management so that patients who have failed metformin monotherapy, if eligible and appropriate, receive add on therapy sooner rather than later. PDB82 PRESCRIBED DAILY DOSES OF ONCE-DAILY LIRAGLUTIDE IN THE GERMAN STATUTORY HEALTH INSURANCE (SHI) Schöffski O1, Mentrup S2, Lund N3 1 Friedrich-Alexander-Universität Erlangen-Nürnberg, Nuremberg, Germany; 2Novo Nordisk Pharma GmbH, Mainz, Germany; 3Novo Nordisk A/S, Soeborg, Denmark OBJECTIVES: To evaluate the prescribed daily dose of liraglutide for patients in German statutory health insurances. The novel once-daily incretin analogue, liraglutide, mimics the effect of endogenous glucagon-like peptide 1 (GLP-1). Liraglutide was launched in Germany in July 2009 in a device allowing 3 different dosages (0.6 mg for initial titration; and the two maintenance doses 1.2 and 1.8 mg). The defined daily dose (DDD) was set by WHO at 1.2 mg. The prescribed daily dose (PDD) has not been evaluated so far. METHODS: Sampled data from German statutory health insurances (SHI) was provided by Insight Health (patient tracking data) for the period from July 2009 to March 2010 (9 months) covering about 11% of all patients in A299 German SHI. All patients with prescriptions of liraglutide were identified (n = 4,284). Patients with only one prescription and less than 4 weeks between first and last date of prescription were excluded to avoid overrepresented influence of low dosed therapy starters (remaining patients: n = 2,118). Total number of prescribed pens was determined from first to second last prescription and total consumption in mg was calculated. Number of days between first and last prescription was determined and mean average consumption in mg per day was calculated. RESULTS: The mean PDD of liraglutide in 2118 patients was 1.28 mg. Sensitivity analyses including only patients with longer periods of use showed decreasing average consumption; patients with at least 10 weeks duration of treatment showed a mean daily dose of 1.25 mg. Stocking effects (prescription before package is empty) could have driven the increase in average use in the starting period and at the end of the calendar year. CONCLUSIONS: The longer patients use liraglutide the lower the observed PDDs. The DDD of 1.2 mg is a valid estimate for real life usage of liraglutide. PDB83 TIME TO TREATMENT MODIFICATION AMONG PATIENTS WITH TYPE 2 DIABETES WHO INITIATED EXENATIDE OR INSULIN GLARGINE Pawaskar M1, Bonafede M2, Johnson BH3, Fowler R3, Hoogwerf B4 1 Eli Lilly & Company, Indianapolis, IN, USA; 2Thomson Reuters, Cambridge, MA, USA; 3 Thomson Reuters, Washington, DC, USA; 4Lilly USA, LLC, Indianapolis, IN, USA OBJECTIVES: To examine time to treatment modification for patients with type 2 diabetes (T2D) initiating treatment with exenatide BID or insulin glargine. METHODS: A retrospective analysis was conducted using the Thomson Reuters Research Databases. Adult patients with T2D who initiated exenatide (N = 9197) or glargine (N = 4499) between October 1, 2006 and September 30, 2007 with 12 months pre- and 18 months post-index continuous enrollment were included. Treatment modification was defined as the first event of change in the treatment and was classified further into three types: intensification, switching or discontinuation of the index medication. The 2 cohorts were 1:1 propensity score matched on baseline demographics, clinical characteristics, and prior health care utilization. Survival analysis was used to examine time to treatment modification. RESULTS: Propensity score matching resulted in inclusion of 3774 exenatide and 3774 glargine patients with a mean age of 57 years, mean Deyo Charlson Comorbidity Index score of 1.6, and with proportionately more males (54%) than females. All of the patients concomitantly used a non-index antidiabetes medication in post-index period. The 18-month discontinuation rates were 38.3% and 40.0% (P = 0.14) and the treatment intensification rates were 15.9% and 26% (P < 0.0001) for exenatide and glargine, respectively. Alternatively, 14.9% of exenatide-treated patients switched therapies, compared to 10.0% of glargine-treated patients (P < 0.0001). Glargine-treated patients were 33% more likely to modify treatment than exenatide-treated patients (Hazard Ratio (HR): 1.33, P < 0.0001). Glargine-treated patients were 72% more likely to intensify their treatment (HR = 1.72, P < 0.0001), 25% more likely to discontinue (HR = 1.25, P < 0.0001), and 29% less likely to switch to new therapy (HR = 0.71, P < 0.0001) compared to exenatide. CONCLUSIONS: This analysis showed that exenatide-treated patients were less likely to modify their treatment suggesting potential longer durability with exenatide therapy. Furthermore, exenatide-treated patients were less likely to discontinue or intensify but more likely to switch their treatment than glargine-treated patients. PDB84 MANAGEMENT OF ACROMEGALY IN CLINICAL PRACTICE CONDITIONS IN SPAIN Roset M1, Merino-Montero S1, Luque M2, López-Mondéjar P3, Salinas I4, Soto A5, Bernal C6, Villabona C7, De Luis D8, Donnay S9, Pascual H10, Pérez-Luis J11 1 IMS Health, Barcelona, Spain; 2Hospital Universitario La Princesa Servicio de Endocrinología y Nutrición, Madrid, Spain; 3Servicio de Medicina Interna-Endocrino. Hospital General Universitario de Elche, Alicante, Spain; 4Servicio de Endocrinología y Nutrición. Hospital Germans Trias i Pujol, Barcelona, Spain; 5Servicio de Endocrinología y Nutrición. Hospital Universitario Virgen del Rocío, Sevilla, Spain; 6Servicio de Endocrinología y Nutrición Hospital 12 de Octubre, Madrid, Spain; 7Servicio de Endocrinología Hospital Universitario de Bellvitge, Barcelona, Spain; 8Sección de Endocrinología y Nutrición Clínica, Unidad de Apoyo a la Investigación, Valladolid, Spain; 9Unidad de Endocrinología y Nutrición de la Fundación Hospital de Alcorcón, Madrid, Spain; 10Sección de Endocrinología y Nutrición Hospital General Universitario Morales Meseguer, Murcia, Spain; 11Servicio de Endocrinología Hospital Universitario de Canarias, Tenerife, Spain OBJECTIVES: The goal of treating acromegaly is directed at removing the tumor, preventing tumor re-growth and reducing long-term morbidity and mortality. For this purpose, different health resources are necessary. This study evaluates the disease resources and costs in Spain. METHODS: An epidemiological, prospective, naturalistic, multicentric study (30 endocrinologists) involving acromegalic patients with micro (≤10 mm) or macro (>10 mm) adenomas was performed. Patients were categorised as Surgical Group (SG) (surgery in the 6 months before inclusion or during follow-up period and somatostatin analogue (SA) treatment for <6 months during the pre-surgery period) and Medical Treatment Group (MTG) (patients receiving SA treatment for ≥6 months, with/without surgery following SA treatment). Resource data were collected from standard visits during a 2-years period. RESULTS: The study included 74 patients (56 SG and 18 MTG). Most patients were women (62%). The mean (SD) age was 49 (14) years. The annual direct acromegaly cost per patient is c9668 (c9223 SG vs. c11,054 MTG). The cost of illness was higher in patients with macroadenoma than in microadenoma (c11,053 vs. c5809), and it is increased in young patients (<40 years). Surgical procedures in acromegaly (involving hospitalization and complementary tests) constitute the 22% and 8% of the illness cost in SG A300 and MTG patients. Additionally, hospitalizations in intensive care units (50% SG vs. 22% MTG patients), emergency visits (21% SG vs. 6% MTG) and the presence of adenoma complications (73% SG vs. 44% MTG) constitute a source of cost increment in these patients. Patients who accomplish with the most strict study clinical control criteria (GH < 1.0 and IGF-1 < 100%) showed the lowest direct cost of illness (c6169 vs. c12,990). CONCLUSIONS: The economic cost of acromegaly is dependent on the clinical control of the disease. Direct cost of illness is the half that the cost in non controlled patients. PDB85 APPROVAL AFTER REJECTION—AN INSIGHT IN HTA RE-EVALUATIONS Sweeney N, Andreykiv M, Wiebinga C Quintiles Consulting, Hoofddorp, The Netherlands OBJECTIVES: To gain insight into the re-evaluation process of HTA agencies after an initial rejection and identify the adaptations that led to the approval of re-submitted dossiers. METHODS: Phase I: manual search of 57 health care agencies’ websites for published diabetes-related assessments (January 2007-May 2010). Phase II: the two most re-assessed drugs for which detailed information was available were selected for further evaluation (insulin glargine and exenatide). For these drugs, all reports published prior to 2007 were also included. RESULTS: Phase I identified 117 relevant single technology appraisals; 18 were re-evaluations. Six agencies performed re-evaluations of the same drug after an initial rejection: CADTH, CVZ, HAS, PBAC, AHTAPol and SMC. To date, SMC evaluated 32 submissions for 13 anti-diabetic drugs, PBAC published 20 (eight drugs), CADTH 13 (four drugs), CVZ 14 (four drugs) and AHTAPol 10 (two drugs). In phase II insulin glargine (four re-submissions to PBAC and 1 to CADTH) and exenatide (two re-submissions to PBAC, 1 to CVZ and 1 to AHTAPol) were evaluated. It became clear that payers do focus on overall cost. The approach that was chosen for those two drugs was to control overall cost either by restricting access or by settling on a lower price. CVZ accepted exenatide for reimbursement only after restricting access to a subgroup of obese type 2 diabetes mellitus patients (with an ICER of c5.231). Instead of patient segmentation PBAC insisted on lowering the price for both medications (rationale for insulin glargine being concern that prescribing cannot be contained within the defined population). AHTAPol limited exenatide reimbursement to 50% to control prescribing rates. CONCLUSIONS: For the diabetes cases analyzed HTA agencies attempted to control health care expenditure by either lowering drug costs or by narrowing the definition of the target population, the latter inevitably allowing fewer patients access to the drug. PDB86 HEALTH TECHNOLOGY ASSESSMENT OF DIABETES COMPOUNDS: THE POLISH PERSPECTIVE Adalsteinsson JE1, Czech M2, Skrekowska-Baran I3, Jasik BM1 1 Novo Nordisk A/S, Copenhagen, Denmark; 2Department of Pharmacoeconomics, Medical University of Warsaw, Warsaw, Poland; 3Novo Nordisk Pharma Sp, Warszawa, Poland OBJECTIVES: The AOTM in Poland was established to give MoH in Poland advice on reimbursement. The aim of this research is to create an overview of HTA reports on diabetes compounds in Poland and the results of the decision making. METHODS: A search was conducted on the webpage of AOTM (http://www.aotm.gov.pl) for HTA reports on the following products: Rosiglitazon, Pioglitazon, Sitagliptin, Vildagliptin, Saxagliptin, Exenatide, Liraglutide, Glargine, Detemir, Aspart, Glulisene and Lispro. RESULTS: Of a total of 163 reports (published between 2007 and 2010), eight reports in Polish language on diabetes were identified and assessed. Two reports can be viewed as secondary assessment of regulatory safety discussions. The other six reports assessed the implementation of new diabetes compounds with assessment of efficacy, safety and cost-effectiveness of the drugs.Two reports assessed safety concerns associated with the risk of cancer and concluded based on EMA and FDA research that no increased risk was associated with these agents. Rosiglitazone and Sitagliptin were not recommended for reimbursement due to availability of other treatments with similar efficacy. Saxagliptin, Exenatide and Liraglutide got the recommendation to be reimbursed due to expected increase in QALYs. The final report was assessing Glulisene which got the recommendation to temporary reimburse (2 years) provided that data on hard endpoints (not specified in public report) and cost-effectiveness should be delivered. CONCLUSIONS: Recommendation by AOTM is supported by assessment of available RCTs, cost per life-year gained, cost per QALY, estimated budget impact for 5 years and also in some cases reports from EMEA, FDA and other HTA agencies (SMC, PBAC and CADTH). The AOTM’s recommendation is not obligatory for the Polish Ministry of Health. PDB87 STANDARDS FOR THE ASSESSMENT OF ANTIDIABETIC DRUGS—THE IQWIG PERSPECTIVE Schweikert B1, John J2, Ringborg A3, Erhardt W4, Bleckmann A5, Neubauer AS4 1 i3 Innovus, Aschheim, Germany; 2Helmholtz Zentrum München, Neuherberg, Germany; 3i3 Innovus, Stockholm, Sweden; 4Bristol-Myers Squibb, München, Germany; 5AstraZeneca, Wedel, Germany OBJECTIVES: A substantial number of new pharmaceutical treatment strategies have been introduced for the treatment of diabetes mellitus type II. The availability of these drugs for patients in different countries depends on the evaluation standards and methods applied in the various phases of drug assessment. Objective of this research was to review the requirements and criteria applied for the assessment of antidiabetic 13th Euro Abstracts drugs along the regulatory process by EMA (Europe) and FDA (USA) for the assessment of efficacy and safety as well as for reimbursement decisions by NICE (England) and IQWiG (Germany) and to compare their consistency, with a special focus on IQWiG’s procedures. METHODS: A review of relevant current method documents and reports on evaluations of antidiabetic drugs published by IQWiG was conducted. These were compared with guidance documents issued by FDA, EMA and NICE with respect to endpoints considered in diabetes and their definition, criteria for the type of evidence, and potential comparators. RESULTS: Consistently, across all agencies severe and non-severe hypoglycemias were considered highly relevant. There was, however, a substantial heterogeneity in the definition of hypoglycaemias. The surrogate parameter HbA1C as primary endpoint was accepted by all agencies investigated apart from IQWiG. In its assessments, evidence from randomized as well as from observational studies was accepted by NICE. For safety evaluations preclinical studies were taken into consideration by EMA and FDA in addition to randomized controlled trials. IQWiG on the other hand focused exclusively on randomized controlled trials for the assessment of effectiveness as well as safety. CONCLUSIONS: There is a substantial variation of criteria applied and evidence considered relevant within the assessment process of IQWiG compared to other agencies. This might lead to regional variations in the availability of drugs. It is important to be aware of the different requirements of agencies, when designing trials and planning market access. PDB88 LEARNING FROM DISEASE MANAGEMENT PROGRAMMES: HOW MEDICAL TREATMENTS AND QUALITY OF DIABETIC CARE (TYPE II) IN GERMANY ARE DIRECTLY AND INDIRECTLY IMPROVED BY DMPS Frenzel A1, Reuter A2 1 IMS Health GmbH & Co. OHG, Frankfurt, Germany; 2Freie Universität Berlin, Berlin, Germany OBJECTIVES: Disease Management Programmes (DMP) aim at improving care quality by implementing standards for medical practices. In the case of Diabetes Mellitus Type II (DM II), care improvements can be assessed by the duration between the first diagnosis and the occurrence of the first related complication. The aim of this longitudinal study is to investigate the direct influence of the DMP-based treatments on patient outcomes, measured as the postponement of diabetes related complications in a large population of DM II patients. The study also investigates how DMP inscriptions of some patients of a medical practice indirectly influence patient outcomes of DM II patients, who are not inscribed in a DMP, but are treated in the same practise. We argue that this indirect effect is due to physicians’ learning from the DMP-based treatments in their clinics. METHODS: Using consultation data from IMS Health from a period of 25 years (1984–2009) a survival analysis is applied. The data set includes 161,747 DM II patients from >1100 practices. Applying a Kaplan–Meier– Method we test for direct effects of DMPs on patient outcomes. By pooling patients by the registration year of the practice-leading physician and by focussing on their quarterly consultation rate, we test for indirect effects of DMPs on patient outcomes. RESULTS: The mean survival time (duration between first diagnosis and first complication) of the medical treatment of diabetics in a DMP is 14,82 years, differing significantly from the 15,76 years without a DMP. These tests are controlled for following patient variables: sex, age, HbA1C, BMI and the insurance status. Learnings from DMPs, indirectly affecting DM care, significantly postpone complications for younger physicians and practices with fewer diabetics. CONCLUSIONS: Contributing to assessments of DMPs, the study discusses policy implications, as it is shown that care quality is improved by physicians learning from DMPs. PDB89 PREDICTORS OF ROUTINE MONITORING OF DIABETES CARE AMONG THE US NON-INSTITUTIONALIZED POPULATION: A RETROSPECTIVE ANALYSIS OF THE MEDICAL EXPENDITURE PANEL SURVEY (MEPS) IN 2007 Zhao Y, Fonseca V, Campbell C, Shi L Tulane University, New Orleans, LA, USA OBJECTIVES: To examine the rate and predictors of diabetes monitoring in the US. METHODS: This cross-sectional retrospective study was conducted on a representative, non-institutionalized sample of the US population, using the self-reported information from the 2007 Household Component (HC) of the MEPS. According to the American Diabetes Association (ADA) 2007 practice guidelines, proper provider monitoring is defined as at least two A1c tests, one eye and one foot examination annually. Health status was measured by SF-12®Version2. a logistic regression model was used to examine the predictors of proper monitoring. Differences in health status and medical expenditures between patients with and without proper monitoring were examined using t-tests. Estimates were weighted to the total population (WTP). RESULTS: Among 1,747 (WTP: 19,320,394) patients with diabetes, 80.64% had at least two A1c tests; 63.29% had an eye examination; and 67.51% had a foot examination. Thus, 63.36% patients (WTP: 14,065,289) received proper diabetes monitoring. Older patients (OR:1.021, 95% confidence interval [CI]: 1.012–1.030), non-Hispanic Caucasians compared with African American patients (OR: 1.236, 95% CI: 0.933– 1.636), patients with a higher education level (OR:1.211, 95% CI: 1.056–1.390), insurance coverage (OR:2.216, 95% CI: 1.408–3.486), use of oral anti-diabetic drugs (OR:2.935, 95% CI: 2.131–4.042) and insulin (OR:3.453, 95% CI: 2.477–4.814) were more likely to undergo the proper monitoring. Well monitored patients had a higher Mental Component Summary score (50.09 ± 0.37 vs. 48.51 ± 0.45, P < 0.05), but a lower Physical Component Summary score (39.95 ± 0.34 vs. 42.28 ± 0.47, P < 13th Euro Abstracts 0.05) on the SF 12. Properly monitored patients spent significantly more on total health care services (+$5243), outpatient visits (+$1023), and medications (+$1204), respectively (all P-values<0.05). CONCLUSIONS: In the US, nearly 40% patients with diabetes do not receive the proper diabetes monitoring controlling for racial and socioeconomic disparities. Anti-diabetics/insulin use, mental/cognitive status, physical health status, and health care expenditure may also interact with performing monitoring. Barriers and cost-benefit for long-term monitoring should be studied. PDB90 IMPROVING PATIENT CARE: RESULTS FROM THE TEXAS NEWBORN SCREENING PERFORMANCE MEASURES PROJECT Tiwana SK1, Tanksley S2, Williams D2, Douglas M2 1 Health Partners, Bloomington, MN, USA; 2Texas Department of State Health Services, Austin, TX, USA BACKGROUND: Advances in screening technology have led to rapid expansion of newborn screening as a public health initiative in the United States. Due to variation across states in program implementation, there is a lack of standardization and accountability, which can affect quality of care. Texas currently screens for 28 disorders. OBJECTIVES: The Texas Newborn Screening Performance Measures Project was initiated with the objective of developing evidence-based performance measures to improve quality, accountability and uniformity in the care. METHODS: A three step approach was used for identification and development of key measures for seven most critical disorders: systematic reviews; impact assessment; and feasibility assessment. Impact assessments (likert scale) were based on National Quality Measures Clearinghouse™ (NQMC) criteria: scientific soundness, relevance, health importance, impact on quality of care, and ability to address disparities in care. A likert scale was also used for feasibility assessment on key aspects: data availability, ease of collection, infrastructure and human resource needs, overall cost, and time constraints. RESULTS: A total of 50 performance measures were supported by scientific evidence. Impact and feasibility assessments led to the approval of 33/50 measures. “Time to initiate treatment” received the highest scores on potential impact on patient outcomes (mean impact score 86.67/100, SD 1.5). Other measures with potentially high impact (score >80/100) were: compliance with oral prophylactic medication and age at first Prevnar® vaccination in sickle cell disease; screening of at-risk family members in fatty acid disorders; frequency of growth assessments in congenital adrenal hyperplasia and phenylalanine levels in phenylketonuria. “Time to treatment” for individual disorders was also ranked very high on feasibility (mean feasibility score 88.67/100, SD 3.88). CONCLUSIONS: This is one of the first efforts to identify and develop evidence-based performance measures in newborn screening and can pave the way for system wide changes and development of national guidelines. PDB91 TREATMENT PATTERNS AND ACHIEVEMENT OF THERAPEUTIC GOALS IN A COHORT OF TYPE 2 DIABETES MELLITUS PATIENTS TREATED IN THE BRAZILIAN PRIVATE HEALTH CARE SYSTEM (PHCS): INITIAL REPORTS OF DIAPS 79 STUDY DIAPS79 study group1, Eliaschewitz FG2, Bahia L3, Cintra M3, Franco L4, Nita ME5, Rached R5 1 Diabetes Mellitus Type 2 in the Brazilian Private Health Care System DIAPS79, São Paulo, SP, Brazil; 2CPCLIN—Centro de Pesquisas Clínicas, São Paulo, SP, Brazil; 3MedInsight, Rio de Janeiro, RJ, Brazil; 4FMUSP, Ribeirão Preto, SP, Brazil; 5Bristol-Myers Squibb S/A, São Paulo, SP, Brazil OBJECTIVES: Investigate the type 2 diabetes mellitus (T2DM) treatment practice and achievement of therapeutic goals in a cohort of T2DM patients treated in the Brazilian PHCS. METHODS: This is a cross-sectional analysis of 383 T2DM outpatients treated in 5 cities covered by the PHCS. Data was collected using information from the previous year by interviewing patients using a validated questionnaire complemented by medical chart review. Therapeutic goals suggested by the Brazilian Diabetes Society were used as standard treatment. RESULTS: Mean age was 60.5 ± 9.6 years, mean BMI 29.1 ± 5.3 Kg/m2 and mean duration of diabetes of 12.2 ± 8.75 years. The frequency of associated arterial hypertension was 66.8% (n = 256); obesity 39% n = 144); dyslipidemia 69.6% (n = 267) and heart failure 5.5% (n = 21). Microvascular complications were present in 32.4% (n = 124); 8.1% (n = 31) had macrovascular complications; 18% (n = 69) had at least one micro and one macrovascular complication and 41.5% (n = 159) had no complications. Current treatment practice was: 3.6% (n = 14) diet only; 30% (n = 115) monotherapy with oral anti-diabetic (OAD); 37.8% (n = 145) combined therapy with more than one OAD agents; 23.2% (n = 89) combined therapy with insulin and 5.2% (n = 20) just insulin. The most prescribed drug was metformin (41.4%), followed by insulin (19.1%), sulphonylureas (18.6%), DPP4 inhibitors (8.7%), TZDs (5.5%), and others (6.7%). Medications for dyslipidemia was used by 47.5% (n = 178); cardiovascular drugs in 77.4% (n = 287) and antiobesity drugs in 4% (n = 15). Treatment targets were achieved in: 76.8% for systolic and diastolic blood pressure (<130 × 85 mmHg), 19.5% for BMI < 25 kg/m2, 28% for HDL cholesterol >55 women, 37.3% for HDL cholesterol >45 men and 28.2% reached the goal for HbA1c (within normal range). Only 6.8% of the group (n = 26) reached all the treatment targets. CONCLUSIONS: The national goals for glycemic control, blood pressure and lipid levels are rarely achieved in real-world clinical practice, even with the high use of medications. A301 DIABETES/ENDOCRINE DISORDERS – Conceptual Papers & Research on Methods PDB92 CLINICAL AND ECONOMIC CONSEQUENCES OF THE PHARMACOLOGICAL HYPOGLYCEMIC TREATMENT OF TYPE 2 DIABETES IN CROATIA Saric T1, Benkovic V2, Poljicanin T3, Sekerija M3 1 Promeritus savjetovanje d.o.o, Zagreb, Croatia; 2Croatian Society for Pharmacoeconomics, Zagreb, Croatia; 3Vuk Vrhovac University Clinic, Zagreb, Croatia OBJECTIVES: Diabetes mellitus type 2 (T2DM) is a chronically progressive disease and the treatment must be selected according to the pathophysiological phase of the disease at the time the treatment is begun. The Croatian public diabetology sector takes care of approximately 150,000 adults treated with oral hypoglycemic agents (OHA) alone or in combination with insulin. Our objective was assessment of the clinical and economic consequences of OHA treatment in T2DM from a Croatian health care system perspective. METHODS: The target population defined for the study was diabetic patients treated with OHA alone. Medication consumption was quantified by using Pharmis and CroDiab data. a clinical expert panel provided resource-use information not available in published literature or health care databases. RESULTS: Current consumption data is showing that 62.8% patients are using OHA as monotherapy. Within this group, majority is using either biguanides or sulphonylureas. Patients treated with sulphonylureas are represented with almost the same percentage as those treated with biguanides (25% vs. 29%). Combination of two OHA is used in 34.2% while 3% of patients are treated by triple therapy. The most often choice in dual therapy is combination of biguanides and sulphonylureas while biguanides, sulphonylureas and thiazolidinediones in combination as the most favourable treatment option in triple therapy. Biguanides are used as a one of OHA in 61% of patients. CONCLUSIONS: Considering current clinical guidelines, lifetime benefits of biguanides and facts that they are low-cost agent, relatively small proportion of T2DM patients are treated with this agent in Croatia. Findings of this investigation revealed real life pattern of T2DM treatment, which enables directing in better treating and more cost-effective management in Croatia. PDB93 THE RELIABILITY OF PROPORTION OF DAYS COVERED CALCULATIONS USING DEFINED DAILY DOSE ESTIMATES Levin R, Papademetriou E, Aubert RE Medco Health Solutions, Franklin Lakes, NJ, USA OBJECTIVES: Using a large, US administrative pharmacy claims database, calculate proportion of days covered (PDC) using actual days supply, and compare estimated PDC (ePDC) using days supply derived from drug-specific Defined Daily Dose (DDD) criteria. METHODS: Continuously eligible patients filling non-insulin anti-diabetic medication were targeted from a large sample of pharmacy claims during 2008 and 2009. Medications were grouped into ATC diabetes drug classes. Proportion of days covered (PDC) was calculated as the number of days a patient had medication in their possession divided by the number of days in the period. PDC was first calculated using actual days supply, then ePDC was calculated using an estimated days supply from DDD, strength, and pill quantity. The percent of patients adherent to therapy was defined by a PDC > 0.80. The reliability of each method was assessed by Pearson correlation coefficients and agreement above chance was assessed using Kappa statistics. RESULTS: Adherence was calculated for 163,750 patients taking non-insulin anti-diabetics. Overall, the PDC was 0.69 and ePDC 0.57. The two measures were also highly and significantly correlated (r = 0.73; P < 0.0001). The percent adherent was 48.8% (PDC) and 34.4% (ePDC), (Kappa = 0.50; P < 0.0001). At the medication class level, differences in PDC and ePDC ranged from 0.01 to 0.35, with correlation coefficients ranging from 0.40 to 0.93. Differences in the percent adherent metric ranged from −2.3 to 23.2, and kappa values from 0.22 to 0.89. CONCLUSIONS: Applying DDD estimates for the purposes of diabetes adherence estimation when lacking days supply values may provide reasonable estimates of adherence based on results presented here. At the medication class level there is greater variability in the reliability measures. Including claims from the U.S. only is a limitation of this analysis, as local treatment patterns may vary, and DDD values were not available for all U.S. medications. PDB94 MARKOV AND MONTE-CARLO MODELS IN THE PROGRESSION OF DIABETES MELLITUS: A LITTERATURE REVIEW TO IDENTIIFY THE FACTORS INFLUENCING THE CHOICE OF THE TYPE OF MODEL Renard L1, Borget I2 1 CRP-Santé, Strassen, Luxembourg; 2Institut Gustave Roussy, Villejuif, France OBJECTIVES: Markov and Monte-Carlo (MC) models are often used to simulate Diabetes Mellitus (DM) and its complications over time, but reasons to choose the model type are poorly documented. a systematic literature review was performed to identify factors influencing the choice of the model type. METHODS: Models simulating the progression of DM and its complications were selected from Medline and Embase databases. Literature reviews, methodological articles and non original models were excluded. Each full-paper selected went through a 31-item checklist via a doublereading process. a qualitative analysis was performed to evaluate the accuracy of the model with the study question. RESULTS: Sixty-one models were selected, including A302 42 Markov and 19 MC. Median time since publication was 6 years. Most models concerned Type 2 DM (57%) and prevention or disease-management programs (54%). In average, models had 11 states and 2.8 complications. MC were significantly more used than Markov when both type 1 and 2 DM and a prevention or disease management program were concerned. Models had, significantly, a higher number of states and complications. The choice of the model type was shown to be influenced by the number of DM and of complications considered. The justification to perform a Monte-Carlo was more often documented than those to perform a Markov (74% vs. 38%). The model was considered accurate with the study question in respectively 86% of Monte-Carlo and 64% of Markov. CONCLUSIONS: This study allowed to statistically identify factors influencing the type of model used to simulate DM. It is an interesting tool for modelers in their decision process to build a model. Information such as data and time availability or financial context could not be collected. PDB95 SIMPLIFICATION OF PATIENT LEVEL SIMULATIONS TO COHORT MODEL FOR SCENARIO ANALYSIS de Nigris E, Graham Roberts GR Double Helix Consulting Group, London, UK OBJECTIVES: In the last 15 years, models built to appraise health technologies have grown in complexity to accurately reflect the natural history of disease and calculate costs and benefits accordingly. The advantage of using patient level simulation models (PSM) over cohort models (CM) is that CM may become unwieldy with thousands of branches. This usually happens if the patient characteristics influence the risks of complications and these vary over time, or if there are many comorbidities to take into account and the risk of each complication is time dependent. In this case, a typical Markov cohort may be unsuitable. However a PSM can take a long time to run in order to produce results. Aim of this research is to present a case study where a PSM is simplified with a Markov structure and to compare the results of the two models. METHODS: We describe how a PSM for diabetes may be adapted to a CM, stating all the limitations. RESULTS: Provided that the PSM gives similar answers to the CM, the CM may be used as a surrogate for conducting complex sensitivity analysis (e.g., 3-way analysis or tornado diagrams). The advantage is that this model would produce results “instantaneously.” CONCLUSIONS: The adaptation of a PSM model to a cohort model may be a desirable feature if the model is required by a non technical audience. This is the case for “due diligence” models that are increasingly required by investors to assess the value of assets that a pharmaceutical company considers to buy or sell. Further research is needed to have a powerful test to assess whether the difference in results between a simplified CM model and the PSM are statistically significant. MUSCULAR-SKELETAL DISORDERS – Clinical Outcomes Studies PMS1 HIGHER COMORBIDITIES IN PATIENTS WITH GOUT VS. WITHOUT GOUT IN THE US GENERAL POPULATION: THE NATIONAL HEALTH AND NUTRITION EXAMINATION SURVEY (NHANES) 2007–2008 Pandya BJ1, Zhu Y2, Choi H2 1 Takeda Pharmaceuticals International, Inc., Deerfield, IL, USA; 2Boston University of School of Medicine, Boston, MA, USA OBJECTIVES: While the comorbidity burden of gout in the US has been considered substantial and may have been rising over the past decade, no contemporary national data are available. We estimated the prevalence of major comorbidities in patients with gout compared with those without gout based on a recent, nationally representative sample of US men and women (National Health and Nutrition Examination Survey [NHANES] 2007–2008). METHODS: Using data from 5707 participants in NHANES 2007–2008 (2797 men and 2910 women) aged 20 years and older, we determined the prevalence of major comorbidities among individuals with gout, including hypertension, renal impairment, nephrolithiasis, diabetes, myocardial infarction, heart failure, stroke, and obesity. We also compared the prevalence with those without gout using age- and sex-adjusted logistic regressions. Case definitions of comorbidities were based on an affirmative answer to a question asking if a physician or a health professional had diagnosed the corresponding condition. RESULTS: Among US adults with gout, 74% had hypertension, 53% obesity, 26% diabetes, 24% nephrolithiasis, 14% myocardial infarction, 11% heart failure, 10% stroke, and 9% renal impairment (Table). Prevalence of these comorbidities among individuals with gout was substantially higher than among individuals without gout. Age- and sexadjusted odds ratios (95% confidence interval [CI]) were 4.19 (2.75–6.39) for hypertension, 2.35 (1–55–3.57) for obesity, 2.36 (1.49–3.73) for diabetes, 2.10 (1.39–3.18) for nephrolithiasis, 2.37 (1.54–3.65) for myocardial infarction, 2.68 (1.88–3.83) for heart failure, 2.02 (0.98 to 4.19) for stroke, and 3.50 (2.05–5.98) for renal impairment. CONCLUSIONS: These findings from the latest nationally representative sample of US adults in NHANES 2007–2008 confirm that the prevalence of comorbidities among individuals with gout is substantial and considerably higher than among individuals without gout. 13th Euro Abstracts PMS2 DISABILITY OUTCOMES AND DOSE ESCALATION IN RHEUMATOID ARTHRITIS PATIENTS TREATED WITH TUMOR NECROSIS FACTOR BLOCKERS: A COMPARATIVE EFFECTIVENESS ANALYSIS Schabert VF1, Bruce B2, Ferrufino CP1, Fries JF2, Globe D3, Harrison DJ3 1 IMS Health, Falls Church, VA, USA; 2Stanford University, Division of Immunology & Rheumatology, Palo Alto, CA, USA; 3Amgen Inc., Thousand Oaks, CA, USA OBJECTIVES: Previous cohort studies using US commercial insurance claims of rheumatoid arthritis (RA) patients have observed higher dose escalation rates for patients treated with tumor necrosis factor-α (TNF) blockers infliximab (INF) or adalimumab (ADA) than for etanercept (ETN). However, such databases cannot be used to determine whether dose escalation is associated with improved clinical outcomes. We compared functional disability and dosing history between patients treated with TNF blockers using data from a large, prospective observational registry (Arthritis, Rheumatism, and Aging Medical Information System; ARAMIS) of RA patients in the US. METHODS: ARAMIS enrolled adult physician-diagnosed RA patients. Patients who met following inclusion criteria were selected: treated with a TNF blocker for ≥9 months; had no biologics 6 months before index TNF treatment; and reported a Health Assessment Questionnaire Disability Index (HAQ-DI) at index and 9–15 months after index. Self-reported dosing information was validated by telephone follow-up, medical, billing, or retail pharmacy prescription records. Dose escalation was defined as dosing above first ADA/ETN or third INF dose. Multivariate models compared HAQ-DI change and dose escalation rates, controlling for variables that differed among INF, ADA and ETN patients at index. RESULTS: Approximately 351 patients (93 INF, 40 ADA, 218 ETN) met study criteria. Subjects were mostly female (83%), Caucasian (88%), and had mean disease duration of 18.9 years. HAQ-DI change scores at 9–15 months did not differ by treatment (−0.10, −0.08, and −0.12 points for INF, ADA, and ETN, respectively; P = 0.52). Dose increases were observed in 1.4% of ETN patients, as compared with 10.8% of INF (P < 0.001) and 12.5% of ADA patients (P = 0.004, both vs. ETN). HAQ-DI change was associated with pre-index HAQ-DI score and disease duration (both P < 0.05). CONCLUSIONS: This study showed dose escalation in fewer ETN than INF or ADA patients, but similar improvements in functional disability. PMS3 IMPROVEMENT IN EFFICACY AND SAFETY OUTCOMES AFTER CERVICAL ARTHROPLASTY VERSUS STANDARD ANTERIOR CERVICAL DISECTOMY AND FUSION (ACDF) SURGERIES: A META-ANALYSIS OF PUBLISHED RCTS Gill S1, Maetzel A2, Alvares L3, Birinyi-Strachan L3 1 Frenchay Hospital, Bristol, UK; 2Stratas Partners, Basel, ON, Switzerland; 3Medtronic International Trading Sàrl, Tolochenaz, VD, Switzerland BACKGROUND: Cervical arthroplasty has been used successfully worldwide for over a decade in patients with cervical degenerative disc disease (CDDD). Several RCTs have documented the clinical advantage of arthroplasty vs. fusion for this patient population. Synthesis of outcomes from published RCTs is needed to consolidate the evidence-base for arthroplasty in CDDD. OBJECTIVES: To examine current evidence of safety and efficacy in the use of arthroplasty to treat CDDD and conduct a pooled analysis of high level comparative outcomes for arthroplasty versus fusion. METHODS: A systematic search of the PubMed, EMBASE, MEDLINE, CRD York, and Cochrane Library databases, plus hand searching of grey literature was undertaken in January 2010. RESULTS: A total of 554 non-duplicate citations were retrieved. Only three studies of arthroplasty versus fusion with 2 years follow-up met all inclusion criteria and were of reasonable quality (according to GRADE criteria) to pool. a meta-analysis was conducted using RevMan5 software (Cochrane 2008). Persons undergoing single-level arthroplasty had significantly improved neurological success (RR 1.14, 95% CI [1.07, 1.21] P < 0.0001), Neck Disability Index (NDI) success (RR 1.13, 95% CI [1.05, 1.22] P = 0.001) and overall success (RR 1.22 95% CI [1.12, 1.33], P < 0.00001) at 2 years post surgery compared to fusion. Secondary surgeries (any revision, removal, or re-operation of the implant or supplemental fixation) after 2 years were also significantly lower with arthroplasty (13/621) compared to fusion (39/592), [0.32 95% CI [0.17, 0.59) P = 0.0003]. Total number of patients undergoing reoperations at adjacent levels was lower in arthroplasty-treated patients (9/621) vs. ACDF (17/592), however it was not statistically significant (P = 0.09). Given the small numbers of patients undergoing reoperations at adjacent levels, longer term results are needed to increase precision of this estimate. CONCLUSIONS: A formal pooled analysis of 2-year RCT data demonstrates clinically significant improvements in efficacy and safety outcomes in CDDD patients who have undergone arthroplasty versus standard fusion. PMS4 NUMBER NEEDED TO TREAT FOR PATIENTS TO REPORT BROAD RELIEF FROM THE BURDEN OF RHEUMATOID ARTHRITIS WHEN TREATED WITH CERTOLIZUMAB PEGOL PLUS METHOTREXATE Strand V1, van Vollenhoven R2, Kvien T3, Combe B4, Khanna D5, Burmester G6, Nikai E7, Coteur G7, Smolen JS8, Schiff MH9 1 Stanford University, Portola Valley, CA, USA; 2Karolinska University Hospital, Stockholm, Sweden; 3University of Oslo, Oslo, Norway; 4Hôpital Lapeyronie, Montpellier, France; 5 University of California, Los Angeles, CA, USA; 6Charité University Hospital, Berlin, Germany; 7UCB, Brussels, Belgium; 8Medical University of Vienna and Hietzing Hospital, Vienna, Austria; 9University of Colorado School of Medicine, Denver, CO, USA OBJECTIVES: To determine in patients with RA the number needed to treat (NNT) to achieve minimum clinically important differences (MCIDs) in multiple patient 13th Euro Abstracts reported outcomes (PROs) simultaneously following treatment with certolizumab pegol (CZP) 200 mg + methotrexate (MTX), in the RAPID 1 and RAPID 2 trials. METHODS: The proportions of patients reporting improvements ≥MCID in RAPID 1 (Week 24 and Week 52) and RAPID 2 (Week 24) were determined for the following PROs: arthritis pain (0–100 mm visual analogue scale [VAS], MCID ≥ 10 mm), fatigue (Fatigue Assessment Scale, 0–10 numeric rating scale, MCID ≥ 1 point), physical function (Health Assessment Questionnaire-Disability Index, MCID ≥ 0.22 points), patient’s global assessment of disease activity (PtGA, 0–100 mm VAS, MCID ≥ 10 mm), and HRQoL (SF-36 physical and mental component summaries [PCS, MCS], MCID ≥ 2.5 points). NNT to achieve improvements ≥MCID in at least 1, 2, 3, 4, 5, or 6 out of the 6 considered PROs simultaneously with CZP + MTX compared with PBO + MTX were also calculated. RAPID 1: NCT00152386; RAPID 2: NCT00160602. RESULTS: The NNT to report clinically meaningful improvements in up to 5/6 PROs following CZP + MTX treatment was very low; approximately 2–3 additional patients after 24 weeks (RAPID 1 and RAPID 2). The NNT remained similar at 52 weeks (RAPID 1). The NNT to report improvements in all 6 PROs was 5 additional patients. In patients who achieved MCIDs in at least 5 of 6 PROs by Week 52, clinically meaningful improvements were more likely to be reported in pain, fatigue, physical function and PtGA than SF-36 PCS or particularly MCS. CONCLUSIONS: Patients with active RA report meaningful and broad relief in PROs following treatment with CZP. The low NNTs indicate that few patients need to be treated with CZP + MTX to report relief from the multiple burdens of RA. PMS5 PRIMARY ANTI-TNF FAILURES EXPERIENCE SUPERIOR CLINICAL RESPONSES TO A SECOND ANTI-TNF AGENT THAN SECONDARY FAILURES: ANALYSIS OF THE ALBERTA RHEUMATOID ARTHRITIS BIOLOGICS REGISTRY Ohinmaa A1, Martin L2, Russell AS1, Barr S2, Sholter D1, Penney C2, Yan C3, Jacobs P3, Maksymowych W1 1 University of Alberta, Edmonton, AB, Canada; 2University of Calgary, Calgary, AB, Canada; 3 Institute of Health Economics, Edmonton, AB, Canada OBJECTIVES: A Provincial prospective observational cohort study observes consecutive Rheumatoid Arthritis (RA) patients starting anti-TNF therapy (infliximab, etanercept or adalimumab). We aimed to assess the impact of switching anti-TNF agents at different time points. METHODS: Data in the registry is collected at baseline, 3 months, and every 6 months thereafter up to 3 years on patients receiving anti-TNF agents. We grouped patients as infliximab or adalimumab starters, and etanercept starters. Health-related quality of life was measured with the EQ-5D and single index scores were calculated using US value sets. Clinical outcome measures assessed were the HAQ and DAS28 scores. We analyzed the mean EQ-5D index score, the HAQ score, and the DAS28 score over time within baseline starter groups according to the time of switch of treatment. RESULTS: From 1222 patients in the registry, 649 patients had 27 months follow-up assessment and 76.7% remained on the first antiTNF. For these patients, significant improvement was observed by 3 months in the EQ-5D (mean change 0.296, P < 0.001), the HAQ (mean change −0.822, P < 0.001) and DAS28 (mean change −2.84 P < 0.001) that was sustained over the follow-up period. Those who switched at 3 months (primary failures) had significantly responded by the next measurement and the 27 months changes in outcomes were comparable to non-switchers (all P < 0.01). Patients who failed the first anti-TNF after the 3-month time point (secondary failures) obtained from 50% (HAQ) to 68% (EQ-5D) of the mean change compared to non-switchers (P < 0.05). Switching from Infliximab or adalimumab to etanercept or vice versa produced similar outcomes. CONCLUSIONS: The results show that primary failures to anti-TNF show similar responses to patients responding to their first anti-TNF agent. Secondary failures obtain about two-third of the outcomes compares to no-switchers and primary failures. Longer follow up is needed to see if the secondary failure patients can sustain the obtained improvement. PMS6 COMPARISON OF RECENTLY REGISTERED BIOLOGICAL DRUGS WITH AVAILABLE THERAPIES IN RHEUMATOID ARTHRITIS: METHODOLOGICAL ISSUES TO CONSIDER FOR META-ANALYSIS Péntek M1, Gulácsi L1, Érsek K1, Baji P1, Boncz I2, Orlewska E3, Brodszky V1 1 Corvinus University of Budapest, Budapest, Hungary; 2University of Pécs, Pécs, Hungary; 3 Centre for Pharmacoeconomics, Warsaw, Poland OBJECTIVES: Rheumatoid arthritis (RA) is a chronic, progressive, disabling autoimmune disease. The first biologic drug (etanercept) for the treatment of RA was registered in 1998, the latest in 2009 (EMEA). Study design of randomized controlled trials (RCT) of RA has changed in the past years, specially in point of reducing the time on placebo. Our aim was to compare the efficacy, safety and tolerability of available biologic drugs and to review the challenges of comparing RA studies with different structures, taking the example of the lately registered drug certolizumab pegol. METHODS: Systematic literature search was performed (closed on 16.08.2009) for RCTs with minimum duration of 24 weeks, involving drugs that have been registered for the treatment of RA after traditional disease modifying drug failure. Improvement of symptoms by 20%, 50% and 70% (defined by the American College of Rheumatology, ACR), adverse events and withdrawals were considered as endpoints in the metaanalysis. RESULTS: Altogether 14 RCTs (adalimumab 3, certolizumab pegol 2, etanercept 2, infliximab 4, tocilizumab 2, golimumab 1) involving 6739 patients were selected for the analysis. Efficacy of certolizumab pegol was superior to other drugs at ACR20, 50 and 70 but heterogenity was observed at certain safety endpoints. A303 Contrary to previous trials, patients on placebo not achieving ACR20 response at week 16 were withrawn in the certolizumab pegol studies. Thus patient number in the placebo group decreased dramatically for the remaining 8 weeks. To exclude this bias, we compared only the active arms and there was no significant difference in terms of safety and tolerability among the drugs. Analysis by patients years confirmed this finding. CONCLUSIONS: Effectiveness and safety of recently registered biologic drugs in RA are similar to the previous agents. Differences in protocol of former and present RCTs should be carefully considered for the analysis. PMS7 DETERMINANTS OF TNF INHIBITOR DOSE ESCALATION IN PATIENTS WITH RHEUMATOID ARTHRITIS Baser O1, Gust C2, Wang L2, Xie L2 1 STATinMED Research/University of Michigan, Ann Arbor, MI, USA; 2STATinMED Research, Ann Arbor, MI, USA OBJECTIVES: Switching among tumor necrosis factor-α inhibitor (TNFi) agents or to another biologic is relatively common, but determinants of dose escalation are largely unknown. Among individuals who are escalating the dose from their first TNFi, we identified factors associated with the choice of dose escalation in a retrospective analysis of medical and pharmacy claims and eligibility data. METHODS: Eligible patients were ≥18 years of age, diagnosed with rheumatoid arthritis (RA, 2 diagnoses ≥2 months apart) from January 2003–March 2008, had initiated a new TNFi prescription after a 6-month biologic-free period and had at least 6 months of continuous enrollment. Kaplan-Meier (KM) analysis and Cox regression were used to analyze time to dose escalation. Multinomial logistic regression was used to determine factors affecting dose escalation. RESULTS: A total of 11,903 (6.9%) of 173,533 RA patients were identified. Among these patients, 16% (n = 1,903) after a mean of 261 days escalated their dose. Comorbidity scores, such as the severity index for rheumatoid arthritis (SIFRA) and Elixhauser, were higher for patients who escalated their dose. The likelihood of escalating the dose was increased by female sex, younger age, and baseline use of corticosteroids or cytotoxic agents. Baseline use of methotrexate raised the likelihood of dose escalation (relative risk ratio, 1.58). CONCLUSIONS: After initiating TNFi treatment, many RA patients failed to remain on therapy and escalated doses or switched to a second TNFi or another biologic ~1 year. Factors influencing whether patients increased their dose included gender, age, and use of certain agents at initiation of the TNFi. a limitation of this study is that patients with <6 months of TNFi treatment were not included. PMS8 OCCURRENCE OF SECUNDARY HIP AND FEMUR FRACTURES ACCORDING TO RISK FACTORS FOLLOW-UP 8 YEARS Sebestyén A1, Lipp S1, Gresz M2, Varga S3, Oláh A3, Boncz I3 1 National Health Insurance Fund Administration, Pécs, Hungary; 2National Health Insurance Fund Administration, Budapest, Hungary; 3University of Pécs, Pécs, Hungary OBJECTIVES: The aim of this study is to evaluate the occurrences of secundary hip and femur fractures according to its different risk factors with follow up 8 years. METHODS: In this retrospective study the data derive from the financial database of the Hungarian National Health Insurance Fund Administration, based on the 10th revision of the International Classification of Diseases (ICD) with ICD code S7200. The patients included into the study had femur neck fracture and being discharged from the hospital after the primary treatment in 2000. The patients with polytrauma were excluded from the study. We evaluated data according to sex, age, type of fracture, type of surgery, local complications, type of residence. RESULTS: A total of 3783 patients met selection criteria of primary femoral neck fracture. During the postoperative 8 years 347 patents (9,2%) had secondary hip and femur fractures. Average age of patient with primary hip fracture 78 years, with secundary hip fracture 80,6 years. We demonstrated the following incidence rate of secundary hip fractures according to risk factors: Sex: female: 10,3%, male: 5,8%. Age groups: 60–69 y: 5,3%, 70–79 y: 7,5%, 80–89 y: 12,4%, 90 y-: 14,1%. Type of fracture: lateral: 6,9%, medial 9,5%. Type of surgery: osteosynthesis: 8,5%, arthroplasty: 13,6%. Local complications: yes: 10,6% no: 9%. Type of residence: capital 11,5%, village: 7,9%, city: 10%, town: 8,2%. CONCLUSIONS: The secondary hip fracture rate increased with the age of patients. The secondary hip fracture rate was the highest in female, in patient with primary intracapsular femoral neck fracture, after arthroplasty, in patient with local complication, in patient with capital residence. Many other factors can influence this secondary hip fractur, which will be analyzed in our further studies. PMS9 TREND OF HIP FRACTURE INCIDENCE IN BELGIUM BETWEEN 2000 AND 2007 AND FUTURE PROJECTIONS Hiligsmann M1, Bruyère O1, Detilleux J1, Gillet P1, Parmentier Y2, Dercq JP2, Carton J3, Reginster JY1 1 University of Liège, Liège, Belgium; 2INAMI, Brussels, Belgium; 3FPS Public Health, Brussels, Belgium OBJECTIVES: The primary aim of this study was to assess the incidence of hip fractures in Belgium between 2000 and 2007 and to examine secular changes within this period. a secondary aim was to estimate the expected number of hip fractures in Belgium until 2050. METHODS: The incidence of hip fractures was determined using the national database of hospital bills, which fully cover the annual hospital stays in the whole of the country. Population data and projections were derived from official sources. Logistic regression including year and age classes was performed for both A304 genders to assess the secular change between 2000–2002 and 2005–2007. Hip fracture projections were made until 2050, assuming the gender-specific secular change observed between 2000–2007 will be maintained. RESULTS: A total of 113,101 hip fractures were recorded in Belgium between 2000 and 2007, of which 76.4% occured in women. The annual number of hip fractures increased from 13,512 in 2000 to 14,744 in 2007, with a more marked increase in men (+20.4%) than in women (+5.7%). Between 2000–2002 and 2005–2007, the age-adjusted incidence of hip fractures significantly decreased by 1.08% per year (95% CI: 0.78 to 1.38) in women, but declined non-significantly by 0.30% per year (95% CI: −0.24 to 0.84) in men. The female/male ratio of hip fractures decreased between these periods from 3.19 to 2.92 (P < 0.01). By the year 2050, the number of hip fractures is expected to increase by 37.8% in women and by 110.0% in men. The female/male ratio would decrease to 1.81. CONCLUSIONS: Despite a significant decrease in the age-adjusted incidence of hip fractures in Belgian women and a non-significant decline in men, the number of hip fractures is expected to substantially increase in Belgium. Appropriate public health strategies are therefore needed and should also focus on populations of men. PMS10 PREVALENCE OF GOUT IN THE US GENERAL POPULATION: THE NATIONAL HEALTH AND NUTRITION EXAMINATION SURVEY (NHANES) 2007–2008 Zhu Y1, Pandya BJ2, Choi H1 1 Boston University of School of Medicine, Boston, MA, USA; 2Takeda Pharmaceuticals International, Inc., Deerfield, IL, USA OBJECTIVES: While the disease burden of gout in the US has been considered substantial and may have been rising over the past decade, no contemporary national data are available. We estimated the prevalence of gout based on a recent, nationally representative sample of US men and women (National Health and Nutrition Examination Survey [NHANES] 2007–8). METHODS: Using data from 5707 participants in NHANES 2007–8 (2797 men and 2910 women) aged ≥20 years, we estimated overall, gender-specific, and age-specific prevalence of gout. During the home interview of NHANES, all participants were asked: “Has a doctor or other health professional ever told you that you had gout?” We estimated the number of individuals with gout by applying our prevalence estimates to the corresponding 2008 US population estimate of 214.8 million adults over age 18. RESULTS: Overall prevalence of gout among US adults was 3.9%, which corresponded to an estimated 8.3 million individuals with gout. Prevalence of gout was 5.9% (6.1 million) among men and 2.0% (2.2 million) among women. Prevalence of gout increased with age, with the lowest (0.4%, 0.2 million) in those aged 20–29 years and the highest (12.6%, 1.2 million) in those aged ≥80 years. Prevalence of gout among those aged <64 years was 2.7% and was higher (9.8%) among those aged ≥65 years. Those ≥65 years corresponded to an estimated 3.5 million adults with gout. CONCLUSIONS: These findings from the latest nationally representative sample of US adults in NHANES 2007–2008 suggest that the prevalence of gout is substantial, particularly among older (aged ≥65 years) individuals. Additional confirmation studies are needed. PMS11 LONGITUDINAL ASSESSMENT OF GOUT IN A LARGE SAMPLE OF ITALIAN PATIENTS: PREVALENCE AND DISEASE MANAGEMENT IN GPS OFFICES Heiman F, Katz PM BKL Consulting SRL, Milan, Italy OBJECTIVES: The main goals of this retrospective study were to analyze the Prevalence of Gout in GPs offices in Italy and to describe the management of gouty patients by the GPs in terms of types and number of Specialistic visits requests during 12 months of follow-up. METHODS: Data were obtained from CSD LPD, a General Practitioner’s longitudinal database. Patients with at least one diagnosis of Gout from January 2008 to December 2008 have been selected and followed-up for 12 months. RESULTS: From 1.126.847 active patients, 17.241 had at least one diagnosis of gout, resulting on a prevalence of 1.53% in the general population. Among the 17.241 patients only 1.42% had received a specialistic visit request related to gout; of these the 42.8% were Reumatology visit requests, the 20.4% Ortopedics visits, the 8.1% Nephrology visits. Comparing the group of patients with at least one visit request related to gout (244) to the rest of the patients (16.997) in terms of specialistic visits request (related to all kind of diagnosis) during follow-up period, it could be seen that the group of 244 patients had received in average 4.5 specialistic visits request while the rest of the patients had received 2.3 visit requests. CONCLUSIONS: During the study we have identified a group of patients (244 patients 1,4% of the total) with at least one specialistic visit request related to the diagnosis of Gout, and we assumed that this group represents “Complicated patients” , maybe refractory to the therapy, that GPs need to manage in cooperation with specialist physicians. On support of our hypothesis the average number of specialistic visits (related to all kind of diagnosis) among this group is significatively higher in comparison with the rest of the patients (4.5 vs. 2.3). PMS12 PROJECTION OF SURGICAL LOADS OF HIP AND KNEE 1 2 Gaiser S , Kapoun M 1 Heraeus Medical GmbH, Wehrheim, Germany; 2Hochschule Niederrhein, Krefeld, Germany OBJECTIVES: Recent trends in the number of primary total hip arthroplasty (THA) and total knee arthroplasty (TKA) in the US has suggested a massive future demand for THA and TKA. It is unknown if this trend is the same in France as France has a 13th Euro Abstracts slower growth rates in obesity compared to US. The purpose of our study was to describe the recent trend of primary THA and TKA in France. METHODS: Registry data on THA and TKA, collected between 2000 and 2006, were obtained from the Agence technique de l’information sur l’hospitalisation (ATIF 2009). RESULTS: In 2006, approximately 139,000 hip replacements and 66,000 knee replacements were performed in France. Hip procedures increased by 6% and knee procedures by 30% since 2001. The economic burden associated with joint replacements is estimated approx. 2 billion Euro in 2006. The mean cost for primary hip replacement in France was c7,467 in 2006 and c8,690 for knee replacement procedures. Arthroplasties represent the medical specialty with the highest expenses in France. In France, only 41% of knee replacements were cemented. This proportion is relatively low in comparison to others European countries. CONCLUSIONS: This trend could accelerate in the near future due to changes in French demographics: increase of prevalence of overweight and obese patients as well as population ageing. These two categories of population have the highest incidence of joint replacement procedures. Obesity prevalence increases by 5% every year since 1997 to reach 12.4% of the total population in 2006 and the elderly population will reach 30% of the French population by 2050. In the context of the current financial crisis reducing unnecessary health care expenditure is inevitable. In the French market implant price represent 30% of hip replacement procedure and 38% for knee replacement. PMS13 RECENT TRENDS IN NUMBERS OF PRIMARY HIP/KNEE ARTHROPLASTIES PERFORMED IN GERMANY AND THE UNITED KINGDOM Gaiser S1, Kim S2 1 Heraeus Medical GmbH, Wehrheim, Germany; 2University of California, Davis, Sacramento, CA, USA OBJECTIVES: The number of primary total hip arthroplasties (THAs) and total knee arthroplasties (TKAs) continues to increase steeply in the US. It is unknown if this trend is the same in Germany and the United Kingdom (UK) as these populations age and become heavier. The purpose of our study was to investigate recent trends of primary THA and TKA in Germany and the UK. METHODS: We analyzed data from the German Bundesgeschäftsstelle für Qualitätssicherung and the UK’s National Joint Register. RESULTS: Between 2004 and 2008 in Germany, THAs increased 14% and TKAs increased 32%. During the same time in the UK, there were 60% and 70% increases in THAs and TKAs, respectively. In the UK, more TKAs than THAs have been performed since 2005, and a similar trend is expected in Germany after 2010. Nearly 20% of TKAs in Germany were performed on non-elderly patients (64 or younger) and patient age distributions remained approximately the same over the study period. In the UK, however, approximately 30% of TKAs were on non-elderly patients and the proportion of non-elderly patients that underwent TKA has increased every year. CONCLUSIONS: The number of arthroplasties in the US has increased dramatically among non-elderly adults. We observed no increase in this patient group in Germany but a slow and gradual increase in the UK. As obesity has risen over the last decades, the US has experienced greater inclines in TKAs than THAs; the annual number of TKAs is now twice that of THAs. Although the number of TKAs and the number of THAs had been approximately the same in Germany and the UK, recent trends suggest a growing gap due to a steeper rise in TKAs. Obesity, a strong risk factor for TKA, may be a bigger contributor than aging to the recent growth in joint arthroplasty. PMS14 RECURRENT FRACTURES AFTER FIRST HIP FRACTURES AND PREVENTION OF SECOND HIP FRACTURES IN JAPANESE WOMEN Yamabe K1, Hagino H1, Sawaguchi T2, Endo N3, Nakano T4, Watanabe Y5, Abe M1, Ito Y1 1 Tottori University, Yonago, Japan; 2Toyama City Hospital, Toyama, Japan; 3Niigata University Graduate School of Medicine and Dental Science, Niigata, Japan; 4Tamana Chuo Hospital, Tamana, Japan; 5Teikyo University School of Medicine, Tokyo, Japan OBJECTIVES: The aim of this study was to elucidate the incidence of recurrent fractures and the current status of osteoporosis treatment within one year after the first hip fracture. METHODS: This study was conducted for all female patients who were 65 years or older who had experienced hip fracture due to trauma for the first time and were treated within five geographic areas in Japan during the study period (January 2006 to December 2007). The data on demographics, treatments, and health outcomes were collected from medical records. a patient questionnaire was mailed to every patient about the health outcomes. RESULTS: The analysis was conducted in 2330 patients (average age 83.6 years). During the one-year observational period 158 fractures occurred in 152 patients. The incidence for all fractures among patients with first hip fracture was 71.7 (per 1000 person-year), and that for hip fracture was 36.7. The relative risk of recurrent hip fracture was 4.4 for women aged 65 years and over with first hip fracture. During this one-year period, anti-osteoporosis pharmacotherapy was performed in 437 patients (18.8%) while 1240 patients (53.2%) received no treatment. CONCLUSIONS: This study demonstrated the high risk of recurrent fracture in patients with first hip fracture and inadequate treatment for fracture prevention after first hip fracture. Since hip fracture patients are the most plausible candidates in the prevention of recurrent fractures, especially second hip fractures with high burden, appropriate treatment of osteoporosis is essential. 13th Euro Abstracts PMS15 FREQUENCY AND RISK FACTORS OF GOUT FLARES IN A LARGE POPULATION-BASED COHORT OF INCIDENT GOUT Rothenbacher D1, Primatesta P1, Ferreira A1, Cea Soriano L2, Garcia LA2 1 Novartis Pharma AG, Basel, Switzerland; 2CEIFE—Centro Español de Investigación Farmacoepidemiológica, Madrid, Spain OBJECTIVES: To describe the frequency and risk factors of gout flares in a large cohort of incident gout. METHODS: A cohort study was conducted in an UK general practice database (The Health Improvement network (THIN)) in the study period January 2000 through December 2008. All patients aged 20–89 years diagnosed with incident gout between the years 2000–2007 were included; their history of flares recorded over time. We studied the association between several factors and the number of flares. a Cox proportional hazards model was used to identify determinants of risk of first flare. RESULTS: The overall incidence rate of gout was 2.68 (95% confidence interval (CI) 2.65–2.72) per 1,000 person-years, and 23,857 incident gout patients (mean age 61.9 years) were included in this study. The incidence rate of a first flare was 13.7 (95% CI 13.4–14.0) per 100 person-years. The proportion of patients with at least one flare during the follow-up period (mean 3.8 years) was 36.9% (n = 8,806). Male gender, alcohol consumption and history of cardiometabolic diseases were positively associated with number of flares (P < 0.001). The hazard ratio estimate of first flare associated with antecedents of renal failure disorders was 1.33 (CI 95% 1.20– 1.48), 1.41 (CI 95% 1.26–1.58) for drinking more than 42 alcohol units per week, and 1.22 (CI 95% 1.14–1.30) for a body mass index greater than 30 kg/m2. Also, a higher risk of a first flare was associated with a history of ischemic heart disease, and hypertension (HR: 1.12 (95% CI 1.06–1.19) and 1.15 (95% CI 1.10-1-20), respectively). CONCLUSIONS: Gout flares are the most common clinical manifestation in patients suffering from gout disease. Male sex, modifiable life-style factors as well as comorbid conditions are important independent risk factors for first flare. PMS16 EARLY MORTALITY RATE OF BILATERAL HIP FRACTURES OVER 60 IN HUNGARY Sebestyen A1, Lipp S1, Boncz I2 1 Baranya County Health Insurance Fund, Pécs, Hungary; 2University of Pécs, Pécs, Hungary OBJECTIVES: The aim of the study is to analyze the early mortality rate after bilateral hip and femur fractures over 60 in Hungary. METHODS: In this retrospective study the data derive from the financial database of the National Health Insurance Fund of Hungary. For the analysis we used code S7200 of the International Classification of Diseases (ICD) tenth revision. The patients included into the study had femoral neck fracture and being discharged from the hospital after the primary treatment in 2000. The patients with polytrauma were excluded from the study. The follow up was 8 years. The patients were categorized according to sex, age, fracture type, surgical methods, presence of accompanying diseases and day of hospital admission. We analyzed the early mortality rate within 30 days after treatment of bilateral hip and femur fracture. RESULTS: Altogether 347 patients were included into the study. The national average of mortality was 7.5%. We demonstrated the following early mortality rate according to risk factors: Sex: male: 13.2%, female: 6.5%. Age groups: 60–69 y: 5.4%, 70–79 y: 5.9%, 80 y-: 8.9%. Type of fractures: femoral neck fractures: 4.2%, pertrochanteric fractures: 15.5%, femur fractures: 5.9%. Surgical techniques: arthroplasty: 3.1%, osteosynthesis: 7.8%. Accompanying diseases: yes: 8.1%, no: 2.6%. Day of hospital admission: weekday: 7.4%, weekend: 7.7%. CONCLUSIONS: The early mortality rate increased with the age of patients. In the analysis according to sex, type of fracture and presence of accompanying diseases we found higher mortality in male, in patient with pertrochanteric fracture and in patient with accompanying diseases. According to surgical intervention early mortality rate is higher in patient with osteosynthesis. Many other factors can influence this mortality, which will be analyzed in our further studies. MUSCULAR-SKELETAL DISORDERS – Cost Studies PMS17 BUDGET IMPACT OF NEW RHBMP-2 FORMULATION IN PATIENTS UNDERGOING POSTEROLATERAL SPINAL FUSION PROCEDURES FOR DEGENERATIVE DISC DISEASE IN RANDOMIZED CONTROLLED TRIAL (RCT) McInnis MM1, Olchanski N1, Kemner JE2, Goss T1 1 Boston Healthcare Associates, Inc., Boston, MA, USA; 2Medtronic Spinal and Biologics, Memphis, TN, USA OBJECTIVES: A new formulation of recombinant human bone morphogenetic protein-2 (rhBMP-2) on a Compression Resistant Matrix (CRM) evaluated in posterolateral spinal fusion (PLF) procedures for degenerative disc disease (DDD) has the potential to improve clinical outcomes and decrease costs. The rhBMP-2/CRM pivotal study demonstrated significant improvement in spinal fusion rates compared to procedures using iliac crest bone graft (ICBG), (Dimar 2009). This study evaluated the likely budget impact of rhBMP-2/CRM on U.S. payer costs compared to ICBG procedures in PLF. METHODS: A budget impact model was developed to quantify clinical and economic outcomes for patients with single-level symptomatic DDD and the U.S. payer costs of rhBMP-2/CRM in PLF procedures. The model results compare use of rhBMP-2/CRM to the use of ICBG in these procedures. Both study arms use posterior supplemental fixation. The 24-month clinical data was derived from 463 A305 patients in the rhBMP-2/CRM pivotal study trial and included independent radiographic assessment of fusion success rates, re-surgery rates, and complications rates (infections, bleeding, cardiovascular complications, GI complications). Payer costs and clinical practice data were derived from published sources and from the 2007 Medicare Inpatient Standard Analytic File and inflated to 2010 costs using the medical care component of the U.S. Consumer Price Index. RESULTS: The average 2-year costs per patient treated with ICBG were $53,740. Comparatively, treating patients with rhBMP-2/CRM is estimated to reduce average costs by 12% ($6,252) per patient. Cost reductions were primarily driven by a reduction in subsequent surgeries. Fusion rates at 2 years were greater in the rhBMP-2/CRM group compared to the ICBG group (96% vs. 89% at p = 0.014). CONCLUSIONS: This budget impact model based on RCT data demonstrates that the use of rhBMP-2/CRM in treatment of DDD patients undergoing PLF procedures is associated with substantial cost savings to payers when compared with traditional ICBG procedures. PMS18 A BUDGET IMPACT MODEL FOR THE ECONOMIC ASSESSMENT OF TOCILIZUMAB IN THE TREATMENT OF RHEUMATOID ARTHRITIS PATIENTS IN ITALY Iannazzo S1, Zaniolo O1, Giuliani G2 1 Adres srl, Torino, Italy; 2Roche S.p.A., Monza, Italy OBJECTIVES: A Budget Impact Model (BIM) was developed to assess the impact on Italian National Health Service (NHS) expenditure caused by the use of tocilizumab (TOC) , monoclonal antibody approved by EMA in early 2009, as a first-line biologic treatment vs. traditional anti-TNFα therapies in the treatment of moderate to severe rheumatoid arthritis (RA) patients. METHODS: A Markov model was used to simulate the progression of a cohort of RA patients through three lines of biologic treatments with anti-TNFα and TOC, palliative therapy and death. The horizon was 5 years. The target population was defined by the number of incident RA patients eligible to first-line biologic treatment based on Italian epidemiological and market data. First-line TOC followed by a second- and a third-line anti-TNFα treatment was compared a more traditional strategy based on anti-TNFα cycling. Direct medical costs were considered: drug acquisition, administration and monitoring tests, according to current prices and tariffs. To estimate the relative effectiveness of use of economic resources, the average number of patients that can be treated with an hypothetical annual budget of c1,000,000 was assessed. RESULTS: Italian RA patients eligible to a first-line biologic therapy were estimated in 750 per year. Firstline TOC strategy induced a decrease of costs of c805,000 (−8.6% with respect to anti-TNFα cycling), and c1,348,000 (−3.1%) at year 1 and year 5 respectively. The average annual number of patients that can be treated with c1,000,000 was 75.7 at year1 and 78.5 at year5 under the anti-TNFα cycling strategy, while it was 82.9 at year 1 and 81.8 at year 5 with first-line TOC, with a 9.5% to 4.3% increase. CONCLUSIONS: The introduction of TOC among the biologic therapies available for Italian RA patients represents a valuable option as it is expected to induce a reduction in costs and a more efficient use of health care resources. PMS19 THE INTRODUCTION OF A GERMAN ARTHROPLASTY REGISTER: COST SAVINGS FOR THE HEALTH CARE SYSTEM Held R1, Gaiser S2 1 Hochschule Niederrhein, Krefeld, Germany; 2Heraeus Medical GmbH, Wehrheim, Germany OBJECTIVES: Worldwide the numbers of first use—implantations and revisions in the hip—and knee arthroplasty are rising. In many countries corresponding arthroplasty—registers have been introduced in order to increase the quality of such operations. Arthroplasty—registers in countries like Sweden have been improving the practical operations and both realizing significant cost savings at national level by reducing the revision burden results. In Germany a similar arthroplasty—register has not yet been introduced. METHODS: This study examines the trends in the German health care system which influences the development of clinical numbers in the field of arthroplasty and makes forecasts to 2020 on the basis of the BQS—data. In a further step this study takes the experiences and the results of existing arthroplasty— registers especially from Sweden into account and makes use of it for the situation in Germany. The aim is to identify possible cost savings when introducing such a register in Germany. RESULTS: The introduction of a German arthroplasty—register for hips seems to make an important contribution to the quality of medical care. In addition considerable cost savings on a national basis might be realized by reducing the revision burden, which exceed the cost of an arthroplasty—register by far. CONCLUSIONS: If the revision burden for hip arthroplasty could be reduced by one percentage point cost savings of about c14–23 million per year could be realized. a reduction of about 50 per cent like in Sweden cost savings about c81–216 million per year could be expected. The results of this study might be a basis of discussion among policymakers. PMS20 COST AND QUALITY OF LIFE IN PATIENTS WITH RHEUMATOID ARTHRITIS IN SOUTH KOREA Lee TJ1, Park BH1, Sohn HG1, Song YW2, Shin K2 1 Seoul National University, Seoul, South Korea; 2Seoul National University Hospital, Seoul, South Korea OBJECTIVES: This study aims to estimate health care cost and productivity loss and to measure the quality of life in patients with rheumatoid arthritis (RA), particularly A306 by its functional class (FC) in the Korean health care setting. METHODS: RA-associated costs were estimated from a societal perspective. Health care cost was estimated using data from National Health Insurance (NHI) Claims Database between January 1 and December 31, 2008. In order to estimate other health care costs, like traditional medicine and physiotherapy, that are not covered under the NHI and non-health care costs including transportation, health food and housekeeping services, face-to-face interviews were conducted for 202 patients with RA in a large teaching hospital in Seoul for 3 months starting from November 18, 2009. RA-related productivity loss and quality of life using EQ-5D were also estimated from the survey. Factors associated with costs as well as quality of life in RA patients were indentified using multiple regression and logistic regression. RESULTS: Total annual cost that RA patients incurred in 2009 was estimated to be $1.6 billion, of which almost a half was attributable to productivity loss. As the function of RA patients deteriorates, annual per capita cost showed an increasing trend (FC I: $3534, FC II: $6057, FC III: $6722, and FC IV: $6856) while quality of life was lowered (FC I: 0.673, FC II: 0.502, FC III: 0.294, and FC IV: 0.227). Annual per capita cost increased with the deterioration of function (P < 0.05) and co-morbidity (P < 0.1). The quality of life was positively correlated with functional class (P < 0.05). CONCLUSIONS: RA patients not only incur high health care cost but also suffer from significant productivity loss and reduced quality of life in Korea. RA patients’ poor function is highly correlated with higher costs and lower quality of life. PMS21 THE ACUTE CARE COSTS OF OSTEOPOROSIS-RELATED FRACTURES IN CANADA Tarride JE1, Hopkins R1, Leslie WD2, Morin S3, Papaioannou A1, Adachi R1, Bischof M1, Goeree R1 1 McMaster University, Hamilton, ON, Canada; 2University of Manitoba, Winnipeg, MB, Canada; 3McGill University, Montreal, QC, Canada OBJECTIVES: Osteoporosis is a prevalent disease, characterized by low bone mass and increased fracture risk. In 1993 the economic burden of osteoporosis-related fractures in Canada was $1.3 billion dollars, of which $437 million was acute care. Our objective was to update the1993 acute care estimates for Canada for non-traumatic fractures. METHODS: We used national administrative databases from Canadian Institute for Health Information for fiscal-year ending March 31, 2008 for Canadians over age 50. Data sources included the Discharge Abstract Database for all acute care hospitalizations in Canada except Quebec; National Ambulatory Care Reporting System for all same day surgery, emergency visits/ambulatory services in hospital for Ontario. Both data sources were projected nationally using Statistics Canada census data. Non traumatic fractures included hip, humerus, vertebral, wrist, other sites (ribs/sternum, pelvis, trunk, clavicle, scapula, femur, patella, tibia/fibula), and multiple sites (more than 1 of the preceding). Costs were based on resource intensity weights for direct medical costs less physician billings. Physician fees billed directly to provincial plans included visits in hospital, diagnostic and surgical interventions. RESULTS: Osteoporotic non-traumatic fractures were responsible for 57,404 acute care admissions, 112,749 emergency room visits, and 3,433 same day surgeries in Canada in 2008. Hip fractures accounted for 50.3% of admissions, wrist fractures accounted for 80.8% of same day surgeries, and emergency visits mostly occurred in other sites (30.0%), wrist (29.7%) and hip fractures (22.9%). The cost of acute care was estimated at $1.2 billion with hip fractures alone costing $619 million. CONCLUSIONS: Osteoporosis is a chronic disease that affects a large segment of the adult population in Canada and increases the risk for expensive health care utilization for Canadians. Since 1993, the acute care cost of osteoporosis-related fractures has risen from $437 million to $1.2 billion dollars per year, which represents an increasing economic burden for Canada. PMS22 ECONOMIC BURDEN OF OSTEOPOROSIS-RELATED FRACTURE HOSPITALIZATIONS IN FRANCE Maravic M1, Vainchtock A2, Jouaneton B3, Tochon V4 1 Hôpital Léopold Bellan, Paris, France; 2HEVA, LYON, France; 3HEVA, Lyon, France; 4Amgen, Neuilly Sur Seine, France OBJECTIVES: To estimate the burden of post-menopausal osteoporosis-related fracture hospitalizations in France for acute and rehabilitation care. METHODS: Data were obtained from the 2008 French Hospital National Database for women aged ≥50 years. For acute care, criteria were established according to ICD-10 codes related to osteoporosis. As the rules of coding are not strictly followed in real practice, an additional database analysis was performed to include stays related to surgical management of hip fractures. Duplicate hospitalizations were excluded. Among women that were identified for acute care, we selected those who went in rehabilitation care related specifically to osteoporosis. We assessed number of hospitalizations and patients, proportion of surgical management, length of stay in acute care and number of rehabilitation day and costs. Hospital costs were calculated according 2009 national hospital tariff and 2000–2001 National Scale of Costs, respectively for acute and rehabilitation care (2009c). RESULTS: There were 67,807 acute hospitalizations (64,793 patients) and 1,359,863 days for rehabilitation care (31,458 patients) associated with osteoporosis-related fractures. In acute care, a total of 80% of the hospitalizations were associated with the surgical management of fractures and 89% of them were transferred in rehabilitation care. The mean (SD) length of stay and days of rehabilitation were 12 ± 8 days and 43 ± 31 days, respectively. The overall cost of osteoporosis-related fracture hospitalizations was 415,429,993c for acute care, of 13th Euro Abstracts which 6.5% was related to medical devices, and 331,755,797c for rehabilitation care. The mean (SD) hospital cost including medical devices, was 6127c ± 2352c per stay for acute care and 10,546c ± 7,869c per stay for rehabilitation care. CONCLUSIONS: In 2008, osteoporosis-related fracture hospitalizations were associated with a substantial economic burden in France. PMS23 BURDEN OF ILLNESS OF OSTEOPOROSIS IN AUSTRIA 1 2 3 4 5 6 Dimai H , Redlich K , Viernstein H , Siebert U , Schneider H , Mahlich J 1 Med Uni Graz, Graz, Austria; 2Med Uni Wien, Vienna, Austria; 3Uni Wien, Vienna, Austria; 4 UMIT—Univ. of Health Sciences, Medical Informatics & Technology, Hall, Austria; 5IWI, Vienna, Austria; 6Austrian Economic Chamber, Vienna, Austria OBJECTIVES: On the basis of current international prevalence, it can be estimated that approximately 740,000 of people in Austria over 50 years are affected by osteoporosis, of whom around 617,000 are women. With a fracture rate of 19.7 fractures per year per 10,000 inhabitants over the age of 65 years, Austria lies within the peak for Europe. Within the framework of a burden of illness study we examined the financial burden of osteoporosis in Austria. METHODS: We took both direct and indirect costs into consideration. Direct costs encompass medical costs such as expenses for pharmaceuticals, inpatient and outpatient medical care costs, as well as other medical services (e.g., occupational therapies). Non medical direct costs include transportation costs and medical devices (e.g., wheel chairs or crutches). Indirect costs refer to costs of productivity losses due to absence of work. Moreover, we included costs for early retirement and opportunity costs of informal care provided by family members. While there exist similar studies for other countries (e.g., Germany), this is the first comprehensive study for Austria. For our analysis, we combined data of both, official statistics, expert estimates as well as unique patient surveys that are currently conducted in the course of an international osteoporotic fracture study in Austria. RESULTS: Our estimation of the total annual costs in the year 2008 imposed by osteoporosis in Austria is c470 million. The largest fraction of this amount is incurred by acute hospital treatment. Another significant figure, accounting for 20% of total costs, is the opportunity cost of informal care. CONCLUSIONS: The financial burden of osteoporosis in Austria is substantial. Economic evaluations of preventive and therapeutic interventions for the specific context of Austria are needed to inform health policy decision makers. PMS24 OSTEOPOROSIS COSTS IN ITALY. INTERIM RESULTS OF THE BLOCK STUDY Veronesi C1, Adami S2, Iolascon G3, Nardi A4, Pietrogrande L5, Resmini G6, Rossini M2, Tarantino U7, Tranquilli Leali P8, Trevisan C9, Buda S1, Intorcia M10 1 CliCon S.r.l. Health, Economics & Outcomes Research, Ravenna, Italy; 2Rheumatology Unit, Dpt Medicine, University of Verona, Verona, Italy; 3Dpt. Orthopedics and Rehabilitation , Second University of Naples, Naples, Italy; 4SOS Dpt Patologia Osteoarticolare, Azienda Sanitaria ULSS 18, Rovigo, Italy; 5Università degli Studi Polo San Paolo, Milano, Italy; 6Centre for the Study of Osteoporosis and Metabolic Bone Disease, Section of Orthopaedic and, Treviglio (BG), Italy; 7Division of Orthopaedics and Traumatology, Tor Vergata Foundation University Hospital, University of Rome Tor Vergata, Rome, Italy; 8Orthopaedic Clinic, Sassari University, Sassari, Italy; 9UO di Ortopedia Ospedale di Lovere, AO Bolognini, Lovere (BG), Italy; 10Amgen (Europe) GmbH, Zug, Switzerland OBJECTIVES: Calculate the cost of osteoporosis in postmenopausal women after a fracture-related hospitalization for over a 24 months follow-up period. METHODS: Retrospective analysis of databases of 8 Italian Local Health Units (4 million beneficiaries). We included female patients (≥65 years) hospitalized for a typical osteoporotic fracture or osteoporosis between January 1, 2000 and December 31, 2005. Data were collected on hospitalizations, visits, medication and diagnostic tests in the 24 months after the first hospitalization. Over this period, treatment compliance measured as Medical Possession Ratio was calculated according to the proportion of days covered by osteoporosis-related drug treatments and classified as low (≤80%) and high (>80%). Relative unit costs for resource use were collected from DGRS, National Tariffs and Drugs National Formulary. RESULTS: A total of 10,158 patients were included (average age ± SD, 79.0 ± 7.5 years) (overall cost was c70.5 million for the first hospitalization). Of these, 98% (n = 9978) were hospitalized for a typical osteoporotic fracture and 180 (2%) for osteoporosis. Most were hip fractures (5189 (52%); 76% of total costs). During 24 months follow-up, 6654 patients (66%) had resource utilization data (overall cost was c15.0 million; average cost per patient was c2254). Among all 9978 women with fracture, 21% (n = 2119) had a subsequent hospitalization for fracture (66% were hip) and this accounted for 89% of the c15.0 million (11% other, diagnostic tests). Only 15% (n = 1470) of all evaluated patients received osteoporosis-related drug treatment within the six months following discharge and bisphosphonates were prescribed to 92.6% of them. Among bisphosphonate-treated patients, 33% (n = 459) and 66% (n = 921) displayed a high and low compliance level, respectively. CONCLUSIONS: In this database analysis, costs of osteoporosis after a first hospitalization for fracture were relevant and mostly due to re-hospitalization for a new typical osteoporotic fracture, mainly at the hip site. Exposure and compliance to osteoporosis treatment were suboptimal in this patient population. 13th Euro Abstracts PMS25 DIRECT MEDICAL COSTS OF OSTEOARTHRITIS IN SPAIN Darba J1, Restovic G2, Kaskens L2 1 Universitat de Barcelona, Barcelona, Spain; 2BCN Health, Barcelona, Spain OBJECTIVES: Osteoarthritis (OA) is the most common musculoskeletal disease, with around 1.6 million patients in Spain. The aim of this study was to determine health care resource utilization and direct medical costs of patients suffering OA of the hip, knee and wrist in Spain in 2009. METHODS: A cost-of-illness analyses was performed to estimate direct medical costs of patients suffering OA. Prevalence data on OA by the Spanish Society of Rheumatology showed a value of 18% for the Spanish population. a semi-structured questionnaire was sent to rheumatologists to collect data on health care resource utilization and costs. Inpatient costs were considered from the perspective of the public health care system. Direct medical costs included were ambulatory, diagnostic tests, physiotherapy, surgery, drug and administration costs. All costs referred to 2009. RESULTS: Patients diagnosed with OA of the hip, knee and wrist in Spain were estimated to be 1.6 million in 2009. Total medical costs for the Health Care System resulted as following: drug use 95c million, health care personnel 251c million, diagnostic tests 20c million, physiotherapy 93c million, surgery 105c million and adverse effects due to OA medication 25c million. Due to demographic changes and increasing prevalence, the number of patients is estimated to grow to 1.7 million with total costs of 717c million in 2014. CONCLUSIONS: Direct medical costs for OA were estimated at 598c million for Spanish patients in 2009. Health care personnel represented 43% of the total direct medical costs. Aging of the population, development of new drugs and increasing patient expectations are likely to increase the future economic impact of OA, which remains a major public health burden. PMS26 HIP FRACTURES IN THE ELDERLY: COST OF ILLNESS STUDY UNDER A PUBLIC HOSPITAL PERSPECTIVE IN RIO DE JANEIRO, BRAZIL Fernandes RA1, Takemoto ML2, Araujo D2, Sauberman MV3 1 Instituto Nacional de Ciência e Tecnologia para Avaliação de Tecnologias em Saúde (IATS), Rio de Janeiro, RJ, Brazil; 2State University of Rio de Janeiro, Rio de Janeiro, RJ, Brazil; 3 Hospital Municipal Lourenço Jorge, Rio de Janeiro, RJ, Brazil OBJECTIVES: To assess direct medical costs associated to hospital treatment of hip fractures in the elderly in a public hospital in Rio de Janeiro, Brazil and their association with demographic and clinical variables. METHODS: Observational, prospective study to assess resource utilization and direct medical costs associated to elderly hip fracture hospitalization in 2007 and 2008, under the health care provider perspective. a standard data collection instrument was used to register identified resources during prospective medical charts review. The resource utilization was converted into Brazilian Real (BRL), based on 2010 prices. Descriptive analysis of costs and resource utilization and their association with clinical and demographic variables were performed. RESULTS: Eighty two patients were included, 81.7% female, mean age of 76.96 years, hospitalization mean time of 12.66 days. Median total costs per patient were 3,064.76 BRL (IC95%: 2,817.63–3,463.98). Clinical hospitalization and surgical procedure were responsible for 65.61% and 24.94% of costs, respectively. Median costs for patients submitted to surgical procedure until the fourth day of hospitalization were lower than median costs for patients submitted after the fourth day (2,136.45 BRL and 3,281.45 BRL, respectively, P < 0.00001). A significant difference in average costs per type of surgical procedure was also observed. Variables sex, age over 80 years, fracture site and presence of cardiovascular disease were not associated with statistically significant differences in total costs. CONCLUSIONS: Clinical hospitalization and surgical procedure were the main cost components observed. Higher cost associated to inpatient treatment of hip fractures in patients who performed surgery after the fourth day of hospitalization added to available evidence about an increased risk of mortality after this period reinforce the need of priority establishment to treat elderly patients with hip fracture. PMS27 ANALYSIS OF DIRECT MEDICAL AND NON-MEDICAL COSTS FOR CARE OF RHEUMATOID ARTHRITIS PATIENTS USING LARGE COHORT DATABASE, IORRA Igarashi A1, Kikuta K1, Tanaka E2, Hoshi D2, Inoue E2, Seto Y2, Nakajima A2, Momohara S2, Taniguchi A2, Yamanaka H2, Tsutani K1 1 Tokyo University Faculty of Pharmacy, Tokyo, Japan; 2Tokyo Women’s Medical University, Tokyo, Japan OBJECTIVES: To examine annual direct medical & non-medical cost in large-scale rheumatoid arthritis (RA) patient cohort (IORRA) in Japan. METHODS: From patients’ perspective, we calculated direct medical (out-of-pocket costs to hospital & pharmacy and cost for complementary & alternative medicine(CAM)) and non-medical costs (caregiving, transportation,self help devises, house modification) of RA patients, participants of the 15–17th IORRA Studies in Oct. 2007- Oct. 2008. We also assessed correlations between these costs and RA diseaseactivity, disability level and QOL. RESULTS: Data from 5204 RA patients were extracted. Annual direct medical costs were JPY132,000 (out-of-pocket to hospital, USD1 = JPY90), JPY84,000 (out-of-pocket to pharmacy) and JPY 146,000 (CAM)), respectively. Annual direct non-medical costs were JPY105,000 (caregiving), JPY22,000 (transportation), JPY30,000 (self help devises) JPY188,000 (house modification), respectively. Considering utilization rates for each cost component (hospital/pharmacy: 100%, CAM: 31.6%, caregiving: 10.5%, transportation: 100%, self help devices: 21.4%, house modification: 21.4%). We assumed that annual medical/non-medical cost per RA patient was JPY264,000 and JPY61,000, respectively. These costs increased progres- A307 sively with worsening RA diseaseactivity, disability level, or QOL. For example, patients with lower Eq-5D score (less than 0.5) spent more money than those with higher one (more than 0.8). Average medical and non-medical costs among them were JPY 30,802 vs. JPY17,887 and JPY229,519 vs. JPY19,536, respectively. CONCLUSIONS: Heavy economic burden lies in RApatients and grows heavier as the disease state is exacerbated using IORRA database. The results also suggest that the increase in medical/non-medical cost may be suppressed by proactively controlling RA. PMS28 COST-EFFICACY ANALYSIS OF TNF ALPHA ANTAGONISTS IN THE TREATMENT OF RHEUMATOID ARTHRITIS Dominguez Gil-Hurlé A1, Costi Ruiz M2, Campo Sien C2 1 Hospital Clínico Universitario de Salamanca, Salamanca, Castilla y León, Spain; 2Abbott Laboratories, Madrid, Spain OBJECTIVES: Estimate the efficiency of Tumor Necrosis Factor α (TNFα) antagonists in moderate to severe rheumatoid arthritis (RA). METHODS: The analysis, performed from the Spanish Health Care System perspective, considers the annual cost of the drugs, and their efficacy, measured through the number needed to treat (NNT) to gain an additional patient who achieves ACR20, ACR50 and ACR70 response. Drug costs were obtained from a Spanish database. Data relative to efficacy was derived from a metaanalysis, which evaluated the anti–TNFα drugs adalimumab (ADA), etanercept (ETA), and infliximab (INF). Efficiency was estimated in terms of incremental cost-efficacy ratios (ICER). RESULTS: Annual treatment cost per patient-year with ADA, ETA and INF is of 13.116c, 12.314c, and 14.047c, respectively. Applying the: 1) ACR20 criteria, the NNT with ADA, ETA, and INF were 4.2 (95% CI, 3.4–5.3), 6.5 (5.2–8.8), and 4, 5(3.4–6.8), respectively; 2) the ACR50 criteria, the NNT with ADA, ETA, and INF were 4.1 (3.4–5.1), 4.4 (3.7–5.5), and 6.6 (4.7–11.5), respectively; and 3) the ACR70 criteria, the NNT with ADA, ETA, and INF were 5.7 (4.6–7.5), 6.8 (5.3–9.4), and 8.6 (5.8– 16.7), respectively. The incremental annual cost per additional patient who achieves ACR20 response with ADA, ETA, and INF is of c54.871 (95% CI, c45,080–70,095), c80,598 (c64,288–c107.999), and c63,329 (c47,268c–c95,923), respectively. The incremental annual cost per additional patient who achieves ACR50 response with ADA, ETA, and INF is of c53,368 (c44,407–c66,828), c54,596 (c45,691–c67,813), and c93,072 (c65,451–c161,030), respectively. The incremental annual cost per additional patient who achieves ACR70 response with ADA, ETA, and INF is of c74,537 (c59,759–c99,025), and c83,661 (c65,481–c115,815), c120,652 (c81,242–c34,319), respectively. CONCLUSIONS: The incremental cost per patient who achieves an ACR20, ACR50, and ACR70 response is lower with ADA, though quite similar to ETA, being with both (ADA and ETA) lower than with INF, in the Spanish setting. PMS29 RETROSPECTIVE CHART REVIEW TO ASSESS COSTS RELATED TO OSTEOPOROTIC FRACTURES IN SLOVENIA AND SERBIA Wladysiuk M1, Wilk D1, Jedynasty K2, Bumbasirevic M3, Kozlevcar Zivec M4 1 HTA Consulting, Krakow, Poland; 2Amgen GmbH, Headquarters Office for CEE, Vienna, Austria; 3Institute of Orthopedic Surgery and Traumatology at Clinical Center Serbia, Beograd, Serbia; 4Ambulanta za osteoporozo Medicus, Ljubljana, Slovenia OBJECTIVES: To evaluate direct medical costs of treatment for osteoporotic fractures in Slovenia and Serbia from a public payer and patient perspective directly after fracture and up to 1 year follow-up. METHODS: A medical chart review, examining medical resources used to treat the 3 most common osteoporotic fractures (proximal femur, vertebral and distal radius) in the first year after the event. Collection of data from 1 osteoporotic center in Slovenia and 3 in Serbia was carried out by local investigators between December 2009 and March 2010. The treatment costs for each fracture type from the public payer and patient perspective were calculated. The analysis was divided into 2 parts: intervention directly after the fracture (including cost of hospitalization, ambulatory visits, procedures, examinations, and medications) and follow-up for up to 1 year after the event (including costs of hospitalization, outpatient visits, examinations, rehabilitation, medications and devices). RESULTS: A total of 240 patients aged >50 years with low-trauma fractures occurring within 5 years before study initiation were included. Average annual costs of treatment of a proximal femur fracture in Slovenia were estimated at c4727 (costs directly after fracture = c4088 and follow-up period = c639) and in Serbia c3002 (c2359 and c642, respectively). The cost of treatment of a vertebral fracture was c4319 in Slovenia (c3762 and c557, respectively) and in Serbia c390 (c103 and c287, respectively). Treatment of the distal radius fracture was c1567 in Slovenia (c1046 and c521, respectively) and in Serbia c163 (c57 and c106, respectively). CONCLUSIONS: Treatment of proximal femur fractures vs. vertebral and distal radius fractures generated the highest costs. The treatment costs were significantly higher in Slovenia compared with Serbia. Large disparities between the costs of hospitalization in both countries were the major reason for the observed differences. PMS30 MEDICINE TREATMENT COST OF RHEUMATOID ARTHRITIS BEFORE AND AFTER TREATMENT WITH BIOLOGICAL DRUGS Roux I, Lubbe M, Burger JR, Lamprecht J North-West University, Potchefstroom, South Africa OBJECTIVES: To investigate the medicine treatment cost of rheumatoid arthritis (RA) before and after treatment with biological drugs in the private health care sector of South Africa. METHODS: A quantitative retrospective drug utilization review was performed on medicine claims data of a pharmacy benefit management company (PBM) in South Africa. Data for a four-year period (January 1, 2005 to December 31, A308 2008) were used to determine the medicine treatment cost of 141 RA patients before and after treatment with biological drugs (namely infliximab, adalimumab and etanercept). [RSA Rand (R)/$US = 6.38112 (2005); 6.78812 (2006); 7.06926 (2007) and 8.27505 (2008)]. RESULTS: Biological drugs represented 81.43% of the total medicine treatment cost of RA patients (n = R25,432,294.04). Other medication (excluding biological drugs) prescribed to RA patients before starting with biological items represented 8.86% (n = R2 254 330.44) of their total medicine treatment cost; those prescribed after treatment with biological drugs, represented 3.91% (n = R992,533.62). The number of prescriptions for other medication (excl. biological drugs), decreased from the period before to the period after treatment with biological drugs from 6271 to 2120. The average number of the other medicine items (excl. biological) per prescription decreased from 2.79 ± 2.30 before to 2.35 ± 1.86 after treatment with biological drugs. The average cost per biological drug (R8 073.61 ± 2 210.46) was practically significantly (d > 0.8) higher than the average cost of other medication prescribed before (R128.45 ± 155.93) and after (R198.66 ± 888.31) treatment with biological drugs. CONCLUSIONS: Although biological drugs used in the treatment of RA are very expensive, it seems that the number of other medication prescribed to RA patients, as well as the average number of items per prescription decreased after treatment therewith. Further research is needed to investigate future medicine treatment cost of RA patients treated with biological drugs. PMS31 THE IMPACT OF CHANGES IN ADALIMUMAB, ETANERCEPT, AND INFLIXIMAB DOSES ON THE COSTS OF TREATING RHEUMATOID ARTHRITIS Lawson RW1, Heatley RM1, Johnson KI1, Khandker RK2, Singh A2 1 Complete Market Access, Macclesfield, UK; 2Pfizer, Collegeville, PA, USA OBJECTIVES: To review and analyze evidence on the changes in dose of adalimumab, etanercept and infliximab over time in adult patients with rheumatoid arthritis (RA) and the associated impact on treatment costs. METHODS: MEDLINE, EMBASE and NHS-EED were systematically searched to identify English language randomised controlled trials, cohort studies and observational studies published between January 1993 and December 2009. Conference abstracts were also hand searched from EULAR (2002 onwards) and ACR (2006 onwards). Studies were selected using predefined criteria, using two independent reviewers. Data pertaining to dose change were then analyzed through pair-wise, random effects meta-analyses carried out in a frequentist framework with heterogeneity assessed using the I2 statistic. Associated cost data were extracted and the impact of change in dose on cost was investigated. RESULTS: Forty-five articles met the selection criteria with 23 containing dose change data and 26 containing cost data. a significantly greater proportion of patients on infliximab had a dose escalation compared to those on etanercept (odds ratio 0.17 95% CI 0.07, 0.43; P < 0.001) or adalimumab (odds ratio 0.25 95% CI 0.2, 0.3; P < 0.001). On average, 43.3% of infliximab patients, 7.3% of etanercept patients and 10.9% of adalimumab patients had their dose increased. RA related costs were on average 36% higher in patients who had their infliximab dose increased compared to 4% in patients on etanercept. No suitable data for adalimumab were available. CONCLUSIONS: A significantly greater proportion of infliximab patients required dose escalation compared to etanercept and adalimumab patients. There is some evidence to suggest that the escalation in dose required to maintain clinical benefit, results in substantially higher costs of treating RA. PMS32 MICRO-COSTING ANALYSIS AND TARIFF COMPARISON: THE INTERSPINOUS PROCESS DEVICE CASE Beccagutti G1, Corbo M1, Pantaleoni M2, Surace MF3 1 Medtronic Italia, Sesto San Giovanni, Milan, Italy; 2IMS Health S.p.A, Milano, Italy; 3Ospedale di Circolo—Fondazione Macchi, Varese, Italy OBJECTIVES: In Italy the recent update of the DRG system has led to evaluate the effect on the diffusion of new therapies. The Interspinous Process Device (IPD) implantation represents an innovative strategy for different degenerative spinal pathologies with potential clinical and economic advantages. The aim of this study is to evaluate the hospitalization costs for IPD procedure according to a micro-costing approach and to compare it with current regional DRG tariffs. METHODS: The project, conducted from the hospital perspective, is performed in one pilot centre (Varese hospital), regional benchmark for this kind of procedure in which learning curve is considered completed. The cost analysis is based on the clinical pathway drawn up from the information provided by the medical team. Resource use including staff time, diagnostic tests, drugs, consumables and technology equipment utilization are collected from interviews to the team. Operating room costs, administrative and general costs and follow up hospital resource consumption are derived from hospital accounting data. Unit costs are collected either from hospital accounting or regional tariffs for specialist services. RESULTS: The total average cost estimated for a patient submitted to an IPD implantation is c5644, with an average LOS of 2.7 days. The average cost for the implantation of 1 IPD is c4515, value assigned to increase to c7087 for multilevel approaches with the implantation of 2 devices in the same procedure (42% of cases). Excluding general costs and the number of IPDs implanted, the main key cost driver are consumables and devices (62%), and operating room costs (16%). CONCLUSIONS: The regional tariff of the DRG related to this procedure (Lombardia Region, DRG 500, version 24) does not cover the hospitalization costs estimated, especially for the multilevel approaches. This leads to consider the effects of current reimbursement on the adoption of innovative therapy. 13th Euro Abstracts PMS33 COST-EFFECTIVENESS OF TOCILIZUMAB FOR THE MANAGEMENT OF PATIENTS WITH ACTIVE RHEUMATOID ARTHRITIS DESPITE PREVIOUS DMARD THERAPY IN MEXICO Carlos F1, Aguirre A2, Peláez-Ballestas I3 1 R a C Salud Consultores S.A. de C.V., México City, D.F., Mexico; 2R a C Salud Consultores S.A. de C.V., Mexico City, D.F., Mexico; 3Hospital General de México, Secretaría de Salud, Mexico City, D.F., Mexico OBJECTIVES: Rheumatoid arthritis (RA) is a chronic, progressive, inflammatory disease that affects physical functioning and quality-of-life and is associated with premature mortality and substantial economic burden. We aimed to assess costeffectiveness of tocilizumab added to disease-modifying antirheumatic drugs (DMARD) in patients with active RA despite DMARD therapy from the Mexican public health care system perspective. METHODS: Two models were designed to compare tocilizumab 8 mg/kg every 4 weeks; infliximab 3 mg/kg (weeks 0, 2, 6, 14 and 22); etanercept 25 mg twice a week and adalimumab 40 mg every other week. First model included only 6-month acquisition costs of drugs and infusion-related cost for infliximab and tocilizumab; the second was a Markov model with four states defined according to Disease Activity Score (DAS-28). Indirect comparison techniques were needed to adjust American College of Rheumatology (ACR) responses rates found in 10 clinical trials with biological agents. ACR70 at week 24 and overall days spent in remission during 5 years were main outcomes. Unit costs of medications were gathered from public bids; an expert panel was constituted to estimate 3-month resource use by health state. All costs are expressed in 2009 US dollars. RESULTS: First six-month costs were lower with tocilizumab (USD$4418) than with etanercept (USD$5,020), infliximab (USD$5484) and adalimumab (USD$5655). Adjusted ACR70 response rate was higher for tocilizumab than for anti-tumor necrosis factor (TNF) agents: 26% vs. 19%, 18% and 12% for adalimumab, etanercept and infliximab, respectively. Markov model estimates show savings of USD$623 up to USD$1321 per patient treated with tocilizumab instead of anti-TNF. Tocilizumab was also associated with mean gains of 9, 12 and 20 days in remission compared to etanercept, adalimumab and infliximab. CONCLUSIONS: When used instead of antiTNF agents, add-on treatment with tocilizumab brings both health benefits and costsavings for RA patients with inadequate response to previous DMARD therapy. PMS34 COST-EFFECTIVENESS OF GOLIMUMAB IN PSORIATIC ARTHRITIS FROM THE UK PAYER PERSPECTIVE Cummins E1, Asseburg C2, Prasad M3, Buchanan J4, Punekar Y5 1 McMaster Development Consultants Ltd., Glasgow, UK; 2ESiOR Ltd, Kuopio, Finland; 3 Merck & Co, Kenilworth, NJ, USA; 4Johnson & Johnson Pharmaceutical Services, LLC, Malvern, PA, USA; 5Schering Plough, Welwyn Garden City, UK BACKGROUND: Golimumab is a novel TNF-α inhibitor licensed to treat patients with active PsA. Although its clinical efficacy has been proven in clinical trials, its costeffectiveness is yet to be established. OBJECTIVES: To estimate the cost-effectiveness of golimumab among patients with active PsA from the UK NHS perspective. METHODS: A decision analytic model was used to simulate progression of a hypothetical cohort of active PsA patients on golimumab and other TNF-α inhibitors as well as palliative care. The clinical evidence was derived from clinical trials of TNF-α inhibitors and compared using mixed treatment models. The primary outcome measure was quality adjusted life-years (QALYs) estimated based on change in Health Assessment Questionnaire (HAQ) and Psoriasis Area Severity Index (PASI) from baseline. The annual acquisition cost of golimumab was assumed to be identical to annual cost of other subcutaneous TNF-α inhibitors. The resource use costs and outcomes were discounted at 3.5% over a period of 40 years. The uncertainty surrounding important variables was further explored using probabilistic sensitivity analyses (PSA). RESULTS: TNF-α inhibitors were significantly superior to palliative care but comparable to each other on Psoriatic Arthritis Response Criteria (PsARC), HAQ and PASI response. The incremental cost-effectiveness ratio (ICERs) for golimumab compared to palliative care was £16,811 for PsA patients and £16,245 for a subgroup of PsA patients with significant psoriasis. At an acceptability threshold of £30,000 per QALY, the probability of golimumab being cost-effective is 89%. CONCLUSIONS: Once monthly, golimumab is a cost-effective treatment alternative for patients with active PsA. With its patient focussed attributes, golimumab is likely to offer additional choice in PsA treatment. PMS35 COST-EFFECTIVENESS OF TERIPARATIDE IN PATIENTS WITH GLUCOCORTICOID-INDUCED OSTEOPOROSIS IN SWEDEN Smolen LJ1, Gregor Z2, Barrett A3, Myrén KJ4, Toll A5 1 Medical Decision Modeling Inc., Indianapolis, IN, USA; 2Eli Lilly & Company, Prague, Czech Republic; 3Eli Lilly & Company Ltd, Windlesham, Surrey, UK; 4Eli Lilly & Company, Windlesham, UK; 5Eli Lilly Sweden AB, Solna, Sweden OBJECTIVES: Glucocorticoid induced osteoporosis is the most common cause of secondary osteoporosis. The objective of this study was to estimate the cost-effectiveness of teriparatide in patients with Glucocorticoid induced osteoporosis in Sweden. METHODS: A cost-effectiveness analysis was developed to evaluate the direct medical and tertiary care costs and clinical outcomes of an 18-month regimen of daily teriparatide in patients with glucocorticoid induced osteoporosis (GIO). a Monte Carlo simulation was used to model the cost and effects of a simulated cohort of 100,000 patients with GIO treated with teriparatide compared to no teriparatide treatment. The model simulated the course of events in 6-month cycles in individual patients over a lifetime horizon. During each cycle the patients were at risk of experiencing clinical 13th Euro Abstracts vertebral, hip and wrist fracture or death (either natural or excess mortality due to fracture). Swedish data on fracture costs, utility reductions after fracture, fracture risks and mortality rates were used. Uncertainty was investigated using one-way and probabilistic sensitivity analyses. Costs and utilities were discounted at annual discount rates of 3%. RESULTS: The analyzed cohort comprised patients aged 69 years (80% female) with a BMD T-Score of −2.5 SD and an historical vertebral fracture (5 years previous) and an incident vertebral fracture. In the base-case analysis of this cohort the costs in the teriparatide treatment group were 558,918 SEK per patient compared to 552,026 SEK in the no teriparatide group. The cost per QALY gained of teriparatide compared to no teriparatide was estimated to be SEK 25,000. The results were robust under a wide range of assumptions. CONCLUSIONS: For the analyzed cohorts, the base-case and one-way sensitivity analyses performed indicate that an 18-month teriparatide regimen versus no treatment in patients with glucocorticoid induced osteoporosis is cost-effective from the perspective of the Swedish payer. PMS36 COST-EFFECTIVENESS OF TOCILIZUMAB FOR THE MANAGEMENT OF PATIENTS WITH ACTIVE RHEUMATOID ARTHRITIS DESPITE PREVIOUS DMARD THERAPY IN COSTA RICA Carlos F R a C Salud Consultores S.A. de C.V., México, D.F., Mexico OBJECTIVES: Rheumatoid arthritis (RA) is a chronic, progressive, inflammatory disease that affects physical functioning and quality-of-life and is associated with premature mortality and substantial economic burden. We aimed to assess the costeffectiveness of tocilizumab added to disease-modifying antirheumatic drugs (DMARD) in patients with active RA despite DMARD therapy from the perspective of public health care system in Costa Rica. METHODS: A decision analysis was carried out to compare tocilizumab 8 mg/kg given every 4 weeks; infliximab 3 mg/kg (weeks 0, 2, 6, 14) and 5 mg/kg (every 8 weeks from week 22); etanercept 25 mg given twice a week and adalimumab 40 mg given every other week. The model included acquisition costs of biological agents during first year of treatment besides infusion-related costs for infliximab and tocilizumab. Indirect comparison techniques were needed to adjust American College of Rheumatology (ACR) responses rates found in 10 placebo-controlled clinical trials with biological agents used as add-on therapy to DMARD. ACR70 response rate, which can be regarded as a close measure of remission, was selected as primary efficacy outcome. Unitary costs were gathered from the 2010 Official Price List of the Public Health Care System in Costa Rica. All costs are expressed in 2010 US dollars. RESULTS: First-year costs for an average 70 kg weight patient were lower with tocilizumab (US$12,272) than with etanercept (US$13,000), adalimumab (US$13,650) and infliximab (US$14,340). Adjusted ACR70 response rate was higher for tocilizumab (26%) than for adalimumab (19%), etanercept (18%) and infliximab (12%). Incremental cost per patient achieving an ACR70 response with tocilizumab instead of anti-tumor necrosis factor (TNF) agents were estimated at −US$9,100, −US$14,771 and −US$19,686 for etanercept, adalimumab and infliximab, respectively. CONCLUSIONS: When used instead of anti-TNF agents, add-on treatment with tocilizumab brings both health benefits and cost-savings for RA patients with inadequate response to previous DMARD therapy. PMS37 THE COST-EFFECTIVENESS OF ABATACEPT IN COMBINATION WITH METHOTREXATE FOR THE TREATMENT OF PATIENT WITH ACTIVE RHEUMATOID ARTHRITIS AFTER AN INADEQUATE RESPONSE TO METHOTREXATE IN THE UNITED KINGDOM Lebmeier M1, Pericleous L1, Drost P2, Dequen P3, Ouwens M4, Bergman G5, Baig H1 1 Bristol-Myers Squibb Pharmaceuticals Ltd, Uxbridge, Middlesex, UK; 2Bristol-Myers Squibb, Braine-l’Alleud, Belgium; 3Mapi Values, Bollington, Cheshire, UK; 4Mapi Values, Houten, The Netherlands OBJECTIVES: Abatacept in combination with MTX has recently been granted a positive opinion from the European Medicines Agency for use for the treatment of moderate to severe active rheumatoid arthritis in adult patients who responded inadequately to previous therapy with one or more conventional disease-modifying antirheumatic drugs (cDMARDs) including methotrexate (MTX). This analysis explores the cost-effectiveness of abatacept in this new indication. METHODS: A patientsimulation treatment-sequence economic model was constructed to estimate the incremental cost per quality adjusted life-year (QALY) for patients with RA in the United Kingdom. Abatacept with MTX , followed by a sequence of DMARDs was compared against a sequence of cDMARDs. Treatment-specific efficacy in terms of Health Assessment Questionnaire (HAQ) was used to calculate the patient’s utility medical resource use and cost over a lifetime. Mortality was HAQ dependent. The analysis is performed from a National Health Service. Costs and outcomes were discounted at 3.5% each. RESULTS: Abatacept with MTX was estimated to yield 1.09 QALYs per patient (6.42 vs. 5.33) over lifetime, compared to DMARDs. The total lifetime costs associated with abatacept with MTX were £110,094 and total costs for cDMARDs were £79,933 resulting in an incremental cost-effectiveness ratio (ICER) of £27,657 per QALY gained. Sensitivity analysis confirmed the robustness of the model findings. CONCLUSIONS: This study has demonstrated that abatacept with MTX is a costeffective treatment option compared to cDMARDs for patients with rheumatoid arthritis after an inadequate response to MTX. A309 PMS38 COST-EFFECTIVENESS OF RITUXIMAB VERSUS ALTERNATIVE ANTI-TUMOR NECROSIS FACTOR (TNF) THERAPY AFTER PREVIOUS FAILURE OF ONE ANTI-TNF AGENT FOR TREATMENT OF RHEUMATOID ARTHRITIS IN MEXICO Carlos F R a C Salud Consultores S.A. de C.V., México City, D.F., Mexico OBJECTIVES: About 30% of patients treated with an anti-TNF agent failed to achieve an improvement of 20% in American College of Rheumatology (ACR) response. Recent clinical practice guidelines recommend the use of rituximab after previous failure of one anti-TNF. This study aims to assess the cost-effectiveness of rituximab compared to cycling between anti-TNF agents in this population from the perspective of the public health care system in Mexico. METHODS: A decision analysis was carried out to compare 2 rituximab courses (1 course, consisting of 2 infusions of 1 g each) given 6 months apart; infliximab 3 mg/kg (weeks 0, 2, 6, 14) and 5 mg/ kg (weeks 22, 30, 38 and 46); etanercept 25 mg twice a week and adalimumab 40 mg every other week. Only direct medical costs cumulated during a one-year timeframe were accounted for and these included acquisition cost of biologic drugs besides infusion costs for rituximab and infliximab. Primary efficacy outcome was defined as an improvement of 70% in ACR response (ACR70), which is a close measure of remission. Indirect comparison techniques were used to adjust ACR responses rates found in 9 clinical trials. Number needed to treat (NNT) to obtain an ACR70 was then calculated. All costs are reported in 2009 US dollars (USD). RESULTS: For a 70 Kg patient, annual mean costs were estimated at USD$13,025 for rituximab, USD$12,938 for infliximab, USD$12,226 for adalimumab and USD$10,850 for etanercept. Adjusted ACR70 rates were higher in rituximab (12.4%) than in adalimumab (9.0%), etanercept (8.2%) and infliximab (5.4%). Average cost to achieve an ACR70 was lower with rituximab (USD$105,047) than with anti-TNF therapies, leading to savings of USD$27,270; USD$30,797 and USD$134,543 compared to etanercept, adalimumab and infliximab, respectively. CONCLUSIONS: This study suggests that rituximab treatment after previous failure of one anti-TNF agent is a cost-effective strategy compared to cycling between anti-TNF agents. PMS39 COST-EFFECTIVENESS OF RITUXIMAB VERSUS ALTERNATIVE ANTI-TUMOR NECROSIS FACTOR (TNF) THERAPY AFTER PREVIOUS FAILURE OF ONE ANTI-TNF AGENT FOR TREATMENT OF RHEUMATOID ARTHRITIS IN COSTA RICA Carlos F R a C Salud Consultores S.A. de C.V., México City, D.F., Mexico OBJECTIVES: About 30% of patients treated with an anti-TNF agent failed to achieve an improvement of 20% in American College of Rheumatology (ACR) response. Recent clinical practice guidelines recommend the use of rituximab after previous failure of one anti-TNF. This study aims to assess the cost-effectiveness of rituximab compared to cycling between anti-TNF agents in this population from the perspective of public health care system in Costa Rica. METHODS: A decision analysis was carried out to compare 2 rituximab courses (1 course, consisting of 2 infusions of 1 g each) given 6 months apart; infliximab 3 mg/kg (weeks 0, 2, 6, 14) and 5 mg/ kg (weeks 22, 30, 38 and 46); etanercept 25 mg twice a week and adalimumab 40 mg every other week. Only direct medical costs cumulated during a one-year timeframe were accounted for and these included acquisition cost of biologic drugs besides infusion costs for rituximab and infliximab. Primary efficacy outcome was defined as an improvement of 70% in ACR response (ACR70), which is a close measure of remission. Indirect comparison techniques were used to adjust ACR responses rates found in 9 clinical trials. Number needed to treat (NNT) to obtain an ACR70 was then calculated. All costs are reported in 2009 US dollars (USD). RESULTS: For a 70 Kg patient, annual mean costs were estimated at US$15,040 for rituximab, US$14,340 for infliximab, US$13,650 for adalimumab, and US$13,000 for etanercept. Adjusted ACR70 rates were higher in rituximab (12.4%) than in adalimumab (9.0%), etanercept (8.2%) and infliximab (5.4%). Average cost to achieve an ACR70 was lower with rituximab (US$121,290) than with anti-TNF therapies, leading to savings of US$30,377; US$37,247; and US$144,266 compared to etanercept, adalimumab and infliximab, respectively. CONCLUSIONS: This study suggests that rituximab treatment after previous failure of one anti-TNF agent is a cost-effective strategy compared to cycling between anti-TNF agents. PMS40 COST-EFFECTIVENESS OF DENOSUMAB COMPARED WITH GENERIC ALENDRONATE IN THE TREATMENT OF POSTMENOPAUSAL OSTEOPOROTIC WOMEN Hiligsmann M, Reginster JY University of Liège, Liège, Belgium OBJECTIVES: Denosumab represents a new therapeutic opportunity for the treatment of osteoporosis, that received a positive opinion from the European Committee for Medical Products for Human Use in December 2009. This study aims to evaluate the cost-effectiveness of denosumab compared with the most relevant alternative (i.e. generic alendronate) in the treatment of postmenopausal osteoporotic women. METHODS: The cost-effectiveness of treatment for 3-years with denosumab was compared with generic alendronate using an updated version of a previously validated Markov microsimulation model (Value Health 2009:12:687–96). The model was A310 populated with relevant cost, adherence and epidemiological data for Belgium from a health care perspective and the results were presented in cost (c2009) per qualityadjusted life-year (QALY) gained. Analyses were performed in populations (over 60 years) where osteoporosis medications are currently reimbursed in many European countries, i.e. bone mineral density T-score is below or equal to −2.5 or presence of prevalent vertebral fracture. Univariate and probabilistic sensitivity analyses were conducted to assess the robustness of the results. RESULTS: The cost-effectiveness of denosumab compared with generic alendronate was estimated at c38,875, c20,690 and c26,153 per QALY for women with T-score ≤−2.5 aged 60, 70 and 80 years, respectively. The equivalent values were c37,856, c18,764 and c17,309 per QALY for women with prevalent vertebral fractures. Discount rates, fracture risk and patient’s adherence to generic alendronate were found to be particularly sensitive when varied within the model. At 70 years of age, probabilistic sensitivity analyses showed that the probability of the ICERs remaining below c40,000 is 84% in women with prevalent vertebral fracture and 74% in those with T-score ≤−2.5, confirming the cost-effectiveness of denosumab. CONCLUSIONS: This study suggests that denosumab is a cost-effective strategy (cost per QALY gained ≤c40,000) compared with generic alendronate for the treatment of postmenopausal Belgian osteoporotic women, aged 60 years and above. PMS41 THE IMPACT OF MEDICATION ADHERENCE ON THE COST OF OSTEOPOROSIS FRACTURES IN GERMANY: A MODELLING APPROACH Stephens JM1, Brotherton SA1, Gitlin MD2, Steinle T3, Leidl R4, Kochevar JJ1 1 Kochevar Research Associates, Charlestown, MA, USA; 2Amgen (Europe) GmbH, Zug, Switzerland; 3Amgen (Europe) GmbH, München, Germany; 4Ludwig Maximilian University, Munich, Germany OBJECTIVES: Current osteoporosis therapies can reduce the risk of fractures and thus costs, but adherence to these medications is often poor. We estimated yearly potential savings in fracture-related costs from improved adherence to osteoporosis therapies in Germany from a societal perspective. METHODS: The model was a deterministic cohort model estimating annual treatments and outcomes for German women aged ≥50 years old in 2010 (n = 17,689,849), by f5 year cohorts. Demographic parameters were drawn from government databases, medication prescription share from industry databases, fracture incidence (hip, clinical vertebral, and non-hip, non vertebral (NHNV)), medication adherence (medication possession ratio MPR, one year intervals) and efficacy from published literature. As a base case we estimated 4.2% treated (n = 745,313) with oral bisphosphonates (94.6%), strontium (2.7%) and raloxifene (2.6%); 54% with effective adherence (MPR > 50%). Fracture related costs (excluding medication costs) included direct medical, long-term nursing care, and work loss. The model parameters are adjustable, allowing real time calculation of outcomes. We estimated current costs with 54% effectively adherent and 100% effectively adherent. RESULTS: The model calculated total fracture-related costs for all patients in 2010 at c4.4 billion (with 69.6% for direct medical costs, 20.2% for long-term nursing care, and 10.2% for work loss). For the estimated 4.2% of treated patients, the model calculated total fracture-related costs at c184.9 million. Increasing the percentage of patients with effective adherence from 54% to 100% decreased costs for hip (16.2%), vertebral (24.4%) and NHNV fractures (6.3%). CONCLUSIONS: Adjustable parameters allow users to calculate yearly fracture-related costs and savings for different economic perspectives and decision options. According to one run using the above parameters, achieving full adherence to medication among women currently treated for osteoporosis in Germany would reduce fractures, and cut down annual fracture related costs by c27.8 million (15.1%). PMS42 COST-EFFECTIVENESS OF CERTOLIZUMAB PEGOL PLUS METHOTREXATE OR AS MONOTHERAPY FOR THE TREATMENT OF ACTIVE RHEUMATOID ARTHRITIS IN THE UNITED KINGDOM Purcaru O1, Taylor PC2, Emery P3, Palmer S4 1 UCB, Brussels, Belgium; 2Imperial College London, London, UK; 3University of Leeds, Leeds, UK; 4University of York, York, UK OBJECTIVES: To evaluate the cost-effectiveness of certolizumab pegol (CZP) either as an add-on therapy to methotrexate (MTX) or as monotherapy versus other biological DMARDs, in the treatment of patients with active RA who did not respond adequately to DMARDs, including MTX, in the United Kingdom (UK). METHODS: A cost-utility model with a Markov structure was developed following BSR and NICE guidelines. Comparators considered in the analysis were licensed and recommended TNF-α inhibitors in England and Wales (etanercept [ETA], adalimumab [ADA] and infliximab [IFX]), + MTX or as monotherapy (ETA and ADA only). Clinical efficacy, history and resource use data came from published literature. Unit costs were taken from routine sources/published references. Utilities were derived from EQ-5D data collected in CZP RA clinical trials. Cost for CZP assumed the first 12 weeks free of charge, as per the patient access scheme. Base-case analysis was conducted from the NHS perspective, with lifetime horizon, costs and outcomes discounting rates of 3.5%. Sensitivity analyses were conducted. RESULTS: Base-case analysis indicated that CZP is cost-effective compared with all combination and monotherapies considered. CZP+MTX dominated the comparators considered. CZP+MTX was associated with an incremental gain in quality-adjusted life-years (QALYs) of 0.242, 0.065 and 0.458 when compared with ADA+MTX, ETA+MTX and IFX+MTX, respectively. Total costs for CZP+MTX were lower than comparator combination therapies by £451, £582 and £10,016, respectively. CZP+MTX had the highest probability of being cost-effective of 13th Euro Abstracts the TNF inhibitor + MTX therapies considered (at £20,000/QALY, 53.6% for CZP+MTX, 25% for ETA+MTX, 21.2% for ADA+MTX, 0.2% for IFX+MTX). Among monotherapies, CZP dominated ADA (QALY incremental gain: 0.215, total costs lower by £2,352 vs. ADA). CZP was cost-effective below a willing-to-pay threshold of £31,582/QALYs vs. ETA. CONCLUSIONS: This analysis indicates that CZP is a cost-effective treatment for RA in the UK. PMS43 COST-UTILITY AND MULTINOMIAL EXPECTED VALUE OF PERFECT INFORMATION OF SEQUENCED TREATMENT WITH BIOLOGICS IN MODERATE-TO-SEVERE RHEUMATOID ARTHRITIS: ANALYSIS BASED ON CLINICALLY MEANINGFUL RESPONSES Soini EJ1, Hallinen T1, Kauppi MJ2, Vihervaara V3, Puolakka K4 1 ESiOR Ltd, Kuopio, Finland; 2Päijät-Häme Central Hospital, Lahti, Finland; 3Roche Oy, Espoo, Finland; 4Lappeenranta Central Hospital, Lappeenranta, Finland OBJECTIVES: To assess the cost-utility and multinomial expected value of perfect information (mEVPI) of sequenced treatment with biologics in moderate-to-severe rheumatoid arthritis (msRA) after failure of traditional disease modifying antirheumatic drug(s) (tDMARD). METHODS: A probabilistic, individual sampling (microsimulation) model was developed to compare ten different treatment sequences among 3,000 hypothetical Finnish msRA patients in a lifetime scenario. Adalimumab+methotrexate (MTX), etanercept+MTX or tocilizumab+MTX were used as first biologics followed by up to three other biologics. Best supportive care including tDMARDs was assumed to be used after exhaustion of treatment options with biologics. Treatment with MTX alone was added as further comparator. The clinical outcomes (no ACR50, ACR50 and ACR70 responses conditional to the use of biologic drugs) were obtained from a recently published mixed treatment comparison and quality-adjusted life-years (QALY) were estimated based on disease severity scores (HAQ) using a nonlinear equation: EQ5D = 0.82-0.11*HAQ-0.07*HAQ2. Resource use was estimated from published references and valued with Finnish unit costs (year 2009). Analyses were performed from payer perspective (productivity losses were excluded) using 3% annual discount rate. RESULTS: Compared to MTX alone, treatment with tocilizumab+MTX was more cost-effective than treatment with etanercept+MTX. Both tocilizumab+MTX and etanercept+MTX dominated adalimumab+MTX. An additional QALY gained with tocilizumab+MTX costs c15,478 (mEVPI c1,258/patient) compared with MTX alone. An additional QALY gained with tocilizumab+MTX followed by etanercept+MTX costs c35,543 (mEVPI c748/patient) compared with tocilizumab+MTX. According to cost-effectiveness acceptability frontier, only MTX alone, tocilizumab+MTX or tocilizumab+MTX followed by etanercept+MTX should be considered, if willingness to pay is c0–50,000 per QALY gained. With c30,000 per QALY gained, tocilizumab+MTX had 97.6% probability of being cost-effective. The results were relatively robust in sensitivity analyses. CONCLUSIONS: After tDMARD failure, tocilizumab+MTX or tocilizumab+MTX followed by etanercept+MTX were the most cost-effective biologics for patients with msRA. mEVPI indicated that the value of additional research information is low. PMS44 IMPACT OF OBESITY IN WORKING ADULTS WITH ARTHRITIS IN TERMS OF MEDICAL AND PRDUCTIVITY COSTS Suh DC1, Kwon JW2, Kim CM3, Choi IS2, McGuire M2, Barone J2 1 School of Pharmacy, Rutgers University, Piscataway, NJ, USA; 2Rutgers University, Piscataway, NJ, USA; 3Catholic University School of Medicine, Seoul, South Korea OBJECTIVES: To measure the impact of obesity on annual medical and productivity costs among working U.S. adults with arthritis. METHODS: We conducted a cross sectional study using Medical Expenditure Panel Survey data from 2003–2007. Working adults with arthritis (18–64 years old) were selected if they did not have pregnancy, malignancy, kidney dialysis, immunodeficiency, low body mass index (BMI < 18.5 kg/m2), or unemployed status. Patients with arthritis were identified by ICD-9 codes of 714–715 or via patient self-report. Obese and normal- weight were defined as BMI of ≥30 kg/m2 or BMI of 18.5–<25 kg/m2, respectively. Loss of productivity was estimated by loss of workdays due to illness or injury and standard hourly wage by occupation. Medical costs were estimated using a generalized linear model with a log link function and gamma distribution. Costs of productivity loss were calculated using a two-part model to adjust for patients with zero costs. Using Oaxaca decomposition, differences in treatment costs between obese and normal-weight patients were decomposed into two parts: a) differences in characteristics (endowments) across groups, and b) differences between obese and normal parameters (coefficients). Costs attributable to obesity were defined as the costs by coefficients component. All costs were converted to 2009 U.S. dollars using price indices. RESULTS: Among the 7345 working adults with arthritis, prevalence of obesity and normal-weight was 24.8% vs. 40.7%, respectively. The difference in medical costs between the groups was US$2380 (95% CI: US$1934–2825) due to endowments and US$379 (95% CI: US$367–392) due to coefficient components. Productivity loss costs in the obese patients were higher, at US$46(95% CI: US$ 41–51) due to endowments and US$441(95%CI: US$435–447) due to coefficient components. CONCLUSIONS: Use of the Oaxaca decomposition method suggested that the economic burden, particularly productivity loss costs, of obesity in patients with arthritis was substantial, adjusting for characteristics across groups. 13th Euro Abstracts PMS45 PRODUCTIVITY COSTS IN PATIENTS WITH RHEUMATOID ARTHRITIS IN SOUTH KOREA Kwon JM, Lee EK SookMyung Women’s University, Seoul, South Korea OBJECTIVES: The objectives of this study are to examine socio-demographic and clinical characteristics of patients with rheumatoid arthritis(RA) and to estimate productivity costs related to RA using data from the Fourth Korea National Health and Nutrition Examination Survey conducted in 2007(KNHANES IV-1) and 2008(KNHANES IV-2). METHODS: A total of 7559 adults of working age(19 ≤ age <65) were available for analyses, including 99 individuals with RA. Socio-demographic and clinical characteristics of subjects with RA were compared with those of subjects without RA. Multiple logistic regression was used to estimate adjusted employment rates of RA population. Percentage of work disability due to RA was estimated using the difference between employment rates of general population without RA and the adjusted employment rates of RA population. All statistical analyses were performed with SURVEY procedure of SAS/STAT® 9.1. RESULTS: The mean age of patients with RA was 50.2 years and 73.0% of them were female. Individuals with RA were older and had lower rates of subjective health awareness, higher rates of activity limitation and lower employment rates than individuals without RA. The percentage of sick leave due to RA was 2.5%. According to the results of multiple logistic regression, subjects with RA were less likely to be employed than subjects without RA (OR 0.63, 95% CI 0.41–0.97). The estimated percentage of work disability due to RA was 11.1%. By using the human capital approach, per capita productivity costs due to work disability were estimated to 3.6 million won($3594) per year. CONCLUSIONS: In South Korea, the percentages of work disability and sick leave for RA patients were lower than those reported in previous studies. Although other aspects of productivity costs related to RA could not be analyzed, this study could provide several important results concerning productivity costs in patients with RA. MUSCULAR-SKELETAL DISORDERS – Patient-Reported Outcomes Studies PMS46 DOSING PROFILE COMPARISON BETWEEN PATIENTS WITH FIBROMYALGIA INITIATING DULOXETINE AND PREGABALIN IN 2007: A LARGE NATIONAL RETROSPECTIVE COHORT STUDY Sun P1, Zhao Y2, Wu A3, Sun S4 1 Kailo Research Group, Indianapolis, IN, USA; 2Eli Lilly & Company, Indianapolis, IN, USA; 3 Kailo Research Group, Los Angeles, CA, USA; 4Kailo Research Group, Fremont, CA, USA OBJECTIVES: To compare dosing profiles between duloxetine and pregabalin among patients with fibromyalgia. METHODS: A retrospective cohort study was conducted on a large U.S. administrative claims database to examine commercially insured individuals aged 18 to 64 who had fibromyalgia and initiated (a 90-day clean period before initiation) duloxetine or pregabalin in 2007. Continuous enrollment in the 12 months pre- and post-index periods and at least 30-day supply of duloxetine or pregabalin over the 1-year follow-up were required for all selected patients. Average daily doses of all duloxetine or pregabalin prescriptions per person, average daily dose and average daily costs in each of the first 12 prescriptions, percent of patients with increased or decreased dose, and percent of daily dose change from previous prescription were compared between duloxetine and pregabalin cohorts. RESULTS: Both the duloxetine (n = 1699) and pregabalin (n = 2560) cohorts had a mean age of 51~52 years. Duloxetine-treated patients had an average initial daily dose of 55.6 mg versus 159.8 mg for the pregabalin-treated patients. The average daily dose per patient was 55.7 mg for duloxetine and 195.4 mg for pregabalin. The average duloxetine daily doses and daily costs ranged from 55.0 mg–60.2 mg and $3.9-$5.0 for the first 12 prescriptions and 159.8 mg–265.7 mg and $4.4–$5.0 for pregabalin. The percentages of daily-dose changes from previous prescription were −5.4%~3.0% for duloxetine and −1.0%~16.6% for pregabalin. Over the 1-year follow-up, both the percentages of patients with a dose increase (17.3% for duloxetine vs. 51.8% for pregabalin, P < 0.01) and a dose decrease (24.8% for duloxetine vs. 10.0% for pregabalin P < 0.01) were statistically different between cohorts. CONCLUSIONS: Among patients with fibromyalgia, duloxetine and pregabalin initiators had very different dosing profiles. The average daily dose for duloxetine was relatively stable over time, while pregabalintreated patients had significant dose escalation over the 12-month follow-up period. PMS47 COMPARISON OF DOSING PATTERNS BETWEEN FIBROMYALGIA PATIENTS WITH HIGH VERSUS LOW DULOXETINE OR PREGABALIN COMPLIANCE Zhao Y1, Sun P2, Sun S3, Wu A4 1 Eli Lilly & Company, Indianapolis, IN, USA; 2Kailo Research Group, Indianapolis, IN, USA; 3 Kailo Research Group, Fremont, CA, USA; 4Kailo Research Group, Los Angeles, CA, USA OBJECTIVES: To compare dosing patterns between patients with fibromyalgia who had high versus low duloxetine or pregabalin compliance, respectively. METHODS: Using a large U.S. administrative claims database, commercially insured patients aged 18–64 who had fibromyalgia and initiated duloxetine or pregabalin in 2007 were examined. Medication compliance was measured via medication possession ratio (MPR), with MPR¡Ý (<) 0.8 as high (low) compliance. All patients selected had A311 continuous enrollment in the 12-month pre- and post-index periods, at least 30 days¡− supply of duloxetine or pregabalin in the 12-month post-index period, and were classified into 4 mutually exclusive cohorts based on their initiation agents (duloxetine vs. pregabalin) and compliance levels (high vs. low). Average daily doses of the first 8 prescriptions and duration between the initial dose and the first increased dose were compared across cohorts. RESULTS: Compared to the low compliance patients, patients with high compliance had significantly higher initial (duloxetine [N = 804 vs. 895]: 56.0 mg vs. 55.2 mg; pregablin [N = 710 vs. 1,850]: 164.9 mg vs. 157.9 mg) and average (duloxetine: 57.0 mg vs. 54.6 mg; pregabalin: 216.8 mg vs. 187.1 mg) daily doses (all P < .05). The average daily doses for the first 8 prescriptions ranged from 55.0 mg–58.3 mg and 56.0 mg–60.6 mg for duloxetine high and low compliance patients, respectively, while the pregabalin-treated patients had 157.9 mg–238.8 mg and 164.9 mg–248.3 mg, correspondingly. The duration from the initial dose to the first dose increase was 263.5 vs. 161.1 days (P < .05) for the duloxetine high versus low compliance patients and 268 vs. 135.4 days for the pregabalin high versus low compliance patients, respectively. CONCLUSIONS: Among patients with fibromyalgia, duloxetine and pregabalin initiators had different dosing patterns by compliance levels. Both duloxetine- and pregabalin-treated patients with high compliance had higher initial and average daily doses than those with low compliance. The average daily dose remained stable over time for duloxetine, however, increased for pregabalin for both compliance groups. PMS48 PSYCHOMETRIC PROPERTIES OF THE GREEK MODIFIED HEALTH ASSESSMENT QUESTIONNAIRE FOR ASSESSING PHYSICAL DISABILITY IN RHEUMATOID ARTHRITIS Kontodimopoulos N1, Bozios P2, Raftakis I2, Elmatzoglou I2, Yfantopoulos J3, Niakas D1 1 Hellenic Open University, Patras, Greece; 2Asklipieio Voulas General Hospital, Voula, Greece; 3National and Kapodistrian University of Athens, Athens, Greece OBJECTIVES: To test basic psychometric properties (validity, reliability and responsiveness) of the Greek version of the Modified Stanford Health Assessment Questionnaire (MHAQ) for assessment of physical disability in rheumatoid arthritis (RA). METHODS: The sample consisted of 120 RA patients (60.0% female, mean age 59.0) starting on biological antirheumatic drugs. Outcome measures were the MHAQ, the Disease Activity Score (DAS28) and the EQ-5D completed at baseline and at 3 months post-intervention. Item response frequencies and ceiling/floor effects were examined. Internal consistency reliability was assessed with Cronbach’s alpha. Construct validity was examined by correlating baseline item and total scores of the MHAQ, with the EQ-5D and DAS28. MHAQ items were also tested for their correlation with the principal component. Effect sizes between MHAQ baseline and three-month data were compared to respective values for the other outcomes for evidence of responsiveness. RESULTS: Floor and ceiling effects for the eight MHAQ items ranged between 0–21.7% and 0.8–4.2% respectively, and the full range of responses was used in all but one item. Cronbach’s alpha was 0.89 at baseline and 0.86 at three months post intervention, indicating good internal consistency of the instrument. Moderate correlations were noted (Spearman’s ρ of 0.3 to 0.5) between most MHAQ and EQ-5D domains, and between the overall scores, and strong correlations (ρ > 0.5) between MHAQ and DAS28. Principal component analysis resulted in one factor explaining 57.1% of the total variance of the scale and all 8 items had high correlation coefficients (range 0.561–0.862) with this principal component. Responsiveness was satisfactory as the MHAQ effect size was 0.41, comparable to EQ-5D and DAS28 effect sizes (0.46 and 0.33 respectively). CONCLUSIONS: Evidence has been provided to support the reliability, validity and sensitivity to change of the Greek version of the MHAQ in the evaluation of functional status of RA patients. PMS49 PSYCHOMETRIC PROPERTIES OF THE RHEUMATOID ARTHRITIS DISEASE ACTIVITY INDEX (RADAI) IN A COMMUNITY POPULATION IN THE US Bharmal M1, Cascade E2 1 Quintiles, Rockville, MD, USA; 2iGuard, Inc, Rockville, MD, USA OBJECTIVES: The Rheumatoid Arthritis Disease Activity Index (RADAI) was developed to provide an easy to use self-administered assessment of rheumatoid arthritis (RA) to complement physician assessment. This study evaluates the psychometric properties of a web administered version of the RADAI in a community sample in the US. METHODS: A random sample of iGuard.org members in the US treated with medications for RA completed the RADAI and a series of other questions related to their disease. iGuard.org is a free medication monitoring service that is introduced to patients through multiple sources including physician, pharmacy and online referrals. Internal consistency of the RADAI was evaluated using Cronbach’s alpha and itemtotal correlations, and factor analysis was used to confirm the domain structure. Convergent validity was established using correlations with patient global assessment of pain and number of painful joints. RESULTS: A total of 153 RA patients completed the study. The mean (SD) age of respondents was 52.7 (10.9) years, 71.2% were females with 54.3% diagnosed with RA 1 to 5 years ago and 39.2% diagnosed 5 to 15 years ago. The mean (SD) RADAI score was 4.59 (2.16), patient global assessment of pain was 52.11 (24.88) on a 100-point scale, and number of painful joints were 7.93 (4.22). RADAI items had good internal consistency with Cronbach’s alpha of 0.89 and all item-total correlations >= 0.56. Factor analysis confirmed one factor with factor loadings of all the items on the factor of >= 0.57. As expected, RADAI scores were significantly correlated with patient global assessment of pain (0.646; P < 0.0001) A312 and number of patient joints (0.649; P < 0.0001). CONCLUSIONS: The study confirms the reliability and validity of a web administered version of the RADAI questionnaire in a US community sample. It is a useful measure to assess RA disease status for patient recruitment in interventional studies and for observational study designs. PMS50 PATIENTS’ STATED HEALTH-OUTCOME PREFERENCES FOR CONFOUNDED PATIENT-REPORTED OUTCOME DOMAINS FOR OSTEOARTHRITIS Johnson FR, Gonzalez JM, Hauber AB RTI Health Solutions, Research Triangle Park, NC, USA BACKGROUND: Patient-reported outcome (PRO) instruments such as the WOMAC Index of Osteoarthritis provide scores for outcomes that cannot be measured objectively. However, such scores do not necessarily indicate patients’ perception of the relative importance of the domains and scales used in such instruments. OBJECTIVES: To obtain valid estimates of patient preferences for osteoarthritis domains and potential adverse outcomes. METHODS: Residents of the United Kingdom aged 45 years or older with a self-reported physician diagnosis of osteoarthritis (OA) completed a web-enabled conjoint-analysis or discrete-choice experiment survey. Patients evaluated 10 pairs of hypothetical OA treatment-outcome profiles. Each hypothetical OA treatment was defined by 4 benefit attributes derived from the WOMAC instrument (ambulatory pain, resting pain, stiffness, and daily activities) plus 3 adverse-event risk attributes (bleeding ulcer, stroke, and coronary infarction). Choice-model parameter estimates were obtained using random-parameters logit. RESULTS: A total of 294 subjects provided usable data. Valid parameter estimates should indicate that better outcomes are preferred to worse outcomes. a main-effects choice model yielded correctly ordered, risk-level, preference-weight estimates for the 3 adverse-event attributes, but disordered severity estimates for all 4 PRO attributes. Implausible combinations of ambulatory-pain and daily-activity levels may cause patients to focus on only one of the attributes. a model that excludes the daily-activity attribute and includes a dummy variable to account for implausible combinations yields correctly ordered preference-weight estimates. An improvement in ambulatory pain from 25 mm to 0 mm on a 100 mm scale is approximately 1.7 times as important to patients as the same improvement in resting pain. CONCLUSIONS: Confounding among PRO domains can interfere with estimating valid preference weights and limit researchers’ ability to populate a model based on PRO outcomes with a full set of preference weights Controlling statistically for implausible outcome combinations successfully corrects the effect of the confounding and yields plausible trade-off values. PMS51 PATIENT PREFERENCES FOR SEVERE OSTHEOPOROSIS IN SPAIN: A DISCRETE CHOICE EXPERIMENT Darba J1, Restovic G2, Kaskens L2, Martinez E3 1 Universitat de Barcelona, Barcelona, Spain; 2BCN Health, Barcelona, Spain; 3Nycomed Pharma, Madrid, Spain OBJECTIVES: This study was designed to identify patient preferences for different aspects of severe ostheoporosis (SO) treatments in Spain. METHODS: Main attributes of SO treatments were determined from a review of the literature and consultations with nurses and patients. The discrete choice experiment included 3 attributes: type of drug administration, place of administration, plus a cost attribute in order to estimate willingness to pay (WTP) for improvements in attribute levels. a pilot study with 50 patients was carried forward to identify areas of misunderstanding. One hundred sixty-six patients with a diagnosis of SO were presented with pairs of hypothetical treatment profiles with varied levels of type of administration, place of administration and cost. Questions were also included to collect socio-demographic data. Data were analyzed using a random effects probit model. RESULTS: All attributes had the expected polarity and all were significant predictors of choice. Patients were willing to pay c183/month to have a subcutaneous injection once per day rather than an intravenous injection once per year. Patients were also willing to pay c121/month to have medical support when administering the drug treatment at home rather than being admitted several hours to a hospital for drug administration. CONCLUSIONS: Spanish SO patients have well-defined preferences among treatment attributes and are willing to accept tradeoffs among attributes. Participants indicated that they are willing to accept self medication with medical support rather than being admitted to a hospital for several hours. The perspective of the patients should be taken into account when making treatment decisions. PMS52 PREFERENCES OVER BENEFITS AND RISKS ASSOCIATED WITH THE USE OF NSAIDS: EVIDENCE FROM PATIENTS DIAGNOSED WITH OSTEOARTHRITIS (OA) IN UK Bridges JF1, Taylor SD2, Arden N3, Hauber AB4, Johnson FR4, Watson D2, Mavros P2, Pellissier JM2, Peloso P5, Sen S2, Mohamed A4, Gonzalez JM4 1 Johns Hopkins University, Baltimore, MD, USA; 2Merck & Co., Inc., Whitehouse Station, NJ, USA; 3University of Southampton, Southampton, UK, UK; 4RTI Health Solutions, Research Triangle Park, NC, USA; 5MRL, Whitehouse Station, NJ, USA BACKGROUND: The premise of patients’ informed consent presupposes a preference structure over treatment related outcomes. OBJECTIVES: To assess patient preferences for treatment-related benefits and risks associated with the NSAID use in the management of OA and examine differences in patient preferences across patient subgroups. METHODS: Eligible study participants, identified in a chronic disease 13th Euro Abstracts panel in UK, were >44 yrs old with self-reported diagnosis of OA. Each patient was randomized to receive one of four blocks of choice questions, each block consisting of 10 choice tasks comparing different treatment profiles of benefits and risks consistent with NSAID use. Treatment profiles were defined by four benefits (ambulatory pain, resting pain, stiffness, difficulty doing daily activities) and four medicationrelated risks (bleeding ulcer, stroke, heart attack, hypertension), each varying across four clinically meaningful levels. Preference weights were estimated using mixed-effects logistic regression models and were standardized in a 0–10 (low-high) importance scale. Differences in patient preferences were assessed by stratifying age, hypertension diagnosis, and use of proton pump inhibitors (PPI). RESULTS: For the 294 study participants, average age was 59 years, 65% were female, 62% married, 56% diagnosed with OA > 4 years, 49% with hypertension, 76% on prescription OA medications and 36% on PPIs. Reductions in ambulatory pain (6.32; 95%CI:5.0–7.6) and difficulty doing daily activities (6.32; 95%CI:5.0–7.6) were the most important benefit attributes followed by resting pain (2.80; 95%CI:1.8–3.8) and stiffness (2.65; 95%CI:0.9–4.4). Incremental changes (3 percentage-point) in the risk of heart attack or stroke were assessed as the most important risk outcomes (10.00; 95%CI:8.2–11.8; and 8.90; 95%CI:7.3–10.5, respectively). a 2.5 percentage-point incremental change in one-year ulcer risk (3.61; 95%CI:2.6–4.6) and the risk of hypertension (3.02; 95%CI:2.8–3.2) were valued less. The hypothesis of preferences homogeneity in patient subgroups was not rejected. CONCLUSIONS: Patients diagnosed with OA have well defined preferences over NSAID-related benefits and risks. PMS53 QUALITY OF LIFE, PATIENTS’ PREFERENCES AND INFORMED DECISION IN TREATING RHEUMATOID ARTHRITIS IN GREECE Karageorgopoulou K1, Seretis G1, Papachristou E1, Antoniadis C2, Aslanidis S3, Boki K4, Boura P5, Boumpas D6, Ioakeimidis D7, Kandili A8, Sakkas L9, Settas L10, Solioti F11, Tzanakakis M6, Tzioufas A12, Vassilopoulos D13, Tountas Y1 1 Centre for Health Services Research, Department of Hygiene and Epidemiology, Medical School, University of Athens, Athens, Greece; 2Department of Rheumatology, Asklepieion Voulas Hospital, Voula, Greece; 3B’Propaedeutic Department of Internal Medicine, Rheumatology, Hippokration Hospital of Thessaloniki, Thessaloniki, Greece; 4Department of Rheumatology, Sismanoglio General Hospital, Athens, Greece; 5Clinical Immunology Unit, 2nd Department of Internal Medicine, Hippokration Hospital of Thessaloniki, University of Thessaloniki, Thessaloniki, Greece; 6Department of Rheumatology, Clinical Immunology and Allergy, Medical School, University of Crete, Heraklion, Greece; 7Department of Rheumatology, General Hospital of Athens, Athens, Greece; 82nd Department of Internal Medicine, Hippokration General Hospital, Athens, Greece; 9Department of Rheumatology, Thessaly University, School of Medicine and Hospital, Larissa, Greece; 10Rheumatology Unit, 1st Department of Medicine, AHEPA Hospital, University of Thessaloniki, Thessaloniki, Greece; 11Euroclinic Hospital, Athens, Greece; 12Department of Pathophysiology, Medical School, University of Athens, Athens, Greece; 13Hippokration General Hospital, Medical School, University of Athens, Athens, Greece OBJECTIVES: Rheumatoid Arthritis (RA) is a seriously debilitating disease affecting almost 0.7% of the adult population in Greece. Treatment options include, besides the conventional treatment with DMARDs, the use of newer biologic agents, which are available in the hospital (intravenous-iv) or the outpatient setting (subcutaneoussc). Patients are usually perceived to be averse to the use of iv treatments that may be essential to control their disease. We assessed quality of life, patients’ preferences, participation in decision making, when selecting the appropriate treatment for RA, and expected outcomes from the prescription medication. METHODS: Structured face-to-face and telephone interviews were conducted among 290 patients taking biologic agents. Patients were recruited from public hospitals (69.7%), office based rheumatologists (22.8%) and private clinics (7.6%). Data were analyzed using SPSS v17.0. RESULTS: Median patient age was 55 years. 3/4 of the sample were women (74.8%). Approximately 1/3 of respondents reported they were not informed at all on adverse events (28.9%), treatment duration (31.4%) and prognosis (24.5%). For the greater part of the patients, the main source of information was their doctor. Treatment decisions were reached by the physician without patients’ involvement in most cases (70.6%). There was no statistically significant difference in patients’ satisfaction between iv and sc treatments, whereas 62% of patients on sc treatment would be willing to switch to iv following their rheumatologist’s suggestion. Finally, “feeling better overall” was rated as the most important expected outcome by 63,2%, followed by reduced joint pain. Only 16% rated the long term remission of the disease as a very important outcome. CONCLUSIONS: An important proportion of patients with RA in Greece feel that they are not informed properly on treatment options, while their expectations may be understated. To improve outcomes, it is essential that rheumatologists overcome perceptions, inform and involve patients in the treatment decision. 13th Euro Abstracts PMS54 GAP BETWEEN PATIENT-REPORTED ESTIMATION ON FRACTURE RISK AND THE WHO ASSESSMENT TOOL (FRAX®) IN POSTMENOPAUSAL OSTEOPOROSIS Pentek M1, Horváth C2, Érsek K1, Baji P1, Bors K3, Csupor E4, Furi J5, Hacsuncz M6, Horváth B7, Kárpáti S8, Boncz I9, Korányi A10, Rápolthy I11, Tamási L12, Tóth E13, Gulácsi L1, Brodszky V1 1 Corvinus University of Budapest, Budapest, Hungary; 2Semmelweis University, Budapest, Hungary; 3Rehabilitation Hospital and Spa of Visegrád, Visegrád, Hungary; 4Budavár Local Government Healthcare Service, Budapest, Hungary; 5Ferencváros Health Care Service, Budapest, Hungary; 6Saint John’s Hospital, Budapest, Hungary; 7Thermal Spa Hospital, Sopron-Balf, Hungary; 8Zugló Health Care Service, Budapest, Hungary; 9University of Pécs, Pécs, Hungary; 10Saint Borbála Hospital, Tatabánya, Hungary; 11Health Care Service Centre II. of Székesfehérvár, Székesfehérvár, Hungary; 12Saint Francis’ Hospital, Miskolc, Hungary; 13Flór Ferenc County Hospital, Kistarcsa, Hungary OBJECTIVES: The FRAX® model recently developed by the WHO calculates 10-year probability of major osteoporotic and hip fracture, offering new basis for medical decision making and health economic analysis in osteoporosis (OP). [http://www.shef. ac.uk/FRAX/]. Patients’ ideas regarding fracture risk and longevity might influence their perception of 10-year fracture risk data and compliance, and as a consequence, successful implementation of FRAX® based thresholds in everyday OP care. Our aim was to study whether FRAX® data differ from OP patients’ and attendant especialists’ self-estimations on fracture risk. We also assessed expectations on longliving which is a crucial point to perceive the 10-year farcture risk projection of FRAX® relevant. METHODS: In 2009 a cross-sectional survey was performed in 10 Hungarian rheumatology centres. Postmenopausal OP patients appearing on routine visit and switching to second line antiporotic drug were involved. Demographics, main clinical characteristics were registered. Patients’ estimation on 10-year fracture risk and longliving were surveyed. Physicians’ expectations on patients’ perspectives were likely detected. FRAX® was calculated and matched with the estimations. RESULTS: 224 patients were involved, mean(SD): age 69.5(8.9)years, duration of OP care 6.7(5.1) years, lumbar Tsc-3.17(0.82), femoral Tsc-2.69(0.87), 133(59.4%) patients had OP fracture previously. Health status VAS was 59(17)mm. FRAX®-major OP fracture was 26(15.7)%, patients estimated 32.7(25.8)%, physicians marked 29.9(21.4)%. Results for hip fracture were: FRAX® 12.6(15.3)%, patients 26.7(25.3)%, physicians 21.5(19.1)% (P < 0.01). Patients’ expected to live until age 82.4(8.2)years, physicians’ estimation was nearly same (82.5, SD7 yrs). Less than 10-year survival was expected by 42,4% and 33.7%, respectively. CONCLUSIONS: Both patients and physicians overestimate fracture risk, especially for hip. Patients often expect to live shorter than the time-frame of the fracture risk projection. These aspects should be highly considered in health communication and also in the application of the FRAX® method in clinical practice. PMS55 ANALYSIS ON THE POSSIBLE ASSOCIATION BETWEEN RHEUMATOID ARTHIRITIS AND DIABETES MELLITUS TYPE 2 Laires P1, Fonseca JE1, Garcia EB2 1 Faculty of Medicine, University of Lisbon, Lisbon, Portugal; 2BioEPI, Clinical and Translational Research Center, Lisbon, Portugal OBJECTIVES: Rheumatoid Arthritis (RA), a chronic inflammatory disease, may predispose to the development of Type 2 Diabetes Mellitus (T2DM). We aimed to compare incidence rates of T2DM in RA versus non-RA populations and study the influencing factors. METHODS: The study population consisted of participants in the National Data Bank for Rheumatic Diseases (NDB), where patients with rheumatologic disorders completed semiannual questionnaires, from 1998 through 2008. Osteoarthritis (OA) patients were used as controls. T2DM was determined based on self-reports of disease and on the use of hypoglycemic medication. The association between RA and T2DM was investigated using COX logistic regressions adjusted for relevant clinical and demographic covariates. RESULTS: A total of 14,481 participants diagnosed with RA (79.5% female; mean age 58.1 years) and 3,441 participants diagnosed with OA (84.5% female; mean age 63.6 years) were followed during 69,943 person-years. RA subjects had lower Body Mass Index (BMI) and less major comorbidities than controls. In patients with RA the T2DM incidence rate was 10 per 1000 person-years, while the incidence rate for OA was 15 per 1000 person-years. Both rates are superior to most estimates calculated among the US population. However, the covariate-adjusted risk of T2DM in patients with RA versus OA was not significant (HR = 0.94; 95% CI: 0.79–1.12, p = NS). Male gender, age, BMI, non-caucasian ethnicity, major comorbidities, low education level and prednisone intake were significantly associated with the incidence of T2DM. Some RA drugs had a clear protective role on TDM2, mainly methotrexate (19% risk reduction; p = 0.022) and hydroxicloroquine (47% risk reduction; P < 0.001). CONCLUSIONS: RA per se is not associated with increased risk of T2DM. Both RA and OA are associated with known TDM2 risk factors, such as BMI and some comorbidities, increasing the incidence rates of TDM2. Some drugs for RA treatment have a significant protective effect on the TDM2 risk. A313 PMS56 DETERMINING THE OPTIMAL TIMING FOR TOTAL KNEE REPLACEMENT Ko Y1, Lo NN2, Yeo SJ2, Yang KY2, Yeo W2, Chong HC2, Thumboo J2 1 National University of Singapore, Singapore; 2Singapore General Hospital, Singapore OBJECTIVES: Total knee replacement (TKR) is a commonly used surgical procedure for patients with severe joint damage caused by arthritis; however, there remain difficulties in establishing criteria to define the optimal timing for TKR. The aim of this study was to identify the preoperative threshold HRQoL scores that were associated with better postoperative outcomes and those associated with greater improvement. METHODS: Data were collected from 1715 patients undergoing TKR between 2001 and 2006. Patients were interviewed at baseline and at 6 months and 2 years after surgery. At all three interview sessions, patients were asked to complete the ShortForm (SF-36) and the Oxford Knee Score (OKS). As physical function and pain have been identified as the key domains in osteoarthritis, the OKS and the physical functioning (PF) and bodily pain (BP) scales of the SF-36 were selected as the outcome measures of this study. Summary statistics were computed for preoperative, postoperative, and improvement in outcome scores. Visual inspection of the simple error bar charts were used to identify the threshold preoperative scores that were associated with better postoperative scores and those associated with more improvement. RESULTS: Patients with poorer preoperative HRQoL had worse postoperative outcomes but experienced greater improvement after TKR compared to those with better preoperative HRQoL. The highest postoperative PF scores are associated with a preoperative PF score of 50 points and above, whereas the baseline threshold OKS score was between 35 and 40 points. a baseline PF score of <30 points and an OKS score of >40 points are associated with the greatest improvement. No clear leveling off pattern was observed in the BP scores. CONCLUSIONS: PF and OKS threshold scores associated with optimal outcomes of TKR were identified in this study. Future research is needed to examine the predictive value of the scores identified in improving patients’ post-operative outcomes. PMS57 HEALTH-RELATED QUALITY OF LIFE AFTER TOTAL KNEE REPLACEMENT OR UNICOMPARTMENTAL KNEE ARTHROPLASTY IN AN URBAN ASIAN POPULATION Ko Y1, Narayanasamy S1, Wee HL1, Lo NN2, Yeo SJ2, Yang KY2, Yeo W2, Chong HC2, Thumboo J2 1 National University of Singapore, Singapore; 2Singapore General Hospital, Singapore OBJECTIVES: To examine health-related quality of life (HRQoL) following total knee replacement (TKR) or unicompartmental knee arthroplasty (UKA). METHODS: Asian adult patients undergoing either TKR or UKA in a hospital of Singapore between 2001 and 2006 were interviewed before surgery and 6 and 24 months postoperatively to obtain demographic information and HRQoL scores using the ShortForm (SF-36) and the Oxford Knee Score (OKS). RESULTS: Data were collected from 2243, 1715, and 1113 patients at baseline, 6, and 24 months, respectively. TKR patients had a lower preoperative score than UKA patients on OKS and four subscales of the SF-36 (p < 0.01). Both TKR and UKA patients’ OKS and SF-36 subscale scores improved six months postoperatively except in the general health domain. SF-36 role physical (RP) and bodily pain (BP) scores showed the most improvement (40.9 and 33.0 points in TKR and 36.9 and 31.4 points in UKA patients, respectively). The most substantial improvements between baseline and two years after surgery were in the physical domains of HRQoL (RP, BP, and physical functioning (PF)). In addition, in both groups, five domains of SF-36 (RP, BP, PF, social functioning, and role limitations due to emotional problems) reached the proposed minimal clinically important difference of ten points. TKR patients’ SF-36 and OKS scores were not significantly different from those of UKA patients two year after surgery, except PF scores. Multiple regression analysis adjusting for sociodemographics showed that baseline scores were a significant predictor of the postoperative scores of OKS and all SF-36 subscales (p < 0.01), whereas the type of surgery was not associated with the postoperative scores. CONCLUSIONS: Both TKR and UKA patients experienced significant improvements in HRQoL, particularly in the RP and BP domains. After controlling for potential confounding variables, the type of surgery was not a significant predictor of patients’ postoperative HRQoL scores. PMS58 CLINICAL OUTCOMES AND CHANGES IN QUALITY OF LIFE IN WOMEN WITH OSTEOPOROSIS TREATED WITH TERIPARATIDE: 36 MONTH RESULTS OF THE FRENCH PATIENTS PARTICIPATING IN THE EUROPEAN FORSTEO OBSERVATIONAL STUDY (EFOS) Rajzbaum G1, Tcherny-Lessenot S2, liu-Leage S2, Gehchan N2, Barrett A3 1 Saint Joseph Hospital, Paris, France; 2Lilly France, Suresnes, France; 3Eli Lilly & Company Ltd, Windlesham, Surrey, UK OBJECTIVES: To describe back pain, and HRQoL in postmenopausal women with osteoporosis treated with teriparatide (Forsteo®) in France. METHODS: European, prospective, observational study of 36 months duration (18-month active treatment phase and 18-month post-treatment follow up) in postmenopausal women with osteoporosis who initiated teriparatide. HRQoL measured by EQ-5D and back pain measured by a Visual Analogue Scale (VAS) and a questionnaire were collected at each follow-up visit. RESULTS: Among 309 patients enrolled in France (18.8% of the total EFOS cohort), 290 (94.5%) had any follow-up data, and 201 (65.0%) completed the last post-treatment visit. At baseline, patients were 73.8 (7.4) years old (mean (SD)), A314 98.6% of patients had 2 or more fractures after age 40 years, the mean (SD) number of fractures was 4.2 (1.7) and median number was 4.0 (interquartile range 3.0–5.0). During the study, the mean (SD) duration of treatment by teriparatide was 443 (203) days; at the end of 17th month, 67.9% of patients were still on treatment. The main reasons of treatment discontinuation were treatment completion (69.9%), adverse events (14.5%), patient decision (14.0%), and physician decision (2.2%). Between baseline and end of study, the rate of women with back pain decreased from 93.8% to 83.3% and 57.1% had an improvement in the severity, mean (SD) back pain intensity VAS decreased from 55.9 (24.8) to 35.0 (24.2), and mean (SD) EQ-5D VAS increased from 52.6 (19.4) at baseline to 57.8 (21.4) at end of study. CONCLUSIONS: French patients with severe osteoporosis treated with teriparatide in a routine setting had an increase in quality of life and a decrease in back pain during the teriparatide treatment period and post-treatment follow-up. The results should be interpreted in the context of a non-controlled observational study. PMS59 THE EFFECT OF TNF THERAPY, SOCIODEMOGRAPHIC AND CLINICAL FACTORS ON SLEEP DISTURBANCES AND FATIGUE AMONG RHEUMATOID ARTHRITIS—RESULTS FROM THE NDB-PORTUGAL COHORT Marques R1, Chaves I1, Vasconcelos J1, Pedro S1, Rodrigues A1, Michaud K2, Wolfe F3, Garcia E1 1 BioEPI, Clinical and Translational Research Center, Oeiras, Portugal; 2University of Nebraska Medical Center, Omaha, NE, USA; 3National Databank for Rheumatic Diseases, Wichita, KS, USA BACKGROUND: The prevalence of sleep disturbances and fatigue among patients with rheumatoid arthritis (RA) is high. TNF therapy reduces disease activity and disability in RA, but few studies have analyzed the impact of TNF vs. traditional DMARD therapy, sociodemographic and clinical factors on sleep and fatigue, in prospective cohorts. OBJECTIVES: We assessed the effect of TNF therapy, sociodemographic and clinical factors on sleep disturbances and fatigue in RA patients. METHODS: A total of 1,082 RA patients from the NDB-Portugal cohort participated in this prospective study. Patients´ last observation was used. Univariate (UV) and multivariate (MV) linear regression models (β, 95% CI) assessed the impact of the following on sleep disturbances (measured by the sleep disturbance scale (VAS 0–10, 10 is worst) and the insomnia severity index (0–28, 28 is worst)) and fatigue (fatigue scale (VAS 0–10,10 is worst)): traditional DMARD and TNF therapy, age, sex, education level, marital status, number of major comorbidities, RA duration, disability (HAQ 0–3, 3 is worst), quality of life (VASQOL 0–1, 1 is better), emotional distress (Hospital Anxiety and Depression Scale—HADS 0–21, 21 is worst) and prednisone use. RESULTS: In MV, TNF therapy seemed to decrease fatigue (−0.60 (−1.08, −0.11)) when compared to traditional DMARDs and although not statistically significant, decreased sleep disturbances (−0.46 (−1.04, 0.13)) and insomnia (−0.22 (−1.16, 0.73)). An increase in sleep disturbances was seen with worse HAQ (0.64 (0.25, 1.02)), lower VASQOL: (−1.46 (−2.63, −0.28)), more anxiety symptoms: (0.19 (0.09, 0.29)) and higher fatigue (0.26 (0.17, 0.36)). These results were also seen for insomnia and fatigue. More sleep disturbances (0.18 (0.12, 0.25)) increased fatigue and higher fatigue increased insomnia (0.39 (0.23, 0.54)). CONCLUSIONS: We found that the use of TNF therapy improved fatigue when compared to traditional DMARDs. Higher disability and worse quality of life increased sleep disturbances, insomnia and fatigue. PMS60 TNF THERAPY REDUCES THE ODDS OF WORSENING DISABILITY TRENDS IN RHEUMATOID ARTHRITIS OVER AT LEAST 2 YEARS—DATA FROM THE NDB-PORTUGAL COHORT Pedro S1, Vasconcelos J1, Marques R1, Chaves I1, Rodrigues A1, Michaud K2, Wolfe F3, Garcia E1 1 BioEPI, Clinical and Translational Research Center, Oeiras, Portugal; 2University of Nebraska Medical Center, Omaha, NE, USA; 3National Databank for Rheumatic Diseases, Wichita, KS, USA OBJECTIVES: Many studies have identified predictors of the health assement questionnaire (HAQ), but few have evaluated the predictors of HAQ trends among RA patients. To investigate the predictors of worsening disability trends compared to other patterns among RA patients over at least 2 years. METHODS: A total of 646 RA patients from the ongoing biannual NDB-Portugal cohort with at least four consecutive HAQ scores per patient during their follow-up were used.The proportion defined by the number of 6-month positive increments in HAQ scores (worsening function) divided by the total number of differences was computed per patient and used to define a patient’ trend. The outcome was then defined as the presence of a trend of worsening disability (when proportion >0.5). This meant that a patient’s tendency of worsening was higher than their tendency of improving during their own follow-up. Univariate (UV) and multivariate (MV) generalized estimating equations (GEE) were used to study the predictors of a worsening disability trend. Age, education, disease duration, paid work, retirement, number of total major comorbidities, SF-36 mental component, RADAI, the VAS scales of sleep, fatigue and pain, the use of current TNF (with or without concomitant DMARDs) vs. traditional DMARD therapy and steroids, were used as possible predictors. RESULTS: A total of 26% patients had worsening disability trends. The UV analyses showed that all of the following factors were statistically relevant: age, educational level, number of major comorbidities, sleep disturbances and fatigue, RADAI and the use of TNF therapy. The final MV model included pain (OR: 1.003 (95%CI: (1.000; 1.005))), age (OR: 1.02 (1.01; 1.02)) and the use of TNF 13th Euro Abstracts (OR: 0.94 (0.91; 0.97)). CONCLUSIONS: In our study, we showed that older age and more pain predicted worsening HAQ disability trends. The use of TNF therapy was the only factor that decreased the odds of having a worsening HAQ trajectory. PMS61 RELATING OSTEOARTHRITIS AFFECTATION, FUNCTIONAL DISABILITY AND QUALITY OF LIFE: A STRUCTURAL EQUATION MODEL. THE EXPECT STUDY Cordero J1, Darder A2, Santillana J3, Caloto T4, Nocea G4, Sanchez I4 1 Hospital Universitario La Princesa, Madrid, Spain; 2Hospital Arnau de Vilanova, Valencia, Spain; 3Hospital Verge de la Cinta, Tortosa, Spain; 4Merck Sharp & Dohme, SA, Madrid, Spain OBJECTIVES: To establish a conceptual model which relates osteoarthritis (OA) affectation, functional disability in daily activities, and Quality of Life (QoL). METHODS: The present is an observational, cross-sectional, multicenter study. OA presence/absence, by location, was clinically recorded. Disability was assessed with the Health Assessment Questionnaire Disability Index (HAQ-DI), QoL through the EuroQoL-5D questionnaire. Descriptives were used for sociodemographic and clinical variables; relationship between OA, disability and QoL was estimated through Structural Equation Modeling (SEM). This multivariate analysis technique allows to hypothesize multiple relationships among latent, unobserved variables and tests the model with a equation system. RESULTS: A total of 965 OA patients were included [mean age = 64 years (SD = 11); 75% women]. Mean body locations affected by OA was 2.81 (median = 2; SD = 1.613). The most frequently affected locations were knees (67% of the patients), lumbar (60%) and cervical (45%) spine. Regarding EuroQoL5D, most patients reported not having severe problems in the five areas assessed. ‘Other activities’ (mean = 1.172; SD = 0.957) and ‘reach’ (mean = 1.127; SD = 0.912) were the HAQ-DI categories that showed higher disability. The SEM presented OA, disability and QoL as latent, related variables. 92% of QoL was accounted for disability (R-squared = 0.92). The global model that depicts OA as causing disability, and disability affecting QoL, has a marginal adjustment (CMIN/DF = 5.42; RMR = 0.026; RMSEA = 0.069). CONCLUSIONS: With the available data, the functional disability can account for the decrease in QoL. Theoretically, OA is strongly related with disability and QoL, but the model fails to fully explain this link. As statistical techniques need good measurement models to correctly estimate relationships, standard clinical records seem insufficient for this purpose. Additional valid measurements of OA affectation would be needed, to give evidence of its direct effect on disability and QoL. PMS62 CERTOLIZUMAB PEGOL MONOTHERAPY PROVIDES SUSTAINED IMPROVEMENTS IN HOUSEHOLD PRODUCTIVITY AND DAILY ACTIVITIES IN PATIENTS WITH ACTIVE RHEUMATOID ARTHRITIS OVER 2 YEARS Strand V1, Purcaru O2, van Vollenhoven R3, Choy E4, Fleischmann R5 1 Stanford University, Portola Valley, CA, USA; 2UCB, Brussels, Belgium; 3Karolinska University Hospital, Stockholm, Sweden; 4King’s Musculoskeletal Clinical Trials Unit, King’s College, London, UK; 5University of Texas, Dallas, TX, USA OBJECTIVES: To evaluate the impact of certolizumab pegol (CZP) monotherapy on household work and daily activities in RA patients over 2 years. METHODS: Patients in the FAST4WARD Phase III trial were randomised to CZP 400 mg administered every 4 weeks (Q4W) or placebo for 24 weeks. Those who completed or withdrew at/after Week 12 were eligible to enter an open-label extension (OLE) study of CZP 400 mg Q4W. This analysis focuses on CZP completers who entered the OLE study and had 2 years (100 weeks) of CZP exposure from baseline (BL). Household productivity and impact on family/social/leisure activities were assessed using the validated Work Productivity Survey (WPS-RA). Analyses were conducted on observed data. FAST4WARD:NCT00548834; OLE:NCT00160693. RESULTS: Sixty-nine CZP completers entered the OLE. At BL: mean disease duration: 9.5 years; mean HAQ-DI: 1.42; mean DAS28-3(CRP): 5.76. Burden of RA on household productivity at BL was substantial: mean 10.1 household work days missed/month, mean 12.1 household work days with reduced productivity/month, mean 5 days missed/month of family/social/leisure activities. At Week 100, compared with BL, patients receiving CZP monotherapy reported on average fewer household work days missed per month (1.0 vs. 10.1), fewer days with reduced productivity in the home (1.1 vs. 12.1), reduced interference of RA on household productivity (2.0 vs. 5.8 on a 0–10 scale), fewer missed days of family/social/leisure activities (0.3 vs. 5.0). Improvements were seen as early as Week 4 and were sustained until Week 100. Over 12, 52 and 100 weeks, mean annualised cumulative gains from BL were 20.5, 108.4 and 199.3 household work days, respectively, 25.1, 136.0 and 244.9 more productive days within the household, and 11.9, 57.7 and 107.2 days gained of family/social/leisure activities. CONCLUSIONS: CZP 400 mg Q4W monotherapy provides sustained improvement in productivity within the home and in RA patients’ abilities to engage in family/social/ leisure activities. 13th Euro Abstracts MUSCULAR-SKELETAL DISORDERS – Health Care Use & Policy Studies PMS63 THE IMPACT OF THE DRUG CATEGORIZATION ON THE PRESCRIPTION AND CONSUMPTION OF NONSTEROIDAL ANTIINFLAMATORY DRUGS (NSAID) Gatialová K, Petrová L, Bellová K, Majtás J, Foltan V Faculty of Pharmacy, Comenius University, Bratislava, Slovak Republic OBJECTIVES: This study deals with the impact of regulatory measures represented by the drug categorization on the prescription and at the use of nonsteroidal and antiinflammatory drugs (NSAID) and asses measures of the state health policy and their application in the Slovak Republic within year 2009. METHODS: Analysis was conducted with key data represented by the medicine name and code, name of registration decision owner, maximum price of producer and importer, final medicine price, maximum copayment, prescription and indication restriction, the number of medicine packages. As data sources were used The List of Drugs. Analysis of categorizations in ATC group of drugs M01- NSAID, comparison of single changes and its significance. RESULTS: In year 2009 were 9,087,217 medicine packages from the group of NSAID expended. From this amount of medicine packages were 2,477,352 packages expended during first quartal of 2009, which represents the highest consumption of all four terms. The lowest consumption was noticed in period of second term reaching expenditure of only 2,020,514 medicine packages. Outstanding changes appeared by indometacine where prescription decreased from 9782 medicine packages to zero. Lornoxicam has appeared in third term as new medicine in categorization and his consumption increased to 3091 medicine packages in the fourth term. In all four terms dominated the same drugs on first ten places of medicine consumption (ibuprofen, diclofenac, nimesulid, combinations of ibuprofen, meloxicam, ketoprofen, flurbiprofen, piroxicam, naproxen, aceclofenac). CONCLUSIONS: Current state of categorization and functioning of professional bodies represent initial assumption of health policy asserting. Patient registries are helping tool in assessing effectiveness of medicines and cost efficacy and should be legislated. There is a disadvantage of administrative burden. Patient co- payment has increasing tendency due to removing responsibility for own health on patients. PMS64 CONSUMPTION OF BISPHOSPHONATES FOR THE TREATMENT OF OSTEOPOROSIS Vytrisalova M, Blazkova S Charles University, Faculty of Pharmacy, Hradec Kralove, Czech Republic OBJECTIVES: The objective was to analyze trends in the consumption of bisphosphonates for the treatment of osteoporosis in 2002–2006. METHODS: The prescription-based database of the General Health Insurance Company of the Czech Republic (VZP CR), the largest health insurance company of the Czech Republic that covers about 65% of the Czech population, was used as the source of data from 2002–2006. Health insurance is compulsory under Czech law. In 2002–2006, alendronate and risedronate (in both once-daily and once-weekly formulations) were available for the treatment of osteoporosis. The prescription of the two bisphosphonates is restricted to specialists in internal medicine, orthopaedy, rheumatology, gynecology and endocrinology. RESULTS: In 2002–2006, as many as 63 251 (0.94%) and 21 064 (0.31%) patients (insured persons) refilled at least one prescription for alendronate and risedronate, respectively. Men accounted for 3% of the bisphosphonate consumption. The age of bisphosphonate consumers was 69 years (median). Bisphosphonates were prescribed most often by physicians specializing in internal medicine, orthopaedy and rheumatology. The mean alendronate and risedronate consumption rates were 5.2 and 1.5 defined daily doses (DDD)/1000 insured persons/day, respectively. Mean DDD per year/patient was higher in women than in men (204 versus 186 for alendronate and 179 versus 172 for risedronate). The number of patients on alendronate increased by 83% from 2002 to 2006. CONCLUSIONS: Signals of inadequate osteoporosis treatment in the Czech population in 2002–2006 were found. The consumption of bisphosphonates is particularly low in men. The increasing consumption rates of alendronate could be considered as a positive trend. The project was fully supported by grant No. 103107 (2007–2009) from the Charles University Grant Agency. PMS65 RETROSPECTIVE OBSERVATIONAL DATABASE ANALYSIS OF TNF-Á INHIBITOR SWITCHING PATTERNS IN PATIENTS IN THE UNITED STATES WITH RHEUMATOID ARTHRITIS IN TWO DISTINCT “NATURALISTIC” TREATMENT SETTINGS USING A PRE-PROGRAMMED DATA ANALYSIS TOOL Chiappinelli R1, McNeeley B1, Byrd J2, Ollinger E2 1 HealthCore, Wilmington, DE, USA; 2Dymaxium, Inc, Toronto, ON, Canada OBJECTIVES: Evaluate 12 month switch patterns among patients taking anti-TNFs for rheumatoid arthritis (RA) in the context of validating a Rheumatoid Arthritis Outcomes Analyzer (RAOA); a data analysis tool incorporating pharmacy, medical claims, and member eligibility information. METHODS: The study was conducted utilizing claims data from a two large commercial data sets: Cohorts #1 and #2. Facilitated by the RAOA, medical and pharmacy claims were entered into two distinct data sets for inclusion in the analysis. Patients were ≥18 years of age, received ≥ one traditional (non-biologic) or biologic DMARD between January 2005 and December 2007, and had ≥ two RA diagnoses (ICD-9 CM 714.0X) ≥ two months apart. For the switch analysis, patients had at least 18 months continuous eligibility; 6 months prior to index date (initial anti-TNF) and treatment naïve, and 12 months post. RESULTS: A315 A total of 2177 (Cohort #1) vs. 1113 (Cohort #2) patients entered the analysis. In Cohort #1, 426 (19.6%) received adalimumab, 1123 (51.6%) received etanercept and 628 (28.8%) received infliximab as initial treatment. In Cohort #2, 355 (31.9%) received adalimumab, 509 (45.7%) received etanercept and 249 (22.4%) received infliximab. In both Cohorts, 75% were female. During the twelve months following the index date, 161 (7.4%) vs. 91 (8.2%) switched to another biologic DMARD; 133 (6.1%) vs. 87 (7.8%) switched to another anti-TNF biologic and 28 (1.3%) vs. 4 (0.4%) to a non anti-TNF biologic DMARD. Twenty-seven (6.3%) vs. 44 (12.4%) of patients in the adalimumab sub-group switched to an anti-TNF biologic compared to 99 (8.8%) vs. 37 (7.3%) for etenercept and 7 (1.1%) vs. 6 (2.4%) for infliximab. CONCLUSIONS: Analytic tools such as the RAOA will allow payers and policy makers to better understand utilization and treatment patterns easily and quickly. Replication and validation of outputs from these tools are important to establish the precision of results. PMS66 PENETRATION OF MONOCLONAL ANTIBODIES ONTO FINNISH PHARMACEUTICAL MARKET Kannisto HE1, Jormanainen V1, Happonen P2 1 Finnish Medicines Agency (Fimea), Helsinki, Finland; 2Finnish Medicines Agency (Fimea), Kuopio, Finland OBJECTIVES: Monoclonal antibodies (MAB) are biomedicines used for treatment of cancer, rheumatoid arthritis, psoriasis and inflammatory intestinal diseases. Since the first market entry in 1996 in Finland, their sales have increased continuously. We describe the penetration of MABs onto the Finnish pharmaceutical market in 1996– 2009. METHODS: The sales of MABs (wholesale figures in million euros, c M) with marketing authorization were extracted by calendar year from the market database SLD Pharma (Finnish Pharmaceutical Data Ltd.) based on active substances. Annual figures were converted into 2009 values with the annually-updated Finnish National Pensions index (Social Insurance Institution). RESULTS: In 2009, the sales of the 17 MABs were c109 M or 5.6% of the total pharmaceutical market value (c1900 M). On the hospital market (73% of MAB sales in 2009), however, the respective share was 16% following a steady increase in 1999–2009. Up to the end of 2009, the overall cumulative MAB sales exceeded c479 M. Infliximab showed the highest cumulative sales of c120 M (entry in 1999). Adalimumab penetrated the market even more rapidly, with cumulative sales of c109 M (entry in 2003), and became the most-sold active substance in 2009 in Finland. The third MAB exceeding cumulative sales of c100 M was rituximab (entry in 1998). In terms of first-year sales, adalimumab was the most successful MAB, amounting to over c6 M. Bevacizumab achieved the secondhighest first-year sales (c1.8 M in 2005), while the third MAB exceeding c1 M sales was ranibizumab (c1.7 M in 2007). CONCLUSIONS: The penetration of several MABs onto the Finnish pharmaceutical market has been very successful within a short period of time. PMS67 CONJOINT ANALYSIS OF REDUCTION OF CO-PAYMENT RATE OF NATIONAL HEALTH INSURANCE SYSTEM IN JAPAN Igarashi A1, Kikuta K1, Hoshi D2, Tanaka E2, Yamanaka H2, Tsutani K1 1 Grad. Sch. of Pharm, University of Tokyo, Tokyo, Japan; 2Tokyo Women’s Medical University, Tokyo, Japan BACKGROUND: In Japanese national health insurance system, there is some limit for monthly co-payment amount. In some diseases, AIDS, chronic renal failure with dialysis and hemophilia, co-payment limits are additionally reduced. Patient groups for myeloma, cervical cancer, hepatitis, chronic myelogenous leukemia and rheumatoid arthritis argued that co-payment reduction should be implemented to their diseases. OBJECTIVES: To conduct conjoint analysis to determine how people value various factors of diseases in decision-making process for co-payment reduction. METHODS: We determined six factors, i.e., influence for life-years/QOL (low/high), availability for medicines (yes/no), number of patients (5,000/50,000/500,000), disease duration (short/long) and out-of-pocket expense per 1 month (JPY10,000/ JPY100,000, USD1 = JPY90) and developed questionnaire. Ninety-six patterns are reduced to 26, via orthogonal methods. a total of 1163 participants filled questionnaire via web survey system. We adopted panel-logit model to estimate odds ratios for each factors. RESULTS: All 6 factors significantly influenced peoples’ decisionmaking process. Coefficient for each factors were as follows; out-of-pocket expense: 2.61 > influence for QOL: 0.976 > disease duration: 0.884 > influence for life-years: 0.831> number of patients: 0.105 > availability for medicines: −0.18, respectively. Based on this estimation, when we consider characteristics of anti-rheumatoid biologics, more than 95% people think that out-of-pocket rate should be reduced. CONCLUSIONS: Various factors contribute to people’s attitude for co-payment reduction. This quantitative result would be helpful for decision-making process in national health insurance system. A316 PMS68 PHYSICIANS’ STATED PREFERENCES OVER BENEFITS AND RISKS ASSOCIATED WITH NSAID USE IN PATIENTS WITH OSTEOARTHRITIS IN UNITED KINGDOM Bridges JF1, Taylor SD2, Arden N3, Hauber AB4, Johnson FR4, Watson D2, Mavros P2, Pellissier JM2, Peloso P5, Sen S2, Mohamed A4, Gonzalez JM4 1 Johns Hopkins University, Baltimore, MD, USA; 2Merck & Co., Inc., Whitehouse Station, NJ, USA; 3University of Southampton, Southampton, UK, UK; 4RTI Health Solutions, Research Triangle Park, NC, USA; 5MRL, Whitehouse Station, NJ, USA BACKGROUND: Treatments for symptom control in osteoarthritis (OA) confer varying degrees of benefits alongside medication-related risks. Physicians’ preferences over benefits and risks of NSAIDS are an important aspect of understanding clinical practice. OBJECTIVES: To estimate physicians’ preferences over benefits and risks associated with NSAID use in the management of OA and examine differences in preferences between general practitioners (GPs) and specialists. METHODS: Participating physicians treated at least 10 OA patients per-month. Each physician was randomized to receive one of four blocks of discrete-choice questions; each block consisting of 12 paired choice tasks comparing treatment profiles. Treatment profiles were defined by four benefits (ambulatory pain, resting pain, stiffness, difficulty doing daily activities) and four medication-related risks (bleeding ulcer, stroke, heart attack, hypertension), each varying across four clinically meaningful levels. Elicitation of preferences was facilitated using standardized patient profiles systematically varying by age, co-morbid conditions and clinically relevant risks of NSAIDs. Preference weights were estimated using mixed-effects logistic regression and were standardized on a 0–10 (low-high) importance scale. RESULTS: 477 physicians participated (61% GPs, 39% specialists). Reductions in ambulatory pain and difficulty doing daily activities were the most important efficacy variables (6.45; 95%CI:4.8–8.2) followed by eliminating resting pain (3.18; 95%CI:1.9–4.5) and stiffness (2.79; 95%CI:1.5–4.1). Ambulatory pain was twice as important as resting pain or stiffness (P < 0.05). Risk of heart attack was the most important medication-related risk outcome (10.00; 95%CI:7.6–12.4) followed by stroke (9.42; 95%CI:7.2–11.6), ulcer risk (4.62; 95%CI:3.5–5.7) and hypertension (3.25; 95%CI:3.2–3.4). There were no statistically significant differences in preferences between GPs and specialists. CONCLUSIONS: Ambulatory pain and the incremental risk of heart attack were the most important NSAID-related attributes that influence physicians’ treatment choices. Preferences did not vary between GPs and specialists. The findings confirm that benefit-risk tradeoffs are important aspects in treament selection for OA management. PMS69 ECONOMY OF NSAIDS IN THE MANAGEMENT OF OSTEOARTHRITIS: SODIUM CHONDROITIN SULFATE VS. CONTROL GROUP Taieb C, Pibourdin JM Pierre Fabre, Boulogne, France OBJECTIVES: Describe the potential economy of NSAIDs in a population of patients with osteoarthritis newly-treated with a sodium chondroitin sulphate (SCS) versus a control group. METHODS: The Disease Analyzer database (IMS), which collects medical data from 1240 representative French GPs was used. The control group consists of patient diagnosed with knee or hip osteoarthritis during the observation period but not treated with symptomatic slow acting drugs for osteoarthritis throughout the study period, no during the year before the study or the following year. RESULTS: In total, 944 patients were included, 472 per group. The characteristics of both groups in terms of age, sex, time since diagnosis and type of osteoarthritis are strictly the same. 80% of patients included suffer from osteoarthritis of the knee. a total of 53.4% of patients included in study received one or more prescriptions for NSAIDs during the year prior to their inclusion. Half of the patients in the SCS group received at least one NSAID prescription during initiation or during the 1 year followup period. This % is significantly higher in the control group (64%; P < 0.01). 18% of patients in the SCS group stopped their treatment with NSAIDs at the initiation of SCS and did not resume it during the follow-up year, versus 11% in the control group. This difference is significant (p = 0.01). Patients in the SCS group require significantly fewer days of treatment with NSAIDs expressed in DDD than patients in the control group: on average 49 days of treatment versus 64 (p = 0.01). CONCLUSIONS: One of the public health goals set by health authorities—to reduce the incidence of iatrogenic complications (serious bleeding or gastroduodenal events) by 20% in osteoarthritic patients could be attained by a prescription of SCS. Indeed, this study highlights the fact that, in actual use, patients in the SCS group are significantly less likely to use NSAIDs (−22%). PMS70 THE MANAGEMENT OF OSTEOARTHRITIS: ECONOMY OF NSAIDS Taieb C, Pibourdin JM Pierre Fabre, Boulogne, France OBJECTIVES: To describe the use NSAIDs in a population of patients with osteoarthritis newly-treated (2009) by a symptomatic slow acting drugs for Osteoarthritis (SYSADOA) METHODS: The Disease Analyzer database (IMS), which collects medical data from 1240 representative French GPs was used. Patients over 50 y of age were included over a period of one year where a diagnosis of osteoarthritis associated with the initiation of an SYSADOA prescription was identified in the DA database (patient had been monitored for at least 6 months). RESULTS: In total, 3141 patients were included in the study. The patient profile is similar regardless of the treatment group considered (68% female, average age of 66, 50% patients the diagnosis dates 13th Euro Abstracts from less than 1 year). The use of NSAIDs at the outset does not significantly differ among the groups of patients (25%). About 30% of patients discontinue their NSAID therapy upon initiation of treatment with SYSADOA. This proportion did not vary significantly with the SYSADOA considered. More than 40% of patients discontinued treatment with NSAIDs during the 6-month follow-up period. 20% of patients who continued treatment with NSAIDs increased the dosage or duration of the NSAID therapy. About 45% of patients benefited from a co-prescription of analgesics during the initiation of treatment with SYSADOA. This rate did not significantly vary among groups of patients. 21% of patients discontinued their treatment with analgesics after initiation of treatment with SSAAO. This proportion did not vary significantly with the SYSADOA considered. CONCLUSIONS: It is generally accepted that an economy of 20% of NSAIDs has a major impact on public health. Regardless of the prescribed SYSADOA, NSAID interruptions at initiation are approximately 30%, and nearly two thirds of patients with co-prescriptions of NSAIDs at the start of treatment stop or reduce this treatment within 6 months. The cost of complications related to antiinflammatory drugs would be around 150 million Euros. MUSCULAR-SKELETAL DISORDERS – Conceptual Papers & Research on Methods PMS71 INDIRECT TREATMENT COMPARISON TO COMPARE EFFICACY IN HEALTH ASSESSMENT QUESTIONNAIRE (HAQ) SCORE FOR BIOLOGIC AGENTS WITH METHOTREXATE IN PATIENTS WITH RHEUMATOID ARTHRITIS AND ACTIVE DISEASE DESPITE METHOTREXATE THERAPY Lebmeier M1, Pericleous L1, Guyot P2, Baig H1, Christensen R3, Bergman G2, Taylor PC4, Drost P5 1 Bristol-Myers Squibb Pharmaceuticals Ltd, Uxbridge, Middlesex, UK; 2Mapi Values, Houten, The Netherlands; 3The Parker Institute: Musculoskeletal Statistics Unit (MSU), Copenhagen, Denmark; 4Imperial College London, London, UK; 5Bristol-Myers Squibb, Braine-l’Alleud, Belgium OBJECTIVES: To compare the efficacy in terms of HAQ score between abatacept and other biologic Disease Modifying Anti-Rheumatic Drugs (DMARDs) in patients with rheumatoid arthritis taking concomitant methotrexate (MTX) who have inadequate response to MTX (MTX-IR). METHODS: A systematic literature review identified controlled trials investigating the efficacy of abatacept (3 studies), etanercept (2), infliximab (3), adalimumab (2), certolizumab pegol (2) ritixumab (2), and tocilizumab (1) in MTX-IR patients. The identified trials were comparable in design, included patients, and concomitant treatment (MTX). Mixed treatment comparison analyses were performed on HAQ change from baseline (CFB) at 24 and 52 weeks. Results were expressed as difference in HAQ CFB score between treatments and expected HAQ CFB and the 95% Credible Interval (CrI) per treatment at 24 and 52 weeks. RESULTS: The analysis of HAQ CFB at 24 weeks showed that abatacept/MTX is more efficacious than MTX monotherapy (−0.30, 95%CrI: −0.40; −0.19) and shows small numeric differences versus other biologics/MTX (range: −0.11:0.08). The expected mean HAQ CFB at 24 weeks for abatacept (−0.58) was superior to placebo (−0.28) and comparable to all the alternative treatments (adjusted mean between −0.47 and −0.66). The findings at 52 weeks are in line with those at 24-weeks, although no data was available for tocilizumab and golimumab. Scenario analyses confirmed the robustness of the findings. CONCLUSIONS: All biologic DMARDs in combination with MTX in the treatment of MTX-IR patients resulted in improvements from baseline in HAQ score compared to MTX monotherapy at 24 and 52 weeks. All biologic DMARDs in combination with MTX are expected to result in a comparable improvement in HAQ score. PMS72 INTERACTIVE ELECTRONIC INTERFACES (IEI): BRIDGING THE COMMUNICATION GAP BY TRANSLATING ECONOMIC ANALYSIS RESULTS TO DECISION-MAKERS EVERYDAY PRACTICE Marinho M1, Schiola A2, Pimentel A2, Santoni N2, Teich V1, Clark LGO3 1 MedInsight-Evidências, Rio de Janeiro, RJ, Brazil; 2Bayer Healthcare, São Paulo, SP, Brazil, 3 MedInsight-Evidências, Campinas, Brazil OBJECTIVES: Interfaces are interactive dashboards built with programs like Visual Basic for Applications, Crystal Xcelsius or Java. We aimed to demonstrate how IEI allow the translation of pharmacoeconomic studies’ results into understandable projections to decision makers. We present the economic evaluation of rivaroxaban in the prevention of thromboembolic events as a case study. METHODS: A model evaluating rivaroxaban in patients undergoing total knee and hip replacement was used. The model was adapted to different decision makers needs, namely Health Maintenance Organizations (HMO), hospitals, and physicians focused only on clinical outcomes. For each perspective, the following parameters could be customized: state taxes, time horizon of the analysis, choice of comparator (enoxaparin, dabigatran or both), duration of hospitalization, unit costs (drugs, treatment and diagnosis resources), eligible population and market share of comparators over the following five years. RESULTS: The IEI design for this case demonstrated that, under the perspective of an HMO with 200.000 lives, considering a base case where 80% of patients are treated with enoxaparin and 20% with dabigatran, and replacing every year 10% of enoxaparin cases with rivaroxaban would result in a budget impact of (-R$46,155) in 4 years for knee and hip replacement cases. The potential impact for cost offsets for the whole private system would be of (-R$8.3 million). Under the perspective of 13th Euro Abstracts a hospital performing 50 knee and 50 hip surgeries every year, this potential for costoffsets would be of (-R$108,850). This would vary a lot for the perspective of HMOs and hospitals of different sizes, as well as depending on the costs of each decision maker, demanding constant customization of parameters. CONCLUSIONS: The use of IEI makes the pharmacoeconomic studies results more intelligible to decision makers, permitting them to foresee their actual application on a real practice, in different scenarios. PMS73 LINKAGE OF ADMINISTRATIVE AND MEDICAL RECORDS DATABASES FOR INVESTIGATING PHARMACEUTICAL USE AND OUTCOMES Huse DM1, Bizier R1, Tomic K2 1 Thomson Reuters, Cambridge, MA, USA; 2Thomson Reuters, Washington, DC, USA BACKGROUND: Use of electronic health care databases to examine pharmaceutical use and effects faces limitations inherent in the source of data. Medical records document prescribing but not consumption of medication; administrative databases document dispensing but not the original intent of the prescriber. OBJECTIVES: To evaluate a patient-level linked database of electronic medical records (EMR) and administrative claim for use in measuring both prescribing and dispensing of pharmaceuticals. METHODS: Claims data from Thomson Reuters MarketScan database were linked to the GE Centricity EMR database using probabilistic methods to overcome the de-identification required of both databases under US privacy laws. Patient-level records were matched based on demographic characteristics and calendar dates of physician visits. Multiple visits were required to reduce the likelihood of mismatches. The agreement between prescribing records in the EMR and claims for dispensing of medication was explored in the context of self-administered medication for osteoporosis, including bisphosphonates and raloxifene. RESULTS: Using data from 2004– 2009, 219,529 patients were matched between the two data sources. Mean age was 43 years and 57% were female. We identified 2,331 patients whose medical record showed new prescriptions for bisphosphonates or raloxifene (no evidence of use in the prior 180 days). Pharmacy claims indicated these prescriptions were filled by 56% of patients within 7 days, 75% within 30 days, and 86% within 90 days. CONCLUSIONS: It is to be expected that there will be some degree of noncompliance, hence incomplete filling of prescriptions, as observed. The lag from dispensing to prescribing among many patients is also consistent with use of samples for initial prescriptions. Overall, the example of osteoporosis therapy shows consistency between prescribing in EMR data and dispensing in claims data among a sample of probabilistically linked patient records. PMS74 CLAIMS-BASED SEVERITY INDEX FOR RHEUMATOID ARTHRITIS FROM HEALTH CARE CLAIMS DATA Baser O1, Gust C2, Akin C3 1 STATinMED Research/University of Michigan, Ann Arbor, MI, USA; 2STATinMED Research, Ann Arbor, MI, USA; 3STATinMED Research/Brigham and Women’s Hospital, Ann Arbor, MI, USA OBJECTIVES: Controlling for disease severity in observational studies is crucial to get an estimate with no selection bias. However, outcomes research studies using claims data, contain no information about disease severity. Therefore, comorbid scores are used for a proxy for the disease severity. There exists no severity score specific for rheumatoid arthritis (RA). The goal of this study was to develop a severity index for rheumatoid arthritis (SIFRA) for private health care claims data. METHODS: We extracted the following variables related to rheumatoid arthritis from the claims data: total number of synthetic disease-modifying anti-rheumatic drugs (DMARDs), total number of biological DMARDs, tests for C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR) ordered, rehabilitation visits, rheumatology visits, Felty’s syndrome and Sjogren’s syndrome, pulmonary, soft tissue nodules, joint surgery, number of platelet counts and chemical panels ordered, and rheumatoid factors testing. a linear regression model was used to create the severity score. The severity score was compared with the rheumatoid arthritis medical records-based index of severity (RARBIS) and currently-used comorbidity scores to proxy severity in outcomes research studies related with rheumatoid arthritis. RESULTS: According to the Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC), log likelihood function, R-squared values and average squared prediction error, SIFRA performed better than RARBIS, Charlson Comorbidity Score (CCI), Elixhauser comorbidity score and Chronic disease score. Spearman correlation with RARBIS was 0.65 and significant. However, the correlation with the Charlson Comorbidity Index (0.1, p = 0.6521), Elixhauser Index (0.15, p = 0.5312) and Chronic disease score (0.13, p = 0.6011) were low and insignificant. CONCLUSIONS: Comorbidity scores (Charlson, Elixhauser or Chronic Disease Scores) commonly used in outcomes research are inadequate to be proxy variable for RA patients. SIFRA, at least for rheumatoid arthritis, controls for disease severity better than any other commonly used measure. PMS75 MODELING OF RHEUMATOID ARTHRITIS: A METHODOLOGICAL APPROACH Taylor MJ1, Righetti C1, Conway P2, Lebmeier M3 1 University of York, York, UK; 2Wyeth Pharmaceuticals, Maidenhead, UK; 3Bristol-Myers Squibb Pharmaceuticals Ltd, Uxbridge, Middlesex, UK OBJECTIVES: Reimbursement agencies like the United Kingdom (UK) National Institute for health and Clinical Excellence (NICE) use mathematical modelling to analyze A317 the costs and benefits of health care technologies. The structure of these models is not pre-specified, but dependent on the requirements of the decision problem on hand. Various decision-analytic models have been developed to assess the potential clinical and economic benefits of several RA treatment regimes, including anti-TNF agents and early use of disease-modifying anti-rheumatic drugs (DMARDs). The main objective is to provide a guide on how to model treatment options for early rheumatoid arthritis, in particular etanercept. Thus, it sought to address the methodological issues relating to the choice of modeling technique. RESULTS: Previous analyses have used a variety of modelling approaches to model treatment options for RA. The rationale for the choice of model structure is rarely discussed in published studies and can affect the results produced. Advances in the treatment of RA make it necessary to develop a model which is flexible enough to account for different attributes (e.g. clinical remission, radiographic non-progression, and functional status), but does not increase the complexity of the model to an amount where it becomes too difficult to populate it with relevant data. Individual level models offer a solution to this problem by simulating the progression of each individual with different characteristics, where nonMarkovian distributions allow greater flexibility in modelling the timing of events. Further, based on the characteristics of the disease, the model should assume independence between individuals, whilst, based on its treatment options, as assessment of adequate treatment response is important, time should be modeled. CONCLUSIONS: This work suggests that future analyses should use individual sampling models (e.g. simulated patient-level Markov models) for modelling the costs and benefits of treatments for RA. PMS76 A LONGITUDINAL COMPARISON OF MAPPING EQUATIONS DERIVED FROM BASELINE AND POST-INTERVENTION DATA Kontodimopoulos N1, Bozios P2, Elmatzoglou I2, Raftakis I2, Yfantopoulos J3, Niakas D1 1 Hellenic Open University, Patras, Greece; 2Asklipieio Voulas General Hospital, Voula, Greece; 3National and Kapodistrian University of Athens, Athens, Greece OBJECTIVES: Mapping from disease-specific to utility measures is a research area gaining increasing attention. However, few studies have assessed the longitudinal validity of mapping models and none, to our knowledge, have compared models derived from baseline and post-intervention patient data. This study examined models derived from time-differing patient data, for mapping the Modified Health Assessment Questionnaire (MHAQ) on to the EQ-5D. METHODS: A total of 120 rheumatoid arthritis patients (60.0% female, mean age 59.0) completed the MHAQ and EQ-5D at baseline, and 3, 6 and 12 months after administration of an anti-TNFa or another biological agent. OLS regression produced mapping equations from baseline and postintervention data. Model predictive ability and explanatory power were assessed by root mean square error (RMSE) and adjusted R2 respectively. Pearson’s r and intraclass correlation coefficient (ICC) assessed association and level of agreement between predicted and reported utilities. RESULTS: R2 (baseline, 3, 6 and to 12 months) was 0.452, 0.418, 0.541 and 0.413 respectively, whereas prediction errors were 13.0%, 11.1%, 8.1% and 7.2%. All equations produced a range of scores comparable to those achieved by the standard EQ-5D scoring algorithm, as well as strong correlations and agreement between reported and predicted utilities. The baseline and 12-month models appeared to under- and overestimate EQ-5D scores respectively, whereas the 6-month model generated the smallest differences, typically less than the minimally important difference for the EQ-5D (0.03). CONCLUSIONS: Baseline data, which typically corresponds to lower patient HRQOL, may give mapping equations with compromised predictive ability, compared to models generated from post-intervention data, and this study showed that this factor is worth examining when longitudinal data are available. The next step in this line of research is to test QALY estimates and cost-utility increments derived from baseline and post-intervention mapping algorithms. PMS77 THE DETERMINANTS OF SOCIAL ROLES AMONG RHEUMATOID ARTHRITIS PATIENTS—DATA FROM THE NDB PORTUGAL COHORT Chaves I1, Marques R1, Vasconcelos J1, Pedro S1, Rodrigues A1, Michaud K2, Wolfe F3, Garcia E1 1 BioEPI, Clinical and Translational Research Center, Oeiras, Portugal; 2University of Nebraska Medical Center, Omaha, NE, USA; 3National Databank for Rheumatic Diseases, Wichita, KS, USA BACKGROUND: An important consequence of rheumatoid arthritis (RA) is disability, traditionally assessed by the Health Assessment Questionnaire (HAQ), which captures the physical component. Disability can also be evaluated by restricted performance of social roles and those predictors haven’t been extensively studied. OBJECTIVES: Determinants of social roles, among RA patients, as evaluated by paid work and social functioning were studied. METHODS: A total of 1,140 RA patients from the NDB-Portugal longitudinal cohort were analyzed. Univariate (UV) and Multivariate (MV) generalized estimating equations were used to assess whether the following factors were determinants of paid work (1 = yes 0 = no, OR; 95%CI) and of SF-36 social functioning (0–100, 100 is best, β; 95%CI): age, sex, marital status, educational level, number of people living in patient’s household, dependence on others, RA duration, number of major comorbidities, TNF, anxiolytic and/or antidepressant use, HAQ, disease activity (RADAI), quality of life (VASQOL), pain, fatigue and sleep disturbances (VAS scales, 0–10, 10 is worst). RESULTS: In MV analysis the odds of having a form of paid work increased with higher education (1.17; 1.12, 1.22) and better VASQOL (1.92; 1.15, 3.20) and decreased with higher age (0.91; 0.90, 0.93), A318 longer disease duration (0.96; 0.94, 0.98) and worse HAQ (0.74; 0.60, 0.92). Social functioning improved with better VASQOL (8.62; 3.32, 13.91), but worsened with more comorbidities (−0.92; −1.49, −0.35), anxiolytic and/or antidepressant use (−5.33; −7.68, −2.98), worse RADAI (−1.54; −2.80, −0.28), HAQ (−8.38; −10.51, −6.25) and fatigue (−1.06; −1.48, −0.65). Furthermore, the mechanism by which RADAI affected social functioning did not appear to be through pain (0.20; −0.43, 0.84), but through HAQ as it absorbed 56% of the explanatory power of RADAI. CONCLUSIONS: Performance of social roles was limited among RA patients with more HAQ disability, and improved by better quality of life. Disease activity appeared to influence social functioning through worse physical disability but not through pain. PMS78 WORSE 6-MONTH BASELINE HAQ AND THE SELF-REPORTED RHEUMATOID ARTHRITIS DISEASE ACTIVITY INDEX PREDICT IMPROVEMENT IN THEIR SCORES 6 MONTHS LATER, AMONG RHEUMATOID ARTHRITIS PATIENTS Vasconcelos J1, Pedro S1, Marques R1, Chaves I1, Rodrigues A1, Michaud K2, Wolfe F3, Garcia E1 1 BioEPI, Clinical and Translational Research Center, Oeiras, Portugal; 2University of Nebraska Medical Center, Omaha, NE, USA; 3National Databank for Rheumatic Diseases, Wichita, KS, USA BACKGROUND: Disease activity and disability should be assesssed at each rheumatoid arthritis (RA) patient visit to monitor response to therapy. Disease activity has been traditionally evaluated by the disease activity scale (DAS-28) and the ACR criteria, which depend on physicians, but few studies have analyzed short-term predictors of the patient self-reported RA disease activity index-RADAI. Disability is routinely assessed by the health assessment questionnaire (HAQ) whose long term predictors, but not short-term, have been extensively studied. OBJECTIVES: To analyze whether the 6-month baseline levels of HAQ and of RADAI and other factors predict change at 6-month intervals. METHODS: RA patients from the biannual NDB-Portugal cohort were used. For each patient, differences between two consecutive 6-month intervals were computed for HAQ (0–3, 3 is worse) and RADAI (0–10, 10 is worse). For each scale, a binary outcome was constructed based upon these differences, where a positive increment meant worsening in function and disease activity. Patients’ observations whose increments were null were excluded. Univariate (UV) and multivariate (MV) generalized estimating equations were used. Factors included age, sex, marital status, disease duration, education level, number of major comorbidities, paid work status and 6-month baseline HAQ and RADAI levels. RESULTS: MV analyses revealed that the main predictors of HAQ were baseline HAQ: (OR: 0.49, 95% CI 0.44, 0.54), number of comorbidities (1.09 (1.05, 1.14)), and age (1.02 (1.01, 1.02)); For RADAI they were, baseline RADAI (0.69 (0.66, 0.73)), comborbidities (1.11 (1.07, 1.16)), and educational level (0.95 (0.93, 0.97)). CONCLUSIONS: Worse baseline levels of HAQ and RADAI predicted their respective improvement 6-months later. This could be due to optimization of treatment strategies when worse baseline scores are detected, but whatever the reason, performing these two patient reported outcomes are a quick and non-rheumatologist dependent way to improve patients’ disease status over 6-month intervals. PMS79 USE OF PATIENT-REPORTED OUTCOMES IN ON-LINE COMMUNITIES TO CONDUCT OBSERVATIONAL COMPARATIVE EFFECTIVENESS RESEARCH: A PILOT STUDY IN RHEUMATOID ARTHRITIS Cascade E1, Bharmal M2 1 iGuard, Rockville, MD, USA; 2Quintiles, Rockville, MD, USA OBJECTIVES: The demand for comparative effectiveness research (CER) data from payers, physicians, and patients is significant, but the cost and time associated with prospective randomized trials is a barrier to rapid decision-making. Use of patientreported outcomes (PROs) collected via on-line patient communities provides one channel for rapid data collection, particularly in conditions such as rheumatoid arthritis (RA), where validated PRO instruments are available. METHODS: A random sample of iGuard.org members in the US treated with non-steroidal anti-inflammatory drug (NSAIDs), oral disease-modifying antirheumatic drugs (DMARDs), or biologics for RA completed the Rheumatoid Arthritis Disease Activity Index (RADAI) and a series of other questions related to their disease. iGuard.org is a free medication monitoring service that is introduced to patients through multiple sources including physician, pharmacy and online referrals. For this study, we report pilot baseline data on patientreported RADAI, pain, and joint counts across the three treatment groups to demonstrate use of on-line communities in supporting CER. RESULTS: A total of 153 RA patients completed the study: 49 treated with NSAIDs only, 51 exposed to oral DMARDs, and 53 exposed to biologics. The mean (SD) RADAI score was 4.59 (2.16). Adjusting for age and gender and multiple comparisons, there were significant differences between the three treatment groups on RADAI scores (p = 0.0045) and patient global assessment of pain (p = 0.0357) but not on the number of painful joints (p = 0.3512). The trend was towards patients on NSAIDs only having worse outcomes compared to patients on biologics or patients on oral DMARDs. CONCLUSIONS: This pilot study demonstrates the possibility of collecting baseline disease severity data directly from patients using the RADAI, which is sensitive to detect differences by treatment modality. The next step in the pilot program will be to investigate the potential for capturing longitudinal disease progress information amongst patients in on-line communities. 13th Euro Abstracts RESPIRATORY-RELATED DISORDERS – Clinical Outcomes Studies PRS1 RESPIRATORY SYNCYTIAL VIRUS PROPHYLAXIS IN SPECIAL POPULATIONS Paes BA1, Li A2, Lanctot KL2, Mitchell I3 1 McMaster University, Hamilton, ON, Canada; 2Sunnybrook Health Sciences Centre, Toronto, ON, Canada; 3University of Calgary, Calgary, AB, Canada OBJECTIVES: To examine the pattern of palivizumab utilization and compliance in infants with pre-existing disease within the Canadian Registry Database (CARESS). METHODS: A prospective, registry of infants across 27 sites who received at least 1 dose of palivizumab during the 2006–2009 RSV seasons. Neonatal and demographic data were collected from the parent/caregiver at enrollment. Data on palivizumab utilization, compliance, and outcomes related to respiratory illness (RI) events were collected monthly. Premature infants ≤35 completed weeks gestational age without medical conditions who met standard approval criteria for palivizumab (Group 1) were compared to those with underlying medical disorders who received prophylaxis (Group 2). RESULTS: Group 1 (n = 3379) Group 2 (n = 489). Male: 56.8% versus 54.6% (P = 0.433). Average Enrollment Age (months) ± SD: 3.6 ± 3.4 versus 9.9 ± 8.8 (P = 0.000). Average GA (weeks) Mean ± SD: 31.0 ± 3.1 versus 37.1 ± 4.3 (P = 0.000). Average # injections ± SD: 3.6 ± 1.5 versus 3.7 ± 1.5 (P = 0.159). Hospitalization Rate (HR) for RI: 4.1% versus 9.2% (P = 0.000). RSV HR: 1.3% versus 2.7% (P < 0.05). On average infants received 86.0% ± 28.3% of the expected number of injections. Group 2 infants comprised Down syndrome (n = 118, 24.1%), upper airway anomalies (n = 112, 22.9%), cystic fibrosis (n = 62, 12.7%), neuromuscular impairment (n = 42, 8.6%), pulmonary (n = 38, 7.8%), multiple system disorders (n = 34, 7.0%), cardiac (n = 17, 3.5%), immunocompromise (n = 8, 1.6%), and miscellaneous (n = 58, 11.9%). From 2006–2009, the proportion of Group 2 infants receiving prophylaxis increased 2-fold from 5.6% (69/1224) to12.2% (245/2016). Overall. Group 2 infants were older at enrollment with more advanced GA and had significantly higher RI and RSV hospitalization rates. No serious adverse events directly related to palivizumab occurred. CONCLUSIONS: Results imply that infants with underlying medical disorders that are not approved for prophylaxis by advisory bodies and current position statements are at greater risk for RSV infections and hospitalization. PRS2 CARESS: THE CANADIAN REGISTRY OF SYNAGIS (2006–2009) 1 2 2 3 Paes BA , Li A , Lanctot KL , Mitchell I 1 McMaster University, Hamilton, ON, Canada; 2Sunnybrook Health Sciences Centre, Toronto, ON, Canada; 3University of Calgary, Calgary, AB, Canada OBJECTIVES: Palivizumab is used for respiratory syncytial virus (RSV) prophylaxis in high risk children. Data on seasonality, risk factors, and outcomes are necessary to evaluate the impact of palivizumab on the incidence of RSV infections, minimize health care resources and identify which infant sub-sets are receiving prophylaxis.To determine current usage of palivizumab prophylaxis, compliance patterns, hospitalization rate (HR) and outcomes in children at high-risk of respiratory syncytial virus (RSV) infection through a Canadian Registry Database (CARESS). METHODS: A prospective, study of infants who received at least 1 dose of palivizumab in the 2006–2009 RSV seasons across 27 sites. Neonatal and demographic data were collected upon enrolment. Parents/caregivers were contacted monthly for data on palivizumab utilization, compliance and outcomes related to any respiratory tract events. RESULTS: A total of 4926 infants aged 2 days–47 months (mean = 5.4 months) were enrolled. Participants were typically male (57.1%) and Caucasian (70.8%). Gestational age (GA) was 32.2 ± 4.6 completed weeks. 3480 (70.6%) premature infants received palivizumab (≤35 completed weeks GA), 403 (8.2%) required O2, 471 (9.6%) had congenital heart disease and 572 (11.6%) were prophylaxed for other risk factors. On average patients received 3.7 ± 1.5 injections, with 17,982 doses given overall. There were no drug related serious adverse events.296 infants required 357 hospitalizations for respiratory tract infections with a hospitalization rate of 6.0%. There were significant differences between indications for palivizumab (chi-square = 71.8, P < 0.005). The overall RSV positive HR was 1.38%. Hospitalization rates were highest in infants of aboriginal descent (15.0%, chi-square = 22.2, P < 0.005). Hospitalized infants had a lower percentage of compliant injections (62.8% versus 68.6%, p = 0.003). CONCLUSIONS: The RSV HR in the 2006–2009 RSV seasons resembled several published reports (range 1.3%–5.3%). RSV HR may be decreasing because of compliance with palivizumab prophylaxis, variability in RSV epidemiology, hospital admission criteria and preventive education. PRS3 COMORBIDITY PROFILING OF COPD PATIENTS IN THE UNITED KINGDOM PRIMARY CARE USING AN INCIDENCE BASED APPROACH TO DETECT ASSOCIATIONS WITH THE DISEASE Kiri VA PAREXEL International, Uxbridge, London, UK OBJECTIVES: Comorbidity is an important factor in any comparative assessment of treatments associated with morbidity and mortality of patients. Many factor such as age, gender and duration of a disease can influence the impact of comorbid diseases on quality of life. In the health care setting (the primary source of data for most observational studies), the decision to give a particular treatment to a particular patient with a given disease is generally based on patient specific characteristics, the most important of which is disease condition. Thus, confounding by indication/disease 13th Euro Abstracts severity is a common source of bias. Consequently, failure to properly control for the bias could lead to serious flaws in the study. METHODS: We used a novel approach for identifying comorbidities associated with COPD by mapping the incident comorbidity patterns in the pre-COPD diagnosis period as well as over the course of the disease using a retrospective cohort of patients aged 50+ in the UK General Practice Research Database. Each patient was matched to another without COPD on year of birth, sex, general practice and completed years of medical records up to at least a year after the index date for COPD between 1990 and 1998. We identified 24,000 such pairs that also satisfied a requirement for at least one medical consultation and at least one prescription for any drug in the year prior to the index date for COPD. RESULTS: Based on trends in rate ratios, we found significant time-dependent associations between the incident COPD diagnosis and incident comorbid conditions such as lung cancer, myocardial infarction, pneumonia, cardiac disorders, osteoporosis, fractures, skin bruises, psychiatric disorders and respiratory infections. CONCLUSIONS: The results indicated interesting associations which could help improve our understanding of the natural history of COPD and its burden. This methodology can be used to identify important comorbidities for effective comparative assessments. PRS4 EFFECTIVENESS OF VARENICLINE COMPARED TO BUPROPION AND NICOTINE REPLACEMENT THERAPY (NRT) FOR SMOKING CESSATION IN TWO SMOKING SPECIALIZED UNITS OF THE SPANISH PRIMARY CARE SETTING Sicras-Mainar A1, Navarro-Artieda R2, Diaz-Cerezo S3, Sanz de Burgoa V4 1 Directorate of Planning, Badalona Serveis Assistencials, Badalona, Barcelona, Spain; 2Hospital Universitari Germans Trias i Pujol, Barcelona, Spain; 3Complutense University of Madrid, Madrid, Spain; 4Pfizer Spain, Alcobendas (Madrid), Spain OBJECTIVES: The objective of this study was to estimate the effectiveness of varenicline, bupropion and nicotine replacement therapy (NRT) in smoking cessation in two specialized smoking units belonging to primary care centers. METHODS: A multicenter longitudinal observational study was designed. Patient’s data were collected retrospectively based on their clinical records. Patients over the age of 18, who initiated treatment of smoking cessation between January 1, 2006 and January 12, 2008 with varenicline, bupropion or NRT were included in the analysis. Patient’s follow-up was conducted from time-baseline (day 1) and assessed at 6 and 12 months. Main variables included in the study were: comorbidities, effectiveness (continuous abstinence) and pharmacological tolerability. Statistical analysis was performed by KaplanMaier survival curves; P < 0.05. RESULTS: A total of 957 smokers patients treated with NRT (53.0%), bupropion (25.1%) and varenicline (21.9%) were included in the analysis. The mean age of participants was 47.6 (11.3) years and 58.6% were men. The average duration of smoking was 19.5 (6.7) years. At 6 months, 61.2% (95% CI: 54.6–67.8%) of participants in the varenicline group were continuously abstinent from smoking compared with 56.9% (95% CI: 50.6–63.2%) in the bupropion group and 52.3% (95% CI: 48.0–56.6%) in the NRT group; p = 0.003. At 12 months, the rate of continuous abstinence was 57.4% (95% CI: 50.7–64.1%) in the varenicline group compared with 52.9% (95% CI: 46.6–59.2%) in the bupropion group and 47.1% (95% CI: 42.8–51.4%) in the NRT group; p = 0.002. Pharmacological tolerability was similar between groups except for symptoms of irritability which were lower in the varenicline group: 4.3% compared to 8.3% in the bupropion group and 10.3% in the NRT group. CONCLUSIONS: Varenicline appeared to be an effective and safety alternative compared with bupropion and NRT on smoking cessation in the primary care setting. PRS5 BUDESONIDE/FORMOTEROL PLUS TIOTROPIUM (BUD/FORM + TIO) VS. SALMETEROL/FLUTICASONE PLUS TIOTROPIUM (SALM/FLU + TIO): A SYSTEMATIC REVIEW AND ADJUSTED INDIRECT COMPARISON BETWEEN TWO ALTERNATIVE TRIPLE TREATMENTS IN CHRONIC OBSTRUCTIVE PULMONARY DISEASE (COPD) Edwards SJ, Gray J AstraZeneca UK Ltd, Luton, Bedfordshire, UK OBJECTIVES: Use of triple therapy (long-acting beta2 agonist [LABA], inhaled corticosteroid [ICS] and long-acting muscarinic antagonist [LAMA]) for the treatment of COPD has doubled in the UK over the past 5 years for all severities of the disease. This research was designed to compare the two most commonly prescribed combination inhalers (BUD/FORM and SALM/FLU) as the basis of triple therapy with TIO, the most widely used LAMA. METHODS: Systematic review of CENTRAL, EMBASE and MEDLINE for randomised controlled trials (RCTs) in patients with COPD treated with BUD/FORM+TIO or SALM/FLU+TIO was conducted in May 2010. Mixed treatment comparison (MTC) using TIO as a common comparator was conducted using a Bayesian Markov Chain Monte Carlo simulation. Fixed- and random-effects models were explored with the preferred model selected based on the Deviance Information Criterion (DIC). Data was extracted from relevant trials on severe exacerbations (a composite of oral corticosteroids, hospitalizations and A&E visits due to worsening symptoms). Summary effect estimate was calculated as odds ratio (OR) with 95% credible interval (95% CrI) where OR < 1 favoured BUD/FORM+TIO and OR > 1 favoured SALM/FLU+TIO. RESULTS: Of the 124 papers identified in the literature search, 3 RCTs had comparable patient populations and were able to supply data for analysis (1 BUD/FORM+TIO [N = 660] and 2 SALM/FLU+TIO [N = 301 and N = 60]). The exclusion of papers was based on not meeting all of the following inclusion critieria: RCTs of the chosen comparators; COPD patient population; reporting exacerbations; English-language full publication; and non-duplicates. When A319 the fixed- and random-effects MTC models were compared the fixed-effects MTC had the lowest DIC. The results indicate a 55% relative reduction in severe exacerbations with BUD/FORM+TIO compared to SALM/FLU+TIO (OR 0.45, 95% CrI: 0.22 to 0.83). CONCLUSIONS: This MTC suggests that BUD/FORM-based triple therapy is significantly more effective at reducing severe exacerbations than SALM/FLU-based triple therapy. PRS6 DEVELOPING AND APPLYING A STOCHASTIC DYNAMIC POPULATION MODEL FOR CHRONIC OBSTRUCTIVE PULMONARY DISEASE Hoogendoorn M1, Feenstra T2, Hoogenveen RT3, Al M1, Rutten-van Mölken MP1 1 Erasmus University, Rotterdam, The Netherlands; 2RIVM /UMCG, Bilthoven, The Netherlands; 3National Institute for Public Health and the Environment (RIVM), Bilthoven, The Netherlands OBJECTIVES: Modeling a chronic disease like COPD is useful to extrapolate (intermediate) treatment effects observed in short-term randomized trials to the medium or long term. This study aimed to extend an existing dynamic population model of COPD progression by including exacerbations and making the model stochastic. METHODS: The COPD model starts with COPD prevalence in the Dutch population specified by age, gender, disease severity (four severity stages) and smoking status. Each following year the model simulates the changes in the prevalence and COPD-related health care costs due to incidence, mortality and disease progression, i.e. annual decline in FEV1% predicted.The structure was adjusted to include moderate and severe exacerbations and the following input parameters were estimated by quantitative meta-analyses: the frequency, case-fatality, lung function decline, quality of life loss and costs of exacerbations. The model was made stochastic by specifying probability distributions around all important model parameters. The adapted model can be used to assess the impact of interventions that influence COPD incidence, disease progression, frequency and/or severity of exacerbations, mortality, quality of life or combinations of these effects. To illustrate the potential use of the model, long- term costs and effects were projected for four different COPD interventions, two on pharmacotherapy, one on smoking cessation therapy and one on pulmonary rehabilitation. RESULTS: Compared to a reference scenario representing minimal treatment, the cost-effectiveness of the four interventions ranged from c6,100 to c12,200 per QALY gained. The probability of the interventions to be cost-effective at a ceiling ratio of c20,000 varied from X to Y%. CONCLUSIONS: The extended COPD model can provide policy makers with information about long-term costs and effects of interventions over the entire chain, from primary prevention to care for very severe COPD. Moreover it describes the uncertainty of the outcomes. PRS7 A COMPARISON OF CLINICAL EFFICACY AND SAFETY OF CICLESONIDE WITH FLUTICASONE IN 1:1 AND 1:2 DOSE RATIOS IN THE TREATMENT OF BRONCHIAL ASTHMA Wojciechowski P, Stozek A, Rogoz A, Rys P, Wladysiuk M, Plisko R HTA Consulting, Krakow, Poland OBJECTIVES: This study compared efficacy and safety of ciclesonide (CIC) with fluticasone (FP) in the treatment of bronchial asthma. METHODS: Comparison was based on randomized controlled trials (RCTs) identified by means of systematic review, carried out according to the Cochrane Collaboration guidelines. The most important medical databases (EMBASE, MEDLINE, CENTRAL) were searched. Two reviewers independently selected trials, assessed their quality and extracted data. Meta-analysis of head-to-head trials was performed. RESULTS: Ten RCTs directly comparing CIC vs. FP were identified and included in the analysis. Comparisons of both interventions in 1:1 and 1:2 dose ratios were assessed in 8 and 5 trials, respectively. Efficacy of CIC was comparable to FP in both, 1:1 and 1:2 dose ratios with respect to reduction in risk of asthma exacerbations, improvement in proportion of symptom-free days, rescue medication-free days and improvement in asthma symptoms. Moreover, CIC and FP showed similar improvement in spirometric parameters. Comparison between CIC and FP in 1:1 dose ratio revealed that treatment with CIC was associated with statistically significant risk reduction of adverse events possibly related to study medication (RR = 0.57 [0.39; 0.83]; NNT = 16.89 [10.24; 48.18]) and candidosis (RR = 0.31 [0.17; 0.56], NNT = 32.74 [22.23; 61.99]), while the differences between groups were not significant for 1:2 dose ratio. There were no statistically significant differences between CIC and FP in either dose ratio with respect to the risk of any adverse events, upper respiratory tract infections, pharyngitis and dysphonia. CONCLUSIONS: Ciclesonide is a therapeutic option for patients with bronchial asthma showing comparable efficacy to fluticasone in both 1:1 and 1:2 dose ratios and provides concomitant risk reduction of candidosis and adverse events related to study medication. PRS8 A COMPARISON OF CLINICAL EFFICACY AND SAFETY OF CICLESONIDE WITH BUDESONIDE IN 1:1 AND 1:2 DOSE RATIOS IN THE TREATEMENT OF BRONCHIAL ASTHMA Wojciechowski P, Stozek A, Rogoz A, Rys P, Wladysiuk M, Plisko R HTA Consulting, Krakow, Poland OBJECTIVES: This study compared efficacy and safety of ciclesonide (CIC) with budesonide (BUD) in the treatment of bronchial asthma. METHODS: Comparison was based on randomized controlled trials (RCTs) identified by means of systematic review, carried out according to the Cochrane Collaboration guidelines. The most important medical databases (EMBASE, MEDLINE, CENTRAL) were searched. Two A320 reviewers independently selected trials, assessed their quality and extracted data. Meta-analysis of head-to-head trials was performed. RESULTS: Seven RCTs comparing CIC vs. BUD were identified and included in the analysis. Comparisons of both interventions in 1:1 and 1:2 dose ratios were assessed in 3 and 4 trials, respectively. CIC as compared with BUD in 1:1 dose ratio was associated with significant improvement in forced vital capacity (FVC) and peak expiratory flow (PEF) by spirometry (WMD = 0.09 [0.03; 0.14] and WMD = 19.00 [2.37; 35.63], respectively) as well as reduction in proportion of symptom-free days (p = 0.018). No statistically significant differences between both drugs in 1:2 dose ratios were observed. CIC and BUD in either dose showed comparable efficacy with respect to the risk of asthma exacerbation, improvement in symptoms of asthma and forced expiratory volume in 1 second (FEV1). CIC-treated patients experienced less upper respiratory tract infections than those treated with BUD in 1:1 dose ratio, however the difference was on the border of statistical significance (RR = 0.65 [0.43; 0.99], NNT not significant). There were no statistically significant differences between CIC and BUD in either dose ratio with respect to risk of adverse events, adverse events related to study medication, pharyngitis and dysphonia. CONCLUSIONS: Ciclesonide provides an improvement in spirometric paramethers and reduction of asthma symptom-free days as compared to budesonide in 1:1 dose ratio, while no differences were noticed between CIC and BUD in 1:2 dose ratio. Safety profile of ciclesonide seems to be comparable with budesonide. PRS9 QUALITY OF LIFE AND ECONOMIC IMPACT OF ASTHMA CONTROL IN FRANCE AND SPAIN. FIRST RESULTS OF THE EU-COAST STUDY Chouaid C1, Calvo Corbella E2, Com-ruelle L3, Brosa M4, Doz M5, Gueron B6 1 Hôpital St Antoine, Paris, France; 2Centro de Salud Universitario Pozuelo I. UAM., Madrid, Spain; 3IRDES, Paris, France; 4Oblikue Consulting, Barcelona, Spain; 5CEMKA, Bourg La Reine, France; 6GSK, London, England OBJECTIVES: Current asthma management recommendations are based on the level of asthma control rather than disease severity. The financial impact associated with asthma control needs to be determined. The primary objective of this study was to estimate quality of life and health care costs according to the patients’ level of asthma control in France and in Spain. METHODS: An observational retrospective bottom-up cost of illness study was conducted in adults patients with asthma. Investigators were general practitioners. Asthma control was evaluated using the validated auto-test Asthma Control Test (ACTTM) for a one month period and 2009 GINA’s asthma control criteria for a three months period. Quality of life (QoL) was assessed using EQ-5D profile. Costs (direct and indirect) were evaluated from a societal perspective. RESULTS: 794 patients (France: 391; Spain: 403) were enrolled in the first quarterly wave of the survey. Based on the ACT, asthma was determined to be well-controlled (ACT ≥ 20) in 48.1% [IC 95%: 42.1%–54%] and 56.2% [IC95: 51.1%–61.2%] of French and Spanish patients respectively. In both countries, EQ-5D scores were 0.9 and 0.7 in well-controlled and not well-controlled patients respectively (P < 0.0001). Total costs of asthma were directly related to asthma control in both countries. The average cost (Euros/month/patient) of well-controlled asthma was c28 (±36) in France and c77 (±105) in Spain compared with c140 (±782) (p = 0.0009) and c252 (±380) (p < 0.0001) respectively for not well-controlled asthma. Similar associations were observed using the latest GINA’s criteria. CONCLUSIONS: We found strong associations between asthma control level, costs and QoL in patients with asthma. Achievement of good clinical control of asthma may result in a significant decrease of the economic burden of asthma and a better health status in adults. PRS10 TRENDS IN ANTI-ASTHMA MEDICATION USE IN DUTCH CHILDREN FROM 1998 UNTIL 2007 Houweling LM, Penning FJ, Meijer WM PHARMO Institute for Drug Outcomes Research, Utrecht, The Netherlands OBJECTIVES: The prevalence of asthma has been reported to be increasing for decades, but recent studies suggest a leveling off or even decrease in prevalence among children. The objective of this study was to assess the trends in anti-asthma medication use in Dutch children from 1998 until 2007. METHODS: From the PHARMO Record Linkage System, an administrative database network of outpatient pharmacies, hospitals, and other settings, including data of ~4 million individuals in defined areas of the The Netherlands, we assessed 1) the proportion of children aged 0–19 years with at least one dispensing for short (SABA) or long-acting β2-antagonists (LABA), inhaled glucocorticoids (ICS), fixed dose combinations (FDC) of LABA and ICS or leukotriene receptor antagonists (LRTA) in the years 1998–2007 stratified by age groups, 2) the proportion of users per type for each year and 3) the age distribution per medication type for the year 2007. RESULTS: Overall prevalence of anti-asthma medication use was approximately 7% and remained stable between 1998 and 2007. Prevalence was highest among children aged 0–1 years and declined with increasing age. Over time, the prevalence remained stable in all age groups, except for the youngest in which it increased from 10% in 1998 to 19% in 2007. Use of ICS and LABA declined over time (user proportions: −17% and −3%), against increased use of LABA-ICS FDC and LRTA (+15% and +3%). Children <6 years made up relatively large proportions of SABA and ICS users, while the older children were more dominant in the other medication types. CONCLUSIONS: This study provides a comprehensive overview of trends in anti-asthma medication use in children showing no overall change in prevalence over time. However, a change in type of medication was observed. 13th Euro Abstracts RESPIRATORY-RELATED DISORDERS – Cost Studies PRS11 COST ANALYSIS OF DIFFERENT PHARMACOTHERAPIES IN SUBJECTS WITH CHRONIC OBSTRUCTIVE PULMONARY DISEASE IN THE PUBLIC SECTOR OF HONG KONG Lee KK1, Yu W2, Ko F3, Lee VW3, Chow DP3 1 Monash University Kuala Lumpur Sunway Campus, Selangor Darul Ehsan, Malaysia; 2Princess Margaret Hospital, Hong Kong, China; 3Chinese University of Hong Kong, Hong Kong OBJECTIVES: To compare the impact of different pharmacotherapies on the cost of exacerbation of COPD in public hospitals in Hong Kong. METHODS: This was a retrospective study comparing the cost of management of chronic obstructive pulmonary disease (COPD) in patients from a public hospital perspective in Hong Kong. Institutional ethics approval was obtained. The patients were above 65 years with a diagnosis of COPD of stage 3 or above based on the GOLD guideline. COPD was defined based on ICD-9-CM codes. Two cohorts of patients were included: 1) Those who received fluticasone propionate and salmeterol, plus ipratropium bromide and salbutamol inhalation, and 2) Those who received ipratropium bromide and salbutamol, and beclomethasone inhalation. Patients were followed for 6 months before and after medications. The date one of the therapies started was the index date. RESULTS: A total of 75 patients were recruited over January to June 2008. All values are expressed as mean ± SD. The number of emergency room visits and cost of visits before and after the index date for both cohorts did not show a significant difference (1.86 ± 2.5 days vs. 1.62 ± 1.93 days and HKD1061 ± 1442 vs. HKD924 ± 1102 for cohort 1, 1.15 ± 1.28 days vs. 1.02 ± 1.29 days and HKD657 ± 730 vs. HKD582 ± 735 for cohort 2, p > 0.1). There was a trend of reduction in hospitalization days and cost of stay before and after the index date for cohort 1 (although not reaching a significant level) but not in cohort 2 (20.6 ± 28.4 days vs. 15.6 ± 18.8 days and HKD68162 ± 93875 vs. HKD51662 ± 66278 for cohort 1, 11.1 ± 13.5 days vs. 10.2 ± 13.2 days and HKD36730 ± 44604 vs. HKD33860 ± 43571 for cohort 2, p > 0.1). CONCLUSIONS: A trend of reduction in hospital stay was demonstrated (although not statistically significant) in patients after receiving combination therapy of fluticasone propionate and salmeterol plus ipratropium bromide and salbutamol. This is consistent with earlier studies in other countries. a larger number of patients is required to prove the significance of this finding. PRS12 THE COST ANALYSIS OF BRONCHODILATOR PRESCRIPTIONS FOR TREATMENT OF STABLE COPD IN UKRAINE Tolubaiev V, Zalis’ka O Danylo Halytsky Lviv National Medical University, Lviv, Ukraine OBJECTIVES: To analyze the GP’s prescriptions of bronchodilators for managing the patients with stable COPD, to compare the prescriptions with Ukrainian National COPD Management Guidelines. And we estimated the annual costs for bronchodilators were used most frequently. METHODS: The analysis of prescription habits based on data from outpatient personal medical cards of 103 patients with COPD (2nd Municipal Clinic, Bila Tserkva, Ukraine), focused on bronchodilators as central medications for symptomatic treatment. Among 103 patients 15 were with mild COPD, 75 with moderate and 13 with severe (according to GOLD). The prescriptions for regular treatment were compared with national guidelines. Annual medications costs were calculated based on the number of medicines units for 24 hour bronchodilation during a year, multiplied by the one unit price in UAH (1 EUR = 10.05 UAH). RESULTS: Regular basis was prescribed to 87 patients. Among 87 prescriptions 13.7% were long-acting inhaled bronchodilators or combinations, 86.3% were shortacting inhaled bronchodilators, methylexanthines or combinations. Ranking to a number of medicines were most frequently prescribed: 56.8% fixed combination Ipratropium/Fenoterol, 10.2% Tiotropium, 9.1% Fenoterol, 7.9% Salbutamol. Annual costs for these bronchodilators were: 1206.0, 6262.0, 1303.0 and 275.0 UAH, respectively. CONCLUSIONS: The most frequently prescribed bronchodilator was Ipratropium/Fenoterol. It’s available during decades and inexpensively. But its basis requires triple daily medication, in consequence of that persistency rate might be reduced with effectiveness reducing, respectively. Tiotropium ranked second in the prescription frequency. It has duration of action of more than 24 hours, high effectiveness and safety levels but gets 6262.0 UAH annually. So the affordability of a bronchodilator is weighty criterion for Ukrainian GP’s prescriptions. But medications costs must be scrupulously compared with effectiveness for their including to the national formulary and reimbursement decision-making. PRS13 ECONOMIC IMPACT OF NON-ADHERENCE TO GOLD GUIDELINES IN COPD PATIENTS IN PRIMARY CARE IN SPAIN Galera J1, Lahoz R1, Lleonart M1, Riera M1, Sicras-Mainar A2, Miravitlles M3 1 Novartis Pharma, Barcelona, Spain; 2Directorate of Planning, Badalona Serveis Assistencials, Badalona, Barcelona, Spain; 3Hospital Clínic de Barcelona, Barcelona, Spain OBJECTIVES: The aim of this study is to analyze the economic impact of nonadherence to GOLD guidelines in COPD patients. METHODS: A retrospective analysis was carried out in a claim database. Patients ≥40 years-old with diagnosis of COPD confirmed with spirometry (FEV1/FVC < 0.7) were eligible for this analysis. Patients were classified into two groups according to whether they fulfill with therapeutics recommendations by severity defined in GOLD guidelines (GOLD group) or those who don’t fulfill (NO-GOLD group). Demographics, medical and use of resources data were collected and direct and indirect costs were analyzed. a probabilistic mul- 13th Euro Abstracts tivariate sensitivity analysis of avoided costs was carried out. All results are presented annualized and costs expressed in c2009. RESULTS: A total of 1365 patients were included, 79.5% males, mean (±SD) age was 71.4 (±10,3), mean FEV1 was 65.34% and a COPD history of 5.5 (±2,9) years. Patients were divided into GOLD group and NO-GOLD group and in stage II (FEV1 < 80% and ≥50%) and stage III (FEV1 < 50% and ≥30%) for both groups. Total annual costs per patient were analyzed. Directs costs analyzed were treatment, medical visits, emergency visits, hospital stay, laboratory, spirometry, radiology and oxigenotherapy. Indirect cost evaluated was sick leave. Avoided costs between both groups will be analyzed, and confirmed with sensitivity analysis. CONCLUSIONS: The results of this study will provide whether the adherence to the recommendations of the GOLD guidelines in COPD patients can lead to significant savings to the Spanish health care system. PRS14 THE POTENTIAL COST IMPACT OF USING A PEG HYDROGEL SEALANT TO PREVENT AIR LEAKS AFTER LUNG RESECTION SURGERY IN SPAIN Ott M1, Larkin M2, Minshall M3 1 Covidien, Zurich, Switzerland; 2PRMA Consulting, Fleet, UK; 3Covidien, Fishers, IN, USA OBJECTIVES: Persistent air leaks after lung resection surgery often lead to increased complications including longer length of hospital stay and resulting treatment costs. a PEG hydrogel sealant (PleuraSeal™) is used as an adjunct to standard closure of pleural air leaks during lung resection surgery and has demonstrated shorter hospitalizations (1.7 fewer days; Wain JC et al, 2008, Conference Abstract) and more rapid removal of chest tubes in a clinical study (Covidien, data on file) when compared with standard closure (e.g. staples and sutures alone). Our study was designed to estimate the potential cost offsets for using PleuraSeal™ compared with standard closure in 100 hypothetical patients from a Spanish hospital who had an air leak after lung resection surgery using the cost offset from shorter hospital stays balanced against the cost for PleuraSeal™. METHODS: We assumed the cost for PleuraSeal™ in Spain to be c420 per treatment applied to all 100 hypothetical patients compared with no added cost for standard closure. We then balanced the added cost for PleuraSeal™ with the reduction in number of hospital days and subsequent costs (c590/day [Varela et al, 2006] @1.7 fewer days per patient) compared with standard closure and calculated the potential cost offset for PleuraSeal™. RESULTS: A Spanish hospital that performs 100 lung resections with air leaks can expect to save c58,300, or approximately c583 per patient when compared to standard closure as patients who had air leaks stayed in hospital an average of 1.7 days longer than those without air leaks. CONCLUSIONS: The cost for PleuraSeal™ is completely offset by the shorter length of hospital stay compared with standard closure inSpain. PleuraSeal™ is a compelling option for Spanish hospitals that perform lung resection surgeries as the cost of the treatment is completely offset by the reduction in air leaks and subsequent shorter hospital stay. PRS15 THE POTENTIAL COST IMPACT OF USING A PEG HYDROGEL SEALANT TO PREVENT AIR LEAKS AFTER LUNG RESECTION SURGERY IN ITALY Ott M1, Larkin M2, Minshall M3 1 Covidien, Zurich, Switzerland; 2PRMA Consulting, Fleet, UK; 3Covidien, Fishers, IN, USA OBJECTIVES: Persistent air leaks after lung resection surgery often lead to increased complications. a PEG hydrogel sealant (PleuraSeal™) is used as an adjunct to standard closure of pleural air leaks and has demonstrated shorter hospitalizations (1.7 fewer days; Wain JC et al, 2008, Conference Abstract) when compared with standard of care (e.g. staples and sutures alone). Our study was designed to estimate the potential cost offsets for using PleuraSeal™ in 100 hypothetical patients with an air leak using the cost offset from shorter hospital stays balanced against the cost for PleuraSeal™. METHODS: We assumed the cost for PleuraSeal™ in Italy to be c420 per treatment applied to all 100 hypothetical patients compared with no added cost for standard of care. We then balanced the cost for PleuraSeal™ with the reduction in hospital days and subsequent costs (c337/ thoracic ward day, a Brunelli et al, 2007), then calculated the potential cost offset for PleuraSeal™. RESULTS: An Italian hospital that performs 100 lung resections with air leaks can expect to save c15,300, or approximately c153 per patient when compared to standard of care as patients who had air leaks stayed in hospital an average of 1.7 days longer than those without air leaks. CONCLUSIONS: The cost for PleuraSeal™ is completely offset by the shorter length of hospital stay compared with standard of care in Italy. PleuraSeal™ is a compelling option for Italian hospitals that perform lung resection surgeries as the cost of the treatment is completely offset by the reduction in air leaks and subsequent shorter hospital stay. PRS16 PITFALLS IN THERAPEUTIC REFERENCE PRICING PRACTICE OF THE MEDICATION OF ASTHMA Kalo Z1, Abonyi-Tóth Z2, Bártfai Z3, Tímár G4, Vokó Z1 1 Eötvös Loránd University, Budapest, Hungary; 2RxPress Ltd, Budapest, Hungary; 3Elisabeth Teaching Hospital Sopron, Sopron, Hungary; 4Syreon Research Institute, Budapest, Hungary OBJECTIVES: Therapeutic reference pricing (TRP) is one of the most frequently employed methods for cost-containment of pharmaceuticals. The Hungarian National Health Insurance Fund (NHIF) applies TRP based on DDD, yet even the WHO objects to using DDD in pricing and reimbursement decisions. Our objective was to compare actual drug use to DDD and to evaluate whether TRP based on DDD could result in A321 savings in the costs of maintenance medication and the total direct health expenditures between asthma patients treated with Symbicort Turbuhaler (SYT) and Seretide Diskus (SED). METHODS: Real world data for the analysis was derived from the NHIF database. Actual drug use and average costs were compared in four groups (high dose SYT, medium dose SYT, high dose SED and medium dose SED). Multiple linear regression was employed to adjust data for the differences in gender and age distribution of patients. RESULTS: A total of 12,260 patients using SED and 15,539 patients using SYT for asthma were involved into the analysis. Average drug use was lower in both groups than DDD. Patients used 1.38–1.95 inhalations in both SED groups, 1.28–1.97 inhalations in the high dose SYT group, and 1.74–2.49 inhalations in the medium dose SYT group. Although the daily therapeutic cost of medium dose SED based on DDD would be much lower than the cost of SYT, no difference was found in the actual cost of the maintenance therapy or the total medical costs. CONCLUSIONS: Our analysis provides evidence about the limitation of employing daily therapy costs according to DDD for TRP. No relationship was found between the daily therapeutic costs calculated using the NHIF methodology and costs of real drug use, in real world TRP did not lead to the reduction of expenditures from the payer’ perspective. Potential confounding factors may limit the generalizability of conclusions. PRS17 COST OF CHRONIC OBSTRUCTIVE PULMONARY DISEASE: A HEALTH CARE ADMINISTRATIVE DATABASES ANALYSIS Fornari C1, Di Stasi F2, Blasi F3, Conti S1, Mantovani LG4, Merlino L5, Cesana G1 1 University of Milano—Bicocca, Monza, MB, Italy; 2Nycomed S.p.a., Milan, MI, Italy; 3University of Milan—IRCCS Fondazione Ospedale Maggiore Policlinico Cà Granda, Milan, MI, Italy; 4 Federico II University of Naples, Naples, Italy; 5Regional Health Authority of Lombardy, Milan, MI, Italy OBJECTIVES: Chronic Obstructive Pulmonary Disease (COPD) currently represents the fifth leading cause of death worldwide and it’s estimated to rise to the third cause by 2030. Few studies have estimated the burden of COPD. The aim of this study was to evaluate consumption and cost of health care resources of severe COPD patients in the general population from the Italian health care system perspective. METHODS: Data were extracted from Healthcare Administrative Databases (HADs) of the region of Lombardy, which covers a population of 9.2 million members. HADs related to eligibility criteria, hospital admissions, pharmaceutical and outpatient claims have been organized in a data warehouse using probabilistic record linkage. The study included subjects aged more than 18 years that in 2003 had at least one hospitalization with COPD as the main diagnosis or as the second one when the main was: pneumonia, respiratory or heart failure, pneumothorax and acute pulmonary heart disease. Subjects were followed for two years collecting information on health care resource and vital status. Severe and moderate exacerbations of COPD were respectively defined as hospitalizations for COPD and as prescriptions of corticosteroids and/or antibiotics after the index hospitalization. RESULTS: We selected 16,476 patients with age mean 76 years and an average follow-up of 19 months. The 38% of patients were women with mean age three years higher than men. Subjects with severe exacerbations had on average 4.4 moderate and 1.9 severe episodes after the index hospitalization. The mean annual cost for these patients was c11,824 (IC95% 11,554–12,079). Patients with only moderate exacerbations had on average 3.9 moderate exacerbations and the mean annual cost was c6413 (IC95% 6288–6549). Costs were steady in time. CONCLUSIONS: The burden of COPD is important both in frequency and cost; it requires an effort in programming new health services particularly devoted to aging population. PRS18 HEALTH CARE COSTS OF INDIVIDUALS WITH AND WITHOUT COPD IN SWEDEN Jansson SA1, Stenling A2, Backman H1, Rönmark E1, Lindberg A1, Lundbäck B1 1 The OLIN Studies, Luleå, Sweden; 2AstraZeneca Nordic MC, Södertälje, Sweden OBJECTIVES: This analysis estimates the health care costs (hospitalizations and outpatient visits) due to respiratory and co-morbid diseases in a cohort of subjects with and without COPD in Sweden. METHODS: A cohort of individuals was identified in earlier clinical examinations of the general population within the Obstructive Lung Disease in Northern Sweden (OLIN) Studies. The cohort consisted initially of 993 subjects fulfilling criteria (GOLD) for COPD and an age- and gender matched control group of same size without obstructive lung function impairment. Since 2005, these subjects have been invited to yearly interviews and examinations. The participation rate has been high, above 85%. This analysis is based on resource use data collected in 2006 (n = 772 COPD, n = 802 non-COPD). Unit costs from the Northern health care region price list of 2010 were used. RESULTS: The mean health care costs of all diseases amounted to c1421 (median c236) annually per subject with COPD compared to c1127 (median c169) for subjects without COPD (p = 0.16, nonparametric Mann-Whitney). The mean costs in relation to disease severity were (GOLD criteria severity stage; p-value COPD vs. non-COPD): c1166 (mild; p = 0.66), c1755 (moderate; p = 0.28), c2223 (severe+very severe; P < 0.01). a minor part of the costs was attributable to respiratory diseases; 5.8% in non-COPD subjects and 10.5% in subjects with COPD. Respiratory disease costs were statistically significantly higher among subjects with COPD in all disease severity stages. CONCLUSIONS: The mean annual health care costs were c294 (median c67) higher for subjects with COPD. The difference in costs between subjects with severe+very severe COPD and subjects without the disease was statistically significant. A322 PRS19 ASSESSMENT OF THE CLINICAL AND ECONOMIC IMPACT OF AIR LEAKS DURING POST-OPERATIVE PULMONARY SURGERY USING THE MEDICARE POPULATION Gemmen E1, Doyle J2, Smith BF3, Garvert W4, Proach J5, Long J5, Nagel MP6 1 Quintiles, Rockville, MD, USA; 2Quintiles Global Consulting, Hawthorne, NY, USA; 3 Quintiles Consulting, Durham, NC, USA; 4Quintiles, Falls Church, VA, USA; 5Triage HealthCom, LLC, Lawrenceville, NJ, USA; 6Neomend, Inc., Irvine, CA, USA OBJECTIVES: Estimate the clinical and economic impact to U.S. hospitals of air leaks during post-operative pulmonary surgery using the Medicare Provider Analysis and Review (MEDPAR) data set. METHODS: The 2008 The Medicare Provider Analysis and Review (MEDPAR) data set contains records for 100% of Medicare beneficiaries who use hospital inpatient services. For all stays with pulmonary surgery, length of stay (LOS), total charges, and in-hospital mortality rates were compared between those stays including an air leak vs. those stays without an air leak. Unadjusted results were calculated using descriptive statistics (mean, median, frequencies, etc.) Adjusted results were calculated using multivariate regression analysis while controlling for age and gender. RESULTS: There were a total of 41,348 hospital inpatient stays with pulmonary surgery in the 2008 MEDPAR data set, of these 8,774 (21.2%) included air leak and 32,574 (78.8%) of which did not. In the MEDPAR data set patients with pulmonary surgery stays including air leak had a similar age distribution to patients without air leak, had a longer LOS on average (10.7 days vs. 7.2 days; P < .0001), had more total charges ($78,830 vs. $63,528; P < .0001) and were nearly equally likely to die during their stay (14.8% vs. 13.94%; p = 0.057). After adjusting for differences in age and gender between the two groups, the incremental LOS and total charges due to the presence of air leak is 3.4 days and $14,532 respectively. The total additional economic impact of having an air leak after pulmonary surgery, estimated by applying patient level adjusted charges to the incidence of air leak, is $127.5 million. CONCLUSIONS: The clinical and economic impact to U.S. hospitals of air leaks during or following major pulmonary surgery is significant. The reduction of these air leaks could save considerable hospital resources, payer dollars and patient lives. PRS20 OUTCOMES ASSOCIATED WITH IATROGENIC PNEUMOTHORAX 1 2 Stemkowski S , Braxton JC 1 Lovelace Respiratory Research Institute, Kannapolis, NC, USA; 2Davidson College, Davidson, NC, USA OBJECTIVES: Iatrogenic pneumothorax occurs when air or gas becomes present in the pleural cavity following medical treatment. Besides subjecting the patient to unnecessary health risks, iatrogenic pneumothorax leads to an increased amount of health care resources including observed through patient cost, length of stay, and inpatient mortality. This research aims to quantify the incremental effects of iatrogenic pneumothorax on these three outcomes. METHODS: Discharge records from Premier’s Perspective database of US inpatients who underwent inpatient pulmonary surgery and were discharged in 2007 were examined. The definition of iatrogenic pneumothorax consistent with ICD-9-CM code 512.1 was used to classify patients. Chi-square tests were used to detect differences between iatrogenic pneumothorax patients and non-iatrogenic pneumothorax patients for three outcomes. Multivariable regression models were used to obtain more precise estimates of the incremental effects of iatrogenic pneumothorax on outcomes while controlling for comorbidities, demographic variables and the patient’s primary treatment. RESULTS: A total of 112,827 patients were analyzed (8,482 with iatrogenic pneumothorax). Chi-square tests demonstrated that patients with iatrogenic pneumothorax were older (P < 0.0001), had lower hospital costs (P < 0.0001), a shorter length of stay (P < 0.0001), and lower mortality rate (P < 0.0001). Log linear modeling demonstrated iatrogenic pneumothorax increases patient costs by 10.49% (95%CI: 8.76%–12.23%). Negative binomial models showed iatrogenic pneumothorax increases patients length of stay by 8.01% (95% CI: 6.21%–9.82%), while no difference was found with respect to mortality. CONCLUSIONS: The incremental effects of iatrogenic pneumothorax are shown to significantly increase patient costs and length of stay but not inpatient mortality. PRS21 BURDEN OF BRONCHIAL ASTHMA AND CHRONIC OBSTRUCTIVE PULMONARY DISEASE IN RUSSIA Omelyanovsky VV1, Avksentieva MV1, Derkach EV1, Tsfasman FM2, Sveshnikova ND1 1 Research Center for Clinical and Economic Evaluation and Pharmacoeconomics, Moscow, Russia; 2Institute of Clinico-Economic Expertise and Pharmacoeconomics, RSMU, Moscow, Russia OBJECTIVES: to assess social and economic burden of a Bronchial asthma and Chronic Obstructive Pulmonary Disease in Russian Federation. METHODS: “Cost of illness” analysis was performed. Available data on epidemiology of bronchial asthma and chronic obstructive pulmonary disease in the Russian Federation has been analyzed. Experts were questioned to describe the common practice of treating patients with bronchial asthma and chronic obstructive pulmonary disease. Direct costs, indirect costs and intangible costs were calculated. RESULTS: According to the state registration data, in 2007 the number of patients with bronchial asthma was 1.3 million and with chronic obstructive pulmonary disease— 2.4 million. The burden of bronchial asthma incorporates direct costs (c220.9 million), indirect costs (c67.4 million) and intangible costs (c69.6 million). The burden of chronic obstructive pulmonary disease is c210.6 million, c212 million, c207 million for direct costs, indirect costs and intangible costs respectively. Cost of BA and COPD is c987.8 million. 13th Euro Abstracts CONCLUSIONS: According to the results of the analysis Bronchial asthma and Chronic Obstructive Pulmonary Disease proved to be an important medical and social problem in Russian Federation. PRS22 COST OF COPD IN POLAND Jahnz-Rozyk KM, Targowski T, From S, Faluta T, Borowiec L Military Institute of Medicine, Warsaw, Poland OBJECTIVES: About two million people suffer from COPD in Poland. The aim of this study was to examine direct, mean costs of COPD in Poland under usual clinical practice form societal perspective. METHODS: It was an observational retrospective and prospective bottom-up-cost-of- illness study, based on a retrospective sample of patients presenting with COPD. Total medical resources consumption of a sample of COPD patients were collected in 2008 year trough physician—lung specialists. Direct costs of COPD were evaluated based on data from different populations of five clinical hospitals and eight ambulatory cares. Medical resource consumptions were categorized by investigators as usual COPD follow up and number and severity of exacerbations.Resources utilization and cost data are summarised as mean values per patient per year; 95% confidence intervals were derived using percentile bootstrapping. RESULTS: In patients studied, number of free –of- exacerbation days was 331, 1, mean number of outpatient exacerbation was 1.27, mean number of exacerbations requiring hospital was 0.24. Average total medical resource consumption of a COPD patient per year was c1006.1. Among this cost c605 was directly related to treatment of stable COPD (costs of drugs, additional exams, costs of medical visits, influenza vaccination and home oxygen therapy), c105.3 to outpatient exacerbation, and c295.8 to exacerbation treated in hospital. CONCLUSIONS: The burden of COPD itself appeared to be considerable magnitude from societal perspective in Poland. Overall, the main cost drivers were inpatient care and prescription medication. PRS23 COMPARISON OF DIFFERENT METHODS FOR ASSESSING ATTRIBUTABLE COSTS: A CASE OF MEDICAL COSTS ATTRIBUTABLE TO OBESE IN PATIENTS WITH ASTHMA Suh DC1, Kim CM2, Choi IS3, Lee DH4, Jang SM5, Kwon JW3, Barone J3 1 School of Pharmacy, Rutgers University, Piscataway, NJ, USA; 2Catholic University School of Medicine, Seoul, South Korea; 3Rutgers University, Piscataway, NJ, USA; 4Ewha Womans Univ College of Medicine, Seoul, South Korea; 5Health Insurance Review Agency, Seocho-gu, South Korea OBJECTIVES: We compared two alternative methods (recycled prediction and Oaxaca-decomposition) to estimate medical costs attributable to obesity in US adults with asthma. METHODS: This study used the 2003–2007 Medical Expenditure Panel Survey to select asthma patients (18–64 years old), excluding patients with pregnancy, malignancy, kidney dialysis, immunodeficiency, or low body mass index (BMI < 18.5 kg/m2). Obesity was defined as BMI >= 30 kg/m2. Medication costs were estimated using a generalized linear model with a log-link function and gamma distribution. For the recycled predictions method, predicted treatment costs for obese patients were calculated assuming that obese patients were normal-weight, holding the distribution of covariates obtained from the entire asthma patient sample. With Oaxaca-decomposition, average treatment costs for each group (obese vs. normal weight) were estimated. The differences in average costs between the two groups were then estimated for two components: a) costs due to patient characteristics (endowments), and b) costs due to obese/normal-weight parameters (coefficient), considered as costs attributable to obesity. To compare the two methods, the difference in costs between obese and normal-weight patients was simulated, after matching for patient demographic and clinical characteristics,. All costs were converted to 2009 US dollars using price indices. RESULTS: The prevalence of obesity and normal-weight among 7340 asthmatic patients was 32.5% vs. 35.1%, respectively. In the recycled prediction method, costs attributable to obesity were US$1798 (95%CI: US$1717–$1878). In the Oaxaca-decomposition, the difference in medical costs between two groups consisted of US$1357 (95%CI: US$1252–$1462) due to endowments and US$1285 (95% CI: US$1229–$1341) due to coefficient components (i.e. costs attributable to obesity). The difference in costs from the simulation was US$1124 (US$1045–US$1203). CONCLUSIONS: Costs attributable to obesity obtained using Oaxaca-decomposition were similar to those of the simulation method, but the costs obtained using the recycled prediction method were higher than those of Oaxaca-decomposition and simulation. PRS24 LONG-TERM EFFECTIVENESS AND COST-EFFECTIVENESS OF SMOKING CESSATION INTERVENTIONS IN PATIENTS WITH COPD Hoogendoorn M1, Feenstra T2, Hoogenveen RT3, Rutten-van Mölken MP1 1 Erasmus University, Rotterdam, The Netherlands; 2RIVM /UMCG, Bilthoven, The Netherlands; 3National Institute for Public Health and the Environment (RIVM), Bilthoven, The Netherlands OBJECTIVES: We aimed to estimate the long-term (cost-)effectiveness of smoking cessation interventions for patients with chronic obstructive pulmonary disease (COPD). METHODS: A systematic review was performed for randomized controlled trials on smoking cessation interventions in COPD patients reporting the 12-month biochemical validated abstinence rates. The different interventions were grouped into four categories: usual care, minimal counseling, intensive counseling and intensive counseling plus pharmacotherapy. For each category the average 12-months continu- 13th Euro Abstracts ous abstinence rate and intervention costs were estimated. a dynamic population model for COPD was used to project the long-term (cost-)effectiveness of one year implementation of minimal counseling, intensive counseling and intensive counseling plus pharmacotherapy for 50% of the smoking COPD patients compared to usual care. Time horizon was 25 years. Uncertainty and one-way sensitivity analyses were performed for variations in (the calculation of) the abstinence rates, the type of projection, intervention costs and discount rates. RESULTS: Nine studies were selected. The average 12 months continuous abstinence rates were estimated to be 1.4% for usual care, 2.6% for minimal counseling, 6.0% for intensive counseling and 12.3% for pharmacotherapy. Compared to usual care, the costs per QALY gained for minimal counseling, intensive counseling and intensive counseling plus pharmacotherapy were c16,900, c8,200 and c2,400, respectively. Results were most sensitive to variations in abstinence rates and discount rates. CONCLUSIONS: Compared to usual care intensive counseling and pharmacotherapy resulted in low costs per QALY gained with ratios comparable to results presented for smoking cessation in the general population. Compared to intensive counseling alone, intensive counseling plus pharmacotherapy was cost saving and dominated the other interventions. PRS25 A COST-UTILITY ANALYSIS FOR TIOTROPIUM BROMIDE IN THE LONG TERM TREATMENT OF SPECIFIC SUBGROUPS OF ITALIAN COPD PATIENTS Zaniolo O, Iannazzo S, Carsi M Adres srl, Torino, Italy OBJECTIVES: The UPLIFT trial demonstrated in 5,993 patients with moderate to very-severe chronic obstructive pulmonary disease (COPD) that 4 years of tiotropium bromide were associated with improvements in lung function, quality of life, and exacerbations compared with placebo. The aim of this study is the economic assessment of tiotropium when included in COPD routine care (RC) for specific groups of Italian COPD patients. METHODS: A probabilistic patient-level simulation Markov model was developed over a lifetime horizon, with one-year cycles and a 3.5% annual discount rate. Patients were characterized by gender, age, height, smoking status and FEV1. FEV1 time trend was modelled based on the decline recorded in UPLIFT. The mortality of the general Italian population adjusted by smoking status and FEV1 was adopted. Health utilities derived from published Italian studies, while their variation from the UPLIFT. Exacerbation rates derived from an Italian observational prospective study and were adjusted for the relative risk (RR) reported in UPLIFT. Direct sanitary costs were considered. Health care resource consumption for RC, exacerbations and SAEs derived from Italian observational studies and were valued according to current price and tariffs. Cost-effectiveness was assessed for the overall cohort and for subgroups of patients by age, sex, GOLD stage and smoking attitude. RESULTS: In the whole cohort, patients treated with tiotropium gained an average (95%CI) 0.50 (−1.63– −6.27) LYs and 0.42 (−0.25– −3.05) QALYs with respect to RC. The incremental lifetime cost was c3,357 (−c10,669– −c29,820). The incremental cost-effectiveness ratio (ICER) was c7,916 /QALY. In the subgroups analysis the ICER ranged from a minimum of c6,627/QALY (females, GOLD III) to a maximum of c13,187/ QALY (age <65 y, GOLD IV). CONCLUSIONS: The inclusion of tiotropium in RC for moderate to very severe COPD patients represents good value for money in Italy. The analysis across subgroups demonstrated a good stability of the model. PRS26 COST-EFFECTIVENESS AND BUDGET IMPACT ANALYSIS OF MOMETASONE FUROATE AS MAINTENANCE TREATMENT IN PATIENTS WITH MILD TO MODERATE ASTHMA FROM THE PUBLIC PAYER PERSPECTIVE IN BRAZIL Fernandes RA, Takemoto ML, Cukier FN, Guerra FC, Passos RB ANOVA—Knowledge Translation, Rio de Janeiro, RJ, Brazil OBJECTIVES: In the Brazilian public health care system, mometasone furoate (MF DPI) is not available and budesonide/formoterol (BUD/FF DPI) association is responsible for 86.9% of pharmacy claims for asthma. This study aimed to conduct costeffectiveness and budget impact analysis (BIA) of MF versus BUD/FF for adult patients with mild to moderate asthma from the public payer perspective. METHODS: A decision tree was developed to compare MF and BUD/FF based on indirect comparison once head-to-head studies were not available. The final FEV1 values were converted into probabilities of hospitalization in the first two years in accordance with observational evidence of association between FEV1 and exacerbation requiring hospitalization. Only direct medical costs were considered and unit costs were obtained from Brazilian official lists. BIA assumed pharmacy claims data from the Ambulatory Information System as current scenario (Beclomethasone: 3.1%; BUD: 9.9%; BUD/ FF: 86.9%) and a 20% initial market share for MF in substitution to equivalent doses of BUD/FF. RESULTS: Indirect comparison indicated 79 hospitalizations per 1000 patients for MF and 82 for BUD/FF during the first 2 years of treatment. Total cost of treatment was 832BRL and 655BRL per patient for MF200 mcg twice a day (bid) and MF400 mcg once a day and 840BRL for BUD/FF 400/12 mcg bid. These findings indicated MF as cost-saving in the proposed scenario with ICER of −2.608BRL and −61.959BRL per avoided hospitalization for MF200 mcg and MF400 mcg, respectively. The estimated budgetary impact for the first year showed a saving of 259,346,480BRL for MF 400 mcg and 10,919,299BRL for MF 200 mcg. CONCLUSIONS: MF is a clinically effective option to treat mild to moderate asthma and indirect comparison showed its clinical and economic benefit when compared to the most used anti-asthma medication in the Brazilian public setting. Further research to A323 directly compare both medications and to measure finalistic outcomes alongside clinical trials is needed. PRS27 COUNTRY ADAPTATION OF A HEALTH ECONOMIC MODEL: THE CASE FOR ROFLUMILAST IN THE NETHERLANDS Vemer P, Goossens LM, Rutten-Van Mölken MP Erasmus University, Rotterdam, The Netherlands BACKGROUND: The phosphodiesterase-4 enzyme (PDE4) inhibitor roflumilast is a new treatment that targets the underlying inflammation associated with COPD. When approved, roflumilast will be registered as an add-on to bronchodilator treatment in adult patients with severe COPD, with a history of frequent exacerbations. a health economic (HE) micro-simulation Markov model was used to support its submission in the United Kingdom (UK). Pharmaceutical companies can save significantly on the process of HE evidence development, if models can be adapted for use in more than one country. OBJECTIVES: To transfer an existing UK HE model to the The Netherlands in order to calculate the cost-effectiveness (CE) of roflumilast in patients with severe COPD from a societal perspective. METHODS: The model structure was adapted to include production loss using the friction cost method, and to separate heterogeneity from parameter uncertainty. All input parameters on health care use, costs, utilities, and COPD epidemiology were obtained from Dutch sources, except for the case-fatality rate of an exacerbation-related hospitalization. a direct comparison was made between a combination of a long-acting β2 agonist (LABA) plus roflumilast (ROFLU) and LABA alone. a second, indirect comparison was between LABA + ROFLU and LABA plus an inhaled corticosteroid (ICS). One-way and probabilistic sensitivity analyses were performed. RESULTS: From a societal perspective, the incremental CE ratio (ICER) for LABA + ROFLU compared with LABA alone, was c7900. The ICER of LABA + ROFLU versus LABA + ICS was c10,000. The probability that LABA + ROFLU was cost-effective when compared with LABA alone at a threshold of c20,000 versus LABA was 97%. Compared with LABA + ICS this probability was 68.3%. CONCLUSIONS: The original UK model was suitable for adaptation to the Dutch setting. The ICERs of roflumilast were below commonly referred threshold values of a QALY. PRS28 ECONOMIC EVALUATION OF FLUTICASONE PROPIONATE/ SALMETEROL COMBINATION THERAPY AND MONTELUKAST IN ADULT PATIENTS WHO ARE SYMPTOMATIC ON SHORT-ACTING BETA 2-AGONIST ALONE Rely K1, Gonzalez SE2, Salinas GE3, Alexandre PK4 1 CEAHealthTech, Mexico City, D.F., Mexico; 2GlaxoSmithKline, Mexico City, D.F., Mexico; 3 Hospital Infantil de México Federico Gómez, Mexico City, D.F., Mexico; 4Johns Hopkins University, Baltimore, MD, USA OBJECTIVES: To estimate the incremental cost-effectiveness of SFC verus montelukast in adult patients with persistent asthma. METHODS: A decision-analytic model was developed from a randomized, double-blind, double-dummy, 12-week clinical trial were analyzed. Efficacy end points included, symptom-free days (SFDs) during the 12-week period. The study assumed the Mexican health care perspective with costs in 2010 US dollars, and hence only direct costs were included in the analysis. Direct costs included those related to study drugs, emergency room department visits, unscheduled physician visits, and rescue medication. The incremental cost-effectiveness ratio (ICER), which is the mean difference in average costs divided by the mean difference in average effectiveness, was calculated for the effectiveness outcomes (SFDs). Issue of uncertainty was addressed by means of a probabilistic Monte Carlo simulation, which attributed stochastic distributions to model inputs. RESULTS: Treatment with FSC resulted in a significantly greater improvement in the mean percentage of symptom-free days compared with MON 48.9 and 21.7 respectively (p 0.001). In the base case, patients initiated on SFC displayed a 45% reduction in overall cost as compared with patients initiated on MON US S186 versus $US258, respectively, respectively). SFC dominated the use of MON because of previously demonstrated lower incidence of Asthma exacerbations and rescue free days. Sensitivity analyses determined that univariate changes in all model variables, including medication cost, and cost of treating exacerbation, did not impact overall results. a Monte Carlo simulation analysis found that use of SFC remains the best overall treatment strategy when taking into consideration the potential variance in all model assumptions. Compared with MON, SFC is estimated to be both more effective and more economically favourable, with a probability of almost 92%. CONCLUSIONS: The decision model indicated that use of SFC as treatment in patients with asthma should result in lower overall treatment costs relative to the cost of MON. PRS29 COST-EFFECTIVENESS OF SALMETEROL/FLUTICASONE PROPIONATE COMBINATION VERSUS LEUKOTRIENE MONTELUKAST FOR THE CONTROL OF PERSISTENT ASTHMA IN CHILDREN Rely K1, Gonzalez SE2, Alexandre PK3, Salinas GE4 1 CEAHealthTech, Mexico City, D.F., Mexico; 2GlaxoSmithKline, Mexico City, D.F., Mexico; 3 Johns Hopkins University, Baltimore, MD, USA; 4Hospital Infantil de México Federico Gómez, Mexico City, D.F., Mexico OBJECTIVES: To assess the incremental cost-effectiveness of SFC compared with MON for the control of persistent asthma in children. METHODS: We conducted an economic evaluation on a 12-week prospective randomized open-label parallel-group A324 comparison of SFC versus MON in children with symptomatic asthma receiving inhaled corticosteroids and short-acting β2-agonists. Asthma-related medication, unscheduled physician contacts and hospitalizations were collected prospectively. The main effectiveness measure was percentage of asthma-controlled week with no shortacting β2-agonist use during the study period. The analysis was conducted from the Mexican health care perspective using 2010 unit cost prices, and only direct costs were considered, all costs are reported in US dollar. The model was made fully probabilistic to reflect the joint uncertainty in the model parameters. RESULTS: Over the whole treatment period, the median percentages of asthma-controlled weeks were 83.3% in the SFC group and 66.7% in the MON group (SFC-MON difference, 16.7%; 95% CI, 8.3–16.7; P < 0.001 in favor of SFC). The mean total cost of the SFC regimen was US$186 compared with US$271 for the MON regimen. The SFC was the dominant strategy (both more effective and less expensive) using the SFC was associated with an incremental cost per additional asthma-controlled of $US (513). Probabilistic sensitivity analysis tested numerous assumptions about the model cost and efficacy parameters and found that the results were robust to most changes. CONCLUSIONS: This analysis demonstrates that, compared with MON, SFC may be cost saving from the Mexican health care perspective for the treatment of pediatric patients with asthma. SFC provided a reduction in the number of severe exacerbations, frequent asthma symptoms and rescue medication use. Incremental cost-effectiveness analysis indicated the dominance of SFC because of both lower costs and greater efficacy. PRS30 DISCREPANCY BETWEEN ANALYTIC APPROACHES IN THE CLINICAL AND ECONOMIC EVALUATION OF THE SAME TRIAL: EXPERIENCE IN COPD Hoogendoorn M, Al M, Rutten-van Mölken MP Erasmus University, Rotterdam, The Netherlands OBJECTIVES: Clinical and economic evaluations of the same trial often use different statistical analyses and methods to handle missing data. This leads to different results for the same health outcome. We aimed to study how the combination of multiple imputation (MI) with frequently used advanced methods of clinical analysis affect estimates of cost-effectiveness. METHODS: Data from a two-year RCT of an INTERdisciplinary COMmunity-based COPD management program (INTERCOM) versus usual care were used. Five outcomes, SGRQ, EQ-5D, 6MWD, total and severe exacerbations measured at 4, 12 and 24 months or continuously (exacerbations) were selected. These outcomes were re-analyzed using the same methods used in the clinical paper, i.e. with repeated measurement analysis or negative bionomial regression, but now after missing data have been imputed using MI. The resulting estimates were compared with 1) the estimates in the original clinical paper before MI and 2) the estimates obtained after MI based on simple averages before any further statistical analyses based on maximum likelihood. RESULTS: A total of 175 patients were included in the analysis of which 158 completed the trial. The cost difference of c2751 between INTERCOM and usual care was kept constant. The number of severe exacerbations avoided varied from 0.014 to 0.077 resulting in ICERs from c35,700 to c196,500, depending on the approach used. The improvement in SGRQ ranged from 2.2 to 2.6 units, but the ICERs were all around c1000. The gain in QALYs varied from 0.062 with an ICER of c44,400 to 0.085 with an ICER of c32,400 per QALY gained. The probability that the INTERCOM program was cost-effective at a threshold value of c50,000 ranged from 56% to 74%. CONCLUSIONS: This study showed that the combination of analytic approaches of the clinical and economic evaluations does alter the cost-effectiveness ratios. PRS31 SHOULD SALMETEROL/FLUTICASONE PROPIONATE (SAL/FP) BE ADDED TO ROUTINE COPD TREATMENT WITH FENOTEROL/ IPRATROPIUM BROMIDE (FEN/IB)? PHARMACOECONOMIC ASSESSMENT OF COPD TREATMENT BASED ON OBSERVATIONAL RESEARCH (PHACTOR) Denisov P1, Belevsky AS2, Chuchalin AG3, Tolkushin A4, Yagudina RI5, Kulikov A5, Zinchuk I5 1 Worldwide Clinical Trials Co Ltd, Saint-Petersburg, Russia; 2Russian State Medical University, Moscow, Russia; 3SRI of Pulmonology, Moscow, Russia; 4GlaxoSmithKline Russia, Moscow, Russia; 5Moscow Medical Academy, Moscow, Russia OBJECTIVES: To assess cost-effectiveness of adding SAL/FP to routine COPD treatment with Fen/IB. METHODS: Depersonalized patient database was created in multicenter observational research of severe and very severe COPD. Patients were subdivided into two groups: 1- Fen/IB-based therapy without SAL/FP (N = 245); 2Fen/IB-based therapy with SAL/FP (N = 84). Prices of drugs were up to Q1 2010 for Moscow city (from Farmexpert market monitoring). Unit cost of inpatient-day, outpatient-visit and emergency-visit was derived from Moscow city government regulation #290 from 04.2010 (about medicare). Direct medical costs within one year time horizon were assessed as health care perspective was taken. RESULTS: Number of COPD exacerbations per patient was 3.9 with and 6.8 without SAL/FP. Sum of yearly direct medical costs was 31,607 RUB (c832) with and 55,179 RUB (c1452) without SAL/FP. Incremental cost per one prevented exacerbation (ICER) was 1237 RUB (c32.5). Average cost of treatment of one exacerbation was 8 056 RUB (c212). Results were sensitive to unit cost of inpatient-day (25% increase leads to cost-saving in with SAL/FP arm). Indirect cost inclusion lead to considerable cost-saving in with SAL/FP arm (7952 RUB = c209). CONCLUSIONS: Adding SAL/FP to routine treatment of severe and very severe COPD with Fen/IB is cost-effective. 13th Euro Abstracts PRS32 REGIONAL DIFFERENCES AS A BASIS FOR SENSITIVITY ANALYSIS OF COST-EFFECTIVENESS OF SALMETEROL + FLUTICASONE PROPIONATE (SAL/FP) VS. INHALED CORTICOSTEROIDS (MONO-ICS) Demko IV1, Tolkushin A2, Yagudina RI3, Kulikov A3, Ogorodova LM4, Chuchalin AG5 1 Krasnoyarsk State Medical University, Krasnoyarsk, Russia; 2GlaxoSmithKline Russia, Moscow, Russia; 3Moscow Medical Academy, Moscow, Russia; 4Siberian State Medical University, Tomsk, Russia; 5SRI of Pulmonology, Moscow, Russia OBJECTIVES: To assess cost-effectiveness of SAL/FP vs. mono-ICS in different Russian regions using OPTIMA pharmacoeconomic model. METHODS: Description and calculation steps of OPTIMA model were published in ISPOR Twelfth Annual European Congress Research Abstract #PRS8. Variable region-specific input data: drug prices and dosage proportion (from Farmexpert monitoring as of Q4 2009), medical tariffs (from regional government regulations), GDP per capita and average salary (from statistics service). Constant disease-specific data: frequency of controlled/ uncontrolled asthma in arms (from clinical trial), number of unscheduled resources utilization and QoL in controlled and uncontrolled asthma (from prof. I.V.Demko’s observational study). Fixed combination SAL/FP (Seretide) was compared with monoICS (Beclomethasone, Fluticasone and Budesonide). ICERs (cost per QALY) were assessed for each 84 Russian regions. Regional WTP was assumed as three regional GDP per capita. 1 EUR = 38 RUB. RESULTS: Weighted average monthly pharmacotherapy cost varied from 1410RUB (in Kostroma) to 3376RUB (in Tula) for SAL/FP, and from 430RUB (in Kostroma) to 1524RUB (in Khanty-Mansi) for MonoICS. The differences were driven by proportion of low/medium/high doses. Medical tariffs varied dramatically as well: tariffs of outpatient visit varied from 107RUB (in Ivanovo and Dagestan) to 975RUB (in Yamal-Nenets), bed-day cost varies from 500RUB (in Kurgan) to 3123RUB (in Yamal-Nenets). GDP per capita were from 38110RUB (in Ingush) to 928374RUB (in Tyumen); average salary—from 9 125 RUB (in Dagestan) to 46480RUB (in Yamal-Nenets). SAL/FP was cost-saving (dominating) in 18 regions, cost-effective in 62 regions (ICER < WTP; in this regions ICERs were from 3210RUB (84EUR) to 639480RUB (16828EUR) per QALY), and disadvantageous (ICER > WTP) in 4 regions (Ivanovo, Kabardino-Balkaria, Ingush, and Dagestan; mainly due to low WTP). CONCLUSIONS: In general case SAL/FP was cost-effective in most Russian regions, in some regions SAL/FP was cost-saving, and in few regions—not cost-effective. To assess cost-effectiveness in particular cohort of patients additional analyses are needed. PRS33 ECONOMIC EVALUATION OF ILOPROST, EPOPROSTENOL AND TREPROSTINIL FOR THE TREATMENT OF PULMONARY ARTERIAL HYPERTENSION Román A1, Barberà JA2, Escribano P3, Sala ML4, Febrer L5, Casado MA6 1 Hospital de la Vall d’Hebron, Barcelona, Catalunya, Spain; 2Hospital Clínic de Barcelona, Barcelona, Catalunya, Spain; 3Hospital 12 de Octubre, Madrid, Spain; 4Hospital de la Santa Creu i Sant Pau, Barcelona, Catalunya, Spain; 5Bayer Healthcare, Barcelona, Catalunya, Spain; 6 Pharmacoeconomics & Outcomes Research Iberia, Pozuelo de Alarcón , Madrid, Spain OBJECTIVES: To analyze the eficiency of three alternative treatments (inhaled iloprost (ILO), intravenous epoprostenol (EPO) and subcutaneous treprostinil (TRE)) for patients suffering from pulmonary arterial hipertensión (PAH) iniciating therapy with a prostanoid. METHODS: A Markov model was built to simulate a PAH patient cohort in functional class III of the New York Heart Association (NYHA). The model had four health states, those of the functional classes, plus death. Treatment changes were allowed when patients worsened from class III to IV. Time horizon was three years and transition cycles were of 12 weeks. Perspective was that of the National Health System (NHS) in Spain. Data sources were: 1) literature review, 2) costs databases and 3) expert opinion. Costs were expressed in euros 2009. Costs and effects were discounted at a 3% rate following Spanish recommendations. Both, deterministic and probabilistic analyses were performed to check for robustness of results. RESULTS: At three years, results for initiating prostanoid therapy with ILO, EPO and TRE were, respectively: total cost—c143,092, c430,271 and c360,387 -; efficacy—2.695 LYG, 2.729 LYG and 2.690 LYG -; —1.737 QALY, 1.780 QALY and 1.728 QALY -; mean cost per LYG—c53,092, c157,678 and c133,997; mean cost per QALY—c82,376, c241,667 and c208,595 -. Incremental cost-effectiveness ratios and cost-utility ratios of EPO vs. ILO were: >8.5Mc/LYG and >6.5Mc/QALY, and vs. TRE were: >1.5Mc/LYG y > 1.3 Mc/QALY, much above the usually accepted threshold in Spain of 30,000 c/LYG or QALY. ILO was dominant vs. TRE. Sensitivity analyses confirmed these results. CONCLUSIONS: Initiating prostanoid therapy in class III PAH patients with intravenous epoprostenol is slightly more efficacious than the alternatives. At a three-year time horizon, inhaled iloprost shows to be the less costly alternative for the NHS in Spain. PRS34 IMPROVED PREDICTION OF FINDING COPD PATIENTS BY LUNG FUNCTION PRE-SCREENING IN PRIMARY CARE Thorn J1, Tilling B2, Lisspers K3, Jörgensen L4, Stenling A4, Stratelis G4 1 Sahlgrenska School of Public Health and Community Medicine, Gothenburg, Sweden; 2 Primary Care, Åtvidaberg, Sweden; 3Department of Public Health and Caring Sciences, Uppsala, Sweden; 4AstraZeneca Nordic MC, Södertälje, Sweden OBJECTIVES: To investigate if easily accessible pre-screening of individuals at risk for COPD leads to a more accurate selection of patients for ordinary spirometry, thereby improving the incidence of pathological test results. METHODS: Primary care 13th Euro Abstracts patients at risk for COPD (aged 45 ≤ years ≤ 85, ≥15 pack years) were included. Lung function (FEV1 and FEV6) was measured by copd-6 (mini-spirometer, Vitalograph), followed by ordinary spirometry (COPD diagnosis FEV1/FVC post bronchodilation ratio <70). Time consumed was recorded and costs calculated (national average nurse wage). Univariate logistic regression and receiver operating characteristic (ROC) curves were utilized. RESULTS: In all, 305 patients (21 centers in Sweden), females 57%, mean (standard deviation, SD) age 61.2 (8.4) years, FEV1/FVC 75.3 (10.2), pack years 30.3 (11.5) were included. COPD was diagnosed in 83 patients (27.2%). Copd-6 screening significantly predicted COPD by FEV1/FEV6 ratio. Patients with COPD had a mean (SD) FEV1/FEV6 of 69.0 (9.1), patients without COPD 77.7 (9.6). Sensitivity and specificity at a FEV1/FEV6 cut-off of 73.0 was 73.5% and 79.7%, respectively; area under the ROC curve was 0.800. Cost of one copd-6 measurement (~4 minutes) was SEK 19 (≈c2), while the cost for an ordinary spirometry (~33 minutes) was SEK 147 (≈c15). Without copd-6 pre-screening the cost of detecting one COPD patient based on age and pack years was SEK 542 (≈c54). Use of copd-6 reduces the cost to SEK 283 (≈c28), but also results in an 8.4% lower COPD detection rate due to copd-6 sensitivity limitations. CONCLUSIONS: Pre-screening with the copd-6 mini-spirometer to select patients (≥45 years; ≥15 pack years) for ordinary spirometry increased the frequency of detected COPD diagnoses from 27.2% to 73.5%. Although copd-6 sensitivity and specificity could be improved, its use in primary care may reduce the costs of detecting COPD patients. PRS35 ASSESSING THE COST-EFFECTIVENESS OF BECLOMETHASONE/ FORMOTEROL IN THE TREATMENT OF MODERATE TO SEVERE PERSISTENT ASTHMA IN SPAIN Darba J1, Restovic G2, Kaskens L2 1 Universitat de Barcelona, Barcelona, Spain; 2BCN Health, Barcelona, Spain OBJECTIVES: To estimates cost and effectiveness of beclomethasone/formoterol extrafine, fluticasone/salmeterol and budesodine/formoterol in the treatment of moderate to severe persistent asthma in Spain. METHODS: A Markov model was developed to represent the transition of a cohort of patients with moderate to severe persistent asthma through different health states: patients free of symptoms, patients with exacerbations and patients without exacerbations. Efficacy data was obtained from observational studies. Deterministic results were estimated and a probabilistic sensitivity analysis was conducted using statistical distributions in order to capture parameter uncertainty in the decision model. Treatment costs were obtained from literature review and a panel of clinical experts. Costs were referred to year 2009 and a time horizon of 12 weeks was chosen. Results were presented as expected cost per QALY and represented in cost-effectiveness acceptability curves (CEACs). RESULTS: In the deterministic analysis, the expected cost per patient was greater in the fluticasone/ salmeterol cohort (c333) and the budesodine/formoterol cohort (c300) in comparison with the beclomethasone/formoterol cohort (c292). The estimated effectiveness was the same in the three cohorts. In the probabilistic analysis CEACs showed that the probability that the treatment with beclomethasone/formoterol was more cost-effective than the treatment with fluticasone/salmeterol and budesodine/formoterol using alternative values for the maximum value that the health service would be willing to pay for an additional QALY gained was greater in the beclomethasone/formoterol cohort. CONCLUSIONS: When beclomethasone/formoterol extrafine has been compared to fluticasone/salmeterol and budesodine/formoterol we may conclude that the first one is a dominant strategy. Results from probabilistic sensibility analysis show that the choice of optimal strategy is independent on the maximum that the health service is prepared to pay per additional QALY gained because beclomethasone/formoterol extrafine has a greater probability of being cost-effective for all threshold values. RESPIRATORY-RELATED DISORDERS – Patient-Reported Outcomes Studies PRS36 ADHERENCE TO MEDICATION AND HEALTH-RELATED QUALITY OF LIFE IN PATIENTS WITH COPD: SGRQ—AN ALTERNATIVE METHOD TO IDENTIFY NONADHERENCE Meszaros A, Agh T Semmelweis University, Budapest, Hungary OBJECTIVES: Medication adherence and health-related quality of life (HRQoL) are important factors in determining success of drug treatments; although both have been studied intensively, limited data is available on the association between these factors. This study aimed: 1) to assess the relationship between adherence and HRQoL in COPD patients using generic and disease specific instruments; and 2) to evaluate the association between adherence and answers of SGRQ part 2-section 5 referring to medication use. METHODS: The cross-sectional study included a post-bronchodilator spirometry and completion of a self reported postal questionnaire. Information on adherence (4-item Morisky Medication Adherence Scale) and HRQoL (generic: EuroQol—EQ-5D, disease specific: St. George’s Respiratory Questionnaire—SGRQ) were obtained. Multiple linear regression model was used to analyze the association between HRQoL as dependent variable and disease severity (FEV1), age and adherence as independent variables. Differences in answers of SGRQ part 2-section 5 between adherence stages were analyzed with Pearson Chi-square test. RESULTS: A total of A325 227 patients were included at baseline, 170 of them completed the study. The mean age was 63.83 (SD = 11.24) years, there were slightly more female patients (58%). a total of 32% (n = 55) of the participants reported themselves as high adherent, 26% (n = 44) were medium, and 42% (n = 71) were low adherent. The mean EQ-5D score was 0.55 (SD = 0.21) and the mean SGRQ total score was 56.22% (SD = 16.19). HRQoL assessed with EQ-5D and SGRQ were neither significantly related to medication adherence. Both generic and disease-specific HRQoL were associated with age and disease severity. We found significant association between the number of the positive answers’ of SGRQ part 2-section 5 and adherence rate (Chi-square = 30.64 p = 0.000). CONCLUSIONS: The study showed no significant relationship between adherence to medication and any HRQoL domain. Our results suggest that, SGRQ part 2-section 5 could be suitable to estimate adherence in patients with COPD in clinical practice. PRS37 ASSOCIATION OF PATIENT-REPORTED INHALER SATISFACTION WITH PHYSICIAN-PERCEIVED COMPLIANCE WITH ASTHMA TREATMENT—RESULTS OF A CROSS-SECTIONAL STUDY IN FIVE EUROPEAN COUNTRIES Small M, Piercy J, Anderson P, Vickers A Adelphi Real World, Macclesfield, Cheshire, UK OBJECTIVES: It is hypothesised that patient satisfaction with the features of an asthma inhaler directly influences compliance behaviour of asthma patients. This study investigated the relationship between physician-reported compliance of 1400 asthma patients with the degree of patient-reported satisfaction with their inhaler. METHODS: Data were drawn from the Adelphi Disease Specific Programme (DSP) in Asthma, a cross-sectional study of consulting patients undertaken in 2009. Data were collected from physicians for each patient and collected directly from consenting patients. Key factors analyzed were physician-reported perceptions of patient compliance on a 1–5 scale (where 1 is ‘not at all compliant’ and 5 is ‘fully compliant’) and data for patientreported satisfaction with their inhaler including specific device features associated with convenience, dosing and inhaler mechanism. The results were tested against potential confounders including age, gender, ethnicity and current asthma severity, to ensure the relationship effect between compliance and satisfaction was distinct. RESULTS: There was a significant relationship between the physician-reported level of compliance and patient-reported satisfaction with their inhaler device. The investigators observed that a better perceived level of compliance was directly associated with a higher level of inhaler satisfaction (P < 0.001). Those inhaler features associated with convenience: “built to last/will not break easily”; “easy to hold and carry,” “instructions are simple and easy to follow,” and “no need to put the drug into the inhaler before use” accounted for four of the top five drivers of overall satisfaction. CONCLUSIONS: Patient satisfaction with features of their inhaler is an important factor in promoting patient compliance. If physicians made particular effort to select an inhaler that satisfies the patient, including attention to features associated with convenience, patient compliance with their asthma treatment would be expected to improve. PRS38 SYMPTOMATIC COMORBIDITIES AND HEALTH-RELATED QUALITY OF LIFE IN COPD IN THE EUROPEAN UNION Langley PC1, Freedman D2, Wagner JS3 1 University of Minnesota, Minneapolis, MN, USA; 2Kantar Health, Princeton, NJ, USA; 3Kantar Health, New York, NY, USA OBJECTIVES: This study assesses the impact of high prevalence comorbidities reported by persons with diagnosed COPD on their health related quality of life (HRQoL). The focus is on (1) the quantitative impact of comorbidities on SF-6D utility scores, and (2) the relative impact of symptomatic vs. non-symptomatic comorbid disease states. METHODS: The study is based on data from the internet based 2008 National Health and Wellness Survey (NHWS) undertaken in the UK, France, Spain, Germany, and Italy. The study identified 3103 persons with diagnosed COPD together with the nine most frequently diagnosed comorbidities: pain, high blood pressure, anxiety, high cholesterol, heartburn, insomnia, depression, migraine, and arthritis. High blood pressure and high cholesterol are considered non-symptomatic comorbidities. The impact of these comorbidities on the SF-6D utility scores (range 0–1) is estimated via an ordinary least squares regression model. All comorbidities enter as categorical variables. The model includes a range of variables which have been shown in previous studies to impact HRQoL. These include: sociodemographic factors, health risk behaviors (BMI, alcohol use, smoking), and a series of country dummy variables (Germany as reference category). RESULTS: The model demonstrates that for symptomatic comorbidites the impact on utility scores is substantial. All symptomatic comorbidities are significant. Depression followed by pain experienced in the past 6 months have the greatest deficit impact (respectively −0.060; 95% CI −0.071 to −0.050 and −0.055; 95% CI −0.064 to −0.045). High blood pressure and high cholesterol are not significant. CONCLUSIONS: Claims for the negative impact of COPD on HRQoL should be seen in the context of a cluster of comorbid disease states; in particular the high prevalence disease states such a pain, anxiety and depression. A326 PRS39 QUALITY OF LIFE, PRODUCTIVITY LOSS, AND RESOURCE USE AMONG EMPLOYED ADULTS AGED 40 TO 64 YEARS WITH CHRONIC OBSTRUCTIVE PULMONARY DISEASE (COPD) IN THE UNITED STATES (US) WORKFORCE DiBonaventura M1, Paulose-Ram R2, Su J3, Mcdonald M2, Zou KH2, Wagner JS1, Shah H3 1 Kantar Health, New York, NY, USA; 2Pfizer Inc., New York, NY, USA; 3Boehringer Ingelheim Pharmaceuticals, Inc., Ridgefield, CT, USA OBJECTIVES: To examine quality of life, work productivity, and health care resource use among employed adults ages 40 to 64 years with COPD. METHODS: Data were obtained from the US 2009 National Health and Wellness Survey (NHWS), an internetbased questionnaire. All employed adults (full-time, part-time, or self-employed) ages 40–64 years with or without a self-reported diagnosis of COPD were included. Impact on quality of life (using the mental (MCS) and physical component summary (PCS) scores and health utilities from SF-12v2), work productivity and activity impairment (using WPAI questionnaire), and resource use were analyzed. Multiple regression analyses for normally-distributed outcomes and negative binomial regressions for skewed outcomes were conducted. All models adjusted for demographic and patient characteristics and were weighted to project to the US population. RESULTS: There were 1,112 workers with COPD (mean age = 51.5 years) versus 18,912 workers without COPD (mean age = 50.0 years). After adjusting for demographic and patient characteristics, adults with COPD reported significantly lower mean levels of MCS (46.8 vs. 48.5, P < 0.0001), PCS (45.6 vs. 49.2, P < 0.0001), and health utilities (0.71 vs. 0.75, P < 0.0001), than adults without COPD. Workers with COPD reported significantly greater impairment while at work (presenteeism) (18.9% vs. 14.3%, P < 0.0001), overall work impairment (a combination of absenteeism and presenteeism) (20.5% vs. 16.3%, P < 0.0001), and impairment in daily activities (23.5% vs. 17.9%, P < 0.0001) than adults without COPD. Employed adults with COPD also reported more mean ER visits (0.21 vs. 0.12, P < 0.0001) and more mean hospitalizations (0.10 vs. 0.06, P < 0.0001) in the previous six months than employed adults without COPD. CONCLUSIONS: After adjusting for various demographic and patient characteristics, employed adults with COPD reported significantly worse quality of life, work productivity, and health care resource utilization than employed adults without COPD. These results highlight the substantial impact and burden of COPD in the US workforce. PRS40 EVALUATION OF SMOKING CESSATION DRUG USE AND OUTCOMES IN THE NETHERLANDS Overbeek JA1, Penning FJA1, van Spiegel PI2, Meerding WJ3, Smulders M1, Herings RMC1 1 PHARMO Institute for Drug Outcomes Research, Utrecht, The Netherlands; 2Slotervaart Ziekenhuis, Amsterdam, The Netherlands; 3Pfizer bv, Capelle a/d IJssel, The Netherlands OBJECTIVES: Several pharmacological therapies are available to help smokers quit. We determined utilization and effectiveness of smoking cessation therapy (SCT) in daily practice in the The Netherlands. METHODS: Subjects ≥18 years with a dispensing of varenicline, bupropion, nicotine or nortriptyline between March 2007 and September 2008 were identified from the PHARMO RLS. Using an encrypted methodology, the corresponding non-person-identifiable dispensing IDs were linked to a web-based system for patient-reported data collection: ePRO-LINK. Corresponding pharmacists asked the subjects to participate in the study and complete a web-based questionnaire on smoking history and cessation. RESULTS: Of 2,684 invited subjects, 612 were included in the analyses. Bupropion was the most frequently used SCT during last quit attempt (35%), followed by varenicline (28%), bupropion + nicotine (12%) and varenicline + nicotine (9%). Overall, 51% of patients also reported behavioural therapy. About 52% of varenicline and bupropion users, 42% of nicotine users and 20–40% of patients using multiple drugs reported not to smoke at the time of questionnaire. The median (IQR) number of days between time of questionnaire and start date of last quit attempt ranged from 271 (104–432) for varenicline + bupropion to 356 (205–518) for bupropion. The mean duration of drug use ranged from 42–53 days among SCT users who remained abstinent and from 19–42 days among SCT users who relapsed. CONCLUSIONS: The results of this study show that up to 50% of SCT users in daily practice quit smoking. More widespread use of these drugs in combination with behavioural therapy will help to further decrease smoking prevalence. PRS41 FACTORS AFFECTING QUALITY OF LIFE (QOL) OF ASTHMATIC PATIENTS IN SPAIN Martinez-Moragon E1, Palop M2, de Diego A3, Serra J4, Caloto T5, Nocea G5, Sabater E6 1 Hospital de Sagunto, Valencia, Spain; 2Hospital Universitario La F, Valencia, Spain; 3Hospital Universitario La Fe, Valencia, Spain; 4Hospital General de Vic, Barcelona, Spain; 5Merck Sharp & Dohme, SA, Madrid, Spain; 6Pharmacoeconomics & Outcomes Research Iberia, Pozuelo, Spain OBJECTIVES: Understanding and identifying impaired QoL is now recognised as an important component of asthma management.To assess QoL and factors influencing it, in asthmatic patients in Spain. METHODS: Observational, multicenter study was carried out in 40 Spanish pneumology units. Asthma diagnosed patients (according to GINA) aged ≥18 were enrolled. Disease severity and treatment information were investigator-reported, and socio-demographic data, symptoms, and QoL (self administered Mini Asthma Quality of Life Questionnaire ,Mini AQLQ) were patientreported. Optimal scaling methods for multivariate categorical data were conducted. 13th Euro Abstracts RESULTS: A total of 536 patients were studied (mean age 54 years; 65% female). Significant differences in Mini-AQLQ total scores were observed by disease severity according to GINA: 4.3 (SD 1.3) for severe persistent, 4.9 (SD1.2) for moderate persistent, 5.4 (SD 1.0) for mild persistent and 5.7 (SD1.0) for intermittent patients (P < 0.0001). As per the final multivariate model (with MiniAQLQ total score as dependent variable and after adjusting for all significant variables in the bivariate analysis), reported QoL was better for patients from Northern and Central Spain as compared with those from the South and the East (P < 0.0001), students and employed patients as compared with housewives and unemployed (p = 0.001), for those who had received information about the disease compared with those who did not (p = 0.004), for those with milder daytime symptoms (p = 0.001) and for patients with higher education (p = 0.021). CONCLUSIONS: Modifiable factors such as symptom control and adequate information about asthma can have a significant effect on patients’ QoL. RESPIRATORY-RELATED DISORDERS – Health Care Use & Policy Studies PRS42 EFFECTS OF ASTHMA MANAGEMENT, SOCIOECONOMIC STATUS AND MEDICATION INSURANCE CHARACTERISTICS ON EXACERBATION FREQUENCY IN CHILDREN WITH ASTHMA Ungar W1, Paterson JM2, Gomes T2, Bikangaga P1, Gold M1, To T1, Kozyrskyj A3 1 The Hospital for Sick Children, Toronto, ON, Canada; 2Institute for Clinical Evaluative Sciences, Toronto, ON, Canada; 3University of Alberta, Edmonton, AB, Canada OBJECTIVES: Asthma symptoms are adequately controlled in fewer than 25% of asthmatic children. The objective was to identify the determinants of severe exacerbation resulting in Emergency Department visits or hospitalization related to health status, socioeconomic status (SES) and drug insurance characteristics. METHODS: In this retrospective cohort study, data were collected on 522 asthmatic children in Toronto, Canada, regarding demographics, SES, drug plan characteristics, health status, health resource use and symptoms. Baseline data were linked deterministically to administrative data on asthma emergency visits and hospitalizations occurring in the year following baseline. Multiple Poisson regression was conducted to identify independent predictors of severe exacerbation in the full cohort and in a sub-group with prescription drug insurance. RESULTS: Younger age, prior emergency visits, nebulizer use, pet ownership, and receipt of asthma education, but no asthma action plan, were significant determinants of more frequent exacerbation. In the full cohort, children with high income adequacy had 28% fewer exacerbations than children with low income adequacy. In the sub-group with drug insurance, girls had 26% fewer exacerbations than boys, and children with food, drug or insect allergies had 52% more exacerbations than children without such allergies. Children of families with annual insurance deductibles greater than $90 had 95% fewer exacerbations. In addition, every percentage increase in the proportion of income spent out-of-pocket on asthma medications resulted in a 14% increase in severe exacerbations. CONCLUSIONS: Asthma history, disease management factors and SES were important determinants of severe asthma exacerbation in children. In families with drug plans, the magnitude of asthma medication cost-sharing as a proportion of household income, rather than income alone, was a highly significant determinant of severe exacerbation. This finding can inform drug policies aimed at reducing the negative consequences of cost-sharing. PRS43 COMPLIANCE TO GOLD GUIDELINES IN THE TREATMENT OF COPD: APPROPRIATENESS AND POTENTIAL COST SAVING IN ITALY Tomic R, Dellavedova E, Colangelo I Novartis Farma S.p.A., Origgio, Italy OBJECTIVES: Recent Italian studies showed that ICS-containing (Inhaled corticosteroids) drugs are prescribed to more than 50% of COPD patients regardless of disease severity and treatment guidelines, creating unnecessary costs for NHS. This study aims at assessing potential saving on drugs expenditure due to increased compliance to GOLD guidelines. METHODS: Novartis Core-model framework was adapted to Italian setting. Patient cohort was defined by applying prevalence rate of treated COPD patients to Italian population (2.5% annual growth rate). Patients severity split and treatment pattern are obtained by combining results of two Italian observational studies (reflecting GPs and pneumologists setting), assumed to be constant over 5 years. Drugs considered: LAMA (Long-acting muscarinic antagonist), LABA (Longacting β2-agonist), Fixed-Dose Combination (LABA+ICS), SABA (Short-acting β2-agonist), add-on ICS. Based on reimbursed net public price, daily treatment costs for drugs are calculated. To calculate annual treatment cost, 50% compliance for all drugs (except 30% for ICS) and annual discounting rate of 3.0% are assumed. Following GOLD treatment scheme, Mild and Moderate patients, receiving long-acting bronchodilators and/or add-on ICS, were assigned respectively to SABA and LABA/ LAMA alone. Difference in drugs costs between current and modified treatment pattern is calculated. RESULTS: Potential saving, calculated for Italy, was c160.2 million (36.4% of total drug costs) in year 1, and c793.4 million (36.4%) over 5 years. Analysis was extended to regional populations, resulting in one-year savings of: c26 million in Lombardy, from c10 million to c15 million in Campania, Lazio, Sicily, Veneto, Piedmont, Emilia Romagna, Apulia, Tuscany, from c3 million to c5 million in Calabria, Sardinia, Liguria, Marche, Abruzzo, Friuli-Venezia Giulia, Trentino-Alto Adige, and up to c2 million in Umbria, Basilicata, Molise, Aosta Valley. Results were 13th Euro Abstracts compliance-sensitive. CONCLUSIONS: Adhering to GOLD guidelines, apart from having patients treated appropriately, National/Regional Healthcare Systems could save significant resources and re-allocate them in more rational manner. PRS44 EFFECTIVENESS AND EFFICIENCY OF ME-TOO LISTS AS ILLUSTRATED BY THE EXAMPLE OF INTRANASAL STEROIDS Schoeffski O1, Becker B2 1 University Erlangen-Nuernberg, Nuernberg, Germany; 2Essex Pharma, Munich, Germany OBJECTIVES: A me-too list has been implemented in a German KV (regional association of SHI-accredited physicians) with the intention of promoting cost-effective prescribing behavior. The aim was to investigate whether the me-too list is an effective steering tool. a mometasone-containing nasal spray (MHN) which has been listed as a me-too, should be substituted by budesonide-containing nasal sprays (BHN), because MHN therapy is assumed to be equally effective but more cost intensive than BHN. MHN’s market share has declined after listing. METHODS: A retrospective study of anonymized patient data was conducted from March to April 2009 in KV North Rhine. ENT physicians documented patients which had received MHN for minimum 12 months and who had then been switched to another intranasal steroid (INS). Evaluation was done with descriptive statistics. RESULTS: A total of 676 patients documented by 76 physicians were eligible for the analysis. a total of 189 patients (28%) received MHN only; a switch to other INS was documented for 487 patients (72%). ENTs cited economic reasons for changing the INS in 86.4% of the cases. Compared to other INS, the average number of consultations for MHN treatment and the administered dose of MHN, regardless of diagnosis, were lower, MHN patients required refill prescriptions later, less co medication, and MHN generated lower daily treatment costs in real-life. CONCLUSIONS: Reasons for switching from MHN were economically. Therefore the me-too list has a heavy impact on prescription behavior. Daily treatment costs of MHN were lower in real-life. Analysis based on PDD (prescribed daily doses) showed that the economic effect of switching a patient to another INS was far less than expected. For instance it was offset by a higher number of consultations in association with the cheaper alternatives in more frequent use. An economic evaluation with a precise quantification would be of major interest in future. PRS45 OPTIMIZATION OF GUIDELINES’ SUSTAINABILITY BY EVALUATING FACTORS PREDICTING TREATMENT RESPONSE IN PATIENTS WITH ALLERGIC RHINITIS Koeberlein J1, Krajewski J1, Schaffert C1, Wieland R1, Moesges R2 1 University of Wuppertal, Wuppertal, Germany; 2University of Cologne, Cologne, Germany OBJECTIVES: To assess predictive factors and risk profiles for treatment response in patients with allergic rhinitis. It was the aim to create a platform giving support to the development of interventions optimizing the daily use as well as the sustainability of the existing national and international guidelines and therewith minimizing the progression to severe chronic upper airway diseases. METHODS: For this purpose, 76,981 case reports of patients with allergic rhinitis from ten post-marketing-studies in Germany were aggregated by the use of an IPD meta-analysis. The data pool was examined in terms of significant predictive factors by univariat methods, logistic regression and discriminant analysis. The predictive ability of the model was assessed by a 10 fold crossvalidation. RESULTS: First, we investigated the data pool as a whole, and the results showed that 20.6% of the patients did not respond to their therapy. Especially patients who were treated according to current guidelines had a higher risk of non-response (28% vs. 19.3%). Next, we examined the data pool in terms of significant predictive factors resulting in non-response of patients receiving an adequate guideline-conform therapy. The primary baseline determinants were a positive history of allergy in the family, duration and severity of disease, the periodicity of an allergen (perennial allergens) and concomitant diseases. Especially patients with asthma as well as patients suffering from more than one concomitant disease had a higher risk of non-response (32.4% and 34%). The main follow up determinants were patient’s compliance, safety and the treatment process. a total of 63.4% of patients with a poor compliance, which was enforced by an inadequate safety were nonresponder. CONCLUSIONS: These results support the findings of current clinical trials describing a non-responder rate of about 20% and encourage efforts to optimize guidelines as well as to achieve a sustainability of guidelines. RESPIRATORY-RELATED DISORDERS – Conceptual Papers & Research on Methods PRS46 ADJUSTING FOR CHRONIC OBSTRUCTIVE PULMONARY DISEASE (COPD) SEVERITY IN DATABASE RESEARCH: FEASIBILITY OF DEVELOPING AND VALIDATING AN ALGORITHM Goossens LM1, Baker CL2, Monz BU3, Zou KH2, Rutten-van Mölken MP1 1 Erasmus University, Rotterdam, The Netherlands; 2Pfizer, Inc, New York, NY, USA; 3 Boehringer Ingelheim GmbH, Ingelheim, Germany OBJECTIVES: When comparing interventions through routine database research, it is important to adjust for COPD severity. GOLD guidelines grade COPD severity according to lung function (FEV1% predicted). Most databases lack data on lung A327 function. Previous database research has approximated COPD severity using demographics and health care utilization. This study aims to derive an algorithm for COPD severity using baseline data from a large, respiratory trial (UPLIFT). METHODS: Several partial proportional odds logit models were used to estimate the probabilities of being in GOLD stages II, III and IV. The predicted GOLD stage for each patient was defined as the state with the highest predicted probability. The concordance between the predicted and the observed (according to lung function) stage was assessed with κ-statistics. Models were estimated in a random selection of 2/3 of all patients enrolled in Western Europe (n = 2439) and validated in the remaining patients (n = 1259). The final model was re-estimated in a subsample with a balanced distribution across severity stages. RESULTS: In the validation set, 47% were in stage II, 44% in III, and 9% in IV. In the final model, a higher risk of more severe COPD was associated with being male, younger, lower BMI, and certain medications (long-acting and short-acting bronchodilators, leukotriene modifiers, oral steroids), oxygen and having quit smoking. No relationship was found between severity and co-morbidities, previous health care resource use (e.g. emergency room, hospitalizations) and inhaled corticosteroids, xanthines, or mucolytics. The concordance between observed and predicted disease states was generally poor (κ = 0.15) and only slightly better in the balanced sample (κ = 0.22). CONCLUSIONS: Data from a well controlled trial setting indicated that COPD severity cannot be reliably predicted from demographics and health care use. This limitation should be considered when interpreting findings from database studies, and additional research should explore other methods allowing to account for COPD severity. PRS47 USE OF PROPENSITY SCORE MATCHING, STANDARD REGRESSION ANALYSIS AND INSTRUMENTAL VARIABLE METHOD IN OUTCOMES RESEARCH STUDIES: A COMPARATIVE ANALYSIS Baser O1, Dysinger A2, Baser E2, Yuce H3 1 STATinMED Research/University of Michigan, Ann Arbor, MI, USA; 2STATinMED Research, Ann Arbor, MI, USA; 3New York City College of Technology-CUNY/STATinMED Research, New York, NY, USA OBJECTIVES: Propensity score matching and standard regression analysis are common ways to control for baseline differences between comparative groups. They control for observable factors. Instrumental variable approach controls not only for observable factors but also unobservable factors. We compared the three methods and showed that using more advanced techniques alters the estimated results in a significant way. METHODS: Using data from U.S. claims databases, the effect of treatment on total health care expenditures among asthma patients was estimated. Reimbursement amounts were dollars paid by the health plan to the provider including patient co-payments and deductibles. Doctors’ prescribing patterns were used as an instrumental variable for treatment choice. Propensity score matching was employed using the nearest neighbor matching algorithm. Generalized linear model was used as an alternative risk adjustment technique. RESULTS: Patients treated with control medication were younger, more likely to live in the northeast and south of the United States and have a higher Charlson comorbidity score, Elixhauser score and chronic disease score relative to patients treated with reliever medication. The difference between one year health care costs for reliever and controller medication was $2,345 by propensity score matching, $2,195 by generalized linear model, and $2,997 by instrumental variable approach. The difference was statistically significant. CONCLUSIONS: After adjusting for patient clinical and demographic characteristics, controller medication was less costly than reliever medication. The choice of risk adjustment was important. The technique that controlled for both observed and unobserved biases (instrumental variable technique) provided a difference of almost 30% higher than the other techniques. PRS48 COMPARISON OF ECONOMETRIC MODELS FOR ESTIMATING COST DIFFERENCES: OLS, GAMMA, AND QUANTILE REGRESSION de Moor C1, Roberts M2, McQueeney K2, Blanchette C3 1 PPD Inc., Wilmington, NC, USA; 2Lovelace Respiratory Research Institute, Albuquerque, NM, USA; 3Lovelace Respiratory Research Institute, Davidson, NC, USA OBJECTIVES: Estimation of cost differences between groups has been assessed using a variety of econometric techniques. Recent literature commonly uses ordinary least squares (OLS) with log transformed costs or generalized linear models with a log link function and gamma error distribution; however recommendations from the field of econometrics suggest that quantile regression models may provide more informative comparisons of health care cost data. The objective of this study was to compare three econometric models for estimating health care cost differences. METHODS: The econometric models were compared using a cross-sectional sample of patients hospitalized during 2007 for COPD exacerbations from Premier’s Perspective Comparative Database, a population-based U.S. inpatient database. Admissions were classified into two groups according to the presence of pneumonia. Total inpatient cost was regressed on pneumonia status and confounders using: 1) OLS with log-transformed costs; 2) generalized linear models with log link and gamma error; and 3) quantile regression with comparisons at every 5th percentile. Cost differences were calculated by subtracting model based estimated costs between COPD patients with and without pneumonia. RESULTS: There were 69,841 COPD exacerbation admissions. Of these, 69,286 admissions had nonzero costs, and 6,840 (9.9%) had evidence of pneumonia. In the OLS model, estimated costs were $3,949 (95%CI: $3,822–$4,076) higher in patients with pneumonia; in the log-gamma model, the estimated costs were $4,442 (95%CI: A328 $4,297–$4,587) higher; while the median quantile difference was $3,756 (95%CI: $3,636–$3,876). The quantile regression differences ranged from a low of $2,078 (95%CI: $2,012–$2,145) for the 5th quantile to a high of $8,691 (95%CI: $8,395– $8,987) for the 95th quantile. CONCLUSIONS: Results from econometric regression models can vary greatly depending on model assumptions. Quantile regression results provide a detailed picture of cost differences across the entire cost distribution, illustrating important non-uniformities of differences between groups. PRS49 TESTING THE USABILITY AND LANGUAGE OF E-PRO TRANSLATIONS DURING LINGUISTIC VALIDATION: COGNITIVE DEBRIEFING ON E-DEVICE VS. PRINT OUTSCREENSHOTS Grataloup G1, Fernandez N1, Dale P2, Muehlhausen W3 1 MAPI Institute, Lyon, France; 2GlaxoSmithKline, Middlesex, UK; 3Cardinal Health Research Services, Hoechberg, Germany OBJECTIVES: Electronic data collection is a prominent evolution in the PatientReported Outcomes (PRO) field. Developing translations for e-formats must involve a rigorous methodology ensuring conceptual equivalence and cultural relevance across languages. Ideally comprehension and acceptability of translations should be tested using the actual e-device. For logistical reasons this may be challenging. Alternative methods include testing of translations using print outs of screenshots. The translation of an Asthma Diary into 20 languages was the opportunity to compare both methods. METHODS: (1) Identification of the languages using e-device testing versus those using screenshots; (2) comparison of the interview guides used for the respective testing methods; and (3) analysis and comparison of the results. RESULTS: A total of 15 languages were tested on the e-device, 5 with screenshots. The interview guide used to test the e-device contained linguistic considerations and questions on instructions and use of the device. The interview guide for the screenshots was purely linguistic in nature, so as not to confuse respondents with questions about the hypothetical use of the e-device. In terms of linguistic observations no differences were seen between the e-device and the screenshot testing methods. Feedback on the device usability differed by country but also across regions. In most countries/languages observations were made around the size of the font/screens and the device instructions. CONCLUSION: Testing translations using the actual e-device not only provides comprehensive feedback on the language of the items, but also information on the practical use of the device, and the interaction of both in the context of each country. Screenshot testing can be considered an alternative in cases where e-device testing is difficult. PRS50 EMPLOYING MULTI-ATTRIBUTE UTILITY THEORY TO DEVELOP THE EXACT-U: A PREFERENCE-BASED MEASURE FOR COPD EXACERBATION UTILITIES Petrillo J1, Cairns J1, Revicki DA2 1 London School of Hygiene and Tropical Medicine, London, UK; 2United BioSource Corporation, Bethesda, MD, USA BACKGROUND: The EXACT (Exacerbations of COPD Tool) is a daily diary used in clinical trials to evaluate frequency, severity, and duration of COPD exacerbations. Reporting utilities from the EXACT allows a more accurate account of preference change for economic evaluations than current methods. Multi-attribute utility theory (MAUT) employs a series of equations to develop a function (MAUF) to report utilities. OBJECTIVES: To develop and validate an MAUF to estimate utilities from the EXACT for use in UK cost-effectiveness studies. METHODS: EXACT-U is comprised of 5 items with 3–5 levels each. Items and levels were grouped to form mixed-level and corner health states. Development group was used for MAUF construction; separate Validation group was used to test functions. UK general public respondents valued 11 health states (including best/worst) using TTO from full health/dead over 10 years. MAUF used the multiplicative model by: (1)-Calculating mean utility of each attribute level; (2)- Calculating group disutilities for each level; (3)-Applying multiplicative model to derive disutility function; and (4)-Converting disutility function to utility function. Performance assessed by: number of inconsistencies predicted, mean absolute error (MAE), and root mean squared error (RMSE). Models validated using a secondary analysis of a separate data set of EXACT patient data to test discriminant validity (statistical significance) and responsiveness (standardized response mean (SRM)). RESULTS: TTO interviews conducted with 400 respondents: 350 Development group, 50 Validation group. Respondents were: 36 yrs old (13.5 SD), 39.2% male, and 46.4% White British. MAUF reported MAE = 0.032, and RMSE = 0.170. Discriminant validity supported by utility differences by clinical severity: stable/acute (P = 0.001); mild/moderate (p = 0.01); moderate/severe (p = 0.0001); and severe/very severe (P = 0.14). Responsiveness by SRM was 0.52 (Day 3), 0.55 (Day 7), 0.66 (Day 10), and 0.76 (Day 13). CONCLUSIONS: The EXACT-U is a condition-specific preference-based measure to report COPD exacerbation utilities with minimal error, good discrimination, and good responsiveness. 13th Euro Abstracts PRS51 ADVANCED PATTERN RECOGNITION METHODS FOR PREDICTING TREATMENT RESPONSE IN PATIENTS SUFFERING FROM ALLERGIC RHINITIS Krajewski J, Koeberlein J University of Wuppertal, Wuppertal, Germany OBJECTIVES: The aim of the present analysis was to optimize the prediction of treatment response in patients with allergic rhinitis. To determine an optimal prediction accuracy, we used different Pattern Recognition (PR) approaches and evaluated their added value compared to standard predictive models as Logistic Regression, and Linear Discriminant Analysis. METHODS: In order to optimize the prediction of treatment response, 76,981 case reports of patient with allergic rhinitis from ten postmarketing-studies in Germany were analyzed by means of PR methodology. The processing steps applied within this study are: (a) feature extraction (genetic algorithm based synthetic feature calculation), (b) dimensionality reduction (correlation filter based feature selection, wrapper based feature selection, Principle Component Analysis based feature transformation), (c) classification (Support Vector Machine, Decision Tree, K-Nearest Neighbor, Random Forest, Artificial Neural Network, Bagging, Boosting Logistic Regression, Linear Discriminant Analysis), and (d) validation (leaveone-sample-out cross validation). RESULTS: The AdaBoost Support Vector Machine classifier with correlation filter based feature selection achieved the highest unweighted mean recall rate (mean of sensitivity and specificity; URR) of 62.8%. The standard Logistic Regression approach yielded 50.4% (−12.4%), the Linear Discriminant Analysis 50.6% (−12.6%). CONCLUSIONS: In comparison to standard learning schemes as e.g. Logistic Regression, and Linear Discriminant Analysis applying advanced PR methods improves substantially the prediction of treatment response in patients with allergic rhinitis. Due to the achieved added value and the superiority of Pattern Recognition methods within several benchmarking studies, advanced PR methods should be primarily considered for modeling and prediction tasks within the field of pharmacoeconomics. POSTER SESSION II CONCEPTUAL PAPERS & RESEARCH ON METHODS – Clinical Outcomes Methods PMC1 BEST PRACTICES IN REPORTING OF PROSPECTIVE OBSERVATIONAL STUDIES Bharmal M, Viswanathan S, Gemmen E Quintiles, Rockville, MD, USA OBJECTIVES: Although a number of guidelines are available for the reporting of clinical trials and interventional studies, there is limited consensus in the structure, content and terminology associated with reporting of prospective non-interventional observational studies. The objective of this study was to describe best practices in reporting of observational studies based on a review of relevant guidelines for reporting data from clinical studies. METHODS: A systematic literature review was conducted and six relevant guidelines that could be adapted for reporting of results from prospective non-interventional observational studies were identified. The guidelines reviewed included the FDA guideline for abbreviated clinical study reports (CSR), the EMEA guideline for CSR, the International Conference on Harmonization (ICH) E3 guideline for CSR, Consolidated Standards of Reporting Trials (CONSORT) guideline for pragmatic trials, the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guideline and the International Society for Pharmacoepidemiology (ISPE) guideline for safety reports. The structure, content, and terminology used in the reporting template recommended by the various guidelines were analyzed. RESULTS: There were substantial differences in the structure, content and terminology recommended by the six guidelines. For example, in contrast to all other guidelines, the CONSORT guidelines mention “Objectives” under “Methods” instead of “ ‘Introduction.” Under the Results section, for reporting of unadjusted and adjusted estimates from outcomes data, the STROBE guideline recommends its inclusion within the Main Results section, whereas the CONSORT guideline recommends its inclusion within the Outcomes and Estimation section. Some guidelines, especially those focusing on interventional studies, were more similar in content. CONCLUSIONS: Based on the evaluation of similarities and differences in the guidelines, we propose a structure and template for reporting of prospective non-interventional observational studies. Recommendations are also provided for adapting the proposed template based on study objectives and design. PMC2 ESTIMATING COST-EFFECTIVENESS BASED ON RESULTS OF UNCONTROLLED CLINICAL TRIALS: OFATUMUMAB FOR THE TREATMENT OF FLUDARABINE- AND ALEMTUZUMAB-REFRACTORY CHRONIC LYMPHOCYTIC LEUKEMIA Batty AJ1, Thompson GJ2, Maroudas PA2, Delea TE3 1 BresMed Health Solutions, Sheffield, UK; 2GlaxoSmithKline, Uxbridge, Middlesex, UK; 3Policy Analysis Inc., Brookline, MA, USA OBJECTIVES: Increasingly, innovative oncology drugs are licensed in settings where randomised controlled trials (RCTs) are ethically and/or practically infeasible. The 13th Euro Abstracts lack of an RCT makes formal technology assessment vs. alternative treatment (e.g., best supportive care [BSC]) challenging. In such instances, naïve indirect comparison based on historical controls is typically employed. We present a method for estimating outcomes for untreated patients when appropriate historical controls are not available, by using data from non-responders in an uncontrolled trial. METHODS: Ofatumumab was licensed for fludarabine- and alemtuzumab-refractory chronic lymphocytic leukaemia (FA-Ref CLL) based on results of an uncontrolled trial (Hx-CD20-406). To evaluate the cost-effectiveness of ofatumumab vs. BSC from the UK National Health Service perspective, a partitioned survival analysis model was developed. Progression free survival (PFS) and overall survival (OS) for ofatumumab were estimated by fitting Weibull survival functions to failure time data for all FA-Ref patients in Hx-CD20-406. Following a literature search, no suitable historical control representing BSC could be identified; therefore hazard ratios for PFS and OS for BSC vs. ofatumumab were estimated by fitting Cox regression models to data for non-responders vs. all FA-Ref patients. Costs and utilities were taken from both published and unpublished sources. RESULTS: BSC patients (approximated by non-responders) were estimated to achieve 4.7 months PFS, 11.3 months OS, 0.50 QALYs, and expected lifetime costs of £4,876. Ofatumumab patients were estimated to reach 6.5 months PFS, 17.9 months OS, 0.77 QALYs, with expected lifetime costs of £43,828. CONCLUSIONS: The novel approach presented permits a practical alternative for estimating cost-effectiveness when neither an RCT nor appropriate historical control can be identified. Further research should be conducted using established data sets to validate the methodology, and to address potential limitations, e.g. unobserved differences between treatment groups, and potential benefits of treatment in patients classified as non-responders. PMC3 A NOVEL APPROACH TO MATCHING ADJUSTED INDIRECT COMPARISON ANALYSIS USING COMMON SAS 9.2 PROCEDURES Malangone E1, Casciano R1, Sherman S1, Berenson K1, Stern L1, Di Lorenzo G2 1 Analytica International Inc., New York, NY, USA; 2University Federico II of Naples, Napoli, Italy OBJECTIVES: While randomized control trials (RCT) are the gold standard for drug approval, there is often a lack of data directly comparing different treatment options. An indirect comparison of the treatment effects may serve as a proxy for a head-tohead RCT, however, naively comparing treatments using published trial data without adjusting for distribution differences in patient characteristics and prognostic factors can result in misleading conclusions. a novel matched-adjusted approach to indirectly compare absolute survival estimates (median overall survival (OS) or progression free survival (PFS)) for competitive treatment options is presented. METHODS: This proposed approach requires patient-level data for one of the treatments and summary data of patient characteristics and survival outcomes for the comparator of interest. Using this proposed method, the researcher would first decide on one or two matching variables that are prognostic for survival, and apply a program involving an extension of a common SAS 9.2 procedure, Proc Surveyselect, to select 1000 random repeated sub-samples from the original population with the same distribution of matched variables. The analysis also requires programming statements using ODS and survival analysis procedures. The median OS or PFS estimates are computed for each bootstrapped sample and a 95% confidence interval (CI) is inferred around the mean of the sampled survival estimates. These absolute survival estimates, based on the adjusted population, can then be compared to the absolute survival estimates reported in published literature of the comparator treatment. CONCLUSIONS: In the absence of head-to-head RCT data, an adjusted indirect comparison accounts for observed differences between populations making them more comparable and results in an effect of treatment exposure on survival outcomes that is less likely due to confounders. CONCEPTUAL PAPERS & RESEARCH ON METHODS – Cost Methods PMC4 COST ESTIMATION IN HEALTH ECONOMIC EVALUATIONS IN GERMANY: A SYSTEMATIC REVIEW Merito M, Breitscheidel L, Eichmann F Kendle GmbH, Munich, Germany OBJECTIVES: The objectives of this study are: 1) to systematically review the methods used in developing cost estimates in the recent German health economic literature; and 2) to examine the methodological approaches in terms of analytical framework, cost components, resource use and cost data sources in light of the national Institute for Quality and Efficiency in Healthcare guidelines. METHODS: The MEDLINE database was searched for studies published between 1-Jan-2006 and 31-Dec-2008 estimating direct and/or indirect costs of health care interventions in Germany. Detailed information on the perspective of the analysis, time horizon, resource use categories, costing approach, valuation of resource use, resource use and unit cost/ price data sources were systematically collected. RESULTS: The literature search returned citations to 122 articles, of which 47 met the inclusion/exclusion criteria. Nearly half of the selected articles (23) adopted the societal perspective and 21 (44.7%) the perspective of German Statutory Health Insurance funds. In nearly three quarters of the studies (35) the time-horizon of the base-case analysis was no longer than one year. Direct medical costs were reported in almost all articles; 25 studies (53.2%) assessed only this cost category. Among the most common unit cost sources of health care services were reimbursement fees for outpatient and hospital services, together with market prices for drugs and medical devices, mentioned in 43 (91.5%) A329 and 39 (83.0%) articles, respectively. CONCLUSIONS: Although the variability in the studies reviewed was very high, some common features emerged. Firstly, most cost estimates focus on direct medical costs only. Secondly, a comprehensive list or at least an exhaustive discussion of the relevant resource use is usually missing. Finally, the time-horizon is often too short to capture all relevant cost drivers and the assessment of the medical resource use not always consistent with the perspective of the analysis. PMC5 DEVELOPMENT OF A WEB-BASED SOFTWARE TOOL TO EVALUATE THE ECONOMIC IMPACT OF LOST PRODUCTIVITY DUE TO PREMATURE MORTALITY IN DEVELOPED AND EMERGING NATIONS Marton JP2, Menzin JA1, Willke RJ2 1 Boston Health Economics, Inc., Waltham, MA, USA; 2Pfizer, Inc, New York, NY, USA OBJECTIVES: Economic analyses that take a societal perspective need to incorporate estimates of lost productivity due to premature death. Such estimates are likely to vary substantially across countries, making it a challenge to assess the value of alternative medical interventions on a global basis. Our goal was to develop a generic, web-based software tool based on rigorous analytic methods that would enable researchers to assess the expected discounted present value of lost productivity for persons who die prematurely at various ages in selected developed and emerging nations. METHODS: An analytic model framework was developed to estimate the expected discounted present value of lost productivity due to premature mortality from a societal perspective using a human capital approach (with value attached to household work, as data permitted). Key model inputs included life tables, labor force participation rates, wages, and discount rates. Default input parameter values were based on each country’s national statistics, as available, or via generic “global” estimates when such data were lacking. RESULTS: Model results were generated for 20 countries around the world, and varied substantially based on patient age at death and the economic region in which a country is located. For illustration purposes, the discounted present values of lost productivity for a person who dies at age 25–29 in the US, Brazil, and Sweden were estimated to be $US 945,162, 124,795, and 805,740, while the corresponding values for those who die at age 45–49 were 622,248, 76,976, and 480,480. The webbased interface allows researchers to select the country of interest, modify default values, and conduct sensitivity analyses. CONCLUSIONS: This generic web-based software tool allows researchers to easily incorporate the value of lost productivity due to premature mortality into economic analyses that take a societal perspective, and provides estimates for many different developed and emerging countries. PMC6 ASSESSING PRODUCTIVITY AND ACTIVITY IMPAIRMENT DUE TO ILLNESS IN POLAND Wrona W, Hermanowski T, Jakubczyk M, Golicki D, Macioch T Department of Pharmacoeconomics, Medical University of Warsaw, Warsaw, Poland OBJECTIVES: The inclusion of loss productivity costs in pharmacoeconomic studies is still a subject of considerable debate. The aim of this study was to quantify the work impairment due to general health status in Poland with the Productivity and Activity Impairment: General Health (WPAI-GH) Questionnaire. METHODS: Data were obtained from a survey that incorporated the WPAI-GH questionnaire and information on burden of care for a sick family member during computer-assisted personal interview in a representative sample of the Polish general population aged more than 15 years. There were 2019 respondents in total, gathered in two waves in January and May 2010. RESULTS: The total population comprised 795 subjects in paid jobs. Subjects reported 4.5% work time missed (absenteeism) during the past 7 days. Impairment while at work (presenteeism) amounted to 13.9% of total time. The overall work productivity loss (absenteeism plus presenteeism) equalled 15.2%. Impairment in performing daily activity was 15.6% in the past 7 days. Observed percentages were in general higher in subject from the first wave of study (January 2010) than from second wave (May 2010)—differences did not reach statistical significance. The general tendency of a higher absenteeism and a lower presenteeism values among men than among women were observed. CONCLUSIONS: Productivity and Activity Impairment measured by WPAI-GH in the Polish population are similar to these observed in other European countries and the U.S. Moderate differences between values estimated in January and May suggest limited impact of seasonal diseases such as influenza on productivity. PMC7 TARIFF LISTS FROM SPANISH AUTONOMOUS COMMUNITIES: AN ASSESSMENT OF ITS STRUCTURE, CONTENTS, AND TARIFF LEVELS De Cock E1, Raluy M2, Rovira J3 1 United BioSource Corporation, Barcelona, Spain; 2United BioSource Corporation, London, UK; 3Barcelona, Spain OBJECTIVES: There exists no standardised list of unit costs (UC) for use in economic evaluations in Spain. Tariffs published by the 17 Autonomous Communities (ACs) are often used as a proxy for costs. We explored the structure and contents of AC tariff lists and tariff ranges for common resource use items. METHODS: Current tariff lists published in the Official Bulletins from 16 Spanish ACs were retrieved. Tariffs for key health services in the following categories were extracted: specialist and A&E visits, hospitalization, investigations, procedures, laboratory tests, and episodes of care (DRGs). We qualitatively assessed structure and contents of tariff lists, item content for selected items. Ranges, normal mean and weighted mean (according to A330 demographic size of ACs) were calculated for selected tariffs. RESULTS: We observed lack of consistency in structure and contents of tariff lists and in tariff levels, as exemplified by: different categorization of services; incomplete listing of services; different terminologies; different levels of detail for common services (e.g. MRI: 2–127 options by body area and/or complexity); wide ranges of tariffs for most items (e.g. specialist visit: c56–c191, general ward per diem: c82–c569, simple MRI: c120–c634). Wide variations were also observed for other diagnostic investigations, (non-)surgical procedures, laboratory tests and DRGs. CONCLUSIONS: Wide ranges in tariffs for health resources commonly used in economic evaluations were identified across ACs, with a difference between minimum and maximum values of at least factor 2. There exists no evidence on how tariffs are calculated and if they reflect real cost. Available AC tariffs should be used with caution and a simple or weighted average across AC tariff lists may be used as proxies for nation-wide costs. Elaboration of a nation-wide list would avoid possible bias from analysts in the selection of cost values to obtain given results. PMC8 BURDEN OF DISEASE AND ECONOMIC EVALUATION: ARE WE INVESTIGATING WHAT IT REALLY MATTERS? Catalá-López F1, Garcia-Altés A2, Álvarez-Martín E3, Gènova-Maleras R4, Morant-Ginestar C4, Parada A5 1 Spanish Medicines and Healthcare Products Agency (AEMPS), Madrid, Spain; 2King’s College London, London, UK; 3Rey Juan Carlos University, Madrid, Spain; 4Regional Health Council of Madrid, Madrid, Spain; 5Agencia de Evaluación de Tecnología e Investigación Médicas (AATRM), Barcelona, Spain OBJECTIVES: We examined the association between economic evaluation studies performed in Spain in 1983–2008 and the burden of disease in the population. METHODS: Cross-sectional observational study. Electronic databases (Pubmed/ MEDLINE, SCOPUS, ISI Web of Knowledge, CRD, IME, IBECS) and reports from public agencies were systematically reviewed. Inclusion and exclusion criteria and a set of variables were defined to analyze the characteristics of the papers selected. Using the Global Burden of Disease (GBD) study classification the following measures were calculated: years of life lost (YLLs), years lived with disability (YLDs), disabilityadjusted life-years (DALYs), and mortality by cause. Correlation and linear regression models were used. RESULTS: Cardiovascular diseases (15.7%), infectious and parasitic diseases (15.3%), and malignant neoplasms (13.2%) were the conditions most commonly addressed. Accidents and injuries, congenital anomalies, oral conditions, nutritional deficiencies and other neoplasms were the categories with a lowest number of studies (0.6% from the total for each of them). The disease sub-categories most prevalent in the studies were lower respiratory infections (5.7%), ischemic heart disease (5.7%), hepatitis B and C (3.3%) and HIV/AIDS (3.1%). For GBD categories (n = 20), a correlation was seen with: mortality 0.67 (p = 0.001), DALYs 0.63 (p = 0.003), YLLs 0.54 (p = 0.014), and YLDs 0.51 (p = 0.018). By diseases sub-categories (n = 51), the correlations were low and non statistically significant. CONCLUSIONS: There is a mild-moderate association of economic evaluations with the main causes of burden of disease. For some conditions, the data show over or under-representation of studies related to their burden generated. The burden of disease is a criterion that, in combination with efficiency and equity, would allow to set recommendations to guide debates on health research priority setting. PMC9 A NEO-RICARDIAN APPROACH TOWARD DISCOUNTING Parouty M, Boersma C, Postma MJ University of Groningen, Groningen, The Netherlands BACKGROUND: The major focus of the history of economic thought has been devoted to defining a scientific theory of value. An even harder task entails formulating a theory of intertemporal value. Work on this theory date back to the birth of modern economic thought. For example, some authors have argued that an intertemporal utility based theory of value involves the explanation of a quantity which can be directly observed and measured in terms of a quantity which cannot. Major current debates on discounting therefore surround the need for a scientific definition and the ethics towards intergenerational justice. OBJECTIVES: We investigate how several current issues in discounting might be irrelevant by adopting a Neo-Ricardian view of intertemporal value by recursively applying valuation of a commodity from value of input commodities, thereby simplifying to the Physiocratic School. We further investigate how our empirical model might be extended to current utilitarian philosophy. METHODS: We adopt a Sraffian approach and devise a pure value-growth matrix that relies only on empirical data. We first derive a 2 × 2 matrix and then a 3 × 3 matrix. We use health effects, income and a third externality for the derivations. RESULTS: It seems that the Neo-Riccardian approach provides the necessary requirements towards satisfying a scientific definition of intertemporal value and allows extension of the classical framework of health and wealth with a 3rd dimension of externalities. Furthermore, by redefining the pure growth term in our 2 × 2 matrix with the Ramsey discount rate, our results simplify to current economic theory. CONCLUSIONS: Although modern economic theory explains value from a utilitarian viewpoint, it seems that it lacks robustness in explaining intertemporal value. Therefore, we suggest that the Sraffian School of economic thought should also be considered when attempting to formulate a discount rate for health effects, within a concept of sustainable growth. 13th Euro Abstracts PMC10 RELIABILITY OF MANUFACTURERS’ BUDGET IMPACT ESTIMATES IN POLAND Iwanczuk T, Szewczyk K, Zagorska A Agency for Health Technology Assessment in Poland, Warszawa, Poland OBJECTIVE: To verify the reliability of the methodology used in the manufacturers’ Budget Impact Analyses (BIAs) submitted in reimbursement dossier to Polish HTA Agency, we compared public payer (National Health Fund—NHF) actual expenditures on selected drugs with their estimated costs. METHODS: BIAs of medicines reimbursed for at least one year and assessed previously in Agency were selected for the analysis. The BI estimated by the manufacturer in the first year after product’s introduction to reimbursement was used. Estimated size of target population was compared with the actual one. The actual data was obtained from the NHF. BIAs for selected medicines were critically appraised to determine any variables that may have affected its reliability. RESULTS: 20 BIAs met inclusion criteria and were included into the study. Forty-five percent (9/20) of the BI expenditures were underestimated. Median and mean difference between actual and estimated expenditure were 17,22% and 30,25% respectively. The overestimation was found in 55% BIAs (11/20); median = 262,26%, mean difference = 2267,25%. Population size was underestimated in 65% (13/20) of the BIAs. Median and mean difference between actual and estimated population were 25,61% and 33,93% respectively. Overestimation of the population was found in 35% (7/20) of the BIAs; median = 566,67%, mean difference = 573,28%. The main factors that could influence differences between predictions and actual spending were: underestimated (65%) or overestimated (35%) number of patients eligible for treatment, overestimated (30%) or underestimated (25%) market share, wrong assumption on 100% compliance (15%). CONCLUSION: The study has demonstrated large variances between predicted budget impact and actual expenditures on drugs. It also revealed significant weaknesses in the quality of submitted BIAs, e.g.: errors in calculations, very limited data provided by manufacturer that unable to complete revision and reproduce of figures in the calculation. PMC11 ARE THRESHOLD RANGES FOR COST PER QALY A BARRIER TO RESEARCH FOR LIFE EXTENDING TREATMENTS Roberts G, de Nigris E Double Helix Consulting Group, London, UK As they are currently used thresholds for cost per QALY may provide a disincentive for companies to invest in research for therapies that prolong life in conditions with an already high treatment cost. Cost per QALY thresholds, although not the sole basis for decision making are a major influence on whether a technology is considered cost-effective by NICE. Discussions have centred on the most appropriate threshold level and how its value should be determined. However a consequence of cost per QALY thresholds that is not discussed is the impact they may have on future health care research. The cost per QALY for renal dialysis has been estimated at £30,0001, the higher end of what NICE considers acceptable. We have therefore reached the ceiling for the cost of treating renal disease. Assuming that utility is not improved a treatment that extends life will be at additional cost and have a cost per QALY greater than £30,000. Manufacturers of health care technologies may consider that the risk of not getting a product approved on cost-effectiveness grounds is not worth the financial investment in its development. As health care costs continue to grow the management costs of more conditions will exceed £20,000 per year and future research may be stifled as manufacturers seek to develop products that replace rather than add to current treatments. Since the background treatment cost would cancel out in an incremental analysis a treatment could be more cost-effective than the one it replaces but perversely can still be at an increased cost which raises the cost per QALY of standard treatment above £20,000 or £30,000. As treatment costs for more conditions increase to threshold values (even if they are raised) manufactures may be advised to consider realigning their portfolio and investment to other diseases. PMC12 ASSESSMENT OF THE WORKLOAD REAL TIME DEDICATED TO EACH PATIENT IN INTENSIVE CARE UNITS (ICU): PRELIMINARY RESULTS OF THE CRRÉA STUDY Garrigues B1, Lefrant JY2, Pribil C3, Bazin J4, Maurel F5 1 Centre Hospitalier du pays d’Aix, Aix-en-Provence, France; 2CHU de Nimes, Nimes, France; 3GSK France, Marly le roi, France; 4CHRU de Clermont-Ferrand, Clermont-Ferrand, France; 5IMS Health, Puteaux, France OBJECTIVES: The objective of the CRRéa study is to assess the real daily cost of a patient’s stay in ICU in France. We present here preliminary results regarding the average time spent per patient by different health caregivers. METHODS: A prospective multicentric health economic study was performed in 23 ICUs of different French hospitals randomly selected from the PMSI database (French National Hospital database). In a one day study, 5 adult patients were randomly selected among patients with a simplified severity score ≥ 15 in each ICU. Data on all the resources used, treatments administered, biological tests performed, etc. and time spent by different health caregiver to take care of each patient over a 24 hour period (direct and indirect interventions) were collected through a time and motion analysis method involving the professionals themselves. RESULTS: A total of 109 patients (median age = 66 years, 65% males) of 22 intensive care units (15 polyvalent, 3 surgical and 4 medical ICUs) were included. 104 of them were followed over 24 hours (there were 2 deaths and 3 early withdrawals). On the day of the study, 84% of patients were mechanically 13th Euro Abstracts ventilated with a median SOFA score = 6. The median cumulated time dedicated to one patient by physicians, nurses and caregivers was 10h20 over the 24-hour period (1h15 by physician, 6h08 by nurses and 2h57 by caregivers). CONCLUSIONS: The median time of more than 10 hours directly dedicated to a patient is a key information for the estimation of the real cost of one day stay in ICU PMC13 METHODS AND IMPACT OF INCORPORATING MEDICATION COMPLIANCE INTO PHARMACOECONOMIC EVALUATIONS Park SY, Lee EK SookMyung Women’s University, Seoul, South Korea OBJECTIVES: This study aims to identify how medication compliance and/or persistence were assessed in the cost-effectiveness analysis, and what the impact was on ICER(Incremental cost-effectiveness ratio). METHODS: Pharmacoeconomic studies with compliance and/or persistence measures, had published from March 2005 to February 2010, were searched through MEDLINE. Articles were included if they integrated medication compliance and/or persistence into the economic analysis model. We reviewed the target diseases, the model designs, and the impact of noncompliance on the treatment costs and effects. After that, the results were compared with a previous review article of ISPOR MCP(Medication compliance and persistence special interest group) had conducted in 2007. RESULTS: The search identified 77 articles, and 10 of them were selected. Even though the overall kinds of target diseases were different, most of them were chronic diseases, which have remission and relapse as common characteristics. Variety of modeling techniques such as decision-analysis and Markov model, DES (Discrete event simulation) were used for the evaluations. In decision-analysis models, the branches of decision trees represented different level of compliance. In case of Markov models, transition probabilities assumed to be higher for those patients who were non-persistent or non-adherent to treatment. Finally, considering the effect of compliance and/or persistence, it may cause decrease of ICER for new intervention. CONCLUSIONS: We found that incorporating compliance and/ or persistence into economic evaluations lead to favorable results to new intervention. However, there was a lack of methodological rigor and consistency in definition. Therefore, development of guidance is needed for measurement, analysis, interpretation, and application of compliance and persistence from variety of data sources. PMC14 DISCOUNTING HEALTH EFFECTS: A REVIEW OF THE SYSTEM Parouty M, Boersma C, Postma MJ University of Groningen, Groningen, The Netherlands BACKGROUND: Discounting health effects remains a matter of great debate these days. Currently, discussion focuses on whether health effects should be discounted at the same rate as costs or not. High discount rates for health effects are impacting negatively on the cost-effectiveness of screening and vaccination programs. Discounting health at a lower rate than wealth has however been argued to result in theoretical inconsistencies and practical unnecessary delays in implementation of health programs. Many authors have therefore assumed that there is a one-to-one relationship between health and wealth. OBJECTIVES: We investigate the rationality of several assumptions involved towards current discounting procedures. We especially investigated the assumption of a one-to-one relationship between health and wealth. METHODS: We performed a literature review to link the issues in current methods of discounting health effects with the assumptions involved. Furthermore, we analyzed other possible linkages of health rather than with wealth only. RESULTS: We noticed that although income might depict the marginal substitution between all commodities, it seems that externalities are not accounted for. Yet, research has shown that all forms of economic growth exert intrinsically negative population health effects among the communities that are most directly involved in the transformations which it entails. This may obviously impact on valuation of health states and preferences, measured using instruments as the EuroQol-5D. These arguments support differential discounting of health effects, and potential further extensions such as differential discounting of life and quality. CONCLUSIONS: Although there might be a relationship between wealth and health, it appears that externalities may play an additional role on the quality of life. Therefore, it seems that we should regard the discounting problem of health effects as an interlinked system, rather than an equation with only health and wealth and allow differential discounting of, and potentially even within, health effects. PMC15 A NOVEL METHOD FOR COMBINING THREE CURRENT ESTIMATION APPROACHES TO PHARMACEUTICAL PRICING Sheng J, Malhotra M, Vincent L, Hoschander S, Doyle J Quintiles Global Consulting, Hawthorne, NY, USA Three of the more common methods of estimating the optimal price for prospective pharmaceutical products are willingness-to-pay, value-based price appraisal, and reference price benchmarking assessments. Each method in itself, however, has inherent limitations. The objective of this study is to present a novel technique in pharamceutical price estimation for arriving at an evidence based price-point. Willingness-to-pay, as assessed through primary research, is limited by lack of knowledge of product prices and the disconnect between respondent answers and real-life price acceptance. Valuebased price appraisals, utilizing cost-of-treatment models to estimate the price at which new products are cost-effective relative to other options, are subject to error and interpretation and are rarely taken at face value by stakeholders who drive price A331 acceptance. Reference price benchmarking, looking to market analogues to gauge appropriate price-points for new products, can be a good starting point but does not take into account unique product differences, perceived or real, of new products. To address gaps and weaknesses of any given individual method, our method of determining optimal product price uses all three pricing methodologies to triangulate on a recommended price-point. Market analogues are used as a base-price starting point. a value-based cost-of-treatment model is used to determine potential cost-savings that can be offset in price. Finally, primary research is used to determine how to modify that price based on perceived differences in the target product, and the how costsavings might be considered in price. CONCEPTUAL PAPERS & RESEARCH ON METHODS – Databases & Management Methods PMC16 HOW DO THIN DEATH DATA COMPARE TO NATIONAL FIGURES FOR EACH UK COUNTRY? Blak BT, Hards M, Lee J CSD EPIC, London, UK OBJECTIVES: Primary care patient data are increasingly used for research and recording of death is important as it helps define end of follow-up and mortality may be a study outcome. This study compared recording of death in a UK primary care database with national death rates for England, Wales, Scotland and Northern Ireland (NI). METHODS: The annual number of deaths (1990 to 2008) for each country was collected after the mortality recording quality threshold from The Health Improvement Network (THIN) database. THIN holds longitudinal anonymised primary care medical records from an increasing number of practices over time and currently contains data from more than 450 practices throughout the UK. Annual age and gender specific person time was estimated and multiplied with annual age, gender and country specific death rates to derive the expected number of deaths. Observed deaths divided by expected deaths provided the standardised mortality ratio (SMR) and Byar’s approximation formula was applied to derive 95% confidence intervals (CI). An SMR close to 1.0 indicates that recorded deaths are near national rates. RESULTS: The average annual number of practices was 263.5 (standard deviation (SD):80.2) for England, 18.8 (SD:8.5) for Wales, 33.6 (SD:17.0) for Scotland, and 14.0 (SD:7.6) for NI. Average annual death rate per 1,000 population in THIN was 10.41 (SD:1.05) overall, 10.38 (SD:1.02) in England, 11.25 (SD:1.45) in Wales, 11.03 (SD:1.54) in Scotland and 8.29 (SD:0.77) in NI. The average annual SMR was 0.88 (SD:0.02) overall, 0.88 (SD:0.02) in England, 0.94 (SD:0.05) in Wales, 0.94 (SD:0.10) in Scotland and 0.91(SD:0.07) in NI. CONCLUSIONS: The observed death rates were slightly lower than expected (SMR < 1) over time and in each country, especially in England. Social deprivation impacts death rates and as unadjusted for this could explain some variation. Reasons for lower rates need to be investigated further. PMC17 ASSISTANCE COST DEPENDING ON COMORBIDITY IN PRIMARY CARE A SPANISH INTERREGIONAL LEVEL Sicras-Mainar A1, Velasco-Velasco S2, Navarro-Artieda R3, Violan-Fors C4, Blanca-Tamayo M2, Vega Martín M5, Prados-Torres A6 1 Directorate of Planning, Badalona Serveis Assistencials, Badalona, Barcelona, Spain; 2 Badalona Serveis Assistencials SA, Barcelona, Spain; 3Hospital Universitari Germans Trias i Pujol, Barcelona, Spain; 4Jordi Gol i Gurina Primary Health Care Research Institute. IDIAP, Barcelona, Spain; 5Son Llatzer Hospital, Palma de Mallorca, Spain; 6Health Sciences Institute of Aragon, Zaragoza, Spain OBJECTIVES: The objective of the study is to obtain behaviour of the cost’s relative average weights of the assistance with the retrospective application of the Adjusted Clinical Groups (ACG’s) in 16 teams of Primary Care with an attended population in the clinical practice use. METHODS: Multicentre, retrospective study based on electronic records of patients seeking care during 2008 in the regions of Aragon, Balears and Catalonia. Main measurements: universal variables (age, sex, health service-family practice/paediatrics) and dependent variables: episodes and total cost (visits, diagnostic test, referrals, drugs). The ACG case-mix System software (version 8.2; N = 106) classified subjects into a single category for a given annual resource consumption. The model of cost per each patient was established differencing the fix cost and the variable. Outlier patients were considered those surpassing T = Q3 + 1.5(Q3-Q1) = c1778.6 for total cost expenditure. Log transformation of the dependent variable was carried out to reduce skewness of the distribution and make it close to normal. Explanatory power was calculated by coefficients of determination (R2). Statistical software: SPSS, P < 0.05. RESULTS: The total number of the studied patients was 227,235 (intensity of use: 75.6%), with an average 4.5 ± 3.2 episodes. The age average was of 44.1 ± 23.7 years, 56.6% women (13.5% paediatrics). The distribution of costs was c148,657,137. The total unitary cost per patient/year c654.2 ± 851.7 (relative weights of reference). Patient’s case-mix: 57.2% of the study population was grouped into 10 ACG. The explanatory power of the ACG classification system was 36.3% (Ln: 41.2%), P < 0,001. a total of 6.2% of patients were considered Outliers (N = 14.066). CONCLUSIONS: The ACG are an acceptable system of classification of patients in situation of clinical practice use. Some ACG classification categories should be separeted due to the high outliers number. A332 PMC18 MULTI-NATIONAL CHART REVIEW STUDIES IN EUROPE: OPPORTUNITIES AND CHALLENGES Payne KA1, Wasiak R2, Stein D1, Chancellor J3 1 United BioSource Corporation, Dorval, QC, Canada; 2United BioSource Corporation, London, UK; 3Chancellor Health Economics Ltd, Beaconsfield, Buckinghamshire, UK OBJECTIVES: In support of European pricing and reimbursement submissions, country-specific health economic evidence is required. In the absence of patient-level databases, multi-national, retrospective chart review studies can provide data related to real-world patient characteristics, patterns of care, clinical outcomes and the natural history of disease. METHODS: A critical review and qualitative analysis of six recent multi-national chart review case studies designed to inform health economic questions was performed to characterize common research aims, objectives, and design parameters, and to delineate main methodological, operational and analytic challenges. RESULTS: Design challenges include defining an eligibility period sufficiently long to obtain enough cases while accessing patterns of care that are current, sample size determination and balancing the need for stable estimates of resource utilization with the need to contain study costs, and delineating sampling frame methodology to minimize chart selection bias. Detailed treatment and clinical outcome data must be weighed in the context of abstractor burden, and data collection methods should be considered separately from data interpretation. a common case report form for multinational studies should allow local area treatment variation while maintaining consistency in data collection across health care settings. Operationally, knowledge of country-specific ethics and privacy protection regulations for retrospective studies is paramount as requirements are diverse and impact timelines differently. Successful study execution is also dependent on the ability to rapidly identify and recruit clinical sites representative of a broad range of clinical practice that also can provide sufficient geographical coverage. CONCLUSIONS: Despite significant challenges, practical, efficient, and scientifically sound approaches to the design and conduct of multinational chart review studies in support of health economic analyses are possible. Main methodological and operational challenges can be identified across studies, thus can be anticipated and planned for. CONCEPTUAL PAPERS & RESEARCH ON METHODS – Modeling Methods PMC19 THE THRESHOLD PRICING MODEL: NOT JUST ANOTHER COSTEFFECTIVENESS MODEL Mladsi D1, Earnshaw S1, Akashi-Ronquest N1, Keith MS2 1 RTI Health Solutions, Research Triangle Park, NC, USA; 2Shire Pharmaceuticals, Wayne, PA, USA BACKGROUND: Threshold analysis as it is typically applied to cost-effectiveness models is an extension of sensitivity analysis in which the threshold analysis may be used to demonstrate the maximum price given different levels of health outcomes resulting in cost-effectiveness. In these instances, the ICER is the primary result of the model, and threshold analysis aids interpretation. As a result, the model is restricted to the intended target indication, population, and line of therapy, reflecting key development decisions that have already been made. In contrast, a threshold pricing model, which is developed early in a drug’s development, is not attempting to evaluate the drug’s cost-effectiveness. Instead, it is built to inform the development plan, pricing strategy, and go/no-go investment decisions. OBJECTIVES: To outline the differences in the underlying mathematical structure, inputs, and outputs of a threshold pricing model compared with a traditional cost-effectiveness model, and to demonstrate its application. METHODS: We present the algebraic manipulations required to convert a decision-analytic model into one used for threshold pricing analysis. Using a hypothetical new pharmaceutical possessing two versions of a product profile, we demonstrate application of the threshold pricing model by generating a table of potential value-based pricing estimates corresponding to the unique combinations of indication, subpopulation, line of therapy, and comparator. We provide graphical depictions of the lost value-based pricing opportunity (reflecting the lost opportunity for the new product to address a greater unmet need) of development strategies that are not valuedriven. CONCLUSIONS: a threshold pricing model is a powerful tool for helping to construct a value-driven development plan and a value-based pricing strategy. Constructed appropriately, threshold pricing models can be used to prioritize among possible indications, identify target subpopulations, select the appropriate line of therapy, and choose and clarify required performance against comparators. PMC20 THE OPTIMAL NUMBER OF MONTE CARLO SIMULATIONS TO BE PERFORMED IN PROBABILISTIC SENSITIVITY ANALYSIS: EMPIRICAL EVIDENCE FROM ECONOMIC MODELS CONSTRUCTED FOR SUBMISSION TO THE NATIONAL INSTITUTE FOR HEALTH AND CLINICAL EXCELLENCE Batty AJ1, Paulden M2 1 BresMed Health Solutions, Sheffield, UK; 2University of Toronto, Toronto, ON, Canada OBJECTIVES: Probabilistic Sensitivity Analysis (PSA) is a common technique to assess uncertainty in economic models, with the majority of economic model publications now containing some form of PSA. Historically the number of simulations performed has been set at arbitrary levels (e.g. 1,000 simulations), however the aim of this research is to rationalise the number of simulations performed, minimizing both 13th Euro Abstracts wasted computational time and the risk of incorrect conclusions being drawn. METHODS: The analysis investigates the number of simulations required in order for a Cost-Effectiveness Acceptability Curve (CEAC) to remain stable at the periphery. Secondary analyses focused on the number of simulations required to give reliable estimates of the mean values in these (non-linear) models. In the UK the National Institute for Health and Clinical Excellence (NICE) require manufacturers to submit PSA as part of the Single Technology Appraisal (STA) process. Models from different contract research agencies that have been constructed in Microsoft Excel, for submission to NICE, were then used to generate 50,000 simulations per model. This data was retrospectively analyzed to determine the number of simulations required such that a cost-effectiveness acceptability curve would remain stable at the periphery (5th and 95th percentiles). Secondary analyses focused on the number of simulations required to give reliable estimates of the mean values in these (non-linear) models. RESULTS: Preliminary analyses suggest that conventional numbers of simulations are sufficient to estimate the CEAC at low levels of precision at the 5% and 95% limits and generate the mean value. However this is not the case if high levels of precision are required. CONCLUSIONS: Research in to the optimum number of Monte Carlo Simulations allows analysts to ground the number performed in empirical data, and suggests that accuracy can be achieved without spurious precision, wasted computing time, or worse, unreliable/unstable conclusions. PMC21 USING GROWTH IN THE COST-EFFECTIVENESS THRESHOLD TO INFORM THE DIFFERENTIAL BETWEEN THE DISCOUNT RATES ON COSTS AND HEALTH EFFECTS: REASON TO BE CAUTIOUS? O’Mahony J Erasmus University Medical Center, Rotterdam, The Netherlands BACKGROUND: Differential discounting of the costs and effects of health care interventions has been extensively debated. Proponents argue that a rising valuation of health as incomes grow justifies discounting health effects at a lower rate than costs. However, despite that fact that the impact of differential discounting is highly sensitive to the difference between the two rates, little attention has been paid what the appropriate differential should be. The existing justification for the differential rests the income and health elasticities of utility and income and health growth rates, all of which are uncertain: estimates of the appropriate differential are speculative at best. The discount differentials recommended by cost-effectiveness analysis advisory bodies vary, being 1.5 percentage points in Belgium, 2.5 in the The Netherlands and 4.5 in the UK prior to 2004. What has not been widely recognised is that the differential should approximate the annual rate of growth in the cost-effectiveness threshold: this has been shown in a simple model published in the literature in which decision makers use a cost-effectiveness threshold in their reimbursement decision rule. ANALYSIS: This link to threshold growth provides a more immediate alternative by which empirically determine the appropriate differential compared to the current estimates. While cost-effectiveness thresholds are often not made explicit, there is no strong evidence that they have risen in recent years. Indeed, there are reasons to expect thresholds to remain constant or even decline, even if income growth is positive. CONCLUSION: This consideration of the plausible discounting differential complements current theoretical debate over differential discounting. Growth in cost-effectiveness thresholds provides an alternative basis for justifying the discount differential. Existing evidence and expectations of threshold growth leads to conservative estimates of the discounting differential of at or near zero. PMC22 REVENUE OPTIMIZATION MODEL TO OPTIMISE POSITION AND INDICATION OF NEW LAUNCHES Mukku S, McConkey D Double Helix Consulting Group, London, UK OBJECTIVES: To assess the impact on revenues from a new drug in different indications and at different positions within a treatment pathway. METHODS: The model was developed by price modeling experts at Double Helix Consulting using a logical flow developed over years. The model is tested by internal and external experts with different products. RESULTS: Drugs that reach to market very rarely are released in only a single indication, especially in chronic disease areas where there may be several different illnesses with a related etiology but different presentation. This is exemplified in the field of immunology and rheumatology, where several different drugs with a similar MOA are being used to treat many conditions that are pathophysiologically related. To explore the impact that multiple indications or use in different lines has or will have on the pricing of a new therapeutic agent, Double Helix Consulting generated a revenue optimization model that can be used to establish the most likely price point for a new drug given several different scenarios. This model will help in optimizing the best position and or indication for long term revenues from the drug. The model uses EPI data, prevalence, incidence, number of competitors, prices of comparators, line of treatment and other inputs. The outputs include NPV over a chosen period of time that can be sorted by line of treatment and indication. CONCLUSIONS: While it is desirable for a drug to be indicated in the largest patient pool possible, such actions can have serious negative consequences for the price at which payers are willing to pay for the treatment. a larger patient pool makes the burden on the health care provider significant if the drug is too expensive, and as a result can lead to significant price erosion or HTA rejection. 13th Euro Abstracts PMC23 PROPOSED METHODS FOR CONDUCTING SENSITIVITY ANALYSES ON THRESHOLD-DERIVED ESTIMATES OF VALUE-BASED PRICE AND PRODUCT PROFILES FOR EARLY-STAGE DRUGS Mladsi D1, Earnshaw S1, Akashi-Ronquest N1, Keith MS2 1 RTI Health Solutions, Research Triangle Park, NC, USA; 2Shire Pharmaceuticals, Wayne, PA, USA BACKGROUND: Established methods exist for evaluating the effects of uncertainty around the model structure and parameters on the results generated by traditional cost-effectiveness analyses (CEAs) and include one-way and probabilistic sensitivity analyses (SAs). In contrast to the primary outcome of a traditional CEA—the ICER— the primary outcomes of a threshold CEA conducted for a product early in development include (1) the value-based price opportunity given a hypothetical or target product profile and (2) the magnitude of effect required to justify a target price. Because the outputs of a threshold model pertain to a new drug or indication where little or no data have been collected, and because the outputs are multiple, representing the set of product attributes, including price, that will define drug value, there is a need to explore the sensitivity of the results to factors that go beyond uncertainty. In analyses that generate potential value-based price or product attribute levels, new methods and applications of SA are required. METHODS: We present example one-way and probabilistic SAs, highlighting problems in interpretation that arise when traditional sensitivity analyses are applied to threshold models. We propose alternative SA methods and analyses and present interpretations of results. a Pricing Contribution Diagram is presented as a means of characterizing the extent to which each product attribute (efficacy, safety, tolerability, quality of life, position in care pathway) influences the value-based price opportunity. Probabilistic SAs are presented to examine the relationship between price (value-based and target) and individual product attributes, and the influence of uncertainty in other model inputs. CONCLUSIONS: Traditional methods of conducting SA are insufficient when applied to the threshold application of CEA. Instead, SAs specific to threshold models supporting decisions regarding early stage development should be employed. PMC24 LAST OBSERVATION CARRIED FORWARD (LOCF) VS. MIXED-EFFECTS MODEL REPEATED MEASURES (MMRM): EMPIRICAL EVALUATION OF TWO APPROACHES TO ANALYZING LONGITUDINAL DATA WITH MISSING OBSERVATIONS Jo H1, Gemmen E2, Bharmal M2 1 Quintiles, Parsippany, NJ, USA; 2Quintiles, Rockville, MD, USA OBJECTIVES: To compare two statistical approaches for analyzing longitudinal data with missing observations: 1)imputation using Last Observation Carried Forward method (LOCF) and 2)Mixed-effects Model Repeated Measures method (MMRM) to analyze the change from baseline in health-related quality of life (HRQoL) by medication adherence level. METHODS: HRQoL via SF-12 Health Survey and medication adherence via a 5-level categorical response was measured monthly for one-year for 184 patients in a U.S. multiple sclerosis observational study. HRQoL was summarized in two continuous variables: Physical Component Score (PCS-12) and Mental Component Score (MCS-12). Categorically collected medication adherence was converted to numeric values and average compliance was calculated over a 1-year period then categorized into two groups: ≥90% (GT90) or <90% (LT90) compliant. For validity of compliance, patients who had completed at least 6 measurements during 1-year on compliance question were included. LOCF used the last available change from baseline to impute the missing values for early drop-out. MMRM is a likelihood-based approach which models all actual observations jointly, with no attempt at imputation for missing values. RESULTS: A total of 131 patients were included in this analysis. The 12-month change from baseline in PCS-12 comparing patients with GT90 compliance vs. LT90 compliance using MMRM was 0.86 (p = 0.277) and using LOCF was 1.30 (p = 0.339). For MCS-12, the improvement among patients with GT90 compliance over LT90 compliance using MMRM was 2.04, while the corresponding improvement using LOCF was 1.97. For MCS-12, only the MMRM method produced statistically significant improvements (p-values: LOCF = 0.234, MMRM = 0.026). CONCLUSIONS: MMRM and LOCF yielded not only different results but also different statistical significance in the 12-month change from baseline in MCS-12. Since the approach to estimate and model is different between two methods, the pattern and shape of data must be investigated to find the right method to produce valid estimates. PMC25 STATISTICAL DISTRIBUTIONS OF COST DATA IN PROBABILISTIC SENSITIVITY ANALYSIS Lacey L Lacey Solutions Ltd, Skerries, Ireland OBJECTIVES: It is generally agreed that calculation of means after non-linear data transformations (e.g., log-transformation) does not result in a comparison of arithmetic means, and so is not appropriate for cost data in pharmacoeconomic evaluations. This would seem to preclude the use of log-normal distributions for cost data in probabilistic sensitivity analysis. The study objective was to investigate the statistical properties of arithmetic mean costs. METHODS: Monte Carlo simulations were use to investigate the statistical properties of arithmetic mean costs derived from an underlying log-normal distribution, loge (X) ~ N(m,s2), where m = loge(c10), s = 1.5 (range 0.5 to 2.5). An underlying log-normal distribution was used because cost data A333 are typically highly positively skewed. Microsoft Excel was used to perform the Monte Carlo simulations generating 1,000 arithmetic means, each from a sample of N = 100, for each value of s investigated. RESULTS: The distribution of arithmetic means increased in positive skewness as s increased. For s ≥ 1.5, the distribution of arithmetic means deviated considerably from normality. The level of skewness was greatly reduced by use of the log-normal distribution. The Gamma distribution was similar to the log-normal distribution in representing the distribution of arithmetic mean costs. CONCLUSIONS: Log-normal distributions for arithmetic mean cost data may have a role for use in probabilistic sensitivity analysis, although this needs further investigation using cost data derived from actual studies. PMC26 EXAMINATION OF TYPE I AND II ERROR RATES IN INTENTION-TOTREAT RANDOMIZED EXPERIMENTS: DO SUBJECTS NEED TO STAY IN THE GROUP IN WHICH THEY WERE ASSIGNED? Wasser T, Eisenberg D HealthCore, Wilmington, DE, USA OBJECTIVES: Intention-To-Treat (ITT) analysis is an established method used in randomized experiments. However, analyzing data where crossover occurs (leaving subjects in the control or treatment arms when they have crossed from one group to another) prevents the true comparison of treatment and placebo effects. When subject’s crossover and ITT analysis methods are used, the true effect of the treatment cannot be determined as data from many groups are included with treatment. The purpose of this research is to determine Type I and II error rates computed by simulation results with and without crossover. METHODS: A simulation study was conducted to determine the impact of Type I and II error given six crossover percentages (1, 3, 5, 7, 9 and 11%), four effect sizes of treatment based on standard deviation (ES = 0.2, 0.4, 0.6 and 0.8 SD), and four sample sizes (n = 50, 100, 200 and 300). Simulations were conducted using “R” and included 1,000 replications for each sample size, effect size and crossover combination. RESULTS: When ES were small (<0.2SD), Type I error rates were below 1%. When ES were larger, and crossover increased Type I rates increased above 4%. Large samples with high crossover and large ES had the highest Type I rates. Type II error rates, which are perhaps more critical, were higher. For 5% crossover, the Type II error rates were 2.4% and for 11% crossover 5.5%. When the ES are very large statistical significance can be observed regardless of crossover percent, even up to 11%. CONCLUSIONS: When crossover rates are low and ES are small, researchers can abandon ITT analysis methods and analyze samples as they were treated with little risk of additional Type I and II errors occurring. The benefit of “as-treated” analysis is that the true treatment effect can be determined with little risk of error. CONCEPTUAL PAPERS & RESEARCH ON METHODS – Patient-Reported Outcomes Studies PMC27 VALUES FOR HEALTH STATES UNDER DIFFERENT LIFE DURATIONS 1 2 3 Scalone L , Milani S , Krabbe P 1 University of Milano—Bicocca, Monza (MI), Monza, Italy; 2University of Milan, Milano, Italy; 3 Radboud University Nijmegen Medical Centre, Nijmegen—The Netherlands OBJECTIVES: Recent research suggests that the value of health depends also on the time of permanence in a health state. This would imply a more complex relationship between quality and quantity of life than the standard linear relationship assumed in the QALY model. To model whether and how life duration affects the value assigned to health states. METHODS: A discrete choice analysis study was conducted comprising health-state scenarios added with a separate duration attribute. Health states were described with the EQ-5D (mobility, self-care, usual activities, pain/discomfort, anxiety/depression), having 3 levels of severity each. Duration was introduced as a sixth domain with six levels (1, 5, 10, 15, 30, 50 years). Sixty choice sets were selected with a Bayesian approach (Stolk et al., Value in Health, 2010). a sample of 209 undergraduate students self-completed the computerized response tasks. Data were analyzed with a conditional logistic regression model. RESULTS: Main effects show negative preferences towards problems with health domains and positive preference toward longer duration. However, preferences are not linearly related with duration (e.g., the next 41st year is valued less than the next 11th year), but a logarithm function describes more accurately this relationship. Negative interactions are found between health-states and duration. Trends of values for health states with increasing duration are represented by diverging curves: when duration increases, the value assigned to good states increases while the value assigned to bad states decreases. However, due to the logarithm relationship with duration, the marginal (negative or positive) value assigned to any state decreases as duration increases. CONCLUSIONS: Duration of health states interacts with the value assigned to them and influences both direction and marginal values through the different durations. Our results show that refinement of the standard QALY framework can be amended. A334 PMC28 CLINICIAN REPORTED OUTCOMES: ISSUES IN THE TRANSLATION AND LINGUISTIC VALIDATION Furtado T, Wild D Oxford Outcomes Ltd, Oxford, Oxon, UK OBJECTIVES: The use of Clinician Reported Outcomes (ClinROs) and cognitive functioning measures in multi-national clinical trials is widespread, but the translation of such measures has often been avoided, with trial managers preferring to keep the measures in English. However, this approach is not ideal and recently the importance of translating and linguistically validating ClinROs has been highlighted. The requirements for translating and linguistically validating ClinROs have not been well documented. Usually, they are translated with a PRO translation procedure (comprising dual forward and back translations and subsequent review) with the addition of a single clinician review. This study investigates the issues surrounding the translation and linguistic validation of these ClinROs and cognitive functioning measures. METHODS: A review was undertaken of translation reports from past projects, including the MMSE and SCID (cognitive functioning measures), TNSN (clinician or nurse guide for assessing neuropathy) 6 Minute Walk Test (6MWT) Worksheet, (guide for clinicians to assess dyspnoea), and EORTC clinician scales. Examples of issues from the translation process were gathered. RESULTS: Clinical review as part of the translation process proved invaluable. Changes typically comprised alterations of specific terminology.—a clinician should also be available to aid project coordinators throughout the translation process.—Clinical review formed an essential role for advising lead translators.—One project included dual clinician reviews; this was particularly valuable since changes could be verified between reviewers. CONCLUSIONS: The translation and linguistic validation of ClinROs and cognitive functioning measures requires further research to determine the optimum methodology. This study found that clinical review was of the utmost importance, and that ideally dual clinician reviewers should be involved. Furthermore, translators should have experience of translating documents intended for clinical use, to ensure they are familiar with the terminology. PMC29 USING DOZENS OF ATTRIBUTES WITHOUT INCREASING RESPONDENT BURDEN: HOW TO ADAPT LATENT VARIABLE MODELING FOR LINKING ATTRIBUTES ON SEPARATE CONJOINT SURVEYS Cole JC1, Dang J2 1 Covance Market Access Services, San Diego, CA, USA; 2University of California, Los Angeles, Torrance, CA, USA Conjoint analysis is a rigorous survey technique used to understand health care preferences in the pharmaceutical and medical device industries. In a traditional conjoint study, respondents are presented with a complete profile of all of the combination of attributes and features (or levels) for a particular product or service. However, research involving a large number of attributes can be too burdensome for respondents and has been shown to elicit inaccurate responses. This study describes a procedure used to link attributes from two or more different conjoint surveys that share at least one attribute. Linked latent conjoint modeling can lessen the burden on respondents while allowing utility parameters to be estimated for a large number of attributes, all on the same interval scale. Conjoint survey data were linked using a partial profile design and parameters were calculated using maximum likelihood estimation for finite mixture modeling. Several examples demonstrating the procedures used to link choice based survey data are provided. In addition, results from a latent variable modeling of the linked survey data are reviewed. Finally, to illustrate the flexibility of latent conjoint analysis, continuous and categorical covariates were simultaneously estimated to demonstrate the usefulness of latent modeling of conjoint data. PMC30 COGNITIVE DEBRIEFING METHODS IN TRANSLATION OF PROS: A MULTI-NATIONAL Furtado T, Gergovich KB, Wild D Oxford Outcomes Ltd, Oxford, Oxon, UK OBJECTIVES: Cognitive debriefing interviews are a key component of the translation and linguistic validation of PROs and a necessity when a PRO is to be used as a primary or secondary endpoint for an FDA label claim. The 2005 ISPOR principles of good practice report describes the objectives of debriefing, but little discussion has been undertaken into the methods for performing debriefing interviews as part of the translation and linguistic validation process. This abstract presents a pilot test of methodologies across countries. METHODS: A literature review was conducted on cognitive debriefing methodology, and different approaches were pilot tested in 10 countries. Three methodologies were selected for pilot testing: 1) Retrospective probing 2) Retrospective think aloud 3) Concurrent think aloud. RESULTS: The literature review highlighted the methodologies and pros and cons of different approaches to cognitive debriefing, but no research was identified that addressed the issues of particular relevance in the translation and linguistic validation process. All three processes listed above proved suitable for a methodology for linguistic validation. The Think Aloud technique provided a true sense of the respondent’s understanding of the translation but its suitability was particularly subject to cultural and individual differences. Questionnaires of a personal nature (e.g. those pertaining to sex or bowel disorders) benefited from Retrospective Probing, since the patients could respond hypothetically. The Retrospective Think Aloud technique elicited more information from some respondents, since it allowed them to discuss their personal experi- 13th Euro Abstracts ence. Where the respondent had difficulty in understanding what was required of them, modifying the interview style proved beneficial. CONCLUSIONS: A combination of Retrospective Probing and Retrospective Think Aloud proved to be the optimum methodology across countries, but this was dependent on the culture, patient, and the nature of the PRO being translated. PMC31 CONSTRUCTION OF PRIMARY HUI3 PERSON-MEAN UTILITY SCORING FUNCTION Furlong W1, Feeny D2, Torrance G3 1 Health Utilities Inc., Dundas, ON, Canada; 2Kaiser Permanente Center for Health Research, Portland, OR, USA; 3McMaster University, Toronto, ON, Canada OBJECTIVES: To assess construction of the primary HUI3 scoring system. METHODS: Mean visual analogue scale (VAS) and standard gamble (SG) scores were collected in 2 face-to-face interview surveys: modeling survey (MS) for the PersonMean utility scoring model to calculate community utilities for health states; direct survey (DS) for validation of the Person-Mean scoring model. Survey results are assessed for completeness and consistency. Completeness is evaluated by response rates and consistency by health state rankings. RESULTS: Completed interviews were obtained from 65% of contacted eligible subjects. Demographic distributions are similar to the underlying general population. There are 256 respondents in MS and 248 in DS. MS and DS respondents use 71% of HUI3 attribute levels in describing their own health status. Each of the 8 HUI3 attributes are reported by 35 or more respondents as being important in their preference measurements: pain (49%); vision (37%); cognition (34%); emotion (28%); ambulation (28%); hearing (17%); dexterity (17%); speech (7%). More than 70% of respondents focused on 2+ attributes. Other important preference measurement factors are self-care ability (89% of respondents), family life (76%), happiness of others (69%), ability to work current job (61%), leisure activities (42%). 83% of respondents report the interviewing did not change their opinions about the health states. Consistency of health state rankings by mean VAS and SG scores between MS and DS was 100%: PH > MA > MB > MC > Dead > Pits. MS had a missing data rate of 0.29% (17/5920) for VAS and 0.10% (1/1024) for SG. CONCLUSIONS: The Person-Mean HUI3 utility function is founded on a survey that was well-constructed in terms of community and attribute representation, consideration of multiple attributes and day-to-day impacts, stable opinions, and consistency of health state rankings. This evidence supports use of the primary HUI3 utility function for group-level analyses, such as allocation of societal resources. PMC32 MINIMALLY IMPORTANT DIFFERENCE OF THE TREATMENT SATISFACTION WITH MEDICINES QUESTIONNAIRE (SATMED-Q) Rejas J1, Ruiz MA2, Pardo A2, Soto J1 1 Pfizer España, Alcobendas/Madrid, Spain; 2Universidad Autonoma de Madrid, Madrid, Spain OBJECTIVES: Treatment satisfaction with drug therapies is an important patientreported-outcome (PRO) that may help clinicians to better impact in patient health care. The Treatment Satisfaction with Medicines (SATMED-Q) questionnaire has shown appropriate psychometric properties for exploring patient’s satisfaction with medicines under routine medical practice in chronic health conditions. The Minimally Important Difference (MID) of the instrument is still unknown. The goal of this research was to determine the MID values of the SATMED-Q questionnaire for the total score and domains. METHODS: The sample of patients (457, mean age 59 years, 53% male) used for testing psychometric properties was also used to assess MID values. Item #14 of the TSQM scale was used as an anchor reference, since it explores directly the satisfaction with medicines in a seven points ordinal response (from extremely satisfied to extremely dissatisfied). Patients were classified into four categories according with responses in this item; extremely satisfied/dissatisfied, very satisfied/dissatisfied, satisfied/dissatisfied, nor satisfied/nor dissatisfied and comparisons were carried out for the overall score and each domain of the SATMED-Q using standardized scores. The mean differences in overall score (and domains) between the neutral category and the satisfied/dissatisfied category were considered the values of MID. Effect sizes (ES) were also computed. RESULTS: MID for total scoring was 13.4 (ES = 0.91), while the value for domains ranged from 10.3 (medical care domain, ES = 0.43) to 20.6 (impact of daily living, ES = 0.85). Mean score differences in overall scale and domains were significant between change in satisfaction categories with respect item#14 with F values ranging from 9.7 to 74.1 (P < 0.001 in all cases). CONCLUSIONS: The SATMED-Q demonstrated to be responsive to different levels of patient’s satisfaction with therapy in chronically ill subjects. Attained MID was 13.4 pts for the overall normalized scoring scale and between 10.3 and 20.6 pts for domains. PMC33 ESTIMATING HEALTH STATE UTILITY VALUES FOR COMORBID HEALTH CONDITIONS Ara R, Brazier J University of Sheffield, Sheffield, UK OBJECTIVES: Health state utility values (HSUVs) for comorbid health conditions (CHC) are frequently estimated using data from single health conditions but there is no consensus on the most appropriate method. The objective of the study is to comparing five techniques in a single data set. METHODS: We use EQ-5D data from the Health Survey for England to compare results generated using the: additive, multiplicative and minimum methods, the adjusted decrement estimator (ADE), and a linear regression model; a baseline of perfect health and an adjusted baseline obtained from 13th Euro Abstracts individuals who indicate they have no chronic health condition. Results are compared using mean errors (ME), root mean squared errors (RMSE) and the proportion of values estimated within |0.05|. RESULTS: Using an age adjusted baseline, we found the additive (and multiplicative) methods underestimate the majority of HSUVs (ME:0.0781(0.0254); RMSE:0.1012(0.0651); 26%(56%) < |0.05|) while the minimum (and ADE) overestimate the majority of HSUVs (ME:−0.0995(−0.0695); RMSE:0.1214(0.0950); 20%(35%) < |0.05|). Although the simple linear model produced the most accurate results (ME:0.0001; RMSE:0.0598; 63% < |0.05|), there were some substantial errors with 20% of errors greater than the minimum important difference (|0.074|). When subgrouping by actual HSUV (range 0.350–0.917) we found the magnitude and direction of errors in the estimated HSUVs are driven by the actual HSUVs being estimated in addition to the technique used. In general the HSUVs estimated using an adjusted baseline were more accurate than those obtained using a baseline of perfect health. CONCLUSIONS: This study makes an important contribution to the evidence in this area as it is the first to compare the five different techniques in the same data set. While the simple linear model gave the most accurate results, the model requires validating in external data and additional research exploring an alternative model specification is warranted. PMC34 ISSUES IN THE TRANSLATION AND LINGUISTIC VALIDATION OF CAREGIVER RATING SCALES REGARDING THE BEHAVIOR AND DEVELOPMENT OF CHILDREN AND YOUNG PEOPLE Furtado T, Wild D Oxford Outcomes Ltd, Oxford, Oxon, UK OBJECTIVES: Caregiver rating scales, intended to evaluate the behaviour of children and young people, are frequently used in clinical trials involving youths. However, the translation and linguistic validation of such scales can be problematic due to the differing cultural markers of behaviour and development. This study aims to document the problems that can occur, with the hope of facilitating future studies and producing guidelines to avoid cultural compromises when such measures are developed. METHODS: Past Oxford Outcomes projects, which included the translation of Caregiver Reported Outcomes, were evaluated to identify problematic items. These included the Vineland-II and ABAS (behaviour development scales), ELDQOL (epilepsy and QoL scale) and WFI-RS (functional impairment rating scale) among others. RESULTS: Numerous cultural and linguistic issues became apparent, including the following:— Many examples of sports and activities were used in the documents, which required thorough cultural adaptation, e.g. types of games.—Logistical cultural differences were marked, e.g. questionnaires mentioned children’s understanding of specific coins or traffic signals, which vary culturally.—Some documents involved markers for identifying speech development, such correct use of irregular verbs. These were problematic in other cultures and speech development specialists were required to find suitable alternatives.— More idiomatic expressions are used than in PROs developed for adults, e.g. “on the go”; these cause difficulties in translation.—Items surrounding activities such as housework or helping look after siblings are not equivalent in some cultures due to differing role expectations. CONCLUSIONS: The validation of caregiver reported outcomes through interviews with caregivers was particularly important with these scales to ensure cultural appropriateness in target languages. Physician or specialist input was sometimes required to find culturally relevant alternatives. When such measures are created, culturally specific markers of behaviour should be avoided if possible. PMC35 RECOMMENDATIONS ABOUT TRANSLATIONS IN THE FINAL FDA GUIDANCE ON PRO MEASURES: WHAT HAS CHANGED AND WHAT HAS REMAINED Conway K1, Mear I2 1 MAPI Research Trust, Lyon, France; 2MAPI Institute, Lyon, France OBJECTIVES: Almost four years were necessary to develop the final FDA guidance on the use of PRO measures in clinical trials. Our objective is to compare how the recommendations about translation and cultural adaptation evolved from the 2006 draft to the 2009 final guidance. METHODS: Both guidances were retrieved on the FDA website and analyzed. RESULTS: Structure and content were modified. Recommendations on translation and cultural adaptation were moved to another section within the Evaluating PRO Instruments Part: from “IV.D. Modification of an existing instrument” to “III.G. PRO Instruments intended for specific populations”. As for the content, the text in the body of the final guidance is more concise compared to the draft. The novelty lies in the stipulation that the FDA will review the process used to translate/culturally adapt the instruments. As a consequence, an appendix (section VIII) was added in which the FDA explains which topics should be addressed in the documents provided to the FDA for review: description of process used, patient testing, rationale for decisions, copies of versions and evidence about validity. They are however key points which did not change: the need for providing evidence that content validity and other measurement properties are similar between all versions. CONCLUSIONS: The recommendations are more concise and precise, especially the expectations of the FDA. The FDA however does not indicate a preference for a specific translation methodology. Interestingly patient testing is clearly indicated as a key point of the process. The need for documenting all decisions is crucial and raises the question of developing standardized system of reporting to structure the evidence to be provided to the FDA. The last point of the Appendix is debatable as we anticipate that it might add a burden in term of costs to provide evidence about the psychometrics of all versions. A335 PMC36 SYSTEMATIC REVIEW OF THE RESPONSIVNESS OF SF-36 HEALTH SURVEY MEASURES TO EFFICACIOUS PHARMACEUTICAL THERAPIES IN WELL-CONTROLLED CLINICAL TRIALS Ware JE, Frendl DM University of Massachusetts, Worcester, MA, USA OBJECTIVES: To determine how often SF-36 Health Survey measures respond to efficacious pharmaceutical treatment benefits in well-controlled clinical trials. METHODS: We conducted a systematic review of randomized, double-blind, placebocontrolled trials published in 124 journals in 1995 through 2009 documenting differences between treatment groups for primary medical endpoints and any of the SF-36 component summaries, or eight subscale scores. Concordance was defined in terms of agreement between primary clinical and SF-36 endpoints (both statistically significant or both non-significant). RESULTS: A review of 2,020 identified clinical trials using the SF-36 confirmed that 162 met study design criteria. For 133 of 162 trials (82.1%), results for primary clinical endpoints and SF-36 measures were concordant. Among the 107 trials achieving medical efficacy (primary endpoint), changes in one or more SF-36 measures were also significant, as hypothesized, for 88 (82.2%). Similar patterns were observed by therapeutic area; for example: rheumatology (29 of 30), neurology (16 of 25), cardiovascular (15 of 18), pulmonary (11 of 13), psychiatry (8 of 10), endocrine (7 of 9), and combined surgical specialties (9 of 9) studies demonstrated concordance. In addition to evaluating characteristics of published reports and scoring methods (subscales, summaries, utility scoring) this presentation will comment on priorities for future studies of patient-reported outcomes (PROs) in evaluations of pharmaceutical and other medical treatments. CONCLUSIONS: In support of their validity as PROs, changes in SF-36 measurements agree with primary endpoints in over 8 out of 10 well-controlled trials of pharmaceutical therapies published to date. In support of pharmaceuticals’ efficacy, when a therapy positively impacted clinical endpoints, it also improved health related quality of life quality of life in over 8 out of 10 clinical trials published to date. PMC37 DOES DATA COLLECTION FROM ONLINE COMMUNITIES RESULT IN BIASED RESPONSE? Vaccarino AL1, Sills TL1, Bharmal M2, Cascade E3, Kalali AH4, Evans KR1 1 OCBN, Toronto, ON, Canada; 2Quintiles, Rockville, MD, USA; 3iGUARD Inc, Rockville, MD, USA; 4Quintiles CNS Therapeutics, San Diego, CA, USA OBJECTIVES: Although the ability to interact with patients in an on-line environment has expanded substantially over the past few years, many researchers are concerned that participants may not be representative from a medication experience perspective (i.e., biased towards complainers). The purpose of this study is to investigate patient responses on treatment satisfaction using a validated PRO measure, the Treatment Satisfaction Questionnaire for Medications (TSQM), collected through a survey of patients with depression from an on-line community. METHODS: A random sample of iGuard.org members treated with an antidepressant were invited to complete an online version of the TSQM, a widely used validated 14-item generic treatment satisfaction instrument. iGuard.org is an online patient community that provides a free medication monitoring service to patients. Non-parametric item response analyses were performed to determine the relationship between scores on individual items and total TSQM scores. RESULTS: Responses from 3641 patients were included in the analyses. TSQM Global Satisfaction scores ranged from 0—100 suggesting a broad spectrum of treatment satisfaction. Non-parametric Item Response analyses of raw scores revealed that individual items of the TSQM discriminated differences in patient satisfaction. That is, as total scores increased the probability of low scores on the individual items decreased and the probability of higher scores increased. As expected, patient satisfaction was related to reported side-effects, with those reporting sideeffects experiencing lower satisfaction with medication than those without reported side-effects. CONCLUSIONS: The results from this analysis suggest that PRO survey data collected through a random sample of members of the on-line patient community iGuard.org can be representative of the spectrum of anticipated treatment satisfaction responses. Continuing to explore the potential of direct data capture from on-line patients will be important as researchers seek faster and cheaper alternatives to traditional physician-based recruitment. PMC38 A COGNITIVE DEBRIEFING METHODOLOGY FOR ESTABLISHING EQUIVALENCE DURING E-PRO MIGRATION Doyle S, Wild D Oxford Outcomes Ltd, Oxford, Oxon, UK BACKGROUND: Most outcomes instruments have been developed and validated as paper versions, but few have been migrated to electronic format. Migration to electronic delivery, without significantly altering format or text, qualifies as a minor modification not requiring a full validation (Coons et al. 2009). However, this does not mean that the two formats are perceived in the same way by patients. We aim here to describe a methodology successfully used to establish equivalence between paper and electronic PROs. METHODS: To demonstrate the equality of these different modes of data collection, we have used a combination of “think-aloud” and retrospective cognitive debriefing techniques, as well as usability testing. The debriefing exercise is designed to assess whether the electronic device changes the way respondents interpret the questions or response options. The usability testing assesses ease of use and identifies issues that may prohibit the use of the ePRO by the target A336 population. We typically recruit between 10–20 patients in which half the participants receive the ePRO first and the other half the paper version. Between administrations participants complete a distraction task. Interviews are recorded and a content analysis conducted to identify key issues. RESULTS: The mix of think-aloud and retrospective probing has worked well in a number of studies across disease areas to ensure equivalence, high usability, and no unforeseen issues unique to ePRO such as screen glare or difficulty holding a PDA device. Some patients have difficulty with the “thinkaloud” approach and so the retrospective probing is a useful check against issues not spontaneously raised by the participant(s). CONCLUSIONS: Increased use of ePRO questionnaires necessitates a robust methodology for demonstrating equivalence during migration from paper versions. a mix of concurrent “think-aloud” and retrospective probing following completion of both PRO formats has shown to be a useful method for establishing validity of electronic outcome measures. PMC39 EXAMINING ITEM RESPONSE PATTERNS OVER TIME IN A HEALTH PROFILE MEASURE USING US NATIONAL REPRESENTATIVE SAMPLES: A MULTI-FACET MODEL APPROACH Gu NY Pharmerit North America, LLC, Bethesda, MD, USA OBJECTIVES: To examine item response patterns over time using the SF-12v2TM from a measurement perspective using US national representative samples. METHODS: Four panel data with two-year repeated measures on each respondent were extracted from the Medical Expenditure Panel Survey (MEPS). Respondents were included if they were ≥18 years, had completed SF-12v2TM and, had at least one of the top ten most prevalent health conditions identified using ICD-9-CM. Three-facet measurement model was used to parameterize time as a distinct facet in the model, in addition to person and item facets. Interactions between time and the twelve items were examined at each time point in all panels. Goodness-of-fit of the items to the model was examined in repeated measures as well as in point-in-time measures. INFIT mean-square (MnSq ≤ 1.40) was used as an item fit indicator. Cross-validations were conducted in each disease groups. RESULTS: Four panels were comparable in their distributions in health conditions, socio-demographics (mean ages were 52–53 years and, about 76–77% were white) and, sample sizes (2003–04, n = 2,124; 2004–05, n = 2,070; 2005–06, n = 2,148 and 2006–07, n = 2,329). Consistently in all panels, significant time and item interaction biases were found at time 1, especially on mental health items (P < 0.01). On the other hand, interaction biases between time and items at time 2 were not significant (p > 0.05). All items fit the model in repeated measures where time was parameterized as a facet (INFIT MnSq ≤ 1.40). The mental health item “Have you felt calm and peaceful?” consistently showed misfit in all point-in-time measures (INFIT MnSq > 1.40). Similar findings were noted in sub-samples. CONCLUSIONS: Findings from this study suggest consistent learned response patterns over time, especially the responses to mental health item, which give rise to the importance of inter-temporal health context in health measurement. Hence, cross-sectional health measures should be interpreted with caution. PMC40 ITEM CALIBRATION OF A GENERIC ROLE FUNCTIONING ITEM BANK 1 2 Anatchkova M , Bjorner J 1 University of Massachusetts Medical School, Worcester, MA, USA; 2National Research Centre for the Working Environment, Copenhagen, Denmark OBJECTIVES: Role functioning (RF) is a key component of social well-being and thus an important outcome in health research. The aim of this study was to calibrate on a common metric newly developed items assessing the impact of health on RF. The items were developed based on review of the literature and focus group interviews and were found to be sufficiently unidimensional for item response theory applications. METHODS: Two thousand five hundred participants completed a battery of measures including 77 items in a RF bank, covering the impact of health on family, occupational and social role functioning. Each new item covered only one of the content areas. Items were evaluated for potential DIF by demographic variables (gender, age, and chronic condition) using a logistic regression approach. To estimate the item parameters for each domain on a common metric we used the generalized partial credit model. Item fit was evaluated using the S-G2 index. Comparison of group mean bank scores of participants with different self-reported general health status and chronic conditions was used to test the external validity of the bank. RESULTS: After excluding items with DIF and poor fit the final item bank had a total of 64 items covering 4 general content areas of role functioning (family, social, occupational, generic). Slopes in the bank ranged between 0.96 and 4.51; the mean threshold range was −0.66 to −1.80. Item bank based scores were significantly different for participants with and without chronic conditions (F(4, 2488) = 31.48, P < 0.0001) and self-reported general health (F(4, 2488) = 233.55, P < 0.0001). CONCLUSIONS: An item bank assessing health impact on RF across 4 content areas has been successfully calibrated. Using computerized adaptive assessment, respondents will only need to answer items regarding relevant roles, while IRT score estimation still allows for scoring all respondents on the same common metric. 13th Euro Abstracts PMC41 PREEMPTING DIFFICULTIES IN LINGUISTIC VALIDATION, THE USE OF FACE VALIDATION TO CREATE MORE SOUND TRANSLATIONS Gawlicki M1, Handa M2 1 Corporate Translations, Inc, East Hartford, CT, USA; 2Corporate Translations, Inc, Chicago, IL, USA OBJECTIVES: The process of linguistic validation is complex especially when working with a variety of languages in widely divergent cultural settings. The ability to clearly delineate concepts and synchronize wording within an instrument before the linguistic validation process begins not only significantly improves the original instrument, but also aids in optimizing its translatability, ensuring greater uniformity between multiple linguistic adaptations and saving time and resources along the way. This paper seeks to explain the benefits provided by the supplemental pre-translation process of face validation. METHODS: As part of a case study, face validated questionnaires were compared to the original homegrown versions of the corresponding instruments— questionnaires that were already psychometrically validated were not eligible. Changes that were made as a result of this analysis will be discussed in-depth to clarify difficulties that each issue would have created for the linguistic validation process had they not been corrected. a cost benefit-analysis was also conducted to confirm the value of this supplemental linguistic validation phase. RESULTS: While standard elements of the linguistic validation process, such as concept elaboration, international harmonization, survey research expert review, in-country clinician review and cognitive debriefing all assist greatly in creating a quality translation, none of their benefits are a substitute for face validation. Furthermore, cost-benefit analysis reveals that the preemption of linguistic or methodological issues prior to translation and the greater uniformity obtained amongst multiple translations created through face validation save time and money later on in the linguistic validation process, justifying the added up-front costs. CONCLUSIONS: As the case studies confirm, taking steps to maximize the translatability of a questionnaire prior to linguistic validation, through face validation in particular, is highly beneficial to the end-products and can also hasten overall project completion and improve the quality of all language versions of the instrument. PMC42 TO WHAT EXTENT CAN TECHNOLOGY IMPROVE THE VALIDITY OF CLINROS? Wild D1, Langel K2 1 Oxford Outcomes Ltd, Oxford, Oxon, UK; 2CRF Health, Helsinki, Finland OBJECTIVES: ClinROs are the most commonly observed endpoint in FDA approved product labels but few have been adequately scrutinized in terms of their suitability as endpoints. This study evaluates two widely used ClinROs (the Expanded Disability Status scale (EDSS), and the Hamilton Rating scale for Depression (HAM-D)) and provides an assessment on how migrating the measures onto an electronic platform might be able to improve their validity and reliability. METHODS: A literature review was conducted on both measures to evaluate the availability of information on their content validity and reliability and validity. An assessment was made on how the measures could be improved if they were to be migrated onto an electronic platform: RESULTS: The EDSS has shown varying results for validity and inter-rater reliability and it involves a complex scoring procedure. The migration of the EDSS onto an electronic format would enable an automated scoring system which could improve its validity. The HAM-D was found to be lacking in evidence of content validity and to have some complexity in the scoring system. Transferring the HAM-D onto an electronic platform could simplify the scoring system which could improve its validity. CONCLUSIONS: This study has highlighted some of the issues with validity and reliability of two widely used ClinROs. The migration of ClinROs to an electronic platform in addition to the ePRO migration cognitive debriefing and usability testing might go some way to improving the clarity of ClinROs which may go some way to improving the validity of the measures. It cannot however resolve all of the issues such as lack of content validity and its impact would vary widely according to the complexity of the ClinRO itself. PMC43 DATA POOLING OF PATIENT-REPORTED OUTCOMES IN CLINICAL TRIALS: EVALUATION OF STATISTICAL TECHNIQUES FOR ASSESSING MEASUREMENT EQUIVALENCE Nixon M Quintiles, Bracknell, Berkshire, UK OBJECTIVES: This analysis describes the development, application and comparison of three different approaches to evaluate measurement equivalence properties of a patient reported outcome (PRO) questionnaire applied to two treatment groups for gastroesophageal reflux disease (GERD). The data used in this analysis was obtained from an on-line patient community, iGuard.org. Patients using either of the two treatments were randomly invited to complete a measure of treatment satisfaction, the Treatment Satisfaction Questionnaire for Medication (TSQM). METHODS: Three statistical approaches were used to evaluate the measurement equivalence of the TSQM across the two patient populations: 1) Classical Test Theory (CTT) to assess the internal consistency of the TSQM items within each of the three factors using Cronbach’s alpha; 2) Confirmatory Factor Analysis (CFA) using a special case of structural equation modelling (SEM); and 3) Item Response Theory (IRT)—based technique of Differential Item Functioning (DIF). RESULTS: All three statistical methods indicated measurement equivalence had been achieved across the two treatment populations for all the three domains of the TSQM. The effectiveness and global 13th Euro Abstracts satisfaction domains exhibited the strongest significant results amongst all three tests. However, while the convenience domain exhibited strongly significant measurement equivalence for the CTT, it only exhibited significant results for the SEM and DIF. CONCLUSIONS: While all three methods indicated the same overall results, there is some suggestion of differing sensitivity amongst the tests. PMC44 EXPERTS’ JUDGEMENT ON PATIENT-CENTRED COORDINATED CARE Mühlbacher AC, Juhnke C, Bethge S Hochschule Neubrandenburg, Neubrandenburg, Germany OBJECTIVES: Delivering care coordination services is often described as the key to effectively meet patients’ needs and expectations. Patient empowerment and patient participation is highly discussed and postulated, but there is a lack of knowledge of how to design patient-centered coordinated care. This study intends to provide health policy and decision-makers with a comprehensive assessment on experts’ priorities in the relative value of different dimensions of coordinated care. METHODS: A questionnaire with 88 items was conducted with N = 251 health care experts. Exploratory and confirmatory factor analysis was performed using SPSS©18. The number of factors to be retained was controlled by Kaiser’s criterion (eigenvalues above 1), validation of the scree plot, and the interpretability of the items. Cronbach’s alpha was used to assess the internal consistency of the subscales identified. RESULTS: The exploratory factor analysis leaded to 25 factors. After analyzing the screeplots and qualitative results confirmatory factor analysis was computed for an 8 factor solution accounting for 42,828 % of the total variance and with KMO of 0.723. Cronbach alpha reliability coefficients were computed for each of the sub-scales and ranged between 0.849 and 0.745. Based on the existing literature and the analysis conducted, coordinated care could be differentiated into eight dimensions: access, knowledge transfer, technical care, interpersonal care, patient-centeredness, continuity, infrastructure and participation in social life. CONCLUSIONS: The aim of the study was to structure the key attributes for future stated preference research. Differences in experts’ judgment and patients’ perspective will be analyzed in upcoming research. If expectations of stakeholders are taken into account adequately, it can be assumed that this will increase the motivation to participate in and the satisfaction with coordinated care programs. PMC45 GLOBAL INDUSTRY USE OF ELECTRONIC PATIENT-REPORTED OUTCOME INSTRUMENTS: PRELIMINARY RESULTS FROM A 2010 EPRO SURVEY Bergstrom F1, McGinley D2, Ackerman S3, Cole J3 1 Covance Market Access Services, Gaithersburg, MD, USA; 2Covance, Conshohocken, PA, USA; 3Covance Market Access Services, San Diego, CA, USA OBJECTIVES: While eClinical Forum’s 2009 survey findings suggest that electronic data capture (EDC) is used in 58% of clinical trials, little is known about the use of electronic patient reported outcome (ePRO) technologies for data collection. The purpose of this survey study was to describe the experiences and perceptions regarding use of ePRO as reported by pharmaceutical, biotech, medical device, and other industry professionals. METHODS: Global industry professionals were invited to complete a web-based survey fielded in early 2010. Participants were asked about their professional demographics, PRO and ePRO experiences, as well as challenges and advantages of using ePROs. Responses were analyzed descriptively. RESULTS: To date, 153 industry professionals completed the survey. Forty-four percent of respondents were from pharmaceutical companies, followed by other (41%), biotech (10%) and medical device (6%). Forty nine percent had previous PRO study experience among which 51% had prior ePRO experience. Among respondents using a PRO measure in an international study, 43% used ePRO for data collection. Hand-held device (tablet, PDA) was the most common ePRO technology (42%), followed by interactive webresponse (29%) or voice-response (29%). Reported advantages of ePROs include accuracy of information collected (79%), increased compliance (73%), and ease of use (64%); challenges include patient training (65%), study start-up costs (64%) or time (54%), and patient burden (54%). Validation of PRO for EDC use was an important factor when considering paper-based versus ePRO data collection (21%). Among those responding, 26% indicated they used ePRO data collection in >50% of their clinical trials, and 82% strongly agreed/agreed they would use ePRO in future studies. CONCLUSIONS: Preliminary results from this survey suggest that among those who use PRO measures in studies, the percent of industry professionals using ePROs is similar to the overall percent of industry using EDC as a data collection method in clinical trials. PMC46 TRANSLATION AND LINGUISTIC VALIDATION OF PRO MEASURES: RESPONSE OPTION ISSUES Griffin AJ, Furtado T, Wild D Oxford Outcomes Ltd, Oxford, Oxon, UK OBJECTIVES: PRO measures use a variety of response scales/options. These vary according to the type of measure, and can include frequency (time-based) scales, severity (intensity) scales, visual analogue scales (VAS) and levels of agreement. The translation and linguistic validation of response options can cause semantic or conceptual difficulties. This research aims to identify the issues raised during the translation of some response scales, with the aim of aiding the translatability of response options. METHODS: Examples of issues in the translation and linguistic validation of response options were collected from past Oxford Outcomes projects. Those A337 response options which were problematic across PRO measures and languages were evaluated. RESULTS: Numerous cultural and linguistic issues became apparent throughout the translation process which require careful attention being paid to the response options during the translation and linguistic validation process: /Some midscale words used in severity scales are particularly difficult to translate, e.g. rather, somewhat. Two response options on a scale, e.g. rather confident, fairly confident can be very small and unclear. /Frequency scales (how often . . .) are often translated as “how many times,” in some languages (particularly Indian). When using “level of agreement” scales the word “strongly” often proves problematic as some languages find it difficult to express levels of agreement. CONCLUSIONS: Various issues with response options were recognised during the linguistic validation of a considerable number of PRO measures. a full translation and linguistic validation procedure can help to overcome such problems, but care should be taken when choosing response scales during the development of PRO measures. In general, response options, which are particularly close in meaning, e.g. somewhat, rather, are usually more problematic to translate than those with clear parameters, e.g. never, rarely, sometimes, often, always. PMC47 FURTHER INSIGHT INTO DESIGNING WEB-BASED PATIENT REPORTED OUTCOMES FOR USE ON PERSONAL COMPUTERS IN GLOBAL STUDIES Tulkki-Wilke RK1, Bushnell D2, Martin ML3, Jewett G4 1 CRF Health, Helsinki, Finland; 2Health Research Associates, Inc, Mountlake Terrace, WA, USA; 3Health Research Associates, Inc, Seattle, WA, USA; 4CRF Health, Lansdale, PA, USA OBJECTIVES: Capturing patient-reported outcomes (PRO) via the web can be an efficient tool in larger-population clinical studies. Enabling patients to use their own computers means that web-based PROs are administered to patients on a wider variety of screen sizes and resolutions. As indicated in the recent ISPOR ePRO Report by Coons SJ, et al, this creates an issue of ensuring equivalency of the instruments across all screen sizes. This research describes a practical approach for maintaining validity of instruments when patients use their own computer. It aims to evaluate whether the PRO is presented as intended on various screen sizes and browsers. The research also takes into consideration variability of computer infrastructure in different geographical areas, as this is a major factor limiting web-based data collection. METHODS: The EuroQoL EQ-5D was programmed for use on smaller screens and scaled up to larger sizes and put into an ongoing usability testing study in the US, UK, Spain, Finland, Singapore, and China for sufficient coverage of languages and technologies. The sample size is 30 healthy volunteers. Screen sizes varied from small mini laptop screens to large LCD screens (9″ to 24″). RESULTS: Early results show that the questionnaire fits on all screens without a need for patients to scroll either in left-right or up-down directions. On 15,4″ screen, the questionnaire occupies 71% of the screen versus 57% on a 24″ display. The difference in relative size is 19%, which indicates that the questionnaire remains usable even if the screen size increases by 56%. Further results are forthcoming. CONCLUSIONS: As the presentation of the PRO is the same on all screens, conducting psychometric validation may be more straightforward. In addition, finding a method to ensure one PRO design works on all computers is a major factor in conducting global studies efficiently. PMC48 UPDATES AND INNOVATIONS IN ELECTRONIC PATIENT-REPORTED DATA CAPTURE: A REVIEW OF THE EVOLUTION AND FUTURE DIRECTIONS OF EPRO Curry C PHT Corporation, Boston, MA, USA OBJECTIVES: In clinical trials, the capture of patient-reported outcome (PRO) data has increased over the past decade. Regulatory guidelines, including the EMEA’s concern of Health-Related Quality of Life (HRQL) measurements and the FDA’s emphasis on PRO data to support labeling claims, have led to a greater inclusion of PRO as clinical trial endpoints. As sponsors include more PRO measurements, many turn to electronic PRO (ePRO) data capture. This presentation summarizes literature from the last 9 years to report changes in the use of ePRO including projections for future applications in clinical trials, disease management and health policy. METHODS: This presentation is a synthesis of literature in peer reviewed journals regarding ePRO from 2000 to 2010. Key search terms include “patient reported outcomes”, “electronic patient reported outcomes”, “electronic diaries”, “interactive voice response system”, “interactive web response system” and “digital pen”. The literature review made use of meta-databases such as PubMed and Medline. RESULTS: The use of ePRO has increased since a decade ago, due to greater movement in health care towards electronic solutions and to regulatory emphasis on PRO collection in clinical trials. The presentation will detail how ePRO is being used and provide a synthesis of recommendations for future ePRO use based on the literature. The session will examine the indications and therapeutic areas, population types, and applications of ePRO within health care and will review evidence showing whether ePRO data quality is higher than that of data captured using paper methods. CONCLUSIONS: The presentation deals with the evolving questions of ePRO—projected limitations and actual scope of experience with ePRO. Due to literature emphasis, the presentation will focus on clinical trials but will also examine other health care fields such as disease management and health policy. Attendees will learn about the evolution of ePRO and forthcoming directions and receive a bibliography of current literature. A338 PMC49 THE ROLE OF PATIENT-REPORTED OUTCOMES (PRO) AS PRIMARY ENDPOINTS IN THE EVALUATION OF MEDICINES APPROVED WITH PRO CLAIMS Caron M, Emery MP MAPI Research Trust, Lyon, France OBJECTIVES: To review the drugs that used PRO as primary endpoints in the evaluation of their clinical efficacy and have been granted a PRO claim by the Food & Drug Administration (FDA) or the European Medicines Agency (EMA). METHODS: The PROLabels database, which centralizes medicinal products with a PRO claim, was searched using the keyword “primary endpoint”. Only drugs approved between January 1, 1995 and May 28, 2010 were included. RESULTS: A total of 303 (66%) out of the 459 products included in the database were retrieved: four were removed because of a withdrawal notice (all EMA), 15 because they were approved before 1995, and two others because the endpoints were not clearly specified. In total, 282 products were analyzed (71 approved in Europe). They represented 81 different indications, with 47 products having more than one indication. The most frequent indications were: Pain (25 products), Rheumatoid Arthritis (25), Menopause (18), Parkinson’s disease (17), Epilepsy (15), Migraine disorders (13), Sleep disorders (12), Rhinitis Allergic Perennial, Seasonal (12), and Ankylosing Spondylitis (10). In total, 140 different PRO endpoints were listed. Seventy-seven percent of them were symptoms (e.g. pain, bowel movement, heartburn symptoms, asthma symptoms, etc). Function (e.g. physical function, functional impairment, etc) represented only 5.7% of all endpoints. Health-related quality of life (HRQL) was a clear primary endpoint in only two products: Duloxetine (Urinary Incontinence) and Alisade (Rhinitis Allergic Perennial and Seasonal). Both approvals were granted by the EMA in 2008 (after the first publication of the EMA and FDA guidances). CONCLUSIONS: Symptoms, as measured by patients, are a key criterion in the evaluation of medicines. Unsurprisingly, pain is the main indication in which the highest number of products with a PRO as a primary endpoint are approved. More sophisticated PRO endpoints, such as function or HRQL, are used less often. PMC50 TRANSLATION, LINGUISTIC VALIDATION AND ELECTRONIC ADMINISTRATION IN THE INDIAN LANGUAGES: A CASE STUDY HIGHLIGHTING THE SPECIAL DIFFICULTIES THAT EXIST AT THIS INTERSECTION Handa M1, Brandt B2, Hasegawa N1, McKown S1, Gawlicki M2 1 Corporate Translations, Inc, Chicago, IL, USA; 2Corporate Translations, Inc, East Hartford, CT, USA OBJECTIVES: Translation and linguistic validation of questionnaires for paper or electronic administration require similar procedures. Both modalities share a common difficulty: how to best utilize existing processes to accommodate Indian languages. Working with Indian language instruments can be daunting, even without the compounded effect of managing both ePRO and linguistic validation together. This paper seeks to enumerate the challenges associated with the intersection of Indian languages, linguistic validation, and electronic administration and recommends the use of a specialized checklist to avoid escalating potential difficulties. METHODS: A case study was performed to explore the complexity of these processes. a Fibromyalgia e-diary and corresponding paper diary were translated into four Indian languages and three non-Indian languages with varying levels of linguistic difficulty. This allowed for a multidimensional analysis of issues faced in translation for ePRO versus paper administration, and translation for Indian versus non-Indian settings. RESULTS: During linguistic validation, difficulties arise pertaining to font and formatting requirements, localization of colloquial terms and phrases and maintaining concept equivalency while respecting cultural appropriateness. Preparing an ePRO for use involves thorough analysis of existing instrument versions and possible modification of instructions, response sets or questionnaire format to accommodate technology or screen size limitations. a holistic approach of quality assurance should be employed, including multiple rounds of proofreading. When preparing an Indian language for electronic administration one must take even more care, accounting for the factors listed above as well as considering software font-testing prior to production, and additional proofreading steps. CONCLUSIONS: Linguistic validation of ePRO questionnaires for use in Indian settings presents special challenges. Evidence suggests that additional consideration must be exercised early on in order to avoid compounded difficulties and delays later in the process. In such cases, the use of a checklist is recommended to assist in preempting future difficulties within the project. PMC51 CONCEPT ELABORATION—AN ESSENTIAL STAGE IN THE TRANSLATION OF PRO MEASURES Verjee-Lorenz A, Clayson D, Two R, Giovanaz M PharmaQuest Ltd, Banbury, Oxfordshire, UK The ISPOR Principles of Good Practice paper for the translation of patient-reported outcome (PRO) measures describes concept elaboration within the scope of project preparation and emphasises that the misinterpretation of items or concepts can result if the step is not undertaken. However, the paper does not discuss how to correctly perform concept elaboration or why it is an essential step in the translation process. Translation companies often harmonise their own translations with each other, but not with existing translations. Over time, a single measure may be translated by any number of companies. Wording choices are often made by various translators and project managers without prior guidance—this can lead to wide variations. a concept 13th Euro Abstracts elaboration document approved by the developer provides consistent clarification and guidance to promote harmonization across all language versions regardless of who is carrying out the work. Inconsistent wording choices across language versions can be due to misunderstandings as a result of ambiguous wording in the source text, misinterpretation of idioms or the influence of personal choice resulting in inconsistent use of alternative phrases for terms that are not directly translatable. Concept elaboration remedies these issues by providing a line-by-line analysis of the source text which clarifies ambiguous wording, provides alternative wording for items with no direct translation, recommends suitable equivalents of culture-bound concepts and provides guidelines on nuance, tone, and emphasis. Concept elaboration ensures that the text is fully understood by all translators before any translation work begins. It encourages standardised decision making for every language and allows the developer to exercise influence over the style and content of the translations even if they are unable to review each translation individually. For studies including developer reviews, these reviews should become more straightforward as the translations will have been carried out according to the developer’s guidance. PMC52 TREATMENT SATISFACTION WITH MEDICATION: A REVIEW OF CONCEPTUAL FRAMEWORKS AND APPLICATIONS Bharmal M, Viswanathan S Quintiles, Rockville, MD, USA OBJECTIVES: The construct of treatment satisfaction with medication (TS-M) provides an avenue for incorporating the patients’ voice in treatment evaluation. In recent years, due to the growing influence of the consumer in health care decisions, there have been a substantial number of both generic and disease-specific TS-M instruments developed. This study provides a review of the available TS-M measures and examples of its applications. METHODS: A literature review was conducted using MEDLINE, ISPOR Research Digest and Google Scholar to identify generic and disease-specific TS-M instruments as well as any applications of these instruments. The review included evaluation of research articles, abstracts, review articles, news articles and editorials discussing instruments that measure TS-M. RESULTS: A total of 39 instruments that measure TS-M were identified. Out of the 39 instruments, four instruments were generic in nature that could be used across therapy areas and included the three versions of the Treatment Satisfaction Questionnaire for Medication (TSQM) (TSQM v1.4, TSQM vII, TSQM-9) and the SATMED-Q. The rest of the 35 measures were disease-specific TS-M instruments covering many diseases including migraine, pain, HIV, diabetes, Crohn’s disease, oncology, osteoarthritis, men’s health and multiple sclerosis. The instruments varied substantially in the number of items and domains they covered and included anywhere from 6 items to 46 items. Treatment satisfaction data has been used to inform clinician decision-making, support label claims, and for supporting the marketing of drugs. For example, in the past decade, TS-M has been used to support label claims for 17 products in Europe and in marketing effort of drugs in many therapy areas including erectile dysfunction, heart disease and diabetes. CONCLUSIONS: Treatment satisfaction with medication can be an important endpoint for product differentiation of novel treatments. Among the existing TS-M instruments, researchers could benefit by identifying instruments that have been developed using psychometrically rigorous standards and procedures. PMC53 COMPARISON OF METHODS FOR ITEM GENERATION AND PRETESTING OF FIVE LYMPHOEDEMA-SPECIFIC HEALTH-RELATED QUALITY OF LIFE QUESTIONNAIRES Williams AE Biomarin Europe Ltd, London, UK OBJECTIVES: A systematic literature review identified four patient-reported outcome (PRO) measures that have been used in the assessment of quality of life for patients with upper limb lymphoedema. These are in addition to the author’s newly devised questionnaire. However the developers had all used different methods to devise the items with the final measures incorporating different items and domains. The objective was to compare the methods that had been used and review in the light of available guidance on best practice. METHODS: Review and tabulate the available information on each PRO regarding its development. Key aspects of the evaluation included specificity of target population, source of item generation, number of initial items generated, drafting of item/question wording, pretesting and defining the domains. RESULTS: Two of the five PROs had been developed to assess the quality of life of patients with upper or lower limb lymphoedema decreasing the specificity of the target population; two of the PROs did not use patients as the source for items relying on clinicians; the number of items initially generated by the remaining three PROs were 98, 134 and 495, but none of the developers evaluated the point of saturation; different approaches were used for analyses of the items generated to create the initial draft of items/questions; not all PROs were pretested. Number of domains ranged from two to seven and number of items/questions in initial drafts ranged from 5 to 70. CONCLUSIONS: Each of the PROs measuring the same concept have been developed using different methods of generating the items and pretesting an initial draft of the PRO. To advance the field of measuring quality of life of lymphoedema patients, it will be important to establish which PRO measure has the best validity, that is, best reflects the impact of lymphoedema on their health and well-being. 13th Euro Abstracts PMC54 TREATMENT SATISFACTION INSTRUMENTS FOR DIFFERENT PURPOSES DURING A PRODUCT’S LIFECYCLE—KEEPING THE END IN MIND Rofail D1, Regnault A2, Taylor F1, Filonenko A3 1 Mapi Values Ltd, Bollington, Cheshire, UK; 2Mapi Values, Lyon, France; 3Bayer Schering Pharma AG, Berlin, Germany OBJECTIVES: There are three main parts during a product’s lifecycle when satisfaction instruments are particularly useful. These are to aid: 1) getting the product to the market; 2) getting the market to the product; and 3) demonstrating value for health care practitioners during daily clinical practice. This study investigates whether the development and implementation of treatment satisfaction instruments during a product’s lifecycle are informed by their purpose. METHODS: A literature review was performed between 2000 and 2010 using electronic databases (PUBMED, PsycINFO, and EMBASE) and keywords such as “satisfaction” and “medication” or “drug” and “questionnaire.” Relevant articles were reviewed in detail to extract information regarding the satisfaction instrument used, its development and validation, and when the instrument was used during a product’s lifecycle. Additional information was collated including the type of studies the instruments were used in, clinical condition/ indication, type of data generated (e.g. descriptive), and whether satisfaction was associated with other endpoints. RESULTS: Of 875 abstracts, 105 articles were further considered. The review indicated similarities regarding the development and validation of satisfaction instruments, such as using patient input to derive the items and exploring classical measurement properties specific to the target population. However, the specificities of the implementation of treatment satisfaction during the three main stages of a product’s lifecycle were rarely considered. CONCLUSIONS: The development and implementation of treatment satisfaction instruments during a product’s lifecycle rarely consider the purpose. By “keeping the end in mind,” data from treatment satisfaction instruments can help three key parts: 1) getting the product to the market thus helping to generate evidence as part of an overall value proposition; 2) getting the market to the product; and 3) demonstrating the value to clinical practice. Furthermore, the development, validation and interpretation of scores from treatment satisfaction instruments should be sensitive to the intended purpose. CONCEPTUAL PAPERS & RESEARCH ON METHODS – Statistical Methods PMC55 SYSTEMATIC REVIEW OF METHODS FOR META-ANALYSIS AND INDIRECT COMPARISON USED IN EXISTING SYSTEMATIC REVIEWS AND HTA REPORTS—RESULTS FROM THE FIRST PART OF EBAYESMET PROJECT Walczak J1, Borowiack E1, Kowalska M1, Nikodem M2, Zapalska A1 1 Arcana Institute, Cracow, Poland; 2CASPolska, Myslenice, Poland OBJECTIVES: To collect information serve as a background for further activities of the eBayesMet project. To perform assessment of the frequency of use the particular statistical methods for meta-analyses and indirect comparisons in existing systematic reviews and HTA reports. METHODS: Database of Reviews of Effectiveness (DARE) was searched for relevant reviews published between January 2009 and March 2010. RESULTS: The majority of meta-analyses were prepared by using direct statistical methods (97%). The most popular expression about fixed model was that fixed effect model with the Mantel-Haenszel was used (36%). In case of random model, information that method DerSimonian Laird was selected (38%) was used most often. Metaanalyses were prepared mostly based on randomized controlled trials (87%) and in 3% only on observational studies. In meta-analyses based on RCTs in 42% of cases authors mentioned that to make strong conclusion more studies were needed. Employing Bayesian methods was generally very rare. For indirect comparisons six methods were recognized and the most common type was MTC Bayesian Model (53%). Observational studies were not included in any identified analysis. In 40% of the review information indicating the need for additional studies was contained. CONCLUSIONS: Our systematic reviews demonstrates a wide range of approaches and methods for conducting meta-analyses and indirect comparison used in current practice. The most popular approach for indirect comparison is Bayesian included network and MTC (over 65%). However performed analysis indicated that Bayesian approach is still marginal methods for performing direct comparison based on head to head studies. Bayesian models have essential advantage: some additional data can be included (as a prior distribution). This extra information can be for instance data from observational studies. However it should be emphasized, that in the analyzed random sample of systematic reviews studies other than RCT were included in only 13%. PMC56 IMPROVED SURVIVAL CURVE FITS TO SUMMARY DATA FOR ECONOMIC EVALUATIONS Hoyle MW University of Exeter, Exeter, UK OBJECTIVES: Estimates of mean cost and quality-adjusted-life-years are central to the cost-effectiveness analysis of health technologies. They are often calculated from curve fits to overall survival and time on treatment, ideally by the method of maximum likelihood applied to individual patient data. However, such data is often not available. Instead, curves are commonly fit to summary Kaplan-Meier estimators, either by regression of the transformed estimator or by minimizing the sums of squares of A339 differences between actual and fitted values. However, the tail of the estimator is often uncertain due to small numbers of patients at risk, and the curve fits do not yield estimates of the true uncertainty in survival times, which is a very important component of overall uncertainty in cost-effectiveness. Here, I describe a new, more accurate method of fitting survival curves to summary survival data. METHODS: First, I estimate the underlying individual patient data from the Kaplan-Meier estimator, numbers of patients at risk and from other published trial-related information. The fitted curve is then estimated by maximum likelihood given the estimated underlying individual patient data. RESULTS: Simulation applied to individual patient data shows that the method tends to give a more accurate curve fit than the traditional methods of fitting to the Kaplan-Meier estimator. Furthermore, the curve fit is often very similar to that derived by fitting to the underlying individual patient data by maximum likelihood. The method naturally yields accurate estimates of the uncertainty in survival times. When applied to economic evaluations submitted to NICE, the method often yields substantially improved estimates of cost-effectiveness compared to estimates based on fitting survival curves in the traditional manner. This highlights the sensitivity of many cost-effectiveness analyses to curve fits. CONCLUSIONS: When only summary survival data is available, I recommend the method for cost-effectiveness analysis. PMC57 BIAS IS WORSE THAN NOISE: HANDLING MISSING DATA FOR CONFOUNDERS IN OBSERVATIONAL STUDIES Elkin EP, Exuzides AK, Pasta DJ, Miller DP ICON Clinical Research, San Francisco, CA, USA OBJECTIVES: Outcomes research often employs observational designs (e.g., disease registries, administrative health care data sets, chart reviews). Researchers using observational data may find various amounts of missing data for confounders when analyzing the association between an exposure (such as treatment use) and an outcome (such as an adverse event). This abstract examines the case when a potentially important confounding variable has a large amount of missing data and compares the analytic methods that may be used in this situation. METHODS: Strategies for handling missing confounder information include: (1) ignore confounders with lots of missing values; (2) exclude cases that are missing a confounder value; (3) impute a value for the confounder; (4) include missing as a separate category in the analysis. Data from a disease registry were used as the basis for simulations to compare the odds ratio for risk of death in patients who received a treatment compared to those without treatment. Both a clinical measurement and a subjective physician assessment are known to confound the relationship between treatment and death. RESULTS: The most problematic pattern of missing data was informative missing data. In one simulation, the clinical measurement was a strong predictor of death; however, it was disproportionately missing in patients who had died. The physician assessment predicted death strongly among patients missing the clinical measurement, but only weakly in patients not missing the clinical data. Different approaches to the missing confounder data either exacerbated or ameliorated the problem. CONCLUSIONS: Excluding cases can create misleading results due to selection bias. Combining all missing values into a separate category can create data “noise” (i.e., classification error); however, this may be the most transparent strategy and least likely to bias results. It is important to include all cases and all potential confounders in the analysis of outcomes research studies. PMC58 SYSTEMATIC REVIEW OF STATISTICAL METHODS OF META-ANALYSIS AND INDIRECT COMPARISON POTENTIALLY AVAILABLE TO USE IN SYSTEMATIC REVIEWS—RESULTS FROM THE FIRST PART OF EBAYESMET PROJECT Walczak J1, Nikodem M2, Siedmiogrodzki K2, Zapalska A1, Borowiack E1 1 Arcana Institute, Cracow, Poland; 2CASPolska, Myslenice, Poland OBJECTIVES: To collect information serve as a background for further activities of the eBayesMet project. To identify statistical methods and approaches for performing meta-analyses and indirect comparisons. To describe limitations and mathematical background, to discuss advantages and disadvantages of identified methods. METHODS: Databases such as MathSciNet, Medline and Cochrane Methodology Group resources, textbooks, guidelines for preparing systematic reviews were searched. We focused on papers with strictly mathematical descriptions of presented methods. RESULTS: Eleven main statistical methods were identified. There were six methods of meta-analysis (Inverse Variance, Mantel-Haenszel, DerSimonian-Laird, Peto, Bayesian and Generalized Linear Mixed Models for direct comparison), and five methods of indirect and mixed comparison (Bûcher, Minimal Squares, Lumley, Bayesian Mixed Treatment Comparison and Generalized Linear Mixed Models for indirect comparison). Moreover, there are different variants of some of these methods. For any type of analysis (direct, indirect, network, fixed, random, etc.) and any given data, there is an adequate version of Bayesian method for performing estimation. We found the Bayesian approach to be generally the most flexible.Analysis of precision and credibility of all identified statistical methods of meta-analysis and indirect comparison will be prepared in the next part of eBayesMet project. CONCLUSIONS: There are several statistical methods for performing proper data analysis in any model. The problem is with choosing optimal one for the given data set. On the other hand, all methods are less reliable and accurate for samples of data, especially for small (or zero) number of events in one or both arms. There is variety of statistical methods combining direct and indirect evidence, from which the most flexible is Mixed A340 Treatment Comparison using the Bayesian approach. Nevertheless, for each of these methods there are specific assumptions which have to be satisfied in order to obtain correct estimation. CONCEPTUAL PAPERS & RESEARCH ON METHODS – Study Design PMC60 INCLUSION OF CONFERENCE ABSTRACT DATA IN SYSTEMATIC REVIEWS OF PHARMACOLOGIC INTERVENTIONS IN DIFFERENT DISEASE AREAS Zhang Y Heron Evidence Development Ltd, Luton, England, UK OBJECTIVES: Conference searching is a common part of systematic review methodology, this study investigates what proportion of study/trial data included in systematic reviews of pharmacologic interventions is derived solely from data published in conference abstracts, for 3 different disease areas. METHODS: The Cochrane Library of systematic reviews (SRs) was searched for SRs on pharmacologic interventions which state that they include conference abstracts in their Specialized Registers and include conference searching as part of the stated SR methods. Included studies lists of completed systematic reviews were reviewed and the total number of included studies and the number of studies for which data was obtained only from conference abstracts were extracted. In disease areas where a large number of SRs met the inclusion criteria, the 10 most recently published SRs were selected. The following disease groups were considered: psychological disease (depression and bipolar disorder), female cancers (breast cancer and ovarian cancer), and arthritis (osteoarthritis and rheumatoid arthritis). RESULTS: 3/10, 9/10 and 4/10 SRs contained studies solely from conference abstracts, for psychological disease, female cancers and arthritis respectively. For psychological disease 5% (5/226) of all studies included in 10 SRs came only from conference abstracts, for female cancers this was 14% (31/220) and for arthritis this was 14% (33/232). CONCLUSIONS: The proportion of studies from conference abstracts only included in SRs varies in the 3 examined disease areas. From these results, there is some evidence to suggest that the disease area of the review should inform the decision of whether to include conference searching as part of the protocol. To answer this question more conclusively, a more expansive review of SRs should be conducted covering a greater number of SRs and disease areas. PMC61 MONITORING OF HEALTH ECONOMIC DATA IN CLINICAL TRIALS Bharmal M, Viswanathan S, Gemmen E Quintiles, Rockville, MD, USA OBJECTIVES: Health care decision makers are increasingly requesting health economic (HE) data, both to support product approval and for marketing purposes. Currently, there is limited information available to aid decisions surrounding the clinical monitoring of HE endpoints when they are collected as part of a clinical trial. It is necessary to understand the current level of monitoring activities surrounding HE data and evolve best practices for monitoring such data. METHODS: To better understand monitoring activities, a literature review was performed and qualitative in-depth interviews were conducted with six clinical research associates (CRA) who had experience collecting HE data as part of clinical trials in a range of therapy areas. The literature review and interviews focused on understanding current clinical trial monitoring practices, monitoring activities specific to HE data, the challenges faced during monitoring and recommendations for the future. RESULTS: All CRAs interviewed reported working either with patient-reported outcome (PRO) measures— quality of life questionnaires, patient diaries—and/or health care resource utilization data, in different therapy areas. Data monitoring activities in clinical trials can include a number of specific tasks ranging from full source data verification (SDV) to partial SDV to just checking for accuracy, legibility and completeness. The most common challenges in monitoring of HE data included incomplete questionnaires, misinterpretation of questionnaire data by the sites, and difficulty in SDV of health care resource utilization data by the CRA. Recommendations for the future included optimizing methods for documentation of health care resource utilization data, improving patient/ site training in PRO use, and selecting the type of PRO and mode of PRO administration based on the patient population being examined. CONCLUSIONS: Health economic endpoints are increasingly being used in clinical trials, and CRAs are becoming familiar with PRO and health care resource utilization data. Monitoring activities for HE data vary by the study design and type of data collected. CARDIOVASCULAR DISORDERS – Clinical Outcomes Studies PCV1 ESTIMATION OF STROKE-RELATED ADVERSE EVENTS, HEALTH CARE UTILITY AND COST OF PATIENTS WITH NON-VALVULAR ATRIAL FIBRILLATION Wang L1, Baser O2 1 STATinMED Research, Ann Arbor, MI, USA; 2STATinMED Research/University of Michigan, Ann Arbor, MI, USA OBJECTIVES: To estimate mortality, health care utility and health care cost burden of patients who suffered a stroke during the 180 days after diagnosis of non-valvular atrial 13th Euro Abstracts fibrillation (NVAF) and compare it with patients who did not suffer a stroke. METHODS: Based on 2005–2007 U.S. Medicare advantage insurance claim files, patients aged 65 years and older who had 2 or more primary diagnoses for NVAF within 30 days of one another were selected. The 180-day follow-up event rates, health care facility use and health care cost for patients with a stroke and those without were compared. Risk adjustment was done by using the propensity score matching method with the ProbChoiceTM algorithm. RESULTS: Out of patients identified with NVAF (n = 18,575), 575 suffered a stroke during the 180 days after NVAF diagnosis. 94% (n = 541) did not have stroke during the baseline period (180 days before NVAF diagnosis). Patients were not significantly different in terms of gender, region, and baseline comorbid conditions. After risk-adjustment for pre-specified covariates, mortality (7.14% vs. 2.09% P < 0.0001), outpatient emergency room (ER) visits (79.97% vs. 46.34% P < 0.0001), acute coronary syndrome (43 vs. 16/100 person years), transient ischemic attack (73 vs. 4 /100 person years), major bleeding (85 vs. 4 /100 person years) and myocardial infarction (32 vs. 9/100 person years) were all higher for patients who suffered a stroke compared to those who did not. Besides inpatient cost ($24,116 vs. $20,828), risk-adjusted outpatient ER costs ($921 vs. $873) were also higher for stroke patients. Overall risk-adjusted difference in health care costs is significant ($33,430 vs. $16,375 P < 0.0001). CONCLUSIONS: Most of the adverse events analyzed were higher for patients who suffered a stroke after NVAF relative to patients who did not. Total health care utility and health care cost were also significantly increased. PCV2 ESTIMATION OF ADVERSE EVENTS IN 3 MONTHS AFTER VENOUS THROMBOEMBOLISM EVENT FOR MEDICARE PATIENTS WHO UNDERWENT HIP FRACTURE SURGERY Wang L1, Dysinger A1, Baser O2 1 STATinMED Research, Ann Arbor, MI, USA; 2STATinMED Research/University of Michigan, Ann Arbor, MI, USA OBJECTIVES: To estimate mortality, re-hospitalization and bleeding 180 days after a venous thromboembolism (VTE) event in patients following hip fracture surgery and to compare the outcomes with patients without VTE. METHODS: Based on 2004– 2006 national Medicare claims, all patients who underwent hip fracture surgery were identified. a total of 180 days follow-up event rates for patients who had a VTE event during their initial hospitalization were calculated. Events were compared between patients who suffered a VTE event and those that did not. Risk adjustment was done using propensity score matching (using the ProbChoiceTM algorithm) controlling for baseline demographic and clinical characteristics between patients with and without VTE. RESULTS: In patients who underwent hip fracture surgery (n = 77,743), 2.23% had post-operative VTE events during their initial hospitalization. Almost 72.96% (n = 1263) of these patients suffered deep vein thrombosis (DVT), 20.97% (n = 363) had a pulmonary embolism (PE), and 6.07% (n = 105) had both DVT and PE. After multivariate adjustment for pre-specified covariates, mortality was almost 50% higher for patients with VTE compared to those without VTE. Differences in mortality rate were more pronounced for PE patients, whom the event was associated with almost two-fold. The VTE group was more likely to be re-hospitalized in one year (odds ratio: 1.18, p = 0.2720). Bleeding was 1.8 times higher (p = 0.0080). CONCLUSIONS: VTE events during initial hospitalization for hip fracture surgery increased patients’ mortality, re-hospitalization and bleeding compared to patients with no VTE events. PCV3 COMPARISON OF MORTALITY, HEALTH CARE UTILITY AND COST OF PATIENTS WITH WARFARIN TREATMENT FOR NON-VALVULAR ATRIAL FIBRILLATION VERSUS PATIENTS WITH OTHER TREATMENT Wang L1, Baser O2 1 STATinMED Research, Ann Arbor, MI, USA; 2STATinMED Research/University of Michigan, Ann Arbor, MI, USA OBJECTIVES: To estimate the economic and clinical burden of patients who used warfarin during the 180 days after diagnosis of non-valvular atrial fibrillation (NVAF) and compare it with patients who did not use warfarin. METHODS: Based on 2005–2007 U.S. Medicare advantage insurance claim files, patients aged 65 years and older who have had 2 or more primary diagnoses for NVAF occurring within 30 days of one another were selected. The 180 days follow-up event rates, health care facility use and cost were compared. Risk adjustment was done by using the propensity score matching method with the ProbChoiceTM algorithm. RESULTS: In patients who identified with NVAF (n = 18,575) 12,186 used warfarin during the 180 days after NVAF diagnosis and 6,389 used other drugs or did not use any drugs. Patients were significantly different in terms of age, gender, comorbid conditions and baseline CHADS score. After risk-adjustment for pre-specified covariates, mortality (0.75% vs. 2.26%), outpatient emergency room (ER) visits (48.79% vs. 53.30%), acute coronary syndrome (105 vs. 154 /100 person years), and myocardial infarction (61 vs. 112 /100 person years) were all lower for patients who had warfarin with non-valvular atrial fibrillation. Even though drug cost is higher for the warfarin group ($1,687 vs. $1,595), risk-adjusted outpatient ER costs ($756 vs. $861) were lower. Overall riskadjusted health care costs did not differ ($12,739 vs. $15,359). CONCLUSIONS: Most of the adverse events analyzed were lower for patients who had warfarin after non-valvular atrial fibrillation relative to patients who did not. However, the economic burden of both groups of patients on the health care system was similar. 13th Euro Abstracts PCV4 CLINICAL AND COST OUTCOMES OF VENOUS THROMBOEMBOLISM IN PATIENTS WHO UNDERWENT MAJOR ORTHOPEDIC SURGERY Baser O1, Wang L2, Xie L2 1 STATinMED Research/University of Michigan, Ann Arbor, MI, USA; 2STATinMED Research, Ann Arbor, MI, USA OBJECTIVES: To evaluate the effects of venous thromboembolism (VTE) occurring within 30 days of hospital discharge after major orthopedic surgery on inpatient costs and mortality within 1 year after initial hospitalization for the surgery. METHODS: U.S. national Medicare data was used for this analysis. All patients who underwent total hip replacement (THR) (n = 51,108) or total knee replacement (TKR) (n = 115,627) surgery were identified. VTE events were diagnosed within the first 30 days and within 1 year post-discharge. Multivariate regression was used to control for differences in baseline characteristics in patients with and without VTE events. Total cost was measured as Medicare cost plus beneficiary out-of-pocket cost. RESULTS: VTE occurred in 0.74% of patients undergoing THR. For patients with VTE versus no VTE, mortality was higher (2.9% vs. 0.4%, P < 0.001). Medicare and total health care cost, including the beneficiary cost share in 1 year, were not different for VTE versus no VTE ($19,300 vs. $19,044, p = 0.817), but the beneficiary cost share was higher for VTE ($3,274 vs. $1,966, P < 0.001). VTE occurred in 0.70% of patients undergoing TKR. For patients with VTE versus no VTE, mortality was higher (2.5% vs. 0.15%, P < 0.001). Total health care cost, including the beneficiary cost share in 1 year, was not different for VTE versus no VTE ($17,970 vs. $16,766, p = 0.142), but the Medicare and beneficiary cost shares were both slightly higher for VTE ($3,274 vs. $1,966 and $2,650 vs. $1,630 for THR and TKR patients, respectively, P < 0.001). CONCLUSIONS: VTE after major orthopedic surgery is associated with higher mortality compared with no VTE. VTE did not affect total health care costs in 1 year, but beneficiary costs were higher. Risk-adjusted total, Medicare, and beneficiary health care costs were significantly higher for major orthopedic surgery patients with VTE (P < 0.001). PCV5 COMPARISON OF ADVERSE EVENTS OF MEDICARE PATIENTS WHO UNDERWENT KNEE REPLACEMENT SURGERY AND EXPERIENCED VENOUS THROMBOEMBOLISM VERSUS NO VENOUS THROMBOEMBOLISM Wang L1, Dysinger A1, Baser O2 1 STATinMED Research, Ann Arbor, MI, USA; 2STATinMED Research/University of Michigan, Ann Arbor, MI, USA OBJECTIVES: To estimate re-hospitalization and bleeding rates during the 30 days after a venous thromboembolism (VTE) event in patients who underwent knee replacement surgery and to compare the outcomes with patients who did not suffer VTE. METHODS: Based on 2004–2006 national Medicare claims, all patients who underwent knee replacement surgery were identified. The 30 days follow-up event rates for patients who had a VTE event during their initial hospitalization were calculated. Events were compared between patients who suffered a VTE event and those who did not. Risk adjustment was done using propensity score matching (using the ProbChoiceTM algorithm) controlling for baseline demographic and clinical characteristics between patients with and without VTE. RESULTS: In patients who underwent total knee replacement surgery (n = 104,952), 1.9% had post-operative VTE events during their initial hospitalization. Almost 69% (n = 1377) of these patients had deep vein thrombosis (DVT), 25% (n = 501) had pulmonary embolism (PE), and 6% (n = 119) had both DVT and PE. The overall likelihood of mortality was four times higher for VTE patients versus those without VTE (1.35% vs. 0.35%). Patients with VTE during their initial hospitalization were more likely to be hospitalized in 30 days compared to patients without an event during the same hospital stay (16.62% vs. 8.00%). In 30 days after the event, patients with VTE were more likely to have bleeding (9.31% vs. 2.18%). CONCLUSIONS: VTE events during initial hospitalization for total knee replacement surgery patients increased the adverse events compared with no VTE events. PCV6 OUTCOME OF ADVERSE DRUG REACTIONS REGISTERED WITH INTENSIVE MONITORING SYSTEM Mugosa S1, Todorovic Z2, Boskovic A3 1 School of Medicine, University of Montenegro, Podgorica, Montenegro; 2School of Medicine, University of Belgrade, Belgrade, Serbia; 3Clinical Center of Montenegro, Podgorica, Montenegro OBJECTIVES: Adverse drug reactions (ADR) appear more frequently then what is actually reported and registered. The aim was to establish intensive monitoring system and to analyze ADR in hospitalized patients. METHODS: Prospective study covered 200 patients hospitalized in Cardiology Center, Clinical Center of Montenegro. ADR were collected in the following way: patients were interviewed on the basis of the symptoms list and any signs which could point to eventual ADR. Secondly, lab tests and other available parameters were observed. RESULTS: At the time when interviews took place, patients received on average 7.96 ± 2.63 medicines (2–17). In total, 67 patients (34%) had 75 ADR. Twenty-one ADR (28%) are classified as serious. Fiftyfour ADR have resulted in the recovery of the patient (72%), eight had as an outcome prolonged hospitalization (11%), another 8 were life threatening (11%), while 5 ADR (6%) were the cause of the hospitalization. The most frequent ADR which had as an outcome admission to hospital were caused by digoxin (40%), prolonged hospital stay A341 by furosemide (38%), while the most frequent registered ADR which were life threatening were caused by streptokinase (50%). CONCLUSIONS: Monitoring ADR in patients using cardiovascular drugs is a matter of importance since this class of medicine is usually used by elderly patients with critical conditions and underlying diseases. PCV7 SPONTANEOUS REPORTING OF ADVERSE DRUG REACTIONS INVOLVING HOSPITALIZED CARDIAC PATIENTS Mugosa S1, Todorovic Z2, Boskovic A3 1 School of Medicine, University of Montenegro, Podgorica, Montenegro; 2School of Medicine, University of Belgrade, Belgrade, Serbia; 3Clinical Center of Montenegro, Podgorica, Montenegro OBJECTIVES: Adverse drug reactions (ADR) in hospital are a significant cause of morbidity and mortality. The aim is to analyze the ADR which have been collected through doctors’ and medical technicians’ spontaneous reporting. METHODS: The prospective study covered 655 patients hospitalized in Cardiology Center of Clinical Center of Montenegro. ADR were collected by spontaneous reporting. According to the criteria of World Health Organization, for every ADR causality assessment, severity, type, outcome, level of intervention and place of manifestation of ADR was established. RESULTS: In the 6 month period of research at the Cardiology Center, 655 patients were hospitalized with the average age of 60.72 ± 11.36 years. With spontaneous reporting, 22 patients had 22 ADR which were noted. Doctors reported a statistically significant larger number of ADR (68%) than medical technicians (32%), P < 0,01. Nine patients (41%) had serious ADR with the following outcome: 21 ADR were the cause of the hospitalization (22%), four ADR resulted in the extended hospitalization (44%), and 3 were life threatening (33%). The most frequent serious ADR were reported spontaneously by doctors and medical technicians and were caused by streptokinase (33%). The most frequent symptoms which the patients had as a consequence of ADR were: fainting (27%), headache (18%), weakness (14%), vertigo (14%) and vomiting (9%). CONCLUSIONS: Considering increased use of cardiovascular drugs and limitations in pre-marketing trials for drug safety evaluation, post marketing evaluation of adverse drug reactions induced by this class of medicinal products seems necessary. PCV8 EVALUATION OF CLINICAL AND DIRECT ECONOMIC OUTCOMES FOLLOWING AN ACUTE CORONARY EVENT AMONG PATIENTS WITH SUBOPTIMAL HDL-C OR TRIGLYCERIDES ADMINISTERED EITHER STATIN AND NIACIN EXTENDED-RELEASE OR STATIN MONOTHERAPY Cziraky MJ1, Quinlan S1, Quimbo RA1, Simko RJ2, Webb S2 1 HealthCore, Wilmington, DE, USA; 2Abbott Laboratories, Abbott Park, IL, USA OBJECTIVES: To compare the risk of subsequent coronary event (SCE) and associated costs between patients maintaining statin monotherapy (SM) or augmenting with niacin extended-release (NER) after an acute coronary event (ACE). METHODS: Patients with an ACE between January 1, 2004 and March 31, 2009 were identified (Index Date was defined as the earliest ACE). Patients age ≥18, ≥6 months eligibility pre- and post-Index Date, ≥1 statin fill and no statin augmenting fill 6 months before the Index Date, and sub-optimal baseline HDL-C (<40 or <50 mg/dL for men and women) or triglyceride (>150 or >200 mg/dL for diabetic and non-diabetic) were included. Cohorts were determined based on NER or SM fills within 6 months of the Index Date (Cohort ID Period). Patients with alternative lipid agents or SCE during the Cohort ID Period were excluded. Outcomes included SCE and event-related medical costs. SCE risk was compared using Cox proportional hazards models, while costs were compared via generalized linear models. Multivariate analyses controlled for age, gender, health coverage, geography, co-morbidities, concomitant medications, baseline lipids, and severity of initial ACE. RESULTS: A total of 1998 patients added NER (n = 145) or maintained SM (n = 1,853) and met all study criteria. NER patients were younger (57 ± 8 vs. 61 ± 11, P < 0.0001), less likely female (8% vs. 33%, P = 0.002), and had lower baseline LDL-C (81 ± 32 vs. 99 ± 41; P < 0.0001), HDL-C (34 ± 7 vs. 40 ± 9; P < 0.0001), and triglycerides (185 ± 118 vs. 208 ± 131; P = 0.0290). Compared to SM, NER patients had a 43% lower SCE risk (hazard ratio: 0.57, 95%CI: 0.35–0.94; P = 0.0274). Clinical findings corresponded with 56% reduction in adjusted annual event-related costs compared to SM (0.44; 95% CI: 0.25–0.80 p = 0.0068). CONCLUSIONS: Among patients requiring comprehensive lipid management, NER demonstrated lower SCE risk compared to SM, consequently reducing annual event-related medical costs following an ACE. PCV9 RATE OF SUBSEQUENT CORONARY EVENTS AND DIRECT ECONOMIC OUTCOMES FOLLOWING AN ACUTE CORONARY EVENT BETWEEN PATIENTS AUGMENTING STATIN TO FURTHER REDUCE LDL-C VS. IMPROVING A COMPREHENSIVE LIPID PROFILE Quinlan S1, Quimbo RA1, Simko RJ2, Cziraky MJ1, Webb S2 1 HealthCore, Wilmington, DE, USA; 2Abbott Laboratories, Abbott Park, IL, USA OBJECTIVES: Compare the annual subsequent coronary event rate (ASCER) and associated costs between patients initiating adjunct therapy with either niacin extended-release (NER) for comprehensive lipid panel management or ezetimibe (EZE) for continued LDL-C reduction after an acute coronary event (ACE). METHODS: Patients with an ACE between January 1, 2004 and May 31, 2009 were identified A342 (Index Date was defined as the earliest ACE). Patients age ≥18, ≥6 months eligibility pre- and post-Index Date, ≥1 statin fill and no statin augmenting fill 6 months before the Index Date, and sub-optimal baseline HDL-C (<40 or <50 mg/dL for men and women) or triglyceride (>150 or >200 mg/dL for diabetic and non-diabetic) were included. Cohorts were determined based on NER or EZE fills within 6 months of the Index Date (Cohort ID Period). Patients with alternative lipid agents or coronary event during the Cohort ID Period were excluded. Outcomes included ASCER and event-related medical costs and were compared via generalized linear models. Multivariate analyses controlled for age, gender, health coverage, geography, co-morbidities, baseline lipids, and severity of initial ACE. RESULTS: A total of 328 patients added NER (n = 145) or EZE (n = 183) and met all study criteria. NER patients were less likely female (8% vs. 29%, P < 0.0001), had lower baseline LDL-C (81 ± 32 vs. 112 ± 40; P < 0.0001) and HDL-C (34 ± 7 vs. 40 ± 9; P < 0.0001). Age (57 ± 8 vs. 58 ± 10; P = 0.2020), Deyo-Charlson comorbidity index (0.6 ± 1.1 vs. 0.9 ± 1.4; P = 0.0923) and triglycerides (185 ± 118 vs. 199 ± 94; P = 0.2531) were not statistically different. Compared to EZE, NER patients had a 57% decrease in ASCER (incidence ratio: 0.43, 95%CI: 0.24–0.77). Clinical findings corresponded with 87% reduction in adjusted annual event-related costs compared to EZE (0.13; 95% CI: 0.04–0.41). CONCLUSIONS: Among patients requiring lipid management, adopting a comprehensive lipid approach demonstrated lower ASCER compared to further reducing LDL-C, consequently reducing annual event-related medical costs following an ACE. PCV10 ANALYSIS OF INAPPROPRIATE SHOCKS IN IMPLANTABLE CARDIOVERTER DEFIBRILLATOR (ICD) DEVICES Neophytou I1, LeReun C2, Diamantopoulos A1 1 Symmetron Ltd., London, UK; 2Independent Biostatistician, Carrigaline County, Cork, Ireland OBJECTIVES: A side effect of implantable cardioverter defibrillator (ICD) devices is inappropriate shocks. Shocks are painful, can potentially induce lethal arrhythmias and cause severe psychological trauma to cardiac arrhythmic patients. The objective of this study was to investigate the risk of inappropriate shocks in patients who receive ICD therapy. METHODS: A comprehensive literature search was conducted to identify studies describing ICD therapy and reporting inappropriate shock data. The analysis included studies of patients with implantation performed after 2002, having a minimum study population of 50 participants and a minimum 12 month follow up period. The database search was augmented by hand searching references of contemporary systematic reviews in the area of ICD therapy, to ensure all relevant studies were captured. Patient numbers from the identified studies were pooled and summary statistics were estimated. RESULTS: A total of 415 published articles were identified via the database search. Following abstract and full text review, 32 studies were included for data extraction and statistical analysis. a further 4 studies were included from the systematic review hand search. The 36 studies provided a combined total of 807 patients. 16.2% (SD: 11.1) of patients experienced at least one inappropriate shock, this was reduced to 15.3% with exclusion of paediatric studies. The median duration of follow up was 2.6 years. The average number of inappropriate shocks per year for all patients is 0.4 (SD: 0.4). The frequency of inappropriate shocks amongst patients who experienced at least one inappropriate shock is 2.2 (SD: 1.7) per year. In 11 studies that present detailed information on the nature of shocks 32.9% (SD: 22.9) of total shock events were recorded as inappropriate. CONCLUSIONS: A high proportion of patients in recent trials experienced inappropriate shocks. Future development of ICD therapy should focus on the reduction of the inappropriate shock risk. PCV11 PLATELET FUNCTION TESTS ARE HIGHLY VARIABLE WHEN USED TO IDENTIFY PATIENTS RESISTANT TO CLOPIDOGREL Beard K1, Folia C1, Liovas A2 1 Agro Health Associates Inc., Burlington, ON, Canada; 2AstraZeneca Canada Inc., Mississauga, ON, Canada OBJECTIVES: Clopidogrel with aspirin is used to prevent thrombosis. However, significant interpatient variability in platelet response to clopidogrel has been reported. As many as 33% of patients are considered to be clopidogrel non-responders (NRs). a recent meta-analysis has shown that clopidogrel resistance is associated with a 3.5fold increased risk of ischemic events. There are several platelet function assays (PFAs) available to assess clopidogrel resistance, the gold-standard being light transmission aggregometry (LTA). Other PFAs have been developed, including vasodilator-stimulated phosphoprotein phosphorylation (VASP), and VerifyNow® (VN), the latter being a point-of-care assay. However PFAs are not frequently used in the clinical setting since they are expensive, and time- and labour-intensive. It is of interest to compare the reliability of various PFAs in the identification of clopidogrel NRs. METHODS: A MEDLINE search was conducted (1966–2010) using the following MeSH terms: Clopidogrel, Resistance, and Platelet Function Assay. Patient-based studies which compared the proportion of clopidogrel NRs using different PFAs were retrieved. RESULTS: Four studies were retrieved. Of these studies, 3 demonstrated significant heterogeneity among PFAs in their assessment of clopidogrel responsiveness. The proportion of clopidogrel NRs identified with the various PFAs ranged from 13%– 39%, 44%–69%, 35%–90%, and 69%–70% in the 4 studies respectively. CONCLUSIONS: The ability to identify clopidogrel NRs by different PFAs varies greatly according to the assay used. In addition, LTA and VASP are associated with poor quality control, expense and long turnaround times, and they are typically restricted to specialized laboratories in research hospitals. The inability of current PFAs to accurately assess clopidogrel resistance may place unidentified clopidogrel NRs at 13th Euro Abstracts increased risk of ischemic events. The challenges and costs associated with such tests may be eliminated by newer therapies which have less response variability than clopidogrel. PCV12 COMPARATIVE EFFICACY OF EZETIMIBE-STATIN COMBINATION THERAPY AND STATIN MONOTHERAPY IN PATIENTS WITH HYPERCHOLESTEROLEMIA: SYSTEMATIC REVIEW AND METAANALYSIS OF RANDOMISED CONTROLLED TRIALS Tunceli K1, Lawson RW2, Sibbring GC2, McCormick AL2, Tershakovec AM3, Davies GM3, Mikhailidis DP4 1 Global Health Outcomes, Merck, Whitehouse Station, NJ, USA; 2Complete Market Access, Macclesfield, Cheshire, UK; 3Merch Research Laboratories, West Point and Upper Gwynedd, PA, USA; 4Department of Clinical Biochemistry (Vascular Disease Prevention Clinics), Royal Free Hospital Campus, University College London (UCL) Medical School, UCL, London, UK OBJECTIVES: To review and analyze evidence for the comparative cholesterollowering efficacy of ezetimibe-statin combination therapy versus statin monotherapy in adults with primary hypercholesterolemia. METHODS: MEDLINE, EMBASE, and the Cochrane Central Register of Controlled Trials were searched to identify parallelgroup, randomised controlled trials of ezetimibe-statin combination therapy versus statin monotherapy, published between January 1993 and March 2010. Studies were selected using predefined criteria including minimum treatment of 4 weeks. Two reviewers conducted screening of articles, critical appraisal and data extraction, with a third reviewer resolving disagreements. Study quality was assessed using the Cochrane Collaboration “risk of bias” assessment tool. The difference between treatments in mean percentage change from baseline in low-density lipoprotein cholesterol (LDL-C) and proportion of patients attaining LDL-C treatment goal was analyzed. Data were combined by pair-wise, direct comparison random effects meta-analysis, with heterogeneity assessed using the I2 statistic. RESULTS: Fifteen articles met selection criteria; 13 studies involving over 5000 patients could be included in the meta-analyses. Data on simvastatin, atorvastatin and rosuvastatin were analyzed. There was a significantly greater percentage reduction in LDL-C levels in patients treated with ezetimibe-statin monotherapy compared with those treated with statin monotherapy (weighted mean difference: 14.11% [95% CI: 16.13, 12.1], P < 0.001), similarly, a greater proportion of patients achieved LDL-C goal when treated with ezetimibe-statin combination therapy compared with those treated with statin monotherapy (odds ratio 2.38 [95% CI: 1.89, 2.94], P < 0.001). The proportion of patients achieving LDL-C goal ranged from 12–83% (median 60%) in ezetimibe-statin arms of studies and from 2–52% (median 23%) in the statin monotherapy arms. CONCLUSIONS: The results indicate that ezetimibe-statin combination therapy is significantly more effective in reducing LDL-C levels than increasing the statin monotherapy dose, thereby enabling more patients to achieve their LDL-C goal. PCV13 META REGRESSION ANALYSIS TO INDIRECTLY COMPARE THE SAFETY AND EFFICACY OF DALTEPARIN TO ENOXAPARIN FOR THE PREVENTION OF VENOUS THROMBOEMBOLIC EVENTS (VTES) IN TOTAL HIP REPLACEMENT (THR) SURGERY Dranitsaris G1, Jelincic V1, Choe Y2 1 Augmentium Pharma Consulting, Toronto, ON, Canada; 2Eisai Inc, Woodcliff Lake, NJ, USA OBJECTIVES: Patients undergoing elective THR surgery are at an increased risk for VTEs. Even though dalteparin and enoxaparin have proven clinical effectiveness in placebo controlled studies, there have been no head-to-head trials to assess comparative effectiveness. Meta-analytic techniques were used for an indirect statistical comparison on the safety and efficacy between dalteparin and enoxaparin in patients undergoing THR surgery. METHODS: A literature search of major databases was conducted from 1980 to 2009 for randomized placebo controlled trials evaluating both agents in THR patients. Binary outcomes (e.g. VTE rates) for the LMWH group relative to control group were statistically pooled using fixed or random effects models in cases of significant heterogeneity. In trials where a common control was used, indirect statistical comparisons between dalteparin and enoxaparin were performed using meta regression analysis with active drug (dalteparin or enoxaparin) as the primary independent variable. RESULTS: A total of nine placebo controlled enoxaparin (n = 5) and dalteparin (n = 4) trials met the inclusion criteria. The meta analysis of all trial data showed that THR patients treated with enoxaparin or dalteparin had a 54% VTE relative risk reduction compared to the placebo (RR = 0.46, p < 0.001). This benefit was achieved without a significant increase in the risk for major bleeds (RR = 1.14, p = 0.80), heparin induced thrombocytopenia (HIT) (RR = 0.83, p = 0.75) or death (RR = 0.77, p = 0.61). The indirect statistical comparison was not able to find significant differences between enoxaparin and dalteparin in terms of VTEs (p = 0.86), risk of major bleeds (p = 0.46), HIT (p = 0.75) and death (p = 1.0). CONCLUSIONS: The findings of the analysis suggest comparable safety and efficacy between dalteparin and enoxaparin in THR patients. Therefore, treatment decisions should be based on other considerations such as patient or physician preference, ease of administration and cost. 13th Euro Abstracts PCV14 ASPIRIN USE IS NOT ASSOCIATED WITH REDUCED MYOCARDIAL INFARCTION IN ELDERLY WITH TYPE 2 DIABETES WITH OR WITHOUT PRIOR HISTORY OF CARDIOVASCULAR DISEASE Sirois C, Moisan J, Grégoire JP Laval University, Québec, QC, Canada OBJECTIVES: Aspirin is recommended for prevention of myocardial infarction (MI) in elderly with type 2 diabetes although the benefits of aspirin therapy have not been well established in this population. We evaluated the association between aspirin use and myocardial infarction (MI) among elderly newly treated with oral antidiabetes drugs, according to whether or not they had prior history of cardiovascular disease. METHODS: Using Quebec administrative databases, we conducted two nested casecontrol analyses within a population-based cohort of 39,680 individuals aged 66 years, newly treated with an antidiabetes drug, who had not used aspirin, antiplatelet or anticoagulant drugs and had no MI in the year before cohort entry. In the first analysis, eligible individuals had no previous cardiovascular disease recorded in the database, whereas in the second analysis, eligible individuals had a known cardiovascular disease. Cases were patients who had a MI. For each case, we selected five controls matched for age, year of cohort entry and sex using incidence density sampling. Exposure to aspirin was defined as current, past or none. Odds ratios (OR) of MI were calculated using paired multivariate conditional logistic regression. RESULTS: We identified 1101 cases of MI among the cardiovascular disease-free individuals. Neither current [OR = 1.03 (95% CI: 0.82–1.28)] nor past users [0.95 (0.74–1.24)] of aspirin had a significant different risk of MI than non users. There were 770 cases of MI among those individuals with a prior history of cardiovascular disease. Compared with non users of aspirin, both past [1.72 (1.30–2.28)] and current users [1.33 (1.03–1.71)] had a significant higher risk of MI. CONCLUSIONS: These results suggest that aspirin does not provide a cardiovascular protection to elderly individuals with type 2 diabetes. PCV15 META-REGRESSION APPROACH TO IDENTIFY THE EQUIVALENCE DOSE RATIOS OF ROSUVASTATIN TO SIMVASTATIN AND ATORVASTATIN Jakubczyk M1, Perkowski P2, Paweska J1, Macioch T1, Niewada M1, Rdzanek M2, Faluta T2 1 HealthQuest sp z o.o., Warsaw, Poland; 2AstraZeneca Pharma Poland Sp. z o.o., Warsaw, Poland OBJECTIVES: To identify the equivalent hypolipidemic doses of rosuvastatin compared to atorvastatin and simvastatin. METHODS: A systematic review of clinical trials was conducted. The percentage reduction of baseline LDL cholesterol was used as a measure of clinical effectiveness. The weighted mean difference (WMD) of effectiveness between comparators and rosuvastatin (separately) was calculated using the meta-regression approach, with logarithm of dosage ratio as an explanatory variable. Due to the heterogeneous baseline levels of cholesterol, the random effects model was used. As several observations came from single trials (various dosages of a given drug) a correlation of error terms may result. In a simplified, benchmark analysis no correlations were assumed. In a full analysis it was assumed that: for two observations from a single trial regarding different patients the covariance equals the common variance of random-effects model; for two observations from a single trial with one group of patients in common the covariance is further increased by the variance of the effect for the shared group. Monte Carlo approach and Cholesky decomposition were used to calculate confidence intervals of equivalent dosage ratios taking into account the correlation of model estimates. RESULTS: In total 41 trials were found and used, allowing for 78 single comparisons of rosuvastatin with atorvastatin, and 25 with simvastatin. Dosage ratios of comparator to rosuvastatin used in the meta-regression amounted to 1:1, 2:1, 4:1, and 8:1. The full analysis yielded WMD = −8.64 + 7.2*ln(dosage ratio) for atorvastatin and WMD = −15.94 + 7.55*ln(dosage ratio) for simvastatin. Benchmark analysis gave similar results. Resulting equivalence ratios amount to 3.32 for atorvastatin (95%CI: 2.93–3.79), and 8.26 (95%CI: 6.86–10.28) for simvastatin. CONCLUSIONS: Meta-regression approach allows for equivalence dosage analysis. Correlation issues need to be addressed for multiple treatment arms in a single study. Rosuvastatin yields equivalent hypolipemic effect for lower dosages than atorvastatin and simvastatin. PCV16 THE EFFECT OF SITAXENTAN ON EXERCISE CAPACITY, HEMODYNAMIC FUNCTION, AND HEALTH-RELATED QUALITY OF LIFE IN ADULTS WITH PULMONARY ARTERIAL HYPERTENSION Hwang LJ1, Liu X2, Teal SA3, Louie M1, Mychaskiw MA1 1 Pfizer Inc, New York, NY, USA; 2Aerotek, Houston, TX, USA; 3Pfizer Ltd, Tadworth, Surrey, UK OBJECTIVES: Pulmonary arterial hypertension (PAH) is progressive and fatal. Hemodynamic parameters, physical functioning (PF), and health-related quality of life (HRQoL) worsen without treatment. The objective was to evaluate the distribution of changes in exercise capacity, hemodynamic parameters, and HRQoL in PAH patients treated with sitaxentan. Because cumulative distribution functions (CDFs) characterize treatment effects beyond conventional statistical significance (which is achievable in large studies with small changes that lack clinical meaning), this method was employed. METHODS: Adult PAH patients (n = 178) received 12 weeks of randomized, double-blind treatment with placebo (n = 60) or sitaxentan 100 (n = 55) or 300 (n = 63) mg once daily. The primary outcome measure was change in percent of A343 predicted peak VO2 (PVO2); 6-minute walk distance (6MWD) measured exercise capacity. Hemodynamic assessments included pulmonary and systemic vascular resistance (PVR and SVR, respectively), pulmonary capillary wedge pressure (PCWP), and cardiac index (CI). The Short Form-36 (SF-36) assessed HRQoL, including the PF domain. CDFs plotted the cumulative percentage of patients against percentage (and, separately, numeric) changes from baseline to week 12 in outcome measures. Kolmogorov-Smirnov tests assessed differences in separation between CDFs for sitaxentan and placebo. RESULTS: Significant differences between CDFs were observed for sitaxentan 300 mg (but not 100 mg) vs. placebo for both numeric and percentage change in percent of predicted PVO2 (P = 0.03). Significant differences occurred for sitaxsentan 100 and 300 mg vs. placebo for 6MWD, PVR, SVR, and CI (P < 0.02 for each); the PCWP difference was significant only for percentage change at the 100-mg sitaxentan dose. No differences occurred for the SF-36 PF domain. CONCLUSIONS: Using CDF analysis incorporating the entire distribution of responses, sitaxentan significantly improved percent of predicted PVO2 (by numeric and percentage changes, 300-mg dose), 6MWD, PVR, SVR, CI, and PCWP (by percentage change, 100-mg dose), but not the SF-36 PF domain in PAH patients. PCV17 RESPONSE PROFILES OF SILDENAFIL CITRATE ON EXERCISE CAPACITY, HEMODYNAMIC FUNCTION, AND HEALTH-RELATED QUALITY OF LIFE IN CHILDREN WITH PULMONARY ARTERIAL HYPERTENSION Cappelleri JC1, Hwang LJ2, Mardekian J2, Mychaskiw MA2 1 Pfizer Inc, New London, CT, USA; 2Pfizer Inc, New York, NY, USA OBJECTIVES: Pulmonary arterial hypertension (PAH) is an ultimately fatal condition that may worsen hemodynamic parameters and contribute to diminished physical functioning and health-related quality of life (HRQoL). Although these relationships are well described in adults with PAH, they are not fully characterized in pediatric populations. METHODS: Children (aged 1–17 years) with PAH received 16 weeks of randomized, double-blind treatment with sildenafil citrate or placebo. The primary outcome measure, peak VO2 (PVO2), assessed cardiopulmonary exercise capacity in children developmentally capable of exercise. Secondary measures (for all patients) included mean pulmonary arterial pressure (mPAP), pulmonary vascular resistance index (PVRI), cardiac index (CI), and the Child Health Questionnaire-Parent Form (CHQ). Cumulative distribution functions (CDFs) plotted the cumulative percentage of patients against percentage changes (and also numeric changes) from baseline to week 16 in outcome measures. RESULTS: Of 234 randomized and treated children, 106 were developmentally able to exercise; 148 to 221 provided non-exercise endpoint data. Descriptive separation between CDFs was evident for sildenafil versus placebo for percentage changes in PVO2, mPAP, PVRI, and CI. For example, 58 of 77 (75%) sildenafil-treated children had increases in the percentage change in PVO2 versus 13 of 29 (45%) placebo-treated children. Descriptive separation between CDFs was not evident for CHQ domains of physical function, bodily pain, behavior, general health, and parental impact-time. Across all measures, an identical pattern of response profiles was obtained when numeric (rather than percentage) changes were assessed. CONCLUSIONS: Cumulative distribution functions, which incorporate the entire distribution of responses, can enhance clinical interpretation of outcome measures. The distribution of responses on PVO2, mPAP, PVRI, and CI for sildenafil was descriptively more favorable compared with placebo, suggesting that sildenafil improves exercise capacity and hemodynamic function in children with PAH. Further research should elucidate the impact of disease and treatment on HRQoL in pediatric PAH. PCV18 COMPARISON OF MIXED DYSLIPIDEMIAS AMONG HIGH CARDIOVASCULAR RISK PRIMARY CARE PATIENTS: RESULTS OF A POOLED ANALYSIS OF PRIMULA DATA FROM CANADA, SPAIN, SWEDEN, AND THE UNITED KINGDOM Ambegaonkar BM1, Bash LD1, Jameson K2, Marentette M3, Nocea G4, Pettersson B5, Sazonov V1 1 Merck & Co., Inc., Whitehouse Station, NJ, USA; 2MSD Ltd., Hoddesdon, UK; 3Merck Frost Canada Ltd, Kirkland, QC, Canada; 4Merck Sharpe & Dohme Ltd, Madrid, Spain; 5MSD, Sollentuna, Sweden OBJECTIVES: Dyslipidemias involving elevated low-density lipoprotein cholesterol (LDL-C), low high-density lipoprotein cholesterol (HDL-C), and/or high triglycerides are correlates of cardiovascular disease (CVD). The degree to which risk persists after lipid modifying therapy (LMT), and varies between high risk populations, has not been well explored. We compare the prevalence of dyslipidemias§ before and after LMT among high CVD risk patients in clinical practice. METHODS: This is a pooled analysis of four observational, longitudinal, studies (PRIMULA) of primary care patients in Canada, Spain, Sweden, and the United Kingdom, ≥35 years old who initiated LMT and continued therapy for >12 months. Data are reported for a total of 12,768 high cardiovascular risk patients, more than half of whom had diabetes mellitus (DM), nearly a third with a history of CVD, and the remainder with a 10-year Framingham risk score (FRS) >20%. RESULTS: The majority (>95%) received statin monotherapy as their initial LMT. At baseline, those with a history of CVD had the lowest, and those with FRS > 20% the greatest prevalence of elevated LDL-C, elevated TG, low HDL-C and elevated LDL-C coupled with low HDL-C and/or elevated TG. After 12 months of follow-up, all high risk groups demonstrated improved lipid profiles, though only among those with FRS > 20% did HDL-C improve. Persistent dyslipidemia was most rampant among those with a FRS > 20% and least among A344 those with a history of CVD. CONCLUSIONS: In each subgroup, there was a substantial improvement in LDL-C, while improvement of TG and HDL-C levels were moderate and negligible, respectively. Despite LMT, both single and mixed dyslipidemias were prevalent among high risk patients, particularly among those with DM and even more so among those with FRS > 20%. While the prevalence and persistence of dyslipidemia varied between groups, observations suggest that all may benefit from other types of LMT in addition to statins. PCV19 EFFECTIVENESS OF ANTIHYPERTENSIVE AGENTS IN THE SECONDARY PREVENTION OF VASCULAR EVENTS AMONG PATIENTS WITH ISCHEMIC STROKE Perreault S1, Cote R2, Dragomir A1 1 Université de Montréal, Montréal, QC, Canada; 2McGill University, Montreal, QC, Canada OBJECTIVES: Antihypertensive agents (AH) have been shown to reduce the risk of major cardiovascular (CVD) events. However, there is no large scale effectiveness studies which have assessed the relationship between adherence to AH medications and major CVD outcomes in high risk individuals that have suffered an ischemic stroke. The aim of the study was to evaluate the relationship between AH drug adherence and vascular outcomes in a cohort of older patients hospitalized for an ischemic stroke and discharged in the community. METHODS: A cohort of 14,227 patients with ischemic stroke was reconstructed from RAMQ and Med-Echo databases. Eligible subjects were 65 years and older and treated with AH agents between 1999 and 2007. A nested case-control design was used to study major CVD outcomes. Every case was matched for age and duration of follow-up. The adherence to AH drugs was measured as the proportion of days’ supply of medication dispensed over a defined period. Conditional logistic regression models were used to estimate the rate ratio of vascular events adjusting for covariables. RESULTS: Mean patient age was 75 years, 54% were male, 23% had diabetes, and 47% had dyslipidemia, 38% had CAD, 6% MI and 14% had atrial fibrillation. Adherence to AH agents ≥80% reduced the risk of vascular events (RR: 0.70; 0.64–0.77) compared to the one of <80%. Male gender, prior CVD, and non adherence to therapies for diabetes and dyslipidemia were risk factors. CONCLUSIONS: Higher adherence to AH therapy is linked with a risk reduction of vascular events among patients with ischemic stroke. PCV20 COMPLIANCE WITH ANTIHYPERTENSIVE AGENTS AND THE ONSET OF END-STAGE RENAL DISEASE Perreault S, Roy L, Lessard M, Dragomir A Université de Montréal, Montréal, QC, Canada OBJECTIVES: The correlation between severity of hypertension and risk of end-stage renal disease (ESRD) is well known. However, the impact of antihypertensive drug adherence on primary prevention of chronic kidney disease (CKD) has never been assessed. Our objective was to evaluate the impact of better adherence to antihypertensive (AH) therapy on ESRD. METHODS: A cohort of 208,128 patients was reconstructed using RAMQ and MedEcho databases. Patients were eligible if they were between 45 to 85 years of age, had a new diagnosis of hypertension and were newly treated with AH drug between 1999 and 2007. a nested case-control design was used to study the occurrence of ESRD. Every case of ESRD was matched for age and duration of follow-up. Adherence level was assessed as a medication possession ratio. Conditional logistic regression models were used to estimate the rate ratio of ESRD adjusting for several covariables. RESULTS: Patients were at 65 years old and 42% male. The mean high adherence level (80%) to AH therapy was 90%. We identified 1026 cases with ESRD during follow-up. High adherence level (80%) to AH therapy compared to lower adherence level (<80%) was associated with a reduction of ESRD (RR: 0.81; 0.70–0.95). Risk factors for ESRD were CKD, gout, diabetes, coronary artery disease, chronic heart failure and peripheral vascular disease. CONCLUSIONS: The study suggests that better adherence to AH therapy is associated with risk reduction of new onset ESRD in hypertensive population. PCV21 THE RELATIVE EFFECTIVENESS PROFILE OF DRONEDARONE USING THE NUMBER NEEDED TO TREAT (NNT) APPROACH de Sauvebeuf C1, Chicoye A1, Hohnloser S2 1 IMS Health, Puteaux, France; 2J.W. Goethe University Hospital, Frankfurt, Germany Dronedarone is a new antiarrhythmic drug developed for the treatment of atrial fibrillation (AF). While numerous large cardiovascular (CV) clinical trials have assessed the efficacy of different compounds on morbidity and mortality endpoints, dronedarone is the first antiarrhythmic drug for AF for which a large randomized clinical trial (ATHENA) was designed to assess morbidity/mortality endpoints. OBJECTIVES: To perform an effectiveness analysis using the Number Needed to Treat (NNT) approach to compare dronedarone with other CV drugs in avoiding major CV events (CV death, CV hospitalization and stroke). METHODS: A literature search for 3 CV active medications (statins, angiotensin-converting enzyme inhibitors, angiotensin receptor blockers) was performed using PubMed from 1995 to present to identify clinical trials with endpoints comparable to those of ATHENA. NNTs were calculated using hazard ratios; different lengths of follow up were adjusted to one year, dividing each rate of events by the study duration. RESULTS: Twenty clinical trials were identified as having at least one endpoint comparable to ATHENA. Based on the ATHENA data, the NNT per year of treatment with dronedarone to avoid one “CV death”, one “CV hospitalization” and one stroke were 157, 11, and 172, respectively. Dronedarone was associated with a lower NNT to avoid a “CV death” compared to ramipril 13th Euro Abstracts (HOPE, NNT = 239), pravastatin (LIPID, NNT = 255), and simvastatin (HPS, NNT = 325; 4S, NNT = 167), but a higher NNT vs. candesartan in selected chronic heart failure patients (CHARM, NNT = 123). NNTs for dronedarone to avoid one “CV hospitalization” and one stroke compared favorably to the other medications evaluated. CONCLUSIONS: Pressure on health care budgets emphasizes the need to demonstrate medical value, notably through NNTs which enable comparisons across interventions. Dronedarone has morbidity/mortality data (“CV death”, “CV hospitalizations” and stroke) that compares favorably in terms of NNT to other CV therapies within their respective trial patient populations. PCV22 THE INFLUENCE OF METEOROLOGICAL FACTORS ON CEREBRAL INFARCTION, INTRACEREBRAL HEMORRHAGE, SUBARACHNOID HEMORRHAGE AND TRANSIENT ISCHEMIC ATTACK Kriszbacher I1, Csoboth I1, Fülöp A2, Boncz I1, Köhalmi A1, Müller Á1, Bódis J1 1 University of Pécs, Pécs, Hungary; 2National Meteorological Service, Budapest, Hungary OBJECTIVES: We have investigated, whether the time of onset of an acute cerebrovascular event demonstrates a seasonal variation, and whether it is influenced by meteorological factors. METHODS: We examined patients admitted to Neurology Departments in Hungary between 2005 and 2007 with the diagnose of cerebral infarction, intracerebral hemorrhage, subarachnoid hemorrhage or transient ischemic attack (n = 178,092). Data was collected from the database of the Hungarian National Health Insurance Fund based on the International Classification of Diseases. Meteorological data was retrieved from the National Meteorology Service. RESULTS: Meteorological analysis showed, that an increase in average temperature on the previous day resulted in a significant drop of intracerebral hemorrhage incidence during Spring, Summer and Autumn (P < 0.01), while in case of cerebral infarction such decrease only occurred during Summer and Winter (P < 0.01), and for transient ischemic attack, only during Summer (P < 0.01).Examining atmospheric pressure, we found variations in case of cerebral infarction, intracerebral hemorrhage and transient ischemic attack. Morbidity rates of cerebral infarction increased during Summer, and decreased during Winter, whenever average atmospheric pressure values on the preceding day were higher (P < 0.01). Decrease in morbidity rates was observed for intracerebral hemorrhage and transient ischemic attack (P < 0.01). Considering relative humidity on the previous day, a marked variation showed for cerebral infarction during Spring (P < 0.01), and for intracerebral hemorrhage during Summer (P < 0.01). No relationship was found with subarachnoid hemorrhage. CONCLUSIONS: Results reveal, that the occurrence cerebrovascular events shows typical variations depending on the season of the year, while certain meteorological factors influence the development these also. PCV23 DISCONTINUATION OF LOW-DOSE ACETYLSALICYLIC ACID TREATMENT FOR SECONDARY PREVENTION OF CARDIOVASCULAR OUTCOMES: INCIDENCE AND PREDICTORS García Rodríguez LA1, Martín-Merino E1, Johansson S2 1 Spanish Centre for Pharmacoepidemiological Research (CEIFE), Madrid, Spain; 2AstraZeneca R&D, Mölndal, Sweden OBJECTIVES: To assess what proportion of patients treated with low-dose acetylsalicylic acid (ASA) for secondary prevention of cardiovascular events discontinue this treatment, and to identify risk factors for discontinuation. METHODS: The Health Improvement Network UK primary care database was used to identify individuals aged 50–84 years with ≥2 prescriptions of low-dose ASA (75–300 mg/day) in 2000– 2007 (n = 35,639). The study cohort was followed from the first day after the initial prescription until the earliest occurrence of one of the following: ASA discontinuation (a period of ≥90 days after the last prescription would have been used up [assuming full compliance], with no refill of the prescription during this time); death; diagnosis of an alcohol-related condition or cancer; or the end of the study period. The mean follow-up time was 2.5 years. RESULTS: Almost one-third of patients discontinued low-dose ASA (n = 11,729; incidence: 13.1 per 100 person-years; 95% confidence interval [CI]: 12.9–13.4). The incidence of discontinuation was higher in the first year of follow-up (26.7 per 100 person-years; 95% CI: 26.1–27.3) than the rest of the study period (6.8 per 100 person-years; 95% CI: 6.6–7.0). The risk of discontinuation was 42–67% higher in patients with an initial indication of unstable angina, ischemic heart disease or cerebrovascular disease than those with an initial indication of myocardial infarction. Current use of proton pump inhibitors (PPIs) was associated with a significant reduction in the risk of discontinuation (odds ratio [OR]: 0.92; 95% CI: 0.87–0.98; compared with no PPI use). Individuals taking PPIs from the same time as their ASA treatment were at particularly low risk of ASA discontinuation (OR: 0.74; 95% CI: 0.69–0.79). CONCLUSIONS: Discontinuation of low-dose ASA treatment is common, especially in the first year of treatment. Concomitant PPI use reduces the risk of ASA discontinuation. PCV24 AN EPIDEMIOLOGICAL EVALUATION OF THE IMPACT OF PERCUTANEOUS CORONARY INTERVENTIONS ON THE LENGTH OF STAY AND MORTALITY OF PATIENTS HOSPITALIZED WITH ACUTE CORONARY SYNDROMES Chevalier P, Lamotte M IMS Health Consulting, Brussels, Belgium OBJECTIVES: Randomized clinical trial comparing percutaneous coronary interventions (PCI) and non-invasive treatment acute coronary syndromes mostly favour the 13th Euro Abstracts invasive approach. The aim of this study was to assess whether in a real life setting performing acute PCI would have an impact on outcomes as length of stay (LOS) and mortality in patients hospitalized for acute coronary syndromes in Belgium. METHODS: The average length of stay (LOS) and mortality among hospitalized patients with acute coronary syndrome were estimated using the longitudinal IMS Hospital Disease Database (year 2007), including data on 34.3% of Belgian hospital beds. Stays were identified based on ICD-9 coding and split in ST-elevated Myocardial infarction (STEMI ICD-9: 410 excluding 410.7), non-STEMI (ICD-9: 410.7-411.89) and unstable angina (ICD-9: 411.1-411.8-413.0). Invasive procedures were identified with ICD-9 codes corresponding to PCI (36.0). First stays in the calendar year were examined. Comparisons were performed using a Wilcoxon non-parametrical test for LOS and a Chi-square for mortality rates. RESULTS: In first hospitalizations, 3,420 STEMI, 2,070 non-STEMI and 954 unstable anginas were retrieved from the database, with respectively 1,206, 504 and 28 of them being treated invasively. LOS of patients undergoing PCI was significantly lower in STEMI (6.0 vs. 9.5 days; P < 0.001) and non-STEMI (5.1 vs. 9.5; P < 0.001). Mortality in patients with PCI was lower in both STEMI (3.5% vs. 17.2%; P < 0.001) and non-STEMI (0.8% vs. 6.2%; P < 0.001). Although LOS and mortality in non-invasively treated patients were lower in hospitals without a cathlab compared to hospitals with a cathlab, both parameters were significantly higher compared to PCI treated patients. CONCLUSIONS: Although information on the baseline characteristics of the different patients is limited, the findings of this observational study seem to support randomized clinical trials. Treatment with PCI decreases the LOS and the mortality significantly in patients with a myocardial infarction. A345 was respectively equal to 51 and 75 within the 2 subgroups, resulting in an incidence of 0.54% in patients with hip replacement and 1.07% in patients with knee replacement. LOS of patients with a DVT/PE episode was more than twice as high after both hip (34.2 vs. 15.9 days; P < 0.001) and knee (24.9 vs. 12.0 days; P < 0.001) replacement. CONCLUSIONS: The incidences of symptomatic DVT and PE reported in clinical trials could be confirmed based on this retrospective search. The occurrence of DVT/PE doubles LOS in patients undergoing hip or knee surgery. PCV27 INCREASING PREVALENCE OF HYPERTENSION IN NIGERIA: A SYSTEMATIC REVIEW FROM 1990 TO 2009 Ekwunife O, Aguwa CN University of Nigeria, Nsukka, Enugu, Nigeria OBJECTIVES: This study aimed at synthesizing population-based studies on prevalence of hypertension in Nigeria from 1990 to 2009. METHODS: Cochrane library and PubMed were searched for similar reviews using filters for systematic review. General databases as well as subject-specific databases were searched for primary studies. Manuscripts of appropriate studies were retrieved. Quality assessment scale was developed to assess the retrieved manuscript and only eligible studies were selected. Formal meta-analysis was not conducted due to heterogeneity of data. RESULTS: A total of 55 abstracts were identified, out of which 10 full manuscripts were retrieved and used for systematic review. There was an increasing trend in prevalence of hypertension. Hypertension prevalence in rural areas increased from 18.7% in 2003 to 22.4% in 2007. In urban areas, it increased from 23.4% in 1997 to 27.1% in 2007. In studies that combined both rural and urban populations, prevalence of hypertension increased from 14.5% in 1994 to 34.8% in 2003. There was gender difference with respect to prevalence of hypertension as males had higher prevalence of hypertension compare to females. CONCLUSIONS: Prevalence of hypertension seems to be on the increase in Nigeria. There is a need to develop strategies to prevent, treat, and control hypertension effectively in Nigeria. PCV25 RISK OF STROKE AND PREVENTATIVE STEPS AMONG PATIENTS DIAGNOSED WITH ATRIAL FIBRILLATION IN EUROPE Gupta S1, Goren A2, Freedman D1 1 Kantar Health, Princeton, NJ, USA; 2Kantar Health, New York, NY, USA OBJECTIVES: This study examined stroke risk and prevention among European patients diagnosed with atrial fibrillation (AF). METHODS: Data were extracted from the European 2008 National Health and Wellness Survey, an annual Internet survey of self-reported health care attitudes and behaviors among 52,524 adults in Germany, Spain, Italy, the UK, and France. Stroke risk was assessed with CHADS2, an index summing the presence of congestive heart failure, hypertension, age of 75 years+, diabetes mellitus, and history of prior stroke or transient ischemic attack (the latter weighted twice). Low- (CHADS2 = 0), moderate- (1), and high- (2+) risk patients reported on what steps, if any, they took to prevent stroke: low fat or low sodium diet, regular exercise, smoking cessation, weight loss, blood pressure or cholesterol reduction, and use of baby aspirin, aggrenox, coumadin/warfarin, plavix, or other medications. RESULTS: Among 508 respondents diagnosed with AF (EU prevalence of 0.97%), 198 (39%) were low, 158 (31.1%) moderate, and 152 (29.9%) high in risk for stroke. Significant differences emerged in the use of any preventative steps among low- (23.7%), moderate- (35.4%), and high- (57.2%) risk patients, P < 0.05. High-risk patients were significantly more likely than the low-risk group to take every preventative step except exercise—16.2% (low), 17.7% (moderate), 20.4% (high)— and aspirin use—11.1% (low), 13.3% (moderate), 15.8% (high). For example, coumadin/warfarin use was higher among high- (21.1%) vs. moderate- (8.9%) and low-risk (3.0%) patients, P < 0.05. CONCLUSIONS: In the EU, 42.8% of diagnosed AF patients at high risk for stroke took no preventative steps, and no significant differences emerged between risk groups on regular exercise and aspirin use. Only 1 in 5 high-risk patients took coumadin/warafin, while many more could benefit from prophylactic anticoagulation therapy. Higher risk correlated with higher prevention, but there remains an unmet need for increased targeted treatment of high-risk AF patients. PCV28 CARDIOVASCULAR RISK FACTORS AMONG ASIAN INDIAN AND WHITE ADULTS IN THE UNITED STATES Mcdonald M, Zhou J, Rubinstein E, Mardekian J Pfizer Inc, New York, NY, USA OBJECTIVES: Current prevalence of cardiovascular risk factors are lacking among Asian Indian adults. This study contributes to our knowledge by examining current national estimates for hypertension, high cholesterol, diabetes and obesity among Asian Indian and white adults aged 20 years and older, by gender and age group. METHODS: Cross-sectional observational study design. Analysis of adults 20 years and older surveyed in the National Health Interview Surveys (NHIS 2006–2008) (Asian Indian, n = 707, white, n = 51,769). RESULTS: The age-standardized prevalence of hypertension among Asian Indian adults is 21%, significantly lower than the 27% prevalence among white adults. At 27%, high cholesterol is equally prevalent in both populations. Diabetes is more prevalent among Asian Indian adults (15% vs. 7%). Asian Indian men have more than twice the rate of diabetes as white men, (17% vs.7%), and Asian Indians aged 65 years and older have a higher prevalence of diabetes than older, white adults (28% vs. 17%). Obesity rates for Asian Indian adults aged 20–39, 40–64, and 65+ are 10%, 16%, and 16%, respectively. Obesity rates for white adults aged 20–39, 40–64, and 65+ are 26%, 34%, and 27%, respectively (p < 0.0001 for all but 65+ obesity differences). CONCLUSIONS: The age-standardized prevalence rate of high cholesterol is similar among Asian Indian and white adults. Hypertension rates are lower among Asian Indians. The prevalence of diabetes is higher in the Asian Indian population, disproportionately affecting men and older adults. Targeted approaches for diabetes disease treatment and reduction are needed. PCV26 AN EPIDEMIOLOGICAL EVALUATION OF THE INCIDENCE OF DEEP VENOUS THROMBOSIS AND PULMONARY EMBOLISM IN PATIENTS WITH HIP OR KNEE REPLACEMENT SURGERY AND OF ITS IMPACT ON THE AVERAGE LENGTH OF STAY Chevalier P, Lamotte M IMS Health Consulting, Brussels, Belgium OBJECTIVES: Recent clinical trials on the new anti-thrombotic agents have shown that the incidence of symptomatic deep venous thrombosis (DVT) and pulmonary embolism (PE) in patients undergoing hip or knee replacement ranges between 0.5 and 2%. This study aimed at assessing the real life incidence of DVT/PE after major orthopaedic surgery in Belgian and the impact of these complications on the length of stay based on a hospital disease database. METHODS: The incidence of DVT/PE and the average length of stay (LOS) among patients hospitalized for hip or knee replacement surgery were estimated using the longitudinal IMS Hospital Disease Database (year 2007), including data on 34.3% of Belgian hospital beds. Stays were searched based on ICD-9-CM codes corresponding to hip replacement (81.51-81.5281.53) and knee replacement (81.54-81.55). Occurrence of DVT/PE was identified with ICD-9 codes 451.1-451.2-453.4. It was also checked whether selected patients were administered low-molecular-weight heparin (LMWH). The impact of a DVT/PE complication on LOS was assessed through a Wilcoxon non-parametrical test. RESULTS: A total of 9377 stays with hip replacement and 6,978 stays with knee replacement were retrieved in the database. More than 99.9% of the patients were administered LMWH in both subgroups. The number of stays with a DVT/PE episode PCV29 PREVALENCE, CHARACTERISTICS, AND IN-HOSPITAL OUTCOMES OF METABOLIC SYNDROME IN ACUTE CORONARY DISEASE PATIENTS IN OMAN Al-Zakwani IS1, Al-Rasadi K1, Sulaiman K2, Panduranga P2 1 Sultan Qaboos University, Al-Khoudh, Oman; 2Royal Hospital, Muscat, Oman OBJECTIVES: To evaluate the prevalence, characteristics, and in-hospital outcomes of metabolic syndrome (MS) in acute coronary syndrome (ACS) patients in Oman. METHODS: Data were analyzed from 1,392 consecutive patients admitted to 15 hospitals throughout Oman with the final diagnosis of ACS during May 8, 2006 to June 6, 2006 and January 29, 2007 to June 29, 2007, as part of Gulf RACE (Registry of Acute Coronary Events). MS was defined, as stipulated by the recent ATP-III guidelines, as having 3 or more of the following abnormalities: waist circumference of ≥94 cm for men and ≥80 cm for women, high triglyceride levels (of ≥150 mg/dL (1.7 mmol/L) or drug treatment), low high density lipoprotein cholesterol (HDL-C) levels (of <40 mg/dL (1.0 mmol/L) for men and <50 mg/dL (1.3 mmol/L) for females or drug treatment), blood pressure (of ≥130 mmHg for systolic and/or ≥85 mmHg for diastolic or drug treatment), and blood sugar (of ≥100 mg/dL (≥5.6 mmol/L) or drug treatment). RESULTS: The overall prevalence rate of MS in ACS patients in Oman was 66% with females being the majority (80% versus 57%; P < 0.001). MS was associated with higher median body mass index scores (27 versus 24; P < 0.001), diabetics (45% versus 19%; P < 0.001), hyperlipidemia (40% versus 23%; P < 0.001), hypertension (62% versus 34%; P < 0.001), renal impairment (9.3% versus 3.4%; P < 0.001), killip score >II (13% versus 8%; p = 0.004) and non-ST segment elevation A346 myocardial infarction (78% versus 68%; P < 0.001). Even after multivariate adjustment, MS was still associated with higher risk of in-hospital heart failure (odds ratio (OR), 1.37; 95% CI: 1.03–1.81; p = 0.028) and mortality (OR, 4.42; 95% CI: 1.25–15.5; p = 0.020). CONCLUSIONS: Prevalence of MS in ACS patients in Oman is high and seen more in females than in males. Furthermore, MS was associated with higher in-hospital heart failure and mortality. PCV30 PREVALENCE AND CONTROL OF TRADITIONAL CARDIOVASCULAR RISK FACTORS AND ANTICIPATED AVOIDABLE CORONARY MORTALITY IN PRIMARY PREVENTION IN EUROPE: THE EURIKA STUDY Banegas JR1, Guallar E2, Borghi C3, Dallongeville J4, De Backer G5, Halcox JP6, Massó-González EL7, Perk J8, Steg PG9, Rodríguez-Artalejo F1 1 Universidad Autónoma de Madrid, Madrid, Spain; 2Welch Center for Prevention, Epidemiology, and Clinical Research, Baltimore, MD, USA; 3Policlinico Universitario Sant’Orsola, Bologna, Italy; 4Institut Pasteur de Lille, Lille, France; 5University of Gent, Gent, Belgium; 6Cardiff University, Cardiff, UK; 7AstraZeneca Farmacéutica Spain S.A, Madrid, Spain; 8 Oskarshamns Hospital, Oskarshamn, Sweden; 9Université Paris VII—Denis Diderot, Paris, France OBJECTIVES: Despite the availability of the ESC European guidelines on cardiovascular prevention, considerable cardiovascular mortality remains throughout Europe. The European Study on Cardiovascular Risk Prevention and Management in Daily Practice (EURIKA) (NCT00882336) investigated the prevalence and degree of control of main cardiovascular risk factors in primary prevention of cardiovascular disease (CVD). METHODS: EURIKA was a cross-sectional study conducted simultaneously in 12 European countries (n = 809 primary care and specialist physicians). Patients aged ≥50 years who were free of clinical CVD, but had at least one risk factor as defined by the 2007 European guidelines on cardiovascular prevention, were eligible for inclusion. Data recorded included smoking status, body mass index, cholesterol levels, blood pressure and presence of diabetes. Cardiovascular risk was assessed by Systematic COronary Risk Evaluation (SCORE) methods. Attributable coronary mortality was calculated based on our prevalence estimates together with hazard ratios for CVD-related mortality (from the Third National Health and Nutrition Examination Survey). RESULTS: In total, 7641 patients (mean age: 65 years; 48% male) were evaluated. Of those aged <65 years, 27% had a SCORE-based absolute risk of CVD of ≥5%. The prevalence of risk factors was 21.3% for current smoking, 71.9% for hypertension, 55.4% for dyslipidemia, and 26.6% for diabetes. Control of risk factors among patients receiving therapy was 49% (between country range: 40–61%) for hypertension, 48% (26–74%) for dyslipidaemia and 40% (26–54%) for diabetes. The adjusted excess mortality attributable to risk factors was 17% (14–25%) for current smoking, 23% (21–25%) for hypertension, 26% (17–31%) for dyslipidaemia, and 30% (20–37%) for diabetes. CONCLUSIONS: The prevalence of traditional cardiovascular risk factors is high and their control is suboptimal. Traditional risk factors such as hypertension, dyslipidaemia and diabetes are responsible for a large fraction of the estimated coronary deaths that could be avoided through primary prevention. PCV31 BLOOD PRESSURE CONTROL AND ANTIHYPERTENSIVE STRATEGY DIFFERENCES ACCORDING TO PATIENTS AGE Font B1, Galera J1, Lahoz R1, Muñoz G1, Sierra C2, Doménech M2, Coca A2 1 Novartis Pharma, Barcelona, Spain; 2Hospital Clínic de Barcelona, Barcelona, Spain OBJECTIVES: This study aims to determine the degree of blood pressure (BP) control and the differences in therapeutic strategies adopted by physicians based on patients’ age and/or their functional status. METHODS: Multicenter and cross-sectional study that included patients with essential hypertension attending primary or specialist care, with at least one year of evolution. The study included 6453 patients, stratified into three main groups: a (<65 years) = 2184; B (65–79 years) = 2079; and C (≥80 years) = 2079. RESULTS: 51.3% of patients were male and mean (SD) age was 55.1 ± 7.5 years (group A), 71.4 ± 4.0 (B) and 83.7 ± 3.2 (C). 49.1% of patients were overweight. Mean systolic BP values were 141.3 ± 15.6 (A), 142.0 ± 16.3 (B) and 142.3 ± 16.8 (C) mmHg, and mean diastolic BP values were 84.5 ± 10.1 (A), 81.5 ± 10.5, 79.9 ± 11.2 (C) mmHg. 33.5% of patients had type 2 diabetes mellitus and 59.1% dyslipidemia. The existence of previous cardiovascular (CVD) or renal disease has been established in 38.9% of cases. The prevalence of CVD was 22.6% (A), 39.1% (B) and 55.9% (C). Target organ damage (TOD) was 26.7% (A), 40.8% (B) and 57.8% (C). The Barthel index that measures a person’s daily functioning reflected that the oldest group had a significantly higher level of moderate or severe dependence (P < 0.0001). Patients with controlled BP were 29.4% (A), 25.5% (B) and 23.7% (C) (P < 0.0001). Clinical inertia—the failure to intensify therapy in patients who do not achieve the clinical objectives—reached 60.3% (A), 61.7% (B) and 66.2% (C) (p = 0.0002), respectively. Global treatment compliance was 94.3%, and the difficulty for taking the medication significantly increased with patients’ age (P < 0.0001). CONCLUSIONS: Older hypertensive patients have a poorer BP control but an increased clinical inertia. These patients have higher prevalence of TOD and CVD, worse functional status and worse treatment compliance. 13th Euro Abstracts CARDIOVASCULAR DISORDERS – Cost Studies PCV32 BUDGET IMPACT ANALYSIS OF PACLITAXEL DRUG ELUTING STENT (DES) FOR THE TREATMENT OF LOWER LIMB PERIPHERAL ARTERIAL DISEASE (PAD) IN FRANCE Lopes S1, De Cock E2 1 Cook Medical, Bjaeverskov, Denmark; 2United BioSource Corporation, Barcelona, Spain OBJECTIVES: The self-expandable paclitaxel DES represents a major development in endovascular treatments for lower limb PAD. Clinical data show improved clinical outcomes compared to bare metal stents (BMS). This budget impact analysis assessed the impact of introducing reimbursement for a paclitaxel DES in France. METHODS: An Excel-based model was developed to estimate the impact of a transition from BMS to DES over a 5-year horizon (15% in 2011 to 35% in 2015). Hospital episode statistics were used to estimate the 2011–2015 patient population. The analysis was conducted from the payer perspective and only direct costs of procedures were considered, based on GHM 2009 tariffs (stenting and revascularization) and LPPR tariffs (BMS and grafts). The main outcome was target lesion revascularization (TLR) in the superficial femoral artery (SFA) after primary stent placement: angioplasty, re-stenting, or bypass surgery. TLR rates were based on the paclitaxel drug-eluting SFA stent registry study and on BMS TLR rates reported in the literature (Years 1 and 2: 6% and 9% for the DES; 16% and 22% for average BMS) and extrapolated for years 3 to 5. Net budget impact was expressed as the difference in cost between the scenario where the DES is progressively adopted versus the status quo (patients treated with BMS only). RESULTS: The base-case results show an incremental cost of c278,526 in year 1, which is more than offset by increasing cost savings in all subsequent years (year 2: −c35,901; year 5: −c510,025), resulting in a cumulative 5-year net budget impact of −c727,649. One-way sensitivity analyses on key inputs continued to show cumulative 5-year cost savings. CONCLUSIONS: Reimbursement and consequent adoption of the paclitaxel DES would result in cost savings for the French health care payer, despite requiring an initial investment. This is due to savings associated with fewer SFA revascularization events after the primary intervention. PCV33 ESTIMATING THE FINANCIAL IMPLICATIONS TO THE UK NHS OF INTRODUCING COREVALVE ALONGSIDE MEDICAL MANAGEMENT FOR THE TREATMENT OF SEVERE AORTIC STENOSIS Watt M, Mealing S, Sculpher M, Eaton JN Oxford Outcomes Ltd, Oxford, Oxon, UK OBJECTIVES: Aortic Stenosis (AS) is a severe cardiovascular condition and treatment often involves a major operation. For a subgroup of patients medical management (MM) is the only treatment option due to procedural risk. A transcatheter aortic valve implantation device “CoreValve,” is less invasive and allows for the implantation of a replacement valve in this patient group. However, CoreValve is more expensive than MM and introducing this technology into routine care would lead to an increase in National Health Service (NHS) spending. We estimated this budgetary impact using existing economic models. METHODS: In an Excel based Markov model CoreValve was compared to MM for inoperable patients. Parameters were derived from published literature. Costs were taken from the most recent published sources. Decrements were applied to age-specific EQ-5D population norms to generate QALYs. Incidence estimates were derived from information in a large national database. Projected five year uptake rates were provided by Medtronic. The outputs from the economic model were used for all relevant parameters. Case mix estimates were elicited from a clinical advisory board. RESULTS: Across all patients, the estimated incidence rate of AS was 165 per million, 40% of these patients were assumed to be inoperable, 66 per million. The uptake rate of CoreValve in the inoperable group rose linearly from 0% to 50% over 5 years. When all inoperable patients were treated with MM the total budgetary cost was £13,454,000. When 50% of patients are treated with CoreValve and 50% with MM, the total budgetary cost increases to £49,706,000. Thus, the total burden to the NHS of 50% of inoperable patients receiving CoreValve would be £36,251,000. With a more optimistic five year uptake rate (75%) this increases to £54,852,000. CONCLUSIONS: Introducing CoreValve would result in an additional £36,251,000 being spent by the UK NHS. PCV34 BUDGET IMPACT ANALYSIS OF PRIMARY TREATMENT OF HYPERTENSION WITH CANDESARTAN/HYDROCHLOROTHIAZIDE OR LOSARTAN/HYDROCHLOROTHIAZIDE IN THE MEXICAN SOCIAL SECURITY INSTITUTE Anaya P, López RJ, Polanco AC AstraZeneca, Naucalpan, Mexico OBJECTIVES: To calculate and compare costs of primary treatment of hypertension with candesartan/hydrochlorothiazide (HCT) or losartan/HCT in the Mexican Social Security Institute (IMSS). METHODS: An adaptation from Kjeldsen observational study of 14,100 patients diagnosed with hypertension and the costs’ sub analysis of Henriksson was made to have an approximation of incurred costs by IMSS when using candesartan/HCT or losartan/HCT in the primary hypertension treatment and its impact in reducing related cardiovascular events. First assumption for adaptation was to use only costs of combined therapies with HCT since monotherapies are not included in the IMSS formulary. Hospitalizations, laboratory tests and physician visits resource consumption reported by Kjeldsen were multiplied by IMSS 2010 unit costs. 13th Euro Abstracts An average of 3.35 years follow-up was used. IMSS population with hypertension was estimated using the national prevalence of 30.8% (20 years and older) reported by the National Health Survey of 2006, the proportion (70%) of the Mexican population over 20 years old reported by the National Population Council and the number of total IMSS affiliates. a 5% discount rate was applied to costs. Results are presented in US dollars (exchange rate 13 MXN/dollar). RESULTS: Primary hypertension annual treatment cost per patient including related cardiovascular events was $556.56 for patients treated with candesartan/HCT and $718.42 with losartan/HCT treatment. Theoretical savings yield by candesartan/HCT are $161.86 per patient. Calculated hypertension patients affiliated to the IMSS was 10 million. Average annual savings given this population are $1.707 million for the institution with 1.8 million bed days, 4.4 million physician visits and 5.3 million laboratory tests avoided. CONCLUSIONS: This analysis showed that the reduction of cardiovascular events with the use of candesartan/HCT for treating primary hypertension could save $1.707 million to the IMSS. PCV35 THE BUDGET IMPACT ANALYSIS OF AMBRISENTAN IN 2ND LINE TREATMENT OF ADULT PATIENTS WITH IDIOPATHIC, FAMILIAR OR ASSOCIATED WITH CONNECTIVE TISSUE DISEASE PULMONARY HYPERTENSION OF III NYHA STAGE Kawalec P1, Holko P2, Krzyzanowska A3, Glogowski C3 1 Jagiellonian University, Kraków, Poland; 2Centrum HTA, Krakow, Poland; 3GSK Commercial Sp. z o.o., Warsaw, Poland OBJECTIVES: To evaluate the costs of ambrisentan within therapeutic health program in 2nd phase line treatment of adult patients with idiopathic, familiar or associated with connective tissue disease pulmonary hypertension of III NYHA stage. METHODS: The analysis compared two scenarios: existing and new. The existing one assumed application of bosentan, iloprost, treprostinil or sildenafil in patients with pulmonary hypertension of III NYHA stage in case of 1st line treatment with bosentan or sildenafil failure. The new one included also administration of ambrisentan. Population’s abundance and market shares of drugs used in presented indication were estimated on the basis of PHPOL Registry. Cost data were collected from public payer’s perspective (National Health Fund) in Poland and calculated in five years horizon. Official ex factore price of Volibris® was provided by the producer. Cost analysis was performed under assumption that ambrisentan would be financed within the framework of Catalogue of active substances used in therapeutic health programs. a range of variables was checked in a sensitivity analysis. RESULTS: Introduction of ambrisentan to Catalogue of active substances used in therapeutic health programs and administration in 2nd line treatment of adult patients with idiopathic, familiar or associated with connective tissue disease pulmonary hypertension of III NYHA stage brought savings from public payer’s perspective amounted to: 178,782PLN (64,719PLN–342,497PLN) in first year, 385,054PLN (132,673PLN–769,277PLN) in second year, 614,006PLN (203,224PLN–1,266,372PLN) in third year, 646,137PLN (206,854PLN–1 366,336PLN) in fourth year and 674,573PLN (210,053PLN– 1,455,310PLN) in fifth following year. CONCLUSIONS: Treatment with ambrisentan leads to savings from public payer’s perspective and constitutes substantial therapeutic achievement (it is administered orally and only once a day unlike other drugs currently used in presented indication). PCV36 BUDGET IMPACT MODEL OF VENOUS THROMBOEMBOLISM PREVENTION AFTER TOTAL HIP AND KNEE REPLACEMENT Krysanov I1, Margieva A2, Omelyanovsky VV1, Malygina MA3 1 Research Center for Clinical and Economic Evaluation and Pharmacoeconomics, Moscow, Russia; 2Institute of Clinico-Economic Expertise and Pharmacoeconomics, RSMU, Moscow, Russia; 3Research institute of ambulance n.a. Sklifosovsky, Moscow, Russia OBJECTIVES: The purpose of this study was to perform comparative pharmacoeconomic analysis of antithrombotic therapy with rivaroxaban (Xarelto®) and enoxaparin (Clexane®) in the conditions of real clinical practice of total hip (THR) and knee (TKR) replacement. METHODS: The method of budget impact modeling was used. RESULTS: At equal duration of venous thromboembolism (VTE) prophylaxis at THR (35 days) rivaroxaban turned out to be more budget saving than enoxaparin by 2399 RUB (62.71c), reducing the quantity of complications cases (pulmonary embolism, deep venous thrombosis, death) by 26 cases per one thousand patients. At various duration of prophylaxis at THR—rivaroxaban (35 days) and enoxaparin (14 days)— rivaroxaban application has also demonstrated budget savings by 100 RUB(2.63c) thus reducing the quantity of complications by 74 cases per one thousand patients. The results of comparison of two weeks VTE prophylaxis after TKR have shown the most sizeable economy after rivaroxaban application by 7734 RUB(202.164c) and simultaneous complications decrease by 92 cases per one thousand patients. CONCLUSIONS: Rivaroxaban demonstrated budget savings compared to enoxaparin at all prophylaxis regiments at THR and TKR by reducing thrombosis complications. A347 PCV37 BURDEN IMPACT OF ATRIAL FIBRILLATION (AF BI) IN RUSSIAN FEDERATION Kolbin A1, Tatarsky B2, Biserova I2 1 Saint Petersburg State University, Saint Petersburg, Russia; 2Almazov Federal Heart, Blood and Endocrinology Centre, Saint Petersburg, Russia OBJECTIVES: AF is a big problem of the Russian Health Care System because cardiovascular diseases are the leaders in the structure of morbidity and mortality. Systemic statistics of AF and AF BI are unknown. Aim: To calculate the AF BI based on regional data and extrapolate the received data for Russia. METHODS: Evaluation of AF incidence and prevalence in dynamic (1995–2005) in the North-West region of Russia, calculation of direct and indirect costs and perspective modeling. RESULTS: As estimation—AF prevalence in Russia is 3.2/1000, incidence is 1,766/100,000, paroxysmal and persistent AF are increased from 0.6 to 1.8 and 0.25 to 0.6/1,000 accordingly. Hypertension (73%), ischemic hart disease (65.2%), diabetes (9%) occurred in the first episode of AF. Hospitalization rate is increased by 66%, estimated rate is 1.23 mln. pts./year, and it’s length is 6.9 days on the average, AF lethality is 1%. Disability rate is 40%, estimated amount of disability days is 3.3 mln/year. Calculated hospitalization cost for one year is 11.35 bln RUR (~c298 mln), cost for out-patients’ day-service—29.6 mln RUR (~c778 ths) and cost of out-patient treatment for all patients is no less than 32.5 bln RUR (c855 mln) including drugs cost—19.6 bln RUR (Euro). Estimated surgery cost is 260 mln RUR (c6.8 mln). Undirected adjusted cost is no less than 3 bln RUR (~c80 mln). CONCLUSIONS: AF BI in the Russian Federation leads to the biggest governmental, insurance and personal expenditures as compared to the developed countries. Annual BI AF without deficiancial GDP per person is no less than c5.5 ths. Decreasing of AF BI can be particularly achieved in case of a better pharmaceutical control of AF aimed at saving budget charges due to the less hospitalization and surgeon rates. PCV38 CLINICAL AND ECONOMICAL BENEFITS OF THE USE OF REMOTE MONITORING WITH CARELINK® IN CENTRO MÉDICO NACIONAL “LA RAZA”, IMSS MEXICO Guevara M1, Cerezo O2, Quiroz ME3, Machado F4, Busca R5 1 National Medical Center (Instituto Mexicano del seguro Social) La Raza, Mexico City, D.F., Mexico; 2Oncology National Institute, Mexico City, D.F., Mexico; 3Medtronic Mexico, Mexico City, D.F., Mexico; 4Medtronic LA, Miami, FL, USA; 5Medtronic International Trading Sàrl, Tolochenaz, Switzerland OBJECTIVES: The majority of cardiac device recipients are routinely followed. Implantable Cardioverter Defibrillators (ICDs) devices must be systematically and continually monitored, with follow-up frequency adjusted for the patient’s underlying medical condition, device-related issues, and patient preferences. In México, in CMNIMSS La Raza, the current implant rate for ICD is 5 per million (equating to 30 new implants per year) and every year an average of 180 patients implanted with ICD are attending follow-up visits at La Raza hospital in Mexico. This simulation provides evidence to support the use of remote follow up and monitoring of ICDs via the CareLink Network as an alternative to follow-up consultations in attending an outpatient clinic setting. METHODS: The simulation considers patients implanted with ICD (without CareLink) will on average require 4 in-clinic follow-up consultations per year whilst those with CareLink will only require 1 in-clinic follow-up consultation (4–6 weeks post implant) and 3 remote follow-up consultations. Reviewing patient data provided via the CareLink Network (remote monitoring consultation) requires an average 8.4 minutes of cardiac physiologist time compared to 25.8 minutes for an in–clinic follow-up consultation. RESULTS: For each patient per year the introduction of a remote monitor to support device follow-ups, lead to a reduction in cardiac physiologist time of 52.2 minutes equivalent to $17,970 MXP ($30,686 MXP without CareLink versus $12,716 MXP on CareLink). The cost of a cardiac physiologist time is $2630.00 MXP (per hour) equating to MXP 368.20 per remote follow-up versus $1130.90 MXP per in-clinic follow-up. CONCLUSIONS: The use of CareLink can potentially generate economic benefits for the health professional, institution, health system and the patient. PCV39 CLINICAL OUTCOMES AND COSTS OF A POCKET-SIZED ULTRASOUND DEVICE FOR SCREENING FOR ABDOMINAL AORTIC ANEURYSM IN ITALY Patel PA1, Shah M2, D’Souza AO3, Min JK4 1 GE Healthcare, Barrington, IL, USA; 2Xcenda, Palm Harbor, FL, USA; 3Xcenda, Cincinnati, OH, USA; 4Weill Medical College of Cornell University, New York, NY, USA OBJECTIVES: To develop a model to assess the clinical and economic impact of a pocket-sized ultrasound (PSU) versus standard ultrasound for abdominal aortic aneurysm (AAA) screening in Italy. METHODS: The model was developed from the perspective of the Servizio Sanitario Nazionale, the national health care system in Italy, during an 8-year time horizon for 65–74-year-old males who had smoked in their lifetime. Our model assumed 100% of eligible males could be screened with the PSU, as compared to 29% actually screened with standard ultrasound, and that screening with a PSU would incur no additional cost. Model inputs were derived from major clinical trials (4-year rates of mortality, aortic rupture, and elective and emergency surgeries) and DRG reimbursement rates (costs). Model outputs of number of deaths, ruptures and costs (2010 euros) were calculated. One-way sensitivity analyses were conducted. RESULTS: A PSU strategy yielded a 33.8% (0.32% vs. 0.48%) and 46.6% A348 (0.42% vs. 0.79%) reduction in rates of AAA-related mortality and rupture, respectively, as compared to the standard ultrasound strategy. During an 8-year follow-up period, the higher cost of elective surgeries by the PSU strategy was partially offset by lower rates and costs of emergency surgeries compared to the standard ultrasound strategy. The higher screening rate with the PSU strategy resulted in a cost savings of c17 per screened eligible male over 8 years (annualized c2/year). Sensitivity analyses demonstrated reductions in deaths and aortic ruptures ranging from 17% to 50% and 2% to 73%, respectively, at a maximum cost-savings of c7 to c50, and only one instance of an incremental cost of c3 per eligible patient over 8 years. CONCLUSIONS: Use of a PSU device to screen for AAA may reduce deaths, aortic ruptures, and surgery costs with a savings of c2/year per screened eligible male. PCV40 CLINICAL AND ECONOMIC IMPACT OF USE OF A POCKET-SIZED ULTRASOUND DEVICE FOR HEART FAILURE HOSPITALIZATIONS IN ITALY Patel PA1, Shah M2, D’Souza AO3, Min JK4 1 GE Healthcare, Barrington, IL, USA; 2Xcenda, Palm Harbor, FL, USA; 3Xcenda, Cincinnati, OH, USA; 4Weill Medical College of Cornell University, New York, NY, USA OBJECTIVES: Patients hospitalized for Heart Failure (HF) continue to have high re-admission and mortality rates in the 30 days following discharge. Cardiac parameters such as left ventricular function (LVF) and inferior vena cava (IVC) size are predictors of mortality and readmission, and are presently assessed by 2-dimensional echocardiography (2DE). Pocket-sized ultrasound (PSU) devices for LVF and IVC assessment offer potential advantages of lower cost, increased availability, and portability as compared to 2DE. The study objective was to develop a model to assess the clinical and economic impact of PSU versus 2DE for HF hospitalizations in Italy. METHODS: The model was developed from an Italian hospital perspective, and timeframe was during HF hospitalization to 30 days post-discharge. The clinical scenario evaluated the impact of LVF assessment on mortality, while the economic scenario incorporated the impact of IVC size on costs (2010 Euros). Our model assumed 100% of PSU patients would undergo evaluation, as compared to only 68% who currently undergo 2DE. We assumed abnormalities identified by PSU would require subsequent 2DE evaluation while lack of abnormalities would not require 2DE. Model outputs of number of deaths, readmissions, and costs were calculated using a sample of 186,000 HF hospitalizations. One-way sensitivity analyses were conducted to evaluate the robustness of model outcomes. RESULTS: A PSU strategy yielded a 19.1% (7.2% vs. 8.1%) reduction in the 30-day mortality rate, as compared to the 2DE strategy, and a 33.4% (26.0% vs. 39.1%) reduction in 30-day readmissions leading to a cost savings of c55 million (c297/ HF admission). Models were robust to changes in model inputs with reduction in deaths from 2–37%, and cost savings from c63–315 per HF admission. CONCLUSIONS: Use of a PSU device to visualize key cardiac parameters for HF hospitalizations in Italy may afford reduction in deaths and cost-savings. PCV41 COST COMPARISON STUDY IN ACUTE CORONARY SYNDROMES AND ACUTE HEART FAILURE PATIENTS Ondrackova B1, Parenica J2, Felsoci M2, Miklik R2, Horakova K2, Spinar J2, Sulcova A1 1 Masaryk University, Faculty of Medicine, Brno, Czech Republic; 2Faculty Hospital Brno, Brno, Czech Republic OBJECTIVES: To assess and compare direct in-hospital costs in patients hospitalized with acute coronary syndromes (ACS; unstable angina or myocardial infarction), acute heart failure (AHF) and acute coronary syndromes complicated by heart failure (ACS with AHF) from health care payer perspective. METHODS: Patients hospitalized from 2005 to 2008 were classified in registry of Faculty Hospital Brno according to the clinical characteristics. Direct in-hospital costs (flat rate of admission and stay; medication and procedures) were evaluated in total 965 ACS (73% male; 64.7 years; 54% STEMI), 962 AHF (56% male; 70.8 years) and 808 AHF with ACS (62% male; 71.2 years; 61% STEMI) patients. The severity of disease was qualified as need of intensive care (ICU days); the predictors of high cost were identified in all groups. (1c = 26CZK). RESULTS: Mean in-hospital cost was 101658CZK/3910c in ACS pts (length-of-stay 5.6 days; 81% pts 2.3 ICU days; 0.3% died), 70201CZK/2700c in AHF pts (lengthof-stay 9 days; 48 % pts 4 ICU days; 11.4 % died) and 122340CZK/4705c in ACS with AHF (length-of-stay 7.2 days; 92% pts 3.5 ICU days; 19.2% died). Percutaneous coronary intervention with stenting (PCI) formed more than 80% of total cost in ACS patients (performed in 78% pts.; 91% STEMI; 62% non-STEMI) and almost 60% costs in ACS with AHF pts (performed in 66% pts; 84% STEMI; 39% non-STEMI); antiarrhythmic interventions (incl. ICD, CRT BiV or PM device) made almost 50% cost in AHF although carried out in 8.6% pts. CONCLUSIONS: According to inhospital costs patients with acute coronary syndromes complicated by heart failure were the most costly with maximum need of intensive care. Since 26.7% pts were transferred to other health care institution (e.g. 9% cardiosurgery) we expect the most significant increase of total amount of direct in-hospital costs in comparison with ACS (16.9% pts transferred) or AHF (11.5% pts transferred). 13th Euro Abstracts PCV42 HOSPITALIZATION COSTS IN SOUTH KOREA FOR PATIENTS WITHOUT IDENTIFIED RISK FACTORS FOR BLEEDING WHO HAVE ACUTE CORONARY SYNDROMES WITH PLANNED PERCUTANEOUS CORONARY INTERVENTION TREATED WITH PRASUGREL VS. CLOPIDOGREL IN THE TRITON-TIMI 38 TRIAL Mahoney E1, Wang K1, Yang BM2, Kim BRM2, Lee HY3, Chae IH4, Kang SY5 1 Mid America Heart Institute of Saint Luke’s Hospital, Kansas City, MO, USA; 2Seoul National University, Seoul, South Korea; 3School of Public Health Seoul National University, Seoul, South Korea; 4Seoul National University, Seongnam, South Korea; 5Eli Lilly & Company, Korea, Gangna-gu, South Korea OBJECTIVES: The TRITON-TIMI 38 Trial demonstrated that in patients with acute coronary syndromes (ACS) undergoing planned percutaneous coronary intervention (PCI), prasugrel compared to clopidogrel significantly reduced the rate of ischemic events over up to 15 months of follow-up, though there was an increased risk of major bleeding. Three risk factors for bleeding identified in the trial were a prior history of stroke/TIA, age ≥75 years, and body weight <60 kg. Patient-level data from TRITON was used to compare hospitalization costs in South Korea for patients without these risk factors for bleeding, who were randomized to prasugrel vs. clopidogrel. METHODS: Details regarding hospitalizations for all patients from 8 high enrolling countries (US, Australia, Canada, Germany, Italy, Spain, UK, France; n = 3373 prasugrel, 3332 clopidogrel) were collected prospectively. The US diagnosis-related group codes were assigned to all hospitalizations, to which 2009 South Korean unit costs were applied. Estimates of incremental costs associated with PCI-associated periprocedural MIs and PCI- and CABG-associated bleeding events were added separately, based on published U.S. data. RESULTS: Prasugrel was associated with a 17,723 KRW per patient increase in costs due to bleeding, and an 8,377 KRW decrease in costs due to reduced periprocedural MI, during the index hospitalization. Per-patient costs from hospitalization for MI (not involving revascularization) during the followup period were 291,022 KRW lower with prasugrel compared to clopidogrel, and costs associated with revascularization hospitalizations during follow-up were 505,032 KRW lower. Costs from follow-up bleeding events were 52,082 KRW higher for prasugrel, while aggregate hospitalization costs were 293,656 KRW lower. CONCLUSIONS: For ACS patients with planned PCI, the lower ischemic event rate with prasugrel yields a reduction in overall hospitalization costs, despite increased costs from bleeding. These results have implications for the cost-effectiveness of alternative antiplatelet strategies for ACS patients undergoing PCI in South Korea. PCV43 OFFSETTING ANNNUAL CARDIOVASCULAR DISEASE-RELATED MEDICAL AND PHARMACY COSTS BETWEEN PATIENTS CONTINUING STATIN MONOTHERAPY OR AUGMENTING STATIN MONOTHERAPY WITH NIACIN EXTENDED-RELEASE Quimbo RA1, Simko RJ2, Cziraky MJ1, Webb S2 1 HealthCore, Wilmington, DE, USA; 2Abbott Laboratories, Abbott Park, IL, USA OBJECTIVES: To compare cardiovascular disease (CVD) associated costs between patients continuing statin (simvastatin 20 mg/day or equivalent) monotherapy (SM) or receiving adjunct therapy with niacin extended-release (NER) for comprehensive lipid management. METHODS: Secondary risk patients age ≥30 continuing SM or augmenting statin therapy with NER between January 1, 2005-November 30, 2008 (index date) with ≥1 pre-index lipid panel (LDL-C < 100 mg/dL & HDL-C < 50 or <55 mg/dL for men & women, respectively) and no hepatic dysfunction were included. SM patients were matched 3:1 to NER patients based on prior exposure to statin. The primary outcome was total (medical and pharmacy) annual CVD-related costs. Medical and pharmacy costs consisted of inpatient, emergency room, and outpatient expenditures associated with CVD and pharmacy charges for anti-hyperlipidemic therapy. Outcome comparison between SM and NER and was conducted via generalized linear models via gamma distribution and log-link function. Multivariate analyses controlled for age, gender, geography, co-morbidities, and baseline lipids. RESULTS: A total of 728 patients were identified; 546 SM patients were matched to 182 NER patients. Compared to SM, there was no difference in baseline LDL-C (P = 0.13) or Deyo-Charlson co-morbidity Index scores (P = 0.77). Baseline HDL-C (35 ± 6 vs. 42 ± 7; P < 0.0001) was lower and TG (190 ± 98 vs. 153 ± 76; P < 0.0001) was higher among NER patients. While annual CVD-related pharmacy expenditure was higher for NER [mean (95% CI), $2261 ($1991–$2587) vs. $1276 ($1186–$1373)], adjusted and unadjusted medical costs were 69% (P = 0.0002) lower for NER ($542 ($378– $799) vs. $1718 ($1373–$2149); P < 0.0001) vs. SM. Offsets in medical and pharmacy costs resulted in total annual CVD-related costs between NER and SM (($2796 ($2331–$3353) vs. $2988 ($2960–$3318); P = 0.5346). CONCLUSIONS: Therapy augmentation with NER for comprehensive lipid management was associated with reduced annual CVD-related medical expenditure leading to total annual CVD-related costs compared to statin monotherapy. PCV44 EVALUATION OF CLINICAL AND DIRECT ECONOMIC OUTCOMES BETWEEN STATIN AND FENOFIBRATE AND STATIN MONOTHERAPY FOLLOWING AN INITIAL ACUTE CORONARY EVENT Simko RJ1, Quinlan S2, Quimbo RA2, Cziraky MJ2, Webb S1 1 Abbott Laboratories, Abbott Park, IL, USA; 2HealthCore, Wilmington, DE, USA OBJECTIVES: Compare the risk of subsequent coronary event (SCE), cerebrovascular event, and associated medical costs between patients maintaining statin monotherapy (SM) or augmenting statin with fenofibrate (FF) after an acute coronary event (ACE). 13th Euro Abstracts METHODS: Patients with an ACE between January 1, 2004 and May 31, 2009 were identified (the Index Date was defined as the earliest ACE). Patients age ≥18, ≥6 months eligibility pre- and post-Index Date, ≥1 statin fill and no statin augmenting fill 6 months before the Index Date, and ≥1 FF or SM fill 6 months after the Index Date (Cohort ID Period) were included. Patients with fills for alternative statin augmenting therapies or SCE during the Cohort ID Period were excluded. The primary outcome was SCE. Secondary outcomes were cerebrovascular event risk and eventrelated medical costs. Event risk was compared using Cox proportional hazards models, while costs were compared via generalized linear models. Multivariate analyses controlled for age, gender, geography, co-morbidities, concomitant medications, statin potency, and severity of initial ACE. RESULTS: A total of 24,625 patients added FF (n = 399) or maintained SM (n = 24,226) 6-months after initial ACE while meeting study inclusion/exclusion criteria. FF patients were younger (59 ± 10 vs. 65 ± 12, P < 0.0001), less likely female (27% vs. 34%, P = 0.002), and had shorter follow-up (358 vs. 443 days, P < 0.0001). FF patients had lower frequency of SCE (20%) vs. SM (25%), with an unadjusted hazard ratio (HR) of 0.88 (95%CI: 0.71–1.10). a similar directional trend was observed for cerebrovascular events (HR: 0.74; 95% CI: 0.40– 1.37). Multivariate analyses of SCE risk demonstrated consistent directional benefit (HR: 0.93; 95% CI: 0.74–1.16). Clinical findings corresponded with 25% lower adjusted annual event-related costs compared to SM (0.75; 95% CI: 0.57—0.97; p = 0.0306). CONCLUSIONS: The positive clinical benefits from this real-world analysis provide evidence supporting ability of FF to reduce both SCE risk and event-related medical costs following ACE. PCV45 COST-EFFECTIVENESS ANALYSIS OF ROSUVASTATIN VERSUS ATORVASTATIN IN TURKEY: AN ANALYSIS BASED ON THE SYSTEMATIC REVIEW OF PUBLISHED RANDOMIZED CLINICAL TRIALS (RCTS) Abaci A1, Kabakci G2, Oguz A3 1 Gazi University Faculty of Medicine, Ankara, Turkey; 2Hacettepe University Faculty of Medicine, Ankara, Turkey; 3Goztepe Training and Research Hospital, Istanbul, Turkey OBJECTIVES: The aims of study were 1)to determine equivalent dosages of Rosuvastin(R) and Atorvastatin(A) in the treatment of dyslipidemia, and 2)to evaluate cost-effectiveness of R vs. A. METHODS: The R and a arms of 29 RCTs with stable dose treatments for minimum 6 weeks (n = 27.588) were included in the analysis. The arms with periods after titration or switch were excluded. The effectiveness parameters chosen were the pooled percent decrease in LDL-C, LDL/HDL, TC/HDL ratios and 10-years’ cardiovascular risk (CVR) based on Wilson equation. Only medication costs were calculated for R and A, since other costs directly related with dyslipidemia were assumed to be same. Dose-effect relationship was analyzed by plotting linear regression line for R and A, separately. Incremental Cost-effectiveness Ratios (ICER) were calculated as added cost per percent decrease in lipid parameters and per absolute reduction in 10-years’ CVR. RESULTS: Equivalent clinical effectiveness of R was provided with 3.9-fold, 3.2-fold and 3.2-fold dosages of a in mg, in terms of percent decrease in LDL-C, LDL/HDL and TC/HDL ratios, respectively. Absolute 10-years’ CVR reductions with R5-40 mg and A10-80 mg were 4.5–6.9% and 3.7–5.7%, respectively. 1(R):2(A) dose pair was selected for cost-effectiveness analysis. The yearly medication costs were as follows: R5 mg vs. A10 mg (c113.67** vs. c128.03), R10 mg vs. A20 mg (c174.68 vs. 226.77c), R20 mg vs. A40 mg (c261.26 vs. 283.20c) and R40 mg vs. A80 mg (c261.26 vs. c294.68). R dominated a with regards to ICER for all effectiveness parameters. Sensitivity analysis proved that dominance of R was stable when the model was run for extreme values for R and a in the literature. CONCLUSIONS: In this analysis, R was found more efficacious than three times higher dose of A. R is more cost-effective than a in reducing LDL-C level, LDL-HDL and TC/HDL ratios and 10-years’ CVR reduction. **1 Euro = 1.9595TL PCV46 COST STUDY OF CAREGIVING FOR PATIENTS WITH CHRONIC SYMPTOMATIC HEART FAILURE IN SPAIN Marti B1, Delgado J2, Oliva J3, Llano M4, Pascual P5, Comin J6, Grillo JJ7, Diaz Molina B8, Culebras J9, Martínez de la Concha L10, Manito N11 1 Medtronic Iberia, Madrid, Spain; 2Hospital 12 de Octubre, Madrid, Spain; 3Univesidad de Castilla La Mancha, Toledo, Spain; 4Hospital Universitario Marqués de Valdecilla, Santander, Spain; 5Hospital Universitario Virgen de la Arrixaca, Murcia, Spain; 6Hospital del Mar (IMIM), Barcelona, Spain; 7Hospital Universitario Nuestra Señora de la Candelaria, Santa Cruz de Tenerife, Spain; 8Hospital Universitario Central de Asturias, Oviedo, Spain; 9Hospital Universitario Insular, Las Palmas De Gran Canaria, Spain; 10Hospital Universitario Infanta Cristina, Badajoz, Spain; 11Hospital de Bellvitge, Barcelona, Spain OBJECTIVES: The objective of this study was to quantify, for the first time in a Spanish population, the time and cost burden of informal care for patients with heart failure. METHODS: A descriptive analysis of a multicenter, prospective observational study was performed. Patients who met inclusion criteria were followed-up for a period of 12 months, with 3 visits programmed at baseline, 6 months and 12 months. a total of 9 Spanish hospitals were involved in the study. Baseline characteristics and caregiver’s information were registered for every patient. Once identified total hours, the replacement cost method was used. RESULTS: A total of 450 patients were included, 76% men, mean age was 62.6 years. 66.1% were in NYHA class II, 32.7% NYHA class III and 1.1% NYHA class IV. Prevalence of ischemic cardiopathy was 33.5%. Thirty-five percent of patients had an implantable device (ICD, RCT or pacemaker). One-third needed support for daily activities. One-hundred and one informal caregivers were identified, mean age of 57.6 years, mostly women (84.1%). A349 Main relationship with caregiver was spouse/couple (74.2%), followed by son/daughter (16.6%). Number of weekly hours of caregiving was estimated at 42.5 hours (39.46 hours for patients NYHA class II and 47.39 hours for patients NYHA class III-IV) and shadow prices values from c9 to c12 per hour. Total costs associated to informal caregiving increases between c19,897 and c26,530 (c18,466–c24,621 for patients in NYHA class II and c22,178–c29,571 for patients in NYHA class III or IV). CONCLUSIONS: Approximately one-third required support from an informal caregiver, which represents a significant burden for society and often has not been accounted for in economic evaluations of treatments for heart failure. Costs for informal care appear to increase with worse disease severity as measured by NYHA class. PCV47 A NATIONAL SURVEY OF SOCIETAL COST OF ACUTE CORONARY SYNDROME IN SOUTH KOREA Kim J1, Lee T1, Lee E1, Kim J2, Yoo W3, Boo Y4 1 Seoul National University, Seoul, South Korea; 2Yuhan University, Seoul, South Korea; 3 Daegu Haany University, Seoul, South Korea; 4Eulji University, Seoul, South Korea OBJECTIVES: Acute coronary syndrome (ACS) is a highly prevalent disease and the third-leading cause of death in Korea, however the economic impact of ACS from societal perspective has not been investigated. This study examined the national epidemiology and economic costs of ACS in Korea for 2009. METHODS: This study used the societal cost of illness framework, consisting of epidemiology of ACS, direct medical costs, direct non-medical costs, and indirect costs such as productivity lost from morbidity and mortality of ACS. National-level health survey results and insurance claim databases were used to analyze annual health care utilization, hospitalization costs and outpatient costs of the total Korean population (48 million people). Using a data mining technique, we identified medical claims with international classification of disease 10 codes for ACS over a one year period and estimated the costs by a macro-costing method. RESULTS: Prevalence rate of ACS is estimated at 6.4 persons per 1000 population in 2009 and its mortality rate is 7.8% in 2008. During 2009, there were 309,384 patients who had at least 1 medical claim related to ACS. The total societal cost of ACS in 2009 was estimated at $1.02 billion. Direct medical cost for ACS was $380.3 million, which includes hospitalization cost of $318.8 million and outpatient cost of $61.6 million. The direct non-medical cost, involving transportation cost and caregiver cost, was estimated at $10.1 million. Indirect costs associated with morbidity and mortality of ACS were $627.2 million. CONCLUSIONS: The study showed that the societal cost of ACS in Korea is markedly high, particularly the indirect cost, followed by hospitalization cost. Early and effective management of the disease is necessary to reduce mortality and morbidity of ACS. Findings of this study suggest for further research to uncover ways to reduce the economic burden of ACS in Korea. PCV48 COST-OF-ILLNESS AND HEALTH-RELATED QUALITY OF LIFE IN SUBARACHNOID HEMORRHAGE: GERMAN LONGITUDINAL STUDY Winter Y1, Ringel F2, Spottke A3, Gharevi N4, Klockgether T3, Schramm J4, Urbach H5, Meyer B2, Dodel R1 1 Neurology, Philipps University, Marburg, Germany; 2Neurosurgery, Klinikum rechts der Isar, Technical University, Munich, Germany; 3Neurology, University of Bonn, Bonn, Germany; 4 Neurosurgery, University of Bonn, Bonn, Germany; 5Neuroradiology, University of Bonn, Bonn, Germany OBJECTIVES: Aneurysmal subarachnoid hemorrhage (SAH) is a cerebrovascular disease with a high mortality and severe disability. Longitudinal studies investigating health-economic burden and health-related quality of life (HRQoL) in SAH are scare and only one previous study analyzed cost-driving factors. The objective was to evaluate HRQoL and first-year costs in German patients with aneurysmal SAH and to identify cost-driving factors. METHODS: A total of 113 incident cases of aneurysmal SAH treated in the Department of Neurosurgery and Neuroradiology at the University of Bonn between January 2004 and December 2005 were eligible for the study. Clinical assessments (Hunt and Hess scale, Barthel Index), evaluation of HRQol (36-Item Short Form Survey, EQ5D, EQ-VAS) and collection of cost data from societal perspective (health-economic questionnaires) were performed at discharge, and at 6 and 12 months follow-up. All costs are expressed in c (year 2009 values). Independent determinants of costs and HRQoL were identified using multiple regression analysis. RESULTS: The total first-year costs were c38,300 (95%CI: 34,490–43,100) per patient. Direct costs accounted for 58.7% of total costs and were mainly paid by the health insurance (92.0%). The major cost-driving factors were younger age and worse functional outcome (Barthel-Index). HRQol in SAH patients was considerably reduced. At discharge, 92.2% of patients had moderate or severe problems on the EQ-5D. The EQ-VAS at discharge was 57.8 ± 19.3 and increased by only 12–14% after 12 months. Independent predictors of decreased HRQol were female gender, severe disease, depression, lower level of education and the lack of a stable partnership. CONCLUSIONS: Aneurysmal SAH is a cerebrovascular disease with considerable health-economic burden and unfavourable long-term HRQoL outcome. Clinical trials in SAH should include HRQol measures. Independent determinants of costs and HRQol identified in this study should be considered in the health care programmes aimed at increasing the HRQol in SAH survivors and decreasing health-economic burden of SAH. A350 PCV49 ESTIMATING THE COST OF ATRIAL FIBRILLATION IN GREECE Athanasakis K, Karampli E, Ollandezos M, Kyriopoulos J National School of Public Health, Athens, Greece OBJECTIVES: Determine the health care resource use and costs attributable to Atrial Fibrillation (AF) in Greece. METHODS: A multi-point data collection procedure, based on patient records of 149 geographically distributed physicians, was used in order to obtain the necessary data for the patient cost model. Patients were categorized according to initial treatment strategy i.e. pharmacological treatment (rhythm control, rate control) and non-pharmacological treatment. Cost categories included costs of consultations, laboratory tests, AF-related medication, anti-thrombotic medication and hospitalization due to AF recurrence, cardiovascular complications or drugrelated adverse events. Calculations were based on 2009 fees and prices from a thirdparty payer perspective. RESULTS: A total of 94.5% of treated patients were on a pharmacological strategy (rhythm control: 56.2%, rate control: 38.3%) and the remaining 5.5% had undergone non-pharmacological treatment, mainly catheter ablation (96% of cases). Mean annual direct cost of treatment per patient, was estimated at c3184, c2095 and c6452 for AF patients under rhythm control, rate control and non-pharmacological treatment respectively and at c2947 on average, regardless of treatment strategy. Hospitalizations were the major cost drivers for patients under rhythm control and for those under non-pharmacological treatment accounting for up to 75.9% of total expenses. Laboratory tests were, also, an important cost driver (the major for rate control patients), attributing from 15.5%—50.4%, (36.4% on average) to total costs, followed by anti-thrombotic medication, ranging from 4.4%– 18.7% (12.3% on average) and consultations ranging from 3.1%–12.4% (9.1% on average). AF-related medication was the smallest portion of expenses accounting for 1.0%–5.4% (4.2% on average) of total costs, across all groups of patients. CONCLUSIONS: Treatment of AF poses a significant health care burden accounting for c275,873,173 annually or 1.3% of health care expenditure in Greece in 2009. Efforts to reduce the main AF cost drivers and hence its economic impact on the health care system, are necessary. PCV50 THE COST OF ATRIAL FIBRILLATION IN SWEDEN Davidson T, Husberg M, Janzon M, Levin LÅ Linköping University, Linköping, Sweden OBJECTIVES: Atrial fibrillation (AF) is the most common arrhythmic disease in Sweden. Persons with AF have a significant increased risk of stroke and the main treatment is therefore intended to prevent stroke by anticoagulation with warfarin or acetylsalicylic acid. However, new anticoagulation treatments will soon challenge the established treatments and there will be a need to know the cost of AF for costeffectiveness comparisons. In this study we estimate the cost of AF in Sweden 2010. METHODS: All relevant costs are identified, quantified and valued. The focus in the calculation is on the complications (ischemic- and hemorrhagic stroke, gastrointestinal bleedings and other major bleedings) caused by AF and treatments intended to lower the risks for ischemic stroke. a societal perspective is used, and therefore productivity loss caused by morbidity is included. Patients with CHADS2 = 1 or higher are included. RESULTS: In Sweden with 9,340,000 inhabitants, there are 118,000 patients with AF and at least one more risk factor of stroke, which comprise 1.26% of the population. Among the patients; 43.3% are treated with warfarin, 28.3% use acetylsalicylic acid and 28.3% are assumed to have no anticoagulation treatment. The cost of AF in Sweden is estimated at SEK 3.4 billion (c354 million and US$439 million) for 2010. This equals SEK 29,000 (c3,023 and US$3,745) per AF patient and year. The highest cost is caused by stroke, the second highest is the cost of monitoring the warfarin treatment. As the prevalence of AF is expected to increase, the AF cost is expected to rise in the future. CONCLUSIONS: AF causes high costs for the society. New treatments that could reduce the risk of stroke and be easily administered might have a potential of being cost-effective. PCV51 THE IMPACT OF ACQUIRED BRAIN DAMAGE IN TERMS OF EPIDEMIOLOGY, ECONOMICS AND LOSS IN QUALITY OF LIFE Mar J1, Arrospide A1, Begiristain JM2, Larranaga I2, Oliva J3 1 Hospital Alto Deba, Mondragon, Spain; 2Government of the Basque Country, San Sebastian, Spain; 3University of Castilla la Mancha, Toledo, Spain OBJECTIVES: Patients with acquired brain damage (ABD) have suffered a brain lesion that interrupts vital development in the physical, psychological and social spheres. The objectives of this study were to estimated its incidence and prevalence of ABD, to calculate the associated cost of the care required and finally to assess the HRQL. METHODS: A cross-sectional survey was carried out, in order to estimate the incidence of ABD and its consequences in terms of costs and loss in quality of life from the evolution of a sample of patients diagnosed with stroke and TBI. On the other hand, a discrete event simulation model was built that enabled the prevalence of ABD to be estimated. Finally, a calculation was made of the formal and informal costs of ABD in the population of the Basque Country and Navarre (2,750,000 people). RESULTS: The cross-sectional study showed that the incidences of ABD caused by stroke and TBI were 61.8 and 12.5 cases per 100,000 per year respectively, while the overall prevalence was 657 cases per 100,000 people. The SF-36 physical and mental component scores were 28.9 and 44.5. The economic burden was calculated to be c382.14 million per year, distributed between 215.27 and 166.87 of formal and informal burden respectively. The average cost was c21,040 per person- year. 13th Euro Abstracts CONCLUSIONS: The main conclusion of this study is that ABD has a high impact in both epidemiological and economic terms as well as loss in quality of life. The overall prevalence obtained is equivalent to 0.7% of the total population. The substantial economic burden is distributed nearly evenly between formal and informal costs. The physical dimensions of quality of life are the most severely affected. The prevalence-based approach showed adequate to estimate the population impact of ABD and the resources needed to compensate the disability. PCV52 COST OF MANAGING CHRONIC THROMBOEMBOLIC PULMONARY HYPERTENSION TO MANAGED CARE Said Q1, Martin B1, Joish V2, Gabriel P3, Kreilick C4, Seal B4, Williamson T4, Mathai S5 1 University of Arkansas for Medical Sciences, Little Rock, AR, USA; 2Bayer, Wayne, NJ, USA; 3 Zenith Healthcare solutions, Indianapolis, IN, USA; 4Bayer Healthcare Pharmaceuticals, Inc., Wayne, NJ, USA; 5Johns Hopkins University, Baltimore, MD, USA OBJECTIVES: To estimate direct medical costs and resource use of chronic thromboembolic pulmonary hypertension (CTEPH) patients. METHODS: Data were from a managed care claims database (2004–09) using a retrospective cohort design. Criteria to identify CTEPH patients were: >18 years old, >2 claims for PH (ICD-9:416.0, 416.8), with first claim occurring at least 6 months after a claim for pulmonary embolism (PE), >1 claim for right heart catheterization (RHC) or echocardiogram before the second PH claim, and continuously enrolled during the study period. Each CTEPH patient was matched to 1 to 5 controls without CTEPH on age, sex, region and payer type. Per-patient-per-month costs were contrasted between CTEPH patients and controls using Wilcoxon rank sum test. RESULTS: A total of 146 cases and 558 controls were identified (mean age 64 yrs, 54.8% female, mean follow up of 22.2 months). Total monthly costs in the period prior to CTEPH diagnosis were higher for CTEPH patients ($3895) than controls ($1177). Most common co-morbidities in CTEPH patients were: Acute pulmonary heart disease (8.7%), symptoms involving respiratory system and other chest symptoms (7.5%), other forms of chronic ischemic heart disease (6.4%). After CTEPH diagnosis, CTEPH patients had significantly higher monthly resource use and costs vs. controls: Total costs $6198 vs. $1579, Outpatient visits 1.2 ($2193) vs. 0.8 ($780), inpatient visits, 2 ($3382) vs. 0.2 ($507), prescriptions 4.2 ($623) vs. 2.8 ($292), all p-values < 0.05. a higher proportion of CTEPH patients had cardiovascular related prescriptions [CCBs (21.2% vs. 18.2%), diuretics (46.6% vs. 29.2%, p = value), oral anticoagulants (56.9% vs. 11.7%), digoxin (8.9% vs. 3.4%), ERAs (3.5% vs. 0.0%)], and procedures [echocardiography (22.6% vs. 4.8%), electrocardiography (43.2% vs. 24.9%), computerized tomography (17.1% vs. 7.5%), using chi-square test all p-values < 0.05, except CCBs. CONCLUSIONS: CTEPH patients used more resources vs. controls. Research is needed to identify factors related to increased utilization. PCV53 THE BURDEN OF ISCHEMIC HEART DISEASES AT A MAJOR CARDIAC CENTER IN RIYADH, SAUDI ARABIA Alsultan MS1, Osman AM1, Al-Mutairi M2 1 King Saud University College of Pharmacy, Riyadh, Saudi Arabia; 2Prince Sultan Cardiac Centre, Riyadh, Saudi Arabia OBJECTIVES: Ischemic Heart Disease (IHD) is the leading cause of death worldwide, including in Saudi Arabia. Cost of illness (COI) studies aiming to explore the burden of IHD are missing in Saudi Arabia. Therefore, the objective of this study is to estimate the direct medical costs associated with IHD at Prince Sultan Cardiac Center (PSCC), a major Cardiac referral center in the country. METHODS: A prospective COI study was conducted from April 2009 to June 2009 from the PSCC perspective. All patients diagnosed or suspected having IHD at admission were included in the study. They were followed up till discharge or performing CABG or changing diagnosis. Clinical data were extracted from the patient computerized database and combined with the unit cost of services (medication, procedures, bed utilization) to calculate direct medical costs. RESULTS: A total of 205 patients were recruited and diagnosed with stable angina (SA) (47.8%), unstable angina (USA) (24.4%), STEMI (19.5%) and NSTEMI (8.3%). Most of the patients were male, Saudi, aged between 40–75 years. 87% of the patients had two or more co-morbidities and 32% of the patients were obese. The average cost is 40,164 SAR/patient (US$10,710). Medication contributed the lowest in the costs (3.2%). ACEI contributed only 0.17% in the total costs. Costs associated to SA, USA, NSTEMI and STEMI were respectively 33,991; 35,107; 46,585 and 58,877 SAR/patient. The lowest hospital length of stay was 6.5 days with SA. The average length of stay increased with the number of co-morbidities from 5.67 (no co-morbidity) to 11.25 (6 co-morbidities). CONCLUSIONS: The study shows that IHD is of high economic burden in the Kingdom. Among the 4 types of IHD studied, the resource consumption associated to SA was the lowest in terms of costs and patient hospital length of stay. Co-morbidities increased the hospital length of stay. 13th Euro Abstracts PCV54 PATTERNS OF THERAPY, HEALTH CARE UTILIZATION, AND HEALTH CARE COSTS IN PATIENTS WITH PULMONARY ARTERIAL HYPERTENSION (PAH) INITIATING THERAPY WITH SILDENAFIL: FINDINGS FROM RETROSPECTIVE ANALYSES OF ADMINISTRATIVE HEALTH CARE CLAIMS DATA Mychaskiw MA1, Berger A2, Mardekian J1, Oster G2 1 Pfizer Inc, New York, NY, USA; 2Policy Analysis Inc., Brookline, MA, USA OBJECTIVES: To examine patterns of therapy, health care utilization, and health care costs among PAH patients initiating therapy with sildenafil in a real-world setting. METHODS: Patients aged ≥18 years with evidence of PAH (ICD-9-CM diagnosis codes 416.0, 416.8) and ≥1 claims for sildenafil between June 1, 2005, and September 30, 2008 were identified from a large administrative health care claims database. Patients with <6 months of pretreatment data were excluded. Patients were followed from index date (first-noted sildenafil claim) until health plan disenrollment or end of study (follow-up). Patterns of therapy with sildenafil were examined, including numbers of prescriptions and associated therapy-days and compliance; the latter was measured using medication possession ratio (MPR, ratio of total therapy-days to total days of follow-up). For the subgroup of patients with ≥6 months of follow-up data, health care utilization and costs were compared between the 6-month period preceding and following the index date. RESULTS: Of 855 PAH patients identified who began therapy with sildenafil and met study inclusion criteria (mean age, 53 years; 69% women), 32% had comorbid lung conditions and 17% connective tissue disorders. Over a mean duration of follow-up of 423 days (median, 357 days), patients averaged 6 prescriptions for sildenafil spanning 209 therapy-days (mean MPR = 0.78). Health care utilization was largely unchanged between the 6-month preindex and postindex periods. Mean costs of outpatient care decreased by $501 ($8321 vs. $7820 during preindex); mean total health care costs increased by $4137 (95% CI, $38,815– $42,952), primarily due to increased costs of PAH-related pharmacotherapy (including sildenafil) (P < 0.01 for all comparisons). CONCLUSIONS: PAH patients initiating sildenafil therapy are relatively compliant with treatment. Although health care utilization is largely unchanged following initiation of sildenafil therapy, total health care costs increase, primarily due to costs of PAH-related pharmacotherapy. Further research is needed to better understand the real-world impact of sildenafil on patient health outcomes. PCV55 HEALTH CARE UTILIZATION AFTER PLATELET AGGREGATION INHIBITORS Löfroth E, Murén KJ, Bruce S IMS Health, Stockholm, Sweden OBJECTIVES: To analyze the annual health care utilization by patients dispensed platelet aggregation inhibitors excluding heparin. METHODS: This was a retrospective longitudinal descriptive database study of the utilization of health care of patients from the South-West region of Sweden (1.5 million inhabitants). All patients who were dispensed platelet aggregation inhibitors, excluding heparin (N05AC) during 2007 and who had a cardiovascular principal diagnosis (chapter 9, ICD-10) were included in the study. Only costs related to cardiovascular care (i.e. visits with a cardiovascular diagnosis) were included. RESULTS: a total of 46,742 patients had at least one dispensed platelet aggregation inhibitor and at least one cardiovascular related health care visit during 2007 in the South-West region of Sweden. a total of 15,779 had at least one in-patient stay, 18,211 had at least one out-patient visit, and 28,432 had at least one primary care visit. The total costs of in- and out-patient care, primary care, and of platelet aggregation inhibitors, excluding heparin was 1.5 billion SEK. The greatest cost contributor was in-patient care, a total of 1.2 billion SEK, 74,501 SEK per patient in in-patient care. The total cost for out-patient care was 98 million SEK, 5371 SEK per patient. The total cost for primary care was 92 million SEK, 3223 SEK per patient. The total drug cost was 43 million SEK, 912 SEK per patient. CONCLUSIONS: For those patients who had at least one health care visit and who were dispensed platelet aggregation inhibitors the drug cost amounted to less than 3% of the total annual cost. In-patient care represented as much as 80% of the total annual costs. PCV56 RENAL MORBIDITY, MORTALITY, AND COSTS IN INDIVIDUALS UNDERGOING INVASIVE CARDIAC CATHETERIZATION PROCEDURES WITH LOW-OSMOLAR CONTRAST MEDIA: A LARGE RETROSPECTIVE DATABASE ANALYSIS Min J1, Ryan A2, Spalding J3 1 Weill Cornell Medical College, New York, NY, USA; 2GE Healthcare Clinical Data Services, Princeton, NJ, USA; 3GE Healthcare, Barrington, IL, USA OBJECTIVES: To investigate in-hospital hemodialysis (HD), length of stay (LOS), mortality, and costs following use of low-osmolar contrast media (LOCM) in patients undergoing invasive cardiac catheterization procedures. METHODS: This retrospective analysis used the Premier PerspectiveTM database, which contains patient-level data. In-patient adults without prior HD who underwent invasive cardiac catheterization procedures with LOCM during 2007–2008 were studied (iohexol: n = 36,118, iopamidol: n = 36,089, ioversol: n = 135,619). Propensity score weighted and multivariate logistic regression analyses were used. RESULTS: In-hospital HD rates were low after exposure to any of the 3 LOCM (0.9% iohexol, 1.0% iopamidol, 1.0% ioversol). Risk of HD did not differ between iopamidol and iohexol, but ioversol demonstrated increased risk of HD compared to iohexol (adjusted Odds Ratio [OR] A351 1.17, 95% CI 1.02–1.35) and iopamidol (adjusted OR 1.17, 95% CI 1.02–1.34). For in-hospital mortality (2.1% iohexol, 2.3% iopamidol, 1.9% ioversol), no betweengroup differences were statistically significant. Similarly, all-cause 30-day readmission rates (10.8% iohexol, 10.4% iopamidol, 10.6% ioversol) did not differ significantly between the groups. Contrast-induced acute kidney injury-related 30-day readmission rates were 0.2% in all 3 groups. There was no significant difference in the mean adjusted LOS between the groups (6.2–6.3 days). Iohexol was associated with a significantly lower (P < 0.001) mean adjusted initial hospitalization cost ($21,591) compared with either iopamidol ($23,482) or ioversol ($23,484) and with a significantly lower (P < 0.001) mean adjusted initial hospitalization cost post-procedure ($10,512) compared with either iopamidol ($11,393) or ioversol ($11,187). CONCLUSIONS: In this large cohort of hospitalized patients, in-hospital HD rates were low after invasive cardiac catheterization procedures with LOCM. The risk of HD was comparable with iohexol and iopamidol, while ioversol was associated with a significantly higher risk. While in-hospital mortality rates, LOS, and all-cause 30-day readmission rates did not differ significantly between the 3 LOCM, iohexol was associated with significantly lower cost. PCV57 COST ANALYSIS OF THE MEDICINE TREATMENT OF SOUTH AFRICAN PATIENTS WITH METABOLIC SYNDROME Burger JR, Serfontein JHP, Lubbe MS North-West University, Potchefstroom, South Africa OBJECTIVES: In the absence of data on the economic impact of the metabolic syndrome amongst South Africans, this study aims to investigate the direct cost of treatment associated with the syndrome using data from a South African Pharmacy Benefit Management company (PBM). METHODS: A non-experimental, retrospective quantitative drug utilization review was performed using prescription claims data. Data for a 4-year period (January 1, 2005 to December 31, 2008) were analyzed. The direct costs of medicine treatment were computed by the addition of the medical scheme contribution and the patient levy. All costs were respectively expressed in 2005, 2006, 2007 and 2008 $U.S. RESULTS: Prescription claims data for a total of 10,567 patients in 2005; 12,123 in 2006; 12,769 in 2007 and 13,201 in 2008 with a diagnosis of metabolic syndrome were included in the study. The total direct cost of medicine treatment for these patients accounted for $7,921,831.10 in 2005, compared to $9,290,773.29 in 2006, $9,780,931.85 in 2007 and $8,515,349.53 in 2008. These costs represented 2.76% (N = $287,538,709.70) of all claims during 2005, compared to 3.18% (N = $292,314,629.60) of claims in 2006, 3.58% (N = $273,086,935.30) in 2007, and 3.87% (N = $220,144,319.80) in 2008. Private insurers contributed 90.91% (N = $7,921,831.10) towards the treatment of patients with metabolic syndrome in 2005, compared to 89.92% (N = $9,290,773.29) during 2006, 89.50% (N = $9,780,931.85) in 2007 and 87.17% (N = $8,515,349.53) in 2008. Patients paid on average $68.15 ± 106.18 (median $17.59) out-of-pocket for levies for 2005, compared to $77.21 ± 110.54 (median $35.30) during 2006, $80.42 ± 98.43 (median $49.15) in 2007 and $82.76 ± 96.64 (median $52.66) in 2008. CONCLUSIONS: Results show that the costs associated with the metabolic syndrome contributes significantly towards the annual expenditure of the South African PBM. Private insurers carry most of this burden, although out-of-pocket expenses for patients show an increasing trend. PCV58 SUBSTITUTION INFLUENCE FROM MARK DRUGS TO GENERICS IN THE ARTERIAL HYPERTENSION THERAPEUTIC COMPLIANCE AND THE DYSLIPEMIA IN A POPULATION SETTING Sicras-Mainar A1, Navarro-Artieda R2 1 Directorate of Planning, Badalona Serveis Assistencials, Badalona, Barcelona, Spain; 2Hospital Universitari Germans Trias i Pujol, Barcelona, Spain OBJECTIVES: To determine the therapeutical compliance in subjects with amlodipine and simvastatin substitution from mark to generic. Secondary objective: to know the professional and patient opinion about its use. METHODS: Patient and methods: Before-after design with control group, realized revision of the medical records, in six centres of primary care. Participants: Patients ≥ 40 years initiating treatment of mark drug (initial period) and after that had a substitution by generic (final period), between January 2003 and June 2009. Study groups: arterial hypertension (amlodopine) and Dislipemia (simvastatin). Main measures: Co-morbidity, compliance, treatment-time, biochemistry objectives and professional-patients opinion (interviews). Continuation: Beginning from the generic drug substitution date, each patient was (Minimum) one year before with a mark drug and a year after with and generic (minim continuation/ patient: 24 months). RESULTS: A total of 1252 patients. Groups: 49.5% amlodipine; 50.5% simvastatin. Average age: 72.4 years; women: 48.9%. The ones treated with amlodipine (comparing periods); show a better compliance (65.8% vs. 61.3%; p = 0.037) and arterial pressure control (48.5% vs. 45.8%; p = 0.039) with mark drugs. With simvastatin, was 62.8% vs. 58.4%; p = 0.041 (cholesterol control: 66.5% vs. 60.6%; p = 0.032) respectively. Interviewed doctors: 73.6% prescribe generic and a 59.2% believe that both have the same efficacy. Interview patients: a 79.8% (CI: 74.3–85.3%) accepted substitution; a 55.3% (CI: 48.5–62.1%) received the appropriate information and 61.5% the different products confuse them. a 18.2% declare that abandon treatment. CONCLUSIONS: In patients with amlodipine/simvastatin substitution from mark to generics, it was observed less treatment compliance light, with minor consecution of control objectives. The realized interviews (professional/patients) reaffirm the results. A352 PCV59 COMPLIANCE INFLUENCE, PERSISTENCE AND THE BLOOD PRESSURE CONTROL GRADE ASSOCIATED WITH THE INCIDENCE OF CARDIOVASCULAR EVENTS AND THE SANITARY COSTS IN CONSUMPTION OF FIXED-DOSES IN THE ARTERIAL HYPERTENSION TREATMENT Sicras-Mainar A1, Galera J2, Muñoz-Ortí G3, Navarro-Artieda R4 1 Directorate of Planning, Badalona Serveis Assistencials, Badalona, Barcelona, Spain; 2Novartis Farmaceutica S.A., Barcelona, Spain; 3Hospital Universitari Germans Trias i Pujol, Barcelona, Spain OBJECTIVES: To determine the incidence of cardiovascular events (CVE) and the sanitary costs in function of the compliance, persistence and blood pressure control; comparing patients consuming fixed-doses (FD) in front of free-doses (FD) in the treatment of HTA. METHODS: Observational-multicentric design. It was included patients >30 years appertaining to six team of primary care and two hospitals, that started pharmacological treatment for hypertension during 2006. It was established two study groups: FD (IECA/diuretics; ARA II/diuretics) and FD (IECA + DIU; ARA II + DIU, separately). Main measures: socio-demographics, co-morbidity, parameters, Charlson-index, compliance, persistence and control therapeutical objects (criteria: ESH-ESC). It was determined the accumulated incidence tax of CVE and a total-cost model (differentiating: sanitary/direct; non-sanitary/indirect). The patients’ pursuit was realized during two years. Statistic analysis: logistic regression, proportional risk model of Cox and the ANCOVA, P < 0.05. RESULTS: It was recruited 1605 patients, 1.112 (69.3%) in FD and 493 (30.7%) in FD, P < 0.001; age average: 69.4 (12.2) years; women: 55.5%. Patients in FD were associated with the isquemic cardiopathy OR = 1.4 (CI of 95%: 1.1–2.0) and organic insufficiencies OR = 1.5 (IC of 95%: 1.2–2.1), P < 0.031. Patients in FD showed a better therapeutic compliance (77.6 vs. 71.9; P < 0.001) and treatment persistence at 24 months (62.1% [CI of 95%: 56.3– 67.9%] vs. 49.7% [CI of 95%: 38.5–60.9%]; P < 0.001. The optimum control of the arterial pressure in FD was higher (48.9% [CI of 95%: 43.0–54.8%] vs. 46.7% [CI of 95%: 35.6–57.8%]; P < 0.001). The accumulated incidence tax of vasculocerebral accident in FD was 4.6% vs. 2.4%; p = 0.041. The total cost in FD was lower (c1650.7 vs. c1674.8; P < 0.001), in specialized care (c316.1 vs. c382.9; P < 0.001) and loses of labour productivity (c44.5 vs. c88.4; P < 0.001). CONCLUSIONS: Treatment of compliance and persistence for hypertension in FD improve the therapeutical control, causing a reduction of CVE and total sanitary costs. PCV60 COST OF ANALYSIS OF THE ADVERSE EFFECTS OF THE ANTIARRYTHMIC DRUGS IN THE CLINCAL PRACTICE IN SPAIN Pastor Fuentes A1, Moreno Reviriego S2, Garcia Coscolin T3, Pérez Alcántara F4 1 Universitary Hospital, Getafe, Madrid, Getafe, Madrid, Spain; 2Universitary Hospital La Paz, Madrid, Madrid, Spain; 3Sanofi-Aventis, Madrid, Madrid, Spain; 4Oblikue Consulting, Barcelona, Spain OBJECTIVES: Antiarrhythmic drugs (AADs) are considered the first-line treatment of atrial fibrillation (AF). However, they present a high adverse effects (AEs) rate. The aim of the present study is to know the health resources and the costs associated with the management of AEs of main AADs. METHODS: The incidence of AE’s of main AADs (amiodarone, sotalol, flecainide and propafenone) has been obtained from the Summary of product characteristics or correspondent clinical trials. The use of health resources associated with the management of the AEs has been evaluated by a panel of cardiology experts. Finally, all these clinical data regarding AE’s incidence derived, has been combined with economic data from the Spanish literature and e- salud data,an Spanish database about health care costs. The study has been carried using National Health System perspective. RESULTS: The most expensive AEs have been pulmonary fibrosis: c2,177.49, cardiac events (mainly tachycardia or bradycardia): c1,422.08, and endocrine disorders (as hyperthyroidism: c244.06c and hypothyroidsm: c239.06 c). When costs are analyzed in relation with the drug to which are associated, results are as follow: sotalol has been associated with the highest costs: c269.42, followed by flecainide: 132.25 c, and amiodarone: 127,82, whereas propafenone has been associated to the lowest cost: c48.09. However, although amiodarone is not the AAD associated with the highest cost, considering current Spanish AADs market, it is the one that has more economic repercussion. CONCLUSIONS: Current AADs may cause AEs and their management is related to health resources consumption. Pulmonary and cardiac events have been associated to the major cost. Considering current clinical practice in Spain, amiodarone is the treatment that supposes a major budget impact for the National Health Service. PCV61 ECONOMICS ESTIMATION OF SIDE EFFECTS CORRECTION ANTIHYPERTENSIVE MEDICAL PRODUCTS OF INHIBITORS-ACE IN RUSSIAN HEALTH Yagudina R1, Protsenko M2 1 Laboratory of Pharmacoeconomics, Moscow Medical Academy, Moscow, Russia; 2Moscow Medical Academy named after I.M.Sechenov, Moscow, Russia OBJECTIVES: To assess the cost of the basic course of pharmacotherapy of arterial hypertension (AH) of inhibitors-ACE, direct costs associated with the correction of side effects (SE), developing on the background of their application with further pharmacoeconomics evaluation of inhibitors-ACE. METHODS: Cost analysis, modeling. a consideration of the basic price of therapy trade names (TN) inhibitors-ACE INN: enalapril, lisinopril and the calculation of direct costs associated with correction of SE in patients diagnosed with stage II AH. Medical patients’ route, the cost structure 13th Euro Abstracts were modeled by questioning physicians, cardiologists of Institute of Gerontology, Clinical Hospital ¹ 55, Moscow. The study takes into account the cost: symptomatic therapy, specialist consultations, emergency care, hospital bed-days, laboratory and instrumental manipulations. Data registered in RF TN obtained from the site www. regmed.ru, information about prices—www.medlux.ru, www.cardioweb.ru. We took into account recommendations of standards of care for patients with AH, developed by Health Ministry of Russia. The cost of the basic course of pharmacotherapy and PE are estimated at 30-day time period for 1 patient. Average daily dose the drugs accounted for enalapril—5 mg/day, lisinopril—10 mg/day. RESULTS: The development of side effects (cough, arterial hypotension, allergies, headache, etc.) was the cause of drug withdrawal enalapril in 5.3% of patients, lisinopril at 3.8% of patients. The study calculated the cost of courses were the main pharmacotherapy TN lisinopril and enalapril, ranging from 30 to 245 Rubles, corrective therapy—from 284 to 583 Rubles. In the economic evaluation of SE was the least expensive scheme with drug enalapril, which is more advantageous from an economic position. CONCLUSIONS: The economic analysis of side effects developing on the background of the basic course of pharmacotherapy, is relevant assessment to improve the quality and credibility of pharmacoeconomics studies of drugs. PCV62 UPTAKE AND COSTS OF IMAGING MODALITIES IN PACEMAKERIMPLANTED AND NON-IMPLANTED PATIENTS Busca MR1, Horin F2, Saal G3 1 Medtronic International Trading Sàrl, Tolochenaz, Switzerland; 2Medtronic, Inc., Minneapolis, MN, USA; 3HealthEcon Write Ltd, Solihull, UK OBJECTIVES: Magnetic resonance imaging (MRI) is considered the gold standard for imaging of the brain, spinal cord, musculoskeletal system, and complex cardiac malformations. However, pacemaker implant is a contraindication to MRI, and thus a barrier to access in pacemaker-implanted patients for this essential diagnostic technology. An analysis of Medicare fee-for-service data was conducted to estimate the difference in MRI uptake rates of pacemaker-implanted and non-implanted Medicare benficiaries. METHODS: The data comprised the fee-for-service portion of the 2008 Medicare patient population. Two issues were examined: the prevalence of the diseases for which MRI is the preferred imaging modality, and the uptake rates of all imaging modalities for MRI-indicated beneficiaries with pacemaker implants compared with those having no implants. For each of diseases for which MRI is the preferred modality we also identified any trade-offs between lower MRI rates and higher rates for other imaging modalities in pacemaker-implanted and non-implanted patients, and potential cost implications substituting MRI with other imaging modalities. RESULTS: The Medicare data indicated MRI uptake was nil in the pacemaker-implanted population where 13% of patients without any implant received MRI in 2008 suggesting that clinical practice is in line with the contraindication for MRI in pacemaker patients. Consequently, uptake of other imaging approaches including CT, ultrasound, nuclear, x-ray was greater in the MRI-indicated cohort of pacemaker patients (54%, 65%, 25%, 82%, respectively) compared to the similar non-implanted cohort (38%, 48%, 18%, 73%, respectively). It was estimated that access to MRI (i.e. patients implanted with MRI-compatible pacemakers) would generate up to 21% reduced diagnostic costs compared with no access to MRI (i.e. patients implanted with older generation pacemakers). CONCLUSIONS: Increased access to the diagnostic superiority of MRI may provide a more efficient allocation of diagnostic resources for pacemaker patients; this increased access can be provided with the availability in the market of new pacemakers engineered for MRI compatibility. PCV63 COST OF ABLATION PROCEDURES FOR ATRIAL FIBRILLATION: RESULTS FROM THE AIAC SURVEY Berto P1, Themistoclakis S2, Tritto M3, Aiac WGAA4 1 Pbe Consulting, Verona, Italy; 2Ospedale dell’Angelo, Mestre (VE), Italy; 3Istituto Clinico Mater Domini, Castellanza (VA), Italy; 4Italian Society of Cardiac Pacing and Arrhythmology, Roma, Italy OBJECTIVES: As part of a broader HTA project on ablation procedures for atrial fibrillation (A-AF), objective of this study was to quantify consumption of resources and calculate cost of A-AF in comparison with current reimbursement in Italy. METHODS: A questionnaire was sent to the Italian electrophysiology (EP) centres identified by AIAC (Italian Society of Cardiac Pacing and Arrhythmology) to collect, 2008 data on: number of procedures; use of pre-post ablation diagnostics, consumables, electro-anatomic mapping, anaesthesia/analgesia, EP laboratory occupancy time, staff employed; EP laboratory equipment; duration of hospitalization. Bottomup costing of resources was performed based on average costs from a sub-sample of 4 hospitals, national tariffs or published data. Average procedure cost was calculated based on mathematical (m-mat) and geometrical (m-geo) means. Final endpoint was the comparison between DRG-518 regional reimbursement values and full hospital costs for ablation treatment. RESULTS: A total of 52/87 (60%) AIAC centres replied to the questionnaire, reporting 33,745 EP procedures, of which 4,561 (13.52%) were A-AF (min = 3; max = 1091; average = 88; SD = 159, median = 49; IQR = 22–81). Production cost for A-AF amounted to c9455 (m-mat) and c8868 (m-geo): consumables (ablation and diagnostic catheters, transeptal needles, transeptal sheaths, and other devices) accounted for 61% of total cost; hospital stay 15%; intra-procedural costs (room occupancy, staff, anaesthesia and equipment) 20%; pre-procedural tests 4%. Difference of DRG-518 tariff vs. hospital incurred costs: m-mat c−4079 (−43%) or m-geo c−3492 (−35%). Cost of production appears more under-remunerated in Lazio (−54%), Lombardia (−46%), Veneto, Umbria (−45%). The imbalance was 13th Euro Abstracts −42%/−44% when centres were grouped by the number of A-AF procedures performed. Total cost of A-AF (including 2 years post-procedure follow-up) was c10,163 (m-mat) or c9,632 (m-geo). Cost of follow-up represented only 8% on total cost of A-AF. CONCLUSIONS: The cost of AF ablation procedures is significant and is largely under-funded by the DRG tariff, which showed inadequate to remunerate hospitals in all Italian Regions. PCV64 COST CONSEQUENCES AND RESOURCE UTILIZATION IN TRANSCATETHER AORTIC VALVE IMPLANTATION (TAVI) AS A SUBSTITUTION FOR HIGH RISK SURGICAL PATIENTS Webb PP1, Carroll GD1, Smith D2 1 Welsh Health Specialised Services Committee, Caerphilly, UK; 2Morriston Hospital, Swansea, UK OBJECTIVES: TAVI is a novel interventional procedure and is considered as an alternative to surgery in high risk patients with severe aortic stenosis. For commissioning purposes, a cost consequences model was derived to look at the impact of substituting TAVI for high risk conventional surgery patients. METHODS: The cost consequence included hospital LOS on CITU, HDU and ward and pacemakers implantation costs. Twenty-five patients were retrospectively identified by clinicans with TAVI experience who might have been eligible to receive TAVI in the previous year of introduction of TAVI on the market. The patient specification was aligned with the TAVI device CE marking criteria. The average TAVI patient cost was obtained from a manufacturer sponsored economic model which used UK patient level data from their sponsored registry. The hospital’s local database was used for unit costs. Costs were calculated based on mean values for CITU (critical intensive care unit), HDU (high dependency unit), ward and pacemaker implantation. We also calculated the mean cost per case based on the total length of stay (LOS) for CITU, HDU ward and pacemaker implants of each of the isolated AVR and AVR plus valvular disease (AVR + CABG) population. RESULTS: The total cost based on mean LOS values was £21,460 for isolated AVR and £23,032 for AVR + CABG. The cost per TAVI case based on total LOS was £24,721 compared to £49,982 for AVR + CABG. CONCLUSIONS: The cost consequence model was sensitive to long LOS on CITU and HDU in the AVR plus CABG group. Using a cost consequences model TAVI was at worst cost neutral and at best most likely to deliver substantial savings in this centre in this well defined patient population. This finding is significant in assessment of the “real” cost impact for substituting TAVI for high risk conventional surgery. PCV65 HEALTH ECONOMIC ASSESSMENT OF FERRIC CARBOXYMALTOSE IN PATIENTS WITH IRON DEFICIENCY AND CHRONIC HEART FAILURE IN THE FAIR-HF TRIAL Gutzwiller FS, Blank PR, Schwenkglenks M, Szucs TD Institute of Pharmaceutical Medicine/ECPM, University of Basel, Basel, Switzerland OBJECTIVES: Chronic heart failure (CHF) is associated with poor quality of life resulting from physical and psycho-social limitations. The FAIR-HF trial showed clinical and quality of life benefits of iron deficiency treatment in CHF patients using ferric carboxymaltose (FCM), an intravenous (i.v.) iron preparation. This study evaluated the cost-effectiveness of iron repletion using FCM in CHF patients, from the perspectives of the statutory health insurance (Germany) and the National Health Service (UK). METHODS: Using data from FAIR-HF, a randomized, double-blind, controlled clinical trial (n = 459), per-patient costs and clinical effectiveness of FCM were estimated. Economic assessment was based on published costs associated with New York Heart Association (NYHA) functional classes. Effectiveness was assessed as the number of QALYs gained, derived from EQ-5D scores. The ICER of FCM was determined compared to placebo. Time horizon of this within-trial analysis was 24 weeks. RESULTS: In the FAIR-HF trial, NYHA classes were significantly improved in the FCM group compared with placebo (P < 0.001). Estimated per-patient costs (excluding iron costs) were c2625 and c2919 (Germany), and £4155 and £4621 (UK), for FCM and placebo arms, respectively. Based on the reimbursed price for FCM in Germany and UK (c28 and £19 per 100 mg iron) and the mean iron dosage of the clinical study (1850 mg) plus administration costs, a net investment of c530 and £68 per patient would be required. FCM resulted in an estimated gain of 0.0254 QALYs over the study period. The ICER of FCM ranged from c20,872 to £2,682 per QALY gained for the FAIR-HF dosing regimen, compared to placebo. ICER-differences result from disparities in administration and medication costs. CONCLUSIONS: From the German and UK payers’ perspective, managing iron deficiency in CHF patients using i.v. FCM can be considered cost-effective. Improved symptoms and better quality of life contribute to economic benefits seen with FCM. PCV66 ECONOMIC PERFORMANCE OF DABIGATRAN ETEXILATE FOR PRIMARY VTE PREVENTION FOLLOWING TOTAL HIP AND KNEE REPLACEMENT SURGERY IN ITALY Zaniolo O, Pradelli L Adres srl, Torino, Italy OBJECTIVES: To estimate cost/effectiveness and cost/utility of dabigatran etexilate (DBG) compared to standard care for the prevention of venous thromboembolism (VTE) secondary to total hip replacement (THR) or total knee replacement (TKR) in Italy. METHODS: A decision analytic, Markov-chain based model originally developed for the UK has been adapted to the Italian context. Clinical outcomes, including A353 incidences of VTE and treatment-related adverse events, were extrapolated from headto-head, phase III trials of DBG vs. enoxaparin. For the other low molecular weight heparins (LMWHs), indirect comparisons were performed on the basis of equal effectiveness assumptions. The adaptation involved cost and demographic characteristics, leaving clinical and utility data unvaried. Costs are taken from national observational studies, where available. Otherwise, current prices and tariffs are applied. Resources consumption where derived from practice guidelines or taken from those estimated for the UK model. According to the prevalent national practice, extended prophylaxis has been considered for both surgical procedures. Time horizon of the analysis is patients’ lifetime. RESULTS: Compared to LMWHs, DBG is associated to an expected increase of 0.019 LYs and 0.014 QALYs per THR patient and of 0.024 LYs and 0.019 QALYs per TKR patient. DBG-related cost is lower than competitors in both procedures, with a mean difference ranging between c82 and c109 for THR, and 100 and 135 for TKR, depending on the considered comparator. Higher acquisition costs for DBG are completely offset and inverted by avoided administration expenses and, less importantly, by savings on VTE management. Probabilistic sensitivity analysis estimates that, for a willingness to pay threshold of £30,000 (~c33,500)/ QALY, DBG is associated to a probability of being cost/effective of about 98% for THR and of 90% for TKR. CONCLUSIONS: On average DBG dominates LMWHs, as it’s expected to be cost-saving and non-inferior in terms of efficacy and safety. PCV67 COST-EFFECTIVENESS OF FONDAPARINUX VS. ENOXAPARIN IN NON-ST-ELEVATION ACUTE CORONARY SYNDROMES IN RUSSIA Vorobyev PA1, Borisenko O1, Lesnicheva M1, Shiganov S2, Lomakin A3 1 Russian Society for Pharmacoeconomics and Outcomes Research, Moscow, Russia; 2Abakan City Hospital, Abakan, Russia; 3GlaxoSmithKline Russia, Moscow, Russia OBJECTIVES: To assess cost-effectiveness of fondaparinux versus enoxaparin in patients with non-ST-elevation acute coronary syndrome in Russia. METHODS: A decision tree model was created incorporating the outcomes associated with 2 antithrombotic approaches from OASIS-5 trial: enoxaparin (1 mg/kg every 12 hours) versus fondaparinux (2.5 mg/day). Probabilities of complications (e.g., myocardial infarction, stroke, major bleeding at 9, 30 and 180 days) were calculated. Data on resource utilization (staff, drugs, materials, laboratory, and equipment), diagnosis and treatment came from the hospital survey. Resources were valued in internal hospital prices based on obligatory medical insurance tariffs (as of 2008). The incremental cost-effectiveness ratio for fondaparinux per death avoided and per major bleeding episode avoided versus enoxaparin was quantified. Sensitivity analysis was performed. RESULTS: The cost analysis showed that fondaparinux would generate a cost saving of 1634 RUR (c43), 1902 RUR (c50) and 2167 RUR (c57) per patient at 9, 30 and 180 days respectively. When cost and clinical results were extrapolated to costeffectiveness, fondaparinux was dominant (less costly and more effective in terms of death avoided and major bleeding episode avoided). The sensitivity analysis proved that costs of pharmacotherapy with fondaparinux and enoxaparin had the biggest impact on results. CONCLUSIONS: Fondaparinux is a more cost-effective option than enoxaparin in treatment of patients with non-ST-elevation acute coronary syndrome from the hospital perspective in Russia. PCV68 POTENTIAL COST-EFFECTIVENESS OF A BIOMARKER TEST TO RECLASSIFY PATIENTS WITH AN INTERMEDIATE RISK BASED ON THE FRAMINGHAM RISK SCORE INTO A LOWER OR HIGHER CATEGORY TO OPTIMIZE STATIN THERAPY Burgers LT, Redekop WK, Severens JL Erasmus University Rotterdam, Rotterdam, The Netherlands OBJECTIVES: The Framingham Risk Score is a well-accepted tool to estimate the 10-year risk of developing coronary heart disease (CHD). However, many patients fall into the intermediate risk category. Improved discrimination within this category is necessary to prevent CHD and side-effects of statins therapy efficiently. This economic evaluation calculated the potential lifetime cost-effectiveness of a novel biomarker test which helps to decide whether patients with an intermediate risk should be treated with statins. METHODS: Prognosis of patients with an intermediate risk was simulated with a Markov chain Monte Carlo model to estimate the potential lifetime costs and effects (life-years (LY)) for three strategies: treat all with statins, treat none with statins or use a test to select patients for statin treatment. Costs were calculated for the The Netherlands using a health care sector perspective. Values for all input parameter were derived from the literature. RESULTS: A strategy using a perfect test for a 55-year old man would be slightly more expensive than the treat-none option (c1966 vs. c1941) but less expensive than the treat-all option (c5374). The test and the treat-all option would be equally effective (24.45 LY) and more effective than the treat-none option (24.3 LY). An ICER of c170 versus treat-none indicates that it is a biomarker test with great potential. Results were sensitive to uncertainties regarding model parameters such as the sensitivity, specificity and costs of the test, as well as CHD risk, and the costs and effectiveness of statins. CONCLUSIONS: A test to reclassify patients in the Framingham intermediate risk group into higher and lower risk categories has the potential to optimize cost-effectiveness by preventing CHD and reduce the risk of drug side-effects. Values used in this model (e.g., test sensitivity and specificity) can be adjusted wherever needed to determine whether continued development of a biomarker is worthwhile. A354 PCV69 COST-EFFECTIVENESS OF EPLERENONE COMPARED WITH SPIRONOLACTONE IN PATIENTS WITH HEART FAILURE AFTER ACUTE MYOCARDIAL INFARCTION Inês M1, Soares M2, Vitorino R1 1 Pfizer Portugal, Porto Salvo, Oeiras, Portugal; 2University of York, York, Yorkshire, UK OBJECTIVES: Aldosterone blockers (AB) are recommended for use in heart failure (HF) and HF post acute myocardial infarction (AMI) in patients with left ventricular dysfunction. Although only eplerenone is licensed for post-AMI HF treatment, less expensive spironolactone is used in some clinical settings. This study aims to evaluate cost-effectiveness of eplerenone compared with spironolactone in Portuguese patients with HF post-AMI. METHODS: A Markov model was developed to predict life-years (LY), quality adjusted life-years (QALY) and associated costs of AB treatment over patients life time. Estimates of all cause mortality relative risks were obtained via a meta-regression of AB published trials. Transition probabilities were derived from EPHESUS (Eplerenone Post-Acute Myocardial Infarction Heart Failure Survival Study), adjusted by a Weibull function with meta-regression estimates. AB adverse events (hyperkalemia, gynescomatia) rates were derived from a published meta-analysis. AB class effect was assumed regarding non-fatal hospitalization rates. Resource use and discontinuation rates were elicited through a panel of six cardiologists with extensive clinical experience. Unit costs and other cause mortality rate were extracted from Portuguese official sources. Utilities were obtained using a regression model published in the literature derived from EPHESUS EQ-5D data. Societal perspective was adopted and both costs and effectiveness were discounted at 5%. RESULTS: Average cost per patient for eplerenone and spironolactone treatment were c25,907 and c20,963, respectively. Average effectiveness gained with eplerenone was 0.44LY and 0.41QALY, meaning the incremental cost-effectiveness ratios were c11,309/LY and c12,040/QALY. Although indirect comparison of eplerenone with spironolactone introduces some uncertainty for relative effectiveness and cost-effectiveness: probabilistic sensitivity analysis revealed eplerenone to be cost-effective in 59% of Monte Carlo iterations at a willingness to pay of c35,000/QALY. CONCLUSIONS: Compared with spironolactone, eplerenone treatment in patients with HF post-AMI is effective in reducing mortality, improving quality of life and is cost-effective by commonly used criteria in Portugal. PCV70 COST SAVINGS OF ENOXAPARIN OVER UFH IN MEDICAL PATIENTS IN CANADIAN HOSPITAL SETTING Wakeford C1, Mittmann N2 1 sanofi-aventis Canada Inc., Laval, QC, Canada; 2Sunnybrook Health Sciences Centre, Toronto, ON, Canada OBJECTIVES: Enoxaparin, a low molecular weight heparin, was approved for use in medical patients in Canada in 2001, however, its use in hospitals has been relatively low. Low utilization is thought to be due to the perception by hospital administration of increased drug costs when compared to unfractionated heparin (UFH). Examination of the cost of clinical events for thromboprophylaxis has demonstrated that enoxaparin may be cost savings. METHODS: A decision tree model was developed to the compare clinical results of four thromboprophylaxis regimens: enoxaparin 40 mg once daily, UFH 5000 International Units (IU) twice daily, UFH 5,000 IU three times daily and no prophylaxis. The main clinical outcomes in the model included DVT, PE, major bleeds, heparin-induced thrombocytopenia and death and were based on published literature Recurrences of the first four endpoints were also modelled. Canadian treatment pattern data was based on one academic institution in Toronto, Canada. Costing (2010 Canadian $) for each state was also developed for direct medical costs. Within the model, parameters such as cost of thromboprophylaxis, number of patients, duration of treatment, length of stay, among many other variables were varied to reflect a hospital specific analysis. RESULTS: When combining clinical and economic data in a Canadian hospital setting, use of enoxaparin was cost-saving in a majority of situations. In the reference case, a typical hospital setting switching from UFH to enoxaparin resulted in an overall cost savings to the hospital, wherein a population of 1000 annual medical patients would realize a savings of $90,000CAD. Sensitivity analysis showed that enoxaparin was not cost savings when medically ill patients did not receive any thromboprophylaxis. CONCLUSIONS: Enoxaparin is a clinically safe and efficacious thromboprophylaxis regimen in medically ill patients. The model and follow-up analysis demonstrate that it is also a cost-saving choice of thromboprophylaxis within the Canadian hospital setting. PCV71 IS COREVALVE A COST-EFFECTIVE TREATMENT FOR PATIENTS WITH SEVERE AORTIC STENOSIS WHO SATISY ALL THE CRITERIA FOR CONVENTIONAL AORTIC VALVE REPLACEMENT? Eaton JN, Mealing S, Watt M, Sculpher M Oxford Outcomes Ltd, Oxford, Oxon, UK OBJECTIVES: Aortic Stenosis (AS) is a severe cardiovascular condition; the treatment of which involves open surgery and cardiopulmonary bypass to replace the damaged aortic valve (AVR). AVR is invasive and carries a significant risk for patients with a high risk of perioperative mortality. Transcatheter aortic valve implantation using “CoreValve,” is less invasive, allowing for the percutaneous implantation of a replacement valve in this patient group. However, the acquisition cost of CoreValve is greater than AVR, therefore it is important to assess whether CoreValve is a cost-effective alternative. METHODS: A 20-year Markov model was developed in Microsoft Excel. 13th Euro Abstracts Treatment options were CoreValve and AVR with parameters derived from published literature. Both products were assumed to have the same long term impact on mortality. All costs were taken from the most recent published sources. Decrements were applied to age-specific EQ-5D population norms to generate QALYs. Extensive probabilistic and deterministic sensitivity analyses were performed to identify key costeffectiveness drivers. All costs and benefits were discounted at 3.5% p.a. RESULTS: The model was very sensitive to changes in the short-term relative risk (RR) of mortality. In the base-case a short term RR of 0.57 was estimated from the available literature, the ICER was £25,229. However, the short-term mortality estimate for AVR was highly uncertain. Assuming a 0.75 RR the ICER is £34,603 and at 1.0 the ICER is £81,011. The model was also highly sensitive to changes in long term RR, device cost and time horizon. Assuming a threshold value of £30,000 per QALY gained, the probability that CoreValve is cost-effective is 0.54. CONCLUSIONS: Assuming that the RR of short-term mortality is at least 0.67 CoreValve is a cost-effective alternative for the treatment of severe AS in high risk patients. PCV72 COST-EFFECTIVENESS OF EXTENDED-RELEASE NICOTINIC ACID/ LAROPIPRANT AS MONOTHERAPY VERSUS EZETIMIBE MONOTHERAPY IN PATIENTS WITH MIXED DYSLIPIDAEMIA IN SCOTLAND Jameson K1, Davies GM2, Ambegaonkar BM3, Sazonov V3, O’Regan C1 1 MSD Ltd., Hoddesdon, UK; 2Merck & Co., Inc., Upper Gwynedd, PA, USA; 3Merck & Co., Inc., Whitehouse Station, NJ, USA OBJECTIVES: Although statins are the mainstay of cholesterol management in Scotland, there remains a group of patients who are ineligible for such treatment due to contraindication or intolerance. This analysis estimates the cost-effectiveness of extended-release nicotinic acid/laropiprant 2 g/40 mg (ERN/LRPT), versus ezetimibe 10 mg, as monotherapy for the treatment of mixed dyslipidaemia in Scotland in patients not currently receiving statin therapy. METHODS: The analysis was based on a previously published Markov model. Risk equations, based on the Framingham Heart Study, were used to model CHD-event rates in patients with established CHD, diabetes mellitus (DM) or a 10-year CHD-risk ≥20%. The model was run on a cohort of 2976 patient profiles extracted from the General Practice Research Database. All patients had abnormalities in low-density-lipoprotein cholesterol (LDL-C ≥ 2.0 mmol/L), high-density-lipoprotein cholesterol (HDL-C < 1.0 mmol/L [men], 1.2 mmol/L [women]) and triglycerides (TG > 1.7 mmol/L). ERN/LRPT efficacy data were taken from the results of a published clinical trial, and ezetimibe data taken from a meta-analysis of monotherapy trials. Results were reported as the cost per qualityadjusted life-year (QALY) gained. RESULTS: When compared to the use of ezetimibe, ERN/LRPT led to an increase in both QALYs and costs across the three patient-risk sub-groups. In patients with CHD, the base-case analysis estimated an incremental cost-effectiveness ratio (ICER) of £7559. Base-case results in patients with DM and patients facing a 10-year CHD risk ≥20% were £9,361 and £11,980, respectively. ICERS in the majority of sensitivity analyses did not vary by more than +/− £3000 from the base case. CONCLUSIONS: When compared to ezetimibe, ERN/LRPT is projected to be cost-effective for the management of lipids in patients for whom the addition of statin is inappropriate or not tolerated, as reflected in the guidance issued by the Scottish Medicines Consortium. PCV73 ECOMOMIC EVALUATION OF COMBINATION THERAPY FOR HYPERTENSION IN HIGH RISK PATIENTS IN SWEDEN Brown R,1 Brede Y,3 Getsios D,2 Tambour M,4 Falvey H3 1 United BioSource Corporation, London, UK, 2United BioSource Corporation, Lexington, MA, USA, 3Novartis Pharma AG, Basel, Switzerland, 4Novartis Sverige AB, Taby, Sweden OBJECTIVES: The ACCOMPLISH trial demonstrated that an ACEi/CCB combination (benazepril/ amlodipine) significantly reduced cardiovascular events in high risk hypertension patients compared to an ACEi/HCTZ combination (benazepril/ hydrochlorothiazide). ACCOMPLISH trial was the first to compare fixed dose combination therapies and examine longer term outcomes in this patient population. This study evaluated health and economic outcomes based on robust findings from ACCOMPLISH, and extended this analysis to an ARB/CCB combination (valsartan /amlodipine). METHODS: ACCOMPLISH evaluated over 11,500 patients with hypertension at high risk for cardiovascular events. Trial data on cardiovascular events, diabetes, and renal failure endpoints, compliance and dosing levels were used to populate a compartmental cohort model which followed patients over 4 years. Cost data from Sweden (reported in 2008 Euros) were assigned to acute events and follow-up care, as well as hypertension-related treatment. a dose equivalency analysis of valsartan and benazepril was used to assign appropriate drug costs for the extension of the analyses to valsartan/amlodipine. The daily cost of benazepril/hydrochlorothiazide was c0.21, c0.16 for benazepril/amlodipine, and ranged between c0.90 and c1.10 for valsartan/ amlodipine, depending on dose. All outcomes were discounted at 3% per annum. RESULTS: Total costs over 4 years averaged c4554 per patient on benazepril/hydrocholorothiazide. With a lower drug cost and 45 fewer acute events per 1000 treated patients, costs were c322 lower per patient on benazepril/amlodipine. Assuming same outcome as for benazepril/amlodipine, valsartan/amlodipine was associated with incremental drug costs of c963 per patient, and c693 in overall costs, resulting in a cost per event avoided of c15,543. CONCLUSIONS: Based on the results of ACCOMPLISH, the combination of benzepril/amlodipine is clearly preferable to benazepril/ 13th Euro Abstracts hydrochlorothiazide. Using the ACCOMPLISH results for valsartan/amlodipine, it is likely to have reasonably low costs per event avoided for the treatment of high risk patients with hypertension in Sweden. PCV74 ONE-YEAR COST-EFFECTIVENESS OF CYTOCHROME P450 2C19 GENOTYPE-GUIDED ANTIPLATELET THERAPY IN PATIENTS WITH ACUTE CORONARY SYNDROMES IN THE UNITED KINGDOM Charland SL1, Agatep B1, Malone D2, Stanek E1 1 Medco Health Services, Inc., Bethesda, MD, USA; 2University of Arizona, Tucson, AZ, USA OBJECTIVES: Cytochrome P450 2C19 (CYP2C19) genotype has been shown to affect cardiovascular (CV) outcomes for clopidogrel but not prasugrel. This study evaluates the incremental cost-effectiveness ratio (ICER) of CYP2C19-guided vs. routine antiplatelet therapy in acute coronary syndrome (ACS) patients in the UK. METHODS: We constructed a literature-based, decision analytic, Markov model to estimate the annual cost-effectiveness of CYP2C19-guided aspirin plus either clopidogrel or prasugrel therapy vs. no genotyping. Post-initial ACS CV events were based on the TRITON-TIMI 38 study and genetic substudy. Cost data sources were: National Health Service (NHS) reference cost for 2008–09—nonfatal MI and stroke, CV death, intracranial hemorrhage, other life-threatening bleed, and minor bleed; Drug Tariff 2009—drugs; www.genetic-health.co.uk/—CYP2C19 genotyping; or USbased reference pricing converted to £ using appropriate exchange rates -monthly CV disease maintenance cost. Disease-state utilities were obtained from published sources. The model allowed for clopidogrel/prasugrel discontinuation and aspirin monotherapy. Model sensitivity was assessed using 1-way analysis of parameters varied by quartile or at least +/− 25%. RESULTS: The analysis demonstrated an increase in incremental cost (£81); greater incremental QALY (0.05); and an ICER £1529/QALY for CYP2C19 genotype-guided therapy over 12 months. The model was most sensitive to monthly CV care cost, NFMI cost, proportion of patients on clopidogrel, and lifethreatening bleeding cost. The model was least sensitive to the cost of clopidogrel, prasugrel, or CYP2C19 genotyping. CONCLUSIONS: The model-based ICER of £1,529/QALY for the CYP2C19 testing strategy is significantly less than the UK threshold of £20,000 that is considered good value. CYP2C19 genotype-guided clopidogrel or prasugrel therapy is cost-effective for up to 1 year in ACS patients in the UK. PCV75 CANADIAN COST-EFFECTIVENESS ANALYSIS OF DRONEDARONE VERSUS OTHER ANTI-ARRHYTMIC DRUGS IN PATIENTS WITH PAROXYSMAL AND PERSISTENT ATRIAL FIBRILLATION Nilsson J1, Åkerborg Ö1, Lindgren P1, Bascle S2 1 i3 Innovus, Stockholm, Sweden; 2Sanofi-Aventis, Paris, France OBJECTIVES: Dronedarone is a novel anti-arrhythmic drug (AAD) that, unlike other AADs, was shown to reduce cardiovascular hospitalizations or death in the ATHENA clinical trial. In addition, dronedarone reduces AF recurrence, maintains rate control, and has a favorable safety profile with low pro-arrhythmic and organ-toxicity profile. The objective of this study was to construct a health economic model to assess the cost-effectiveness of dronedarone vs. other AADs in a Canadian setting. METHODS: A state transition model evaluated through patient-level simulation has been developed using Microsoft Excel. It allows comparisons over varying time horizons and treatment durations, and consists of health states for: treatment, off treatment, symptomatic AF recurrences, stroke, acute coronary syndromes, coronary heart failure and death. Transition probabilities were derived from the patient level data from the ATHENA trial, and relative risks between dronedarone and three commonly used comparators (amiodarone, sotalol and flecainide) identified by clinical experts were derived from a mixed treatment comparison (systematic review) of published clinical trials published between 1980 and 2009. Patients discontinuing treatments were assumed to progress according to the rates in the standard of care arm of ATHENA. Costs of monitoring and initiation were taken into account. Costs were applied to each adverse event (AE) observed (Canadian Costs [C$] 2007). Effectiveness was expressed as QALYs, using preference based utility weights for health states based on published data. Discounting was 5% and a lifetime horizon was taken. RESULTS: The model predicts higher quality adjusted survival for patients on dronedarone: between 1.13 and 2.01 QALYs depending on comparator. In Canada, the resulting ICERs (per QALY) are C$5600 compared to amiodarone, C$5300 compared to flecainide, and C$5300 compared to sotalol. Results were sensitive to differences in risk of mortality between treatment groups. CONCLUSIONS: Dronedarone represents a cost-effective treatment for patients with atrial fibrillation in Canada. PCV76 COST-EFFECTIVENESS ANALYSIS OF DRONEDARONE IN PATIENTS WITH ATRIAL FIBRILLATION IN MEXICO: A WITHIN TRIAL ANALYSIS BASED ON ATHENA TRIAL Martinez E1, Lemus A2, Soto H3 1 Cardiology Hospital CMN Siglo XXI, Mexico City, D.F., Mexico; 2Sanofi-Aventis, D.F., Mexico City, Mexico; 3ESEA, Mexico City, D.F., Mexico OBJECTIVES: To perform a cost-effectiveness analysis (CEA) for the use of dronedarone in patients with atrial fibrillation (AF) in order to prevent hospitalizations due to cardiovascular events or death (HCED), from the public health care system perspec- A355 tive in Mexico. METHODS: A CEA was made based on the clinical information from the multicenter, randomized, clinical study ATHENA, where dronedarone was compared to placebo on top of standard treatment in patients with AF. Overall, 2301 patients were randomized to dronedarone, and 2327 to placebo. The primary clinical endpoint was HCED. Only direct medical health care costs were calculated. The health resource utilization was elicited from the ATHENA trial. The unit costs of each event were obtained from the medical literature and/or validated by local experts, whenever information was not available. Most of the cost information is based on IMSS (Social Security) figures, and updated to year 2009 (1c = MX$17.05). a discount rate of 5% was used. An incremental cost-effectiveness analysis was performed complemented by a deterministic sensitivity analysis (DSA) to assess robustness of the model. RESULTS: Patients randomized to dronedarone experienced 1190 events of HCED (average rate 0.51 [CI 95%: 0.47–0.55]) while patients in the placebo group had 1601 (average rate 0.69 [CI 95%: 0.64–0.74]), or −415 events, 18% less hospitalizations (CI 95%: 12–25%) for the dronedarone group. The average cost per patient in the dronedarone group was c3028 as compared to the placebo group of c2,941, yielding a cost difference of c87.4, and an avoided incremental cost per HCED of c477.00 of dronedarone vs. the placebo group. The DSA shows the analysis is robust. CONCLUSIONS: According to the ATHENA trial, dronedarone is a cost-effective treatment option for the reduction of HCED from the Mexican perspective. Dronedarone’s value could be enhanced if indirect costs averted from the decreased rates of HCED included. PCV77 A SIMULATION MODEL TO ASSESS COST-EFFECTIVENESS OF STATINS IN HIGH RISK PATIENTS WITH ELEVATED LDL-C IN SPAIN Darba J1, Restovic G2, Kaskens L2, Tunceli K3, Plans P4 1 Universitat de Barcelona, Barcelona, Spain; 2BCN Health, Barcelona, Spain; 3Merck, Whitehouse Station, NJ, USA; 4Generalitat de Catalunya, Barcelona, Spain OBJECTIVES: The aim of this study is to estimate cost-effectiveness of lowering lowdensity lipoprotein cholesterol (LDL-C) with statin monotherapy in patients with elevated LDL-C with two or more cardiovascular risk factors or either coronary heart disease (CHD) in Spain. METHODS: A Markov model was developed to represent the transition of a cohort of patients with elevated LDL-C or with CHD at risk of a cardiovascular event (CVE) through four health states: patients with LDL-C, CVE, death by CVE and death by other causes. Probabilities of a CVE in females and males were determined, based on CHD risks estimated trough locally-adjusted Framingham risk equations using data from the DORICA and PRIMULA study. LDL-C lowering efficacy of statins, mortality, and health-state utilities were obtained from published scientific literature. Cardiovascular risk factors included were age, systolic blood pressure, diabetes, smoking and high-density lipoprotein cholesterol (HDL-C). Treatment and CVE direct medical costs were obtained from a medication database and DRGs for public hospitals in 2009 in Spain. Deterministic results were estimated and a probabilistic sensitivity analysis was conducted. Results were expressed as expected cost per quality adjusted life-years (QALYs) gained. RESULTS: In deterministic analyses, expected costs per patient per year at age of 40 were higher for patients with 2 or more cardiovascular risk factors who were not treated than those who were treated (female: c41,300 vs. c40,106; male: c22,160 vs. c18,333). Effectiveness was higher for treated patients in both genders (female: 0.17 QALY; male: 0.37 QALY). Similar results were found for patients with CHD (female: c35,706 vs. c34,664, 0.10 QALY; male: c16,892 vs. c16,073, 0.12 QALY). CONCLUSIONS: From the perspective of the Spanish health care system, treatment with statin monotherapy is considered to be cost-effective versus no treatment in female and male patients with 2 or more risk factors or CHD. PCV78 COST-EFFECTIVENESS ANALYSIS OF IVABRADINE IN CHRONIC STABLE ANGINA PATIENTS IN AN AUSTRIAN SETTING Schwarz B1, Wild R2 1 Center for Public Health, Medical University Vienna, Baden bei Wien, Austria; 2Servier Austria GmbH, Vienna, Austria OBJECTIVES: High resting heart rate (HR) has been progressively accepted as a modifiable cardiovascular risk factor. Ivabradine (ProcoralanÒ) is a specific HR lowering agent. This study aimed estimating the cost-effectiveness of ivabradine in stable angina patients with a normal sinus rhythm and a resting HR above 70 beats per minute (bpm) from the Austrian health care perspective: 1) versus generic diltiazem when beta-blockers (BB) are contra-indicated or non tolerated; and 2) in combination with generic atenolol versus generic atenolol alone. METHODS: A Markov chain Monte Carlo stochastic simulation model was used to estimate the influence of HR lowering in cardiovascular morbidity and mortality and its economic consequences. Treatments considered are ivabradine, 7.5 mg twice a day, diltiazem, 240 mg once a day and atenolol 50 mg once a day. HR distribution, survival and time to hospitalization were modelled as weibull functions. Events considered were acute myocardial infarction, stroke, heart failure, death and revascularization procedures. Only direct medical costs were included. Effectiveness was measured in quality-adjusted life-years (QALYs). Time horizon was set at 20 years and discount rates for costs and effectiveness were 3%/year. RESULTS: The between-group difference in HR reduction was −6.4 bpm and −8.8 bpm in favour of ivabradine strategy in targeted patient populations 1) and 2) respectively. Incremental ivabradine strategy cost was c6789 versus generic diltiazem and c6749 versus generic atenolol. Incremental QALYs were 1.067 and 1.076 respectively. Incremental cost-effectiveness ratios for ivabradine strategy were c5800/QALY and c6273/QALY. Deterministic sensitivity analyses showed that A356 ivabradine strategy is cost-effective in both approaches in all cases at a willingness to pay threshold of c22,000/QALY. CONCLUSIONS: In an Austrian setting, ivabradine is a cost-effective treatment in stable angina patients with resting HR > 70 bpm. PCV79 COST-EFFECTIVENESS OF DRONEDARONE IN SOUTH KOREA 1 1 2 3 Jo MW , Kim SH , Choi Y , Kim J 1 University of Ulsan College of Medicine, Seoul, South Korea; 2sanofi-aventis Korea, Seoul, South Korea; 3Pusan National University Yangsan Hospital, Yangsan, Gyeongnam, South Korea OBJECTIVES: Atrial fibrillation (AF) contributes for stroke, sudden death, heart failure, markedly reduced exercise capacity and degraded quality of life. Therefore, an effective treatment of AF is expected to reduce cardiovascular (CV) events and their costs. In 2010, dronedarone has been approved by Korean FDA for risk reduction of CV hospitalization in patients who are in sinus rhythm or who will be cardioverted with relevant conditions. The purpose of this study is to evaluate the cost-effectiveness of dronedarone compared with amiodarone in Korea. METHODS: We used the Markov simulation model for AF patients consisting with health states for AF treatment, off-AF treatment, symptomatic AF recurrences, stroke, acute coronary syndromes, congestive heart failure and death. Transitional probability was obtained from ATHENA trial and published literatures. Patient baseline, drug cost, initiation and monitoring cost of AF treatment, disease state cost and adverse event cost were obtained from national insurance claim database. The effectiveness of amiodarone vs. dronedarone was drawn from results of mixed treatment comparison. Discount rate for cost and effectiveness were applied as 5%. From the societal perspectives, we evaluated cost for life-year gained (LYG) and quality adjusted life-years (QALYs) until patients become 100 years old. Subgroup analysis and sensitivity analysis was performed to deal with uncertainty. RESULTS: In the base-case analysis, the incremental cost-effectiveness ratio (ICER) of dronedarone versus Amiodarone was approximately c2344/LYG (KRW 3.75 million, 1 Euro = 1600 KRW). Results were robust across subgroups. The ranges of ICER in the sensitivity analysis were from around c1875 ~ c3750/LYG (KRW 3 to 6 million/LYG). Fifty percent of simulations in probabilistic sensitivity analysis fall below a willingness-to-pay of about c3750 per QALYs (KRW 6 million per QALYs). CONCLUSIONS: These results showed that dronedarone is to be cost-effective vs. amiodarone for AF patients in Korea. PCV80 THE COST-EFFECTIVENESS OF AMLODIPINE BESYLATE VERSUS PLACEBO FOR THE PREVENTION OF CARDIOVASCULAR DISEASE IN KOREA Lee HY1, Park HJ2, Ko SK2 1 Seoul National University Hospital, Seoul, South Korea; 2Pfizer Pharmaceuticals Korea Ltd, Seoul, South Korea OBJECTIVES: Antihypertensive therapy is a well-established approach to reducing the risk of cardiovascular disease (CVD). Amlodipine besylate, a calcium channel blocker, has been shown to be an effective antihypertensive agent. The objective of this study was to evaluate the cost-effectiveness of antihypertensive therapy with amlodipine besylate for the prevention of CVD in Korea from a health care perspective over a lifetime. METHODS: To estimate long-term cost and effects, a Markov model consisting of nine health states was constructed: Healthy with hypertension, Angina, Myocardial Infarction (MI), Post-MI, Stroke, Post-stroke, CHD death, CVD death, non-CVD cause death. One health state to another can occur with a certain probability at yearly intervals. The incidence of CVD was obtained from published local sources, whereas the risk reductions associated with antihypertensive therapy were taken from the medical literature, selected studies randomly assigned amlodipine besylate or placebo and followed up for at least 1 year. Utility values for CVD and costs of amlodipine besylate were drawn from published literature based on 2005 Korea National Health and Nutritional Examination Survey (KNHANES) data, and Korean pharmaceutical pricing lists, respectively. Costs for CVD were found in published cost-of-illness studies based on local hospital charge data. Patient outcomes were modeled for 45 years, and incremental cost-effectiveness ratios were calculated for amlodipine besylate compared with placebo. RESULTS: For a 55-year-old patient with hypertension, the incremental cost of amlodipine besylate compared with placebo was 3,213,660 Korean won (KW) per patient, although the incremental effectiveness of amlodipine besylate was 0.210 quality-adjusted life-years (QALYs) gained per patient. Therefore, the incremental cost-effectiveness ratios associated with amlodipine besylate were 15,288,941 KW/QALY, compared to no treatment. Sensitivity analyses indicated these results to be robust. CONCLUSIONS: The results from the model indicate that amlodipine besylate provides a cost-effective antihypertensive treatment strategy for the prevention of CVD in Korea. PCV81 THE COMPARATIVE ANALYSIS OF THE EFFECTIVENESS AND THE COSTS OF USING THE LOW-MOLECULAR-WEIGHT HEPARINS AND THE ORAL ANTICOAGULANTS FOR THE TREATMENT OF THE VENOUS THROMBOEMBOLISM IN POLAND Bochenek T, Nizankowski R Jagiellonian University, Collegium Medicum, Krakow, Poland OBJECTIVES: The currently recommended standard treatment of the venous thromboembolism (VTE) are the oral anticoagulants (OA) in the majority of patients and the low-molecular-weight heparins (LMWH) in the selected subpopulations. The 13th Euro Abstracts titration of the OA doses is difficult and often ineffective. The goal of this research was to compare the OA and the LMWH in the VTE treatment in terms of effectiveness, safety and cost-effectiveness. METHODS: The systematic review of the scientific literature comparing the VTE treatment with OA and LMWH was performed. Among others, the resources of the Cochrane Library, the MEDLINE, the Embase and the Biomed Central were searched. The metaanalysis of the reported treatment outcomes was performed using the RevMan5® software. The cost analysis and the cost-effectiveness analysis were performed. The data on the costs of treatment of the VTE in Poland were collected through the retrospective review of patient records obtained from the three hospitals and the anticoagulation clinic in the Krakow area, the pharmaceutical reimbursement databases and the public payer’s charge tariffs for the medical services. The modeling (decision tree) was performed using the TreeAgePro2009® software. RESULTS: The most important differences between the OA and the LMWH were related to the better prevention of the VTE incidence and the better prevention of the small bleedings. With respect to none of the assessed outcomes the OA were better than the LMWH. Within a six month treatment period the Incremental Cost-Effectiveness Ratio of avoiding the complication incidents was 49,865 zlotys (c12,242) from the payer perspective and 3,609 zlotys (c887) from the patient perspective. CONCLUSIONS: The LMWH offer the better effectiveness and safety than the OA but their cost-effectiveness is still limited by the relatively high prices of the LMWH. PCV82 COMMUNICATING COST-EFFECTIVENSS RATIOS TO DECISION MAKERS—THE CASE OF SWEDISH NATIONAL GUIDELINES FOR HEART DISEASES Eckard N1, Janzon M2, Levin LÅ1 1 Linköping University, Linköping, Sweden; 2Linköping University, Linköping University Hospital, Linköping, Sweden OBJECTIVES: Despite the continuing interest in health economic research, we could find no comprehensive and accessible data set on costs and effects, useful as practical information for decision makers who must allocate scarce resources within the cardiovascular field. The objective of this study was to present the cost-effectiveness league table, used in the Swedish national guidelines for heart disease and highlight issues of importance when communicating and interpreting cost-effectiveness evidence to decision makers. METHODS: A unique systematic literature search for the treatment of heart diseases was conducted. We then compiled all available cost-effectiveness ratios for different heart conditions and treatment strategies, in a so called league table. All cost-effectiveness results were expressed as a cost per quality adjusted lifeyears (QALY) or life-year gained. The league table was broken down to illustrate how health economic results may be communicated and made accessible to decision makers. We have highlighted methodological issues when interpreting cost-effectiveness league tables by using implantable cardioverter defibrillators (ICDs) as an example. RESULTS: More than 200 cost-effectiveness ratios were found and compiled in the league table ranging from dominant to c950,000 per QALY. Using ICD as an example we identified various problems when interpreting league table results. The results are context specific, time dependant, comparator dependent, often based on point-estimates giving a false sense of precision. CONCLUSIONS: League tables provide a means of presenting cost-effectiveness evidence aiding decision makers with valid information within a limited space. We have given examples and presented ways of communicating cost-effectiveness results for e.g. target groups, focusing on how information included in a cost-effectiveness league table may be interpreted and conveyed and used as a tool in the decision-making process. PCV83 PILOTING THE DEVELOPMENT OF A COST-EFFECTIVE EVIDENCEINFORMED CLINICAL PATHWAY: MANAGING HYPERTENSION IN JORDANIAN PRIMARY CARE Chalkidou K1, Lord J1, Obeidat N2, Alabbadi I3, Stanley A1, Bader R2, Momani A2, O’Mahony R2, Qatami L2, Cutler D1 1 National Institute for Health and Clinical Excellence, London, UK; 2Medicines Transparency Alliance, Amman, Jordan; 3University of Jordan, Amman, Jordan OBJECTIVES: The UK’s National Institute for Health and Clinical Excellence (NICE) and the Jordan office of the Medicines Transparency Alliance embarked on a pilot project to design an evidence-based guideline for cost-effective pharmacological treatment of essential hypertension in Jordan. The project’s objectives were to directly address a major health problem for Jordan by producing a guideline; and to delineate the strengths and weaknesses of Jordan’s health care process to allow similar future efforts to be planned more efficiently. METHODS: The pilot spanned a period of approximately 8 months. Activities were overseen by local technical and guideline development teams, as well as experts from NICE. NICE’s hypertension guidelines and economic model were used as a starting point. Parameters in the economic model were adjusted according to input and feedback from local experts with regards to Jordanian physician and patient practices, resource costs, and quality of life estimates. The results of the economic model were integrated with the updated available clinical trial literature. RESULTS: The outputs of the economic model were used to inform recommendations, in the form of a clinical algorithm. a report of the process and the strengths and weaknesses observed was developed, and recommendations for improvements made. CONCLUSIONS: The pilot represented the start of what is intended to be a health care process change for the country of Jordan. Issues emerged which can inform strategies to ensure a more cohesive and comprehensive health care approach 13th Euro Abstracts to the cost-effective use of appropriate drugs in managing chronic disease. The project highlighted how countries with relevant experience in evidence-informed policy making in health care can assist others in strengthening their decision-making systems. PCV84 ECONOMIC EVALUATION OF IMPLANTABLE LOOP RECORDER IN THE DIAGNOSIS OF SYNCOPE, UPDATE FROM THE SPANISH NATIONAL HEALTH SYSTEM García Baena I1, García García2 FJ2, Navascues R3, Marti B3 1 Universitat Pompeu Fabra, Barcelona, Spain, Barcelona, Spain; 2Subdirector Médico Área Médica,, Las Palmas de Gran Canaria, , Spain; 3Medtronic Iberia, Madrid, Spain OBJECTIVES: The aim of this study was to evaluate costs and benefits of diagnosing patients with syncope of unknown aetiology with implanted loop recorders compared to clinical practice (conventional investigation) of an electrophysiology, from the Spanish National Health System perspective. METHODS: Based on an economic decision analysis model, the cost-effectiveness study analyses diagnostic yield results from clinical practice versus Reveal DX®, with a time horizon of one year. Clinical data and resource use was obtained from a randomised controlled trial and expert opinion. Cost data was expressed in Euros 2010. a univariate sensitivity analysis was carried out to analyze the robustness of the model by modifying the number of outpatient visits, the costs and the diagnostic yield for both strategies. RESULTS: The incremental cost-effectiveness Ratio of Reveal DX® versus conventional investigation was c3167 per additional diagnosis, resulting in a confidence interval of 95% of the incremental cost-effectiveness ratio of c2335–c4867 per additional diagnosis made. The results of the univariate sensitivity analyses did not change the main results from our study. CONCLUSIONS: Reveal DX® in the diagnosis of syncope of unknown aetiology after an initial evaluation is a cost-effective strategy compared to conventional diagnosis for the Spanish Health System, since it increases diagnostic yield at a reasonable incremental cost. PCV85 PRASUGREL VS. CLOPIDOGREL IN PATIENTS WITH ACUTE CORONARY SYNDROME UNDERGOING PERCUTANEOUS CORONARY INTERVENTION: A SPANISH MODEL-BASED COST-EFFECTIVENESS ANALYSIS Davies A1, Sculpher MJ2, Barrett A3, Valladares A4, Huete T4, Dilla T4 1 Oxford Outcomes (UK), Botley, Oxford, UK; 2The University of York, York, UK; 3Eli Lilly & Company, UK, Windlesham, Surrey, UK; 4Eli Lilly & Company, Spain, Alcobendas, Madrid, Spain OBJECTIVES: In patients with acute coronary syndrome (ACS) undergoing percutaneous coronary intervention (PCI), the TRITON-TIMI 38 trial (TT38) demonstrated that treatment with prasugrel vs. clopidogrel was associated with significantly reduced rates of atherothrombotic events, though with increased risk of bleeding. The aim of the analysis was to evaluate the long-term cost-effectiveness of a 12-month treatment with prasugrel vs. clopidogrel in the trial population, excluding patients with prior transient ischemic attack or stroke, from the perspective of the Spanish health care system. METHODS: A Markov model was developed based on risk equations for cardiovascular death, myocardial infarction (MI) or stroke, bleeding, and rehospitalization, derived from TT38 data (N = 13,608 patients). Cost variables included were medication and hospitalizations costs. Hospital readmissions captured during TT38 in all patients from 8 countries (N = 6705) were assigned to Spanish diagnosis related groups, and were modelled to accrue over the life-time horizon. Long term survival and quality adjusted survival were estimated for the life-time of each patient. RESULTS: At 12 months, a difference in drug costs between prasugrel and clopidogrel of +c77 (branded clopidogrel) to +c460 (generic clopidogrel) per patient was partially offset by hospital cost savings (−c97 per patient) due principally to reduced rates of revascularization. In the longer-term, prasugrel was associated with higher total costs (+c11 to +c395 per patient), life expectancy gains of 0.07 years, due primarily to the reduced rate of MI, and 0.05 additional QALYs, resulting in incremental costs per life-year saved and per QALY gained of c164 to c5718 and c216 to c7540, respectively. Probabilistic sensitivity analysis indicated that prasugrel has a 68% to 72% probability of being more cost-effective than clopidogrel at a willingness to pay of c30,000 per QALY. CONCLUSIONS: Among ACS-PCI patients, these results showed prasugrel to be within the bounds of reasonable cost-effectiveness for Spain in comparison with clopidogrel. PCV86 ECONOMIC EVALUATION OF DRONEDARONE IN TREATMENT OF ATRIAL FIBRILLATION IN GREECE Maniadakis N, Fragoulakis V, Athanasakis K, Kyriopoulos J National School of Public Health, Athens, Greece OBJECTIVES: Atrial Fibrillation (AF) is a common cardiac arrhythmia and a significant cause of morbidity and mortality worldwide. As it is necessary to maximize value from the money spent in health care, an economic evaluation was undertaken to compare a new therapy, Dronedarone, in relation to Amiodarone, Sotalol and Propafenone already used in the Greek National Health Service (NHS) setting. METHODS: An international Markov model was locally adapted. The model reflects the management and the progression of AF patients through different health states in the course of their life time, including stroke, post stroke, heart failure (HF), post-HF, acute coronary symptom (ACS) post-ACS and death. Clinical and quality of life data to populate the model were derived from a variety of relevant clinical studies and A357 registries including: ATHENA, AFTER, DIONYSOS, AFFIRM and synthesizing analyses undertaken by academic experts. Resource utilization and cost data were derived by means of a large and representative panel of local experts, who utilized patient data files and data which came from a sample of NHS hospitals. Data refer to the year 2010 and all outcomes were discounted at a rate of 3.5%. The model is probabilistic to account for uncertainly and mean estimates are reported with corresponding uncertainty intervals. RESULTS: Mean total treatment costs were: Dronedarone: c12,931 (95%UI: 12,065–c12,495); Amiodarone: c8893 (95%UI: c8685–c9100); Sotalol: c6,185 (95%UI: c5901–c6509); Propafenone: c8433 (95%UI: c8229–c8642). The incremental cost-per-life-year-gained with Dronedarone versus Amiodarone was: c2236 (95%UI: 1897–c2615), versus Sotalol: c2576 (95%UI: c2442–c2822 and versus Propafenone: c2718 (95%UI: c2497–c3395). The incremental cost-per-qualityadjusted-life-year gained with Dronedarone versus Amiodarone was: c3275 (95%UI: 2730–c3838), versus Sotalol: c4319 (95%UI: c4130–c4510) and versus Propafenone: c3138 (95%UI: c2571–c4004). CONCLUSIONS: The newly available treatment Dronedarone appears to be a cost-effective alternative to other already existing therapies, used in the management of Atrial Fibrillation patients in the Greek NHS. PCV87 USING VALUE OF INFORMATION ANALYSIS IN COMBINATION WITH AN EARLY STAGE MODEL OF COREVALVE FOR SEVERE AORTIC STENOSIS TO INFORM FUTURE RESEARCH NEEDS Mealing S, Watt M, Sculpher M, Eaton JN Oxford Outcomes Ltd, Oxford, Oxon, UK OBJECTIVES: Aortic Stenosis (AS) is a severe cardiovascular condition the treatment of which often involves a major operation. For a subgroup of patients medical management (MM) is the only treatment option due to procedural risk. a transcatheter aortic valve implantation device “CoreValve,” is a novel procedure, is less invasive and allows for the implantation of a replacement valve in this patient group. Since information is not yet available on key clinical parameters, we modified an existing early stage economic model to perform a value of information (VoI) analysis to informing the prioritization of future research. METHODS: The underlying model used in the analysis is a 10-year Markov model was developed in Microsoft Excel. Treatment options were CoreValve and MM with parameters derived from published literature. All costs were taken from the most recent published sources. Decrements were applied to age-specific EQ-5D population norms to generate QALYs. a probabilistic sensitivity was used to inform the global Expected Value of Perfect Information (EVPI) calculation. Deterministic one way analyses were used to select the variable groups and individual parameters on which partial EVPI (EVPPI) calculations were performed. Annual incident population estimates were derived from information in a large national database. RESULTS: Assuming a decision horizon of 10 years, an annual incident population of 4052 and a willingness to pay threshold of £30,000 per QALY gained the EVPI is £11.3 million. EVPPI estimates were generated for costs, utilities, overall survival (MM patients) and treatment effects. Of these, the VoI for baseline mortality and long term mortality reduction were greatest (£21,364,000 and £8,151,000 respectively). The VoI for all others was negligible CONCLUSIONS: Further information on long term survival would have the greatest impact on decision uncertainty. Thus, a new clinical trial may not be required and a registry may be more appropriate. PCV89 COST-EFFECTIVENESS OF DRONEDARONE FOR THE TREATMENT OF ATRIAL FIBRILLATION IN THE UK Brereton NJ1, Craig AM2, Akehurst R3 1 BresMed Health Solutions, Sheffield, South Yorkshire, UK; 2Sanofi-Aventis, Guildford, Surrey, UK; 3ScHARR, The University of Sheffield, Sheffield, UK OBJECTIVES: To evaluate the cost-effectiveness of dronedarone for the treatment of atrial fibrillation (AF) compared to current antiarrhythmic drugs (AADs), from a UK NHS perspective. METHODS: A cost-utility analysis was performed, for which an individual patient lifetime discrete event simulation model was constructed. The model predicted a patient’s course for a treatment pathway based on the current National Institute for Health and Clinical Excellence (NICE) AF guidelines and compared treatment with amiodarone, sotalol and Class 1c agents to dronedarone. The model consisted of seven AF-related events; AF recurrence, acute coronary syndromes, stroke, congestive heart failure, treatment discontinuation, AF status change and mortality. Between events patients resided in four health states; normal sinus rhythm, permanent AF with uncontrolled symptoms, permanent AF with controlled symptoms and death. Patient’s baseline event risks were estimated from the non-active comparator arm of the ATHENA trial then adjusted for treatment effects based on a mixed treatment comparison. Cost data were elicited from existing literature and UK reference costs. Quality of life estimates were based on data from the AFTER cohort. Cost-effectiveness was measured in cost per quality adjusted life-year (QALY) gained. Costs and QALYs were discounted at 3.5%. One-way and probabilistic sensitivity analyses (PSA) were performed. RESULTS: Dronedarone was shown to be cost-effective with incremental cost-effectiveness ratios of £2,406 versus amiodarone, £1,911 versus sotalol and £18,737 versus Class 1c agents. One-way sensitivity analysis showed that treatment effect on mortality was the key driver of cost-effectiveness. PSA results estimated that dronedarone was costeffective at an acceptability threshold of £20,000 on 95% of occasions compared to amiodarone and sotalol and on 60% of occasions compared to Class 1c agents. CONCLUSIONS: The results of this analysis demonstrate that in the UK setting dronedarone is a cost-effective treatment of AF compared to current AAD treatment. A358 PCV90 THE COST-EFFECTIVENESS OF ROSUVASTATIN VERSUS SIMVASTATIN FOR THE PREVENTION OF CARDIOVASCULAR MORBIDITY AND MORTALITY IN PATIENTS WITH HIGH BASELINE RISK—A SWEDISH ECONOMIC EVALUATION BASED UPON THE JUPITER TRIAL Olsson AG1, Jensen MM2, Gandhi SK3, Smolen L4, Paulsson T5 1 Linköping University Sweden, Linköping, Sweden; 2AstraZeneca, Lund, Sweden; 3 AstraZeneca LP, Wilmington, DE, USA; 4Medical Decision Modeling Inc., Indianapolis, IN, USA; 5AstraZeneca Nordic MC, Södertälje, Sweden OBJECTIVES: To assess the long-term cost-effectiveness of various doses of rosuvastatin (R) versus relevant doses of simvastatin (S) (R20 mg versus S40 mg (primary analysis), R10 versus S20 and R10 versus S40) in a patient population with high risk of cardiovascular events (10-year Framingham CVD risk ≥ 20%). METHODS: A Monte Carlo simulation model was developed based on the JUPITER (Justification for the Use of statins in Primary prevention: an Intervention Trial evaluating Rosuvastatin, NCT00239681) trial findings and includes modeling of cardiovascular events and death over the lifetime of patients. The relative efficacy of S20, S40 and R10 compared with R20 (as observed in JUPITER) was estimated by computing relative 10-year Framingham cardiovascular event risks based on reported differences in Total Cholesterol/High-Density Lipoprotein cholesterol ratio. Epidemiological data specific for the Swedish setting were utilized to model post-event mortality and long-term overall mortality in event-free patients. Incremental effectiveness was primarily measured as quality-adjusted life-years (QALYs) gained. Cost-effectiveness was assessed based upon direct costs. All effects and costs (in 2008/09 Swedish unit prices) were discounted at annual 3%. RESULTS: The model estimated that treating a cohort of 100,000 patients (66 years, 60% males) with R20 mg avoided 2642 CVD events over lifetime compared with S40 mg. This translated into an estimated gain of 9515 years in full health (QALYs). The incremental cost per QALY gained was SEK161,712 (16,434). The cost per QALY gained was SEK228,655 (c23,237) for R10 versus S20, and SEK234,932 (c23,875) for R10 versus S40. Probabilistic sensitivity analyses supported base-case results. CONCLUSIONS: Treatment with rosuvastatin 10 and 20 mg is cost-effective compared with relevant doses of simvastatin in the primary prevention of CVD for patients with high baseline cardiovascular risk (10-year Framingham CVD risk ≥ 20%). PCV91 THE COST-EFFECTIVENESS OF ROSUVASTATIN VERSUS ATORVASTATIN FOR THE PREVENTION OF CARDIOVASCULAR MORBIDITY AND MORTALITY IN PATIENTS WITH HIGH BASELINE RISK—A SWEDISH ECONOMIC EVALUATION OF THE JUPITER TRIAL Olsson AG1, Jensen MM2, Gandhi SK3, Fox KM4, Paulsson T5 1 Linköping University Sweden, Linköping, Sweden; 2AstraZeneca, Lund, Sweden; 3 AstraZeneca LP, Wilmington, DE, USA; 4University of Maryland School of Medicine, Monkton, MD, USA; 5AstraZeneca Nordic MC, Södertälje, Sweden OBJECTIVES: To assess long-term cost-effectiveness of various doses of rosuvastatin (R) versus relevant doses of generic atorvastatin (A) (R20 mg versus A40 mg (primary comparison), R10 versus A20 and; R40 versus A80) in patients with a high risk of CV events (10-year Framingham CVD risk ≥ 20%). METHODS: A Monte Carlo simulation model was developed based on JUPITER (Justification for the Use of statins in Primary prevention: an Intervention Trial evaluating Rosuvastatin, NCT00239681) trial findings and modeled cardiovascular events and death over the lifetime of patients. The relative efficacy of the A20, A40, A80, R10, and R40 mg compared to R20 mg (as observed in JUPITER) were estimated by computing relative 10-year Framingham cardiovascular event risks based on reported differences in Total Cholesterol/High-Density Lipoprotein cholesterol ratio. Epidemiological data specific for the Swedish setting were utilized to model mortality. Incremental effectiveness was primarily measured as quality-adjusted life-years (QALYs) gained. Cost-effectiveness was assessed based upon direct costs. All effects and costs (2008/09 Swedish unit prices) were discounted at annual 3%. An 80% price reduction was assumed for generic versus branded statin. RESULTS: The model estimated that treating a cohort of 100,000 patients (66 years, 60% males) with R20 avoided 1121 CVD events over lifetime compared with atorvastatin 40 mg. This translated into an estimated gain of 4090 years in full health (QALYs). The estimated incremental cost per QALY gained was SEK366,763 (c37,273). This estimate was SEK428,060 (c43,502) for R10 versus A20, and SEK582,241 (c59,171) for R40 versus A80. Probabilistic sensitivity analyses supported base-case results. CONCLUSIONS: Treatment with rosuvastatin 10 mg, 20 mg and 40 mg is cost-effective compared with relevant doses of generic atorvastatin (20 mg, 40 mg, 80 mg, respectively) for the primary prevention of cardiovascular events for patients with high baseline cardiovascular risk (10-year Framingham CVD risk ≥ 20%) in Sweden. PCV92 THE HEALTH AND ECONOMIC IMPACT OF SWITCHING FROM ATORVASTATIN TO GENERIC SIMVASTATIN IN BELGIUM Liew D1, Webb K2, Marbaix S3 1 The University of Melbourne, Fitzroy, Victoria, Australia; 2Pfizer Limited, Surrey, UK; 3Pfizer, Bruxelles, Belgium OBJECTIVES: Containing pharmaceutical budgets has encouraged generic prescribing policies. Across Europe, many patients have been switched from atorvastatin to generic statins, particularly simvastatin, but often at lower therapeutic doses. This study sought to estimate the potential clinical and economic effect if policy-induced 13th Euro Abstracts switching to first generation statins occurred in Belgium as per previously observed patterns. METHODS: A Markov micro-simulation model was populated with 80 primary prevention Belgian patients from a 2007 observational study. Risks of firstonset cardiovascular disease (CVD) were estimated using a calibrated Framingham risk equation. With a baseline of January 2010, follow-up was simulated for 20 years. Decision analysis estimated the marginal effects of switching all patients from atorvastatin (weighted average daily dose [WADD] 20.7 mg) to simvastatin (WADD 31.6 mg). Dose-specific, lipid-modifying effects of the two statins, CVD costs and utilities were sourced from published data. Annual discount rates of 3% and 1.5% were applied to costs and health effects, respectively. RESULTS: Of the 80 subjects on atorvastatin, 23 (28.9%) were predicted to develop CVD over 20 years. In the switched-to-simvastatin group, the predicted number was 26 (32.0%), equating to a “number needed to harm” of 32. Switching was estimated to lead to a net cost saving of c581 per subject, but also a loss of 0.04 QALYs. These equated to an ICER of c13,608 per QALY gained (atorvastatin vs. simvastatin). Sensitivity analyses indicated the results to be robust. CONCLUSIONS: Our preliminary analyses indicate that there would be an increase in the burden of CVD if Belgian patients are switched from atorvastatin to generic simvastatin at non-equipotent doses, as has happened in neighbouring countries like The Netherlands. This study highlights the need to consider the potential health and health economic impact of population-based switching policies. PCV93 THROMBOPROPHYLAXIS AFTER TOTAL KNEE REPLACEMENT: COST-UTILITY ANALYSIS OF RIVAROXABAN VERSUS ENOXAPARIN IN SLOVAKIA Lukac M1, Bielik J2, Lees M3, Tomek D4, Foltan V5 1 Slovak Medical University, Bratislava, Slovak Republic; 2Trencin University, Trencin, Slovak Republic; 3Bayer, Uxbridge, UK; 4Slovak Society for Pharmacoeconomics, Bratislava, Slovak Republic; 5Faculty of Pharmacy, Comenius University, Bratislava, Slovak Republic OBJECTIVES: To estimate the cost-effectiveness of rivaroxaban against enoxaparin for the prophylaxis of venous thromboembolism (VTE) in patients after total knee replacement (TKR) in Slovakia from payer perspective. METHODS: Previously published cost-utility model based on results of large randomized controlled trial (RECORD 3) has been adapted to Slovakian settings. In RECORD 3, patients received 12 days prophylaxis with rivaroxaban or enoxaparin. Rivaroxaban reduced total VTE (composite: any DVT, non-fatal PE, all-cause mortality) by 49% versus enoxaparin after 12 days prophylaxis. The model was divided into three parts: prophylaxis, postprophylaxis, and long-term complications. The first two parts represents acute phase and were modeled as a decision tree. Third part represents the long-term complications and was developed as a Markov model. The first part of the model is populated by RECORD 3 trial, while published epidemiological and clinical data estimating the risk of further VTE events and post-thrombotic syndrome beyond the trial period were used in second and third part of the model. Local cost data was based on published price lists, clinical guidelines, product labels and expert opinion. VTE related utilities were used from literature. Effectiveness was measured in quality-adjusted life-years (QALY). Time horizon was set at 5 years and payers perspective was used. Discount rate was 5% per year for costs and effects according to valid Ministry of Health (MoH) guidelines for health economic evaluation. One-way and probabilistic sensitivity analyses were performed. RESULTS: Rivaroxaban produced improved outcomes (QALY) and cost savings of c28 per patient versus enoxaparin in Slovakian setting (dominance). Probabilistic sensitivity analysis showed dominance of rivaroxaban compared to enoxaparin in more then 99% of cases. CONCLUSIONS: Prophylaxis of VTE with rivaroxaban following TKR may improve health outcomes and reduce direct medical costs when compared to enoxaparin in Slovakian setting. PCV94 POTENTIAL COST-EFFECTIVENESS OF A BIOMARKER TEST TO STRATIFY PATIENTS INDICATED FOR A CORONARY STENT Burgers LT, Redekop WK, Severens JL Erasmus University Rotterdam, Rotterdam, The Netherlands OBJECTIVES: Patients requiring a coronary stent can receive a bare-metal stent (BMS) or a drug-eluting stent (DES), but both have their advantages and disadvantages. We estimated the potential one-year cost-effectiveness of a test to decide which stent a patient should receive based on the risk of restenosis after a BMS implantation. METHODS: This study was performed as part of a Dutch large-scale five-year study (“Circulating Cells”) now underway to identify blood biomarkers to facilitate the prevention and treatment of coronary heart disease. a Markov chain Monte Carlo model was developed to estimate costs and effectiveness for three strategies: DES for all patients, BMS for all patients, and use of a test (80% sensitivity & 80% specificity). Input values were based on the literature and expert opinion. Costs were calculated according to the health care sector perspective. Scenario and sensitivity analyses were performed to test the robustness of the results. RESULTS: The DES-for-all strategy was the most effective (0.840 QALYs), followed by the test strategy (0.839) and the BMS-for-all strategy (0.838). However, it was also more costly (c8189) than the other two (test strategy, c7475; BMS, c6905). These results meant high incremental costeffectiveness ratios for the test and DES-for-all strategies. Implanting DES in all patients would have an important budget impact (c46,000,000) for the The Netherlands, given 36,000 interventions annually. Various input parameters had an important influence on the results, including the sensitivity, specificity and costs of the test. CONCLUSIONS: Both DES and BMS stents are often used for PCIs in the The Netherlands. a stratifying test has the potential to reduce costs and still achieve accept- 13th Euro Abstracts able effects. The impact of adopting a longer time horizon to capture very late stent thrombosis needs to be examined. The model developed in this study will be used to facilitate further development of the biomarker test. PCV95 ECONOMIC EVALUATION OF LIFESTYLE INTERVENTION IN PRIMARY PREVENTION OF CARDIOVASCULAR DISEASE IN 60-YEAR-OLD MEN IN STOCKHOLM, SWEDEN Saha S1, Lindgren P2, Johansson P2, Gerdtham UG1 1 Lund University, Lund, Sweden; 2Karolinska Institute, Stockholm, Sweden OBJECTIVES: To assess the cost utility analysis of lifestyle interventions (diet, physical activity and diet + physical activity) at primary care for person with high risk of cardiovascular disease by using a Markov model. METHODS: A cost-utility analysis was performed using a 10-state Markov model of CVD (cardiovascular disease) events applied to a cohort of 60 year-old men from the county of Stockholm, Sweden. a previously performed randomized controlled trial of dietary advice, physical exercise and diet + physical exercise on middle aged men of Sweden (The Sollentuna study) was used as effectiveness of intervention. The risk of having a CVD event in primary care was taken from The Framingham Heart Study. Lifetime costs and health effects (QALYs, quality-adjusted life-years) were estimated, in both a health care and societal perspective with 3% discount rate. RESULTS: The model predicts physical exercise to be the best option at zero willingness to pay and dietary advice the best if the decision-maker willingness to pay is higher than 60,000 SEK per QALY (1 Euro = 10.74 SEK (2009). The incremental cost-effectiveness ratio (ICER), compared with no intervention is 77,000 SEK/ QALY from a health care perspective and 57,000 SEK/ QALY from a societal perspective. CONCLUSIONS: Based on the model estimates, dietary advice and physical exercise both are cost-effective interventions, and the choice depends upon the decision-maker willingness to pay. PCV96 PRIORITIZING FURTHER RESEARCH USING THE EXPECTED VALUE OF PERFECT INFORMATION: APPLICATION TO VENOUS THROMBOEMBOLISM PROPHYLAXIS POST TOTAL HIP REPLACEMENT McCullagh LM1, Walsh C2, Barry M1 1 National Centre for Pharmacoeconomics, Dublin, Ireland; 2Trinity College Dublin, Dublin, Ireland OBJECTIVES: To estimate the expected value of perfect information (EVPI) associated with a lifetime Markov-model which evaluated the cost-effectiveness of rivaroxaban and dabigatran compared to enoxaparin for venous thromboembolism (VTE) prophylaxis after total hip replacement (THR). To estimate the expected value of perfect parameter information (EVPPI). To indentify the model parameter subset whose uncertainty drives the decision uncertainty. To decrease the uncertainty, by collecting additional data on this subset. To investigate the impact of threshold change on the EVPI/EVPPI values. METHODS: EVPI was determined directly from PSA and was calculated over the threshold c0–100,000/QALY. Population EVPI (PEVPI) was calculated based on the 5464 THR procedures performed in Ireland in 2006 and the ENDORSE-study which revealed that 64% of at risk surgical patients in Ireland receive recommended VTE prophylaxis. The assumptions were; a 50% policy of choice uptake rate, a 4% discount rate and a 10-year time horizon. Alternative assumptions of 5 and 15 years were explored. EVPPI were calculated for the 3 parameter subsets using the 2-stage Monte-Carlo algorithm. RESULTS: Base-case analysis indicated that rivaroxaban dominated. Probabilistic analysis (PSA) indicated that at c45,000/QALY, the probability that rivaroxaban was the most cost-effective strategy was 67%. The PEVPI was c15.62 million. Costs had the highest EVPPI (c12.43 million), followed by probabilities (c1.45 million) and preference weights (c1.44 million). The costs subset was most sensitive to threshold. Costs were updated with a more detailed costing study. All analyses were repeated. Base-case analysis indicated that rivaroxaban continued to dominate. The probability that rivaroxaban was the most cost-effective strategy increased to 81%. PEVPI decreased to c4.59 million. EVPPI for costs decreased to c2.13 million. EVPPI for probabilities and preference weights were c1.03 million and c0.94 million respectively. CONCLUSIONS: This study demonstrates the application of EVPI/EVPPI analyses to quantify and decrease uncertainty associated with the choice of VTE prophylaxis after THR in the Irish health care setting. CARDIOVASCULAR DISORDERS – Patient-Reported Outcomes Studies PCV97 AN EVALUATION OF ONCE- VERSUS TWICE-DAILY DOSING ON PERSISTENCE WITH PRESCRIPTION MEDICATIONS IN CARDIOVASCULAR PATIENTS Bae JP1, Zagar A1, Dobesh PP2, Klepser DG2, Anderson J1, Tomlin ME1, McCollam PL1 1 Eli Lilly & Company, Indianapolis, IN, USA; 2University of Nebraska Medical Center, Omaha, NE, USA OBJECTIVES: To compare patient persistence with chronic-use prescription medications between once- (QD) and twice-daily (BID) dosing. METHODS: Using a large claims database (MarketScan), prescription medications were defined as QD if ≥80% of claims showed a quantity/day = 1 and as BID if ≥80% of claims had a quantity/ day = 2. Data for patients (18 years or older) with a first claim in 2007 were selected for analysis and were further limited to 4 classes of medications frequently used by A359 cardiovascular patients: antiplatelet, general cardiac, antihyperlipidemic, or antidiabetic medications. Persistence was measured by time-to-discontinuation (TTD), defined as the number of days from the first prescription filled to the first gap of >30 days between exhausting the supplied medication and filling the next prescription. The mean TTD with censoring at 365 days was estimated. a proportional hazards model accounting for within-patient correlations in patients using multiple medications was used to model the TTD. The model was stratified on medication class and adjusted for baseline confounding variables (gender, age, Charlson Comorbidity Index). RESULTS: A total of 1,077,936 patients were included, and the number of patientmedication combinations was 1,440,917 (QD: 1,384,565 and BID: 56,352). The estimated mean TTD for BID compared to QD was −8% overall (P < 0.01) and −32% for antiplatelet agents (P < 0.01), −22% for antihyperlipidemic agents (P < 0.01), −10% for antidiabetic agents (P < 0.01), and 1% for cardiac agents (P < 0.05). CONCLUSIONS: In this unique, recent, claims analysis, investigating multiple therapeutic classes of medications in a large database, persistence with antidiabetic, antihyperlipidemic, and antiplatelet agents was greater with QD compared with BID dosing; however, a wide range in the differences in persistence between QD and BID dosing was noted between therapeutic classes. Persistence may prevent subsequent adverse events. PCV98 IMPACT OF MEDICATION ADHERENCE ON HEALTH CARE EXPENDITURES BY MEDICARE BENEFICIARIES WITH HEART FAILURE Lopert R1, Shoemaker JS2, Stuart B3, Davidoff A3, Shaffer T3, Lloyd J2 1 Therapeutic Goods Administration, Canberra, Australia; 2University of Maryland Baltimore, Baltimore, MD, USA; 3University of Maryland School of Pharmacy, Baltimore, MD, USA OBJECTIVES: Various studies have sought to determine the impact of medication adherence on medical expenditure among the elderly with chronic conditions. To date, none have looked explicitly at adherence to treatment in cardiac failure, a common condition in the elderly. This study had two aims: 1) to measure adherence to treatment for recommended heart failure medications, and 2) to assess whether better adherence is associated with lower spending among Medicare beneficiaries. METHODS: We used data from the 1997–2005 Medicare Current Beneficiary Survey to track utilization rates for cardiac failure medications over three years in a nationally-representative sample of Medicare beneficiaries. The relationship between MPR and Medicare costs was tested in a sequence of multivariate models with extensive behavioral variables to control for healthy adherer bias and reverse causality. RESULTS: Over 3 years, median MPR values were 0.77 for ACE-I/ARB users, 0.74 for beta blockers, 0.82 for diuretics and 0.74 for cardiac glycosides. Higher adherence rates were consistently associated with lower Medicare spending in every model tested. The estimated marginal effect of a 10 point increase in MPR for an ACE-I or ARB was a reduction in cumulative 3-year Medicare spending of $390 (P < 0.05) in 2006 dollars. Similar increases in MPRs for beta blocker, diuretics, and cardiac glycosides were associated with $861 (P < 0.01), $520 (P < 0.01) and $2318 (P < 0.01) in lower Medicare costs respectively over three years. CONCLUSIONS: Higher levels of adherence to recommended CHF regimens by Medicare beneficiaries were associated with lower cumulative Medicare spending over three years, with savings generally exceeding the cost of the drugs. PCV99 COMPLIANCE AND PERSISTENCE WITH ANTIHYPERTENSIVES IN THE NETHERLANDS; A COMPARATIVE ANALYSIS WITH SPECIFIC FOCUS ON RAAS INTERVENING AGENTS Vegter S, Visser ST, Boersma C, Postma MJ University of Groningen, Groningen, The Netherlands OBJECTIVES: Inaccurate use of chronic medication is associated with worse disease prognosis and high health care costs. Differences in drug prices might be outweighted by differences in costs due to compliance and persistence. The aim of this study was to investigate compliance and persistence with antihypertensive agents acting on the renin-angiotensin-aldosterone-system (RAAS). METHODS: A cohort study was conducted based on pharmacy dispensing records (IADB.nl). Subjects were included who initiated treatment with either ACE-inhibitors (ACEi) or angiotensin receptor blockers (ARBs) between 1999 and 2007. a standardized toolbox was used, Refill-Sequence analyses (RS) measured persistence; Proportion of Days Covered analyses (PDC) measured one-year compliance. Differences in persistence were adjusted for age, gender and time of treatment initiation. RESULTS: A total of 27,544 patients initiated treatment with ACEi; 8,649 patients initiated treatment with ARBs. Using a grace period of 60 days, persistence with ARBs was better compared to ACEi, adjusted HR: 0.90 (95%CI: 0.84–0.96). Using a PDC cutpoint of 80%, one-year compliance with ARBa was also better compared to ACEi, 83.9% vs. 86.3%, P < 0.01. Within ACEi’s, differences in persistence between drugs were small. Within ARBs, persistence with losartan or valsertan was higher compared to irbesartan or candesartan, adjusted HR: 0.82 (0.74–0.90). Compliance within ARBs however was similar for all drugs (P = 0.27). Results were consistent when other grace periods or PDC cutpoints were used. CONCLUSIONS: Chronic users of ARBs were more compliant and persistent compared to users of ACEi. Differences in persistence within ARBs might be due to differences in drug tolerability and safety, but also due to differences in drug indications. Further results will be presented further exploring these findings, as well as analyzing differences in persistence between branded versus generic ACEi. A360 PCV100 THE EFFECTS OF DOSING COMPLEXITY ON ADHERENCE WITH PRESCRIPTION MEDICATIONS COMMONLY USED FOR CARDIOVASCULAR PATIENTS Bae JP1, Anderson J1, Zagar A1, Dobesh PP2, Klepser DG2, Tomlin ME1, McCollam PL1 1 Eli Lilly & Company, Indianapolis, IN, USA; 2University of Nebraska Medical Center, Omaha, NE, USA OBJECTIVES: To compare patient adherence with chronic-use prescription medications between once- (QD) and twice-daily (BID) dosing. METHODS: Using a large claims database (MarketScan), prescription medications were defined as QD if ≥80% of claims showed a quantity/day = 1 and were defined as BID if ≥80% of claims had a quantity/day = 2. Data for patients (18 years or older) with the first claim in 2007 were selected for analysis and were further limited to 4 classes of medications frequently used by cardiovascular patients: antiplatelet, general cardiac, antihyperlipidemic, or antidiabetic medications. Adherence was measured by medication possession ratio (MPR) defined as the number of days of medication supplied (between the first prescription fill date and 365 days following)/365 days. a linear model (generalized estimating equation) accounting for within-patient correlations in patients using multiple medications was used to model the MPR. The model was stratified on medication class and adjusted for baseline confounding variables (gender, age, Charlson Comorbidity Index). RESULTS: A total of 1,077,936 patients were included, and the number of patient-medication combinations was 1,440,917 (QD: 1,384,565 and BID: 56,352). The mean MPR for BID compared to QD was −5% overall and −28% for antiplatelet agents, −19% for antihyperlipidemic agents, −7% for antidiabetic agents, and 3% for cardiac agents (P < 0.01 for all comparisons). CONCLUSIONS: This unique, recent, claims analysis, which investigated multiple therapeutic classes of medications in a large database, generally found lower adherence in BID dosing in medications frequently used by cardiovascular patients; however, a wide range in the MPR differences between QD and BID dosing was observed between the therapeutic classes. Further research is needed to investigate the relationship between dosing complexity and therapeutic class of medications and subsequent patient adherence. PCV101 THE IMPACT OF MEDICATION SWITCHES INDUCES BY MEDICATION DISCOUNT CONTRACTS ON CHRONIC PATIENTS’ ADHERENCE Groth A1, Mueller S2, Wilke T1 1 Institute for Pharmacoeconomics and Drug Logistics, University of Wismar, Wismar, Germany; 2University of Wismar, Wismar, Germany OBJECTIVES: Switching of medications due to non-medical grounds may be a critical factor influencing patients’ medication-based adherence (NA). The aim of this analysis is to examine the effects of medication switching on NA for hypertonic/diabetic (type II) patients. Only switches between medications containing the same active ingredient were examined. METHODS: Prescriptions filled at 5 German pharmacies over a period of 2 years were analyzed. Patients who had filled prescriptions for at least 3 medications to control hypertonia (focus on Moxonidin—ATC-Code C02AC05) and/ or diabetes type 2 (focus on Metformin—ATC-Code A10BA02) were included. The NA was calculated using the medication possession ratio (MPR); NA was defined as MPR < 80%. The average MPRs obtained while using the original and the discount products (before/after comparison) were compared using non-parametric tests for independent samples (Wilcoxon-Tests). RESULTS: Of the 344 hypertonia sufferers, 253 received the active ingredient Moxonidin (NA patients’ percentage: 47.4%). In 68 cases, a switch to the discount product took place (NA percentage before switch: 46.4%). The percentage of NA patients was reduced to 39.1% after the switch. Of the 726 diabetic patients, 518 patients received Metformin (NA: 49.4%). a total of 136 patients received a discount article after the discount contract was in place (NA before switch: 42.7%). After the switch, the NA percentage increased to 66.2%. CONCLUSIONS: In the case of hypertonia, medication switches had no significant effect on the NA. By contrast, switchers receiving Metformin exhibited a substantial increase in medication-based NA. In future, the focus of research in this area should be to discover whether specific contracts concerning discount products (as, for example, in the case of Metformin) are related to particularly negative adherence effects and if so which factors influence such effects. PCV102 ASSOCIATION BETWEEN KNOWLEDGE AND DRUG ADHERENCE IN PATIENTS WITH HYPERTENSION IN QUETTA, PAKISTAN Saleem F1, Hassali MA1, Shafie AA1, Bashir S2 1 Universiti Sains Malaysia, Penang, Malaysia; 2University of Sargodha, Sargodha, Punjab, Pakistan OBJECTIVES: Good adherence to long term management of chronic diseases such as hypertension is very vital in order to prevent long term mortality and morbidity. There are many factors associated with non-adherence to prescribed medications and one of it is lack ok knowledge and awareness among patients towards the disease they are suffering from. The aim of the present study is to evaluate the association between patient’s knowledge to hypertension disease state management and medication adherence. METHODS: A cross-sectional study was undertaken with 385 established hypertensive patients who visited out patient departments in two public hospitals of Quetta city, Pakistan. Besides the demographic and disease-related questions, two questionnaires (Hypertension Fact Questionnaire with 15 items for knowledge and Drug Attitude Inventory with 10 items) were used for data collection. Descriptive statistics and Spearman correlation were carried out by using SPSS version 16. RESULTS: Out 13th Euro Abstracts of 385 patients (68.8% male; mean age: 39.02 ± 6.596), 236 (61.3%) of the patients had moderate degree of knowledge about hypertension. a total of 249 (64.7%) were categorized as poor adherent. Noticeably no patient was considered as good adherent in the study. Correlation coefficient between total score of knowledge and total adherence was—0.170 (P < 0.001) indicating an inverse association between knowledge scores and adherence level. CONCLUSIONS: Patient knowledge about hypertension had weak relation with drug adherence in this cohort of study group. This is possibly due to the dominancy of other factors influencing drug adherence such as perception of patients towards disease and drug, income, education, age etc in this cohort of study group. This calls attractive opportunities for further research on drug adherence and also identifies the gaps in providing optimum health care to patients. PCV103 COST IMPLICATION OF UNACCEPTABLE REFILL-BASED ADHERENCE RATES OF ANTIHYPERTENSION TREATMENT IN THE SOUTH AFRICAN PRIVATE HEALTH CARE ENVIRONMENT Steyn R, Lubbe MS, Serfontein JHP, Smit R North-West University, Potchefstroom, South Africa OBJECTIVES: To investigate the cost implication of unacceptable refill-based adherence rates of antihypertensive treatment in the South African private health care environment. METHODS: A non-experimental, retrospective drug utilization study was conducted on medicine claims data of a pharmacy benefit management company for hypertension patients for a four year period (January 1, 2005–December 31, 2008). The refill-based adherence rate was calculated per trade name for all antihypertensive medicine items that were prescribed more than once to 171,815 patients. The following equation was used: Refill–adherence rate = (total number of days of antihypertensive medicine items supplied—number of days supplied at last refill)/(date last claimed—date first claimed). [RSA Rand (R)/$US = 6.38112 (2005); 6.78812 (2006); 7.06926 (2007) and 8.27505 (2008)]. RESULTS: An average refill-based adherence rate of 82.8% ± 125.9 was obtained for 376,954 antihypertensive medicine items. Approximately 54.5% of antihypertensive medicine items (n = 205,476) had acceptable refill-based adherence rates between 80% and 120%, 42.2% (n = 159153) had unacceptably low refill-based adherence rates (less than 80%) and 3.3% (n = 12325) had unacceptably high rates (more than 120%). Those with unacceptably low or unacceptably high refill-based adherence rates accounted for respectively 24.3% (R109,275,001) and 1.3% (R5 958,593.70) of the total cost (N = R448,883,473) of all antihypertensive medicine items (n = 376,954). The following pharmacological groups of antihypertensive drugs had on average unacceptable refill-based adherence rates: diuretics (n = 114 623, 74.6% ± 131.7); central acting sympathetic nervous system inhibitors (n = 5 094; 79.9% ± 112.6); sympathetic nervous blockers (n = 3 090; 76.3% ± 87.5). CONCLUSIONS: Since pharmacy records are generally computerised, the prospect of using a simple algorithm to assess compliance based on refill patterns is enticing, particularly for monitoring the progress of patients with chronic diseases. PCV104 IMPACT OF POLICY-DRIVEN STATIN SWITCH PROGRAMS ON ADHERENCE AND PERSISTENCE Khalil C1, Folia C2, Liovas A3 1 University of Waterloo, Kitchener, ON, Canada; 2Agro Health Associates Inc., Burlington, ON, Canada; 3AstraZeneca Canada Inc., Mississauga, ON, Canada OBJECTIVES: Cardiovascular disease imposes a significant burden on patients and society, and is associated with considerable costs to the health care system. Although statins have been shown to reduce the risk of cardiovascular events, suboptimal adherence and poor persistence with therapy remains a common problem. The need to lower health care costs has led to increased interest in policy-driven therapeutic substitution programs for statins, however their impact on patient adherence and persistence is often not considered. METHODS: A review of published literature describing the impact of policy-driven statin switch programs on adherence and persistence was conducted. PubMed search terms included: hydroxy-methyl-glutaryl-CoA Reductase Inhibitors (MESH), and statin, therapeutic, generic, substitute, switch, persistence, compliance, adherence (all fields); the search was limited to articles published in English between 1989 to May 2010. Reference lists from selected papers were also reviewed. RESULTS: 4 studies were identified. Following therapeutic substitution of one statin for another, 20–25% of patients discontinued therapy within 6–12 months (4 studies). Additionally, one study found that patients were 20% more likely to discontinue treatment within the first year than those who did not switch. Furthermore, switching statins reduced adherence (i.e., probability of taking medication on most days) by 12–33% (3 studies). One study showed that switching statins for economic/ formulary reasons led to decreased filling and utilization of statin prescriptions overall. CONCLUSIONS: Low rates of adherence to statin therapy have been documented, despite strong evidence of clinical benefit with sustained use. As a result, many patients initiating statins may achieve limited clinical benefits from therapy because of premature discontinuation. Policy-driven statin switch programs may further contribute to problems with persistence and adherence. While therapeutic substitution policies for statins may decrease drug plan costs, their impact on adherence and persistence may significantly affect patient outcomes and increase costs to the health care system. 13th Euro Abstracts PCV105 BLOOD PRESSURE PREDICTION MODEL FOR HYPERTENSIVE PATIENTS AT SARABURI HOSPITAL IN 2009 Auamnoy T, Siemmai A, Tantipidoke R Chulalongkorn University, Bangkok, Thailand OBJECTIVES: 1. To compare means of calories burnt by exercise (exercise), compliance, eating behavior, hypertension knowledge, age and blood pressure (BP) between gender and 2. To find correlations between exercise, compliance, eating behavior score, hypertension knowledge, age and blood pressure. 3. To employ Hierarchical Stepwise Multiple Regression Analysis (MRA) to predict blood pressure. METHODS: A retrospective research by face to face interview and clinical outcomes were used to investigate relationships between 6 variables and BP of random 200 hypertensive patients at Saraburi Hospital 2009. RESULTS: Total (N) 200 (100%) patients were mostly female 118 (59.00%) and 82 (41.00%) were male. The average age was 52.21 ± 12.01 years, average exercise per week 2787.24 ± 141.61 kcal, average compliance score 7.41 ± 1.93, average SBP and DBP were 150.24 ± 18.49 and 89.40 ± 9.15 Hgmm. Cronbach’s Alpha coefficient of Sorofman’s Compliance scale for constructs “right time” and “right amount” were 0.7978, and 0.7896 respectively, Auamnoy Eating Behavior Scale was 0.7915. ANOVA confirmed that BP, compliance, hypertension knowledge, age, exercise and eating behavior score between male and female were not significantly different (p > 0.05). Pearson’s correlation confirmed that age, exercise and compliance had significant negative correlation with DBP (r = −0.19, −0.43, −0.60 with p = 0.00, 0.00, 0.00 respectively). Age had significant positive correlation with SBP however exercise and medical regimens compliance had significant negative correlation with SBP (r = 0.16, −0.81, −0.98 with p = 0.02, 0.00, 0.00 respectively). MRA equation demonstrated the three most significant variables those predicted SBP and DBP were: compliance (Beta = −0.90, −0.70), eating behavior (Beta = 0.82, 0.72) and exercise (Beta = −0.11, −0.29) P = 0.00, 0.00, 0.00 R2 = 0.58, 0.55 respectively. CONCLUSIONS: The more patients complied to medical regimens, with good eating behavior and the more exercise the less patients’ blood pressure was. PCV106 THE MISSING PIECE BETWEEN TREATMENT EXPERIENCE AND INTENTION TO PERSIST: TESTING THE INTERNAL CONSISTENCY RELIABILITY AND PREDICTIVE VALIDITY OF ACCEPTABILITY Chretin S1, Viala-Danten M2, Van Ganse E3, Patrick D4, Arnould B2, Longin J1 1 Registrat–Mapi, Lyon, France; 2Mapi Values, Lyon, France; 3Hôpital Neurologique Pierre Wertheimer, BRON Cedex, France; 4University of Washington, Seattle, WA, USA OBJECTIVES: The ACCEPT© questionnaire is a self-administered generic acceptability instrument assessing how patients balance the advantages and disadvantages of long-term chronic therapy. It is made of 2 parts: 1/ several characteristics of patient experience with treatment giving an indication on which treatment attributes are leading to intentions and adherence; 2/ a scale measuring acceptability of the treatment. Our objective was to test the internal consistency reliability and predictive validity of the acceptability scale. METHODS: A survey was conducted in 60 community pharmacies. Patients treated with statin for more than 3 months were asked to complete a preliminary version of the acceptability scale (3 items), as well as questions regarding their compliance with their statin medication. Compliant patients were defined as those who reported never to forget to take their statin medication. Internal consistency of the acceptability scale was assessed by Cronbach’s alpha. The statistical independence between the 3 acceptability items and compliance was investigated by chi-2 tests. The ability of the 3 acceptability items to detect non compliant patients was evaluated by the area under the Receiver Operating Characteristic (ROC) curve. RESULTS: Out of 443 patients included in the analysis, 28% have been treated for 1 to 3 years, and 54% for more than 3 years. 60% had no cardiovascular antecedent, 14% had an angina, 13% a prior myocardial infarction, 6% a prior stroke. 66% declared they never forgot to take their statin medication. Cronbach’s alpha of the acceptability scale was 0.70. The 3 items were statistically related to compliance (p-chi2 < 0.0001). The area under the ROC curve was 0.69. CONCLUSIONS: The acceptability scale showed satisfying preliminary results of reliability and predictive validity. Further work is needed to validate the scale in other long-term treated populations and to evaluate its ability to predict persistence to treatment over time. PCV107 VALUATION OF HEALTH STATE UTILITIES RELATED TO CARDIOVASCULAR PREVENTION WITH ASPIRIN Manson SC1, van Hanswijck de Jonge P1, Palsgrove A2, Gorelick PB3 1 United BioSource Corporation, London, UK; 2United BioSource Corporation, Bethesda, MD, USA; 3University of Illinois College of Medicine, Chicago, IL, USA OBJECTIVES: The primary aim of the study was to determine societal utility scores for health states associated with events either benefiting from aspirin (myocardial infarction, stroke, angina) or potentially caused by aspirin (GI bleeding). The secondary aim was to determine the societal disutility score of taking aspirin every day. METHODS: Draft health states were developed based on a brief literature review and exploratory and validation interviews with cardiovascular (n = 3) and stroke (n = 2) specialists. The final health states were tested in a pilot study with members of the general public (n = 6). In the main study, members of the general public (n = 90 Canadians; n = 86 Americans) completed a chained standard gamble (SG) interview, a Visual Analogue Scale (VAS) rating task (0–100), the EQ-5D and a socio-demographic form. RESULTS: The samples were a reasonable match to the population of Canada and the US in terms of demographics (48% female, 67% white) and quality of life (rated by EQ-5D). Eight percent of participants had experienced one or more A361 forms of CVD or stroke as described in the health states and 18% of participants were taking daily preventive aspirin. In patients who were not taking aspirin, adding daily aspirin to prevent CVD caused a decrement in utility of 0.018. Conversely, patients who were taking daily aspirin reported an increase in utility of 0.021 compared to if they were not taking aspirin. GI bleeding and transient ischemic attack were rated as the states with the highest utility (0.79) followed by angina (0.71 stable, 0.66 unstable). Various stages of MI and stroke were considered least desirable (0.10–0.63). CONCLUSIONS: The study highlights the benefit in utility that could be gained from preventing cardiovascular and stroke events. It also demonstrates that the minor burden of taking aspirin daily changes to a small incremental benefit in wellbeing in participants already taking aspirin. PCV108 METHODS FOR ESTIMATING HEALTH-STATE UTILITIES IN PULMONARY ARTERIAL HYPERTENSION Mychaskiw MA1, Berger A2, Mardekian J1, Hwang LJ1, Oster G2 1 Pfizer Inc, New York, NY, USA; 2Policy Analysis Inc., Brookline, MA, USA OBJECTIVES: To compare health-state utility values obtained using three different instruments administered to patients with pulmonary arterial hypertension (PAH) enrolled in a randomized controlled trial of sildenafil. METHODS: Data for this study were obtained from a large phase III clinical trial in which patients were randomized to receive either sildenafil or placebo for 12 weeks. At each visit (baseline, weeks 1, 4, 8, and 12 of follow-up), patients were administered the Short Form-36 General Health Survey (SF-36) and the EuroQol Health Survey (EQ-5D); additionally they also were asked to provide a direct assessment of their current health state using a visual analog scale (VAS). Responses to the SF-36 and EQ-5D were then converted to healthstate utility values using published algorithms. Patients were pooled across treatment groups, and attention was focused on baseline values. Statistical significance of differences between these three estimates was ascertained using paired t tests. RESULTS: There were a total of 274 patients across both treatment groups in the intent-to-treat population. Proportions of 0–30, 31–60, and 61–100 health-state utility values were 0.4%, 22.6%, and 77.0% for SF-36; 16.2%, 9.6%, and 74.3% for EQ-5D; and 8.6%, 47.2%, and 44.2% for VAS, respectively. Mean (95% confidence intervals) healthstate utility values were 71.4 (69.6–73.1) for SF-36 (P < 0.01 vs. EQ-5D or VAS), 63.2 (60.2–66.3) for EQ-5D (P < 0.01 vs. VAS), and 59.0 (56.9–61.2) for VAS, respectively. CONCLUSIONS: Ratings of current health by PAH patients are significantly worse with a single-item VAS scale than based on responses to the EQ-5D and SF-36 health questionnaires. The EQ-5D appears to yield somewhat lower values than the SF-36. Further research is needed to better understand the reason(s) for these differences. PCV109 SEVENTY-SEVEN HEALTH STATE UTILITIES ESTIMATED IN POLISH CARDIAC PATIENTS Zawodnik S, Golicki D, Hermanowski T Department of Pharmacoeconomics, Medical University of Warsaw, Warsaw, Poland OBJECTIVES: The aim of this study was to create a catalogue of SF-6D health state utility values in the cardiovascular diseases, based on studies conducted in Polish population using the SF-36 questionnaire. METHODS: Five databases: MEDLINE, EMBASE, Cochrane Database, SCOPUS, Polish Medical Bibliography (Polska Bibliografia Lekarska; accessed on April 23, 2010) were systematically searched for SF-36 quality of life studies conducted in Polish cardiac patients. Only studies using original key answer and with published complete data for eight SF-36 dimensions were included. SF-6D utility scores were estimated based on SF-36 population data using the method published by Ara and Brazier in 2009. RESULTS: We initially identified 31 studies using SF-36 in Polish cardiac patients. In 14 studies proper SF-36 scoring algorithm was used and data for all eight domains were available. Data for 77 different health states related to five cardiac disorders were extracted: acute coronary syndrome, coronary artery disease, hypertension, atrial fibrillation and aortic valve disease. Each health state was described as value of 8 dimensions of quality of life and was converted to single-figure utility. Utility of acquired 77 health states ranged from 0.41 to 0.78. CONCLUSIONS: A catalogue of 77 health state utilities derived in Polish cardiac patients was estimated. It can be useful in pharmacoeconomic analyses conducted for cardiologic health technologies and should support reimbursement decision making in Poland. PCV110 HOW PATIENTS’ EXPERIENCE COULD INFLUENCE THEIR PREFERENCES TOWARD TREATMENT FOR ABDOMINAL AORTIC ANEURYSM: RESULTS FROM THE PREFER STUDY Scalone L1, Borghetti F2, Faggioli G3, Stella A3, Cortesi PA1, Mantovani LG4 1 University of Milano—Bicocca, Monza, Italy; 2University of Milan, Milano, Italy; 3Sant’Orsola Malpighi Hospital, University of Bologna, Bologna, Italy; 4CIRFF, Federico II University, Naples, Italy OBJECTIVES: Factors influencing preferences towards treatment of abdominal aortic aneurysm (AAA) were scarcely investigated. To identify how patients’ experience can affect their preferences for treatment of AAA. METHODS: within a multicentre Discrete-Choice-Study aimed to assess preferences toward AAA treatment the participant patients responded to 4 pair-wise choice sets comparing treatment options obtained from a factorial combination of 6 characteristics: type of anaesthesia (general vs. local); recovery time to everyday basic activities (2 vs. 4 days); risk of A362 re-intervention at 5 years (7% vs. 15%); complexity of follow-up (with vs. without CT examination); risk of major complications (2% vs. 5%); additional cost of intervention (c0 vs. c2000). Patients preferences were elicited before treatment (NAIF) and after treatment with open surgery (OPEN) or endovascular procedure (EVAR). a segmentation conditional logistic regression model was applied to estimate relative importance (RI) assigned to each characteristic according to patients’ experience with treatment. RESULTS: A total of 160 patients participated (aged 49–88 years, 91.7% male) from 9 hospitals. Fifty-two percent were NAIF at enrolment. Overall, half patients were treated with OPEN, half with EVAR. One most important characteristic, risk of procedural major complications, was considered more important by NAIF and EVAR than OPEN (RI = 31.7% and 22.0% vs. 11.0%). Additional costs were more important to OPEN (RI = 38.3%) than NAIF (25.2%) and EVAR (19.5%). Local anaesthesia was more important to EVAR (24.8%) than NAIF or OPEN (RI = 8.5%). Risk of repeating the procedure was similarly important to treated patients (RI = 17.5%), less to NAIF (RI = 12.3%). Recovery time was more important to OPEN (RI = 14.3%) and EVAR (RI = 10.3%) than NAIF (8.6%). Type of follow-up tests was least important, with OPEN and NAIF preferring less complex tests (RI = 4–10%), but EVAR preferring more complex tests (RI = 6%). CONCLUSIONS: Preferences depend on experience, knowledge and expectancies. Understanding patients’ preferences can help physicians to optimize benefits of treatments. PCV112 QUALITY OF LIFE AND DEPRESSION AMONG ADULTS WITH TYPE 2 DIABETES MELLITUS, HYPERTENSION, AND OBESITY Grandy S1, Fox KM2 1 AstraZeneca LP, Wilmington, DE, USA; 2Strategic Healthcare Solutions, LLC, Monkton, MD, USA OBJECTIVES: Individuals with type 2 diabetes mellitus (T2DM) are known to have poorer quality of life and more depressive symptoms than those without diabetes, yet the impact may be in part due to comorbid conditions. This study compared quality of life and depression among adults with T2DM and comorbid hypertension (HTN) and obesity with adults reporting T2DM only. METHODS: Respondents to the Study to Help Improve Early evaluation and management of risk factors Leading to Diabetes (SHIELD), a large US survey, self-reported their height, weight, and comorbid conditions, and completed the Short Form- 12 (SF-12) and Patient Health Questionnaire (PHQ-9, depression assessment). Respondents reporting T2DM and HTN and obesity (body mass index [BMI] ≥30 kg/m2) were identified and compared with a T2DM-only group. RESULTS: Respondents with T2DM and comorbid HTN and obesity (n = 1292) were similar to T2DM-only respondents (n = 349) in race, education, smoking, and cardiovascular disease history (all p > 0.05), but were younger and were more likely to be men and have lower income (P < 0.01). Respondents with T2DM, HTN and obesity had significantly lower Physical and Mental Component Summary scores (37.3 and 50.9, respectively) than T2DM-only respondents (45.8 and 53.5, respectively, P < 0.0001). Mean PHQ-9 scores were significantly higher among T2DM respondents with comorbid HTN and obesity (5.0 vs. 2.5, P < 0.0001), indicating greater depression burden. Approximately 16.5% of respondents with T2DM, HTN and obesity had moderate to severe depression (PHQ-9 scores ≥15), compared with 6.1% of respondents with T2DM only (P < 0.0001). CONCLUSIONS: SHIELD respondents with T2DM, HTN and obesity report a lower quality of both physical and mental health and more depression symptoms than the T2DM-only group. More research is needed to determine whether the poor quality of life and greater depression in this population affects selfmanagement of their diabetes and comorbid conditions. PCV113 HEALTH-RELATED QUALITY OF LIFE OF PATIENTS UNDERGOING PERCUTANEOUS CORONARY INTERVENTION COMPARED TO HEALTHY SUBJECTS IN HONG KONG Lee VW, Yan B, Choy A, Yeung S, Yu CM The Chinese University of Hong Kong, Shatin, Hong Kong OBJECTIVES: This study aimed to investigate the health-related quality of life (HRQoL) of coronary heart disease (CHD) patients and healthy subjects in Hong Kong. METHODS: This study was a one-year prospective cohort with two groups of study populations, the patient group and healthy subject group. Thirty-three eligible patients were recruited from Prince of Wales of Hospital. The HRQoL data using Short-Form 36-item Health Survey (SF-36) were collected just after percutaneous coronary intervention (PCI) (baseline) and one year post-PCI. Thirty-five healthy subjects who were recruited from two elderly centres and parents of School of Pharmacy (SOP) students, were also interviewed by SF-36. This study consists of two parts. In the first part, the HRQoL data of healthy subjects were compared with patients at baseline and 1 year post-PCI respectively. In the second part, the improvement in HRQoL of PCI patients at baseline and 1 year post-PCI was compared. RESULTS: In the first part of study, healthy subjects had significantly (P < 0.05) higher scores in role-physical (79.3 ± 36.6 vs. 53.8 ± 48.1), vitality (65.1 ± 23.8 vs. 49.5 ± 33.7) and role-emotional (93.3 ± 15.8 vs. 72.7 ± 39.5) than the patients at baseline. Patients’ HRQoL returned to similar level with healthy subjects after 1 year of PCI treatment. In the second part of the study, patients had significant (P < 0.05) improvements in bodily pain (12.7 ± 30.3) and vitality (16.1 ± 33.1) after 1 year. CONCLUSIONS: This study showed that the HRQoL of CHD patients was poorer than healthy population in Hong Kong, but returned to similar level as healthy population after 1 year of PCI treatment. In addition, significant HRQoL improvements had been seen in patients after 1 year of PCI, especially in bodily pain and vitality. 13th Euro Abstracts PCV114 ABSENTEISM AND IMPAIRED QUALITY OF LIFE IN CHRONIC VENOUS DISEASE IN ROMANIA Alegre P1, Puskas A2, Giurcaneanu C3, Andercou A4, Jantet G5 1 Science Union et cie (Servier), Suresnes, France; 2Targu Mures, Romania; 3Bucarest, Romania; 4Cluj, Romania; 5Paris, France OBJECTIVES: To evaluate the impact of chronic venous disease (CVD) on patients’ professional activity and its consequence on patients’ quality of life (QOL) in Romania. METHODS: A large observational, multicentre, descriptive survey in CVD was implemented in Romania in 2009. Impact of CVD on patients’ professional activities and QOL were documented based on a patient self-administered questionnaire. The CIVIQ-14 was used to assess QOL (score 0 for very good to 100 for bad QOL). Patient data were crossed with their general practitioner evaluations on the severity of the disease. RESULTS: Out of 2542 screened patients, 2294 filled in the questionnaire (90.2%). Among CVD patients, 7% had had hospitalization because of the disease, while 7% had changed their professional activities or had lost working days. Number of lost working days over the previous 5 years exceeded 1 week for most (40% lost >1 week, 35% >1 month, 16% <1 week). Patients who lost less than 1 working week, between 1 week and 1 month and more than 1 month due to CVD had a worsened QOL (respectively 28.71 ± 24.32, 31.4 ± 17.73 and 31.8 ± 16.69 CIVIQ-14 score) QOL scores paralleled the severity of CVD, going from 15.64 ± 14.94 in C0s patients (symptoms only, according to the CEAP) to 51.79 ± 30.43 in C6 (patients with ulcer). The presence of pain decreased QOL (30.49 ± 19.04). QOL scores associated to work absenteeism varied from 21.98 ± 15.05 to 49.49 ± 20.06 depending on absenteeism frequency (from 1 to 3 times respectively) CONCLUSIONS: CVD generates an important negative professional impact in Romania with productivity losses. Quality of life impairment occurs from the first symptoms of CVD and it is correlated to the work absenteeism duration and frequency. Three times out of work is associated to a QOL score comparable to the one quoted between grades 4 and 5 of the CEAP classification. PCV115 ASSESSMENT OF THE ASSOCIATION BETWEEN SUMMARY SYMPTOM MEASURES AND HEALTH STATUS OF PATIENTS WITH ISCHEMIC HEART DISEASE Kaakeh R, Erickson SR University of Michigan, Ann Arbor, MI, USA OBJECTIVES: This study compared five methods of summarizing symptom data and the association of these measures with the outcome health related quality of life (HRQL) in patients who had experienced an acute coronary syndrome event (ACS). METHODS: All patients listed in an ACS registry (treated for ACS and discharged from a university affiliated hospital) during a 3 year period were mailed a questionnaire that assessed symptom status, demographics, cardiac functional status and therapy, comorbidity, ACS type and HRQL (SF-8 and EQ-5D). The symptom survey included 24 symptoms and measured the occurrence, frequency, and distress associated with each symptom during the previous four weeks. Symptom summary measures included summed, averaged, and multiplicative (Symptom Product) values of each symptom characteristic. Bivariate correlations and multivariate regression analysis were used to assess the association of each symptom summary measure with HRQL, controlling for other patient, disease, and treatment characteristics. RESULTS: A total of 490 of 1217 patients (40%) responded. The mean age was 65.7 ± 11.3 years; 67.7% male;94.0% Caucasian;PCS-8 53.7 ± 10.5; MCS-8 49.5 ± 9.5; and EQ-5D VAS 73.6 ± 20.5. The Symptom Sum = 7.8 ± 4.9, Symptom Frequency Average = 2.5 ± 0.72, Symptom Distress Average = 2.2 ± 0.8, Symptom Distress Sum = 18.6 ± 14.8, and Symptom Product = 51.7 ± 48.3. Bivariate correlations demonstrated that the Symptom Sum and Symptom Product had the highest correlation (>0.54) with the HRQL domains, while Symptom Frequency Average was the lowest (0.24–0.33). The same pattern was seen with the multivariate regression model. The model adjusted R-square for the Symptom Sum and Symptom Product measures were higher than the other measures, with the Symptom Product having the highest correlation. The R-square values ranged for these two measures depending on the HRQL domain; PCS (0.38–0.44), MCS (0.33–0.35), EQ-5D (0.34–0.39). CONCLUSIONS: Researchers may consider using Symptom Sum or a Symptom Product measures when attempting to summarize the association of symptoms with HRQL in ischemic heart disease patients. PCV116 QUALITY OF LIFE IN CHRONIC SYMPTOMATIC HEART FAILURE PATIENTS IN SPAIN Marti B1, Delgado J2, Oliva J3, Llano M4, Pascual P5, Comin J6, Grillo JJ7, Diaz Molina B8, Culebras J9, Martínez de la Concha L10, Manito N11 1 Medtronic Iberia, Madrid, Spain; 2Hospital 12 de Octubre, Madrid, Spain; 3Univesidad de Castilla La Mancha, Toledo, Spain; 4Hospital Universitario Marqués de Valdecilla, Santander, Spain; 5Hospital Universitario Virgen de la Arrixaca, Murcia, Spain; 6Hospital del Mar (IMIM), Barcelona, Spain; 7Hospital Universitario Nuestra Señora de la Candelaria, Santa Cruz de Tenerife, Spain; 8Hospital Universitario Central de Asturias, Oviedo, Spain; 9Hospital Universitario Insular, Las Palmas De Gran Canariaý, Spain; 10Hospital Universitario Infanta Cristina, Badajoz, Spain; 11Hospital de Bellvitge, Barcelona, Spain OBJECTIVES: To analyze, for the first time in a large Spanish population of heart failure patients, quality of life according to NYHA class II, III or IV using generic and specific quality of life questionnaires. METHODS: A descriptive analysis of a 13th Euro Abstracts multicenter, prospective observational study was performed. Patients who met inclusion criteria were followed-up for a period of 12 months, with 3 visits programmed at baseline, 6 months and 12 months. a total of 9 Spanish hospitals were involved in the study. Questionnaires used to measure quality of life were: EQ-5D (generic questionnaire), Minessotta living with heart failure- MLWHF (specific questionnaire) and Barthel Index (index of independence). RESULTS: A total of 450 patients were included, 76% men, mean age was 62.6 years. 66.1% were in NYHA class II, 32.7% NYHA class III and 1.1% NYHA class IV. Prevalence of ischemic cardiopathy was 33.5%. 35% of patients had an implantable device (ICD, RCT or pacemaker). Significant differences were observed in Barthel Index’s scores depending on class: class I 97.6 ± 6.7 vs. Class III-IV 91.8 ± 14.6. Related to EQ-5D tariffs, individuals in class II had a mean value of 0.8407 ± 0.1887 (out of 1), and mean VAS value of 60.6 ± 17.39 (out of 100). Individuals in NYHA class III or IV had a mean score of 0.6624 ± 0.2848, and mean VAS value of 51.93 ± 17.15 (out of 100). In a MLWHF questionnaire analyses, patients in NYHA class II showed a mean score of 32.52 ± 20.69 while patient in NYHA class III-IV showed a mean score of 50.44 ± 18.23. CONCLUSIONS: Our findings showed the results from the three questionnaires were generally consistent with each other and values reported from other countries in the literature. PCV117 HEALTH-RELATED QUALITY OF LIFE IN PATIENTS WITH PULMONARY ARTERIAL HYPERTENSION Mychaskiw MA1, Berger A2, Mardekian J1, Hwang LJ1, Oster G2 1 Pfizer Inc, New York, NY, USA; 2Policy Analysis Inc., Brookline, MA, USA OBJECTIVES: To characterize health-related quality of life (HRQoL) in patients with pulmonary arterial hypertension (PAH) in relation to persons of similar age and gender in the general United States (US) population. METHODS: Data were obtained from a large phase III clinical trial in which patients (n = 274) were randomized to sildenafil or placebo for 12 weeks. We characterized HRQoL using responses to the Short Form-36 General Health Survey (SF-36) obtained at baseline; the SF-36 addresses eight HRQoL domains: physical functioning (PF), role functioning–physical (RP), bodily pain (BP), general health (GH), vitality (VT), social functioning (SF), role functioning–emotional (RE), and mental health (MH). Values for these eight domains were then standardized to US population norms for persons aged 45–54 years (mean age of patients in trial was 49 years). RESULTS: A total of 39% and 58% of patients were designated at baseline as Functional Class (FC) II and III, respectively (<1% and 3% were designated as FC I and IV), with 75% women. Standardized mean scores were 30.4 for PF, 27.9 for RP, 47.8 for BP, 34.2 for GH, 41.2 for VT, 38.9 for SF, 34.9 for RE, and 43.7 for MH. In comparison with the US population, whose mean score for each domain = 50, PF, RP, GH, SF, and RE were all substantially worse for PAH patients. Additionally MH and VT were marginally worse among PAH patients, and BP was comparable to US norms. CONCLUSIONS: With the exception of bodily pain and mental health, PAH patients have substantially poorer SF-36 scores than their peers in the US general population, especially in terms of physical functioning, role functioning-physical, role functioning-emotional, and general health. PCV118 TO DETERMINE THE EFFECT OF DIFFERENT DOSES OF POMEGRANATE ON BODY WEIGHT AND BLOOD PRESSURE Malik M, Hussain A Hamdard University, Islamabad, Punjab, Pakistan OBJECTIVES: The aim of the present study was to evaluate the effect of 700 mg and 1400 mg pomegranate seed capsules on body weight and blood pressure for a period of 60 days at time interval 7, 14, 21, 28, 45 and 60 days respectively. METHODS: At the beginning of the study twenty four healthy human volunteers were selected. On the basis of BMI two broader groups, i.e. group 1 taking 700 mg pomegranate capsules and group 2 taking 1400 mg pomegranate capsules were designated and both the groups were subdivided on the basis of diet control and exercise i.e. group1 (IA, IB) and group 2(IIA, IIB). Group IA and IIA took pomegranate capsules without diet and exercise while IB and IIB took the capsules with diet and exercise. Hypertensive patients from these groups were included in both groups who were categorized and evaluated separately and compared with normal individuals. RESULTS: The result of investigation indicated that both doses of pomegranate seeds capsules had significantly reduced the body weight and blood pressure profile (both systolic and diastolic) in all groups when compared with their baseline values. However both doses caused equal reduction of body weight and blood pressure as no significant difference was present in both doses (p-value > 0.05). CONCLUSIONS: Thus on the basis of results obtained after conducting our study it was concluded that pomegranate seed capsules have established their efficacy and safety profile for obesity control and hypertension management. PCV119 SURVEY AT RETAIL PHARMACIES LEVEL ASSESSING QUALITY OF LIFE OF VKAS PATIENTS Delaitre O1, Haim M1, Samama CM2 1 Boehringer Ingelheim France, Paris, Ile de France, France; 2Hotel Dieu University Hospital, Paris, Ile de france, France OBJECTIVES: To measure how VKA’s patients assess their treatment in terms of medication intake constraints, monitoring and psychological impact related to perceived risk. METHODS: Questionnaires to patients under VKAs for at least 3 months. A363 Questionnaires were distributed through 150 retail pharmacies spread over all France from November 2009 to March 2010. a random distribution of questionnaire was made. RESULTS: 1094 questionnaires were administered. Study included 58% men and 42% women aged >65 years for 72%; patients aged more than 75 years represented 43%. Patients were treated in majority by fluindione (82.3%) and for more than 3 years (65.1%). Patients were treated in majority for cardiac rhythm disorder (40%). More than ¾ (77%) of VKAs patients consider their treatment being a burden even though being vital. Treatment is judged as being a constraint because of the obligation to adapt dosages according to coagulation monitoring (42%), and due to the necessity to contact their physician after each blood sample (25%). The requirement to permanently have a monitoring is considered as constantly reminding the associated risk (65%), expectation of coagulation monitoring results generate anxiety for more than ¼ of patients (27%). Majority of patients acknowledge that regular blood sampling is a constraint (53%) and that the treatment would be more convenient without these constant blood samplings (60%). In parallel, 38% of patients interviewed consider that the VKA treatment constraint is reinforced by the fact that they cannot eat what they would like. Patient preference for a new treatment is in priority: a treatment without regular blood sampling (62%); a treatment for which the daily anticoagulant dosage does not need adaptation (60%). CONCLUSIONS: Treatment by VKAs is perceived as burdensome and risky by a majority of patients due to the constant monitoring required and the uncertainty of the good blood dosage. CARDIOVASCULAR DISORDERS – Health Care Use & Policy Studies PCV120 COMBINED MEDICATION CHOICE AND PATIENT PERSISTENCE IN HYPERTENSION THERAPY: EVIDENCE OF REAL-LIFE EFFECTIVENESS Molnár MP1, Dankó D1, Katona L2 1 Corvinus University of Budapest, Budapest, Hungary; 2Hungarian National Health Insurance Fund Administration, Budapest, Hungary OBJECTIVES: While hypertension treatment guidelines emphasize the medical benefits of combined medication choice, consumption patterns of antihypertensive drugs raise worldwide several questions related to real-life therapeutic and cost-effectiveness. Hungary has a relatively poor track record: international comparative studies show Hungarian patients’ adherence to treatment to fall behind not only desirable targets but also international average values. In our analysis, we start from the hypothesis that medical benefits achievable through combined therapy are weakened through inadequate patient adherence, and we investigate into the health loss caused to Hungarian patients by insufficient persistence in combined hypertension therapy. Our goal is to determine how the real-life consumption patterns of combination hypertension therapy impact real-life therapeutic effectiveness and cost-effectiveness. METHODS: We use itemized prescription-level data from the Hungarian National Health Insurance Fund Administration’s (NHIFA) database. Our research covers patients whose therapy was initiated during 2008 or 2009 with high-value fix-dosage (one-pill) and non-fix-dosage (multiple-pill) angiotensin-converting enzyme inhibitor (ACEI) and calcium channel blocker (CCB) combinations, or angiotensin-receptor blocker (ARB) and CCB combinations. RESULTS: Adherence to treatment is measured as persistence on the initiating combination therapy. Firstly, ‘hard’ persistence (which does not allow for add-on therapies) and ‘add-on’ persistence (which allows for additional active substances being added to the therapy regime) are separately calculated, both for fixdosage and non-fix-dosage combinations. Secondly, if the initiating therapy is a fixdosage combination, parallel consumption of monocomponents is also analyzed. Thirdly, we carry out a switch analysis to determine the active substances that patients had been taking before receiving a fix-dosage combination, or take after abandoning the fix dosage combined therapy. CONCLUSIONS: We hypothesize that substantial societal loss is caused by patient non-adherence in hypertension treatment and clinical advantages of combination therapies are significantly impaired. Our research will deliver exact calculations for the extent of the societal loss. PCV121 A SURVEY OF PHYSICIANS’ ATTITUDES TOWARD THE CONTROL OF CARDIOVASCULAR RISK FACTORS. THE EURIKA STUDY. Dallongeville J1, Banegas JR2, Guallar E3, Borghi C4, De Backer G5, Halcox JP6, Massó-González EL7, Perk J8, Steg PG9, Rodriguez Artalejo F2 1 Institut Pasteur de Lille, Lille, France; 2Universidad Autónoma de Madrid, Madrid, Spain; 3 Johns Hopkins University, Baltimore, MD, USA; 4Policlinico Universitario Sant’Orsola, Bologna, Italy; 5University of Gent, Gent, Belgium; 6Cardiff University, Cardiff, UK; 7 AstraZeneca Farmacéutica Spain S.A, Madrid, Spain; 8Oskarshamns Hospital, Oskarshamn, Sweden; 9Centre Hospitalier Bichat-Claude Bernard, Paris, France OBJECTIVES: Cardiovascular risk factors remain poorly controlled across Europe despite clinical guidelines. The European Study on Cardiovascular Risk Prevention and Management in Daily Practice (EURIKA) investigated the use of cardiovascular risk assessments tools and guidelines, and explored factors limiting their use. METHODS: Physicians (n = 806) from 12 European countries answered questions regarding their work setting, their assessment of patients with cardiovascular risk factors, and their use of risk calculation tools and clinical guidelines. RESULTS: Participating physicians worked in primary care centres or outpatient clinics; 63.8% were GPs. The majority (69.3%) reported using global risk calculation tools. Written charts were the preferred method (69.9%), followed by software (33.0%). The most popular tools were SCORE (European Society of Cardiology [ESC]; 52.4%), A364 Framingham (16.9%) and the European Hypertension Chart (31.3%). Reasons for not using risk tools included time constraints (60.8%), not being convinced of their utility (22.1%) and lack of knowledge (20.0%). a high percentage of physicians believed the algorithms have limitations (72.4%); 92.1% believed they lead to overlooking other risk factors and 69.3% believed they cannot be used to calculate risk in elderly patients. The most commonly used clinical guidelines were those from the ESC (CVD Prevention in Clinical Practice [56.3%] and Management of Arterial Hypertension [29.9%]) or from local authorities (17%). In total, 12.9% of physicians reported not using guidelines; reasons included the wide choice or uncertainty of which to use (47.5%), time constraints (33.7%), lack of knowledge (27.7%) and a perception that they are unrealistic (23.8%). CONCLUSIONS: Time constraints, perceived utility and inadequate knowledge were common factors limiting the use of cardiovascular risk evaluation tools and assessment guidelines. Better compliance with risk assessment tools may reduce the high proportion of patients with poorly managed cardiovascular risk factors. PCV122 THE IMPACT OF A CLINICAL PHARMACIST INTERVENTION ON LIPID-LOWERING IN A PRIMARY CARE SETTING Triki N1, Shani S1, Rabinovich-Protter D2, Mossinson D2, Kokia E2, Greenberg D1 1 Ben Gurion University of the Negev, Beer-Sheva, Israel; 2Maccabi Healthcare Services, Tel-Aviv, Israel OBJECTIVES: The Adult Treatment Panel III guidelines suggest that the goal of lowdensity lipoprotein cholesterol (LDL-C) in patients with both cardiovascular disease and diabetes is <100 mg/dL. Many patients remain poorly controlled despite various interventions in primary care, including statin therapy and health behavior modification. We evaluated the impact of adding a clinical pharmacist intervention to usual care on LDL-C control and treatment costs in diabetic cardiac patients. METHODS: We prospectively compared a clinical pharmacist intervention in 138 patients with a matched control sample of 353 patients receiving usual care in Maccabi Healthcare Services (MHS) in Israel. Patients with cardiovascular disease and diabetes and LDL-C levels >100 mg/dL were identified from the MHS’s computerized database. The clinical pharmacist reviewed patients’ clinical charts and discussed the recommendations to improve hyper-lipidemic control with the patients’ primary-care practitioners. The recommendations were given every three months for a one-year period. The primary clinical endpoint was reaching LDL-C goal. Clinical outcomes and overall treatment costs in both groups were evaluated at the end of the study year. RESULTS: During the study year, 67% of the patients in the intervention group reached the LDL goal vs. only 54% in the control group (p = 0.014). LDL target was reached three months earlier in the intervention group as compared with control patients (0.710 year vs. 0.992 year, respectively; log-rank test: p = 0.015). However, at the end of the study year, LDL target was maintained in approximately 50% of patients in both groups. Overall treatment costs (physician visits, hospital and emergency room admissions, lab tests, medications) were 14% lower in the intervention group and 11% higher in the control group as compared to the year prior the intervention. CONCLUSIONS: A clinical pharmacist intervention in high-risk patients may result in clinical improvements and lower treatment costs. These results demonstrate the high-value of clinical pharmacist involvement in patient treatment. PCV123 EPIDEMIOLOGICAL STUDY OF EUROPEAN CARDIOVASCULAR RISK PATIENTS: DISEASE PREVENTION AND MANAGEMENT IN USUAL DAILY PRACTICE—TURKISH RESULTS OF EURIKA STUDY Abaci A Gazi University, Ankara-on behalf of EURIKA Study Group, Turkey OBJECTIVES: To assess management of cardiovascular risk factors (CVRF) in daily clinical practice and to identify areas of potential improvement in primary prevention of CVD. METHODS: A total 663 patients >50 years of age (59.4 ± 7.6 years; 47.2% males) with at least one additional CVRF and 67 physicians (mean age: 40.7 ± 8.6 years; 82.1% males) were included from Turkey in the multicentre, multinational, cross-sectional epidemiological EURIKA study (NCT00882336) conducted across Europe. Management and control of classical, emergent and psycho-social CVRF, use of CV risk assessment by the physicians as well as barriers for estimating and using global cardiovascular risk scores were identified. RESULTS: Total CV risk assessment in Turkish patients was stated to be performed by 48.5% of the physicians mostly by chart (71.9%) and mainly for an advice on healthy lifestyle (84.4%) and to decide on antihypertensive (78.1%) or lipid-lowering treatment (75.0%). Time constraint for global CV risk evaluation was the main reason (73.5%) for the lack of assessment identified by the physicians. a total of 514 patients (77.5%) were classified to have high CV risk by the physicians using a local (7.6%) or the recent European Guidelines on Cardiovascular Disease Prevention in Clinical Practice (ESC 2007) (80.0%). Although global cardiovascular risk was said to be under control in 75.5% of the patients, satisfying control of CV risk factors was evident in only 56.7% while the overall percentage of the patients who were aware of their CV risk was 69.5%. CONCLUSIONS: Apparently targets defined in guidelines are not sufficiently met and there is clear need for better management of high risk patients. Development of better structured and more realistic, simple and credible national guidelines adapted to suit local medical and economic conditions should be encouraged. 13th Euro Abstracts PCV124 IMPACT OF REGIONAL MEASURES IN THE SALES OF THE RENINANGIOTENSIN SYSTEM ANTAGONISTS IN SPAIN Gimenez E, Garrido E, Lindner L, Sabater J, Vieta A IMS Health, Barcelona, Spain OBJECTIVES: To analyze the impact of regional measures in the sales of the reninangiotensin system antagonists in Spain. METHODS: Regional measures from each of the 17 Spanish Autonomous Regions (AR) were identified by searching on health services’ websites. a Regional Measurement Scale (RMS) in rational use of drugs (RUD) was constructed. Values ranged from 0 to 24 considering the number and the specificity of the applied politics. The 2009 market share of the Angiotensin-II Receptor Antagonists (AIIRA) vs. Angiotensin-Converting Enzyme (ACE) inhibitors was analyzed. AR market shares and RMS correlations were calculated. Correlated AR Market shares were adjusted according to the values of RMS using a linear regression model for the AR with correlation between the RMS and the market shares. The decrease in market share on the AIIRA by RMS point was determined by the slope coefficient of the regression (β). RESULTS: Health plans, clinical guidelines, pharmacotherapeutic guides, health technology assessments and therapeutic newsletters, promoting the prescription of ACE inhibitors rather than AIIRA, were identified. In 14 out of 17 AR, the correlation between the RMS and the market shares was statistically significant (r = 0.55, p = 0.004). Of these, three AR scored 0, five 1 to 10, four 10 to 20, and two 22 and 23 points respectively. a 0.52% decline in the market share AIIRA was observed for each point in the RMS. The maximum impact observed in the market share was 11.96%. Considering that the variability in the AIIRA market share was 24.1%, half of it was explainable by the establishment of RUD measures. CONCLUSIONS: Results support the fact that not only the establishment of regional measures is important but also the way they are implemented. PCV125 PROTEIN-C-REACTIVE AS A MARKER OF INFLAMMATION AND CARDIOVASCULAR DISEASE IN PATIENTS WITH SCHIZOPHRENIA: A CROSS-SECTIONAL ANALYSIS OF A HEALTH CARE PROVIDER ADMINISTRATIVE CLAIM DATABASE Sicras-Mainar A1, Blanca Tamayo M2, Rejas J3, Navarro Artieda R4 1 Directorate of Planning, Badalona Serveis Assistencials, Badalona, Barcelona, Spain; 2 Department of Psychiatry, Badalona Serveis Assistencials, Badalona, Barcelona, Spain; 3Pfizer España, Alcobendas/Madrid, Spain; 4Department of Clinical Documentation, Hospital Germans Trias i Pujol, Badalona, Barcelona, Spain OBJECTIVES: Interest in cardiovascular diseases (CVD) in patients with schizophrenia has recently become growing among psychiatrists due to its documented incremental mortality for these causes. Identification of markers for such disorders seems, therefore, reasonable. Serum Protein-C-Reactive (PCR) levels have been determined as a marker of inflammation in individuals with CVD and/or at high risk for developing it. However, it is unknown the role of this protein in schizophrenics. Thus, the goal of this research was to explore the use of PCR as a marker of inflammation and CVD in patients with Schizophrenia. METHODS: A cross-sectional analysis of the BSA administrative claim database was conducted including all men and women, >18 years, with a schizophrenia spectrum disorders (by DSM-IV criteria) diagnosis. PCR measurement together with socio-demographics, evolution, medical history, 10-years CVD risk (Framingham equation) and biochemistry data was extracted for analysis. RESULTS: A total of 705 patients [53.0% men, 48.2 + 15.8 years (mean + SD), 5.9 + 3.2 years of evolution, 79.7% on atypical drugs] met criteria for analysis. Mean 10-year CVD risk was high; 11.9% + 5.7% and mean PCR levels were 2.6 + 2.5 mg/L with 30.4% showing values above normal’s (>3 mg/L). Unadjusted PCR slightly correlated with CVD risk; r = 0.171, P < 0.001. After adjusting by age, sex, evolution, smoking and anti-inflammatory drugs treatment, PCR was linearly associated with 10-year CVD risk stratified by its level of risk (low, moderate, high/very high); respectively, 2.3 (95% CI: 2.1–2.5), 3.1 (2.6–3.5) and 3.7 (3.2–4.1) mg/L; F = 13.5, P < 0.001. Patients with known CVD showed also higher PCR levels; 3.7 (2.9–4.5) vs. 2.5 (2.4–2.7) mg/L, p = 0.008, and higher probability of values above normal’s; Odds Ratio = 4.71 (2.01–11.04), P < 0.001. CONCLUSIONS: High PCR levels (above normals) were associated with both known CVD and high/very high 10-year risk of CVD event in patients with schizophrenia. Then, PCR might be a marker of inflammation and CVD in this psychiatric disorder. PCV126 USE OF PROTON PUMP INHIBITORS (PPIS) IN ACUTE CORONARY SYNDROME PATIENTS TREATED WITH CLOPIDOGREL IN GERMANY, FRANCE, AND THE UNITED KINGDOM Mccollam P1, Nasuti P2, Rex J2, Gaskin M2 1 Eli Lilly & Company, Indianapolis, IN, USA; 2IMS Health, London, UK OBJECTIVES: There is controversy in the literature regarding the effects of concomitant use of proton-pump inhibitors (PPIs) potentially reducing the clinical effectiveness of clopidogrel. Also in question is whether the effect occurs with all PPIs or only omeprazole The purpose of this study is to assess the usage of PPIs in combination with clopidogrel in Acute Coronary Syndrome (ACS) patients in the primary care setting. METHODS: This was a retrospective study using IMS Disease Analyzer databases in France, Germany and the UK. These longitudinal patient databases provide information from continuing physician and patient interaction on consultations, diagnoses and treatments within primary care. RESULTS: From April 2008 to April 2009, 57% of post-ACS clopidogrel patients also had a PPI prescribed (France 13th Euro Abstracts 54% and Germany 45%). An overall increase in co-prescribing of clopidogrel and PPIs was observed in all countries from April 2006 to April 2009 except for France where omeprazole co-prescribing decreased in 2008–2009. The most frequently prescribed PPI in combination with clopidogrel was omeprazole in all three countries (58% in the UK, 36% in France and 65% in Germany) The second most frequently used PPI was lansoprazole in the UK (36%), pantoprazole in Germany (21%) and esomeprazole in France (28%). The proportion of clopidogrel patients who were co-prescribed lansoprazole was 36% in the UK and only 11% in France and 2% in Germany. CONCLUSIONS: Overall prescribing of PPIs with clopidogrel after ACS is common in the UK, France and Germany. Since the publication of the OCLA study in 2008 no decrease in concurrent prescribing of clopidogrel and omeprazole was observed except in France. Revisions of national guidelines and the clopidogrel label in 2009 may further affect prescribing of PPIs, and especially omeprazole, with clopidogrel in the future. PCV127 INCREASING NURSE STAFFING LEVELS IN BELGIAN CARDIAC SURGERY CENTERS: A COST-EFFECTIVE PATIENT SAFETY INTERVENTION? Van den Heede K1, Simoens S2, Vleugels A1, Sermeus W1 1 Katholieke Universiteit Leuven, Leuven, Belgium; 2K.U. Leuven, Leuven, Belgium OBJECTIVES: A previous study indicated that increasing nurse staffing levels in Belgian general cardiac postoperative nursing units was associated with lower mortality rates. The aim of this study is to conduct a cost-effectiveness analysis of increasing nurse staffing levels to the level of the 75th percentile in Belgian general cardiac postoperative nursing units from a hospital perspective. METHODS: The intervention was an increase in the number of nursing hours per patient day to the 75th percentile for nursing units staffed below that level. The comparator was a “do nothing” alternative. Data on nurse staffing levels were extracted from the Belgian Nursing Minimum Data set and data on in-hospital mortality from the Belgian Hospital Discharge Database. The number of life-years gained was calculated by multiplying the number of avoided deaths by the life expectancy of patients having a coronary artery bypass graft and patients having heart valve procedures. To this effect, survival rates were derived from the literature. National cost estimates for the year 2007 were used. Results were expressed in the form of the additional costs per avoided death and the additional costs per life-year gained. RESULTS: The costs of increasing nurse staffing levels to the 75th percentile in Belgian general cardiac postoperative nursing units amounted to c1,211,022. Such nurse staffing levels would avoid an estimated number of 45.9 (95% CI: 22.0–69.4) patient deaths per year and generate 458.86 (95% CI: 219.93–693.79) life-years gained annually. This corresponds with incremental cost-effectiveness ratios of c26,372 per avoided death and c2,639 per life-year gained. Sensitivity analyses indicated that the incremental cost-effectiveness ratio was robust to changes in input parameters. CONCLUSIONS: Increasing nurse staffing levels appears to be a costeffective intervention as compared to other cardiovascular interventions. PCV128 MODELING OF HEALTH SERVICE RESEARCH RESULTS WITH UPDATED COST DATA—THE GERSHWIN STUDY EXAMPLE Brüggenjürgen BH, Willich SN, McBride D University Medicine Berlin, Charité, Berlin, Berlin, Germany OBJECTIVES: Health services research is an expanding field of interest for recently introduced medical technologies. Complementing efficacy results with effectiveness data illustrates the clinical contribution of a new technology in the real-life setting. With effectiveness results being relatively stable over time, the change in reimbursement and price decreases have a substantial impact on initial cost-effectiveness considerations. The three-year GERSHWIN study (GERman Stent Health outcomes WIthin Normal practice) was designed to determine long-term clinical outcome and economic consequences of Sirolimus-eluting stents (SES) versus bare-metal stents BMS in the treatment of CAD from a societal prospective. Economic analysis resulted in an ICER of c29,868 per avoided major adverse coronary events (MACE) based on 2003 to 2005 prices. Due to substantial price reductions, a remodelling with current prices would increase applicability of results by decision-makers. METHODS: Complete intention-to-treat study data were valued and analyzed with 2009 prices for the main cost drivers (hospital case rate, incremental stent cost and generic clopidogrel price). Sensitivity analyses were conducted to evaluate the robustness of the results and included a low- and high-cost scenario. RESULTS: Initial hospitalization costs were c1157 higher per patient receiving SES, driven primarily by the incremental acquisition cost for SES compared BMS. Follow-up direct costs were similar in both groups of patients, however the indirect costs in SES patients were significantly lower. Overall 36-month MACE-related costs were c283 not significantly higher in the SES group (p = 0.62). The ICER based on 2009 cost data resulted in c5320 per MACE avoided. Variation of incremental SES prices between c650 and c750 showed no significant impact. CONCLUSIONS: Incorporating updated cost data can drastically change the ICER generated from health services research projects. Modelling with current cost data is an essential contribution for iterative cost-benefit assessment procedures. A365 PCV129 TELEHEALTH—EARLY EVALUATION FINDINGS Beale S1, Tatlock S1, Wheeler K2, Ryan J2 1 University of York, York, North Yorkshire, UK; 2NHS, York, North Yorkshire, UK OBJECTIVES: NHS North Yorkshire and York commissioned York Health Economics Consortium to carry out an evaluation of their recent Telehealth pilot involving patients with long-term conditions. The findings from the evaluation will be used to inform plans for the further rollout of Telehealth in North Yorkshire. METHODS: Three approaches were employed: a questionnaire was used to obtain details about the Telehealth experiences of users and their carers; Case Managers and Community Matrons with Telehealth patients on their caseloads were interviewed about referral criteria and the impact of Telehealth on their role and patient care; Health care resource use data were analyzed. Some caution should be used when generalizing findings due to the small sample size and the specific characteristics of pilot Telehealth recipients. RESULTS: Forty-eight questionnaires were handed out and twenty returned (42%). Overall, respondents were happy with Telehealth (90%). Results showed that the installation process had been smooth (75%); individuals had received sufficient tuition (95%); they were confident using the equipment (95%); and were happy with the service received from the monitoring centre (95%). Telehealth gave users peace of mind and helped them to manage their own health condition. However, a number of users had experienced technical issues (during installation and when taking daily measurements). Although there had been some teething problems, mainly in relation to the installation process and the monitoring system, clinicians were broadly supportive of Telehealth. There was a repeated view that individual patient characteristics needed to be taken into account when identifying patients who would benefit from Telehealth. The impact of Telehealth on health care resource use was difficult to determine within the eight month timescale of the pilot. CONCLUSIONS: Telehealth gives users peace of mind and helps them manage their health; however, its impact on health care resource use is still unclear. PCV130 PRACTICE PATTERNS AND QUALITY OF LIFE IN ACUTE CORONARY SYNDROME PATIENTS IN 2008–2009: BASELINE RESULTS FOR AUSTRIA FROM THE ANTIPLATELET TREATMENT OBSERVATIONAL REGISTRY II (APTOR II) Ammer M1, Paget MA2, Hronkova M3, Norrbacka K4, Eber B1 1 Klinikum der Kreuzschwestern Wels, Wels, Austria; 2Lilly France S.A., Suresnes Cedex, France; 3Eli Lilly & Company, Vienna, Austria; 4Eli Lilly & Company, Vantaa, Finland OBJECTIVES: This analysis aims to explore management of acute coronary syndromes (ACS) from acute event to hospital discharge in Austria, and to measure Quality of Life (QoL) at discharge. METHODS: This 12-month international, prospective, observational study recruited ACS patients in selected hospitals undergoing percutaneous coronary intervention (PCI), April 2008–March 2009, capturing practice patterns, resource use and QoL. RESULTS: A total of 148 ACS-PCI patients (out of the 152 recruited) were eligible: median age 60 yrs (IQR 51–69), median weight 80 kg (IQR 70–89), 20% female, 28% Type II diabetics, and 17% prior myocardial infarction (MI). Index diagnosis was: unstable angina or non-ST-elevation MI (UA/ NSTEMI)-44% and ST-elevation MI (STEMI)-56%. Almost all patients (96%) received stents: 28% bare metal stents only, 70% drug eluting stents only and 2% both. Time from start of ACS symptoms to PCI was ≤3 days in 86% of UA/NSTEMI patients and ≤1 day in 98% of STEMI patients. Oral antiplatelet medications with loading dose (LD) used: aspirin-97% and clopidogrel-91%. Clopidogrel LD was administered in the ambulance-9%, previous hospital-10%, emergency room-41%, CCU or ICU-31%, catheterization lab-5%, or other ward-5%. LD was administered between 6 hours before to 6 hours after PCI in 82% of cases. The first clopidogrel LD was 600 mg in 85% and 300 mg in 10% of cases and in-hospital maintenance dose was 75 mg in 97%. At time of hospital discharge, 97% of the discharged patients were prescribed clopidogrel (discharge dose 75 mg for all patients except one). QoL in discharged patients was good: median EQ-5D health state index at 1.00 (IQR 0.81–1.00). CONCLUSIONS: These real life data reflect treatment patterns among ACS patients managed by PCI in selected hospitals in Austria in 2008–2009. Timing and place of loading of antiplatelet agents differ. The QoL of patients at discharge was high. PCV131 AN INTERNATIONAL COMPARISON OF DUAL ANTIPLATELET USE BY STENT TYPE AT 6 MONTHS FOLLOWING HOSPITAL DISCHARGE AFTER ACUTE CORONARY SYNDROME: RESULTS FROM THE ANTIPLATELET TREATMENT OBSERVATIONAL REGISTRY II (APTOR-II) Pavlides G1, Coufal Z2, Mohacsi A3, James S4, Zeymer U5, Paget MA6, Goedicke J7, Norrbacka K8, Berkenboom G9 1 Onassis Cardiac Surgery Center, Kallithca, Greece; 2Bata’s Regional Hospital, Zlin, Czech Republic; 3Hungarian Institute of Cardiology, Budapest, Hungary; 4Uppsala University Hospital, Uppsala, Sweden; 5Herzzentrum Ludwigshafen, Ludwigshafen, Germany; 6Lilly France S.A., Suresnes Cedex, France; 7Eli Lilly & Company, Hamburg, Germany; 8Eli Lilly & Company, Vantaa, Finland; 9U.L.B Erasme Hospital Brussels, Brussels, Belgium OBJECTIVES: Current European Society of Cardiology Guidelines recommend dual antiplatelet therapy for 12 months for patients with acute coronary syndrome (ACS); however, reimbursement for antiplatelet therapy differs by EU country and is dependent upon the use of bare metal (BMS) or drug-eluting (DES) stents during percutaneous coronary intervention (PCI). Dual antiplatelet (clopidogrel + aspirin) treatment A366 patterns were assessed at 6 months following hospital discharge for patients who received only BMS and for patients with at least 1 DES. METHODS: A prospective, observational registry of 11 European countries grouped into 6 clusters recruited ACS patients undergoing PCI from 2008–2009. Interventional cardiologists collected data from ACS events up to hospital discharge, and primary care physicians and cardiologists collected 6-month data. RESULTS: Of 3042 eligible patients at baseline, 2964 (97%) had 6-month data, of which 2842 (96%) received a stent at index PCI. The percentage of patients with at least 1 DES was 17% in the Czech Republic, 31% in Germany, 33% in Austria-Hungary, 33% in Belgium-The Netherlands, 37% in the Nordic countries, and 81% in Greece. The percentage of patients taking dual antiplatelet therapy at hospital discharge ranged from 93% in Belgium-The Netherlands to 100% in Greece for BMS only and from 93% in Belgium-The Netherlands to 100% in Czech Republic for patients with DES. At 6 months, the percentage of patients taking dual antiplatelet therapy ranged from 45% in Czech Republic to 97% in Greece for BMS-only patients and from 83% in Belgium-The Netherlands to 97% in Greece for patients with DES. Aspirin use ranged from 89% to 100% across the countries and type of stent. CONCLUSIONS: Use of DES versus BMS varied among the European countries. In patients with BMS, there was marked variability with clopidogrel use at 6 months; whereas, in patients with DES, clopidogrel was used more frequently and with less variability. PCV132 AN INTERNATIONAL COMPARISON OF ANTIPLATELET USE AT 6 MONTHS FOLLOWING HOSPITAL DISCHARGE IN UA/NSTEMI AND STEMI PATIENTS UNDERGOING PCI: RESULTS FROM THE ANTIPLATELET TREATMENT OBSERVATIONAL REGISTRY II (APTOR-II) Coufal Z1, Berkenboom G2, Pavlides G3, Mohacsi A4, James S5, Paget MA6, Goedicke J7, Norrbacka K8, Zeymer U9 1 Bata’s Regional Hospital, Zlin, Czech Republic; 2U.L.B Erasme Hospital Brussels, Brussels, Belgium; 3Onassis Cardiac Surgery Center, Kallithca, Greece; 4Hungarian Institute of Cardiology, Budapest, Hungary; 5Uppsala University Hospital, Uppsala, Sweden; 6Lilly France S.A., Suresnes Cedex, France; 7Eli Lilly & Company, Hamburg, Germany; 8Eli Lilly & Company, Vantaa, Finland; 9Herzzentrum Ludwigshafen, Ludwigshafen, Germany OBJECTIVES: Current European Society of Cardiology Guidelines recommend dual antiplatelet therapy for 12 months for all patients with acute coronary syndrome (ACS). The variations in antiplatelet treatment patterns at 6 months following hospital discharge were assessed in patients with unstable angina (UA), non-ST-elevation myocardial infarction (NSTEMI), and ST-elevation MI (STEMI) undergoing percutaneous intervention (PCI). METHODS: A prospective, observational registry of 11 European countries grouped into 6 clusters recruited acute coronary syndrome (ACS) patients undergoing PCI from 2008–2009. Interventional cardiologists collected data from ACS events up to hospital discharge, and primary care physicians and cardiologists collected 6-month data. RESULTS: Of 3042 eligible patients at baseline, 2964 patients (97%) had 6 month data; 51% of patients presented with STEMI and 49% with UA/NSTEMI. The median age was 62 yrs. At 6 months following hospital discharge, the percentage of UA/NSTEMI and STEMI patients using clopidogrel, respectively, was 70% and 40% in the Czech Republic, 84% and 89% in Germany, 98% and 99% in Greece, 78% and 81% in the Nordic countries, 93% and 94% in Austria and Hungary, and 81% and 88% in Belgium and the The Netherlands. The percentage of UA/NSTEMI and STEMI patients taking aspirin, respectively, was 98% and 99% in the Czech Republic, 89% and 90% in Germany, 97% and 100% in Greece, 95% and 93% in the Nordic countries, 94% and 96% in Austria and Hungary, and 91% and 94% in Belgium and the The Netherlands. CONCLUSIONS: Use of clopidogrel varied considerably between European countries at 6 months after discharge, whereas the majority (≥90%) of patients continued to use aspirin. Results suggest that both physician and patient education on continuing dual antiplatelet therapy may be needed. PCV133 PRACTICE PATTERNS AND QUALITY OF LIFE IN ACUTE CORONARY SYNDROME PATIENTS IN 2008–2009: BASELINE RESULTS FOR GERMANY FROM THE ANTIPLATELET TREATMENT OBSERVATIONAL REGISTRY II (APTOR II) Zeymer U1, Norrbacka K2, Paget MA3, Goedicke J4, Clouth J5 1 Herzzentrum Ludwigshafen, Ludwigshafen, Germany; 2Eli Lilly & Company, Vantaa, Finland; 3 Lilly France S.A., Suresnes Cedex, France; 4Eli Lilly & Company, Hamburg, Germany; 5Lilly Deutschland GmbH, Bad Homburg, Hessen, Germany OBJECTIVES: This analysis aims to explore management of acute coronary syndromes (ACS) from acute event to hospital discharge in Germany, and to measure Quality of Life (QoL) at discharge. METHODS: This 12-month international prospective, observational study recruited ACS patients undergoing percutaneous coronary intervention (PCI), April 2008–March 2009, capturing practice patterns, resource use and QoL. RESULTS: 500 ACS-PCI patients (out of the 508 recruited) were eligible: median age 63 yrs (IQR 53–72), median weight 82 kg (IQR 74–92), 23% female, 24% Type II diabetics, and 27% prior myocardial infarction (MI). Index diagnosis was: unstable angina or non-ST-elevation MI (UA/NSTEMI)-53% and ST-elevation MI (STEMI)-47%. Almost all patients (96%) received stents: 68% bare metal stents (BMS) only, 29% drug eluting stents (DES) only and 3% both. Time from start of ACS symptoms to PCI was 3 days or less in 91% of UA/NSTEMI patients and 1 day or less in 97% of STEMI patients. Oral antiplatelet medications with loading dose (LD) used: aspirin-86% and clopidogrel-93%. Clopidogrel LD was administered in 13th Euro Abstracts the ambulance-7%, previous hospital-8%, emergency room-32%, CCU or ICU-23%, catheterization lab-25%, other ward 5%. LD was administered just before or direct after PCI in 79% of cases. The clopidogrel LD was 600 mg in 68% and 300 mg in 24% of cases and in-hospital maintenance dose (MD) was 75 mg in 98%. At time of hospital discharge, 92% of discharged patients were receiving clopidogrel (discharge dose 75 mg in all cases). QoL in discharged patients was good: median EQ-5D health state index at 0.85 (IQR 0.73–1.00). CONCLUSIONS: These real life data reflect treatment patterns among ACS patients managed by PCI in Germany in 2008–2009. Timing and place of antiplatelet loading differ between patients. The QoL of patients at discharge was high. PCV134 THERAPEUTIC VERSUS ACTUAL MARKET SHARES FOR STATINS Gorevski E, Bian B, Kelton C, Boone J, Guo JJ University of Cincinnati, Cincinnati, OH, USA OBJECTIVES: This study estimated market shares for six statins based on 1) the literature-determined lipid-lowering potential of the drugs, and 2) the distribution of U.S. adults depending on the amount of lipid-lowering they required. Estimated market shares were compared with actual market shares. METHODS: Two panels (1999–2000 and 2001–2002) from the National Health and Nutrition Examination Survey (NHANES) provided interview, demographic, and laboratory data for 3324 people >20 years of age who were not taking a statin. In accordance with the ATP III Guidelines, coronary heart disease (CHD) and CHD-risk-equivalent events including diabetes, other risk factors such as smoking and high blood pressure, and Framingham scores to obtain 10-year risk estimates were identified and calculated, respectively for each person, along with percentage LDL-C lowering required to reach his or her LDL-C goal. Depending on the amount of lowering needed and on whether the individual had a liver condition, hence, enhanced risk of rhabdomyolysis, a particular statin or particular statins with equal probabilities were assigned to the individual; e.g., a person without a liver condition requiring 50% lowering of LDL-C would be assigned rosuvastatin or atorvastatin, each with probability = 0.5. RESULTS: Of 1643 men surveyed, 57.88% were in the lowest-risk group (0–1 risk factors); 14.36% were in the highest-risk group (CHD or CHD-risk-equivalent). For the 1681 women, these percentages were 73.11% and 11.30%, respectively. Of people needing a statin, 73.74% needed LDL-C lowering by <25%; only 1.16% required LDL-C lowering of >60%. Therapeutic market shares were estimated at 20.11% (rosuvastatin); 18.94% (atorvastatin); 17.89% (simvastatin); 16.96% (lovastatin); 13.05% (fluvastatin); and 13.05% (pravastatin). CONCLUSIONS: Actual market shares are significantly different from market shares based on lipid-lowering considerations. Explanations include direct-to-consumer advertising; varying pricing strategies; and clinical trials for additional indications and subpopulations conducted only by the market leaders. PCV135 CREATING AN EFFICIENT HOSPITAL PAYMENT SYSTEM 1 2 Baser O , Gust C 1 STATinMED Research/University of Michigan, Ann Arbor, MI, USA; 2STATinMED Research, Ann Arbor, MI, USA OBJECTIVES: By creating incentives around quality and efficiency, there exists consideration to bundle hospital and physician payments around the episodes of inpatient surgery. We provided current payments around surgical episodes including the degree to which different types of providers are involved. In particular, we looked at the most expensive procedure performed commonly on the elderly: coronary artery bypass graft (CABG) surgery. METHODS: The study was based on complete national U.S. claims data. Managed care patients were excluded (16% in 2005) from the study because service provided to them was not consistently captured in the data set. We also excluded patients who were less than 65 years of age or over 99, and those not enrolled in the data set at the time of their procedures (4%). Also excluded were patients who were nursing home residents before surgery. Patients undergoing CABG were identified using ICD-9 codes. Price-standardized payments from the date of admission for the index procedure to 30 days post-discharge were determined and categorized by payment type (hospital, physician, and post-acute care) and sub-type. RESULTS: The average total payment around an inpatient surgery episode was $45,358 for CABG. Hospital payments accounted for the largest of total payments (60% of the procedure), followed by physician payments (13%). Diagnosis-Related Group (DRG) payments and surgeon and anesthesia professional payments together accounted for 65% of total payments. Among payment types potentially leveraged by bundled payments, 30-day readmissions accounted for 10% of total payments around surgical episodes. Post-acute care, including home health care and extended care facilities, accounted for 7% of total payments. CONCLUSIONS: Payments for potentially mutable services, including outlier payments, 30-day readmissions, and post-acute care are considerable and might be reduced by incentives for hospitals and physicians to improve quality and efficiency. PCV136 ATRIAL FIBRILLATION MANAGEMENT PATTERNS IN GREECE Karampli E, Athanasakis K, Ollandezos M, Kyriopoulos J National School of Public Health, Athens, Greece OBJECTIVES: Construct the disease management model for patients with Atrial Fibrillation (AF) in Greece, according to initial therapeutic approach. METHODS: The analysis was based on data from patient records of 149 geographically distributed cardiologists, extracted via strictly structured questionnaire-based interviews. Partici- 13th Euro Abstracts pants were requested to categorize patients according to initial AF approach, i.e. pharmacological or non-pharmacological (interventional) treatment option. Discrete sets of answers were provided for each patient group. RESULTS: A pharmacological therapy was the initial treatment approach for 89.6% of patients (rhythm control: 59.5%, rate control: 40.5%), whereas 5.2% had initially undergone non-pharmacological therapy, mainly catheter ablation (96% of cases). The remaining 5.2% received no treatment. Patients on rhythm control were mainly prescribed a Class-III (28.9%), Class-Ic (28.9%) or Class-II anti-arrhythmic agent, whereas 13.4% received a combination. In the rate control group, the treatments of choice were b-blockers (37.0%) and digoxin (24.2%). The majority of patients, also, received anti-thrombotic medication (89.0%, 91.3% and 100% for rhythm control, rate control and non-pharmacological treatment, respectively). The most frequently performed diagnostic procedures across groups, on an annual basis, were INR monitoring (8.3–11.5 tests/year), echocardiogram (4.0–4.4), liver (1.8–2.0) and renal function tests (1.7–1.9) and Holter monitoring (1.4–1.6). On average 4.3, 4.0 and 3.1 consultations/year with a cardiologist are necessary for the monitoring of patients on rhythm, rate control and nonpharmacological treatment respectively. Annual hospitalization rates for patients under rhythm control were 20% due to AF recurrence, 13% due to cardiovascular complications and 5.63% due to drug adverse reactions. For patients under rate control, the corresponding percentages were 6%, 6.13% and 4% and for those under non-pharmacological therapy 18.75%, 3% and 4.86% respectively. CONCLUSIONS: Pharmacological rhythm control appears to be the prevailing initial AF treatment option in Greece. AF management requires an intense monitoring pattern, given that it can lead to increased frequency of hospitalizations. PCV137 A SYSTEMATIC REVIEW ON DISEASE BURDEN AND UNMET NEEDS FOR VENOUS THROMBOEMBOLISM IN HOSPITALISED MEDICAL PATIENTS IN EUROPE SHOWS UNDER-UTILIZATION OF PREVENTIVE THERAPIES Khoury H1, Welner S1, Kubin M2, Folkerts K2, Haas S3 1 BioMedCom Consultants inc., Dorval, QC, Canada; 2Bayer Schering Pharma AG, Wuppertal, Germany; 3Technical University Munich, Munich, Germany OBJECTIVES: To assess the incidence of venous thromboembolism (VTE) and current practice patterns for VTE prophylaxis in hospitalised acutely ill medical patients in Europe. METHODS: A systematic literature search was conducted in major databases on the epidemiology and treatment practices of VTE prevention among adult patients treated in hospital for major medical conditions. Relevant studies published between 1999 and April 2010 were captured. RESULTS: Thirty-five multinational and country-specific studies were retrieved, including randomized clinical trials, registry and chart audits, meta-analyses, and cross-sectional, retrospective, prospective, and observational studies. Among patients admitted for an acute medical illness, the incidence of VTE diagnosed during hospitalization ranged from 3.01% (mean hospital stay: 4.5 days) to 15% (day 14 from index hospitalization). While clinical guidelines recommend pharmacological VTE prophylaxis to all patients hospitalised for an acute medical illness who are bedridden, a clear identification of specific risk groups who would benefit from VTE prevention is lacking. In the majority of studies captured, prophylaxis was underused among medical inpatients (all diseases); 33% to 82% of all patients hospitalised for acute medical illnesses did not receive any VTE prophylaxis. Furthermore, among patients who did receive prophylaxis, a considerable proportion received medication that was not in agreement with guidelines, due to short duration, suboptimal dose, or inappropriate type of prophylaxis. Of all prescriptions, low molecular weight heparin was the most widely prescribed anticoagulant. In most cases, the duration of VTE prophylaxis did not exceed hospital stay duration, and varied between 5 and 33.8 days (mean duration). CONCLUSIONS: VTE imposes a substantial burden among hospitalised medical patients. Despite the proven efficacy of prophylaxis, utilization remains suboptimal among medical patients at risk for VTE, stressing the necessity for improved or easier access to proven preventive therapies among these patients. CARDIOVASCULAR DISORDERS – Conceptual Papers & Research on Methods PCV138 MEASUREMENT PROPERTIES OF PEAK VO2 IN CHILDREN WITH PULMONARY ARTERIAL HYPERTENSION Cappelleri JC1, Hwang LJ2, Mardekian J2, Mychaskiw MA2 1 Pfizer Inc, New London, CT, USA; 2Pfizer Inc, New York, NY, USA OBJECTIVES: Although the 6-minute walk test (sub-maximal exercise test) historically has been used to evaluate the effect of pharmacologic intervention in adults with pulmonary arterial hypertension (PAH), it was not widely regarded as appropriate for children at all ages. Thus, Peak VO2 (maximal exercise test) was used in a pediatric PAH trial (A1481131) to evaluate the effect of treatment. As this is the first large controlled trial using the Peak VO2 endpoint in this population, we investigated its performance in terms of correlational analyses. METHODS: Relationships between change in Peak VO2 and other endpoints were evaluated using correlation coefficients and regression analyses with the data from the 16-week randomized, placebo-controlled, clinical trial of sildenafil assessing Peak VO2 at both baseline and week 16 in 106 evaluable pediatric patients. a Bland-Altman plot was used to assess the reliability of screen and baseline visit data. RESULTS: The intraclass correlation was 0.79 for A367 the screening and baseline visit Peak VO2 data. Additionally, agreement between these visits was supported by a Bland-Altman plot. Correlations of percentage changes in Peak VO2 from baseline with the change from baseline in the Physician Global Assessment (overall and for sildenafil doses) and in World Health Organization Functional Class (WHO FC, with baseline of I and III/IV) correlated well (correlations of ≥0.40) and were responsive to change, while the low correlation with the Subject Global Assessment (0.12) suggested influence by factors associated with child and parental proxy response and instrument administration. Percentage changes in Peak VO2 gave no real association (0.04) with the Family Cohesion of the Child Health Questionnaire. CONCLUSIONS: In pediatric PAH, Peak VO2 measurements exhibited good reliability. Improvements in Peak VO2 were shown to be associated with improvements in qualitative clinical endpoints, including WHO FC and physician global assessment. PCV139 INTEGRATING PATIENT PREFERENCES IN EFFICIENCY FRONTIER ANALYSES USING THE ANALYTIC HIERARCHY PROCESS Hummel JM, Steuten L, Van Til J, Ijzerman MJ University Twente, Enschede, The Netherlands OBJECTIVES: In comparative effectiveness research and economic evaluations, benefits of technologies are measured using multiple outcomes measures. Information lacks however about the importance of these endpoints for patients. We propose a new methodology to integrate patient weighted outcomes in a cost-efficiency frontier. We illustrate this methodology by means of an efficiency frontier analysis of five alternative treatments of patients with equinovarus deformity poststroke. METHODS: The Analytic Hierarchy Process (AHP) is a technique for multi-criteria analysis. The AHP supported 140 patients to prioritize the outcome measures of treatments of equinovarus deformity poststroke, and 10 professionals to prioritize the treatments regarding the outcome measures. These outcome measures include functional outcomes, risk and side effects, comfort, daily effort, cosmetics, and impact of the treatment. Sensitivity analysis is based on bootstrapping of the participants’ priorities. Relative costs include the device related costs and the care related costs of the treatments. RESULTS: The overall effectiveness of soft-tissue surgery (.41) is ranked first, followed by orthopedic footwear (.18), ankle-foot orthosis (.15), surface electrostimulation (.14), and finally implanted electrostimulation (.12). Implanted electrostimulation (.35) and soft-tissue surgery (.34) are considered to be most expensive, followed by surface electrostimulation (.26), orthopedic footwear (.03) and ankle-foot orthosis (.02). Based on these priorities of the treatments’ overall effectiveness and costs, an efficiency frontier was drawn that includes decision uncertainty. CONCLUSIONS: The results suggest that the cost-effectiveness of implanted electrostimulation and surface electrostimulation are unfavourable. This new methodology for efficiency frontier analysis allows decision makers to integrate the outcomes about the diverse values and costs of health care technology, and can be applied broadly. It is particularly suitable in the field of early technology assessment, since the AHP supports a systematic estimation of priors about the effectiveness of alternative treatments. PCV140 HOSPITAL QUALITY INDEX AND ITS USE ON RATES OF VENOUS THROMBOEMBOLISM Baser O1, Akin C2, Wang L3, Dysinger A3 1 STATinMED Research/University of Michigan, Ann Arbor, MI, USA; 2STATinMED Research/ Brigham and Women’s Hospital, Ann Arbor, MI, USA; 3STATinMED Research, Ann Arbor, MI, USA OBJECTIVES: We used the composite quality index (CompQualTM) to examine the association between venous thromboembolism (VTE) events and hospital quality for patients who underwent major orthopedic surgery. METHODS: Using the empirical Bayes approach to combine mortality rates with information on hospital volume at each hospital, we created an index to weight observed mortality according to how reliable it is estimated, with remaining weight placed on hospital volume (CompQualTM scoring algorithm). Then, using a national database, all patients who underwent major orthopedic surgery were identified. Rates for patients who had a VTE event during their initial hospitalization were calculated. By using the algorithm, every provider in the Medicare data set was ranked according to their quality. Multivariate regression was used to see the effect of quality on VTE event rates. RESULTS: We obtained a sample that included 2,745 patients in the VTE group. After controlling for patients’ demographic and clinical factors, VTE events in low quality hospitals were almost 2.5 times higher than VTE events in high quality hospitals (p = 0.000). We also compared the ranking with only volume and only for hospitals. Our combined measure to rank the hospital was more effective and explained more variation than the individual measures (pseudo R-square: 0.46 for composite score, 0.15 for volume only, 0.09 for mortality only). CONCLUSIONS: Using national Medicare data for patients with VTE, we found that a simple composite measure was a strong predictor of subsequent performance for operations. In this regard, it was more effective than individual measures. Such measures would be useful to help patients and payers identify low quality hospitals for major surgery. Any policy implementation that would decrease the variation in hospital quality would have a direct effect on the rates of VTE events. A368 PCV141 A NOVEL METHOD FOR ESTIMATING THE EFFECTS OF ADHERENCE TO ANTIHYPERTENSIVES AND DURATION OF DRUG ACTION ON CARDIOVASCULAR RISK Lowy A, Brede Y Novartis Pharma AG, Basel, Switzerland Most patients miss occasional doses of antihypertensive treatment. The consequent loss of blood pressure (BP) control, and thus increased cardiovascular risk, may be mitigated by prescribing “forgiving” drugs that have a duration of antihypertensive effect beyond the 24-hour dosing interval. However, the consequences of missed doses for cardiovascular outcomes, and additional impact of duration of drug action, have not been evaluated quantitatively. We developed a method to quantify the chain of factors linking adherence and duration of drug action with outcomes; an important consideration given that adherence and hence BP reduction in real-world clinical practice are typically much lower than in randomised controlled trials. The method involves simulating 256-day dosing histories for 1250 individuals, incorporating realistically distributed gaps in dosing based on a study of electronically monitored dosing records from 4783 hypertensive patients in clinical trials. Percentage adherence is adjusted by altering the proportion of doses missed while preserving a realistic distribution of gap length. Systolic BP (SBP) reduction for each patient is estimated, with rises and falls according to the individual’s dosing behaviour. By averaging SBP reductions over time and individuals, population mean SBP reduction is estimated. Cardiovascular disease (CVD) risk over, for example, 10 years may then be predicted using the Framingham Risk Equation, with baseline characteristics from NHANES 2005– 2006 (average data: age 65.2 years; 52.9% women; SBP 155.95 mmHg; total cholesterol 5.41 mmol/L) and baseline absolute 10-year CVD risk of 27.0%. The method allows variable inputs including adherence level (%), rate of loss of antihypertensive effect when treatment is interrupted (mmHg/day), rate of BP reduction when treatment is (re)started (mmHg/day) and SBP-lowering effect of uninterrupted treatment (mmHg). In conclusion, we describe a novel method that allows quantification of the effects of treatment adherence and antihypertensive drug forgiveness on clinical outcomes (i.e. SBP and CVD risk reduction). PCV142 AN APPLICATION OF IMPUTATION TECHNIQUES TO IMPROVE DATA AVAILABILITY FROM ELECTRONIC MEDICAL RECORDS Exuzides A, Colby C ICON Clinical Research, San Francisco, CA, USA OBJECTIVES: Analytic data sets based on electronic Medical Records (eMRs) offer long and detailed patient follow-up, large sample sizes, and a rich set of variables. However, the existence of a particular variable within an eMR is no assurance of actual data availability for all patients. Patients may receive lab tests irregularly or miss visits. Reporting standards can vary within an eMR system, creating large numbers of patients with missing data. We present details from a three step imputation process to improve eMR data availability from an observational study of hypertension outcomes. METHODS: We used an analytic data set comprised of 227,257 patients from 10 participating medical centers. One important element of the study was the computation of a cardiovascular risk score based on 11 demographic and clinical variables available from eMRs. However, only 43,676 patients (19%) had complete data for all 11 variables. To increase the sample size of patients with complete data for all 11 variables, we used three imputation techniques: we first allowed for lab and blood pressure values to be carried forward to replace missing values; we then augmented a missing diabetes status with pharmacy data by using diabetes-specific medications to indicate a positive diabetes history; and, finally, we applied a multiple imputation (MI) technique, based on a Markov Chain Monte Carlo (MCMC) approach, implemented using SAS® PROC MI. RESULTS: After application of the three data imputation techniques, the sample size of patients with complete or imputed data for all 11 variables increased to 75,209 (33%). This amounts to a substantial increase (72%) in the number of patients with all critical variables available to be used for further analysis. CONCLUSIONS: Proper application of imputation techniques, including MI, can yield a substantial increase in the number of patients available for analysis. GASTROINTESTINAL DISORDERS – Clinical Outcomes Studies PGI1 A SYSTEMATIC REVIEW OF THE DIAGNOSTIC ACCURACY OF SEROLOGIC TESTS IN THE DIAGNOSIS OF CELIAC DISEASE Costa V1, Chandra KM2, McCurdy B1, Ieraci L3, Levin L1 1 Ontario Ministry of Health and Long-term Care, Toronto, ON, Canada; 2McMaster University, Hamilton, ON, Canada; 3Toronto Health Economics and Technology Assessment (THETA) Collaborative, Toronto, ON, Canada OBJECTIVES: To evaluate the diagnostic accuracy of IgA (immunoglobulin A) and IgG serologic tests in the diagnosis of celiac disease including anti-tissue transglutaminase (tTG) antibody, anti-deamidated gliadin peptide antibody (DGP), anti-endomysial antibody (EMA), and anti-gliadin antibodies (AGA). METHODS: A systematic review of the peer-reviewed literature was conducted to identify studies that evaluated the sensitivity and specificity of serologic celiac disease tests using small bowel biopsy as the gold standard. The patient population consisted of untreated, undiagnosed subjects with symptoms consistent with the disease. Pooled estimates of sensitivity and 13th Euro Abstracts specificity were calculated using a bivariate, binomial generalized linear mixed model (SAS 9.2) which takes into account the negative correlation between sensitivity and specificity. Statistical significance was defined by p-values less than 0.05, where “false discovery rate” adjustments were made for multiple hypothesis testing. RESULTS: In total, 17 studies were eligible for the analysis. The pooled analysis showed that IgA tTG has a sensitivity of 92.1% (95% CI 88.0, 96.3), compared to 89.2% (83.3, 95.1, p = 0.12) for IgA DGP, 85.1% (79.5, 94.4, p = 0.07) for IgA EMA, 74.9% (63.6, 86.2, p = 0.0003) for IgA AGA. Combining more than one serologic test slightly increased the sensitivity, 95.1% (92.2, 98%, p = 0.039 vs. IgA tTG). The pooled sensitivity of IgG DGP was 88.4% (95% CI: 82.1, 94.6), 44.7% (30.3, 59.2, p = 0.0003) for IgG tTG, and 69.1% (56.0, 82.2, p = 0.34) for IgG AGA. Specificity ranged from 90.1% to 93.9% and was similar among the different tests. The quality of the evidence according to the GRADE Working Group criteria was considered “moderate” for IgA tTG and EMA, “low” for IgA DGP and test combinations, and “very low” for IgA AGA. CONCLUSIONS: The evidence available suggests that IgA tTG has a higher accuracy than other serologic tests. Combining more than one serologic test contributes little to the diagnostic accuracy. PGI2 IS THERE ANY DIFFERENCE BETWEEN THE EFFECTS OF THERAPY WITH PEGINTERFERON-ALPHA-2A VERSUS STANDARD-DOSE PEGINTERFERON-ALPHA-2B? A META-ANALYSIS COMPARING BOTH TREATMENTS PLUS RIBAVIRIN IN GENOTYPE 1/4 CHRONIC HEPATITIS C VIRUS (HCV) INFECTION PATIENTS Barros FMR1, Cheinquer H2, Borges LG3, Santos E3 1 Hospital Português de Beneficência em Pernambuco and Hospital das Clínicas—UFPE, Recife, Brazil; 2Hospital das Clínicas da Universidade Federal do Rio Grande do Sul, Porto Alegre, Brazil; 3Roche Brazil, São Paulo, Brazil OBJECTIVES: Until recently, few head-to-head trials evaluated the efficacy of peginterferon-alpha-2a versus peginterferon-alpha-2b and none meta-analysis was available so far. In 2010, a meta-analysis performed by Awad et al was able to demonstrate that peginterferon-alpha-2a was more efficacious in achieving sustained virological response as compared to peginterferon-alpha-2b. McHutchison 2009, one of the studies included in the meta-analysis, evaluated two different doses of peginterferonalpha-2b (standard-dose: 1,5 mcg/kg; low-dose: 1,0 mcg/kg), raising doubts about possible negative impact of lower dose on the analysis results.We aim to compare treatment outcomes in genotypes 1/4 HCV infection treated with peginterferon-alpha2a versus peginterferon-alpha-2b at standard-dose (per package insert) both plus ribavirin in order to assess efficacy, performing a systematic review and meta-analysis of published randomized controlled trials (RCTs). METHODS: An extensive search through main medical information databases (MEDLINE, Cochrane Library, Embase and Lilacs) and specialized websites was conducted on the second trimester of 2010. We aim to indentify RCTs which evaluate treatment efficacy of peginterferon-alpha-2a versus standard-dose peginterferon-alpha-2b both plus ribavirin for HCV treatment in genotypes 1/4 patients not co-infected with HIV. According to heterogeneity test, a fixed or a random-effect model was adopted. RESULTS: Only 6 papers out of 623 found (databases citation without duplicates plus manual searches) met our inclusion criteria: 1)Ascione, 2009; 2)Rumi, 2009; 3)McHutchison, 2009; 4)Yenice, 2006; 5) Scotto, 2008; 6)Berak, 2007. For genotypes 1/4 chronic HCV infection patients, peginterferon-alpha-2a showed higher SVR as compared to standard dose of peginterferon-alpha-2b: 42.09% versus 33.44% (RR = 1.10, IC 95% 1.006–1.202, assuming a fixed-effect framework). CONCLUSIONS: According to these results, no significant impact on already published meta-analysis results was observed when lower dose peginterferon-alpha-2b patients were excluded from the analysis. In summary, peginterferon-alpha-2a is more effective in treating genotypes 1/4 HCV infected patients as compared to standard-dose peginterferon-alpha-2b, both in association with ribavirin. PGI3 EFFICACY OF ANTIBIOTICS AND PROBIOTICS IN MANAGEMENT OF POUCHITIS; A META-ANALYSIS Abdollahi M1, Nikfar S2, Darvish Damavandi M3 1 Faculty of Pharmacy; and Pharmaceutical Sciences Research Center, Tehran University of Medical Sciences, Tehran, Iran; 2Tehran University of Medical Sciences, Faculty of Pharmacy and Food & Drug Laboratory Research Center, Iran MOH, Tehran, Iran; 3Faculty of Science, National University of Singapore, Singapore OBJECTIVES: Pouchitis is the most frequent long-term complication of ileal pouchanal anastomosies (IPAA) surgery for ulcerative colitis (UC) which is a nonspecific inflammation of the ileal reservoir. Its clinical frequency varies depending on the definition and the follow up but is approximately 50% after a decade. Antibiotics and probiotics are currently the most widely accepted treatment in pouchitis patients. Objective of this study was to meta-analyze efficacy of probiotics and antibiotics in the management of pouchitis. METHODS: All databases specially Pubmed, Web of Science, Scopus, Cochrane, and Google Scholar were searched between 1965 and December 2009, and relevant controlled clinical trials were extracted, reviewed, and validated according to the study protocol. The outcome of interest was defined by a pouchitis disease activity index (PDAI) < 7. Thirteen clinical trials were included in the meta-analysis. RESULTS: Pooling of the results from eight trials yielded a relative risk (RR) of 5.33 with a 95% CI of 2.12–13.35 and a significant RR (P = 0.0004) in the treatment group in comparison with the placebo group. Summary RR for clinical improvement in 6 trials was 14.17 with a 95% CI of 1.19–168.93 (P = 0.036) in 13th Euro Abstracts efficacy of VSL#3 (all doses) comparing to placebo and slightly more effective for VSL#3 (6 g/day) comparing to placebo with RR of 20.35 with a 95% CI of 6.16– 67.22 (P < 0.0001). Efficacy of antibiotics comparing to placebo showed a summary RR of 2.68 with a 95% CI of 0.4–17.99 and P = 0.3107 for clinical improvement in three trials. The summary RR for efficacy of ciprofloxacin comparing to metronidazole was 0.68 with a 95% CI of 0.44–1.06 (P = 0.8913). CONCLUSIONS: In conclusion, alongside the benefit of probiotics and antibiotics in the management of pouchitis, effects of probiotics and antibiotics on pouchitis vary according to different mixtures of microorganisms strains in probiotics and different spectrums of antibiotics. PGI4 PREVALENCE AND INCIDENCE OF HEPATITIS A IN A UNITED STATES MANAGED CARE CLAIMS DATABASE Changolkar AK1, Eisenberg D2, Misurski DA1 1 GlaxoSmithKline, Philadelphia, PA, USA; 2HealthCore, Wilmington, DE, USA OBJECTIVES: To evaluate the prevalence and incidence of hepatitis a in a managed care population, segmented by a range of characteristics including age, gender, geography, and health insurance plan. METHODS: This was an observational retrospective cohort study utilizing medical and pharmacy claims data for January 1, 2005 through December 31, 2008 from the Impact National Managed Care Database (Benchmark Database). The index date was defined as the date of the first and only diagnostic claim for hepatitis a within the intake period. All prevalence results reported below are per 100,000 and incidence is presented as per person years. RESULTS: Out of a total of 94,985,124 subjects in the database, females and males were represented equally (51% and 49% respectively). The 45–64 year age group had the greatest percentage of subjects. The prevalence of hepatitis a among men and women were almost equal (9 and 8, respectively). The 45–54 years age group had the highest hepatitis a prevalence (4) followed by the 25–34, 35–44, and 55–64 age groups (3 for each group). The largest hepatitis a prevalence rate was from the northeast region (7) followed by the south (6). In the 2005–2008 timeframe, hepatitis a prevalence was similar in each year (4). Incidence rates across age, region and specific zip codes in person years were less than 1 person per year. The incidence was slightly higher in males. Among three digit zip codes, the zip codes “100” showed higher incidences each year with higher prevalence. CONCLUSIONS: In a large commercially insured US population, we observed similar prevalence rates of hepatitis a for males and females. The disease was most prevalent in the 45–54 years age group. The northeast had the highest prevalence rate from 2005–2008. GASTROINTESTINAL DISORDERS – Cost Studies PGI5 A BUDGET IMPACT ANALYSIS OF THE HEPATITIS C TREATMENT WITH PEGYLATED INTERFERON IN BRAZIL Fonseca M, Araújo GTBD AxiaBio, São Paulo, Brazil OBJECTIVES: Cost-effectiveness analyses of new hepatitis C drugs have been used to select and approve drugs that should be financed by the public health care system. However there are still a huge amount of uncertainties within these studies. The fact of having pharmacoeconomic and budget impact data of these new pharmacological alternatives will help to select the most efficient alternative. With this, the objective is to perform a budget impact analysis (BIA) of the treatment with pegylated interferon (pegIFN), alfa-2a or alfa-2b, plus ribavirin in, economically active genotype I patients, aging from 30 to 59 years, with chronic hepatitis C (CHC). METHODS: An interactive model has been designed from the inputs obtained from the medical literature. Both strategies have been considered as therapeutic equivalents, without significative difference in side effects, and as having the same price. RESULTS: The number of patients with CHC evaluated in the model has been of 355,611, 15.2% between 30 and 39 years, 13% between 40 and 49 years and 9% between 50 and 59 years, mimicking the Brazilian population and with an average weight of 70.6 kg. The duration of treatment was 48 weeks, with virologic response measured in weeks 2.4,12.24 and 48. a total of 91% and 99.8% of the patients receiving pegIFN alfa-2a or alfa-2b, respectively achieved sustained virological response at 72 weeks. Total cost was R$ 4,012,531,131.29 and R$3,752,238,973.99 for the treatment with pegIFN alfa-2a + RIB and pegIFN alfa-2b + RIB, respectively. CONCLUSIONS: Although cost-effectiveness analysis appears to be favorable to pegIFN alfa-2a + RIB, a BIA seems to favour pegIFN alfa-2b + RIB due to its better predictability in the 12nd week of treatment. In this case, the treatment with pegIFN alfa-2b + RIB (in comparison with pegIFN alfa-2a + RIB) is an efficient strategy. PGI6 COST-EFFECTIVENESS OF MR ELASTOGRAPHY FOR DIAGNOSING LIVER FIBROSIS: PRELIMINARY THRESHOLD ASSESSMENT Lee DW1, Palathinkara V1, Dekoven M2 1 GE Healthcare, Waukesha, WI, USA; 2IMS Health, Falls Church, VA, USA OBJECTIVES: MR Elastography (MRE) is a non-invasive test under clinical trial evaluation for effectiveness in helping diagnose liver fibrosis. This study estimated the accuracy needed for MRE to be cost-neutral from a US payer’s perspective. METHODS: We constructed a decision-analytic model comparing diagnostic costs under two scenarios for patients with suspected liver fibrosis: 1) biopsy, or 2) MRE followed by biopsy when the MRE test was positive. We conducted a targeted litera- A369 ture review and consulted with a leading hepatologist and pathologist to identify the appropriate procedure codes associated with liver biopsy; we assumed MRE would be reimbursed using CPT-4 code 74,181 (magnetic resonance (eg., proton) imaging, abdomen; without contrast material(s)). We assigned the appropriate allowable charges to the identified procedure codes using the 2010 Medicare Physician Fee Schedule. Finally, we assumed negative predictive values (NPV) of 0.8, 0.9 and 0.95, and that patients with a false-negative MRE ultimately received a biopsy. RESULTS: The cost of a liver biopsy was $1424 (ultrasound $164, surgical $881, pathology $347, laboratory $32) and the cost of an MRE (without contrast) was $946 (hospital setting) or $666 (non-hospital setting). In a hospital setting, MRE is potentially cost saving if the test-negative rate is greater than 83%, 74% and 70% for NPVs of 0.8, 0.9 and 0.95, respectively. In a non-hospital setting, MRE can reduce diagnostic costs when more than 58%, 52% and 49% or patients have negative MRE results for the corresponding NPV values. CONCLUSIONS: The cost of liver biopsy is substantial as compared to MRE. MRE offers the potential of reducing the cost of diagnosing liver fibrosis by avoiding unnecessary biopsies. Cost saving potential increases with MRE’s negative predictive value and negative test rate. PGI7 ECONOMIC BURDEN OF OPIOID INDUCED CONSTIPATION IN SPAIN 1 2 1 1 Guijarro P , Alonso-Babarro A , Viqueira A , Fernandez G 1 Pfizer Spain, Alcobendas, Madrid, Spain; 2La Paz Hospital, Madrid, Spain OBJECTIVES: To analyze the use of health care resources and the associated costs in patients with opioid induced constipation (OIC). METHODS: An observational, retrospective, multicenter study was carried out in Spanish National Health System hospitals. All patients were free of constipation at baseline and received opioids for at least 2 months. In order to determine patient resource utilization, a review of the patient records and patient-interviews were performed to all patients diagnosed with OIC. Patients were evaluated depending on response to oral treatment for OIC. The observation period was 2 months. The unit costs were obtained from Spanish databases (c 2009). RESULTS: A total of 744 patients were included. Patients had a mean (SD) age of 64.2 (13.4) and 48% were male. During the study period, 46.6% of the patients developed OIC. Most of these were treated first with oral laxatives exclusively (67.7%), 63.8% responded to the treatment and 36.2% did not. Of the total number of patients with OIC, 61.5% required a visit to health care services. Forty-four patients required a visit to the Emergency Room (ER) and 26 were hospitalized with a median length of stay of 3 days. Overall mean total cost (SD) of constipation management was c271.08 (c621.22). Resource utilization in responders to oral laxatives was statistically significant lower than in non-responders. a higher percentage of non-responders required health care visits, ER visits, and hospitalizations (P < 0.001). The mean cost (SD) for each responder was c115 (c230) compared to c442 (c810) for each non-responder (P < 0.001). CONCLUSIONS: OIC increases health care resource utilization particularly in patients with a poor response to oral laxatives. The economic burden of OIC was 3.8 times higher in non-responders than in responders. PGI8 HIDDEN COSTS OF ROTAVIRUS DISEASE—A RETROSPECTIVE MATCHED ANALYSIS OF HOSPITAL EPISODE STATISTICS (HES) DATA Morgan C1, Adlard NE2, Carroll SM2 1 Cardiff Research Consortium Ltd, CARDIFF, UK; 2Sanofi Pasteur MSD, Maidenhead, UK OBJECTIVES: To determine the impact of rotavirus (RV) disease as a secondary diagnosis on hospital length of stay (LOS) and cost for children under five years of age in England. METHODS: The study was based on Hospital Episodes Statistics for England 2001–2007 using a matched cohort design. Admissions with a secondary diagnosis of RV were extracted and matched by age, year of admission and, where relevant, therapeutic procedure, to other admissions from HES with the same primary diagnosis. Primary admissions were classified as potentially related or unrelated to RV. LOS was compared using the paired Wilcoxon test. RESULTS: Of 2126 admissions with a secondary diagnosis of RV, 707 (33.3%) had no other secondary diagnoses. Of these, 385 admissions were matched for primary diagnosis to controls. Median LOS for admissions was 2.0 days (Inter-Quartile Range [IQR] 1.0–4.0), mean (sd) 3.2 (4.1) compared with median 1.0 (IQR 0.0–1.5), mean 1.4 (sd 3.5) for control admissions (p-value < 0.001). Median cost/admission was £771 (IQR £604–£1090), mean cost was £1105 (sd £1064) compared with median £650 (IQR £564–£1090), £902 (sd 1004) (p-value < 0.001). a total of254/385 admissions were defined as unrelated to RV. For these admissions, median LOS was 2.0 days (IQR 1.0–4.25), mean 3.7 (sd 4.5) compared with median 1.0 (IQR 0.0–2.0), mean 1.6 (sd 4.0) for control admissions (p-value < 0.001). Median cost/admission was £918 (IQR £564– £1090), mean £1132, (sd £1132) compared with median £771 (IQR £564–£1090), mean £980 (sd £1187), (p-value < 0.001). CONCLUSIONS: As a secondary diagnosis in HES, RV has a significant incremental impact on LOS and cost. Variations in testing and reporting of RV mean that HES data may significantly under represent the overall burden of RV disease. This implies that standard modelling methodologies may underestimate the true burden of RV disease in the context of health care acquired infection. A370 PGI9 HOSPITAL COST FOR TREATMENT OF PATIENTS WITH PEPTIC ULCER BLEED (PUB) IN SWEDEN—DATA FROM THE KPP (COST PER PATIENT) DATABASE Sörngård H1, Sverdén Eklund E2, Söderlund C2, Marschall HU3 1 AstraZeneca AB, Södertälje, Sweden; 2Södersjukhuset, Stockholm, Sweden; 3Karolinska Unviersitetssjukhuset, Stockholm, Sweden OBJECTIVES: Treatment with esomeprazole or placebo was investigated in patients with PUB (NCT00251979). In all, 102 of 764 patients from 16 countries in Europe, Asia and Africa were Swedish. Following successful endoscopic hemostasis, patients were randomized to 72 hours intravenous esomeprazole or placebo with subsequent oral esomeprazole 40 mg for 27 days. Rebleeding was the primary variable and occurred in 7.2% and 12.9% in the esomeprazole and placebo group, respectively. The objective of this study was to describe hospital costs in Sweden for patients with and without rebleeding. METHODS: The KPP database includes 60% of all Swedish episodes of somatic in-hospital care and reports total hospital costs. Six out of the 12 participating Swedish hospitals utilize KPP. Here, individual cost data was collected and matched with the in-hospital episode and the study definition of rebleeding. RESULTS: Data was collected from all 60 patients in the six KPP reporting clinics. Information structure and level of detail varied considerably between clinics, why only total costs are reported. Six patients (10%) were defined as rebleeders, and accounted for 20% of the total costs for all patients. The mean total cost/patient was SEK 39,822 (median 29,082, range: 10,377–263,520) for all patients without rebleeding and SEK 88,928 (median 82,273, range: 52,464–160,090) for all patients with rebleeding (c1 = SEK9.84). The mean cost per patient without rebleeding varied between clinics from SEK 14,791 to 45,636. CONCLUSIONS: Hospital costs for patients with PUB vary considerably. The cost of patients with rebleeding was more than double that of patients without rebleeding, and in all the 10% of patients who rebled accounted for 20% of the total hospital costs. In conclusion, a treatment which successfully prevents rebleeding is important not only from a clinical perspective, but also from a cost point of view to reduce hospital costs. PGI10 MODELLING THE IMPACT OF TREATMENT WITH ENTECAVIR ON HEALTH CARE COSTS OF CHRONIC HEPATITIS B IN FRANCE Zambrowski JJ1, Ratziu V2, Bourlière M3, Pol S4, Zarski JP5, Woolmore A6, Bregman B7, Sahraoui S7, Bronowicki JP8 1 Service de Droit & Economie de la santé—Faculté des sciences pharmaceutiques et biologiques—Université Paris Descartes Paris, France, 2Service Hépato-Gastro-Entérologie, Hôpital de la Pitié-Salpêtrière, Paris, France, 3Hôpital St-Joseph, Marseille, France, 4Hôpital Cochin, Paris, France, 5Clinique Universitaire d’Hépato-Gastroentérologie—CHU de Grenoble, Grenoble, France, 6Monitor Group France, Pairs, France, 7Bristol-Myers Squibb, Rueil-Malmaison, France, 8Service Hépato-Gastro-Entérologie, Centre Universitaire de Nancy-Brabois, Vandoeuvre-les-Nancy, France OBJECTIVES: Chronic hepatitis B (CHB) treatment necessitate, according to European guidelines, to use potent antiviral agents with optimal resistance profiles. Increased health care financial burden means physicians, payers and decision makers need to evaluate CHB treatment cost-effectiveness. This model aims to estimate the medical cost savings of treating nucleoside-naïve CHB patients with a potent antiviral agent, from a French payer’s perspective. METHODS: CHB was simulated using a disease-state transition model with states defined as mild fibrosis (Ishak F0/F1), significant fibrosis (F2–F4), advanced fibrosis/cirrhosis (>F4) and complicated states (decompensated cirrhosis (DC), hepatocellular carcinoma (HCC), liver transplant and death) based on available natural history data. The model assumed a 5-year entecavir treatment and 30-year follow-up and was based on available clinical data. The transition probabilities between states increased with detectable viral load levels and varied by HBeAg status. Direct medical costs included CHB and liver complications management. The primary model output is the estimated cost avoided per patient per day of treatment, compared to no treatment in nucleoside-naïve CHB patients. RESULTS: Progression to HCC, liver transplant or death was estimated at 76% for untreated patients compared to 31% for entecavir patients, while the progression to DC, HCC, liver transplant or post-liver transplant resulted in annual costs/patient of c9,718 [95% confidence interval (CI): 8,260; 11,175], c5,066 [4,306; 5,826], c87,105 [74,039; 100,171] and c19,421 [16,508; 22,335], for 2008, respectively. Cost of not treating CHB patients was estimated at c16.4/day (average over patient lifetime). Entecavir treatment translated into specific patient benefit with an estimated cost saving of c1.2/day of entecavir treatment (95% CI: −9.6; 4.6). CONCLUSIONS: Treatment of CHB using a potent antiviral agent with high genetic barrier to resistance, such as entecavir, is costeffective as associated with improved clinical outcomes and lower health care costs compared with no treatment. PGI11 RISK FACTORS AND COSTS ASSOCIATED TO NSAIDS THERAPIES IN ITALY: AN ADMINISTRATIVE DATABASE ANALYSIS Berti A1, Rossi E1, Sacripanti C2, Poluzzi E2, De Rosa M1 1 Cineca, Casalecchio di Reno (BO), Italy; 2Università di Bologna, Bologna, Italy OBJECTIVES: Describe treatment patterns of NSAIDs, risk factors for gastrointestinal injuries and associated costs. METHODS: Prescriptions of reimbursed NSAIDs in 2007 were selected from ARNO database and analyzed for 12 months in 23 Italian Local Health Units. Intensity of NSAID use and co-medications were analyzed. Cluster analysis and risk stratification were performed to assess the influence of risk factor in NSAID prescriptions and gastroprotective (GP) co-medications. Risk factor analysis 13th Euro Abstracts was based on prescribing limitative AIFA (Agenzia Italiana del FArmaco) note 1 for GP: presence of chronic NSAID treatment, age ≥60 years, concomitant antithrombotics, cortisonics and/or previous gastric events. RESULTS: Among 1,028,100 new NSAIDs recipients (76% of overall treated) 31% received nimesulide, 33% diclofenac, 20% ketoprofen, 13% ibuprofen, 12% COX2i and 10% piroxicam. NSAID prevalence appeared increased with age, whereas decreased in >80 years except for ibuprofen. Fifty-eight percent of patients <60 yo received <40 posologic units/year. Meanwhile, 62% of patients ≥60 yo received >80 posologic units/year. Patients on ibuprofen + COX2i seemed to have more risk factors (at least 2 risk factors consistent with AIFA prescribing note 1 in >33% of patients treated with COX2i or ibuprofen) in comparison with <30% for those treated with the other NSAIDs (P < 0.001). Moreover, ibuprofen, COX2i and diclofenac showed statistically significant higher frequency of GP co-medication (16%) compared to other NSAIDs, (<13%, P < 0.001). Drug expenditure varied among treatments. Nimesulide and COX2i were associated with the highest NSAIDs + GP total costs/year, respectively 167 and 212 euros. Despite the high frequency of GP co-medications with ibuprofen and diclofenac, lower expenditures were observed, 138 and 118 euros, respectively. CONCLUSIONS: This analysis showed that cost for gastroprotective co-medications were higher than the cost of NSAIDs alone. However, NSAIDs + GP expenditure differed among NSAID treatment groups. PGI12 ECONOMIC CONSEQUENCES OF POORLY CONTROLLED PATIENTS WITH GASTROESOPHAGEAL REFLUX DISEASE IN GERMANY, ITALY AND SPAIN Darba J1, Restovic G2, Kaskens L2, Plans P3 1 Universitat de Barcelona, Barcelona, Spain; 2BCN Health, Barcelona, Spain; 3Generalitat de Catalunya, Barcelona, Spain OBJECTIVES: The aim of this study was to estimate the implications of poorly controlled GERD for patients and the economic implications for health care providers and employers in Germany, Italy and Spain. METHODS: Based on population figures, prevalence and incidence for GERD and its implications and cost data, the number of patients with poorly treated GERD and their implications, as well as the economic consequences for health services and employers were calculated for each country. RESULTS: The amount of patients with poorly treated GERD that have severe esophagitis are estimated to be 740,364 in Germany, 240,559 in Italy and 225, 054 in Spain per year. The number of patients with Barrett’s oesophagus are estimated to be 29,678 in Spain, 19,327 in Germany and 10,079 in Italy. The number of patients with adenocarcinoma are around 483 patients in Germany, 142 in Spain, and 53 in Italy. Costs for poorly treated GERD patients for the health services were estimated to be c18 million for Spain, c12 million for Germany and c7 million for Italy. Absenteeism and presenteeism costs due to poorly controlled GERD for employers were almost none. CONCLUSIONS: Costs for complications in patients with poorly controlled GERD added costs for health care systems for all three countries but almost no extra costs were found for employers. PGI13 MODELING THE LONG TERM CONSEQUENCES OF SUPPRESSING VIRAL REPLICATION IN CHRONIC HEPATITIS B: A COSTEFFECTIVENESS ANALYSIS OF ENTECAVIR (BARACLUDE®) IN TURKEY Lescrauwaet B1, Saylan M2, Malhan S3, Öksüz E3, Korkmaz M3, Kursun E3, Eminsoy G3 1 Bristol-Myers Squibb, Braine-l’Alleud, Belgium; 2Bristol-Myers Squibb, Istanbul, Turkey, Turkey; 3Baskent University, Ankara, Turkey OBJECTIVES: To evaluate the cost-effectiveness of entecavir(ETV) vs.lamivudine(LVD) in the treatment of nucleoside-naive CHB patients and vs. adefovir(ADV) in LVD refractory CHB patients in Turkey. METHODS: A decision-tree model compared cost and effect of treating CHB patients over a 10-year period. Treatment effect in terms of viral load(VL) reduction predicted risk of long-term liver complications. Two CHB patient populations were studied: 1) nucleoside-naïve patients treated for 2 years with ETV(0,5 mg/day) vs. LVD(100 mg/day) and ADV as salvage therapy in case of LVD resistance; and 2)LVD- refractory patients treated for 10 years with ETV(1 mg/day) vs. ADV(10 mg/day). Effectiveness was measured as LYS and QALY. Efficacy data were obtained from pivotal trials, relative-risk estimations were derived from the R.E.V.E.A.L.-HBV Study cohort. a Turkish health care payer perspective was considered and a 3% discount rate was used for costs and outcomes. RESULTS: In the cohort of HBeAg(+/−) CHB patients, net incremental drug cost of ETV was higher compared to LVD. However, ETV was associated with a lower projected number of liver events (−73 and −55CC events, −43 and −33HCC events respectively, −8DC events in each cohort) compared with LVD. In nucleoside-naïve HBeAg(+) patients, ETV compared with LVD and ADV as salvage therapy gained 0.82LYS and 0.68QALYs at an incremental cost of 5384TL(6571TL/LYS and 7865TL/QALY). In nucleoside-naïve HBeAg(-) patients, ETV generated 0.66LYS and 0.55QALY at an incremental cost of 7257TL(11039TL/LYS and 13,203TL/QALY). In LVD-refractory patients, ETV generated 0.48LYS and 0.40QALY with a saving of 14,117TL (−29,416TL/LYS and −35,178TL/QALY). CONCLUSIONS: This cost-effectiveness analysis suggests that in the Turkish health care system, ETV is a cost-effective treatment option compared with LVD in the treatment of naïve CHB patients. In LVD refractory patients, ETV is a cost-saving treatment option relative to LVD with ADV as salvage therapy. Cost-benefits with ETV therapy are explained lower overall CHB treatment costs due to slower disease progression rate and avoidance of resistance associated with LVD. 13th Euro Abstracts PGI14 COST-EFFECTIVENESS ANALYSIS ON THE USE OF TEGASEROD COMPARED TO STANDARD THERAPY IN PATIENTS WITH IRRITABLE BOWEL SYNDROME WITHIN AN INSTITUTIONAL SETTING IN MEXICO Briones B, Garcia F Novartis Farmaceutica S.A., Mexico City, DF, Mexico OBJECTIVES: Irritable bowel syndrome (IBS) is a functional gastrointestinal disorder prevalent in up to 35% of the adult population in Mexico, thus rendering high costs for health institutions. Tegaserod, a 5-HT4 receptor agonist approved for the treatment of IBS, could potentially render a beneficial economic impact. An economic evaluation was performed to assess the economic impact of treating IBS patients with tegaserod in comparison to standard therapy in an institutional setting in Mexico. METHODS: A cost-effectiveness analysis was performed within an institutional setting (Mexican Institute of Social Security, IMSS). The drug comparator used was butylhyoscine, as it is the recommended standard therapy available in the National Formulary, as published in the institution’s IBS clinical guidelines. Resource utilization data was obtained from previous studies; physician consults, drug usage, laboratory tests, endoscopic procedures, non-endoscopic diagnostic procedures, and surgical procedures were used for the analysis; a triangular probability distribution was used. Efficacy data was obtained from published meta-analysis studies. The source of the unit costs was the institution, current for 2009. Discount rate was not used because the time horizon was for 1 year. a probabilistic sensibility analysis was obtained through a Monte Carlo simulation with 100,000 iterations in the weakest parameters. RESULTS: The expected annual cost with standard therapy was $19,431.92 per patient in local currencies (Mexican Pesos, MXP), as compared to $18,394.68 with tegaserod therapy. Tegaserod was more effective than the standard therapy. Probabilistic sensibility analysis showed the same conclusions, regardless of the variability in resource utilization and efficacy. CONCLUSIONS: From an institutional perspective in Mexico, tegaserod is a cost-effective therapeutic option (dominant) for the treatment of IBS in comparison with the current standard therapy featured in the National Formulary. PGI15 COST-EFFECTIVENESS OF SEROLOGIC TESTING FOR DIAGNOSING CELIAC DISEASE Chandra KM1, Costa V2, Blackhouse G1, Goeree R1 1 McMaster University, Hamilton, ON, Canada; 2Ministry of Health and Long-Term Care, Toronto, ON, Canada OBJECTIVES: To assess the cost-effectiveness of tests for diagnosing celiac disease (CD). The following tests were assessed: biopsy; endomysial antibody (EMA); antigliadin antibody (AGA); deaminated gliadin peptide (DGP); tissue transglutaminase (TTG); and panel of tests. METHODS: A decision analysis was constructed to compare costs and outcomes based on sensitivity, specificity and prevalence estimates from a primary systematic literature review. Study results were pooled using a bivariate, binomial, generalized linear mixed model (SAS 9.2). The target population was adults and children experiencing symptoms consistent with CD. Biopsy was assumed to have sensitivity and specificity of 1 since it is considered the gold standard. In serologic testing strategies the positive cases (true or false) were assumed to be confirmed with biopsy. a family physician consult was incurred with serologic testing and a gastroenterologist consult was incurred with biopsy. The outcome of the analysis was expected costs and false negatives (FN). Costs were reported in 2010 CAD$. All analyses were performed using TreeAge Pro Suite 2009. RESULTS: Four strategies made up the efficiency frontier; IgGTTG, IgATTG, EMA and biopsy. All other strategies were either strictly or extendedly dominated. IgGTTG was the least costly and least effective strategy ($178.95, 0.1553 FNs). Biopsy was the most costly and most effective strategy ($396.60, 0 FNs). The cost per FN avoided moving from IgGTTG to the other strategies sequentially on the efficiency frontier were $293, $369, $1401 for EMA, IgATTG and biopsy respectively. One-way sensitivity analyses did not change the ranking of strategies. CONCLUSIONS: All testing strategies with biopsy were cheaper than biopsy alone however they also resulted in more FNs. If a decision maker’s willingness to pay to avoid a FN result is $1401 or greater, then biopsy alone is the most cost-effective strategy. Otherwise, either IgGTTG, IgATTG or EMA is the most cost-effective strategy. PGI16 COST-EFFECTIVENESS OF ON-TREATMENT MONITORING OF HBSAG LEVELS TO PREDICT RESPONSE TO PEGINTERFERON-ALPHA-2A FOR HEPATITIS B TREATMENT UNDER THE PUBLIC PAYER PERSPECTIVE IN BRAZIL Ferreira PRA1, Borges LG2, Santos E2 1 Universidade Federal de São Paulo, São Paulo, SP, Brazil; 2Roche Brazil, São Paulo, SP, Brazil BACKGROUND: Chronic hepatitis B (CHB), affects 350–400 million people worldwide, being associated with increased risk of liver failure and approximately 300,000 cases of hepatocellular carcinoma per year. Treatment with pegylated interferon (48 weeks) is associated with the possibility of sustained disease remission, with no need for maintenance with any other drug. Furthermore, on-treatment monitoring of HBsAg can predict response to peginterferon–alfa-2a (Lau 2009). Currently, only entecavir is available for HBeAg-positive chronic hepatitis B cirrhotic patients under Brazilian Public Healthcare System. OBJECTIVES: To compare costs and outcomes of on-treatment monitoring of HBsAG levels during peginterferon–alfa-2a therapy versus standard entecavir treatment of CHB under Brazilian Public Healthcare System perspective. METHODS: To project disease progression, a Markov model was built A371 based on clinical stages of CHB. a stopping rule was applied to peginterferon–alfa-2a scenario at week 12 based on efficacy results from Lau 2009: if HBsAg > 20000 UI/ ml, treatment is switched to 2nd line treatment (tenofovir). Medical resources related to each stage were obtained from Castelo, 2007. Efficacy of entecavir was obtained from Veenstra, 2007 & 2008. Tenofovir was also available as 2nd line treatment for entecavir scenario. Discount rate for costs and outcomes was 5%, according to Brazilian guidelines for HTA. Costs were reported in 2010 Brazilian Reais (US$1 ≈ $Brz1.8). RESULTS: Assuming a lifetime perspective, expected costs and outcomes for peginterferon-alfa-2a were: $Brz 38,605, 15.37 LYs and 13.64 QALYs. For entecavir: $Brz 39,630, 15.20 LYs and 13.41 QALYs. Cost-effectiveness analysis estimated an ICER of (-$Brz37,614/LYs) and (-$Brz15,556/QALYs) for peginterferon-alfa-2a, being the dominant therapy. CONCLUSIONS: Considering on treatment monitoring of HBsAg, findings suggest that peginterferon–alfa-2a shows higher clinical responses and is a cost saving therapy for the management of CHB patients in the Brazilian Public Healthcare System. PGI17 COULD AN INNOVATIVE SURGICAL PROCEDURE SUCH AS STAPLED HEMORRHOIDOPEXY BE COST-EFFECTIVE AND EVEN COST-SAVING IN UK CLINICAL PRACTICE? Ribaric G1, Kofler J1, Jayne DG2 1 Ethicon Endo-Surgery (Europe) GmbH, Norderstedt, Germany; 2St. James’s University Hospital, Leeds, England OBJECTIVES: Stapled hemorrhoidopexy (SH) offers several benefits over conventional excisional hemorrhoidectomy, which makes it attractive to patients with symptomatic prolapsing piles, yet its provision remains disparate. a full economic evaluation of stapled hemorrhoidopexy was undertaken to establish its cost-effectiveness and take into consideration recent literature on clinical outcomes to show that the SH procedure can become cost-saving compared to CH (conventional hemorrhoidectomy). METHODS: A cost–utility analysis was undertaken to compare the use of SH with CH.The analysis used a probabilistic, cohort-based decision tree within a 1-year timehorizon. Costs were analyzed from hospital and UK NHS perspective and utility was calculated using Quality Adjusted Life-years (QALY). Sensitivity analyses were used to show how variation in the clinical parameters affects cost-effectiveness. RESULTS: The decrease in time in the operating theatre and shorter hospital stay with SH led to a cost saving of £27 per procedure compared to CH on a hospital level. From the NHS perspective the total treatment costs inclusive of the cost of recurrent prolapse, led to an incremental cost of £33 after one year. Calculation of QALYs induced an incremental QALY of 0.0076 and showed an ICER of £4,316, which is a highly costeffective result. Sensitivity analysis taking into account recent publications showed that the SH procedure becomes cost saving with a reduction in recurrence rate of prolapse from 10% to 5% or a 50% reduction in re-surgery rate. CONCLUSIONS: Stapled hemorrhoidopexy is a cost-effective procedure with an incremental costeffective ration (ICER) of £4136. Furthermore the analysis also showed that stapled hemorrhoidopexy could even be cost-saving in UK clinical practice. This economic analysis, in line with the 2007 NICE guidance, supports its widespread implementation. PGI18 ECONOMICS OF A SYMPTOM BASED PATIENT MANAGEMENT APPROACH FOR GASTRO-ESOPHAGEAL REFLUX DISEASE (GERD), USING GERDQ, COMPARED TO AN INVASIVE APPROACH USING GASTROSCOPY AND PH-METRY Jonasson C1, Langkilde LK2, Hatlebakk JG3 1 AstraZeneca AS, Oslo, Norway; 2Wickstrøm & Langkilde ApS, Vejle, Denmark; 3University of Bergen, Bergen, Norway OBJECTIVES: In Norway, proton pump inhibitor (PPI) therapy for GERD can only be reimbursed for patients with a diagnosis verified by gastroscopy or pH-metry. The purpose of this study (NCT00842387) was to evaluate the economic aspects of an alternative, structured treatment pathway based on GerdQ—a validated PRO tool for the diagnosis and treatment decisions in GERD. METHODS: Two treatment pathways in patients presenting with reflux, but no alarm symptoms were evaluated in an open randomized, parallel-group 8 week study. In the New Structured Pathway (NSP) diagnosis and treatment were based on GerdQ scores dividing patients into high impact GERD (treated with esomeprazole 40 mg), medium impact GERD (treated with generic PPI) and low probability of GERD (treated at the physician’s discretion). In the Ordinary Clinical Pathway (OCP), the physician was kept blinded to the GerdQ score and diagnosis was based on findings at gastroscopy or, if needed, pH-metry. Treatment was given according to normal clinical practice. Direct medical cost was estimated based on diagnostic procedures, medication and unscheduled health care visits. RESULTS: Treatment success with NSP diagnostic method was not inferior to OCP. The response rates were 86.5% and 80.1%, respectively. The health economic evaluation could thus be performed as a cost-minimization analysis. More patients in the NSP arm were treated with esomeprazole 40 mg than in the OCP arm (28.4% vs. 10.2%); however, fewer health care visits and diagnostic procedures were seen. The average direct medical cost in the NSP arm was c51 per patient (NOK 410) compared with c390 (NOK 3139) in the OCP arm. CONCLUSIONS: By integrating the New Structured Pathway into clinical practice, direct health care resource use may be reduced without any loss in clinical effectiveness. A372 GASTROINTESTINAL DISORDERS – Patient-Reported Outcomes Studies PGI19 PRESCRIPTION RATES AND ADHERENCE TO PROTON PUMP INHIBITOR THERAPY AMONG PATIENTS WHO REQUIRE LOW-DOSE ACETYLSALICYLIC ACID FOR CARDIOVASCULAR PREVENTION Sörstadius E1, Herlitz J2, Nauclér E3, Naesdal J1 1 AstraZeneca, Mölndal, Sweden; 2Sahlgrenska University Hospital, Göteborg, Sweden; 3 AstraZeneca R&D, Göteborg, Sweden OBJECTIVES: Low-dose acetylsalicylic acid (ASA; 75–325 mg daily) is a mainstay of cardiovascular (CV) prevention. However, some patients taking low-dose ASA may experience upper gastrointestinal (GI) symptoms that are associated with poor adherence to and discontinuation of low-dose ASA. Established gastroprotective strategies, e.g. concomitant proton pump inhibitor (PPI) therapy, may ameliorate these symptoms and thus improve low-dose ASA adherence. METHODS: This subanalysis of a multinational, observational, non-interventional study (NCT00681759) conducted in the United States, Canada and France assessed PPI prescription rates (one-time retrospective survey) and daily PPI adherence rates (prospective 3-month eDiary phase) in adult patients with increased GI risk who had been prescribed low-dose ASA for management of CV risk. Here, increased GI risk was defined as a history of peptic ulcer and/ or complications or additional antiplatelet use (clopidogrel, ticlopidine, dipyridamole). RESULTS: A total of 195 of the 1770 patients in the survey were identified as having increased GI risk (history of peptic ulcer and/or complications, n = 109; concomitant antiplatelet therapy, n = 74; both factors, n = 12); 119 (61%) of whom were not prescribed a PPI. a total of 340 patients entered the eDiary phase, of whom 110 were prescribed a PPI before the first diary day; of these, 79 patients were prescribed a daily PPI for the 3-months. Among these patients, fewer than half (n = 37) took >75% of prescribed daily PPIs. Almost one-third (n = 25) did not take their prescribed daily PPI at all during the 3-month phase. CONCLUSIONS: PPI prescription and adherence rates are low among patients with increased GI risk receiving low-dose ASA for CV risk management. Strategies that deliver gastroprotection with improved adherence rates during low-dose ASA therapy in patients with increased GI risk may be warranted. PGI20 MEDICATION ADHERENCE AND PERSISTENCE IN THE TREATMENT OF ULCERATIVE COLITIS: ANALYSES WITH THE RAMQ DATABASE Lachaine J1, Beauchemin C1, Hodgkins P2, Yen L2 1 University of Montreal, Montreal, QC, Canada; 2Shire Pharmaceuticals, Wayne, PA, USA OBJECTIVES: Non-adherence with oral mesalazines has a significant impact on treatment outcome which is one of the most important predictors for relapse in ulcerative colitis. The objective of this study was to assess adherence and persistence with oral mesalazines, particularly to analyze whether adherence with oral mesalazines is linked to the use of once daily high strength Mezavant® compared to more frequent dosing and/or low strength oral mesalazines. METHODS: A retrospective prescription claims analysis of a random sample of patients from the Quebec provincial public health plan (RAMQ) database was conducted. New users of a mesalazine formulation during the period from January 2005 to December 2009 and with no diagnosis of Crohn’s disease were eligible for inclusion in the analysis. Treatment adherence was estimated using medication possession ratio over a one-year period. For the analysis of persistence to treatment, patients were considered non-persistent if they had not used the mesalazines medication for a period of twice the median duration of prescriptions. Proportion of patients who were persistent was estimated at 3-, 6-, and 12 months after index prescription. RESULTS: The mean age of the study sample was 55.7 years (SD = 18.2) and the proportion of males and female were similar (48.8% vs. 51.2%). The proportion of patients ≥80% compliant on the mesalazine long acting formulation (Mezavant®) (46.3%) was significantly higher compared with all other mesalazine formulations (1.6% to 26.0%) (P < 0.001). The proportion of patients who were persistent at 12 months on Mezavant® (70.2%) was higher when compared with those on any other mesalazine formulations (14.5% to 42.6%) (P < 0.001). Similar trends were observed at all time points examined. CONCLUSIONS: Results of these prescription claims analyses indicate that adherence and persistence to mesalazine formulations are relatively poor, however improved adherence and persistence are observed with the long acting formulation (Mezavant®). PGI21 RELATIONSHIP BETWEEN PATIENT PREFERENCES FOR 5-ASA THERAPIES AND SELF-REPORTED ADHERENCE Hodgkins P1, Swinburn P2, Solomon D1, Yen L1, Dewilde S2, Lloyd A2 1 Shire Pharmaceuticals, Wayne, PA, USA; 2Oxford Outcomes Ltd, Oxford, Oxfordshire, UK OBJECTIVES: The effectiveness of 5-aminosalicylic acid (5-ASA) therapy for mild to moderate ulcerative colitis (UC) is commonly affected by poor medication adherence. The present study was designed to determine if adherence behaviour could be explained by differences in patient preferences for 5-ASA therapies. METHODS: A discrete choice experiment (DCE) survey was used to explore patient preferences for different aspects of oral 5-ASA therapy. The DCE survey captured trade-offs that patients were willing to make and was based on a literature review, clinician interviews, and in-depth interviews with UC patients (followed by cognitive debriefing). Six attributes were identified: Ease–of–swallowing, Number of administrations per day, Number of pills (per administration), Symptom flare resolution; Likelihood of flare occurrence and Cost (to estimate willingness-to-pay (WTP) for improvements in 13th Euro Abstracts attributes). Adherence behaviour was assessed using the Modified Morisky Scale. Participants (mild to moderate UC patients, n = 400) in the UK, US, Germany, and Canada were recruited through specialist patient recruitment agencies and the survey was administered via the internet, following IRB approval. Data were analyzed using the conditional logit procedure. RESULTS: Clinical effectiveness was most highly valued by participants independent of country of origin (e.g. reduction in annual flare risk to 10%: WTP = £78.81 per month) and a return to normal bowel functioning with mucosal healing (WTP = £29.24). Significant interaction terms identified that people who reported good adherence placed greater value on symptom control compared with self reported poor adherers to therapy (P = 0.011). CONCLUSIONS: Data suggest that the most highly valued aspect of therapy was effectiveness. Therefore, patients may adhere to a medication better if they place greater value on its ability to effectively treat their UC. Furthermore, these data suggest a possible avenue for physicians to explore with their patients to improve adherence to the treatment regimen by highlighting the risk-benefit profile of 5-ASAs in the treatment of UC. PGI22 MAPPING PAC-QOL SCORES ONTO THE EQ-5D 1 1 2 Parker M , Haycox A , Dubois D 1 University of Liverpool Management School, Liverpool, UK; 2Patient Value Solutions, Huldenberg, Belgium OBJECTIVES: The clinical trial programme for Prucalopride, a selective and high affinity 5-HT4 receptor agonist, incorporated a constipation specific HRQoL measure (PAC-QOL) and SF-36 but not EQ-5D. a mapping relationship was developed to link EQ-5D to PAC-QOL using established algorithms linking EQ-5D and SF-36. METHODS: Trial responses (n = 5488) on a common patient data set enabled an empirical link to be established between the SF-36 and PAC-QOL which was extended to EQ-5D through the established algorithm. Having established this relationship a range of functional forms for mapping PAC-QOL onto EQ-5D were tested ranging from a simple linear relationship to more complex mapping structures incorporating quadratic and interactive terms. RESULTS: The relationship between PAC-QOL and EQ-5D was generally good. The estimated equation for deriving EQ-5D from PAC-QOL in its’ simplest linear functional form was: EQ-5D = 97.7–9.8 (PAC-QOL). This implies that a one point change in PAC-QOL overall score would lead to a 9.8% change in EQ-5D. As expected, the mapping was largely limited to the upper health states of EQ-5D given that chronic constipation by itself is unlikely to lead to the severest forms of disability. This initial regression analysis displayed elements of nonlinearity and hence a more complex analysis was undertaken which incorporated square and interaction terms. This new functional form facilitated a more accurate relationship to be established between PAC-QOL and EQ-5D. CONCLUSIONS: Mapping is required whenever a preference based generic HRQoL measure is not directly collected in clinical trials. Very limited theoretical guidance is available to structure such analyses and therefore empiricism largely rules. However by testing the robustness of the results to different assumptions and functional forms a robust mapping can be developed. This process was employed here to convert PAC-QOL into EQ-5D utility scores for incorporation into Cost Utility analyses. PGI23 IDENTIFYING ENDPOINTS FOR IRRITABLE BOWEL SYNDROME (IBS) CLINICAL TRIALS: INCORPORATING THE PATIENT’S VOICE Fehnel SE1, Ervin CM1, Lewis BE2, Carson RT3, Johnston JM2 1 RTI Health Solutions, Research Triangle Park, NC, USA; 2Ironwood Pharmaceuticals, Cambridge, MA, USA; 3Forest Research Institute, Jersey City, NJ, USA OBJECTIVES: 1) Identify a comprehensive set of symptoms experienced by patients with irritable bowel syndrome with constipation (IBS-C), and 2) Identify the most important symptoms for measurement in clinical trials for IBS-C. METHODS: Two iterative sets of in-depth interviews were conducted in different US cities, with a total of 27 participants meeting modified Rome II criteria for IBS-C. a semi-structured interview guide was used, beginning with a series of open-ended questions to elicit all relevant symptoms, followed by interviewer probes to fully understand the relationships among the concepts. Multiple rating and ranking methods were used to develop a subset of IBS-C symptoms of greatest importance to patients. For example, participants were asked to identify their most bothersome IBS-C symptoms, as well as those in which they would most like to see an improvement with treatment. RESULTS: When asked to describe their IBS-C symptoms, patients reported 54 potentially distinct concepts: 8 abdominal symptoms, 12 bowel symptoms, 31 additional physical symptoms (e.g., nausea, headache), and 3 emotional issues (e.g., irritability, depression). Some symptom terms were highly related (e.g., abdominal pain and stomach ache) and others could be considered consequential to IBS-C (e.g., hemorrhoids, vomiting). Results of the subsequent rating and ranking tasks suggest that abdominal pain, abdominal discomfort, bloating, stool frequency, stool consistency, straining, and incomplete evacuation were distinct and represent patients’ most bothersome symptoms. Further, according to the patients, improvements in these symptoms would constitute an improvement in IBS-C overall. CONCLUSIONS: Patient input is vital to identify the full spectrum of symptoms and to determine an optimal set of clinical trial endpoints. Within and across the two separate rounds of interviews, participants consistently reported the importance of abdominal pain, abdominal discomfort, bloating, stool frequency, stool consistency, straining, and incomplete evacuation, demonstrating concept saturation and supporting the measurement of these symptoms in IBS-C clinical trials. 13th Euro Abstracts PGI24 A PILOT STUDY OF PATIENT PREFERENCES FOR MID-THERAPY ASSESSMENT TIMING IN CHRONIC HEPATITIS C TREATMENT Kauf T1, Nelson DR1, Schelfhout J1, Zeigler L1, Bhula M1, Grant WC2 1 University of Florida, Gainesville, FL, USA; 2James Madison University, Harrisonburg, VA, USA OBJECTIVES: For many courses of therapy, assessments of treatment effectiveness are used to inform treatment continuation decisions. If a “mid-therapy” assessment (MTA) is positive, the patient is indicated to continue treatment; if negative, treatment may be discontinued. Using expected utility theory, we demonstrated previously that the availability and timing of such assessments may influence patients’ treatment initiation decisions. We conducted a pilot study among chronic hepatitis C (CHC) patients to examine preferences over MTA timing and treatment initiation. METHODS: A stated preference survey was developed and pre-tested among 10 community volunteers and then administered to 49 CHC patients. The survey described two MTAs for CHC treatment: one at 4 weeks (rapid virologic response, RVR) and the other at 12 weeks (early virologic response, EVR). Test characteristics varied between the two MTAs, but the decision algorithm (stated above) was held constant. Multiple response formats were used to elicit preferences across MTA and treatment initiation. Results are summarized as means and proportions. RESULTS: Average age was 54.6 yrs; 55.1% were male; 32.7% were treatment-naïve; 67.3% reported their general health as good or better. Three subjects failed to answer one or more MTA or treatment choice questions. More subjects had previously heard of EVR compared to RVR (43.8% vs. 14.3%, p > 0.01). Given a choice between MTAs, 55.3% of subjects indicated a preference for RVR. Treatment utilizing RVR was somewhat or strongly preferred by 57.4% of subjects, and 61.2% said they would choose RVR if only one test was available. However, more subjects responded that they would accept treatment with EVR than with RVR 60.9% vs. 55.3% (difference not significant). CONCLUSIONS: Patients with CHC consistently indicated MTA preferences that are at odds with current treatment guidelines. Further research is needed to understand the relationship between MTA preference and treatment initiation. PGI25 CHILD AND PARENT REPORTS OF SYMPTOMS OF IRRITABLE BOWEL SYNDROME WITH CONSTIPATION (IBS-C): RESULTS OF QUALITATIVE INTERVIEWS Arbuckle R1, Lewis BE2, Carson R3, Abetz L1, Johnston JM2 1 Mapi Values Ltd, Bollington, Cheshire, UK; 2Ironwood Pharmaceuticals, Cambridge, MA, USA; 3Forest Research Institute, Jersey City, NJ, USA OBJECTIVES: The Rome III criteria define pediatric irritable bowel syndrome with constipation (IBS-C) as abdominal pain or discomfort associated with constipation symptoms. Historically, few pediatric IBS-C trials have used symptom measures that were developed with patient input. This study aimed to develop pediatric IBS-C symptom measures through qualitative interviews with children with IBS-C and their parents/caregivers. METHODS: Children diagnosed with IBS-C (aged 6–8 [n = 10], 9–11 [n = 10] and 12–17 [n = 10]) and their parents were interviewed. Thematic analysis of interview transcripts identified concepts. Age appropriate items and response options, developed to measure each concept, were reviewed by expert clinicians. RESULTS: IBS-C symptoms identified as being bothersome to children included: abdominal symptoms such as abdominal pain (“stomach hurts”) and bloating (“tummy like a balloon”), and bowel symptoms such as infrequent bowel movements (“don’t go often”), difficulty defecating (“it won’t come out”), straining on defecation (“have to push hard”), rectal pain during defecation (“butt hurts”), hard stools (“hard and bumpy”), large stools (“it’s like a log”), and a feeling of incomplete evacuation (“some that won’t come out”). Saturation was achieved for the above concepts. Parents relied on behavioural signs of IBS-C (such as inactivity or irritability) and the child telling them about symptoms when assessing their child’s condition. In general the majority of parents and children agreed in their reports of symptoms, though there were some minor inconsistencies. CONCLUSIONS: Consistent with guidelines for patient-reported outcomes, these results were used to develop age-appropriate questions to measure both abdominal and bowel symptoms. The instrument is currently undergoing testing to assess patient understanding and relevance. Results from qualitative interviews with children with IBS-C suggest abdominal and bowel symptoms are both important and bothersome to pediatric IBS-C patients and should be included in treatment assessments. PGI26 PREVALENCE AND IMPACT IN WORK PRODUCTIVITY OF GASTROESOFAGEAL REFLUX DISEASE (GERD) IN PRIMARY CARE PATIENTS WITH UPPER GASTROINTESTINAL (GI) SYMPTOMS. THE GREEK GERDQ STUDY Rokkas T1, Panitti E2, Nikas N2 1 Henry Dunant Hospital, Athens, Attiki, Greece; 2AstraZeneca, Athens, Greece OBJECTIVES: GERD is a common condition in daily clinical practice associated with reduced quality of life and impact on productivity. The aims of the current study were to estimate the prevalence of GERD in primary care by using a novel diagnostic tool (GerdQ questionnaire) and to assess the impact of the disease in productivity. METHODS: The Greek GerdQ study was a cross-sectional, single-visit, epidemiological study in patients presenting to their physician with upper GI symptoms. Data on patients’ demographics, medical/GI history and upper GI symptoms were recorded by the investigators. All subjects completed the GerdQ, and those scored ≥ 8 also completed the WPAI-GERD questionnaire for the evaluation of GERD impact on produc- A373 tivity. RESULTS: Overall 889 (887 evaluable) patients were enrolled by 91 primary care physicians. 47% of patients were male. Mean (± SD) age was 51 (17) years with 37.5% of patients presenting no previous history of GI tract disorders. The most prevalent GI symptoms in the week prior to study visit were heartburn, regurgitation and belching, presenting a high frequency (≥2 days/week or daily) in 62.4%, 47% and 50.2% of the patients, respectively. Antisecretory treatment was reported by 62% of patients. Based on GerdQ, 71.8% patients had GERD (GerdQ score ≥ 8) while 45.1% of them were suffering from disrupting disease (GerdQ impact score ≥ 3). Mean (± SD) absenteeism due to GERD was 2.3 (4.9) hours/week with a mean (± SD) of 11.8 (9.6) additional hours/week lost due to presenteeism. The observed reduction in daily life productivity was 37.4%. CONCLUSIONS: These data suggest that in Greece, GERD is a highly prevalent condition in primary care patients with upper GI symptoms, posing a significant burden to patients in terms of reduced productivity both in work and in daily life. PGI27 STRUCTURED MANAGEMENT STRATEGY BASED ON THE GERDQ QUESTIONNAIRE VERSUS USUAL PRIMARY CARE FOR GASTROESOPHAGEAL REFLUX DISEASE: META-ANALYSIS OF FIVE EUROPEAN CLUSTER RANDOMIZED TRIALS Ponce J1, Garrigues V1, Agréus L2, Tabaglio E3, Gschwantler M4, Guallar E5, Tafalla M6, Nuevo J6, Hatlebakk JG7 1 Hospital Universitario La Fe, Valencia, Spain; 2Karolinska Institute, Stockholm, Sweden; 3 Società Italiana Medicina Generale, Firenze, Italy; 4Wilhelminenspital, Wien, Austria; 5Welch Center for Prevention, Epidemiology, and Clinical Research, Baltimore, MD, USA; 6 AstraZeneca, Madrid, Spain; 7University of Bergen, Bergen, Norway OBJECTIVES: Gastroesophageal reflux disease (GERD) has substantial impact in primary care, but the optimal approach to management is uncertain. a structured management strategy may improve the diagnosis and therapeutic management of GERD, but individual studies may be limited by their focus on local strategies that may not be valid for other countries. METHODS: We conducted a meta-analysis of five cluster randomised clinical trials comparing a new management strategy with usual care in patients with GERD conducted in Austria, Italy, Norway, Spain and Sweden (NCT00842387). The intervention strategy was based on the self-administered validated GerdQ questionnaire to stratify adult patients with classical symptoms of GERD (heartburn or regurgitation) according to the frequency and impact of symptoms. The most effective acid-suppressive therapy (esomeprazole 40 mg once daily) was used only in patients with the highest GerdQ symptom impact score (≥3). The primary outcome was non-response to treatment defined as a total GerdQ score ≥ 8 at the end of follow-up. Odds ratios for the primary outcome were combined using a random effects model. RESULTS: A total of 2400 patients were enrolled and average follow-up ranged from 4 to 18 weeks. The odds ratios for lack of treatment response with the structured management strategy compared with usual care ranged from 0.22 to 0.84, across studies. The random-effects combined odds ratio for non-response to treatment was 0.56 (95% CI 0.22–0.90; p = 0.001), with significant between-study heterogeneity (p < 0.001). CONCLUSIONS: Stratification of patients according to the GerdQ questionnaire, using a locally adapted primary care management strategy for GERD significantly increased the likelihood of a response to treatment compared to usual clinical practice, although significant between-country heterogeneity suggests that GERD management can still be improved. PGI28 HEALTH-RELATED QUALITY OF LIFE IN OPIOID INDUCED CONSTIPATION PATIENTS IN SPAIN Guijarro P1, Viqueira A1, Alonso-Babarro A2, Fernandez G1 1 Pfizer Spain, Alcobendas, Madrid, Spain; 2La Paz Hospital, Madrid, Spain OBJECTIVES: To analyze the impact of opioid induced constipation (OIC) on patients’ health related quality of life (HRQoL). METHODS: An observational, multicenter study was carried out in Spain. All patients were free of constipation at baseline and received opioids for at least 2 months. The impact of OIC on patients’ HRQoL was determined in a cross-sectional phase of the study. Patients were evaluated depending on response to oral treatment for OIC. HRQoL was assessed by the following tools: the Spanish version of the EuroQoL (EQ-5D) and EuroQoL Visual Analogue Scale (EQ-VAS). The CVE-20 questionnaire and a specific question were used to assess the perceived health status related to constipation. RESULTS: Patients with OIC showed a mean (SD) overall CVE-20 score of 48.0 (18.6), the time spent defecating and discomfort caused by a bloated stomach were highly rated in the general physical dimension (56.4% in both items). Laxative dependence was the most rated item (51.9%) in the social dimension. Regarding general QoL, EQ-5D, the dimension in which patients were more affected was the pain/discomfort dimension (89.7%). Mean EQ-VAS score (SD) was 51.3 (19.3), mean EQ-5D VAS tariff (SD) was 0.45 (0.25) and mean EQ-5D TE tariff (SD) was 0.38 (0.40). Overall, responders to oral laxatives showed a better HRQoL than non-responders. The CVE-20 questionnaire score in responders was higher than in non-responders (50.8 vs. 40.6, respectively, P < 0.001). Regarding general QoL, the EQ-5D anxiety/depression was the only dimension significantly more affected in non-responders (p = 0.003). Accordingly, responders obtained statistically significant higher EQ-VAS score (SD), 53.0 (18.9) vs. 45.5 (18.4) (p = 0.004). CONCLUSIONS: OIC is one of the most distressing opioid adverse events and as such, has a negative impact on HRQoL. The differences observed in the CVE-20 and EQ-VAS scores between responders and non-responders are statistically significant and may also have a clinical impact. A374 GASTROINTESTINAL DISORDERS – Health Care Use & Policy Studies PGI29 PPI GENERAL PRACTITIONERS’ PRESCRIBING IN THE TREATMENT OF ACID-RELATED DISORDERS: THE IMPACT OF GENERIC PPI LAUNCH INTO THE PHARMACEUTICAL MARKET IN ITALY Cammarota S1, Sarnelli G2, Citarella A1, Menditto E1, Riegler S1, Bruzzese D2, Cuomo R2 1 CIRFF, Federico II University, Naples, Italy; 2Federico II University Hospital School of Medicine, Naples, Italy OBJECTIVES: To investigate the phenomenon of PPI switching in the acid related disorders treatment in an Italian primary care setting during 2005–2008 (before and after PPIs’ patents expired) and to estimate the costs of health care resources consumption associated with this phenomenon. METHODS: Retrospective cohort study was performed analyzing data from 127 GPs of Naples in the south of Italy. PPI users for ARD treatment within each study year were selected from source population. Switchers were defined patients changing from one PPI to another within each study year. Multivariate logistic regressions were used to assess the potential predictors of PPI switching and to investigate the factors influencing the direction of the switch. Cost was expressed as Euro 2008 per PPI user. RESULTS: The phenomenon of PPI switching rose from 13.0% in 2005 to 16.7% in 2008 with a peak of 18.8% in 2006. Calendar years, long-term treatments and GERD diagnosis were positive predictors of PPI switching. All years analyzed (versus 2005) were associated with switching to lansoprazole while the 2008 year was strongly associated with switching to omeprazole and pantoprazole. Very long-term treatment (>11 pack/years) group accounted for 66.3% of the total primary care cost. Switchers increased primary care costs by c61.1 compared with no switchers, reaching an incremental cost of c133.1 per useryear in the case with more than 1 switch. CONCLUSIONS: In Italy the launch of generic PPIs in the national market generates the increasing amount of chronic treatments and therapeutic substitution that will probably have had a negative impact on the total savings on PPIs, achievable by the introduction of generic products. Policy rules favouring generic PPIs prescribing can often influence physicians’ decision to select the appropriate treatment for each patient. PGI30 ASSESSMENT OF FEASIBILITY OF THE METHODOLOGICAL APPROACH DESCRIBED IN THE MODEL FOR THE REGULATION OF REIMBURSEMENT PRICES IN GERMANY Aidelsburger P1, Benkert D1, Neumann A2, Niebuhr D3, Wasem J2 1 CAREM GmbH, Sauerlach, Germany; 2University of Duisburg-Essen, Essen, Germany; 3 University of Applied Sciences Fulda, Fulda, Germany OBJECTIVES: The generic drug market within the statutory health insurance (SHI) in Germany is affected by price regulations like fixed amounts or rebats. Legal amendments led to an increasing importance of price discounts with uncertain impact on pharmaceutical supply of the patients. To assure a justifiable price for the pharmaceutical companies and high quality a model has been developed by the Federal Association of Medicines Manufacturers e. V. (BAH) for the regulation of reimbursement prices, e.g. cost-benefit based price corridors. Aim of the study is to test the described methods for feasibility. METHODS: The prior published methods for the development of reimbursement price corridors are applied and tested for feasibility by use of the indication Gastroesophageal Reflux Disease (GERD). The methods intend the derivation of one or more corridors following a defined decision algorithm based on primary and secondary benefit criteria. Therefore a systematic assessment of benefits of proton pump inhibitors (PPIs), H2-Blockers and sucralfat is the foundation of further decisions concerning the number of corridors to consider. Several methodical approaches for the definition of height of each corridor are tested. RESULTS: PPIs are more effective than H2-Blocker or sucralfat concerning primary benefit endpoints. The assessment is based on secondary literature for financial and time reasons that will be relevant when implementing the model in real decision problems, too. Based on the results of the benefit assessment two corridors were implemented into the model. The height of each corridor can be derivated by use of angles as well as by other mathematical parameters like mean, quartiles etc. Both approaches show specific limitations. CONCLUSIONS: The feasibility of the reported concept can be shown. Several limitations, e.g. assessment of benefits based on secondary literature should be considered. Use of angles or mathematical parameter should be discussed with decision makers before implementation. PGI31 EVALUATING THE POST OPERATIVE PRESCRIBING PRACTICES FOR APPENDICITIS IN PUBLIC HEALTH FACILITIES IN PAKISTAN Malik M, Hussain A, Abassi N, Khan J Hamdard University, Islamabad, Punjab, Pakistan OBJECTIVES: To investigate the pattern of prescribing practices for postoperative appendicitis in public health facilities. METHODS: The study population consisted of the largest public hospital Federal Government Services Hospital (Polyclinic) from Punjab, Pakistan; Islamabad. a sample of 100 prescriptions of post operative appendicitis was collected from the public hospital. RESULTS: A total of 13.3% prescriptions out of 100 contain 2 drugs, 40% contains 3 drugs, 40 % contains 4 drugs & 6.7% contains 5 dugs. Only 6.7% prescriptions contain drugs prescribed by generic while 93.3% prescriptions were not prescribed by generic name. a total of 3.3% prescriptions contain no antibiotics, 76.6% prescriptions contain 1 antibiotic, while 20% prescriptions contain 2 antibiotics. On the other hand 23.3% prescription does 13th Euro Abstracts not contain any injection, 10% prescriptions contain 2 injections, 23.3% prescriptions contain 3 injections, while 43.3% prescriptions contain 4 injections. One hundred percent of prescribers suggested that standard treatment guidelines and essential drug list should be available and updated regularly in hospital and must be strictly followed. a total of 33.3% of prescribers think the essential drug list should be properly regulated and updated should by the hospital management, 33.3% prescribers think by Ministry of Health while 33.3% prescribers thinks it is the duty of Hospital pharmacist. This difference was noted with respect to age of the prescriber (p = 0.020) in the prescribing practice of physicians in the public health facility. CONCLUSIONS: The major reasons for irrational drug use in case of post operative appendicitis were due to polypharmacy, overuse of antibiotics and injection and lack of standard treatment guidelines in the hospital. Thus the extent of irrational drug use in the public sector calls for in-depth investigation of the system factors and motivations that underlying these problems in the practice and the development of interventions that target the causative factors of inappropriate prescribing practice in Pakistan. GASTROINTESTINAL DISORDERS – Conceptual Papers & Research on Methods PGI32 USING DATA ENVELOPMENT ANALYSIS TO ESTABLISH THE EFFECTIVENESS EVALUATION OF DRUGS Wu ML Taichung Veterans General Hospital, Taichung City, Taiwan OBJECTIVES: The development and research in pharmaceutical industry has constantly created new listing drugs which provide more therapeutic effect and lower side-effects than former drugs, but the price become more expensive relatively. In order to balance the trade-off between medical quality and cost reduction, hospitals have to make the decision-making between the former drugs and the new listing drugs, which have similar therapeutic effectiveness. In this regard, the most primary issue in this research to solve is to establish a more objective and efficient decision-making analysis that consider both effectiveness and cost for the drugs which have the similar therapeutic effect. METHODS: We use data envelopment analysis (DEA) to solve this problem, We screen six critical factors, daily drug expenses, drug profits, side-effect, consumption, patent duration and lead time. Evaluation of relative efficiency of the option with DEA method must be built on the relative performance data of every input or output attributes of each decision-making unit. RESULTS: This study uses four kinds of proton pump inhibitors (PPI) for effectiveness evaluation (the code A, B, C, D drugs to replace its original name), first according to definition of the factors to collect relevant data. We use the DEA method to treat these data, Drug a in the DEA method analysis results is the most effective (6.6542E-06), Drug D is the second (4.2381E-06), Drug B is the third (2.1165E-06) and Drug C is the worst (1.2786E-06). CONCLUSIONS: Through this research model, we transfer complex decision-making goals into various measurable or comparable factors that can compare the relative importance. The evaluation result may not let all decision-makers and users to adopt at all. But it is more comprehensive and objective to evaluate the effectiveness of decision-making models than that in the past. INDIVIDUAL’S HEALTH – Clinical Outcomes Studies PIH1 EMERGENCY CONTRACEPTION FOR UNINTENDED PREGNANCY: ROLE OF ULIPRISTAL, A NOVEL PROGESTERONE RECEPTOR MODULATOR Aggarwal A Heron Health Private Ltd, Chandigarh, India OBJECTIVES: Unintended pregnancies represent a global health problem with over 80 million annual pregnancies. These are associated with an increased risk of morbidity and a considerable impact on women’s quality-of-life. Globally, 38% of all pregnancies are unintended resulting in 42 million induced abortions and 34 million unintended births. This corresponds to 43% of all child-births worldwide. a novel drug “Ulipristal” was introduced in Europe in May 2009 with efficacy beyond 72 hours after unprotected coitus. The objective of this review was to determine the safety and efficacy profile of ulipristal compared to levonorgestrel. METHODS: A number of databases (PUBMED, EMBASE, POPLINE, CENTRAL and clinicaltrials.gov) were searched and the citations screened to identify randomised controlled trials (RCTs) reporting efficacy and safety outcomes of ulipristal. Grey literature was searched to identify the cost associated with unintended pregnancies. Summary-statistics (Random-Effects: DerSimonian-Laird) were used to assess pregnancy and adverse-events outcomes. RESULTS: Seven RCTs were identified and only two reported pregnancy outcomes. In a pooled analyses when compared to levonorgestrel, ulipristal showed better efficacy results in preventing pregnancy following unprotected coitus on day 1 and day 3 (Day 1, RR = 0.435, [95%CI:0.148, 1.279] and Day 3, RR = 0.376, [95%CI:0.110, 1.283]). However, these results were non-significant. Conversely, subgroup analysis showed that if administered on day 2, this trend was reversed with administration of levonorgestrel being associated with fewer pregnancies (RR = 1.309, [95%CI: 0.572, 2.996]). The frequency of adverse events like nausea, headache, fatigue and dizziness was similar with both the drugs. Unintended pregnancies were found to be associated with huge costs that amounted up to US$ 5 billion in US during 13th Euro Abstracts 2002 and £335 million in UK during 2001. CONCLUSIONS: Ulipristal appears as effective as levonorgestrel in preventing unintended pregnancies with similar safety profile. Substantial cost-savings can be expected with appropriate awareness programs about emergency contraceptives among women. PIH2 PEDIATRIC INTENSIVE CARE UNIT (PICU) ADMISSIONS FOR RESPIRATORY SYNCYTIAL VIRUS (RSV) INFECTION IN THE ERA OF PALIVIZUMAB PROPHYLAXIS Paes BA1, Butt M1, Janes M2, Symington A2, Elliott L2, Steele S1 1 McMaster University, Hamilton, ON, Canada; 2McMaster Children’s Hospital, Hamilton, ON, Canada OBJECTIVES: To examine the characteristics of patients admitted to PICU with RSV infection following provincial approval of RSV prophylaxis in June 2002. Secondly, to determine if patients had received palivizumab and document incurred morbidities. METHODS: A retrospective, hospital medical records review of all PICU admissions for RSV infection from January 1, 2003 to December 31, 2009. RSV infection was identified by ICD codes and cases were confirmed by RSV IFA test, culture, or PCR. Data was collected on baseline demographics, underlying disease, criteria for hospitalization, type of respiratory illness and management, complications and palivizumab prophylaxis. Group 1 patients (≤2 years) were compared to Group 2 (>2–18 years). RESULTS: A total of 181 patients were admitted with RSV infection over 7 years. Group 1 (n = 152); Group 2 (n = 29) had a mean admission age in months (SD); 3.7 (5.7) versus 59.9 (37.7). Majority (79.6%) Group 1 versus only 20.7% Group 2 (P < 0.001) had no underlying disease. 97.4% versus 93.1% were admitted with respiratory distress and most had bronchiolitis; 88.8% versus 55.2 % (P < 0.001), decreased oxygen saturation; 77% versus 75.9% and inability to maintain intake; 75.7% versus 44.8% (P < 0.01). Fifty percent versus 41.3% required mechanical ventilation and 81.6% versus 69.0% received antibiotics. Median days (range) in PICU was 5 (1–73) versus 2 (1–36). Median length of hospital stay days (range) was 9 (1–113) versus 7 (1–64) (P < 0.05). The overall RSV positive, PICU hospitalization rate was 5.7%. There were few complications: respiratory (2.8%); concurrent bacterial sepsis (5.5%). Only 3.3% children had received prophylaxis. 1 death was attributed to RSV infection. CONCLUSIONS: Majority of admissions comprised infants ≤2 years of age. 63.3% were ≥36 weeks gestation without underlying disease and in total 88.4% would not qualify for prophylaxis. Current Canadian RSV prophylaxis guidelines have significantly impacted PICU admission rates in high-risk infants. PIH3 PARENTERAL ALANYL-GLUTAMMINE IN CRITICALLY ILL PATIENTS: A BAYESIAN META-ANALYSIS OF PUBLISHED TRIALS Pradelli L1, Zaniolo O1, Iannazzo S1, Eandi M2 1 Adres srl, Torino, Italy; 2University of Torino, Torino, Italy OBJECTIVES: Glutamine, although abundant in human tissue, can become conditionally essential in clinical conditions with hyper-catabolism and glutathione depletion (burns, pancreatic necrosis, surgical complications), but has not been added to parenteral nutrition solutions for a long time, for its alleged non-essentiality and the low solubility and stability in aqueous solutions, which have been solved by conjugation with alanine. In 2002 a meta-analysis of available trials conducted with alanylglutamine dipeptide revealed significant reductions of mortality, infections and ICU length of stay. Since then, data from other trials have become available. Aim of the present study is to update the treatment effect estimates by means of a series of Bayesian random effects models. METHODS: We searched EMBASE and Medline for clinical trials of standard total parenteral nutrition (TPN) vs. TPN + parenteral alanylglutamine in critically ill patients reporting hospital mortality, relative ICU-incident infection rate, and relative hospital length of stay. For each outcome, a series of Bayesian random effects models was specified, in which the treatment effect observed in the individual trials is assumed to be drawn from a common distribution and expressed as a relative risk or duration. RESULTS: Outcomes from 15 trials and 781 patients were retrieved. The main models, i.e. simple hierarchy random effects models with neutral priors, estimate a relative mortality of 0.70 (95% CrI: 0.46–0.97), a relative infection rate of 0.71 (95% CrI: 0.49–0.97), and a relative length of stay of 0.91 (95% CrI: 0.76–1.00). Secondary analyses indicated some heterogeneity in the magnitude and reliability of the benefits in sub-groups of the wider critically ill patient population. The incorporation of prior knowledge leads to significantly more precise estimates and permits to obtain comfortable reliability even on subgroup-specific treatment effects estimates. CONCLUSIONS: In conclusion, the available evidence supports a highly credible beneficial effect of alanylglutamine on mortality, infections and hospital length of stay in ICU-admitted critically ill patients. PIH4 COMPARING THE MODEL PREDICTED VACCINE IMPACT AGAINST ROTAVIRUS HOSPITALIZATION WITH OBSERVED DATA IN BELGIUM Standaert B1, Strens D2, Van Bellinghen LA3, Van Vlaenderen I2 1 GlaxoSmithKline Biologicals, Wavre, Belgium; 2Deloitte, Diegem, Belgium; 3Deloitte, Brussels, Belgium OBJECTIVES: In reimbursement files models estimate the events avoided over time with new interventions. Simulations are performed in the absence of long-term observational data. In this study observed hospitalization data prior- and post-rotavirus vaccination in Belgium were compared to predicted (i.e. modeled) results after 1,2 and 3 years. METHODS: A Markov cohort model estimated over time the number of A375 hospitalizations potentially avoided with rotavirus vaccination in Belgium. We adjusted the model to the observed vaccine coverage and to RotarixTM vaccine efficacy from clinical trials. The obtained modeling results were compared with observed data collected from 9 Belgian hospitals (2 years pre- and 3 consecutive years post-vaccination). The outcomes of both studies are expressed as a percentual decrease in hospitalizations for 2 age-groups (<2 and <5 years old), after each year post-vaccination. The differences are presented in absolute and relative (%) differences for each subsequent year post-vaccine launch. RESULTS: The observed data provided slightly better outcomes than the modeled results. After 1 year the absolute difference in decrease of hospitalizations between observed and modeled data was 3% (11%) for both agegroups. After 2 years the absolute differences were 3.5% (6%) and 6% (12%) for the 2 y and 5 y age-group respectively. In the last observation year the absolute differences were respectively 2.8% (4%) and 9.3% (13%). CONCLUSIONS: Compared with the observed data the model estimates are conservative. The more favorable observational results are explained by the indirect vaccine effect on non-vaccinated age-groups which is not captured by the static model. The relatively better model fit in the <2 y old with increasing time is explained by the accumulated vaccine impact over time: from 26% (1st year) to 68% (3rd year). Rotarix is a trademark of the GlaxoSmithKline group of companies. PIH5 PHARMACOEPIDEMIOLOGICAL BURDEN OF PREGNANCY IN BELARUS Gavrilenko L, Romanova I, Kozhanova I Belarusian State Medical University (BSMU), Minsk, Belarus OBJECTIVES: Application of medicines during pregnancy is a challenge of clinical pharmacology. We have analyzed priorities of physicians while choosing medications in the course of various diseases during pregnancy and adherence of pregnant women to the appointed treatment in Belarus. METHODS: It was prospective multi-center cross-sectional epidemiological research. We interviewed 1334 pregnant women and 619 physicians with diverse specialties in 6 regions of the country using a structured questionnaire from January to December 2009. All questionnaires have undergone statistical analysis. RESULTS: We found that 34.5% of the pregnant women who took part in questioning have had chronic diseases. Medical products were prescribed for 91.7% (1223) women. a total of 23.8% of them have received more than 5 medical products. Vitamins and drugs with microcells (magnesium, iodine, iron and calcium), herbal drugs (valerian, leonurus and eleutherococcus) were in the lead in structure of prescription. 91,8% of the pregnant women took all prescribed medicines. 7.7% (103) women took antibacterial medicines following recommendations of the doctor. a total fo 365 women (27.3%) took medicines without prescription (625 cases). Non-prescribed drugs during pregnancy had mostly been recommended by family or friends (24.1% of women) and pharmacists in drugstores (24.1%). CONCLUSIONS: Pregnant women in Belarus are active consumers of medicines (vitamins, microcells, herbal drugs, bioadditives) themselves and with doctors’ prescription. The widespread use of medicines indicates an increased need for documentation and education about the safety of medicines in the course of pregnancy. Application of these medications creates additional financial burden (in addition to necessary medicines and products), and it does not necessarily positively influence health. Educational programs for pharmacists, doctors and pregnant women are necessary for improvement of drug administration. PIH6 IMPACT OF ROTAVIRUS VACCINATION ON ACUTE GASTROENTERITIS RELATED EMERGENCY ROOM VISITS IN CHILDREN ≤FIVE YEARS OLD IN BELGIUM Morel T1, Strens D1, Raes M2, Standaert B3 1 Deloitte, Diegem, Belgium; 2Jessa Ziekenhuis, Hasselt, Belgium; 3GlaxoSmithKline Biologicals, Wavre, Belgium OBJECTIVES: Belgium is one of the few countries to implement rotavirus vaccination within its universal paediatric immunization programme. This study was designed to measure the impact of rotavirus vaccination on acute gastroenteritis (AGE) related emergency room (ER) visits in children aged ≤5 years in Belgium. METHODS: A retrospective observational study was performed across 11 Belgian participating centres. Each centre provided an anonymised list of all ER visits of children aged ≤5 years during the periods June 2004–May 2006 (pre-vaccine period) and June 2007– May 2009 (post-vaccine). a sample of 7620 ER visits (3699 pre-vaccine; 3921 postvaccine) was randomly selected from the 180,253 ER visits reported. Patient demographics and information on the reason for ER visit were collected from all randomly selected visits. If the visit was AGE-driven, additional data were collected on vaccination status, symptoms, tests performed, treatment, and visit outcome. RESULTS: The proportion of AGE-driven ER visits was 11% in the 2-year pre-vaccine period (N AGE-driven visits/Total N ER visits = 412/3,699). After insignificant change to 10.5% during the 1st year after vaccine introduction (N = 208/1,977), this proportion significantly declined to 8.2% (26.6% reduction; p-value < 0.001) in the 2nd year (N = 159/1,944). a strong seasonal effect exists in the proportion of AGE-driven ER visits on the total number of ER visits, peaking in February and March. The number of patients who needed to be hospitalised because of AGE declined over time (from a yearly average of 94 pre-vaccine to 62 and then 44 in the 2 consecutive years post vaccine introduction). Oral and IV rehydration are the most commonly used treatments. Their use also declined after vaccine introduction. CONCLUSIONS: A significant decline in AGE-driven ER visits and subsequent hospitalizations was observed in Belgium after introduction of a universal mass vaccination program against rotavirus. AGE cases appeared less severe. A376 PIH7 BODY COMPOSITION CHANGES IN THE PREOPERATIVE PERIOD AND AFTER SURGERY Hock M, Domány B, Garai J, Kránicz J, Kriszbacher I, Boncz I, Bódis J University of Pécs, Pécs, Hungary OBJECTIVES: Our aim was to examine the changes of body composition in the perioperative period. METHODS: The authors examine the body composition of 11 postmenopausal women (Total Body Water (TBW), Extracellular Water (EW), Intracellular Water (IW), Fat Mass (FM), Fat Free Mass (FFM)) 1month and 1 day before as well as 1day and 1 month after the operation. The performed operations were anterior colporrhaphy with or without posterior colporrhaphy or enterocele repair. Preoperative medical preparation ordered the use of oral (1 mg estriol/die; T. Ovestin) and local (U. Ovestin; 1 mg/g estriol) estrogen for a month. Fluid replacement was adequate as ordered for patients with first-degree severity (average operative load, prohibition of fluid intake is not longer than 6–12 hours, maximum blood loss is 250–500 ml) during the operation and in the postoperative period. Body composition parameters were measured by MF HUMAN—IM SCAN Master 1.0. The statistical data were calculated according to mean, standard deviation, median, range, Fisher’s exact test, Student’s t-test methods and the results were considered to be significant at P < 0,05. RESULTS: The mean age of the subjects was 59.4 years (8.58 SD, median 57 years, range 48–72 years), the average number of preceding pregnancies was 3.33 (1.68 SD) with 2,46 deliveries (1.46 SD). In the preoperative period (between the 1st and the 2nd measurements) a significant decrease (p = 0,056) was found in EW. In comparison of the 2nd and the 3rd measurements a significant decrease (ECW p = 0,027; TBW p = 0,04; FM p = 0,019; FFM p = 0,019) was observed except for the parameters of ICW (p = 0,15). Between the 1st and 30th day after the surgery changes in findings were not significant (ECW p = 0,81; TBW p = 0,83; FM p = 0,5; FFM p = 0,3; ICW p = 0,616). CONCLUSIONS: The results suggest that body composition changes in the perioperative period should be followed. PIH8 CHANGES IN PELVIC FLOOR MUSCLE FUNCTION DURING THE PERIOPERATIVE PERIOD OF COLPORRHAPHY Hock M, Domány B, Rácz S, Garai J, Kránicz J, Kriszbacher I, Boncz I, Bódis J University of Pécs, Pécs, Hungary OBJECTIVES: Our goal was to follow up the changes of pelvic muscle function during the perioperative period. METHODS: Fifteen women in postmenopausal period were included in the study. The surgical management included anterior colporrhaphy with or without posterior colporraphy or enterocele repair. Preoperative medical preparation ordered the use of oral (1 mg estriol/die; T. Ovestin) and local (U. Ovestin; 1 mg/g estriol) estrogen for a month. The mean age of the subjects was 59.4 years (8.58 SD, median 57 years, ranging from 48 to 72 years), the average number of preceding pregnancies was 3.33 (1.68 SD) with 2,46 deliveries (1.46 SD). The pelvic muscle function was measured 1 month before (at the beginning of the preoperative preparation), 1 day before and 1 month after the operation. The following factors were measured: pelvic floor muscle activity, ability to relax and differences between left and right sides. The measurement of muscle parameters was performed by intravaginal surface EMG equipment. The statistical data were calculated according to mean, standard deviation, median, range, Fisher’s exact test, Student’s t-test methods and the results were considered to be significant at P < 0.05. RESULTS: In the preoperative period (from the 1st to the 2nd measurement) a significant recovery (p = 0.03) was found in the ability to relax pelvic floor muscle. In comparison of the 2nd and the 3rd measurement a significant decrease occurred in the average muscle activity (p = 0.054) and on the left (p = 0.034) side. The difference in the results of the 1st and the 3rd measurement was even more significant (p = 0.005), there was a 20.8% average muscle activity decrease during this period. CONCLUSIONS: The results suggest to follow up the function of pelvic floor muscles in the perioperative period of colporrhaphy, and if it is necessary to begin the proper physiotherapeutic treatment. INDIVIDUAL’S HEALTH – Cost Studies PIH9 COST COMPARISON OF MIDWIFE-LED VERSUS OTHER MODELS OF CARE FOR CHILDBEARING WOMEN FROM THE PRIVATE PAYER PERSPECTIVE IN BRAZIL Takemoto ML, Fernandes RA, Passos RB, Cukier FN, Tolentino ACM ANOVA—Knowledge Translation, Rio de Janeiro, Rio de Janeiro, Brazil OBJECTIVES: Efficacy of midwife-led models of care (MLMC) has been already assessed in a Cochrane systematic review (Hatem 2008) but there is a lack of cost comparisons in Brazil. This study aims to compare costs and consequences of MLMC versus other models of care (OMC) for women with low or mixed risk of complications from the private payer perspective. METHODS: Efficacy data was obtained from Hatem 2008 which compared MLMC with OMC including models centered in obstetricians, family doctors or shared models. As major maternal and neonatal events had not shown statistically significant differences between groups, efficacy of both models was considered similar. Significant differences in favor of MLMC (analgesia, episiotomy, instrumental birth, etc.) were used only for cost estimation. Resource use was obtained through local published data and input from clinical experts. Only direct medical costs related to the delivery room setting were considered. Unit costs were obtained from Brazilian official sources. RESULTS: Risk Ratios for regional analgesia 13th Euro Abstracts (RR = 0.81) and episiotomy (RR = 0.82) were selected as important for the adopted setting. The current episiotomy rates in Brazil varied from 76.20% to 94.50% due to differences between studies and regional analgesia rate was 99.40% in the only identified study. The incremental estimated costs for one additional episiotomy and analgesia were 42.82 BRL and 359.95 BRL, respectively. For each 1000 deliveries in the private setting, 137 to 170 episiotomies and 179 analgesias would be prevented with the adoption of MLMC, resulting in cost savings of 70,274 to 71,685 BRL. CONCLUSIONS: MLMC has shown similar efficacy and safety profiles when compared with OMC and our findings indicated potential cost savings with MLMC in the Brazilian private system. Further researches including other clinical outcomes, longer follow-up and complications as results of medical interventions could result in higher savings for the private payer. PIH10 TREATMENT AND COST FOR PATIENTS WITH ENDOMETRIOSIS IN THE UNITED KINGDOM Wasiak R1, Manson S1, Ryan J2, Petri JC3 1 United BioSource Corporation, London, UK; 2Bayer Schering Pharma AG, Newbury, UK; 3 Bayer Schering Pharma AG, Berlin, Germany OBJECTIVES: Endometriosis is defined by the growth of endometrial tissue outside the uterine cavity. It is a common and debilitating benign gynecological condition affecting mainly reproductive-age women with prevalence estimated between 7%–10% of women. The purpose of this study was to understand the cost associated with endometriosis in a secondary care setting in the UK. METHODS: A retrospective analysis was performed using secondary care data from the Cardiff and Vale National Health Service Hospitals Trust including 411 patients diagnosed with endometriosis. a supplementary postal survey was completed by a subset of 80 patients. Data included information on hospital stays, surgical procedures, prescriptions, as well as survey data on primary care treatment, quality of life and productivity loss. Patients were classified into surgical or clinical diagnosis groups depending on type of diagnosis: laparoscopy/laparotomy, or clinician’s assessment only. RESULTS: For hospital visits with endometriosis as primary diagnosis, patients with a surgical diagnosis (n = 89) experienced an average hospital stay duration of 0.4 days (SD = 0.7) compared to 2.2 days (SD = 3.4) for patients with a clinical diagnosis (n = 322). The average cost per hospital visit was £871 (surgical group) and £1525 (clinical group). Patients with clinical diagnosis also had more GP visits in the 6 weeks prior to their hospital visit (mean 2.0 (SD = 2.9) vs. 1.4 (SD = 1.4)). The most common endometriosis-specific pharmacotherapy used in a secondary setting was GnRH agonists (22%) followed by progestins (11%) and oral contraceptives (10%) with drug use in the surgical group exceeding that of the clinical group. Patient utility as measured by the EQ-5D was similar for surgically and clinically diagnosed patients (0.70, SD = 0.32 vs. 0.71, SD = 0.27). CONCLUSIONS: This study demonstrates that there are substantial costs involved in treating patients in a secondary care setting, with a different pattern of endometriosis-related costs for patients without a confirmed diagnosis of endometriosis. PIH11 SOCIETAL IMPACT OF DIARRHEA IN YOUNG CHILDREN IN THE NETHERLANDS Wolleswinkel J1, Speets A1, Hovels AM2 1 Pallas, Health Research and Consultancy, Rotterdam, The Netherlands; 2GSK, Zeist, The Netherlands OBJECTIVES: Diarrhoea is one of the most common childhood diseases. This study examined the medical resources use and societal impact related to diarrhoea in young children in the The Netherlands. METHODS: In 2007 an Internet survey questionnaire on childhood diseases was administered to a representative sample of parents with children aged <5 years. The survey covered socio-demographic data, characteristics of the most recent disease episode, medical resources use, productivity loss by the caregivers, and travel-related costs. Data were available for 2425 respondents. This study focused on diarrhoea episodes. Diarrhoea was defined as symptoms and/ or diagnosis of diarrhoea only or diarrhoea in combination with symptoms of fever and/or vomiting; i.e. diarrhoea without concomitant diseases. RESULTS: In total 128 episodes of diarrhoea were reported. In 27% of these episodes a general practitioner had been consulted, in 2% a paediatrician, and in 4% an emergency department had been visited. In 59% of the diarrhoea episodes no medical doctor was consulted. Hospital admission occurred in 3% of the cases (mean duration 2 days). Over-thecounter drugs were bought in 41% of the episodes. In 17% of the diarrhoea episodes caregivers lost days from a paid job (mean 16.4 hours per episode; SD 15.8). In 19% of the episodes parents reported productivity loss at work during their child’s illness (mean 6.3 hours per episode; SD 5.9). Leisure time loss was reported in 29% of the episodes (mean 2.3 hours per episode; SD 2.5). The mean costs were estimated to be c208,- per diarrhoea episode, of which 23.3% direct medical costs, 10.3% direct non-medical, and 66.4% indirect non-medical costs. CONCLUSIONS: The medical and economic burden of diarrhoea is considerable to individual families as well as to society in the The Netherlands. This study has filled a gap in the knowledge base on the impact of diarrhoea on society. 13th Euro Abstracts PIH12 GOLD STAGE AND DURATION OF HOSPITAL ADMISSION DETERMINES SIZE AND STRUCTURE OF COPD RELATED DIRECT MEDICAL COSTS IN ELDERLY Jakovljevic M1, Lazic Z2, Jankovic S1, Verhaeghe N3, Annemans L4 1 The Medical Faculty University of Kragujevac, Kragujevac, Serbia; 2Clinical Center Kragujevac, Kragujevac, Serbia; 3Ghent University, Ghent, Belgium; 4Ghent University—Brussels University, Ghent, Belgium OBJECTIVES: COPD cost of illness assessment and its dependence on GOLD stage or duration of hospital admission. METHODS: The sample consisted of 186 elderly COPD patients. Chest physicians conducted follow-up and financial data were obtained from administrative registry. Time horizon was one year and Health Insurance Fund perspective was selected. Included clinical endpoints were disease exacerbation and hospital admission. Economic data included all inpatient COPD-related medical goods and services consumption, and outpatient drug utilization. We excluded medical costs which did not arise from COPD. RESULTS: An average elderly COPD patient imposed c1745.41 of costs annually to the national health care budget. Severity grade (GOLD stage) and duration of all hospital admissions were significantly and directly correlated with overall cost. Expenses structure per year was: c14.54 outpatient care, c126.97 inpatient drug consumption, c377.98 hospital admission (Intensive Care Unit admissions and specialist consultations included in daily price), c133.5 imaging diagnostics, c111.43 laboratory analysis, c26.67 therapeutic interventions, c70.32 consumables, c884 outpatient drugs consumption. Most expenses originating from drug acquisition were due to antibiotics prescribed for curing infections and mucolitics and antihistamine drugs consumption. Average number of hospitalizations per person was 1.45 with a duration of 12.84 days. Fifteen persons deceased. Average number of outpatient visits was 2.71 per person. CONCLUSIONS: Overall burden of COPD is mostly driven by outpatient drug consumption and exacerbations leading to hospital admissions. Relative relevance of drug acquisition expenses in our country is higher than in high-income societies, because of still significantly lower human labour wages in the area. Pattern of diagnostic procedures requested and ATC drug classes consumed remains similar and comparable in most countries. More in-depth research of indirect COPD attributable costs e.g. lost productivity, absentism, premature death etc, will be needed in the future. PIH13 PATTERNS AND COSTS OF HOSPITALIZATIONS IN ELDERLY PATIENTS IN RIO DE JANEIRO, BRAZIL Takemoto ML, Fernandes RA, Passos RB, Tolentino ACM, Cukier FN, Takemoto MM, Moretti AIP ANOVA—Knowledge Translation, Rio de Janeiro, RJ, Brazil OBJECTIVES: Ageing population implies greater demand of hospital services and consequent health care costs. This study aims to describe patterns of hospitalization and estimate hospitalization costs for aged patients in public hospitals in the city of Rio de Janeiro, Brazil. METHODS: Retrospective analysis of Rio de Janeiro hospital admissions for elderly people (aged ≥60 years) as reported in Brazilian Hospital Information System (SIH/DATASUS) database from January 1st to December 31st 2007. The costs informed in the database represent federal reimbursement values for hospitalizations placed in public hospitals in the city. Admissions were categorized according to ICD-10 groups and age categories. RESULTS: 287,972 hospital admissions were identified for all ages in the city in 2007 and 24.3% were for patients ≥60 years (70,077). The five more common reasons for hospitalization were cardiovascular diseases (22.4%), cancer (17.5%), mental and cognitive disorders (11.2%), gastrointestinal (9.3%) and respiratory diseases (7.6%). The in-hospital mortality rate was 14.76% for all causes. When disease groups were analyzed, infectious and respiratory diseases had the higher mortality rates (38.6% and 38.3%, respectively). The mean length of stay was 14.3 days and the average cost for elderly patients hospitalizations was 960,62 BRL. The total amount paid for hospitalization of elderly patients in 2007 was 67,317,348 BRL, 30.7% of all hospitalization costs. CONCLUSIONS: The findings showed the significant contribution of the elderly to public hospital expenditures with leading causes of admission including cardiovascular diseases, cancer and mental and cognitive disorders. Additionally, the higher in-hospital mortality rate was seen for preventable diseases. These results reinforce the need of prevention strategies among the elder population for targeted conditions. PIH14 COST-BENEFIT OF FLUCONAZOLE FOR VAGINAL CANDIDIASIS 1 2 1 Sura MV , Avksentieva MV , Gostishev R 1 Institute of Clinico-Economic Expertise and Pharmacoeconomics, RSMU, Moscow, Russia; 2 Research Center for Clinical and Economic Evaluation and Pharmacoeconomics, Moscow, Russia OBJECTIVES: To perform cost-benefit analysis of fluconazole for VC in Russia. METHODS: Willingness-to-pay (WTP) for treatment of vaginal candidiasis was assessed. 240 women took part in the survey. a special questionnaire was designed to reveal women’s preferable methods of antifungal drug application and WTP for fluconazole in comparison with other drugs. Women were proposed: 1) to chose the preferable method of drug application; 2) to chose the preferable drug among 4 options with similar efficacy but different methods of application and price (no concrete drug names were mentioned, only prices and costs of treatment and application methods); and 3) to give the maximal affordable price for treating vaginal candidiasis. RESULTS: More than 60% of women preferred oral treatment once with the mean A377 costs of therapy 350 rubles ($11) among 4 proposed alternatives. The median WTP for antifungal therapy was 500 rubles ($16) (range from 50 to 5000 rubles, $1.6– 161.0) that is less than real cost of treatment with original fluconazole preparation. The cost-benefit ratio (WTP/cost of treatment) was 1.30 (range 0.09–20.6). CONCLUSIONS: Fluconazole is a cost-benefit alternative for VC treatment. PIH15 ECONOMIC EVALUATION OF DARBEPOETIN ALFA (ARANESP) COMPARED TO EPOETIN ALFA (ERYPO) AND EPOETIN BETA (NEORECORMON) IN THE TREATMENT OF CHEMOTHERAPY-INDUCED ANEMIA (CIA) IN AUSTRIA Walter E1, Ribnicsek E2, Kutikova L3 1 Institut fuer Pharmaoekonomische Forschung, Vienna, Austria; 2Amgen, Vienna, Austria; 3 Amgen (Europe) GmbH, Zug, Switzerland OBJECTIVES: Anemia is a common side effect observed in patients receiving myelosuppressive chemotherapy [1]. The purpose of this pharmacoeconomic analysis was to evaluate the cost-effectiveness of the long-acting erythropoiesis-stimulating agent (ESA) darbepoetin alfa (DA) 500 mcg once every 3 weeks (Q3W) and 150 mcg weekly (QW), and short-acting ESAs epoetin-alfa (EA) 40,000 IU QW, epoetin-beta (EB) 30,000 IU QW and 3-times weekly (TIW) for the treatment of CIA. METHODS: A cost-consequence model was constructed using a decision-analysis tree. The treatment period considered was based on 12 weeks and was aligned with routine chemotherapy regimen administration. Model inputs included: medical treatment, outcomes, and health care service utilization from published clinical trials and summary of product characteristics recommendation. Effectiveness of therapeutic alternatives was determined by comparing hemoglobin response rates. Costs included direct medical costs(intervention drug, inpatient and outpatient) and transportation costs. Costs presented reflect 2010 data. The analysis was performed from the perspective of the Austrian health care system. RESULTS: The average expected direct costs per patient were c3,675 for DA Q3W, c3830 for DA QW, c4290 for EA QW, c4240 for EB QW and c4745 for EB TIW. Cost-savings associated with DA Q3W were 4% relative to DA QW, 14% to EA QW, 13% to EB QW and 23% to EB TIW. The cost per hemoglobin response rate (therapeutic success) amounted to c5035 for DA Q3W, c5247 for DA QW, c6309 for EA QW, c6235 for EB QW and c6977 for EB TIW. CONCLUSIONS: In the treatment of CIA among cancer patients in Austria, darbepoetin alfa Q3W and QW are projected to provide more efficient use of health care resources compared to alternative treatment strategies such as epoetin-alfa and epoetin-beta. [1] Ludwig et al. Eur J Cancer 2004;40:2293–306. PIH16 SURGICAL REPAIR OF ANTERIOR VAGINAL WALL PROLAPSE USING PROLIFT AND COLPORRHAPHY. A COST-EFFECTIVENESS ANALYSIS FOR A THIRD LEVEL HOSPITAL IN MEXICO Nasciben V1, Zanela OO2, Cabra HA2, Rodriguez S2 1 Johnson & Johnson, Sao Paulo-SP, Brazil; 2Johnson & Johnson Medical, Mexico City, D.F., Mexico OBJECTIVES: Estimate the clinical and economic outcomes of the surgical repair of anterior vaginal wall prolapse using Prolift® vs. colporrhaphy to assess its costeffectiveness under a third level hospital perspective in Mexico. METHODS: A multistate Markov model was developed to assess the evolution of a patient with anterior vaginal wall prolapse under a conventional surgical repair strategy (colporrhaphy) and a mesh grant (Prolift®) after 2, 4, 10 and 20 years. a specialist panel was conducted to collect current clinical practice, resource utilization and complication data. After the first surgical prolapse repair, the base-case patient could face two additional surgeries if prolapse relapsed. Two scenarios were modeled to compare the clinical and economic impact of Prolift® as second line treatment; the mesh grant was the assumed third line treatment for both scenarios. The effectiveness variable was the objective recurrence (failure) rate. Clinical data, transition probabilities and mortality rates were taken from published sources. Only direct medical costs were considered, and cost data was retrieved from IMSS official cost list. The price of Prolift® was internally estimated. Results are expressed as 2010 inflation-adjusted Mexican pesos (MXN) and MXN/QALY for ICERs. a 4.5% discount rate was used for costs and outcomes. RESULTS: Total costs for year 1 were higher for the Prolift® strategy ($75,688 vs. $74,544), but yielded an overall 22% complication reduction. For year 2, total costs were lower for Prolift® ($81,675 vs. $104,007), yielding better outcomes (0.04 incremental QALYs), thus being a dominant strategy (−$528,682/QALY). For the following years, Prolift® showed to be a consistent cost-saving alternative vs. colporrhaphy: ICERs for years 4, 10 and 20 were −$273,598/QALY, −$29,942/QALY and −$1,212/QALY. CONCLUSIONS: Findings suggest Prolift® is a safe and costeffective alternative to treat anterior vaginal wall prolapse when compared to colporrhaphy. Savings are due to lower failure and complication rates. PIH17 COST-EFFECTIVENESS OF SEASONAL INFLUENZA VACCINATION IN PREGNANCY IS DEPENDENT ON VACCINATION EARLY IN FLU SEASON Myers ER1, Misurski DA2, Swamy GK1 1 Duke University, Durham, NC, USA; 2GlaxoSmithKline, Philadelphia, PA, USA OBJECTIVES: To estimate the impact of vaccination timing on the cost-effectiveness of seasonal influenza vaccination in pregnancy for prevention of disease in women and infants under 6 months of age. METHODS: We constructed an open cohort A378 Markov simulation of US pregnant women and infants under 6 months of age for 1 calendar year, from July 1 through June 30. National statistics on live births and infant mortality were used to generate probabilities of delivery and infant death based on gestational age. Each week, women entered the cohort with new pregnancies, and left by delivery; infants entered by delivery and left by death or reaching 6 months of age. Annual influenza-attributable rates of outpatient visits and hospitalizations for pregnant women and infants were obtained from the literature, and adjusted to weekly incidences using CDC data on influenza-positive respiratory isolates. In the base case, we assumed vaccination would begin in calendar week 40, be administered at routine prenatal visits across all gestational ages, and continue throughout the rest of the year, with vaccine efficacy of 73% for prevention of maternal flu and 63% for prevention of flu in infants under 6 months. Costs and maternal utilities were obtained from previously published analyses; we used published parental utilities for relevant infant health states. RESULTS: Base-case cost-effectiveness was $65,112/QALY gained. Fifty-six percent of the reduction in morbidity was attributable to prevention of disease in infants, and all of this benefit was accrued by vaccination within the first 4 weeks of vaccine availability. CONCLUSIONS: Vaccination of pregnant women against seasonal influenza is cost-effective, with much of the benefit derived from prevention of infant disease. Vaccination is most efficient when the majority of pregnant women are vaccinated within 4 weeks of vaccine availability. PIH18 COST-EFFECTIVENESS OF COMBINATION THERAPY WITH DUTASTERIDE AND TAMSULOSIN FOR THE TREATMENT OF MODERATE TO SEVERE BENIGN PROSTATIC HYPERPLASIA IN SPAIN Antoñanzas F1, Cozar JM2, Brenes F3, Molero JM4, Fernández-Pro A5, Palencia R6, Martín I6, Huerta A6 1 Universidad de la Rioja, Logroño, La Rioja, Spain; 2Hospital Universitario Virgen de las Nieves, Granada, Spain;, 3Centro de Atención Primaria Llefià (ABS6), Badalona, Spain; 4 Centro de Salud San Andrés, Madrid, Spain; 5Centro de Salud de Menasalbas, Menasalbas, Toledo, Spain; 6GlaxoSmithKline, Tres Cantos, Madrid, Spain OBJECTIVES: To evaluate the cost-effectiveness of the combination therapy with dutasteride and tamsulosin (D+T) as initiation treatment versus tamsulosin (T), considered the standard of care, in the treatment of moderate to severe benign prostatic hyperplasia (BPH) in Spain. METHODS: A semi-Markov model was developed using 4-year and 35-year time horizons. Data were obtained from the CombAT and literature review, and the analysis was made from the National Healthcare Service perspective. Effectiveness was measured in terms of reduction of acute urinary retention (AUR) events and/or BPH-related surgeries and quality adjusted life-years (QALYs). Health care resources were defined by an experts’ panel, and unitary costs were obtained from published Spanish sources and expressed in 2010 Euros. The model calculates costs, AURs and/or surgeries, and QALYs related to each therapy. Costs and effectiveness outcomes were discounted at 3.0%. One-way and probabilistic sensitivity analyses were conducted to test the robustness of the model. RESULTS: Combination therapy with D+T improves patients’ outcomes. At 4 years, D+T patients have a 9.9% reduction in AURs and/or surgeries over T, reaching 43.9% at 35 years. At 4 and 35 years, total costs related to T treatment add up to 1,373.90c and 5,187.37c, and total costs related to D+T are 2,184.43c and 8,630.99c, respectively. Therefore, at 4 years, treatment with D+T presents an additional cost of 81.66c per AUR and/or surgery avoided and 14,023.32c per QALY gained compared to T. At 35 years, results were 79.10c per AUR and/or surgery avoided and 8,750.15c per QALY gained. Sensitivity analyses showed that results are robust. CONCLUSIONS: Given the assumptions, combination treatment with D+T not only represents a more effective alternative versus T due to the reduction in AURs and/or surgeries, but also is a cost-effective treatment in patients with moderate to severe BPH in Spain. PIH19 TRENDS IN COST-EFFECTIVENESS STUDIES OF HIGH BUDGET IMPACT DRUGS Aggarwal S1, Stevens CA2 1 PAREXEL Consulting, Bethesda, MD, USA; 2PAREXEL Consulting, Waltham, MA, USA OBJECTIVES: The recently made coverage decisions by UK’s NICE, Scotland’s SMC and the allocation of $1.1Billion for comparative effectiveness research by the United States, are strong indicators of trends in pricing and reimbursement that are likely to be observed in the future. To gain an additional insight into these trends, we analyzed the cost-effectiveness studies for the top 10 highest selling drugs (~$80–95B worldwide sales). METHODS: The Top 10 drugs were selected based on their worldwide sales. For this analysis, we segmented these drugs into categories as primary care, specialty, small molecules, biologics, therapy areas and availability of generic alternatives. We analyzed the cost-effectiveness studies that were published in peer-reviewed journals. Search was conducted using generic names of the drugs and the phrase “cost-effectiveness” in abstract of the published study. RESULTS: During 2003–2008, the number of published studies on “cost-effectiveness” have increased by more than 30%. Almost half of the published studies belong to—Remicade, Plavix and Enbrel. There is a large variability in CERs for same drugs for different indications, in some cases also varying by biomarkers. Primary care drugs had lower and less variable CERs than specialty drugs. Variations also exist in methodology used by different groups in modeling cost-effectiveness, especially for time horizon and comparator. Majority of primary care drugs were modeled for a time horizon of 35–40 years or lifetime to demonstrate cost-effectiveness. Among the top 10 drugs, quetiapine and erythropoietin had the highest variability across different studies, and atorvatstatin, salmetrol/ 13th Euro Abstracts fluticasone and clopidogrel had the most consistent ICER values across studies. CONCLUSIONS: This analysis shows the range, variability and methods used for calculation of ICER values for these high budget impact drugs and provides lessons for executives and policy makers. PIH20 EXPLORATORY COST-EFFECTIVENESS ANALYSIS OF THE ANTERIOR REPAIR OF THE PELVIC ORGAN PROLAPSE COMPARING ANTERIOR COLPORRHAPHY (CONVETIONAL SURGERY) VERSUS PROLIFT® MESH, UNDER THE BRAZILIAN PRIVATE PAYER PERSPECTIVE Nasciben V Johnson & Johnson, Sao Paulo-SP, Brazil OBJECTIVES: To assess the cost-effectiveness of the anterior repair of the Pelvic organ prolapse (POP) with Prolift® versus the colporrhaphy (COLP), under private payer perspective, in Brazil. METHODS: A multi-state Markov model was developed to assess the evolution of a patient with anterior vaginal prolapse after COLP and Prolift® after 2, 4, 10 and 20 years. a panel of specialist was conducted to obtain the local practice and to collect the complication rates. Only direct medical costs were considered (SIMPRO, 2010; CBHPM 5th Ed). Clinical data, transition probabilities and mortality rates were taken from published sources (IBGE, 2008; Jia, X 2007). The base-case patient could face up to three surgical repairs if prolapse relapsed. Two scenarios were modeled to compare the clinical and economic impact of Prolift® as second line treatment; mesh grant was the assumed third line treatment for both scenarios. Discount rate of 5% for costs and outcomes was taken following the Brazilian HTA guidelines (Vianna, 2007). One-way sensitivity analysis was conducted to assess the robustness of the results. RESULTS: The total costs for the first year were higher for Prolift® (R$8119 versus R$4777; incremental R$3342). For the 2nd year the difference in the total costs reduced (R$8659 vs. R$7435; additional R$1223) with better outcomes for Prolift® (incremental QALYs: 0.04) with ICER ranging from R$19,099 (non discounted) to R$33,207 (discounted). For the next years the ICERs reduced, showing long term benefits of the adoption of Prolift®. For the fourth year ICER ranging from R$419 to R$532, 10th year from R$578 to R$827 and for the 20th year ranging from R$832 to R$1291, for non discounted and discounted results. CONCLUSIONS: Findings suggest Prolift® as a cost-effective intervention, under the Brazilian private payer perspective. PIH21 IS ROUTINE IMMUNIZATION OF ELDERLY WITH THE 13-VALENT PNEUMOCOCCAL CONJUGATE VACCINE LIKELY TO BE CONSIDERED AS COST-EFFECTIVE? Rozenbaum MH1, Hak E2, van der Werf TS3, Postma MJ2 1 Rijksuniversiteit Groningen, Groningen, The Netherlands; 2University of Groningen, Groningen, The Netherlands; 3University Medical Centre Groningen (UMCG), Groningen, The Netherlands OBJECTIVES: To estimate the cost-effectiveness in relation to the efficacy of PCV-13 among elderly (both the total population and those at increased risk) aged 65 years and older for the The Netherlands, for various levels of efficacy assumed. METHODS: Our previously published cost-effectiveness model was updated to include, age-specific epidemiological data and health care utilization and costs for a hypothetical cohort of elderly aged over 65 years of the population of the The Netherlands. This cohort was followed twice- once as unvaccinated and once as a vaccinated cohort- over a time period of 5 years. Outcome measures included costs, life-years (LYs), qualityadjusted life-years (QALYs) and cost-effectiveness ratios (CERs). All analyses were performed from a societal perspective. RESULTS: Vaccination remained well below the c80,000 per LY except if the vaccine was only assumed to be protective against bacteraemic pneumonia with a relatively low effectiveness (40%) in combination with a high vaccine price (c65) and indirect effects of serotype replacement would largely offset the direct effect of vaccination. For various assumptions, introduction PCV-13 (assuming a 60% efficacy against invasive and non invasive disease due to vaccine serotypes, and a cost of c50 per vaccinated person) the incremental cost-effectiveness ratio varies over from cost-saving to c50,676 per LY. More probable scenarios generated cost-effectiveness ratios which would be labeled as cost-effective. CONCLUSIONS: In the The Netherlands, vaccination with PCV-13 is likely to be considered cost-effective both for the total and for the high-risk population over 65 years of age from a societal perspective over a five-year time horizon. The main limitation of this study was the uncertainty regarding the share of pneumococcal related pneumonia. PIH22 EVALUATING THE COST-EFFECTIVENESS OF CERVICAL CANCER SCREENING AND HUMAN PAPILLOMAVIRUS VACCINATION STRATEGIES USING A MATHEMATICAL MODEL Taylor DC1, Pawar V1, Gilmore K1, Sanon M1, Kruzikas D2, Kohli M3, Arondekar B4, Demarteau N5, Weinstein M6 1 i3 Innovus, Medford, MA, USA; 2Lovelace Respiratory Research Institute, Kannapolis, NC, USA;, 3i3 Innovus, Burlington, ON, Canada; 4GlaxoSmithKline, Philadelphia, PA, USA; 5 GSKbio, Wavre, Belgium; 6Harvard School of Public Health, Boston, MA, USA OBJECTIVES: To assess the cost-effectiveness of various cervical screening strategies in a cohort of 12-year-old US women with (V) and without (NV) a human papillomavirus (HPV) 16/18 vaccine with efficacy against vaccine and non-vaccine oncogenic HPV types. METHODS: A lifetime Markov model simulating the progression of HPV infection and subsequent cervical disease (cervical intraepithelial neoplasia (CIN), A379 13th Euro Abstracts cervical cancer, and death) was developed. Thirty cervical screening strategies were investigated, including the American College of Obstetricians and Gynecologists’ recommendation: screening women every 2 years starting at age 21 years, increasing to 3 years following three consecutive negative tests after age 29 (denoted “21/2–3”). Other strategies varied initial screening age (16, 21, or 25 years) and screening frequency before and after age 30 years (every 1, 2, 3, 4, or 5 years). Vaccination was assumed to provide 95% efficacy against CIN2+ due to types 16/18 and 37.4% efficacy against CIN2+ excluding co-infections due to HPV types 16/18 (extrapolated to efficacy against infection). The primary outcome was incremental cost-effectiveness ratio (ICER) per quality-adjusted life-year (QALY) gained. Full screening compliance and vaccination coverage were assumed; all outcomes discounted at 3% per year. RESULTS: The 25/5–5NV strategy (every 5 years starting at age 25 years, without vaccination) was least costly. No other screening scenario without vaccination was cost-effective. The 25/5–5V strategy was cost-effective compared with the 25/5–5NV strategy ($9,000/QALY). Other non-dominated strategies (25/3–4V, 25/2–3V, and 25/1–2V) produced ICERs above $100,000 per QALY. CONCLUSIONS: Model results suggest initiating screening at age 25 with a 5-year frequency in the presence of HPV vaccination is cost-effective compared with increasing screening frequency without vaccination or lowering the age of screening initiation. Increasing HPV vaccination in 12-year-old women would allow for less frequent screening initiated at older ages, constituting an efficient use of health care resources. PIH23 COST-EFFECTIVENESS ANALYSIS OF 5-ALPHA REDUCTASE INHIBITORS IN PATIENTS WITH BPH AND AN ELEVATED PSA IN SWEDEN Sobocki P1, Björkstedt K1, Svedman C2 1 GlaxoSmithKline, Solna, Sweden; 2Karolinska Institute, Stockholm, Sweden OBJECTIVES: Dutasteride and finasteride are indicated for treatment of BPH and both agents have demonstrated efficacy as risk reduction (RR) of PCa. The present work assesses the cost-effectiveness of the combined effect of each drug on BPH progression and RR of PCa, in the treatment of BPH patients with a risk of developing prostate cancer (PSA ≥ 3). METHODS: A previously validated Markov model was populated to simulate the cost-effectiveness of treatment with dutasteride or finasteride in men with BPH and a PSA-level ≥3. Equal effect was assumed of the two drugs on BPH progression, and an indirect comparison was performed to compare RR effect on PCa for dutasteride (the REDUCE trial) and finasteride (the PCPT trial). Due to different study design and populations in the two trials, an adjustment was made to the REDUCE trial results to match the PCPT trial design. Data on incidence, costs and quality-of-life for BPH and PCa were retrieved from registries, publications and expert opinions. The outcome is presented as incremental cost per QALY and results were presented in c2009. Extensive sensitivity analyses were performed to present the uncertainties in the results. RESULTS: Estimated RRR of clinically detectable PCa in REDUCE was 33.3%–73%, depending risk level chosen as trigger for detection of PCa, compared to 19% for finasteride. Incremental cost per QALY gained was estimated at c28,970 with the most conservative risk reduction of 33%. Sensitivity analysis gave a range from c4,940 to c54,200 per QALY. Results were most sensitive to changes in level of risk-reducing benefit of PCa and quality-of-life weights attributed to the diagnosis of PCa. CONCLUSIONS: Use of dutasteride to reduce the risk of PCa and BPH progress is expected to be cost-effective compared to finasteride when used in 60-year-old men with BPH and PSA ≥ 3. PIH24 INDIRECT COSTS OF PREGNANCY, PARTURITION AND PUERPERIUM IN POLAND Szkolut S1, Jakubczyk M1, Hermanowski T2 1 Medical University of Warsaw, Warsaw, Poland; 2Department of Pharmacoeconomics, Medical University of Warsaw, Warsaw, Poland OBJECTIVES: The aim was to estimate the indirect costs of absenteeism generated by women during pregnancy and in post-natal period. The secondary objective was to quantify the indirect costs in relation to the maternity benefits and examine the shares of absenteeism caused by pregnancy, childbirth, and puerperium in a total disease-related absenteeism in Poland. METHODS: The indirect costs of pregnancy were estimated using human capital approach. Social Insurance Office reports on the disease-related absenteeism in Poland in the years 2006–2008 were used as a data source. The sickness benefits (equal to wages) were used to estimate the indirect costs of monthly absence. Indirect costs were compared with the forecasted production value to be generated by a child in a life-time horizon (discounted with a rate of 5%, accounting for survival probabilities, changing wages and rate of activity rate). RESULTS: The pregnancy, parturition and puerperium related absenteeism represented in years 2006–2008 successively 15.1%, 15.8% and 16.7% of the total diseaserelated absenteeism in Poland. The cost of absenteeism during pregnancy, childbirth and puerperium in 2006 amounted to c242 million (1c = approx. 4.10 PLN), and in 2 successive years increased by an average of c66 million yearly. Indirect cost associated with maternity leaves in years 2006–2008 amounted to c267.7 million, c341.8 million, and c441.6 million respectively. The total indirect costs of pregnancy in years 2006–2008 achieved an average of c651.1 million. The indirect cost related with single pregnancy was estimated at c1650 representing only 3.5% of future estimated production value generated by a child (c47,083). CONCLUSIONS: However pregnancy is obviously not an illness, the indirect costs thereof can be calculated and should be used when evaluating the impact of family-friendly policies. The indirect costs are negligible comparing to future benefits. PIH25 THE IMPACT OF EMPLOYEES WORK ABSENTEEISM DUE TO CARE FOR A SICK PERSON ON PRODUCTIVITY IN POLAND Wrona W, Hermanowski T, Jakubczyk M, Czech M Department of Pharmacoeconomics, Medical University of Warsaw, Warsaw, Poland OBJECTIVES: The work absenteeism due to care for a sick person can be considered as indirect costs of illness. This study aimed at assessing the global burden of caring for the ill on productivity in Poland. METHODS: Data were obtained from a survey that incorporated information on a burden of caring for the sick and the WPAI-GH questionnaire during computer-assisted personal interview in a representative sample of the Polish general population aged more than 15 years. There were 2019 respondents, gathered in two waves in January and May 2010. Human capital approach was used with GDP per capita as a measure for the monetary value of a time unit (c4.2 per hour). RESULTS: The final population comprised 795 employees among whom 167 respondents (21%) took care for a sick person in the last 12 months—78 (10%), 74 (9%), and 19 (2%) responders took care for a child, other family member and non-family member, respectively. On average a respondent spent 9 hours a day (h/d) taking care of a sick person for an average period of 21 days (11 h/d for 14 d caring for a child; 7 h/d for 23 d caring for other family member; 13 h/d for 22 d caring for non-family member). These 9 hours were reduced working time (3,4 h) and off-work activities (5.6 h). Estimated productivity lost equals c303 on average per year for individuals looking after the ill and c11 on average per year for each employee. CONCLUSIONS: Caring for the sick, mostly family members, influences working time of Polish employees. Productivity loss of a single employee has a moderate impact on production, but in global scale or in specific health condition it might be perceived as an important factor modifying indirect costs of illness. PIH26 DEVELOPMENT OF A SEGMENTATION MODEL TO IDENTIFY HEALTH CARE DEMAND AND RELATED COSTS ATTRIBUTABLE TO POPULATION’S CHARACTERISTICS Madotto F1, Fornari C1, Riva MA1, Scalone L1, Ciampichini R2, Bonazzi MC1, Cesana G1 1 University of Milano—Bicocca, Monza (MB), Italy; 2Charta Foundation, Milan, Italy OBJECTIVES: The assessment of individuals’ demand and related costs can guide to allocate health care resources and improve access to health care. We created a segmentation demand model and we estimated health care costs and their determinants within specific and distinguished groups of general population. METHODS: The target population included subjects living in an area of northern Italy registered at one Local Healthcare Unit in 2005 (1,031,684 subjects). On the basis of clinical judgments and literature, we identified eight different segments: subjects unknown to HS, maternity and infancy, elderly, people with one chronic disease (CD), people with more CDs, people with probable or not severe CDs, subjects with acute event, healthy people. To describe these groups and their health demand, we used demographic and health care demand data (hospital admissions, drug’s prescriptions, medical specialist visits, diagnostic tests) from administrative databases available at the Lombardy Health System (HS). These were merged adopting probabilistic record linkage (DENALI Data Warehouse) to optimize correct matching of data. RESULTS: Overall, health care cost c834 million in 2005 (c809 per-capita). Healthy people were 53% of the sample and cost c180 per-capita. Next more frequent subjects were those with one CD (14%) and cost c916 per-capita, then those with more CDs (13%) who cost c3457 per-capita. Hospitalizations were the cost driver in 5 segments (maternity and infancy, elderly, one CD, more CDs, acute event), accounting for 42–89% of total costs. Diagnostic tests and medical visits contributed to 42–45% of total costs among healthy subjects and those with probable CD. Drugs accounted for 24% of costs of sample. CONCLUSIONS: This pioneer demand segmentation model shows an example of how merging different administrative databases makes possible understanding effects of their characteristics on health care demand and costs. Research is encouraged to improve model and study specific variants that can be applied in different health care contexts. INDIVIDUAL’S HEALTH – Patient-Reported Outcomes Studies PIH27 ADHERENCE TO PRESCRIPTION MEDICATIONS IS HIGHER WITH USE OF A GENERAL PRACTITIONER Goren A1, DiBonaventura M1, Gupta S2, Wagner JS1, Freedman D2 1 Kantar Health, New York, NY, USA; 2Kantar Health, Princeton, NJ, USA OBJECTIVES: Non-adherence to treatment regimens results in risks and health care costs due to increased emergency care. One reason for non-adherence, especially with multiple prescriptions, may be the lack of coordination of care among multiple specialists (Cutler & Everett, 2010). Assuming that general/family practitioners (GPs) help coordinate patient care, the current study examined the impact on adherence of visiting a GP. METHODS: Adherence data from the European 2008 National Health and Wellness Survey (a self-reported online survey sampling respondents from the UK, France, Germany, Italy, and Spain) were analyzed, with ANOVAs or negative binomial regressions (for resource utilization), as a function of seeing a GP (yes/no) crossed with number of prescriptions (1–2, 3–4, and 5+). Adherence, the main outcome, was a Morisky Score ranging from 0 (perfect adherence) to 4 (low adherence). Health care resource utilization measures included number of hospitalizations and visits to the emergency room (ER) in the past six months. RESULTS: There were 17,690 A380 respondents (of 53,524 total) using 1–2 (n = 6,434), 3–4 (5,529), or 5+ (5,727) prescription medications for symptomatic conditions. Adherence was higher among those who saw a GP in the past 6 months (M = 1.19, SD = 1.24, n = 14,802) vs. those who did not (M = 1.36, SD = 1.29, n = 2,888), P < 0.001, and higher among those with a higher number of prescriptions (for 5+, 3–4, and 1–2: Ms = 0.99, 1.26, and 1.38; SDs = 1.18, 1.26, and 1.27, respectively), P < 0.001. Hospitalizations were lower among GP visitors (M = 0.30, SD = 2.39) than non-visitors (M = 0.34, SD = 2.07), P < 0.001, ER visits were lower among GP visitors (M = 0.32, SD = 1.20) than nonvisitors (M = 0.33, SD = 2.09), P < 0.001, and both were significantly higher among those with more prescriptions, ps < 0.001. The benefits of a GP were largely independent of number of medications. CONCLUSIONS: Among European patients taking prescriptions, visiting a GP was associated with greater adherence and lower health care resource utilization. PIH28 WHY DO PATIENTS NOT ADHERE TO PRESCRIBED MEDICATION REGIMES? RESULTS OF TWO GERMAN SURVEYS Wilke T1, Mueller S2 1 Institute for Pharmacoeconomics and Drug Logistics, University of Wismar, Wismar, Germany; 2University of Wismar, Wismar, Germany OBJECTIVES: The aim of this study is to answer the following questions: 1) How high is the self-reported nonadherence (NA) of German patients with the need to regularly take medication? and 2) Which factors capable of explaining this self-reported NA can be identified by multivariate analysis? METHODS: Two cross-sectional surveys (phone survey with 1177 patients; face-to-face in-depth survey with 340 patients in 17 German pharmacies) were conducted. Self-reported NA was measured by the generic Morisky scale (either as 4 items or 8 item MMAS). Identification of explanatory factors was conducted on the basis of multivariate logistic regression analysis (including the calculation of additive risks by dichotomization of significant explanatory factors). RESULTS: 1) Approximately 35–40 % of the patients can be described as non-adherent (38.8 %/35.3 %); 2a) Survey 1: Only a few socio-demographic factors are able to explain the NA (chronic disease, some aspects of age, and low number of required medications to take); and 2b) Survey 2: Most results of the first survey can be replicated. However, intentional NA explanations have considerably more influence: positive medication belief, a positive mood, and a good patientdoctor relationship reduce the NA risk. Furthermore, patients who are easily able to recognize the correct medication on the basis of the identification of the packaging have a significantly reduced NA probability. When additive risk is considered, patients who are chronically ill but display no other risk factors have an NA probability rate of 10.4 %, for patients displaying all identified risk factors this rate increases to 93.9 %. CONCLUSIONS: Our surveys are the largest and most detailed to have been conducted in Germany concerned with the theme of medication-based NA. Our results show that approximately one-third of patients can be classified as non-adherent. Intentional NA factors explain the NA considerably better than do socio-economic factors. PIH29 RETROSPECTIVE EVALUATION OF THE IMPACT OF COST-SHARE INCREASES FOR SPECIALTY MEDICATIONS ON ADHERENCE AND PERSISTENCE Kim YA1, Prasla K2, Rascati K3, Goel NS2, Dunlop D4, Knisely E4, Godley PJ2 1 The University of Texas at Austin, Austin, TX, USA; 2Scott & White Health Plan, Temple, TX, USA; 3University of Texas, College of Pharmacy, Austin, TX, USA; 4Novartis Pharmaceuticals Corporation, East Hanover, NJ, USA OBJECTIVES: A regional health plan implemented a specialty pharmacy benefit with increased copayment/coinsurance for specialty medications. Our objective was to measure and compare the change in adherence and persistence after implementation of the specialty benefit. METHODS: Pharmacy claims for patients who chronically used anti-inflammatory, immunosuppressant, cancer, and multiple sclerosis medications were assessed. The intervention group consisted of those whose out-of-pocket cost-share for specialty medications increased by at least $25 (per 30-day fill), and the control group consisted of those whose out-of-pocket amounts did not change. Adherence, defined by proportion of days covered (PDC), was measured every 3 months for 12 months before and after the copay change. Paired t-tests compared the adherence in the pre- vs. post-periods. Individual growth model analysis evaluated the changes in adherence throughout the time periods. Cox regression analysis determined the difference in persistence between groups after the copay change, controlling for age, gender, copay level, group, and pre-period persistence. RESULTS: There were 237 patients in the intervention group and 211 patients in the control group. The PDC results varied by drug type; adherence decreased after the copay change in the intervention group for patients on anti-inflammatory, immunosuppressant, and multiple sclerosis medications, but remained consistent for patients on antineoplastics. The growth model showed a small, but statistically significant decrease in PDC of 0.01 after copay changes (P = 0.014). The Cox regression analysis indicated that the estimated risk of discontinuing therapy increased for patients in the intervention vs. control groups (hazard ratio = 2.35, 95% CI: 1.43–3.58). CONCLUSIONS: The move to the specialty pharmacy benefit allows for closer scrutiny of specialty utilization by pharmacists who actively monitor utilization and access. Despite the minimal adherence decrease and significant persistence changes, the results indicated relatively more stability with the use of specialty medications than that reported with copayment/coinsurance increases for traditional pharmaceutics. 13th Euro Abstracts PIH30 PERSISTENCE TO POSTMENOPAUSAL OSTEOPOROSIS (PMO) TREATMENTS IN A REGION OF SPAIN Sicras-Mainar A1, Navarro-Artieda R2, Gutierrez L3, Sorio F3, Intorcia M3 1 Directorate of Planning, Badalona Serveis Assistencials, Badalona, Barcelona, Spain; 2Hospital Universitari Germans Trias i Pujol, Barcelona, Spain; 3Health Economics Department, Amgen S.A, Spain OBJECTIVES: Women with PMO are most often treated with bisphosphonates (BPs), as well as with strontium ranelate (SR) and raloxifene (RLX). However, a high percentage of patients are not persistent after 1 year of therapy, which can compromise treatment effectiveness (Imaz I, Osteoporos Int 2010) and thus increase the risk of fractures. The objective was to estimate the persistence to PMO treatments in a region of Spain (Catalonia) representative of the Spanish population. METHODS: This database analysis included women with PMO from 6 primary care centers, aged ≥50 years who initiated BPs (alendronate, ibandronate, risedronate), SR or RLX between January 1, 2004 and June 30, 2008. Patients with cancer, other bone diseases, hospitalization >30 days or <1 year follow-up data were excluded. Persistence was measured at 1, 2 or 3 years according to prescriptions dispensed at office-based pharmacies. Three patients’ cohorts were analyzed: patients with ≥1 year (cohort 1), ≥2 years (cohort 2) and ≥3 years (cohort 3) follow-up. Patients with no access to medication for >1 month were considered non-persistent and a switch to another PMO treatment was considered a discontinuation. In a secondary analysis, switch was not considered a discontinuation. Statistics were performed using Kaplan-Meier methodology. RESULTS: A total of 3,049 patients (mean age ± SD; 68.3 ± 9.7 years) were included in cohort 1; 30% were persistent after 1 year (95%CI:27.5–32.5). In cohort 2 (n = 2698; 68.9 ± 9.3 years) persistence was 35% (CI:32.6–36.2) and 16% (CI:13.6–19.2) after 1 and 2 years respectively. In cohort 3 (n = 2163; 68.4 ± 9.5 years) persistence was 36% (CI:33.9–37.9), 20% (CI:18.3–21.7) and 9% (CI:5.9–12.1) after 1, 2 and 3 years, respectively. The results were similar by drug and frequency of administration, and also when switch was not considered a discontinuation. CONCLUSIONS: In this Spanish population of postmenopausal women with osteoporosis, persistence to treatment was poor, even when a switch was not considered a discontinuation. PIH31 STATIC AND DYNAMIC BALANCE EXAMINATION ON STABILOMETER DURING PREGNANCY Hock M, Pálfi T, Kránicz J, Kriszbacher I, Boncz I, Bódis J University of Pécs, Pécs, Hungary OBJECTIVES: Our study focused on revealing difficulties in maintaining balance during pregnancy. METHODS: Balance testing was performed on 150 volunteers (50 healthy pregnant, 50 pathologically pregnant, 50 healthy young non—pregnant women). One hundred pregnant women were allocated into 4 subgroups based on physical activity performed prior to and during pregnancy. During stabilometer measurements static balance was examined with Romberg test while dynamic balance with 5 programs on a stabilometer. Statistical data were calculated according to mean, standard deviation and T-test while the results were considered to be relevant at P < 0,05. RESULTS: Average results of Romberg test were worse in groups of pregnant women compared to non-pregnant women although the difference was not significant. Concerning the subgroups: those who had previously been involved in professional sports and were healthy and physically active during pregnancy performed better on the open–eye test (P < 0,05) compared to pathologically pregnant patients. Between the second and third trimesters of healthy pregnancies no significant results were found in Romberg test while a significant difference was detected (P < 0,05) in the performance of the 5th exercise of the dynamic balance test. Regarding the 5 programs on the stabilometer examining dynamic balance non-pregnant women performed all exercises significantly better (P < 0,05) than healthy pregnant women. However when compared to pathologically pregnant women apart from the exercises No.2 the result showed a significant difference (P < 0,05). Concerning the same exercises healthy pregnant women performed significantly better (P < 0,05) in exercises No.1 and No.4. Pregnant women who had previously been and remained physically active during pregnancy performed the exercise of moving the centre of body weight in a significantly shorter time than those who were inactive and hospitalised. CONCLUSIONS: Statistical analysis of data proves that physical activity (especially professional sports) prior to becoming pregnant and physical activity during pregnancy have positive effects. PIH32 TREATMENTS TO PREVENT PROSTATE CANCER AND THEIR IMPACT ON HEALTH RELATED QUALITY OF LIFE Lloyd A1, Kerr C1, Rowen D2, Maslen T3, Brazier J2 1 Oxford Outcomes Ltd, Oxford, UK; 2University of Sheffield, Sheffield, UK; 3GlaxoSmithKline UK Ltd, Uxbridge, UK OBJECTIVES: Treatments to prevent prostate cancer are associated with side effects and the impact of these on health related quality of life (HRQL) should be considered in any assessment of cost-effectiveness. However our systematic review identified a lack of robust data. a prospective assessment of the impact of side effects was undertaken using generic and condition-specific measures of HRQL. METHODS: Participants were recruited from an online panel. Men over 55, with either erectile dysfunction (ED), loss of libido, gynecomastia or ejaculatory disorders (without other chronic disease such as renal failure, diabetes, multiple sclerosis etc), and an age matched control group completed assessments of symptom severity and HRQL (EQ-5D and Sexual QoL survey—SQoL-3D). Both HRQL measures can provide 13th Euro Abstracts utility estimates. Regression models (tobit, OLS, Censored least absolute deviation) were used to estimate specific changes in HRQL associated with the side effects. RESULTS: Many participants reported more than one symptom, with ED most common (n = 139); reduced libido (n = 99); ejaculatory disorder (n = 98), and gynecomastia (n = 20). From the OLS regression EQ-5D and SQoL-3D disutilities were estimated for ED (−0.042; −0.075); reduced libido (−0.053; −0.047); ejaculatory disorder (−0.046; −0.028), and gynecomastia (−0.045; −0.037) respectively. EQ-5D and SQoL-3D were weakly correlated (r = 0.296). CONCLUSIONS: The condition-specific and generic estimates indicate the impact of prostate-related preventative treatment on HRQL. While the magnitude of disutilities is similar the poor correlation between the two measures suggests they are measuring different aspects of HRQL. The value of condition-specific versus generic methods for estimating utilities will be discussed. PIH33 PSYCHOMETRIC VALIDATION OF AN ABBREVIATED VERSION OF THE SEXUAL FUNCTION QUESTIONNAIRE (ASFQ) Williams K, Abraham L, Symonds T Pfizer Ltd, Sandwich, Kent, UK OBJECTIVES: The Sexual Function Questionnaire (SFQ-28) is a well established and validated self-reported screening and outcomes measure of female sexual dysfunction (FSD). In order to reduce patient burden and focus on symptoms of FSD, two domains (partner and enjoyment) have been removed to create an abbreviated version (ASFQ). The objective of this study was to ensure that the removal of these domains had not changed the psychometric properties of the measure. METHODS: Forty seven premenopausal women with diagnosed female sexual arousal disorder (FSAD), primarily cognitive arousal difficulties, completed the ASFQ, the Women’s Sexual Distress Questionnaire (WSDQ), an FSAD daily diary and a meaningful benefit question (MBQ) as part of a double-blind, placebo-controlled, 3-way crossover trial. Baseline data were used to assess internal consistency and convergent validity (with the FSAD diary). Known groups validity (comparing those with high versus low levels of distress on the WSDQ at end of treatment) and responsiveness of the arousal domains of ASFQ were also assessed. RESULTS: The ASFQ domains demonstrated excellent internal consistency with Cronbach’s alpha ranging from 0.73 to 0.89. All ASFQ domains showed excellent convergent validity with the respective items on the FSAD diary. Excellent known groups validity was demonstrated for the desire and arousal cognitive domains with those with higher distress scores showing poorer scores on these dimensions of the ASFQ. Moderate effect sizes were observed in the arousal domains for those who indicated they had a meaningful improvement in their arousal disorder during the trial. CONCLUSIONS: The results confirm that the removal of two domains from the SFQ-28 has not impacted on its psychometric properties or responsiveness. The ASFQ is recommended for use in studies where patient burden needs to be minimized. PIH34 CONTENT VALIDITY OF THE BENIGN PROSTATIC HYPERPLASIA IMPACT INDEX (BII): RESULTS FROM CONCEPT ELICITATION AND COGNITIVE INTERVIEWS Naegeli A1, Martin ML2, Kingery LR3, Viktrup L1 1 Eli Lilly & Company, Indianapolis, IN, USA; 2Health Research Associates, Inc, Seattle, WA, USA; 3i3 Research, Basking Ridge, NJ, USA OBJECTIVES: The objective of this qualitative interview study was to assess the content validity of the BII in a sample of men with signs and symptoms of BPH using concept elicitation (CE) and cognitive interviewing (CI) methods. METHODS: Fifty men with BPH participated in the study; 27 completed the CEs and 23 completed the CIs. CEs were semi-structured asking subjects to describe the BPH symptoms they experience and how symptoms impact their life. CIs were semi-structured, asking subjects to describe their 1) difficulties completing the BII; 2) understanding of item meaning and terminology; 3) understanding of response options; and 4) ability to respond appropriately to the specified recall period. All interviews were audio recorded and transcribed. Data was analyzed using Atlas.ti. a saturation table was used to identify when no new concepts were forthcoming. Inter-rater agreement (IRA) was evaluated by having three coders independently dual-code 3 (11%) transcripts. Consistency of coding was characterized by agreement in the identification of concepts, and agreement in assignment of codes. RESULTS: Saturation of concepts was reached by the completion of 21 CE interviews. High agreement on coding consistency was achieved at 69.4 to 89.4% for identification of concepts, and 87.8 to 96% for assignments of codes. The BII was shown to be readily understandable and easily completed in a short period of time, and supported by the qualitative results as measuring the relevant impacts related to BPH. CONCLUSIONS: The BII shows strong evidence of content validity and provides an assessment of disease-related, clinically meaningful impacts of BPH symptoms and treatment outcomes in BPH studies. PIH35 DEMONSTRATING CONCEPTUAL EQUIVALENCE ACROSS MULTIPLE CULTURES: TRANSLATION AND LINGUISTIC VALIDATION OF THE IPAQ Arnold BJ1, Kimel M2, Goddard A3, McCormack J2, Parks-Vernizzi E1, Dhar J1, Pleil A4 1 FACIT.org Translations, Elmhurst, IL, USA; 2United BioSource Corporation, Bethesda, MD, USA; 3Pfizer Ltd, Sandwich, Kent, UK; 4Pfizer Inc, San Diego, CA, USA OBJECTIVES: Translation and linguistic validation of patient reported outcomes (PRO) measures is an essential component of research methodology in preparation A381 for multinational clinical trials. The Injection Pen Assessment Questionnaire (IPAQ) is a dyadic (parent and child together) or self-report of ease of use and preference tool that allows objective and normative comparisons across and between injection pens used to administer human growth hormone (hGH). The purpose of this work was to translate and linguistically validate the IPAQ, which was developed in English in the United States, for use in 7 countries: Czech Republic, Germany, The Netherlands, Slovakia, Sweden, Turkey and United Kingdom. METHODS: The IPAQ was translated according to industry standard methodology. Five parent-child dyads (children 8 to 18 years old) per country completed the respective translated questionnaire and participated in a cognitive interview. Interviews were conducted using a standardized guide to assess the relevance, understandability, and appropriateness of the wording of the translations. Qualitative analyses were performed to ensure equivalence and that the content validity of the IPAQ was maintained across language versions. RESULTS: The study sample consisted of 35 parent-child dyads who use injection devices to administer hGH (42.8% male). Mean age of the children was 12 years. The sample consisted of patients who speak 7 languages collectively. All IPAQ items were well understood and proved relevant to the patients in this sample. Of interest, terms such as, “cartridge”, “needle guard” and “injecting the medicine” were understood similarly by participants across countries. CONCLUSIONS: The results indicate the IPAQ translations were conceptually equivalent to the English source version and easily understood by the target population in all countries. We consider these translations acceptable for PRO assessment in international research, clinical practice and clinical trials. PIH36 PATIENT SATISFACTION WITH FOLLITROPIN ALFA PREFILLED PEN IN WOMEN UNDERGOING OVARIAN STIMULATION: ELABORATION AND VALIDATION OF VENUSF QUESTIONNAIRE Espallardo O1, Ruiz-Balda J2, Carrera M2, Caballero J3, Polanco C1, Bruna I4, Hernández C5, Cuadrado C6, Carrillo E7, Ordóñez D7, García-Velasco J8, Serna J8, Galera F9, Zafra J10, Caballero P11, Hernández E3, Fuente A12 1 Merck Serono, Madrid, Spain; 2Hospital Universitario Doce de Octubre, Madrid, Spain; 3 Clínica FIV, Madrid, Spain; 4Hospital de Madrid Montepríncipe, Madrid, Spain; 5Fundación Jiménez Díaz, Madrid, Spain; 6Hospital Universitario La Paz, Madrid, Spain; 7Hospital Ruber Internacional, Madrid, Spain; 8IVI Madrid, Madrid, Spain; 9Instituto Madrileño de Fertilidad, Madrid, Spain; 10Hospital General Universitario Gregorio Marañón, Madrid, Spain; 11Clínica Tambre, Madrid, Spain; 12Instituto Europeo de Fertilidad, Madrid, Spain OBJECTIVES: To elaborate and validate a questionnaire on patient satisfaction with the follitropin alfa prefilled pen, compared with previously used injectable gonadotropins, in women undergoing ovarian stimulation. METHODS: VENUsf questionnaire with 14 items was constructed after a four-stage process (concept identification and item generation; item review by an expert on female infertility; patients interview; and editing of questionnaire according to interviews results) in which both experts on health outcomes research and fertility professionals and patients were involved. a post-authorization observational study (non-randomised patients, national setting) was undergone in order to validate the questionnaire. After assessing sample adequacy (Kaiser-Meyer-Olkin index and Bartlett’s test of sphericity), Rasch analysis (infit and outfit statistics) provided a reduced version of the questionnaire. Its psychometric properties were assessed: feasibility (omitted response and time to fulfil it), variability (floor and ceiling effects), validity (factor analysis and nonparametric tests) and reliability (Cronbach’s alpha). RESULTS: Initial questionnaire (14 items and 6 additional questions) was answered by 91.1% of sample (n = 107) in an average of 5.04 minutes (SD = 3.3). Statistical correlation was observed between scores from two dimensions of questionnaire and overall treatment satisfaction, as well as variables Number of previous treatments, Time since last treatment and Minutes required to prepare pen. Rasch analyses yield 9 items in the final version of the questionnaire. This process was confirmed by sample adequacy (KMO = 0.72, statistically significant value of Bartlett’s test of sphericity) and reliability parameters (Cronbach’s alpha = 0.78). CONCLUSIONS: The reduced version of the questionnaire VENUsf a feasible, valid and reliable tool for assessing patient satifaction with follitropin alfa prefilled pen in women undergoing ovarian stimulation. PIH37 DISPENSED MEDICATIONS LABELING IN MALAYSIA: VIEWS FROM GENERAL PUBLIC Al-Haddad M, Hassali MA Universiti Sains Malaysia, Penang, Malaysia OBJECTIVES: To assess the perceptions of the general public in Malaysia about the importance of drug labeling. METHODS: A cross sectional study using pre-validated questionnaire was undertaken with a convenient sample of general public in the State of Penang, Malaysia. All data was analyzed using SPSS for Windows version 12.0. Inferential statistics were used whenever appropriate at alpha value of 0.05 or less considered significant. RESULTS: A total of 365 respondents had participated in the survey. Majority of them agreed that drug labeling is important to any person dispensing medicines (74.8%), caretakers (76.26%), health care professionals (77.8%) and patients (80.6%). Besides, they believed that it is important to ensure all drug labels are not vaguely claimed to improve quality of life (54.2%). Majority of respondents agreed that incorrect dosage instructions on the drug label would result in worsening the quality of life (73.2%). Majority of respondents (58.4%) believed that drug labeling is highly important to ensure safe and effective drug use. Majority also expected that all drug products and controlled medicines should be labeled with product name (97.5%), active ingredients (78%), date of manufacture and expiry (87%), and dosage A382 (95%). From their past experience, 21% of them mentioned that they seldom get their drugs labeled whereas 14% of the total respondents were not sure whether they get their drugs labeled or not. In this study, gender, age, and level of education showed significant associations with most responses (P < 0.05). CONCLUSIONS: General public in Malaysia are aware about the importance of drug labeling and about one fifth of the population surveyed rarely receive their medicines labeled appropriately. Therefore, decision makers have to strictly enforce the existing drug labeling requirements for dispensed medications. PIH38 STUDY OF THE DEGREE OF SATISFACTION OF PATIENTS WITH URINARY DISORDERS, EVOCATIVE OF BPH Perrin P1, Auges M2, Taieb C2 1 Lyon Sud, Pierre Benite, France; 2PFSA, Boulogne, France OBJECTIVES: Patient satisfaction, in response to a treatment, is an element of the medical service rendered. We quantified the satisfaction of patients treated medically for urinary disorders. METHODS: A pragmatic cohort (France, Italy, and Portugal) of 420 patients treated with Serenoa Repens, α-blocker or 5 α-reductase inhibitor, was followed-up for 6 months. RESULTS: A total of 175 patients were evaluated. Satisfaction was expressed as the differential between the expectation of the patients recorded before the start of treatment and the status declared at 6 months. In addition a binary (yes/no) question regarding general satisfaction was used as the primary evaluation criterion. We observed positive satisfaction in 61.7% of subjects in terms of the “effort or force needed to start urinating”, 51.1% for the “size and force of the stream of urine”, 54.35% for “sensation of not emptying the bladder after urinating”, 52.38% for “interrupting the flow”, and 50% for the “need to urinate”. We observed negative satisfaction in 68.18% of subjects with respect to the progression of “getting up in the night to urinate”. At 6 months, the response to the general satisfaction question confirms these initial results—indeed, nearly 98% of subjects were satisfied with the treatment of their BPH. We did not see any significant difference between the 3 treatment groups. CONCLUSIONS: The individualised expectation of the patient will undoubtedly be one of the major preoccupations of the next few decades. Medical treatment for BPH is accompanied by a satisfaction that is compatible with long term compliance with the treatment by the patient. PIH39 PATIENTS WITH URINARY DISORDERS, EVOCATIVE OF BPH WHAT ARE THEIR EXPECTATIONS? Taieb C1, Auges M1, Perrin P2 1 PFSA, Boulogne, France; 2Lyon Sud, Pierre Benite, France OBJECTIVES: The individualised expectations of the patient will undoubtedly be one of the major preoccupations in the next few decades to guarantee optimal treatment through compliance. METHODS: A pragmatic, European cohort (France, Italy, and Portugal) of 420 patients presenting with urinary disorders, evocative of BPH, was followed-up over 6 months. a questionnaire regarding expectations was handed out at the first consultation. RESULTS: A total of 317 patients were evaluated. The symptoms that 30.7% of patients wished to see improved with the highest priority were “getting up in the night to urinate”, then for slightly less than 20%, “sensation of not emptying the bladder after urinating”. Amongst the symptoms that patients were the least concerned about were “the effort or force needed to start urinating” for 23% of responders, then “the interruption of the flow of urine” for 16% and the “size and force of the stream of urine”. “Getting up in the night” was the principal complaint in all 3 countries (39% in France, 26 and 25% in Italy and Portugal), similarly “the effort or force needed to start urinating” is the symptom that preoccupies the patients the least in France and Italy, the “size and force of the stream of urine” preoccupies the Portuguese the least. Nearly 90% of the Italians claimed that they would only be satisfied if they never had to get up in the night again, (35% for the French, 50% for the Portuguese). Overall, 60% of the subjects questioned said that they would be satisfied if they were “markedly” improved. CONCLUSIONS: The expectation of patients in the treatment of BPH is very important, and undoubtedly difficult to satisfy entirely. These results are probably due to the fact that our population was composed of patients that had been diagnosed recently. PIH40 METHODOLOGICAL CONSIDERATIONS WHEN ASSESSING WORK PRODUCTIVITY (WP) AND ACTIVITIES OF DAILY LIVING (ADL) OUTCOMES IN MULTINATIONAL CLINICAL TRIALS IN WOMEN WITH HEAVY AND/OR PROLONGED MENSTRUAL BLEEDING (HPMB) TREATED WITH ESTRADIOL VALERATE/DIENOGEST (E2V/DNG) Wasiak R1, Filonenko A2, Stull DE1, Kreif N1, Raluy M1, Ryan J3, Jeddi M4, Uhl-Hochgräber K2, Vanness D5 1 United BioSource Corporation, London, UK; 2Bayer Schering Pharma AG, Berlin, Germany; 3 Bayer Plc, Newbury, UK; 4Bayer Inc., Toronto, ON, Canada; 5University of Wisconsin School of Medicine and Public Health, Madison, WI, USA OBJECTIVES: To evaluate the effect of E2V/DNG, an oral contraceptive, on WP (presenteeism) and ADL outcomes in HPMB sufferers using an appropriate analytical strategy. METHODS: This was a post-hoc analysis of patient-reported outcomes from two multicenter, randomized, placebo-controlled trials in North America and Europe/ Australia that evaluated the efficacy of E2V/DNG in women with HPMB. Data were collected using a modified Work Productivity and Activities Impairment questionnaire. WP and ADL outcomes were measured on a 10-point Likert scale. The analytical 13th Euro Abstracts strategy was developed to determine and apply the most appropriate statistical methodology given the data and methodological challenges, including highly-skewed, incomplete, multi-country data, unbalanced enrolment across countries, and the auto-regressive nature of the outcomes. The analyses progressed from descriptive statistics to Bayesian regression in several sequential steps. The underlying model chosen for Bayesian analyses was simultaneous equation modeling to incorporate temporal aspects and potential cross-country heterogeneity. RESULTS: The data set included 416 patients (E2V/DNG, n = 265; placebo, n = 151) across 12 countries. In all analytical approaches, E2V/DNG vs. placebo treatment showed significantly positive effects on WP and ADL at a magnitude of a one-point change on the Likert scale (based on linear regression analysis). In Bayesian analyses, Gamma distribution yielded a better model fit (DIC = 2129.01 vs. 2460.12 for normal distribution [presenteeism] and DIC = 2313.33 vs. 2640.36 [ADL]). The average treatment effect on presenteeism for Gamma distributed models was −0.82 (95%CI: −1.37, −0.34) at treatment day 84 and −1.06 (95%CI: −1.67, −0.56) at treatment end (EOT; day 196). The average treatment effect on ADL was −1.07 (95%CI: −1.70, −0.52) at day 84 and −1.09 (95%CI: −1.63, −0.62) at EOT. CONCLUSIONS: E2V/DNG has a statistically significant and positive impact on presenteeism and ADL impairment. The robustness of these findings was confirmed by the application of several methodological approaches, with Bayesian analyses appropriately dealing with identified methodological challenges. PIH41 USE OF QUALITATIVE RESEARCH TO IDENTIFY DETERMINANTS OF PERSISTENCE FOR ANTI-OSTEOPOROTIC TREATMENTS Alegre P1, Arnould B2, Spizak C2, Marrel A2 1 Science Union et cie (Servier), Suresnes, France; 2Mapi Values, Lyon, France OBJECTIVES: A large proportion of patients are non-adherent to anti-osteoporotic treatments within 12 months of initiation. In the health psychology literature, several behavioural models have been developed to explain non-adherence, most notably the Health Belief Model (HBM). Based on these, our study aimed to explore determinants of persistence (i.e., continuation) with anti-osteoporotic treatments. METHODS: A psychologist carried out face-to-face interviews with patients who had taken antiosteoporotic treatment for 2 years. Interviews were based on a guide including questions on beliefs and experiences with osteoporosis and its treatment. Saturation was reached with 16 patients providing over 95% of information. Content analysis of interview transcripts was performed to highlight criteria that determine intentions regarding treatment continuation. RESULTS: Through patient interview analysis, determinants with an effect on persistence were categorized either as barriers or facilitators. While general health behaviour was identified in both type of determinants, barriers to persistence included: Disease perceived as non-severe (“osteoporosis is not a severe disease, perhaps not a disease at all”); Treatment side-effects (“there are side-effects, I suffer from allergies, I am hesitant to take treatment”) and Constraints (“you have to take treatment every day”; “you have to wait 2 hours after dinner before taking treatment”). Facilitators included: Perceived benefits (“my treatment works, bone mineral density results are good”); Habit/maintenance (“I take my treatment because I am used to it, I do not question it”) and Relationship with doctor (“I have confidence in my doctor”). CONCLUSIONS: Results are consistent with the health psychology models, in particular the HBM. They add new concepts to HBM. Results suggest that areas for improvement in persistence in osteoporosis lay in beliefs about disease and treatment efficacy, rather than treatment convenience. PIH42 THE USE OF PATIENT-REPORTED OUTCOME MEASURES FOR DRUG APPROVAL IN KOREA Park IS, Lee EK SookMyung Women’s University, Seoul, South Korea OBJECTIVES: Patient-Reported Outcomes(PRO) is the information on treatment evaluations that comes directly from a patient. PRO is derived from every endpoint from patient reports. These reports have been recorded in a patient diary or other reporting systems. PRO is an important tool for measuring the impact of diseases, treatments, health and social polices like pharmacoeconomic polices. The objective of this study is to determine the level and nature of the use of PRO compared to other endpoints for Korea’s approved new drugs. METHODS: A survey has been conducted for 121 newly approved products in Korea during the period of 2005 to 2009. Different survey forms were prepared for each product. Each survey form provided information on products and clinical trials: title, objectives, types of endpoints, types of PRO and the information on PRO sent by e-mail to the manager of each pharmaceutical company. RESULTS: Primary endpoints were measured in 200 clinical trials whereas secondary endpoints were measured in 183 clinical trials. The two both studied 111 products. For PRO, a primary endpoint was reported in 33(14.6%) trials and a secondary has been reported in 66(25.1%) trials. For Clinician-reported outcomes, the figures were 104(46.0%) and 108(41.1%), respectively. For Laboratory/ device endpoints, the figures were 89(39%) and 89(33.7%), respectively. For 11 products, PROs have been used as the only type of endpoint. About 40 percent of PROs have been used as primary and secondary endpoints in clinical trials for drug approval during the last five years(2005~2009). However, the results of the PRO secondary endpoints have not been reflected in product labeling. CONCLUSIONS: PRO has been used as a useful endpoint that can be employed in clinical trials for the development and evaluation of new drugs. Therefore, it is necessary to develop PRO instruments and guidelines to evaluate Korean patients’ PRO for Koreans and clinical trials at home. A383 13th Euro Abstracts PIH43 USE OF QUALITY-ADJUSTED LIFE-YEARS FOR THE ESTIMATION OF EFFECTIVENESS OF SCREENING PROGRAMMES: A SYSTEMATIC REVIEW Aäklin S1, Räsänen P2, Laitinen R1, Kovanen N1, Autti-Rämö I3, Sintonen H4, Roine RP5 1 National Institute for Health and Welfare (THL), Helsinki, Finland; 2Helsinki and Uusimaa Hospital Group, Group Administration and National Institute for Health and Welfare (THL), Helsinki, Finland; 3The Social Insurance Institution, Helsinki, Finland; 4University of Helsinki, Dept. of Public Health and National Institute for Health and Welfare, Helsinki, Finland; 5 Helsinki and Uusimaa Hospital Group, Helsinki, Finland OBJECTIVES: To identify and characterise studies that have used quality-adjusted life-years (QALYs) based on actual measurements of patients’ health-related quality of life (HRQoL) as an indicator of effectiveness of screening programmes. METHODS: Systematic search of the literature until March 2010, using several electronic databases. Initial screening of articles was based on abstracts, and evaluation of full-text articles was done independently by at least two of the authors. The final inclusion decision was based on consensus. RESULTS: The search identified 1610 articles. The use of QALYs in evaluation of screening programmes has expanded in recent years as 617 of the articles had been published between 2007 and 2010, while the corresponding figure between 1966 and 2010 was 993. Based on review of abstracts, 431 full-text articles were obtained for closer inspection. Of the reviewed full-text articles, 78 reported QALYs based on patient-derived data using a valid HRQoL assessment. The most frequently used method was Time Trade-Off (55%) followed by Standard Gamble (21%) and EQ-5D (26%). The most frequently studied medical conditions were malignant diseases (22%) followed by cardiovascular diseases (19%). The majority of the studies came from the USA (46%) or the UK (27%). All studies employed some kind of modelling with the Markov model being the most prevalent type (62%). Most articles (71%) concluded that the screening programme studied was cost-effective. Most of the studies included in the review were of high quality, but there were still a lot of variation as to which outcome measures were chosen, and only few studies reported incremental analysis. CONCLUSIONS: The use of QALYs in the evaluation of cost-effectiveness of screening programmes has expanded during the last few years. However, only a minority of studies have employed actual patient- derived HRQoL data. PIH44 THE INFLUENCE OF INDIVIDUAL FACTORS ON PELVIC PAIN AND QUALITY OF LIFE IN ENDOMETRIOSIS Koppán Á1, Oehmke F2, Hámori J1, Kriszbacher I1, Vranics I1, Garai J1, Bódis J1, Boncz I1, Tinneberg HR2, Koppán M1 1 University of Pécs, Pécs, Hungary; 2University of Giessen, Gießen, Germany OBJECTIVES: The aim of the study was to assess potential individual factors influencing the efficacy of combined surgical and medical therapy, as well as quality of life and pain scores in endometriosis patients with pelvic pain. METHODS: In a prospective study we used a specifically designed questionnaire among patients suffering from persistent pelvic pain and undergoing laparoscopy followed by GnRH analogue therapy. The questionnaire gathered information on the following groups of variables: age, marital status, education, reproductive and medical history including previous pregnancies and parity, knowledge of accompanying pelvic disorders, regular sport activity, concurrent cigarette smoking, as well as general quality of life estimates including self-image. Pelvic pain was scored using a visual analogue scale. RESULTS: Eighty-one patients were eligible and willing to participate by filling out the questionnaires before surgery and upon completing medical therapy. After cessation of therapy, 53.0% of patients reported absence of pain. Among these, 11.6% were smokers, corresponding to 35.7% of all smokers in the study. However, 56.7% of non-smoker participants reported a positive outcome that proved to be significantly larger than the ratio of pain-free smoker participants (p = 0.02). Improvement in quality of life was reported by 73.7% of all patients, and only 9.4% of them were smokers. Fourteen patients among regular exercisers and 33 patients among those without physical activity reported the effectiveness of painkillers for pelvic pain, corresponding to 45,1% and 66% of these subgroups, respectively (P < 0.05). CONCLUSIONS: Based on our results, we can conclude, that regular smoking might have a disadvantageous impact on the success rate of combined surgical and medical therapy for endometriosis related pelvic pain. Also, it appears that painkillers might be less effective among endometriosis patients performing regular daily sport activities, and, thus it might impose them to an unnecessary burden of possible side effects. PIH45 HEALTH-RELATED QUALITY OF LIFE (HRQOL) OF FERTILE WOMEN USERS OF CONTRACEPTIVE METHODS Perulero N1, Roset M1, Gómez MÁ2, Dueñas JL3, Lete I4, Lertxundi R5, Pérez-Campos E6, de la Viuda E7, Sánchez-Borrego R8 1 IMS Health, Barcelona, Spain; 2Centro de Planificación Familiar, Alicante, Spain; 3Hospital Universitario Virgen Macarena, Sevillaý, Spain; 4Hospital Santiago Apóstol, Vitoria, Spain; 5 Clínica Euskalduna, Bilbao, Spain; 6Hospital General de Requena, Valencia, Spain; 7Hospital Universitario de Guadalajara, Guadalajara, Spain; 8Clínica Diatros, Barcelona, Spain OBJECTIVES: Women’s choices of contraceptive method are largely affected by the safety and efficacy of contraceptive methods as well as by change in women’s quality of life. The SEC-QoL questionnaire has been recently elaborated and validated in fertile women users of contraceptive methods in Spanish population. METHODS: Observational, prospective, multicenter study involving 41 contraceptive gynaecolo- gist and GPs. Three patients’ groups were identified at baseline visit: 1) Fertile women users of efficacy contraceptive methods; 2) Fertile women users of non-efficacy contraceptive methods but will begin to use efficacy contraceptive methods; 3) Fertile women user of non efficacy contraceptive methods. All women were ≥18 years. Follow up and final visit was at 4 month since baseline visit. Socio-demographic and clinical information was gathered. The specific questionnaire SEC-QoL (19 items) was administered. RESULTS: A total of 453 women were included. Mean (SD) age was 29.3 (7.7). The most frequently used contraceptive methods were: pill (43% A), inconsistent condom (79% B and 63% C). The factorial analysis confirmed a total of 5 dimensions (social/daily activities, menstrual and breast symptoms, psychological and sexual dimension). HRQoL according to SEC-QoL score was higher (better HRQoL) in group a (59.2) than in group B (46.6) and C (49.8) (P < 0.01). Considering a and B groups, HRQoL (SEC-QoL total score) was better in women with low bleeding intensity, low pain, and low androgenic and breast symptoms (P < 0.01). No differences were shown according to days between menstrual intervals. Hormonal contraceptive method users showed better HRQoL (61.2) than women using other method (53.3) (P < 0.05). Both, in group a and B an increase in HRQoL was shown in the final visit, being statistically significant in all dimensions of SEC-QoL questionnaire in Group B (P < 0.05). CONCLUSIONS: SEC-QoL is a valid and reliable specific measure to assess HRQOL in fertile women users of contraceptive methods, in routine clinical practice and clinical studies. PIH46 A SUBJECTIVE APPROACH TO FACTORS INFLUENCING RECOVERY AMONG MEN WHO ARE ACTIVE IN THE LABOR MARKET Benkő B, Tancsics D, Boncz I, Kriszbacher I, Rébék-Nagy G, Warta V, Éliás Z University of Pécs, Pécs, Hungary OBJECTIVES: The survey was meant to find out how much importance males attribute to various factors including patients’ compliance, faith and financial resources, physicians’ expertise and personality, nurses’ attitudes, relatives’ and God’s help and the appropriate medication. Another major aim was to reveal any possible differences between males’ and females’ attitudes to the factors listed and any gender specific differences in the evaluation of individual factors and their relationship with background variables such as health as a value, past major disease or operation, subjective health status, age, education. METHODS: In the non-randomised research quantitative approach was used to investigate males’ personal opinions concerning recovery. The cross-sectional examination included a survey administered in three workplaces. The target population included employees 18–65 years of age. There were 188 questionnaires distributed of which n = 147 were assessable, which included n = 73 completed by males. RESULTS: There was a significant difference concerning the importance attributed to relatives’ (p = 0.000,), faith (p = 0.005) and God’s help (p = 0.005). Males considered the role of these factors as less important. Those judging their own health status very bad or quite bad thought that financial resources would have a great impact on recovery. This opinion was significantly characteristic among males with lower levels of education. Physicians’ personality was regarded as less important by males than by females. CONCLUSIONS: There was no significant difference detected between the two genders with regard to the priority order of the examined factors including patients’ compliance, physicians’ expertise and patients’ financial resources. Age proved to have a strong influence on evaluating the different factors. INDIVIDUAL’S HEALTH – Health Care Use & Policy Studies PIH47 LONG-TERM EFFECTS OF CHILDREN PNEUMOCOCCUS VACCINATION: AN AGENT BASED APPROACH Zauner G1, Miksch F2, Popper N1, Endel G3, Schiller-Frühwirth I3, Breitenecker F2 1 Dwh Simulation Services, Vienna, Austria; 2Vienna University of Technology, Vienna, Austria; 3 Main Association of Austrian Social Security Institutions, Vienna, Austria OBJECTIVES: Dynamical modeling and simulation of the epidemical influence of infant immunization with PCV7 against pneumococci induced serious disease and time depending behavior of carrying rates of the pathogens in nasopharynx using an agent based approach. METHODS: The implemented strategy is a multi agent based model approach with detailed implementation of individual social contacts and direct person to person transmission of the pneumococcal bacteria. To test the influence of different vaccination strategies and the induced non linear effects like herd immunity effects and serotype replacement, strongly influencing the over all effectiveness, additional assumptions are made. Instead of 90 known pneumococcal pathogens only 2 relevant classes are modeled: the serotypes included in the vaccine and others. The dynamic model is simulated over a long period (25 years) to find out long term effects for stability tests. Additionally population dynamics are integrated because of their influence on long time behavior. Induced change in the demographic shape of the population combined with knowledge of different carrier rates depending on the persons age lead to extra effects on immunization. RESULTS: For standard immunization program implementation as advocated by EPAR/EMEA the direct influence of the immunization including herd immunity and serotype replacement is measured. For basic simulation runs with realistic parameters for Austria a decrease of 30% of serious diseases is measured. Results for serotype replacement (40–60%) are calculated. In contradiction to other modeling methods this values are dynamically generated by the simulation and no input parameters. CONCLUSIONS: Agent based modeling leads to detailed A384 information on behavior of the pathogen distribution for all age groups and their changes over time and is thereby a real extension to classical static decision tree based methods in HTA. The used validated simulation system is dealing with transparent parameters and is better understandable for decision makers in HTA. PIH48 ANTIDEPRESSANT PRESCRIBING TO PATIENTS AGED 18 YEARS AND YOUNGER: A PHARMACOEPIDEMIOLOGICAL STUDY Truter I Nelson Mandela Metropolitan University (NMMU), Port Elizabeth, Eastern Cape, South Africa OBJECTIVES: Antidepressants carry an increased risk of suicidality in children, adolescents and young adults. The primary aim was to determine antidepressant prescribing and cost to patients aged 18 years and younger in a South African private health care sector patient population. METHODS: A retrospective, cross-sectional pharmacoepidemiological study was undertaken on prescription data of a private pharmacy group in South Africa for 2009. The study focused on antidepressant prescribing. No diagnoses were available. a total of 814540 antidepressants were prescribed, of which 12549 were prescribed to patients aged 18 years or younger. RESULTS: A total of 4269 patients (53.74% females) received an average of 2.94 (SD = 3.31) antidepressants over the year. The total cost of antidepressant prescribing was R1 581480, with an average cost of R126.02 per product (item). Two-thirds (66.67%) of antidepressants were prescribed to patients from 12 to 18 years of age. Selective serotonin re-uptake inhibitors (SSRIs) accounted for 55.91% of antidepressant prescribing, followed by tricyclic antidepressants (32.32%), together accounting for 79.49% of antidepressant cost. Lithium was only prescribed to 1.04% of patients in this age group. The highest average cost was for serotonin and noradrenaline re-uptake inhibitors (R322.94 per product), selective mono-amine oxidase inhibitors (R177.20) and SSRIs (R155.52). Tablets were the preferred dosage form. The most frequently prescribed active ingredients were imipramine (19.77%), escitalopram (14.60%), citalopram (14.09%) and fluoxetine (13.92%), together accounting for 62.38% of prescribing frequency. CONCLUSIONS: Diagnoses were not available, therefore only prescribing trends could be investigated. Fluoxetine generally has a favourable riskbenefit balance (prescribed “off label”), yet it only had a prescribing frequency of 13.92% in this study. Active ingredients such as escitalopram and citalopram are not preferable in this age group. The risk of prescribing antidepressants in children, adolescents and young adults should be balanced against the clinical need. PIH49 PRESCRIBING PATTERNS OF SELECTIVE SEROTONIN RE-UPTAKE INHIBITORS AND SELECTIVE SEROTONIN AND NORADRENALINE RE-UPTAKE INHIBITORS IN CHILDREN AND ADOLESCENTS IN A SOUTH AFRICAN PRIVATE PRACTICE SETTING Van Schalkwyk K, Truter I Nelson Mandela Metropolitan University (NMMU), Port Elizabeth, Eastern Cape, South Africa OBJECTIVES: There has been much controversy over the prescribing of selective serotonin re-uptake inhibitors (SSRIs) and selective serotonin and noradrenaline reuptake inhibitors (SNRIs) regarding their safety and efficacy in patients younger than 19 years with limited guidelines to prescribers from South African authorities. The primary aim was to investigate the prescribing patterns of SSRIs and SNRIs to children and adolescents, younger than 19 years, in a primary practice setting in a South African patient population. METHODS: Data from 2009 of patients younger than 19 years who were prescribed antidepressants belonging to a private medical aid scheme were retrospectively analyzed. RESULTS: There were 440 patients younger than 19 years receiving antidepressant drugs accounting for 6.20% of patients of any age receiving antidepressant therapy. SSRIs and SNRIs were prescribed to 50.68% of these patients. Female patients were prescribed SSRIs and SNRIs most frequently (62.33%). The average age of patients was 14.21 (SD = 3.05) years while the majority of patients were older than 12 years (77.13%). Fluoxetine was prescribed most frequently (36.50%), followed by citalopram (22.14%) and escitalopram (17.15%). The total cost of SSRIs and SNRIs amounted to R81201.69. Escitalopram accounted for 30.28% of the cost while fluoxetine only accounted for 11.11% of the cost. Venlafaxine accounted for 6.32% of prescriptions although being contraindicated in this patient population. Paroxetine accounted for 5.96% of prescriptions although its use is not recommended due to possible suicidality side effects. a total of 72.65% of patients received drug therapy for less than 6 months while 41.26% of patients received a single prescription for an SSRI or SNRI. General practitioners prescribed 67.76% of prescriptions while psychiatrists only prescribed 14.96%. CONCLUSIONS: The prescribing of these agents to children and adolescents needs to be further investigated using retrospective data which include clinical information and studies conducted over a longer time period. PIH50 MATERNAL MORTALITY RATE IN INDIA: A COMPREHENSIVE LITERATURE REVIEW Guri S, Kumar J, Saini P, Ghasolia D JRDS ePharma Health Pvt Ltd, Dehli, India OBJECTIVES: Maternal mortality rate in India: a comprehensive literature review. METHODS: Information was retrieved from websites of World Health Organization, Pubmed and grey literature retrieved using key terms like maternal mortality, MMR 13th Euro Abstracts etc. The relevant prevalence data were extracted and compiled. RESULTS: The maternal mortality rate (MMR) is maternal death per 100,000 live births in one year. WHO estimates show that out of the 529,000 maternal deaths globally each year, 136,000(25.7%) are contributed by India. According to registrar general of India estimates for the year 2000, MMR for India was 407/100,000 live births. The trend has not changed significantly in the last 5 years. In the Armed Forces the scenario is very encouraging with a MMR of 36 per 100,000 live births. In 1980, India reported about 677 maternal deaths per 1, 00,000 live births but in 2008, this was down to 254 deaths. MMR has reduced by 4% a year since 1990. High MMR was reported in Hindu (573) than Muslims (384) and in Illiterate people (574) (1994). India has climbed from bottom to 127th place in the MMR ratings, above Pakistan and just below Nepal. Hemorrhage (25.6–38%) ranks first as the cause of maternal death, followed by anemia (14–24%) sepsis (13–16%), toxemia of pregnancy (11.9%), abortions (8–9%) and obstructed labor (6.2–10%) while other causes together total 35.3%. Tamil Nadu has taken initiatives to improve maternal health services leading to reduction in maternal mortality from 380 in 1993 to 90 in 2007 due to increased rate of institutional deliveries from 20% in 1971 to 97.7% in 2007. CONCLUSIONS: A consistent decline in the maternal mortality rate has been observed in India but it attributes to a few states. Initiatives need to be taken by government in regard to proper facility in hospitals, efficient supply of blood, and iron containing medicines. PIH51 AN ESTIMATE OF COSTS AND BENEFITS OF ALTERNATIVE METHODS OF DELIVERY: AN EMPIRICAL ANALYSIS IN AN ITALIAN HOSPITAL Pizzo E Imperial College London, London, UK OBJECTIVES: The recent large increase in caesarean sections (CSs) in Europe seems not to be completely justified: CS is practiced independently of epidemiological evidence. This work analyzes costs and benefits effectively involved in alternative methods of delivery- vaginal delivery (VD), with and without epidural analgesia, and planned caesarean. METHODS: The empirical analysis has been conducted in an Italian University hospital, through direct collection of data, questionnaires and interviews to patients and staff. a logistic regression has been used to model the probability of the event “delivery with planned caesarean section” occurring as a function of a set of clinical and socio-economic characteristics of the women. a micro–costing analysis has been used to assess the direct health costs, following an activity–based costing approach. From a societal perspective we consider also the indirect and the intangible costs of each method. Patients’ wellbeing is measured through appropriate anonymous instruments—the State-Trait Anxiety Inventory, the Italian Questionnaire of Pain and the Childbirth Perception Questionnaire- to measure the changes in clinical and psychological dimensions due to the delivery experience. RESULTS: The results confirm the hypothesis that CS is widely performed for non-medical reasons (Osborn, 1995). The analysis shows that CS is, on average, more expensive than VD, but the difference is marginal if we take into account the opportunity-cost of labour time. CONCLUSIONS: Since CS is generally reimbursed more than VD to cover the supposed higher costs of surgery, differences between the real costs and the DRG tariffs may induce opportunistic behaviour in terms of clinical practice. We show that, in general, VD with analgesia provides better results both in terms of costs and, but the final effect of its introduction is not clear: it may reduce the frequency of inappropriate caesarean sections, but it may also increase the costs due to complications PIH52 EVALUATING THE PERCEPTION OF HOSPITAL PHARMACIST TOWARD THEIR CURRENT CLINICAL ROLE IN HEALTH CARE SYSTEM OF PAKISTAN Azhar S1, Hassali A2, Izham M1 1 Universiti Sains Malaysia, Penang, Malaysia; 2Discipline of Social & Administrative Pharmacy, Universiti Sains Malaysia, Pinang, Palau Pinang, Malaysia OBJECTIVES: To investigate hospital pharmacists’ perception towards their current clinical role in Pakistan’s health care system. METHODS: The study population consisted of hospital pharmacists from three cities of Punjab, Pakistan; Islamabad, Faisalabad and Lahore. a sample of 116 hospital pharmacists was selected from government hospitals from these three cities. RESULTS: At least 77.6% of the hospital pharmacists involved in patient counseling in Pakistan. Although they are willing to take personal responsibility to resolving any drug related problem but pharmacist (n = 98, 84.5%) conveyed that their current role is more focusing towards the record keeping of hospital pharmacy. Only (n = 67, 57.8%) of the pharmacist expressed that they are involve in making and designing of hospital formulary. In this case significant difference was noted with respect to age (p = 0.020) and gender (p = 0.056).As far as experience with other health care profession is concern, only (n = 60, 51.7%) of the respondents were agreed of having collaborative relationship with other health care professional. Moreover only (n = 65, 56.0%) of the hospital pharmacist agreed regarding their suggestion taken into consideration by physician. CONCLUSIONS: The findings suggest that the hospital pharmacists in Pakistan do have concerns about their present professional role but they are facing significant barriers in terms of increasing clinical services. Moreover, pharmacists need to be proactive in collaboration with other health care professionals. This paves the way for the concept of pharmaceutical care in health care system of Pakistan. 13th Euro Abstracts PIH53 DO HIGH-INCOME GROUPS IN PAKISTAN UNDERSTAND THE CONCEPT OF GENERIC MEDICINES? Jamshed SQ1, Hassali MA1, Ibrahim MIM1, Shafie AA1, Babar ZUD2 1 Universiti Sains Malaysia, Minden, Penang, Malaysia; 2University of Auckland, Auckland, Malaysia OBJECTIVES: High income groups in Pakistan resort to private health care settings for consultations and treatment modalities but only 5% of them have Social Health insurance coverage. As 77% of medicine expenditures are out-of-pocket payments in Pakistan, it is imperative to assess the understanding, perception, and attitude of this privileged group towards cost-effective alternatives. METHODS: In order to have an in depth evaluation of the issue qualitative methodology was adopted. a combination of purposive and snowball sampling was used to conduct face-to-face semi-structured interviews, which were then audio-taped, and transcribed verbatim. As sample size in qualitative research revolves around the attainment of point of saturation, no new themes emerged after the interview of 8 respondents. RESULTS: Thematic content analysis identified four major themes; appropriate knowledge of generic medicines, negative perception towards generic medicines, negative attitude towards generic medicines, and future recommendations which should be directed towards the maintenance of quality and efficacy of generic alternatives. Interestingly, all the respondents managed to explain generic medicines with reference to patent expiry. Regarding perception all of them expressed negative views and considered quality and safety as questionable domains in generic medicines. All the respondents cited negative concerns towards generic medicine utilization and harbored the notion that “low cost relates compromised quality.” Majority of the respondents expressed that future strategies should be directed towards educated high income group to build confidence for generics and this should be possible until and unless the local manufacturers in Pakistan make themselves compliant with WHO Good Manufacturing Practices (GMP). CONCLUSIONS: The respondents showed excellent understanding towards generic medicines. They exhibited negative concerns and attitudes towards generic medicine utilization. Still there is room for improvement, provided the confidence is built in consumers regarding generic medicines quality and efficacy, which in turn will pave the way for their quality utilization. PIH54 METFORMIN FOR THE TREATMENT OF CHILDHOOD OBESITY: A SYSTEMATIC REVIEW AND META-ANALYSIS Bouza C, Gutierrez L, Lopez-Cuadrado T Instituto de Salud Carlos III, Madrid, Spain OBJECTIVES: Childhood obesity associates with significant morbidity and premature death; its prevalence has increased greatly during the past three decades; and it is recognized as a worldwide public health problem. However, the efficacy of treatments for childhood obesity remains unclear. In recent years the use of metformin, an insulin sensitizer, has aroused a great interest for the treatment of obesity in adults. Our aim was to assess the efficacy and safety of metformin for childhood obesity. METHODS: Systematic review of literature and meta-analysis of randomized controlled trials in obese subjects age ≤19 years without diabetes or other morbidities. Structured electronic searches of published studies until March 2010 were performed. Changes in the Body Mass Index (BMI) were considered our main outcome measure of efficacy whereas metabolic parameters such as insulin levels, glucose, HOMA, lipid profile and leptin levels were considered as secondary outcomes. Safety parameters included adverse events and losses from adverse effects. Individual studies were graded using published methodologies. Pooled estimates of effect and confidence intervals were derived using a fixed effects model and tested for heterogeneity. Consistency across studies was evaluated by means of the I-square statistic. RESULTS: Seven trials met the inclusion criteria. All trials compared metformin with placebo and used behavioural co-interventions. Average follow-up was six months. Though with small sample sizes, methodological quality of trials was adequate. Meta-analysis showed that compared to placebo, metformin provided a significant decrease in BMI (−1.90 (−3,−.8). No statistical significant differences were found in secondary outcomes. Main adverse effects were digestive, no serious adverse events were reported. CONCLUSIONS: Available evidence suggests that, added to behavioural interventions, metformin is a relatively safe and effective treatment for childhood obesity in the short term. Further research with longer follow-up periods is needed to solve this important health issue. Partially supported by Spanish National I+D Program. PIH55 BPH PATIENTS TREATED WITH PHYTOTHERAPY: RESULTS AT 6 MONTHS Perrin P1, Auges M2, Taieb C2 1 Lyon Sud, Pierre Benite, France; 2PFSA, Boulogne, France OBJECTIVES: Assess the impact of the treatment of urinary disorders of the lower urinary tract related to benign prostatic hypertrophy (BPH) using medical treatment under actual conditions of use. METHODS: A pragmatic cohort of 420 patients treated medically, was followed up for 6 months, using several validated questionnaires: IPSS, MSF4, and SF12. RESULTS: 267 patients treated with Serenoa Repens were evaluated, the mean age was 64.3 ± 8.6 years, and on average the diagnosis had been made 18 months previously. At 6 weeks, the IPSS was significantly improved in this group (P < 0.0001). This improvement in the IPSS score between 6 weeks (11.98 ± 5.11) and inclusion (14.58 ± 5.65) was 2.6 points. An improvement was also observed at 3 months. At 6 months, the p-value was also significant (P < 0.0001). The improvement in the IPSS score between 6 months (8.20 ± 4.12) and inclusion (14.65 A385 ± 7.01) was 6.6 points. The physical dimension (50.97 ± 6.45 at inclusion) of the SF12 improved significantly (P < 0.001) from the 6th week (52.63 ± 5.20), an improvement that was confirmed at 6 months (53.21 ± 5.16) (P < 0.001) in comparison with inclusion (49.09 ± 6.58). The mental dimension (50.9 ± 7.04 at inclusion) of the SF12 improved significantly (P < 0.001) from the 6th week (52.16 ± 7.31), an improvement that was confirmed between 5.45 and 6 months (52.50 ± 6.69) (P < 0.001) in comparison with inclusion (47.09 ± 10.82). The MSF4 was unchanged. CONCLUSIONS: We observed an improvement in the IPSS score from the 6th week; this statistical improvement was confirmed by a significant clinical improvement in the 6th month. This favourable progression is consistent with the improvement observed for both dimensions of the SF12. PIH56 COMPARATIVE PRICING AND REIMBURSEMENT ANALYSIS BETWEEN BULGARIA AND THE CZECH REPUBLIC Peev S1, Petrikova A2, Petrova G1, Stoimenova A1 1 Medical University Sofia, Faculty of Pharmacy, Sofia, Bulgaria; 2VFU Brno, Brno, Czech Republic OBJECTIVES: Comparison of regulatory rules and procedures for pricing and reimbursement of pharmaceuticals between Bulgaria and the Czech Republic (CR). METHODS: Medicine laws, health insurance laws, corresponding regulations stating the pricing and reimbursement procedures were reviewed. Special emphasis was devoted to the requirements for pharmacoeconomic evidences in the procedures. RESULTS: Both countries apply the reference pricing system for prescription medicines and the CR is among the 8 reference countries for Bulgaria, while the opposite is not the case. In Bulgaria the lowest reference price for ex-factory price setting is used, while the CR employs the average among the three lowest prices. There is also a regressive margin scale in both countries applied. In Bulgaria margins are stated for wholesalers and retailers separately, while in the CR they are negotiated between wholesalers and retailers. In both countries health insurance is obligatory. In the CR there are 11 insurance companies and in Bulgaria—only one fund. In Bulgaria reimbursement is in the form of positive drugs lists. For reimbursement of pharmaceuticals in the CR a complex external and internal referencing is employed and evaluation of innovativeness is necessary. There are 111 internal therapeutic groups defined at the 4th ATC level (e.g. statins). The reimbursement base is set as the cheapest price of medicine in the group in all 27 EU countries. There is a bonus of maximum 30% if a medicine shows superiority. In Bulgaria the reimbursement level is defined as the cheapest price per DDD for every INN. Both countries require pharmacoeconomic evidences, but there are guidelines in the CR, while Bulgaria applies only criteria for evaluation. CONCLUSIONS: We consider the Czech system more flexible and providing freedom for the manufacturers and distributors due to its negotiating practice and therapeutic level of reimbursement, but the process is more prolonged. PIH57 CHALLENGES IN THE ADHERENCE AND ADOPTION OF INTERNATIONAL GUIDELINES: A SYSTEMATIC REVIEW OF THE ADHERENCE TO THE HEALTH CARE INFECTION CONTROL PRACTICES ADVISORY COMMITTEE (HICPAC) GUIDELINES FOR THE APPROPRIATE USE OF VANCOMYCIN IN CHILDREN Bouza C, Lopez-Cuadrado T, Saz-Parkinson Z Instituto de Salud Carlos III, Madrid, Spain OBJECTIVES: One of the most pressing problems faced by health care services is the increasing prevalence of antimicrobial resistance to vancomycin, an important antimicrobial for the treatment of infections caused by gram-positive pathogens in the severely ill patient. Inappropriate use of vancomycin encompasses serious public health consequences linked to the development of resistant species. In 1995, the HICPAC developed guidelines that delineated specific criteria for appropriate vancomycin use to reduce the improper use of antimicrobials and improve patient safety. Our aim was to evaluate and summarize the level of compliance of clinical practices with the HICPAC guidelines by assessing the appropriateness of vancomycin therapy for hospitalized critically ill children. METHODS: Systematic review of literature (Jan 1996Feb 2010) and meta-analysis of studies performed in subjects aged 0–18 years. Structured electronic and manual searches were performed. The main outcome measure was the proportion of vancomycin prescribed appropriately according to the HICPAC guidelines. Appropriate/inappropriate uses were subdivided into several categories. Summary rate ratios and confidence intervals were estimated using a fixed effects model and tested for heterogeneity. Furthermore, we explored the potential reasons for non-adherence to guidelines. RESULTS: From 24 candidate studies, 13 publications met the inclusion criteria. Meta-analysis showed that in only 21% of patients (95%CI: 18%–23%), vancomycin prescribing and dispensing practices were consistent with the recommendations. Non-adherence to HICPAC guidelines included surgical prophylaxis, empirical use and length of therapy. Lack of awareness, lack of agreement, lack of outcome expectancy and inertia of previous practice, emerge as specific reasons for non-adherence. CONCLUSIONS: This study reveals that adherence to international guidelines is far from optimal with a variety of potential barriers that undermine the process. Because physician adherence is critical in translating recommendations into improved outcomes, there is an urgent need for strategies aimed at improving physician compliance with guidelines to optimize antibiotics utilization. Supported by the Spanish National I+D Program (STPY 1346/09). A386 INDIVIDUAL’S HEALTH – Conceptual Papers & Research on Methods PIH58 HEALTH AND ECONOMY: A GOVERNMENTAL PERSPECTIVE, NATIONAL ACCOUNTING MODEL FOR ASSESSING INVESTMENTS IN ROTAVIRUS VACCINATION Connolly M1, Schey C1, Standaert B2 1 Global Market Access Solutions, St Prex, Switzerland; 2GlaxoSmithKline Biologicals, Wavre, Belgium OBJECTIVES: The WHO repeatedly stresses the importance of human capital and investing in health as a determinant of future economic growth. We describe a health care investment model that reflects the government perspective attributed to investing in rotavirus vaccination in Egypt, and how changes in morbidity and mortality influence government expenditure (education, health, allowances) over many generations. METHODS: The model applies a generational accounting approach for estimating the inter-temporal fiscal impact of policy changes. It accounts for direct fiscal transfers between age cohorts and the State during different life stages—childhood, school-age, working-age, and retirement—while simultaneously accounting for rotavirus medical costs, and how rotavirus mortality and morbidity influence government fiscal transfers. Costs are expressed in Egyptian Pounds (EGP; 1c = 7EGP). The model is constructed using Egyptian life tables, rotavirus related and unrelated health care costs, employment earnings adjusted for age and social parameters. The model compares vaccinated and unvaccinated cohorts against rotavirus using discounted net tax revenues (gross taxes—transfers). RESULTS: Based on variations in rotavirus vaccine price, the model predicts health service savings mostly attributable to averting rotavirus treatment costs that could be achieved within 3–5 years and reaching EGP178 million obtained at year-5. The discounted net tax revenue between vaccinated and unvaccinated cohorts was EGP5.2 billion and EGP27.3 billion at year 25 and 50, respectively. Investing in rotavirus vaccination represented a 15% rate of return for government at year-50. Long-term government net tax revenues were insensitive to vaccine prices, although sensitive in short-term. CONCLUSIONS: Health investment models are complementary to conventional economic evaluations of health care technologies. But they illuminate how government accounts, tax revenues, and expenditures are influenced by investing in health care programs. Investing in rotavirus vaccination could deliver early cost-offsets associated with reduced health care expenditure. It could increase future government net tax revenue attributed to lives saved. PIH59 PUBLIC HEALTH ECONOMICS—AN EMERGENT SUBDISCIPLINE? 1 2 3 3 Ehlers L , Kruse M , Højgaard B , Søgaard J 1 Health Economics & Management, Institute of Business Studies, Aalborg University, Denmark, Aalborg, Denmark; 2Centre for Applied Health Services Research, University of Southern Denmark, Odense, Denmark; 3DSI Danish Institute for health services Research, Copenhagen, Denmark BACKGROUND: Are we witnessing the beginning of the development of a new subdiscipline of health economics? The purpose of the paper is to outline the main theoretical arguments for the establishment of a new research tradition on “public health economics” combining public economics with health economics. METHODS: The results from a systematic literature search in PubMed and NHSEED on the term “public health economics” were presented and discussed at a workshop at the Danish public health conference in Nyborg September 22, 2009. Among the invited participants were health economists, public health researchers, HTA-advisors, econometricians, and decision makers. RESULTS: Three main theoretical arguments were identified: 1) There is a need for developing new methods for the economic evaluation of public health interventions. 2) Economic evaluation of health care may be seriously misleading if public health research is ignored. 3) The entire health economic research tradition may benefit from encompassing a more socioeconomic model of health. CONCLUSIONS: There is limited tradition yet for health economists and researches within public health to work together and meet regularly, no scientific journal specifically oriented towards public health economics, and no textbook in health economics that include research from all these related areas. Thus, the research environment is to some extent characterised by lack of cooperation, which may constitute an obstacle for the development of a consistent and coherent line of economic research in the field of public health. PIH60 ESTIMATING THE CROSS-SECTIONAL NUMBER AND GESTATIONAL AGE AT BIRTH DISTRIBUTION OF INFANTS FOR ECONOMIC AND HEALTH IMPACT ANALYSIS Myers ER1, Misurski DA2, Swamy GK1 1 Duke University, Durham, NC, USA; 2GlaxoSmithKline, Philadelphia, PA, USA OBJECTIVES: In order to estimate the size of the population of infants at risk for seasonal influenza because of ineligibility for vaccination, we developed a method for estimating the cross-sectional number of infants under 6 months of age at any point in time. METHODS: Data on the monthly number of deliveries at gestational ages 23–42 weeks for 2006 (the most recent available) from from publicly accessible birth certificate data from the US National Center for Health Statistics (NCHS) were used. The number of deliveries at each gestational age each calendar week were derived based on the number of days in each month. Conditional probabilities for survival for each week after birth were derived from the linked birth and death certificate data set from 2006 for each gestational age. The number of infants younger than 6 months 13th Euro Abstracts with a given gestational age at birth at any point in time was calculated using the following formula: Number of infants (gestational age) born (current week) + Number of infants (gestational age) born (current week-1)*1 week survival (gestational age) + . . . Number of infants (gestational age) born (current week-25)*26 week survival These results were then summed for gestational ages 23–42 weeks to obtain the total number of infants. Separate calculations were performed for single pregnancies, twins, and higher order multiples. RESULTS: In 2006, the average number of infants 0–6 months of age in any given week was 1.98 million, with seasonal variation reflecting seasonal variability in deliveries. CONCLUSIONS: Estimates of the number of infants within a given age range alive at a given point in time can be derived from routinely collected administrative data. These estimates, together with estimates of the number of pregnant women, can be used for evaluation of interventions targeting pregnant women and infants. PIH61 CALIBRATION OF A DISCRETE EVENT SIMULATION MODEL OF NATURAL HISTORY OF HPV-RELATED DISEASES Quon P1, Vanness DJ1, Kansal A1, Hillemanns P2, Remy V3, Quilici S3 1 United BioSource Corporation, Bethesda, MD, USA; 2Hannover Medical School, Hanover, Germany; 3Sanofi Pasteur MSD, Lyon, France OBJECTIVES: Develop a mathematical model simulating the clinical and economic impact of different cervical cancer screening options for German females alongside HPV vaccination. METHODS: We developed a discrete event simulation (DES) model, describing the natural history of cervical cancer and genital warts for five categories of HPV types (HPV 16, 18, 6/11, other high risk and other low risk). To fit German epidemiological data, we manually calibrated model parameters for natural history to match a number of calibration targets including genital warts incidence by age, cervical cancer incidence and mortality by age, prevalence of HPV type by age, and distribution of HPV types by disease stage. The model also calibrates against prevalence of CIN by age as reported in previous HPV disease models. The fit of model outputs to calibration targets was represented by a calibration score computed using normalized residuals weighted by the quality of the available data (e.g., prioritizing epidemiological data over model results). Our model employs lognormal distributions for time to progression and two-piece exponential distributions for time to regression, enabling us to simulate the long separation in peak times between HPV infection, cervical cancer precursors and cancer incidence. RESULTS: Predicted type-specific HPV-prevalence and disease incidence are close to epidemiological data. HPV prevalence differed from reported prevalence by <1.55% for each 5 year age group for HPV 16, 18, and other high risk HPV. The model predicted genital warts incidence of 159 per 100,000 (target: 167), cervical cancer incidence of 17.7 per 100,000 (target: 15.5), cervical cancer mortality incidence of 4.3 per 100,000 (target: 4.1), and agreed well with incidence age distributions. CONCLUSIONS: We developed an individual-based, fully calibrated model that is ready for cost-effectiveness analysis of cervical cancer vaccination and screening strategies. Time-to-event distributions, an inherent feature of DES, facilitate realistic modelling of disease progression. PIH62 ESTIMATING THE CROSS-SECTIONAL NUMBER AND GESTATIONAL AGE DISTRIBUTION OF PREGNANT WOMEN FOR ECONOMIC ANALYSIS Myers ER1, Misurski DA2, Swamy GK1 1 Duke University, Durham, NC, USA; 2GlaxoSmithKline, Philadelphia, PA, USA OBJECTIVE: To develop a method for estimating the cross-sectional number of pregnant women, distributed by gestational age for cost-effectiveness and budget impact analysis. METHODS: We obtained data on the monthly number of deliveries at gestational ages 20–42 weeks for years 2003 through 2006 (the most recent available) from publicly accessible birth certificate data from the US National Center for Health Statistics (NCHS). Estimates of the number of deliveries at each gestational age each week were derived based on the number of days in each month, under the assumption that deliveries were randomly distributed throughout the month. Cohorts of women who became pregnant during the same calendar week were constructed using the following formula: Deliveries at 20 weeks (calendar week X) + Deliveries at 21 weeks (calendar week X+1) +. . . . Deliveries at 42 weeks (calendar week X+22). This procedure was repeated for each week from January 2003 through July 2006. Estimates of the number of pregnancies between 6 and 20 weeks were generated by applying gestational age-specific loss rates from a large prospective cohort study of early pregnancy. RESULTS: Over the 3.5 year period, there were approximately 2.5 million pregnant women ranging from 6 to 42 weeks gestational age in the US; the mean weekly number increased from 2.5 to 2.6 between 2003 and 2006. There was also substantial seasonal variation in singleton pregnancies, with the highest number of pregnant women observed in March and April of each year, but no seasonal variation in multiple gestations. CONCLUSIONS: Routinely collected data on births allows estimation of the number of pregnant women at any point in time. Such data, in conjunction with estimates of probabilities of delivery and pregnancy and neonatal complications, can be useful for cost-effectiveness and budget impact analyses of interventions during pregnancy. A387 13th Euro Abstracts PIH63 TURKISH CULTURAL ADAPTATION AND VALIDATION OF GLASGOW HEALTH STATUS INVENTORY Oksuz E, Malhan S, Tekin R, Yilmaz F, Baytar S Baskent University, Ankara, Turkey OBJECTIVES: The Glasgow Health Status Inventory, (GHSI) measures the effect of a health problem on the quality of life of a person and allows cross-comparison among many health conditions, among different health interventions, and among demographic and cultural subgroups. The GHSI contains 18 health status questions, which ask specific questions about how the health problem has affected their quality of life. This study aims to adapt the GHSI into Turkish culture and check the reliability and validity of the inventory culturally. METHODS: The original instrument was translated and back translated by two independent translators. a small sample consisting of 40 people was used to check the initial comprehension and factibility. Cronbach’s Alfa was used to assess reliability and factor analysis to assess dimensionality. The EuroQol questionnaire and corresponding Visual Analogue Scales were used for concurrent validity. RESULTS: A total of 163 people participated in this study. 52% of them were female, 48% of them were male. Mean age was 24.3. The internal consistency coefficient (Cronbach’s alpha) of GHSI was 0.86. Factor analysis of the scale revealed that it was composed of four factors with Eigenvalues >1.0, accounting for 62.2 % of the total variance. Correlations were moderate with EuroQol and VAS. CONCLUSIONS: The culturally adapted GHSI has good validity and reliability, making it a potentially useful outcome measure in the determining the effect of health problems on the quality of life of people in Turkey. PIH64 A CONCEPTUAL FRAMEWORK FOR A PATIENT REPORTED OUTCOMES MEASURE OF ATTITUDES TOWARD COMPLEMENTARY AND ALTERNATIVE MEDICINE IN MENOPAUSE Bunsiriluck S, Williamson EM, Donyai P Reading University, Reading, UK OBJECTIVES: Menopause may cause hot flashes, night sweats, mood swings and other symptoms. Although hormone replacement therapy (HRT) is prescribed in severe cases, it is not suitable for all women, and increasing numbers of women refuse or discontinue HRT because of side effects, or real or perceived health risks. The use of complementary and alternative medicine (CAM) in menopause has been increasing, with women perceiving CAM as safer than HRT, but this is not supported by clinical evidence. Biased decisions occur when an individual’s cognition is affected by factors that include attitudes, preferences, and moods. Cognitive bias affecting women’s perception of CAM can impact on decisions and ultimately CAM usage in the menopause. Yet, no validated questionnaire exists to measure attitudes towards CAM in menopause, in line with accepted patient reported outcome (PRO) methodology. Our aim was to develop a conceptual framework of attitudes to CAM in the menopause to inform the development of a PRO tool. METHODS: A systematic review of the literature was conducted April-May 2010 using Medline and PsycINFO databases with the MESH terms attitude, menopause, complementary therapy, and perception. a total of 122 papers were retrieved, with 91 excluded on the basis of abstract review and 13 after full-text analysis, resulting in 19 studies suitable for inclusion. The papers were analyzed in line with qualitative review methodology to produce a number of themes that were later grouped together to produce the dominant categories. RESULTS: A conceptual framework was developed to reflect attitudes to CAM in the menopause. The dominant categories of perception related to menopausal symptoms, optimism about CAM, pessimism about HRT, cure-control with CAM, and relationship with physician. CONCLUSIONS: This research has identified key domains relevant to women’s attitudes towards CAM in the menopause, from which a relevant PRO instrument could be developed. PIH65 USING A GENERALIZED ADDITIVE MODEL TO EXAMINE THE RELATIONSHIP BETWEEN BODY MASS INDEX AND HEALTH-RELATED QUALITY OF LIFE IN THE ELDERLY POPULATION—RESULTS FROM THE POPULATION-BASED GERMAN KORA-AGE STUDY Hunger M, Thorand B, Döring A, Peters A, Schunk M, Holle R Helmholtz Zentrum München—German Research Center for Environmental Health (GmbH), Neuherberg, Germany OBJECTIVES: Numerous cross-sectional studies reported that high body mass index (BMI) is associated with poorer health-related quality of life (HRQL), while only few studies also provide evidence for impaired HRQL in underweight individuals. Our objective was to investigate the nonlinear relationship between BMI and HRQL in an elderly general population sample using semiparametric regression methods. METHODS: We analyzed data from 4562 individuals aged 65 years or older, living in the region of Augsburg, Southern Germany. The data come from the KORA-Age study which is based on a postal follow-up of individuals who participated in the population-based MONICA/KORA surveys S1-S4. Health-related quality of life was measured using the German EQ-5D index. For our multivariable regression analyses, we used a generalized additive model (GAM) to estimate the functional form of the relationship between BMI and HRQL after adjusting for confounding factors. RESULTS: We found a significant nonlinear (inverse U-shaped) relationship between BMI and HRQL after adjusting for gender, age, sociodemographic factors and comorbid conditions. The maximum HRQL was observed at a BMI of about 25 kg/m2. Our estimates indicate that going from a BMI of 25 to 35 is associated with a EQ-5D utility loss of about 5.2 units (corresponding to 0.28 of standard deviation). On the other hand, underweight individuals with a BMI of 18 had an average impairment of 5.8 units compared to a BMI of 25. Subgroup analyses showed that the inverse U-shaped relationship is more pronounced in individuals ≥75 years than in individuals aged 65–74 years. In particular, the effect of low BMI on HRQL is more important in the older age group. CONCLUSIONS: Generalized additive models are an adequate method to estimate the nonlinear relationship between BMI and HRQL in the elderly population. With increasing age, low BMI has similar impairments in HRQL as overweight. PIH66 IDENTIFYING APPROPRIATE COMPARISON GROUPS IN GLOBAL PREGNANCY REGISTRIES Albano JD1, Roberts SS2, Brown VD3 1 Kendle International Inc., Durham, NC, USA; 2University of North Carolina, Wilmington, Wilmington, NC, USA; 3Kendle International Inc., Wimington, NC, USA BACKGROUND: Pregnancy exposure registries are valuable for studying the teratogenicity of drugs used in the post-marketing setting. However, their limited sample size and potential for selection bias can make data interpretation challenging. While enrolling an internal comparison group is ideal, this is often not feasible. To ease these limitations, pregnancy registries commonly use a population-based background rate as the primary comparison group. OBJECTIVES: The objectives of this research are to evaluate appropriate population-based birth defect surveillance systems, cohorts, and other studies from a variety of geographic areas and to systematically assess their utility as a comparison group for global pregnancy exposure registries. RESULTS: Among the 14 studies evaluated, study designs include observational cohort studies, case control studies, population-based surveillance registries, and population-based active surveillance systems. Catchment areas are somewhat diverse geographically and range in scope from a metropolitan city, single state/province, or individual country, to multi-national/multi-continental networks of organizations contributing to a single data source. All identified studies have been ongoing for at least a decade. Studies range in size from several thousand participants in total to more than 1 million participants enrolled annually. Data reporters vary and include patients, physicians, and other health care personnel. Some studies ascertain infant outcome and birth defect status at the time of birth only, while others involve pediatricians and include followup for one to several years. CONCLUSIONS: There are many potential sources of comparison data for overall birth defect rates; however, the sources vary greatly in design, data collection methods, birth defect coding systems, clinical review procedures, and equation for calculating birth defect prevalence rate. Utilizing a populationbased background rate can be useful, but care should be taken in selecting an appropriate comparison group. The differences between the individual pregnancy registry and the population-based comparator should be evaluated and recognized. NEUROLOGICAL DISORDERS – Clinical Outcomes Studies PND1 INDIRECT COMPARISON OF ADVERSE EVENTS AND DROPOUT RATES FOR EARLY PARKINSON’S DISEASE (PD) MONOTHERAPY TRIALS: PRAMIPEXOLE, ROPINIROLE AND RASAGILINE Zagmutt FJ1, Tarrants ML2 1 Vose Consulting, Boulder, CO, USA; 2Teva Pharmaceuticals, Kansas City, MO, USA OBJECTIVES: Pramipexole, Rasagiline, and Ropinirole are approved for monotherapy in early PD. Understanding the comparative safety profiles of these products can be important information to choose an optimal therapy. As no direct comparisons of the safety profile of these products are available, our objective was to perform an indirect comparison of Adverse Events (AEs) and Dropout Rates (DRs). METHODS: Articles were selected and reviewed via Cochrane Guidelines. Placebo-controlled randomized clinical trials were eligible for review. Data collected for analysis included total AEs, Cognitive, Gastrointestinal (GI), and Sleep/Fatigue AE categories and dropout rates. We used indirect meta-analysis to calculate the pooled Relative Risk (RR) of each product against placebo, and then used pairwise comparisons. Frequentist and Bayesian methods were used to compare sensitivity of findings. RESULTS: 208 studies were identified and reviewed, 6 were determined eligible via established criterion. The RRs and [95%CIs] from the fixed-effects model for Rasagiline, Pramipexole, and Ropinirole respectively were total AEs: .97 [.87, 1.07], 1.05 [1.00, 1.1], 1.07 [1.00, 1.14]; Cognitive: .78 [.43, 1.44], 5.56 [2.35, 13.13], 1.56 [.83, 2.94]; GI: .9 [.49, 1.63], 2.00 [1.57, 2.53], 2.43 [1.81, 3.27]; Sleep/Fatigue: .84 [.53, 1.35], 1.63 [1.35, 1.97], 3.24 [2.08, 5.05]; and DR: .60 [.39, .91], 1.01 [.66, 1.56], 1.77 [1.14, 2.76]. AEs were no worse than placebo for all Rasagiline AEs and for Ropinirole cognitive AEs. DRs for Rasagiline were significantly lower than placebo. Rasagiline had the lowest RRs with ≥90% confidence for all categories. Results were comparable across statistical models tested. CONCLUSIONS: This indirect treatment comparison suggests that subjects with early PD treated with Rasagiline have less risk for adverse events and treatment dropouts than patients treated with Pramipexole or Ropinirole. Ropinirole exhibits the highest risk for GI AEs, Sleep/Fatigue AEs and DRs, while the risk for Cognitive AEs is higher for Pramipexole. A388 PND2 THE INFLUENCE OF PARENTERAL NUTRITION ON THE PLASMA PROTEIN BINDING OF THERAPEUTIC DRUGS Salih M, Bahari MB, Abd AY, Shafie AA Universiti Sains Malaysia, Minden, Penang, Malaysia OBJECTIVES: Parenteral nutrition (PN) is used in various clinical situations. However, some interactions might occur with drugs that are administered concurrently with PN. Therefore, this search reviews the potential interactions of PN with therapeutic drugs (mainly antiepileptics), especially in respect to the plasma proteinbinding of the drug. METHODS: The articles related to the topic were identified through the Medline and PubMed. Full text of the articles were then traced from the Universiti Sains Malaysia (USM) library subscribed databases, including Wiley-Blackwell Library, Cochrane Library, EBSCOHost, OVID, ScienceDirect, SAGE Premier, Scopus, SpringerLINK, and Wiley InterScience. The articles from journals not listed by USM library were traced through inter library loan. RESULTS: There were interactions between PN and drugs, including antiepileptics. Several guidelines were designed for the management of illnesses such as traumatic brain injuries or cancer patients, involving the use of PN and antiepileptics. Moreover, many studies demonstrated the in vitro and in vivo PN-drugs interactions, especially with antiepileptics. CONCLUSIONS: Alteration in the drug-free fraction result from PN-drug (i.e. antiepileptics) interactions may necessitate scrupulous reassessment of drug dosages in patients receiving these therapies. This reassessment may be particularly imperative in certain clinical situations characterized by hypoalbuminemia (e.g., burn patients). PND3 EFFICACY AND TOLERABILITY OF INTERFERON-BETA IN ALL TYPE OF MULTIPLE SCLEROSIS Nikfar S1, Abdollahi M2, Rahimi R3 1 Faculty of Pharmacy, Tehran University of Medical Sciences and Food & Drug Laboratory Research Center, MOH, Tehran, Iran; 2Faculty of Pharmacy; and Pharmaceutical Sciences Research Center, Tehran University of Medical Sciences, Tehran, Iran; 3Faculty of Traditional Medicine, and Pharmaceutical Sciences Research Center, Tehran University of Medical Sciences, Tehran, Iran OBJECTIVES: The aim of this meta-analysis was to evaluate the efficacy and tolerability of IFN β for maintenance of remission in MS. METHODS: Pubmed, Scopus, and Cochrane Central Register of Controlled Trials were searched for studies investigated efficacy and/or tolerability of interferon β (IFN β) in MS. Data were collected from 1966 to 2009 (up to July). RESULTS: Nine randomized placebo controlled clinical trials met our criteria and were included. Summary relative risk (RR) for at least one relapse including all types of MS and all types of IFN β was 0.86, a significant RR (95% CI: 0.76–0.97). Summary RR for at least one relapse in secondary progressive MS (SPMS) patients received all types of IFN β was 1.11(95% CI: 0.79–1.55). RR for at least one relapse in RRMS patients received all types of IFN β was 0.77(95% CI: 0.57–1.05). RR for at least one relapse in SPMS patients received IFN β-1b was 0.93 (95% CI: 0.75–1.14). RR for at least one relapse in patients with all types of MS received IFN β-1a was 0.97 (95% CI: 0.57–1.67) and for IFN β-1b was 0.92 (95% CI: 0.85–1). The summary RR for discontinuing because of adverse events in nine trials was 2.76, a non-significant RR (95% CI: 1.97–3.89, P < 0.0001). The summary RR death in three trials was 1.53 (95% CI: 0.45–5.15). RR for suicides or suicide attempts in five trials was 0.86 (95% CI: 0.41–1.79). RR for different adverse events of all types of IFN β comparing to placebo in all types of MS including flu-like symptoms, injection site reactions, injection site inflammation, myalgia, leucopenia, lymphopenia, increased alanine aminotransferase were significant except for depression that its RR was non-significant. CONCLUSIONS: It can be concluded that INF β’s effectiveness in MS is dependent to administration of different kind of interferon β and type of MS. NEUROLOGICAL DISORDERS – Cost Studies PND4 BUDGET IMPACT ANALYSIS OF NATALIZUMAB FOR MULTIPLE SCLEROSIS TREATMENT IN BRAZIL: A 5-YEAR PROJECTION Ferreira Da Silva AL1, Finkelsztejn A1, Ribeiro R2, Polanczyk CA1 1 Federal University of Rio Grande do Sul (UFRGS), Porto Alegre, Brazil; 2Hospital de Clínicas de Porto Alegre, Porto Alegre, Brazil OBJECTIVES: To estimate the annual and 5-year budgetary impact of including natalizumab in the Brazilian public health care system drug formulary for the treatment of relapsing remitting multiple sclerosis (MS). METHODS: Brazilian public health care system perspective was adopted. The baseline scenario comprised four treatment options available in Brazil in the year 2009, and its corresponding market shares: 1) interferon beta 1A intramuscular (23%); 2) interferon beta 1A subcutaneous (32%); 3) interferon beta 1B (23%); and (4) glatiramer acetate (22%). Data were retrieved from the Brazilian health system database. a claims data based approach was used to estimate the baseline population size, which was estimated to be 6935 patients. In addition to the base-case scenario, three alternative scenarios were created to estimate the budgetary impact of including natalizumab to the drug formulary at 3 possible market shares: 3%, 10% or 22%. Costs for currently available treatments comprised average purchase prices paid by the Brazilian government in 2009. The adopted natalizumab dose cost was R$,652 (c1660), obtained from public data issued by the Brazilian’s regulatory health agency (ANVISA). Costs were adjusted for 4% 13th Euro Abstracts annual inflation. No discounts were considered. Sensitivity analysis was performed. RESULTS: Base-case costs for treatment of MS within a 5-year period were R$1280,698,041 (c582,135,473). Relative to this value, at 3% market share, the inclusion of natalizumab produced a 6% (R$78,858,437/c35,844,744) increase in expenditures. At 10% market share, the corresponding expenditures increased 9% (R$108,902,983/c49,501,356). At 22% market share, there was an increase of 13% in total costs with MS treatment (R$161,903,806/c73,592,639). CONCLUSIONS: Entry of natalizumab into the market is likely to result in a significant increase in MS treatment costs for the Brazilian health care system. However, this must be weighted against the need for new treatment options as well as potential savings from better disease control rates. PND5 BUDGET-IMPACT MODEL TO TEST EFFECTS OF CHANGING RUFINAMIDE TIER STATUS FROM A PAYER PERSPECTIVE Powers A1, Faria C1, Buchner D1, Shaul A2, Cragin L3 1 Eisai, Inc., Woodcliff Lake, NJ, USA; 2United BioSource Corporation, Bethesda, MD, USA; 3 United BioSource Corporation, Lexington, MA, USA OBJECTIVES: Lennox-Gastaut Syndrome (LGS) is an epileptic encephalopathy characterized by intractable seizures of many types. LGS has its onset in early childhood, and often persists into adulthood. It is estimated that the prevalence of LGS among children and adults with epilepsy is 6.5% and 1.5%, respectively. This budget impact model was developed to help US payers make informed decisions regarding the addition of rufinamide for the management of LGS. To analyze overall budget impact of changing rufinamide from tier 3 to tier 2 status for a large U.S. payer. METHODS: The budget impact model was constructed based upon the indicated population for rufinamide, which is children 4 years and older and adults (18+) requiring adjunctive treatment of seizures associated with LGS. The model follows patients over three years and estimates the direct health care costs associated with LGS treatment before and after the introduction of rufinamide. The three main categories of data inputs informing the model include: plan population inputs, treated prevalence inputs, and treatment and cost inputs. The base-case model results are based on data derived from published literature, publicly available data sources, or assumptions. The results are presented as follows: total costs over 3 years, cost per member per year (PMPY), cost per treated member per year (PTMPY), cost per member per month (PMPM), and cost per treated member per month (PTMPM). RESULTS: Assuming a one million member plan and a discount rate of 3% shifting rufinamide from tier 3 to tier 2 resulted in $1,001 in additional total costs over 3 years. This translates to a change in a PTMPY of $1.00 and a PTMPM of $0.13. PMPY and PMPM were not notably affected. CONCLUSIONS: Based on this analysis changing rufinamide from tier 3 to tier 2 does not exhibit a significant cost increase. PND6 VALUE OF BUDGET IMPACT ANALYSIS BASED ON EPIDEMIOLOGIC DATA: INSIGHTS FROM MULTIPLE SCLEROSIS IN SAO PAULO, BRAZIL Ferreira Da Silva AL1, Filkelsztejn A1, Ribeiro R2, Polanczyk CA1 1 Federal University of Rio Grande do Sul (UFRGS), Porto Alegre, Brazil; 2Hospital de Clínicas de Porto Alegre, Porto Alegre, Brazil BACKGROUND: Despite its relatively low prevalence in Brazil, multiple sclerosis (MS) is responsible for significant health care expenditures for the public health care system. OBJECTIVES: To compare actual MS treatment expenditures and expected expenditures based on an epidemiological projection. METHODS: We retrieved actual treated prevalence and treatment costs from Brazil’s public health system database for the city of Sao Paulo, in 2009. Then, current prevalence of MS for the city of Sao Paulo was estimated based on the best epidemiological evidence available, a prevalence study conducted in 1997, which reported a prevalence of 15/100,000 inhabitants. Data were analyzed comparing actual and potential expenditure related to drug treatments, assuming that 85% of prevalent cases comprised the relapsing form of MS. RESULTS: We retrieved the total number of doses for each one of the available treatment options for MS in the city of São Paulo for the year 2009 (interferon beta-1A, interferon beta-1B and glatiramer acetate). Considering the expected number of doses required to treat one patient for one year with each treatment option, we estimated the number of patients covered within one year of treatment to be 1,319 patients. This corresponds to an estimated treated prevalence of 11.9 cases per 100,000 inhabitants, considering the estimated population for the year 2009 (11,037,593 inhabitants). Actual costs incurred for the city of Sao Paulo in 2009 with the treatment of MS were R$40,993,666/c18,633,484. Assuming a prevalence of 15/100,000 inhabitants, 1,666 cases of MS would be expected in the city, of which 1,416 would be relapsing MS. Their treatment would require R$43,811,800/c19,914,454. CONCLUSIONS: There are marginal differences between treated and expected prevalence of relapsing MS in the city of Sao Paulo. This indicates that local programs for MS diagnosis and treatment are successful. Likewise, current budget impacted analyses based on epidemiologic information seems accurate. PND7 ECONOMIC IMPACT OF PERSISTENCE TO DISEASE MODIFYING THERAPIES FOR THE TREATMENT OF MULTIPLE SCLEROSIS Szkurhan AR1, Dembek C2, Malik S1, Agarwal SS2, Rajagopalan K2, Rashid N1 1 Dymaxium, Inc, Toronto, ON, Canada; 2Biogen Idec GmbH, Wellesley, MA, USA OBJECTIVES: Multiple sclerosis (MS) is a chronic incurable disease with a progressive course. Studies have demonstrated that disease progression can be slowed by treatment with disease modifying therapies (DMTs). Poor DMT adherence is associated with a 13th Euro Abstracts higher risk of relapse, lower quality of life and higher medical costs. The adherence measure, persistency, refers to the duration a patient continues with therapy. This study aims to evaluate the non-pharmacy medical costs associated with the persistency of DMT’s in patients with MS. METHODS: A decision-analytic model was designed using persistency rates, health care resource use and costs from the published literature. The model evaluated the medical costs associated with the persistence, switching, and discontinuation rates of patients on interferon β1a IM, interferon β1b SC, and interferon β1a SC over a 2-year time horizon. Using 6-month intervals in the model, patients could switch to a comparator treatment, discontinue treatment, or persist/ remain on their current treatment, and would incur the non-pharmacy medical costs associated with each. For patients who switched treatments, the model assumed an equal probability of switching to the three remaining DMTs. The model also assumed that patients who discontinued therapy would initiate on a different therapy after one month of discontinuing. The overall medical costs related to persisting on treatment, switching treatment or discontinuing treatment were calculated for all treatment arms. RESULTS: Based on the results from the model, an annual per-patient medical cost of $7423 was observed for interferon β1a IM patients, showing a medical cost advantage over patients on interferon β1b SC ($8144; 8.9%), and interferon β1a SC ($7552; 1.7%). CONCLUSIONS: MS is a lifelong disease that requires continuous treatment. The results of this model show that interferon β1a IM is a cost-saving treatment for the medical costs associated with MS patients on disease modifying therapies. PND8 THE POTENTIAL COST IMPACT OF USING A PEG HYDROGEL SEALANT COMPARED WITH FIBRIN SEALANT TO PREVENT CEREBRAL SPINAL FLUID LEAKS AFTER CRANIAL SURGERY IN THE UK Ott M1, Grieve J2, Larkin M3, Minshall M4 1 Covidien, Zurich, Switzerland; 2University College London, London, UK; 3PRMA Consulting, Fleet, UK; 4Covidien, Fishers, IN, USA OBJECTIVES: Cerebrospinal fluid (CSF) leak is an unavoidable consequence of cranial surgery with leak rates between 0% to 25% and resulting costs ranging from £9,000 to £36,000. The use of polyethylene glycol (PEG) hydrogel dural sealant as an adjunct to sutured closure has been shown to reduce CSF leak rates as compared with sutures alone in cranial procedures (GR Cosgrove et al, 2007). Our hypothetical analysis applied the potential cost offsets of using PEG hydrogel sealant as an adjunct to sutured closure in 200 cranial procedures assuming CSF leak rates of 4.5% (9/200) (GR Cosgrove et al, 2007), compared with 10% (20/200) for fibrin sealant (JA Grotenhuis, 2005). METHODS: The incremental additional cost for treating CSF leaks using total patient costs for those with CSF leaks (£25,253) compared to those without CSF leaks (£10,497) was estimated at £14,756 in a Dutch study (JA Grotenhuis, 2005). We applied this CSF leak cost to estimate potential UK hospital cost offsets on 200 hypothetical cranial patients using PEG hydrogel sealant (£300/treatment), compared with fibrin sealant (£133/treatment) on all 200 patients. RESULTS: Use of a PEG hydrogel sealant compared with fibrin sealant could potentially save £137,611 (or £688/patient) for a hospital that performed 200 cranial surgery procedures using a PEG hydrogel sealant compared with using fibrin sealant on all 200 procedures. CONCLUSIONS: This study demonstrates the potential economic advantages of using a PEG hydrogel sealant in cranial procedures. Future clinical direct comparative studies would be beneficial to confirm these findings and understand the possible economic advantages for other types of dural surgeries. PND9 ASSESSMENT OF DISEASE BURDEN ASSOCIATED WITH EPILEPSY IN HUNGARY, BASED ON A CROSS-SECTIONAL QUESTIONNAIRE SURVEY OF 100 PATIENTS Érsek K1, Mikudina B1, Arányi Z2, Juhos V3, Baji P1, Péntek M1, Brodszky V1, Gulácsi L1 1 Corvinus University of Budapest, Budapest, Hungary; 2Semmelweis University, Budapest, Hungary; 3St. István and St. László Hospital, Budapest, Hungary OBJECTIVES: Epilepsy is the second most frequent neurological disease among adults; approximately 50–60 thousand people suffer from epilepsy in Hungary. With our survey we aimed to assess the cost of illness from a societal viewpoint as well as the quality of life associated with epilepsy in Hungary. METHODS: We conducted a cross-sectional questionnaire survey of 100 consecutive patients in 2 epilepsy centres in Hungary in collaboration with epilepsy specialists. The self-developed questionnaire consisted of general and disease-specific parts to determine costs associated with epilepsy. Questions related to basic demographic characteristics, disease progression and health and social services used in the past 12 months, disease-associated nonmedical services, as well as the productivity losses of epilepsy. Two generic quality of life questionnaires, EQ-5D and SF-36 were also used. RESULTS: A total number of 100 patients were assessed (60% female). The average age was 36.7 years (SD. 12.5), average disease duration was 15 years (SD. 12). Quality of life calculations resulted in mean of 0.83 (SD. 0.24) in EQ-5D, 74 (SD. 16) in VAS scale and 72.3 (SD. 20.2) in SF-36. Cost-of-illness calculations were carried out and direct (15%) and indirect costs (85%) were determined. The annual total cost of epilepsy per capita was estimated to be a mean of c5179 (SD. c10,822) with the human resource method, while c2,552/capita/year (SD. 8659) by the friction cost method. CONCLUSIONS: Comparing the 100 patients’ data with data from previous EQ-5D based population-sample survey shows that quality of life among people with epilepsy was lower than among the average Hungarian population. We found significant correlation (at P < 0.05) among all of the quality of life results and the yearly total costs (calculating both with human capital and friction cost methods). A389 PND10 LONG-TERM DISABILIY COST IN TUBEROUS SCLEROSIS COMPLEX (TSC) IN BRAZIL Valentim J Novartis, Sao Paulo, SP, Brazil OBJECTIVES: To estimate long-term disability costs associated with tuberous sclerosis complex, a rare multisystem genetic disease, in Brazil. METHODS: Literature review for TSC long-term disability and economic burden was performed (Pubmed, LILACS, SciELO, CRDs). Cost estimation was limited to epilepsy (most prevalent TSC disorder). Cost of lost productivity in patients since epilepsy onset in childhood and carried in adulthood as well as caregivers’ productivity costs were estimated. The Human Capital Method was adopted and potential lost working years estimated till an active age of 65 years. It was assumed a caregiver for all ages. Average income, unemployment rate due to epilepsy, productivity growth and epidemiology data were obtained from the literature and from the Brazilian Institute of Geography and Statistics. Mean annual productivity cost per patient, total productivity cost per patient and the total productivity burden of TSC-related epilepsy were calculated. Costs were estimated in 2008 Reals and discounted at 5%. Univariate sensitivity analysis was conducted for epidemiology data, employment status rate, productivity growth, discount rate and time horizon. RESULTS: Productivity loss was 47, 30 and 65 years for epilepsy onset in childhood and adulthood and caregivers, respectively. The discounted and not discounted mean annual productivity cost per patient were R$1,970 and R$11,323, total productivity cost per patient R$97,882 and R$691,150 and total productivity burden of TSC-related epilepsy in Brazil R$1,568,965,961 and R$11,078,536,271, respectively (1Euro = 3.24Reals). Results were sensitive to all the parameters varied in the sensitivity analysis, especially discount rate. CONCLUSIONS: TSC-related epilepsy is a chronic disorder associated with loss of productivity with a significant economic burden in Brazil. Although significant, the economic burden related to productivity loss is expected to be even higher since there are still costs deriving from absenteeism of patients and caregivers when employed looking for health services to be included in further analysis. PND11 NURSE COSTING SURVEY FOR THE MANAGEMENT OF ADVERSE EVENTS IN PATIENTS WITH MULTIPLE SCLEROSIS TREATED WITH DISEASE-MODIFYING DRUGS IN THE UK Perard R, Parkes L 1 Merck Serono Limited, Feltham, UK OBJECTIVES: To determine the costs and medical resource use for the management of injection-site reactions (ISRs) and flu-like symptoms (FLS) associated with injectable disease-modifying drugs (DMDs) used to treat patients with relapsing–remitting multiple sclerosis (RRMS) in the UK. METHODS: A survey was carried out amongst NHS nurses to understand the management of adverse events occurring with injectable DMD use. Data were collected via a postal questionnaire. Responses were compiled for the frequency and duration of side-effects, support received and medical consultations required for adverse event management. The unit costs were derived from the Personal Social Services Research Unit, the Office for National Statistics and the British National Formulary. Once all parameters and unit prices were characterised with conservative approaches for missing data (e.g. no GP consultation mandatory before antibiotics prescription), computations were processed to determine means and standard deviations (SDs). RESULTS: In total, 19 NHS nurses managing significant numbers of patients with MS for different primary care trusts participated in the costing survey. On average, the nurses reported that patients experienced 27 ISRs per year of those 33% experienced ISRs only at treatment start while 21% experienced ISRs continuously during the year. Seventy-four percent experienced FLS only at treatment start and 11% continuously. The average cost for the management of ISRs per year was £187 (SD £209) equivalent to an average cost of £7 per event. The average cost for the management of FLS per year was £166 (SD £144) equivalent to an average cost of £6 per event. CONCLUSIONS: This is the first UK study to provide data on the costs associated with management of ISRs and FLS occurring with injectable DMDs. These data, alongside prevalence estimates, enabled us to calculate the costs of managing adverse events associated within DMDs in an economic model. PND12 EUROPEAN-HUNTINGTON’S DISEASE BURDEN STUDY (EURO-HDB)— PRELIMINARY RESULTS FOR ITALY AND FRANCE Dorey J1, Toumi M2, Clay E1, Tedroff J3, Squitieri F4, De Nicola N4, Verny C5 1 Creativ Ceutical, Paris, France; 2University Claude Bernard Lyon 1, Lyon, France; 3 NEUROSEARCH, Ballerup, Denmark; 4Neurogenetics and Rare Disease Centre, Pozzilli, Italy; 5Neurology Unit CHU, Angers, France OBJECTIVES: Huntington’s disease (HD) is a rare neurodegenerative disease leading to sustained disability and poor health-related quality of life (HRQOL). As new treatments are in development for HD, data on the burden of disease are required. This study evaluated patient health status, patient and caregiver HRQoL and costs in HD. METHODS: Euro-HDB is a European cross-sectional survey being conducted in eight European countries. Self-reported questionnaires were completed by patients and caregivers. The patient questionnaire includes the Huntington Self-Assessment Instrument, a specific tool that assesses clinical characteristics, HRQoL and health care resource utilization. The EQ5D questionnaire and the SF36 Survey are also included. RESULTS: To date, 201 patients have been enrolled in France and 124 in Italy. All levels of disease severity are represented. Average annual costs from societal A390 perspective are c5,325 (±9588), c18,045c (±34487) and c41,716 (±35211) at mild, moderate and severe stages respectively in France; c3,125 (±4319), c8,457 (±18374) and c5,069 (±11747) in Italy. Associated EQ-5D utilities are respectively 0.79 (±0.18), 0.39 (±0.37), −0.11 (±0.3) in France; 0.59 (±0.44), 0.39 (±0.42), 0.25 (±0.44) in Italy. The primary cost driver is productivity loss. In France, hospitalization and nursing home costs are the main components of direct costs. For more severely affected patients, medical resource utilization diminishes while caregiver involvement increases significantly (the shift is greater in Italy than France). Physical, mental and social HRQoL domains are all seriously affected. CONCLUSIONS: Euro-HDB is the first study to comprehensively assess the cost and HRQOL burden of HD. The 1:5 cost ratio (Italy : France) is consistent across most of the cost items, suggesting that differences in health care systems, access to health care and cultural attitudes towards caring for patients at home have a large impact on a country’s overall costs. Results suggest that HD has greater impact on HRQOL than Parkinson’s disease and Alzheimer’s disease. PND13 A PROSPECTIVE STUDY OF THE FINANCIAL COSTS OF MULTIPLE SCLEROSIS AT DIFFERENT STAGES OF THE ILLNESS IN IRAN Khanizadeh H1, Izham M1, Akmal A1, Nikkhah K2, Ebrahimzadeh S2, Moshiri Tabrizi H3 1 University Sains Malaysia, Penang, Malaysia; 2Mashhad University of Medical Sciences, Mashhad, Khorasan, Iran; 3UKM, Kuala Lumpur, Malaysia The economic burden of Multiple Sclerosis (MS) on society and the individuals concerned is not known. Documenting such costs is essential for several reasons: costs of illness is a key factor of optimal disease management policies, knowledge of cost is useful for allocating research and development. The aim of our study as the first pharmacoeconomic investigation in Iran was to estimate the costs of multiple sclerosis according to severity of disease. Total, direct and indirect costs were compared in 160 patients divided into three groups categorized by disease severity: stage I Expanded Disability Status Scale (EDSS < 2.5), stage II (EDSS 3–4.5) and stage III (EDSS > 5). The majority of these patients (94%) developed relapsing- remitting MS. a minority of the patients (0.2–4%) developed secondary progressive and primary progressive MS. Cost evaluation was performed from the societal perspective and covered the one-year period. The study was carried out at the Division of Neurology at Ghaem Hospital and MS association in Mashhad in northeast of Iran and was approved by the local ethics committee. The mean total cost/patient for one year was estimated at $27,095, $27,997and $31,662 for stage I, II and III, respectively. Both direct and indirect costs increased with MS progression. For indirect cost the main item was productivity loss. The mean extra medicine (treatments for MS symptoms and adverse effects of medications) cost/patient for one year was calculated at $19,036. This study confirms that MS represents a high economic burden to patients and society, with direct costs greatly exceeding indirect costs. As costs increase with disease progression, treatment efforts should focus on patients in the early stages of MS. Disease support system that monitors a variety of common progressive signs for the MS individuals is a key element of a management program as well. PND14 SOCIO-ECONOMIC ASPECTS OF TESTING FOR NEUTRALIZING ANTIBODIES IN MS PATIENTS ON INTERFERON BETA TREATMENT IN AUSTRIA: A COST OF ILLNESS STUDY Walter E1, Brennig C1, Schöllbauer V1, Mair W2, Deisenhammer F3 1 Institute for Pharmaeconomic Research, Vienna, Austria; 2Biogen Idec GmbH, Vienna, Austria; 3University Hospital Innsbruck, Innsbruck, Austria OBJECTIVES: According to EU-guidelines testing all patients on interferon-beta (IFNb) for presence of neutralizing antibodies (NAB) is recommended irrespective of clinical course and stop IFNb or switch to alternatives in patients who developed persistent NAB; based on the fact that development in persistently NAB-positive patients equals that of placebo-treated patients. Economic impact of NAB-testing in MS-patients has not been explored yet. METHODS: This analysis estimated the economic impact of NAB-testing versus not testing during IFNb treatment in MSpatients (n = 3590) on Austria’s health care system. a cost of illness model (decision tree combined with an integrated Markov model, time horizon 5 years), based on the cohort of IFNb-treated patients was performed. Two alternatives were compared: Cost-effectiveness of NAB-testing versus no NAB-testing. The NAB-testing arm allowed switching to alternative therapies whereas no-NAB-testing did not. Direct costs comprised all treatment-costs of symptoms due to MS. Indirect costs were not included. All costs represent data from 2010 (discounted at 5%p.a). Clinical data and resource use were determined by literature/experts. Efficacy assessment was based on the outcome measure “relapses avoided.” RESULTS: Total discounted costs for all Austrian MS-patients on IFNb-therapy (incl. testing) from a health care system’s perspective amount to c187,764,180 for 5 years. Total costs for all MS-patients without testing amount to c176,331,610. The difference of costs between tested patients, and therefore switching, and not tested patients values c11,432,570. Considering all IFNb-treated patients and a time horizon of 5 years 1400 relapses can be avoided. Testing for NAB leads to costs per relapse avoided of c24,383p.a. versus c27,569p.a. when no tests are done resulting in a difference of c3,186 per patient in favour of NAB-testing. CONCLUSIONS: General NAB-testing in MS-patients on IFNb-therapy is reasonable and cost-effective. Patients switching to effective and more expensive alternatives do not account for higher health care costs. Furthermore, less relapses increase QoL. 13th Euro Abstracts PND15 PATIENT CHARACTERISTICS AND CHARGES ASSOCIATED WITH EMERGENCY DEPARTMENT VISITS AMONG PATIENTS WITH A DIAGNOSIS OF RESTLESS LEGS SYNDROME Meyers J, Candrilli S RTI Health Solutions, Research Triangle Park, NC, USA OBJECTIVES: Restless legs syndrome (RLS) affects 2 to 15% of the US population. Limited data exist on patient characteristics and charges associated with emergency department (ED) visits among patients with RLS. METHODS: Data from the 2007 Healthcare Cost and Utilization Project’s Nationwide Emergency Department Sample were used. Patients were selected for inclusion if they had a diagnosis of RLS (ICD9-CM code 333.94). Study measures included patient demographics and charges associated with the ED visit. Study measures were reported separately for patients with a primary versus secondary RLS diagnosis. Among patients with a secondary RLS diagnosis, the most common primary diagnoses were reported. RESULTS: A total of 6133 patients with a primary RLS diagnosis and 140,931 patients with a secondary RLS diagnosis were identified. Common primary diagnoses among patients with a secondary RLS diagnosis included respiratory symptoms (7.0%), general symptoms (4.6%), and pneumonia (3.7%). Mean (Std. Err.) age was 54.5 (0.6) years among patients with a primary diagnosis and 64.0 (0.3) years among patients with a secondary diagnosis. In both cohorts, over two-thirds of patients were female, the most common geographic regions were the South and Midwest, and Medicare was the most common primary payer (41.0% of patients with a primary diagnosis and 59.3% of patients with a secondary diagnosis). Over 90% of patients with a primary diagnosis had a routine discharge compared to only 26.6% of patients with a secondary diagnosis, and most patients with a secondary diagnosis were admitted to the facility as an inpatient. Mean (Std. Err.) charges were $816 ($48) for patients with a primary diagnosis and $2,043 ($62) for patients with a secondary diagnosis. CONCLUSIONS: This nationally representative study suggests that patients admitted to the ED with RLS accrue substantial costs during their visit. Further research is needed to more fully assess the total economic burden of the disease. PND16 MODELING THE CLINICAL AND ECONOMIC IMPLICATIONS OF GALANTAMINE IN THE TREATMENT OF MILD TO MODERATE ALZHEIMER’S DISEASE IN GERMANY Guo S1, Hernández L1, Wasiak R2, Gaudig M3 1 United BioSource Corporation, Lexington, MA, USA; 2United BioSource Corporation— Europe, London, UK; 3Janssen-Cilag, Neuss, Germany OBJECTIVES: A reimbursement policy issued by the Federal Joint Committee in Germany to reassess the benefit of cholinesterase inhibitors every six months in order to receive continued coverage by the Statutory Health Insurance triggered an additional need to periodically assess the cost-benefit of galantamine in the treatment of mild-to-moderate Alzheimer’s disease (AD). An economic model specifically designed for the purposes of such an assessment was developed using the most up-to-date IQWiG guidelines for cost-benefit assessment. METHODS: The model uses a discrete event simulation to predict the course of AD through changes in cognition, behavioral disturbance, and function over time, and compare the costs and benefits of galantamine versus no-drug treatment and ginkgo biloba. Clinical data were mainly derived from analyses of pooled data from clinical trials. Epidemiological and cost data were obtained from literature and public data sources. Costs (2009 c) from the perspective of the German Statutory Health Insurance were used. Both costs and benefits were discounted at 5%. Sensitivity analyses were performed to assess the robustness of the model outcomes. RESULTS: Over a 10-year period, galantamine on average delays time to severe stage of the disease by 3.57 and 3.36 months, compared to no-drug treatment and ginkgo biloba, respectively. Galantamine also reduces time spent institutionalized by 2.34 and 2.21 months, compared to no-drug treatment and ginkgo biloba, respectively. The use of galantamine is projected to yield net savings of c3,978 and c3,972 per patient compared to respective treatments. CONCLUSIONS: Our analyses suggest that compared to no-drug treatment and ginkgo biloba, treatment with galantamine not only improves clinical benefits, but also achieves savings in health care costs associated with care for patients with mild-to-moderate AD in Germany. PND17 MEMANTINE DELAYS THE ADMISSION OF ALZHEIMER’S DISEASE PATIENTS TO NURSING HOME: COST-EFFECTIVENESS ANALYSIS IN FRANCE Touchon J1, Lachaine J2, Beauchemin C2, Crochard A3, Rive B3, Bineau S3 1 Gui de Chauliac Hospital, Montpellier, France; 2University of Montreal, Montreal, QC, Canada; 3Lundbeck SAS, Issy-Les-Moulineaux, France OBJECTIVES: To evaluate in the French setting the cost-effectiveness of memantine as adjunct therapy to Cholinesterase inhibitors (ChEI) compared to ChEI monotherapy in Alzheimer’s disease (AD) patients. METHODS: A cost-effectiveness analysis employed a 3-state Markov model (“non-institutionalized,” “institutionalized” and “dead”) and compared the treatment alternatives in terms of time to nursing home admission, Quality Adjusted Life-years (QALYs), and costs over a 7-year time horizon. Annual transition probabilities between states were derived from two observational cohort studies: Lopez et al 2009 (US) for institutionalization probabilities and Helmer et al 2001 (FR) for death probabilities. Costs were valued from health care system and societal perspectives, and included cost of AD medications (French National 13th Euro Abstracts Health Insurance database, 2009), costs of care in community and in institution (French National Assembly on AD management, report 2005). Results were reported in EUR 2009. Health-related utilities were obtained from preceding published economic evaluations in AD (Getsios et al 2001). Costs and QALYs were discounted at annual rates of 0% (base-case analysis), 3% and 5%. Deterministic and probabilistic sensitivity analyses were carried out to test the robustness of model assumptions. RESULTS: Over the seven-year time horizon, patients treated with ChEI monotherapy spent on average 41.6 months before institutionalization. Overall costs were c72,469 (health care system perspective) or c89,735 (societal perspective). QALYs were estimated at 2.36. Memantine as adjunct therapy to ChEI was associated with a longer time to nursing home of 8.9 months, QALYs gains of 0.19 and a cost saving of c5900 (health care system perspective) or c2200 (societal perspective), i.e. a dominant treatment scenario versus ChEI monotherapy. CONCLUSIONS: This economic evaluation suggest that, from both a health care system and a societal perspective, memantine as adjunct therapy to ChEI is a cost-effective strategy in the management of AD patients compared with ChEI monotherapy. PND18 48-HOUR INFUSION OF METHYLPREDNISOLONE IS A COST-EFFECTIVE INTERVENTION FOR TRAUMATIC SPINAL CORD INJURY Schwartz G, Ruiz C, Ferreira JC, Rotstein OD, Hoch JS, Coyte PC University of Toronto, Toronto, ON, Canada OBJECTIVES: Methylprednisolone sodium succinate (MP) is an acute therapeutic option for traumatic spinal cord injury (SCI). a pivotal multicentre randomized control trial reported modest functional improvements and increased clinical complications associated with an extended dose regimen of MP for 48 hours (48h-MP) versus a limited dose regimen of MP for 24 hours (24h-MP), resulting in clinical ambiguity between 48h-MP and 24h-MP. Concerning the health care burden imposed by this devastating form of neurotrauma, an economic assessment comparing the benefits either MP regimen imparts has never been reported. We performed a cost-effectiveness analysis (CEA) of 48h-MP compared with 24h-MP to determine their impact on direct health care costs for this patient population. METHODS: A decision tree model, incorporating motor improvement and complication frequencies reported by the Third National Acute Spinal Cord Injury Study and utility scores (QALYs) obtained from an Australian cohort, measured outcomes and effects at 6 and 12 months post-injury. Survival data, direct health care expenditures and complication costs associated with SCI and MP intervention were obtained from published epidemiological and survey data. CEA was performed from the health care payer’s perspective, discounted at a rate of 4% annually with a lifetime horizon. Distributions of the incremental costeffectiveness ratio between the interventions were determined by Monte Carlo simulation. The model was validated with sensitivity analyses by varying costs and outcome comparators. RESULTS: As a result, 48h-MP dominates 24h-MP, providing higher QALYs at lower costs. The lower costs associated with 48h-MP intervention was $35,703 per patient lifecycle. Earlier motor improvement maintained at 1-year postinjury was a key variable favouring 48h-MP intervention, despite complications associated with this dosing regimen. CONCLUSIONS: To conclude, 48h-MP is the cost-effective intervention for SCI in comparison to 24h-MP, wherein the former results in modestly improved motor function, an effect which is maintained up to at least 1-year post-injury. PND19 A LONG-TERM COST-EFFECTIVENESS MARKOV MODEL COMPARING DISEASE MODIFYING TREATMENTS IN PATIENTS WITH RELAPSING REMITTING MULTIPLE SCLEROSIS IN GERMANY Plesnila-Frank C1, Putzki N2, Eheberg D1, Limmroth V3, Katsarava Z4, Patel SN5 1 IMS Health GmbH & Co. OHG, Munich, Germany; 2Cantonal Hospital, St.Gallen, Switzerland; 3Kliniken der Stadt Köln, Cologne, Germany; 4Universitätsklinikum, Essen, Germany; 5Biogen Idec GmbH, Ismaning, Germany OBJECTIVES: To conduct a German economic evaluation of natalizumab compared to other disease modifying drugs (DMD) in relapsing-remitting Multiple Sclerosis (MS) from a societal perspective. METHODS: A Markov model was designed to compare costs and outcomes of Natalizumab (Nb), other DMD (interferon-beta, glatiramer acetate) and best supportive care (BSC). The expanded disability status scale (EDSS) and the line of treatment were used to define the distinctive Markov States. Transition probabilities for progression, treatment switches and withdrawals were derived from clinical studies and literature. German real-life treatment data of MS-patients under DMD were collected retrospectively (N = 554) and used to validate assumptions and conduct sensitivity analyses. Cost data and quality of life estimates were taken from a European burden of MS study. a time horizon of 30 years and annual discount rates of 3% for costs and outcomes were chosen. RESULTS: Treatment with Nb resulted in 8.13 avoided relapses over 30 years, and in 2.36 avoided relapses under other DMD. After 30 years, the proportion of surviving patients at a low state of disease (incl. EDSS 4) was 59% for the Nb group, 31% for the other DMD and 8% for patients in the BSC group. The average MS related costs over 30 years were estimated at c847,160 for Nb, c816,139 for other DMD, and c627,701 for BSC. Cost per quality adjusted life-year (QALY) was c60,938 for Nb, c64,481 for other DMD and c53,911 for BSC. The incremental cost-effectiveness of Nb compared to other DMD was c24,919 per QALY. CONCLUSIONS: MS is a resource intense disease due to its chronic course and its severe impact on patients’ daily life. Long term analysis suggests that even treatment without DMD is expensive and leads to considerable inferior clinical outcomes. Treatment with DMD improves the situation of patients, with Natalizumab showing the highest efficacy and best cost-effectiveness ratio. A391 PND20 IS ROPINIROLE-PROLONGED RELEASE A COST-SAVING TREATMENT OPTION IN PARKINSON’S DISEASE? Novak A1, Boomsma M2 1 Anovák-Services, Apeldoorn, Gelderland, The Netherlands; 2GlaxoSmithKline bv Nederland, Zeist, The Netherlands OBJECTIVES: Parkinson’s Disease (PD) is both a chronic and progressive neurodegenerative disorder. a 24-hour prolonged release tablet (PR) of the dopamine agonist ropinirole was introduced next to three daily doses of ropinirole immediate release (IR). a randomized controlled trial (PREPARED) was conducted, comparing ropinirole-IR with ropinirole-PR. Ropinirole-PR significantly improved the off-time and this analysis assesses the costs-effectiveness of the ropinirole-PR in PD patients who are not adequately controlled on L-dopa compated to ropinirole-IR. METHODS: A Markov-health-state-transition model was used with health states combining off-time ≤25% and >25% per day, Hoehn & Yahr stages 2–5 and problematic dyskinesias. Time horizons are 5 years and lifelong. Costs and effects were discounted by 4% and 1,5% respectively. Healthcare perspective was taken, covering direct costs related to medication, consults, nursing and patient care including informal care, based on an ongoing Dutch observational study in PD (IMPACT study). Clinical outcomes from the PREPARED-trial are extrapolated based on literature assumptions. Results are presented as incremental costs and QALY’s gained. Both univariate and probabilistic sensitivity analyses (PSA) were performed. RESULTS: Ropinirole-PR was associated with lower L-dopa use, less off-time and less problematic dyskinesias. This resulted in incremental QALY gains of 0.125 and 0.336 over respectively 5 years and lifetime. The health care costs per H&Y-stage increased with disease severity and amounted c916, c1,492, c11,295 and c11.295 for stage 2 to 5 over 6 months. Treatment with ropinirole-PR was more costly than ropinirole-IR with a difference of c7.266 over 5 years and c17.773 over lifetime. Treatment with ropinirole-PR however reduced medical costs by c8,059 over 5 years and c69,532 over lifetime compared with ropinirol-IR, mainly due to reduced dyskinesia occurrence. Sensitivity analysis confirmed the robustness of the model. CONCLUSIONS: Patient-functioning and quality of life were improved with ropinirole-PR realizing cost-savings to the health care budget as compared to treatment with ropinirole-IR. PND21 COST-EFFECTIVENESS OF TRANSDERMAL PATCH (ROTIGOTINE) IN PATIENTS WITH PARKINSON DISEASE IN MEXICO Aguirre A, Benitez A, Bierschwale H UCB MEXICO, Mexico City, D.F., Mexico BACKGROUND: Parkinson’s Disease (PD) is a central nervous system disorder caused by progressive deterioration of brain areas that produce dopamine. Oral dopaminergic therapies control the symptoms of the disease, but these require three or more times daily doses, so it is associated with poor compliance or adherence, which affects the overall efficacy and costs in health. OBJECTIVES: To analyze the cost-effectiveness of rotigotine versus pramipexole in patients with PD in Mexico. METHODS: We conducted an economic evaluation. The alternatives to compare were rotigotine 4, 6, 8 and 12 mg administered once daily versus pramipexole 3 mg/d and another scenario versus pramipexole 4.5 mg/d. The perspective is the Mexican Social Security Institute. The model included the cost of drug acquisition and management of adverse events (AE) for a 22 weeks period. The measure of efficacy was compliance or adherence to treatment, as a direct comparison study of rotigotine versus pramipexole demonstrated non-inferiority between the two alternatives. RESULTS: The compliance rate for rotigotine was 81% vs. 61% pramipexole. The costs were US$748, US$920, US$1113 and US$1701 for rotigotine 4, 6, 8 and 12 mg/d respectively, compared with US$670 and US$967 for pramipexole 3 and 4.5 mg/d. The cost per successfully treated patient was lower for rotigotine 4, 6 and 8 mg/d (US$923, US$1 136 and US$1374, respectively) than with pramipexole 4.5 mg/d (US$1585). Rotigotine 4, 6, 8 and 12 mg/d were found to be a highly cost-effective strategy compared with pramipexole 3 and 4.5 mg/d, according to WHO criteria. CONCLUSIONS: The results of this analysis suggest that the use of rotigotine in patients with PD, represents a highly cost-effective strategy or cost saving for the public health institutions in Mexico. Rotigotine is an innovative alternative for easy administration (transdermal). PND22 COST-EFFECTIVENESS ANALYSIS COMPARING BRIDION® (SUGAMMADEX) WITH NEOSTIGMIN AND SPONTANEOUS RECOVERY IN THE REVERSAL OF NEUROMUSCULAR BLOCKADE INDUCED BY ROCURONIUM/VECURONIUM Ozdemir O1, Bahar M2, Aypar U3, Askar F4, Gura M5 1 Yorum Consultancy Co. Ltd., Istanbul, Turkey; 2Istanbul University Cerrahpaþa Medical Faculty, Istanbul, Turkey; 3Hacettepe University Medical Faculty, Ankara, Turkey; 4Ege University Medical Faculty, Izmir, Turkey; 5Goztepe Training and Research Hospital, Istanbul, Turkey OBJECTIVES: This study aimed to compare the cost-effectiveness (CE) of Bridion® (sugammadex) with neostigmine and spontaneous recovery (SR) approach in the reversal of neuromuscular blockade (NMB) induced by rocuronium/vecuronium, during anesthesia. METHODS: CE analysis (CEA) was performed by solving back the decision tree that included pathways starting with residual NMB and followed by hypoxia and pulmonary complications defined as “aspiration, atelectasis and/or pneumonia” in patients, in whom NMB was induced by rocuronium/vecuronium. Bridion was compared with neostigmine and SR approach. Primary analysis parameters that A392 were used as denominators in the calculation of incremental CE ratios (ICER) are decreased frequencies of mild hypoxia, severe hypoxia and pneumonia. RESULTS: Frequencies of residual NMB were estimated as 51.2% 44.6% and 0.4% in SR, neostigmine and Bridion groups, respectively. Hypoxia rate was calculated as 27.4% and 11.4% in patients with and without residual NMB. Frequencies of pneumonia were 17.2% and 5.9% in patients with and without residual NMB. Percentage of patients without any complication was found to be 88.0% in Bridion group, while it was 73.8% and 75.7% in SR and neostigmine groups. The costs of complications were 126.45c and 114.56c in SR and neostigmine groups, while it was only 34.93c in Bridion® group. When medication cost was added, the total cost rose to 114.62c in Bridion® group, which was still lower than the total costs of SR and neostigmine groups. The rate of severe hypoxia was estimated as 7.9% in Bridion® group, while it was 13.5% and 12.7% in SR and neostigmine groups, respectively. Pneumonia was seen in 1.2% of the patients in Bridion® group, while the frequencies of pneumonia were 9.4% and 8.3% in SR and neostigmine groups. CONCLUSIONS: Bridion® was found to be dominant to both SR and neostigmine groups, in terms of decrease in frequencies of hypoxia and pulmonary complications. PND23 COST-EFFECTIVENESS EVALUATION OF LEVODOPA/CARBIDOPA/ ENTACAPONE IN TREATMENT OF PARKINSON DISEASE Yagudina R, Kulikov AU, Serpik VG Moscow Medical Academy, Moscow, Russia OBJECTIVES: To conduct pharmacoeconomic evaluation of Levodopa/Carbidopa/ Entacapone (LCE) compared to standard therapy in Parkinson disease treatment. Standard therapy was presented by Levodopa and Carbidopa. METHODS: The costeffectiveness study was carried conducted. Effectiveness was measured in DALYs. Time horizont for cost-effectiveness analysis was 2 years. a 3% discount rate was used. 1 EUR = 40 RUB. RESULTS: LCE provided benefits in effectiveness compared to standard therapy. LCE reduced days incapacity index twice from 44 to 21 days. Two year pilot study showed that LCE provided 0,83 DALY, than standard therapy presented 0,5 DALY. Also administration of LCE reduced daily Levodopa dose on 68 mg. Total annual costs for LCE therapy were 158 938 (3 974 EUR) RUB. Standard therapy total costs varied from 129 113 RUB (3 228 EUR) to 145 422 RUB (3 636 EUR) yearly according to disease progression grade. Indirect costs, including GDP losses due to temporary disability and payments for temporary disability, were 14 372 RUB (359 EUR) for LCE and 30 113 RUB (753 EUR) for standard therapy. The costeffectiveness ratio for LCE was 191 492 RUB (4 787 EUR) per DALY, than for standard therapy it was 258 326 RUB (6 458 EUR) per DALY. CONCLUSIONS: LCE demonstrates lower cost-effectiveness ratio (191 492 RUB) compared to standard Parkinson disease therapy (258 326 RUB), therefore LCE has an advantage over standard therapy in terms of pharmacoeconomic evaluation. PND24 THE IMPACT OF COHORT SELECTION ON COST-EFFECTIVENESS RESULTS IN MULTIPLE SCLEROSIS Becker RV1, Dembek C2 1 Russell Becker Consulting, Chicago, IL, USA; 2Biogen Idec GmbH, Wellesley, MA, USA OBJECTIVES: Adherence to cost-effectiveness analytic guidelines requires careful assessment of clinical trial results to ensure the most appropriate cohort data selection. This study examines the impact of cohort selection on the results of the Goldberg, et al. 2009 cost-effectiveness study of disease-modifying therapies (DMTs) in multiple sclerosis (MS). METHODS: Using intent–to-treat (ITT) two-year data from pivotal trials, Goldberg compared cost per relapse avoided of first-line DMTs. However, there are important differences in the ITT cohorts among the trials. Due to lower than expected subject drop-out rates, the interferon beta-1a IM (INF-b-1a-IM) trial was able to meet its primary endpoint with a reduced sample size and a decision was made to terminate the trial early. This resulted in almost half of the 301 ITT (or “allpatient”) cohort being on study drug for less than two years. a second, “two-year” 172 patient cohort consisted of patients who had completed two years on drug therapy. Goldberg used the INF-b-1a-IM “all-patient” cohort in his cost study. To test the impact of this cohort selection, we recreated Goldberg’s model using the “allpatient” relapse rate (0.67 for INF-b-1a-IM vs. 0.82 for placebo) and then substituted the “two-year” cohort data (relapse rate of 0.61 for INF-b-1a-IM vs. 0.90 for placebo) and compared results. RESULTS: This study’s cost per relapse avoided was 45% lower for INF-b-1a-IM while the results for the other DMTs were comparable to those reported by Goldberg. Ranked from most to least cost-effective, the model results were $77,980 for INF-b-1a-IM, $80,121 for interferon beta-1a SC, $86,572 for interferon beta-1b, and $87,767 for glatiramer. CONCLUSIONS: Our analysis demonstrates the importance of cohort selection in cost-effectiveness analyses in MS. By selecting comparable cohort data from patients who had completed at least two years on study drug, we found that INF-b-1a-IM had the lowest cost per relapse avoided of all firstline DMTs. PND25 ECONOMIC EVALUATION OF NATALIZUMAB IN THE TREATMENT OF RELAPSING REMITTING MULTIPLE SCLEROSIS IN 4 EUROPEAN COUNTRIES Brennig C, Walter E, Schöllbauer V Institute for Pharmaeconomic Research, Vienna, Austria OBJECTIVES: Multiple sclerosis (MS) is an inflammatory, neurodegenerative condition of central nervous system with prevalence of 1 Million worldwide (age at first 13th Euro Abstracts manifestation 20–40 years). MS may lead to permanent disability and early retirement even in young adults. The purpose of this analysis was to provide an economic assessment of relapsing-remitting MS-treatment with Natalizumab versus Interferons and Copolymer in 4 European countries (Austria, Czech Republic, Slovakia, Slovenia). METHODS: The analysis was performed using a Markov model. The model allows a treatment switch due to relapse. Efficacy assessment was based on the outcome measure “relapse-free patients”. Costs were captured for the year 2010. Resource use was determined from results of a survey conducted by the Austrian MS Society and accurately reflects the therapeutic approach. Resource use of Czech Republic, Slovakia and Slovenia was determined via country-specific research. Where country specific data could not be captured, Austrian data was adjusted via Purchasing Power Parities (PPP). The study time horizon was 2 years. The analysis was performed from the perspective of the health care system of the 4 European countries. RESULTS: In Czech Republic the average cost of the therapy algorithm Natalizumab amount to c33,835 per patient within the time horizon of 2 years versus c22,896 (Interferon) and c21,256 (Copolymer). a patient successfully treated with Natalizumab accounts for c56,525 compared to c70,190 (Interferon) and for c87,555 (Copolymer). In Slovakia a relapsefree patient values c60,357 (Natalizumab) versus c58,043 (Interferon) and c19,921 (Copolymer). a relapse-free patient in Slovenia values c59,108 (Natalizumab) versus c66,202 (Interferon) and c84,207 (Copolymer). In Austria a relapse-free patient values c58,423 (Natalizumab) versus c71,906 (Interferon) and c89,591 (Copolymer). CONCLUSIONS: In Czech Republic, Slovenia, and Austria, treatment of MS with Natalizumab is more cost-effective than Interferon resp. Copolymer 1 therapy. Switching to effective and more expensive alternatives does not account for higher health care costs. PND26 MEDICO-ECONOMIC EVALUATION OF LACOSAMIDE ADJUNCTIVE THERAPY IN THE TREATMENT OF PATIENTS WITH REFRACTORY EPILEPSY IN SCOTLAND AND SPAIN Benhaddi H1, Gunn A1, Ferro B2 1 UCB Pharma S.A., Brussels, Belgium; 2UCB Pharma S.A., Madrid, Spain OBJECTIVES: To calculate and compare the incremental cost-utility ratios for standard antiepileptic drug (AED) therapy with and without adjunctive lacosamide in patients with uncontrolled partial-onset seizures. METHODS: The model simulated the treatment pathway of a hypothetical cohort of 1000 patients over 2 years from the perspectives of the National Health Service (NHS) in Scotland and the Spanish Healthcare System (SNS) in 2008. a decision tree was split into four phases of six months each during which patients can become seizure free, experience a seizure reduction (responder defined as ≥50% reduction in seizures), or withdraw due to non-response. The standard therapy arm included carbamazepine, lamotrigine, levetiracetam, topiramate, and valproate. The likelihood of being in a particular health state has been estimated from clinical data. The cost of general practitioner visits, outpatient visits, hospitalizations and emergency department visits were included. Costs and utility values attached to various health states were taken from the published literature. RESULTS: Lacosamide adjunctive therapy was associated with 6730 avoided seizures and a gain of 38 quality adjusted life-years (QALYs), compared to the standard therapy arm within the two-year timeframe. Treatment with lacosamide was associated with a cost of £113 and c107 per seizure avoided, and £20,017 and c22,771 per QALY gained versus standard therapy in Scotland and Spain, respectively. Results calculated for 6-, 12- and 18-month follow-up showed respective incremental cost-utility ratios of £23,479, £21,422 and £20,998 in Scotland, and c23,771, c20,703 and c21,778 in Spain. Using a willingness-to-pay threshold of £30,000 per QALY, 80% of the simulations in Scotland and 74.2% in Spain fell below this value after 2 years of treatment. CONCLUSIONS: Lacosamide was shown to be a cost-effective adjunctive treatment in patients with uncontrolled partial-onset epilepsy in Scotland and Spain. PND27 COST UTILITY ANALYSIS OF ORPHAN DRUGS: CASE STUDY OF DUODENAL LEVODOPA INFUSION VERSUS STANDARD TREATMENT IN PATIENTS WITH ADVANCED PARKINSON’S DISEASE IN SWEDEN Willis M1, Gradl B2 1 IHE, Lund, Sweden; 2Abbott Products Operations AG, Allschwil, Switzerland BACKGROUND: Advanced Parkinson’s disease (APD) severely impacts the quality of life (QoL) for both patients and their caregiver and is associated with high health care and societal costs. The treatment options for those patients are limited. Duodenal levodopa infusion (DLI), an orphan drug, has shown to restore symptom control and to improve QoL substantially for both patients and their carers. OBJECTIVES: To evaluate the cost-utility of DLI versus standard treatment (including oral treatment and subcutaneous dopamine agonists) in patients with advanced PD in Sweden. METHODS: A stochastic Markov-based simulation model was developed. Health was described by 12 health states reflecting 4 categories of “OFF” time and 3 severity stages as measured by the Hoehn & Yahr scale. The clinical effect for both treatment arms was derived from published clinical trials; costs and QoL utilities were taken from an interim analysis of the ongoing DAPHNE study. The base case simulated 5 years with costs/effects discounted by 3%. The societal perspective was adopted. RESULTS: The model estimated an improvement from 0.68 to 1.30 QALYs and a cost increase from 1,410,643 (c147,108) to SEK 1,674,295 (c174,603) for DLI versus standard treatment, leading to an incremental cost-effectiveness ratio (ICER) of around SEK 420,000 (c43,800) per QALY gained. The ICER for nearly 90% of the cohorts fell below SEK 655,000 (c68,306), the willingness-to-pay threshold often cited 13th Euro Abstracts for Sweden [1]. Sensitivity analysis was associated with ICERs ranging up to SEK 900,000 (c93,856) per QALY gained. CONCLUSIONS: Due to the data limitations, HE modeling in the orphan drug setting is challenging. Analysis could be performed as requested by Sweden’s Dental and Pharmaceutical Benefits Agency (TLV), however, providing evidence that health benefits can provide good value for money even for an orphan population. DLI received a positive reimbursement recommendation by the TLV. [1] Persson U, Hjelmgren J (2003). PND28 COST-UTILITY ANALYSIS OF ROTIGOTINE TRANSDERMAL PATCH IN EARLY-STAGE PARKINSON’S DISEASE IN SCOTLAND Benhaddi H, Gunn A UCB Pharma S.A., Brussels, Belgium OBJECTIVES: To evaluate the cost-effectiveness of rotigotine transdermal patch as monotherapy in early-stage Parkinson’s disease (PD) compared to ropinirole and other dopamine agonists (DA) from the NHS perspective in Scotland. METHODS: a decision-analytic model was developed, based on treatment of an early-stage PD patient (Hoehn and Yahr-stage 2); treatment arms were rotigotine, ropinirole, and a DA practice comparator including ropinirole, cabergoline and pramipexole. 5-year and 10-year time horizons were considered for patients who remained on monotherapy only. The economic evaluation is a cost-utility analysis with health outcomes expressed in Quality Adjusted Life-years (QALY) gained in 2006. Costs relating to drug acquisition, PD severity, falls, occurrence of motor complications and other complications/ adverse events/co-morbidities were considered in the model. Efficacy and safety data were estimated with a meta-analysis. Quality of life was measured using EQ-5D. Data on medical resource use was obtained via expert interviews and literature review. Costs and effects were discounted at the rate of 3.5%. RESULTS: After 5 years, treatment with rotigotine transdermal patch resulted in an estimated 2.30 QALYs, slightly higher than with ropinirole (2.26) and the DA practice comparator (2.27). 10-year outcomes were 3.22, 3.17 and 3.17 QALYs for rotigotine, ropinirole and DA practice comparator, respectively. Total costs for rotigotine, ropinirole and DA practice comparator were £34,748, £37,694 and £36,459 respectively after 5 years and £79,477, £84,120 and £81,631, respectively after 10 years. With a willingness-to-pay of £20,000 per QALY gained, there is a 90% probability that rotigotine is cost-effective relative to ropinirole, and a 85% probability of cost-effectiveness relative to the DA practice comparator for both 5-year and 10-year time horizons. CONCLUSIONS: Based on the model, rotigotine may be considered a dominant strategy over ropinirole and DA practice comparator in the treatment of early-stage PD at 5-year and 10-year time horizons. PND29 COST-UTILITY ANALYSIS OF RIZATRIPTAN VERSUS (GENERIC) SUMATRIPTAN IN SWEDEN Lundberg J1, Golden WM2, Insinga RP3 1 MSD Sweden, Sollentuna, Sweden; 2Merck & Co., Inc., Whitehouse Station, NJ, USA; 3 Merck & Co., Inc., Upper Gwynedd, PA, USA OBJECTIVES: In 2005 the Swedish pharmaceutical benefits board published a review concluding that rizatriptan 10 mg was one of the most cost-effective triptans for treating migraine, although efficacy differences among the triptan class were generally small. However, since the review, the price of sumatriptan has declined due to generic product entry. In this study, we sought to investigate the cost-effectiveness of rizatriptan versus generic sumatriptan. METHODS: A published Canadian decision-analytic model (Thompson et al., Pharmacoeconomics 2005) was adapted to estimate treatment costs and effects of rizatriptan 10 mg versus (generic) sumatriptan in a single migraine attack over a 24-hour time-frame in Sweden. Values modified from the published model were the substitution of Swedish health care and productivity costs, and updated data on 2-hour and 2–24 hour pain freedom (including results from more recent head-to-head trials of the comparators). As both sumatriptan 50 mg and 100 mg are marketed in Sweden, and have comparable efficacy profiles, averaged data for the two doses were used as a comparator to rizatriptan 10 mg. RESULTS: From a societal perspective, including both health care and productivity costs, the comparison of rizatriptan 10 mg versus (generic) sumatriptan resulted in cost-savings of ~ 5 SEK, and greater QALYs for rizatriptan 10 mg. Inclusion of health care costs only, yielded a cost-effectiveness ratio for rizatriptan vs. sumatriptan of 343,929 SEK/ QALY. CONCLUSIONS: Given the data and assumptions used, in a single attack model, rizatriptan 10 mg was found to have a cost-effectiveness ratio compared to (generic) sumatriptan within the range typically regarded as cost-effective in Sweden. These findings suggest that there is reason to explore the cost-effectiveness of pharmaceutical and other medical interventions with large differences in acquisition cost even though efficacy differences may appear to be small. PND30 COST-UTILITY ANALYSIS OF LACOSAMIDE ADJUNCTIVE THERAPY IN THE TREATMENT OF PARTIAL-ONSET SEIZURES IN EPILEPTIC PATIENTS IN BELGIUM Simoens S1, Dedeken P2, De Naeyer L2, Benhaddi H2 1 K.U. Leuven, Leuven, Belgium; 2UCB Pharma S.A., Brussels, Belgium OBJECTIVES: This study aims at evaluating the incremental cost-effectiveness ratio (ICER) for lacosamide compared with standard therapy alone from the perspective of the Belgian health care payer in 2010. METHODS: A decision tree simulating the treatment pathway of a hypothetical cohort of 1,000 patients over two years was split A393 into four phases of six months each during which patients can become seizure free, experience a seizure reduction (defined as ≥50% reduction in seizures), or withdraw due to non-response. The antiepileptic drugs (AEDs) included in the standard therapy arm were extracted from the pivotal trials and included carbamazepine, lamotrigine, levetiracetam, topiramate and valproate. Health state probabilities, seizure frequency and utility values were taken from lacosamide trials or from the literature. Costs of general practitioner visits, outpatient visits, hospitalizations and emergency department visits were included. Resource use was estimated by a Belgian panel of eight neurologists. Costs were discounted at a rate of 3% and consequences at a rate of 1.5%. RESULTS: Over a 24-month period, standard AED therapy plus lacosamide led to a reduction of 7 seizures, an increase of 0.038 quality-adjusted life-years, and a cost decrease of c3619 per patient as compared with standard therapy alone. Results were also calculated for a 6-, 12- and 18-month follow-up. Lacosamide plus AED therapy dominated versus standard therapy alone. Using a willingness to pay of c30,000 per quality-adjusted life-year, the net monetary benefit of standard antiepileptic drug therapy plus lacosamide amounted to c4,754. The probability of standard AED therapy plus lacosamide being cost-effective was 97.3% at 6 months, 99.8% at 12 months, 99.9% at 18 months, and 100% at 24 months. CONCLUSIONS: In epileptic patients who are difficult to treat with other AEDs, standard AED therapy plus lacosamide appears to be a cost-effective alternative. NEUROLOGICAL DISORDERS – Patient-Reported Outcomes Studies PND31 IMPACT OF AN ADHERENCE PROGRAM, RUN AS A TELEPHONE INTERVENTION ON COMPLIANCE WITH SUBCUTANEOUS INTERFERON Â-1A FOR MULTIPLE SCLEROSIS PATIENTS USING A MAIL-ORDER PHARMACY Papademetriou E1, Levin R1, Tuennermann M2, Lammers V2, Aubert RE1 1 Medco Health Solutions, Franklin Lakes, NJ, USA; 2Europa Apotheek Venlo B.V., Venlo, The Netherlands OBJECTIVES: Quasi-experimental analysis to determine the effect of an opt-in telephone intervention on adherence in patients using subcutaneous interferon β-1a, which is indicated for relapsing-remitting multiple sclerosis. Drop-out reasons, sideeffects, and expectations of therapy are described. METHODS: Customers of a mailorder pharmacy that services Germany who ordered subcutaneous interferon β-1a were targeted for enrollment in a free program that included an initial counseling call, optional e-mail reminders for the next doctor’s consultation and prescription, and ongoing counseling calls. Patients enrolled in the program for at least 6-months were included in the analysis and compared to patients that did not enter the program over the same time period. Proportion of days covered (PDC) was calculated for each group and compared using analysis of variance. Enrollees in the program were administered a questionnaire at the initial welcome call addressing their expectations of therapy, and again during each counseling call regarding their compliance behavior, and side effects. RESULTS: Patients in the adherence program showed an unadjusted PDC 8.2% higher than the control, F(1.247) = 13.44, p = 0.0003. One program patient switched drugs compared to six control group patients. a total of 21% of enrolled subjects reported missing at least one dose. Side-effects included pain/inflammation at the site of injection (24.41%), fatigue (20.73%), headaches (17.06%), and flu-like symptoms (9.71%). Exacerbations were reported by 15.7% of patients. Patients’ had expectations that therapy would prolong the distance between exacerbations (23.36%) and slow the progression of disability due to the disease (21.28%). CONCLUSIONS: Actively recruiting patients into an optional adherence program significantly increased the compliance rate for relapsing-remitting multiple sclerosis patients using subcutaneous interferon β-1a. Side effects experienced by enrolled patients were consistent with the package insert. Limitations include a potential bias between patients that agree to the program vs. those that do not, as well as the lack of additional questionnaire data from the control group. PND32 INFLUENCE OF AGE ON REFILL-ADHERENCE RATES OF ANTIEPILEPTIC DRUGS IN SOUTH AFRICA Van Zyl T, Lubbe MS, Serfontein JHP, Rakumakoe DM North-West University, Potchefstroom, South Africa OBJECTIVES: To investigate the possible influence of age on the refill-based adherence rates of anti-epileptic drugs. METHODS: A retrospective drug utilization review was performed on medicine claims data of a pharmacy benefit management company in South Africa. Refill-based adherence rates were calculated for 64,457 anti-epileptic drugs that were prescribed more than once during a four-year period (January 1, 2005 to December 2008). The refill-based adherence rate was calculated per trade name by using the following equation: Refill-Adherence rate = (total number of days of antiepileptic drugs supplied—days supplied at the last refill)/(date last claimed—date first claimed). [RSA Rand (R)/$US = 6.38112 (2005); 6.78812 (2006); 7.06926 (2007) and 8.27505 (2008)]. RESULTS: Only 30.5% (n = 19 635) of anti-epileptic drugs had refill-adherence rates between between 90% and 110%. The majority of anti-epileptic drugs (58.9%; n = 37 962) had refill-adherence rates below 90% that accounted for 39.2% (n = R 57 599 838) of the total cost of all anti-epileptic drugs (N = R146 863 755) included in these calculations. Anti-epileptic drugs with refill-adherence rates >110% (10,7%; n = 6 860) accounted for 6.7% (R9 782 864) of the total cost of all anti-epileptic drugs. The average refill-adherence rate decreased with nearly 10% from A394 93.7% in age group 1 (0 ≤ 12 years) to 83.3% in age group 5 (older than 65 years). In general the lowest average refill-based adherence rates were obtained with medicine items containing phenobarbitone and vitamin B1 (52.0% ± 37.8); phenobarbitone (63.5% ± 47.2); primidone metabolites (69.7% ± 47.2%); clonazepam (77.8% ± 184.8) and carbamazepine (80.9% ± 151.1) CONCLUSIONS: Most of the anti-epileptic drugs had unacceptable low refill-adherence rates. Refill-adherence rates of anti-epileptic drugs decreased with an increase in the age of patients. PND33 REFILL-ADHERENCE RATES OF ANTIPARKINSON MEDICATION IN THE PRIVATE HEALTH CARE SECTOR OF SOUTH AFRICA Van der Merwe S, Lubbe MS, Du Plessis JM, Bekker E North-West University, Potchefstroom, South Africa OBJECTIVES: To investigate the prevalence of unacceptable refill-based adherence rates with antiparkinson medicine items. METHODS: A retrospective drug utilization study was performed on medicine claims data of a pharmacy benefit management company in South Africa during January 1, 2005 until December 31, 2008. Refillbased adherence rates were calculated for 8 768 antiparkinson medicine items that were prescribed more than once during a four-year period (January 1, 2005 to December 2008). The refill-based adherence rate was calculated per trade name by using the following equation: Refill-Adherence rate = (total number of days of antiparkinson medicine items supplied—days supplied at the last refill)/(date last claimed—date first claimed). (RSA Rand(R)/$US = 6.8595 on 31 Dec. 2007). [RSA Rand (R)/$US = 6.38112 (2005); 6.78812 (2006); 7.06926 (2007) and 8.27505 (2008)] RESULTS: A majority of antiparkinson medicine items (53.50%, n = 4,691) had unacceptably low refill-adherence rates below 90%, that accounted for 41.62% (n = R16,398,512.00) of the total cost (N = R39,402,898.20) of all antiparkinson medicine items included in this study. Only 36.78% (n = 3225) of antiparkinson medicine items had acceptable refill-adherence rates between 90% and 110%. Those with unacceptable high refilladherence rates accounted for 9.72% (n = 852) of all antiparkinson medicine items and represented 6.5% (n = R2,574,597) of the total cost. No practical significant difference in the average refill-adherence rates was found between male (93.99% ± 186.99) and female (90.83% ± 175.21) patients. Biperidien, carbidopa/levodopa, and levodopa/benserazide containing products had on average unacceptable low refilladherence rates (<90%). CONCLUSIONS: Although poor obedience to treatment schedules adds up to aggravation of Parkinson’s disease leading to death and amplified health care cost, it seems that the refill-adherence rate of antiparkinson medicine items is not very favourable PND34 A PSYCHOMETRIC EVALUATION OF THE REVISED SCOPA DIARY CARD IN PARKINSON’S DISEASE PATIENTS Buck PO1, Castelli-Haley J1, White RE1, Rendas-Baum R2, White MK2 1 Teva Neuroscience, Kansas City, MO, USA; 2QualityMetric Incorporated, Lincoln, RI, USA OBJECTIVES: The Scales for Outcomes in Patients with Parkinson’s disease Diary Card (SCOPA-DC) is a daily diary designed to measure motor impairment in Parkinson’s disease (PD) patients with fluctuating symptoms. Previous qualitative research evaluated the content validity of the SCOPA-DC in the US and expanded it to measure non-motor symptoms. The current research examined the psychometric properties of the revised SCOPA-DC. METHODS: A sample of adults age 30 and older with selfreported doctor-confirmed PD were recruited, screened, and consented online from a Knowledge Networks panel. Eligible patients were mailed a study packet that contained the revised SCOPA-DC as well as a training video. The revised SCOPA-DC included 7 non-motor items: fatigue, memory, anxiety, pain, difficulty swallowing, frequent urination, and sweating. The diary was completed 7 times per day for 3 consecutive days. Consistent with the original SCOPA-DC, 3-day scores were calculated for each item. Higher scores indicated greater symptom severity. RESULTS: A total of 101 PD patients completed and returned the revised SCOPA-DC. The sample was 50.5% male and had been diagnosed with PD for an average of 7.4 years. Frequency distributions showed little missing data (approximately 1.0%), although items were generally right-skewed. Fatigue (29.4) and walking (28.7) had the highest mean scores; sweating (7.3) and difficulty swallowing (9.7) had the lowest mean scores. Factor analysis supported a 3-factor solution: mobility, physical functioning, and psychological functioning. These factors demonstrated good internal consistency (alpha = 0.83–0.87) and correlations with health-related quality of life instruments were suggestive of construct validity. CONCLUSIONS: In this US sample of PD patients with varied disease severity, the revised SCOPA-DC exhibited good psychometric properties, including evidence of reliability and validity. Furthermore, patients reported that the revised SCOPA-DC was clear and easy to complete. The revised SCOPA-DC holds promise for measuring a broad spectrum of fluctuating motor and non-motor PD symptoms. PND35 COMPARISON OF ANALYIC HIERARCHY PROCESS AND CONJOINT ANALYSIS METHODS IN ASSESSING TREATMENT ALTERNATIVES IN STROKE REHABILITATION Ijzerman M1, Bridges JF2, Van Til J1 1 University of Twente, Enschede, The Netherlands; 2Johns Hopkins University, Baltimore, MD, USA OBJECTIVES: There has been increasing interest novel HTA methods that will incorporate patient preferences in a more transparent and scientifically valid way. The fundamental problem of the assessment of benefits in HTA is the identification, 13th Euro Abstracts ranking and valuation of multiple health care outcomes. We used two multi-criteria methods to rank and value five different treatments in stroke rehabilitation. Analytic Hierarchy Process (AHP) stems from operations research and is increasingly being used in health care to weigh patient-reported endpoints. Conjoint analysis (CA) is a stated preference method that often takes the discrete choice format. In CA, hypothetical scenarios are used to generate part-worth utilities for attributes. METHODS: To determine the clinical decision context and related criteria, a paper-and-pencil questionnaire was conducted among a sample of Dutch physiatrists united in a stroke interest group. From the lists of criteria (e.g. clinical benefit, impact of treatment) an expert panel defined the AHP decision structure as well as the conjoint analysis survey format. Finally, the complete questionnaire including the AHP and CA survey was sent out to 184 patients with ankle-foot impairments. Eventually, 89 patients completed both surveys. RESULTS: On average, the prediction of preferred treatment on a group level is similar for both AHP and CA. However. on an inidividual level there seems to be more variation in treatment preference. Using AHP weights, a vast majority preferred soft-tissue surgery where most patients preferred orthopedic shoes if CA weights were used. This may have been caused by labelling effects of the attributes. CONCLUSIONS: Both methods have there pros and cons in ranking and valuing patient-reported endpoints. Of the methods AHP is relatively easy to apply. In prediction of overall outcome, both methods perform equally. However, for individual treatment preference we observed some differences. It may be concluded that the decision structure, framing and labelling of the treatment attributes are more important than the specific elicitation method used. PND36 HEALTH STATUS COMPARISON BETWEEN STABLE PARKINSON’S DISEASE PATIENTS AND THOSE EXPERIENCING OFF-TIME Buck PO1, White RE1, Castelli-Haley J1, Rendas-Baum R2, White MK2 1 Teva Neuroscience, Kansas City, MO, USA; 2QualityMetric Incorporated, Lincoln, RI, USA OBJECTIVES: End-of-dose wearing-off is commonly experienced by Parkinson’s disease (PD) patients who have used dopaminergic therapy for several years. Although investigations of wearing-off have traditionally focused on motor fluctuations, it is increasingly recognized that non-motor symptoms also vary between periods of “ON” (when PD symptoms are minimized due to medication) and “OFF” (when PD symptoms return). This study characterizes the self-reported health status of PD patients who experienced OFF-time as compared to those who were stable. METHODS: Recruited from an online panel maintained by Knowledge Networks, adults with self-reported doctor-confirmed PD were screened, consented, and completed a crosssectional survey. Frequency of OFF-time was measured using the Unified Parkinson’s Disease Rating Scale Part IV. Demographics, PD-specific characteristics, the 9-item Wearing-off Questionnaire (WOQ-9), the Short Form-12v2 (SF-12), and the Parkinson’s Disease Questionnaire-8 (PDQ-8) were also assessed. RESULTS: Data were available for 165 PD patients (mean age = 66.6 years; 52.7% male; mean time from diagnosis = 7.1 years). Twenty-five (15%) of the patients reported experiencing no OFF-time on a typical day and were classified as stable; the remaining 85% reported experiencing OFF-time. There were few significant differences between the two groups in terms of demographics and PD history. Compared to those experiencing OFF-time, stable patients reported fewer motor and non-motor wearing-off symptoms based on the WOQ-9 (P < 0.05), as well as better health on the Physical and Mental Component Summary scores of the SF-12 (P < 0.05) and the Summary Index score of the PDQ-8 (P < 0.01). CONCLUSIONS: PD patients who experienced at least some OFF-time on a typical day reported worse overall physical and mental well-being than stable patients. Furthermore, both motor and non-motor wearing-off symptoms differed between the two patient groups. Additional research to understand the consequences of OFF-time would be useful, especially as it pertains to non-motor symptoms. PND37 PATIENT AND PHYSICIAN GLOBAL PERCEPTION OF LEVODOPA/ CARBIDOPA/ENTACAPONE VS. LEVODOPA/CARBIDOPA IN PATIENTS WITH PARKINSON’S DISEASE EXPERIENCING EARLY WEARING-OFF Ruyra J1, Balañá M1, Lahoz R1, Hernandez B2, Tolosa E2 1 Novartis Farmacéutica S.A., Barcelona, Spain; 2Hospital Clínic de Barcelona, Barcelona, Spain OBJECTIVES: To compare patients’ and physicians’ global perceptions of Parkinson’s disease (PD) in two treatment groups: levodopa/carbidopa/entacapone (LCE) vs. levodopa/carbidopa (LC). METHODS: Multicentre, double-blind, randomised phase IV study. Ninety-five PD patients with early wearing-off (WO) and deterioration of activities of daily living (ADLs) were randomised to receive LCE (n = 46) or LC (n = 49) with a 3-month follow-up. Patient and physician global perception of PD was assessed at the end of the study. The PDQ-39 quality of life (QoL) questionnaire, and the longitudinal course of PD using the different parts of UPDRS (part I, part II, part III, and IV) were evaluated along the study. Differences between health improvement by patient and physician were analyzed by the Mann Whitney U-test. The mean differences from baseline to final visit in PDQ-39 and in UPDRS (part I, II, IIII IV) score were analyzed by an ANCOVA model. RESULTS: Mean (SD) age was 66.4 ± 8.6 years and 50.0% were women. Half percent of patients showed stage II according to the Hoehn and Yahr classification. Patient global perception showed a significant better score in the LCE than in LC group (−0.9 ± 1.0 LCE and −0.4 ± 1.17 LC, p = 0. 0291). Similar results were obtained by the physician (−0.3 ± 0.8 vs. −0.4, p = 0. 0017). The adjusted mean differences in the PDQ-39 showed a trend for a higher improvement in QoL in the LCE group (6.3 ± 20.4 vs. 0.81 ± 15.6), although did not reach statistical significance. The UPDRS evaluation showed a significant higher 13th Euro Abstracts response in part II and III (p = 0.0078 and p = 0.0072, respectively). a trend to better results for the LCE group was observed in part I and IV. CONCLUSIONS: Levodopa/ carbidopa/entacapone shows a significant better patient and physician global perception in the LCE treatment group vs. LC group in PD patients with early WO and ADLs deterioration. PND38 DIFFICULTIES IN IDENTIFYING THE ORIGINAL SOURCE QUESTIONNAIRE FOR USE IN TRANSLATIONS: THE ADAS-COG CASE STUDY Anfray C1, Giroudet C1, Berne C2, Acquadro C2 1 MAPI Institute, Lyon, France; 2MAPI Research Trust, Lyon, France OBJECTIVES: Since its development in the 80’s, variations of the Alzheimer’s Disease Assessment Scale-Cognitive subscale (ADAS-Cog), a Clinician-Reported Outcome (ClinRO) measure, have been used to monitor disease progression and treatment efficacy in Alzheimer’s disease. The objective of this study was to identify all versions used as a basis for translation in Mapi Institute projects and to take stock of existing translations. METHODS: The review was based on all ADAS-Cog translation projects performed by Mapi Institute. RESULTS: Sixteen projects were identified representing a total of 70 languages and 219 translations. Translations were based on 11 source versions which differed in terms of content (number of items, order of items and instructions), and format. The number of items ranged from 11 to 15. Four studies used 13 items, but only in two cases the same items were used although in a different order. Four studies used 12 items: only two studies used the same items (with a different list of words for the Word Recognition Task), but again in a different order. Format and instructions differed in all cases. In most projects the source version provided by the sponsor was a single document mixing instructions with the rater and response forms. Only in 3 cases the original consisted in a separate instruction manual and response forms. With regard to available translations, more than one translation was identified in 56 of the 70 available languages and in one language (Swedish) as many as 7 translations. CONCLUSIONS: The abundance of different versions of the same questionnaire both in its original US English form as in translations makes comparisons between studies or pooling of data difficult for both researchers and users. In the light of FDA’s recent PRO guidance it would be beneficial to demand the same scientific rigor when using ClinROs in international studies. PND39 RESPONSIVENESS OF THE MULTIPLE SCLEROSIS INTERNATIONAL QUALITY OF LIFE QUESTIONNAIRE TO EXPANDED DISABILITY STATUS SCALE SCORE CHANGES IN PATIENTS WITH MULTIPLE SCLEROSIS: MONTH 12 RESULTS FROM AN INTERNATIONAL OBSERVATIONAL STUDY Auquier P1, Fernandez O2, Butzkueven H3, Flachenecker P4, Idiman E5, Pelletier J1, Stecchi S6, Verdun di Cantogno E7, Issard D7, Simeoni MC1 1 Timone University Hospital, Marseille, France; 2Hospital Regional Universitario Carlos Haya, Málaga, Spain; 3Royal Melbourne Hospital, Melbourne, Australia; 4Neurological Rehabilitation Center Quellenhof, Bad Wildbad, Germany; 5Dokuz Eylül University, Ýzmir, Turkey; 6UO Riabilitazione e Sclerosi Multipla, Bologna, Italy; 7Merck Serono S.A., Geneva, Switzerland OBJECTIVES: Quality of life (QoL) is an important measure that is often overlooked in the assessment of multiple sclerosis (MS). The MS International QoL (MusiQoL) questionnaire is a validated, MS-specific instrument. This study aimed to assess the responsiveness of the MusiQoL questionnaire to changes in Expanded Disability Status Scale (EDSS) scores in patients with MS. METHODS: In this ongoing, 24-month, multicentre, observational study, MusiQoL and EDSS scores were recorded at baseline (BL) and at 6-month intervals. The primary endpoint is change in MusiQoL index score from BL to month 24 (including effect size). Secondary endpoints include change in MusiQoL index score from BL to month 12 and change in MusiQoL scale scores. RESULTS: Of 600 patients enrolled, 474 had evaluable BL and month-12 EDSS and MusiQoL index data. At BL, mean (SD) EDSS score was 2.9 (1.9); mean (SD) MusiQoL index score was 68.5 (14.4); and mean (SD) MusiQoL scale scores ranged from 59.9 (24.8) to 85.4 (18.5). At month 12, 68 patients (14.3%) had a worse EDSS score than that recorded at BL. Mean (SD) change in MusiQoL index score was 0.48 (10.99) overall, and −1.00 (9.88) in “worsened” patients; effect sizes were 0.03 and −0.08, respectively. Changes from BL to month 12 in MusiQoL scale scores were mostly small. However, there were larger (mean [SD]) decreases in some subscales at month 12 for “worsened” patients: “Activities-of-daily-living” (−4.20 [18.07]); “Relationship-with-family” (−5.14 [23.43]); ‘Sentimental-and-sexual-life’ (−4.11 [17.56]); and ‘Relationship-with-health care-system’ (−3.56 [14.11]). Effect sizes were −0.17, −0.25, −0.14 and −0.21 respectively. Conversely, the mean (SD) change in ‘Symptoms’ score was 5.36 (17.27) at month 12 for ‘worsened’ patients (effect size: 0.24). CONCLUSIONS: As expected, most MusiQoL scores decreased over 12 months in patients with worsening EDSS scores, indicating poorer QoL and confirming the utility of the MS-specific MusiQoL in rating QoL. A395 PND40 QUALITY OF LIFE OF CAREGIVERS IN HUNTINGTON’S DISEASE—FIRST RESULTS FROM EURO-HDB STUDY Dorey J1, Clay E1, Aubeeluck A2, Verny C3, Aballea S1, Squitieri F4, Toumi M5 1 Creativ Ceutical, Paris, France; 2University of Nottingham, Derby, UK; 3Neurology Unit CHU, Angers, France; 4Neurogenetics and Rare Disease Centre, Pozzi, Italy; 5University Claude Bernard Lyon 1, Lyon, France OBJECTIVES: Huntington’s disease (HD) is a rare neurodegenerative disease leading to sustained disability for patients and poor quality of life (QoL) for patients as well as caregivers. This study investigated the impact of HD on caregivers’ QoL and its drivers. METHODS: The European HD burden study (Euro-HDB) is an ongoing cross-sectional survey among HR patients and their caregivers in six countries (France, Italy, Germany, UK, Sweden and Spain). The Huntington’s disease Quality of Life Battery for Carers (HDQoL-C) short-version, a previously validated questionnaire by Aubeeluck A. and Buchanan H., was administered. Pearson correlations with generic HR-QOL (SF36, EQ5D) and the specific HR-QOL in HD for patients (HQOLI) were evaluated. The determinants of caregiver QOL among drivers among patients’ clinical characteristics (voluntary movement disorders, chorea, depression/anxiety, psychotic disorder, cognition, temper) were studied by regression analysis adjusting on age, sex and occupational categories. The relationship between QOL of patients and caregivers were also explored. RESULTS: To date, 201 caregivers in France and 124 in Italy have been enrolled. For France (respectively Italy) 6% (12%) were completely unsatisfied by their overall QOL and 7% (5%) were totally satisfied. HDQoL-C scores were poorly correlated with generic caregiver HR-QOL: correlation equaled 0.31 for EQ5D utility and varied between 0.04 and 0.45 for the eight domains of SF36. Correlation was quite high (0.59; P < 0.01) between HDQoL-C and HQOLI. Drivers of caregiver’s QoL were voluntary movement disorders (p = 0.049), depression/anxiety (p = 0.02), psychotic disorder (p = 0.01) and cognition (p = 0.01). Temper and chorea were not drivers of caregivers’ QOL independently of other clinical characteristics. CONCLUSIONS: Caregivers QOL worsens as the patient clinical characteristics deteriorate especially voluntary movement disorders, depression/anxiety, psychotic disorders and cognition disorders. Patients and caregiver QoL was indirectly correlated via patient clinical scores. The potential impact on caregiver QOL should be considered in evaluations of innovative HD treatments. PND41 THE IMPACT OF PAIN SEVERITY AND FREQUENCY ON HRQOL IN THE BIG 5 EUROPEAN UNION COUNTRIES Langley PC1, Liedgens H2 1 University of Minnesota, Minneapolis, MN, USA; 2Gruenenthal GmbH, Aachen, Germany OBJECTIVES: This study assesses, for an estimated EU pain population of 50 million patients, the impact of pain severity and frequency on three dimensions of health related quality of life (HRQoL): the SF-12 MCS and PCS scores and (ii) the SF-6D absolute utility scores. METHODS: The study is based on data from the internet based 2008 National Health and Wellness Survey undertaken in the UK, France, Spain, Germany and Italy. This study identified 11,000 respondents (1 in 5 of the estimated big 5 EU countries) who had experienced pain in the last month. The assessment of the quantitative impact of pain status on HRQoL is estimated via three single equation generalized linear (ordinary least squares) models which estimate the impact of pain on PCS, MCS and utility scores. The model includes a range of variables which have been shown in previous population studies to impact HRQoL. These include: sociodemographic factors, health risk behaviors, comorbidity status, medication utilization, duration of medication utilization and satisfaction with care. The experience of pain is captured by a combination of severity and frequency categorical variables. RESULTS: Pain has a substantial impact on all three of the dimensions of HRQoL considered here. Compared to the reference category (mild pain experienced weekly or less) the presence of severe, daily pain has a substantial impact on SF-12 PCS scores (−13.85 points); SF-12 MCS scores (−4.72); and SF-6D utilities (−0.147). There is a marked gradient on scores by severity and frequency of pain experience. The impact of pain on HRQoL is more significant than the impact of socio-demographic characteristics, health risk factors, comorbidities and the experience of pain medication. CONCLUSIONS: The presence of moderate and severe pain imposes a significant burden on persons reporting pain in the big 5 EU countries; the burden increases the greater the frequency and severity. PND42 THE RELATIONSHIP BETWEEN SOME INDICATORS INFLUENCING THE QUALITY OF LIFE OF PEOPLE WITH DOWN’S SYNDROME LOOKED AFTER IN THE FAMILY AND PARENTAL QUALIFICATION IN CONNECTION WITH A STUDY CARRIED OUT IN HUNGARY Harjánné Brantmüller É1, Máté O1, Pál K1, Nagy I1, Kriszbacher I1, Boncz I1, Sándor J2 1 University of Pécs, Pécs, Hungary; 2University of Debrecen, Debrecen, Hungary OBJECTIVES: The aim of the study was to survey some objective and subjective indicators determining the quality of life of people with Down’s syndrome (DS), born between 1975 and 2005, looked after in their families in the Southern-Transdanubian region of Hungary. The relationship between the parental level of education and the above factors was explored. METHODS: On the basis of the VRONY database (National Registry of Congenital Anomalies) health visitors contacted families looking after DS people (N = 107), and conducted anonymous, questionnaire surveys in 2008–2009. Reading was examined from school-age (N = 79), drawing and writing was examined depending on age. RESULTS: The abilities under examination moved A396 on a wide range. The increase of the mother’s educational level is in significantly positive relationship with the DS person’s computing (khi2 = 14,314, p = 0,026), drawing/ writing (khi2 = 21,027, p = 0,002) and reading (khi2 = 22,169, p = 0.001) performance. It has a favourable effect on the development of the basic self-supporting skills (khi2 = 25,571, P < 0.001), and the DS person’s social connections. The majority of those lacking friends live with mothers with the lowest qualification (khi2 = 9.799, p = 0.02). The level of happiness was measured on a four-grade scale. Owing to a mental retardation of medium gravity, surveying was carried out in an indirect way, by questioning the parents. The higher parental qualifications associated with happier DS people (khi2 = 17,344, p = 0.008). The parents’ educational levels are correlated (r = 0.609, P < 0.001). CONCLUSIONS: More qualified parents provide supportive environment rich in stimuli. Higher level of development reduces the DS person’s dependence on others, increases their self-estimation, the level of happiness and through this the quality of life improves. It means a smaller burden on the health care system, the society and the family. However the professionals taking part in family support must be reminded that not every family can perform special child educational tasks independently, and they need more intensive family support. NEUROLOGICAL DISORDERS – Health Care Use & Policy Studies PND43 IMPACT OF MEMANTINE TREATMENT INITIATION ON PSYCHOTROPICS USE: ANALYSES WITH THE RAMQ DATABASE Lachaine J1, Beauchemin C1, Legault M2, Bineau S3 1 University of Montreal, Montreal, QC, Canada, 2Lundbeck Canada Inc., Montreal, QC, Canada; 3Lundbeck SAS, Issy-Les-Moulineaux, France OBJECTIVES: Behavioural and psychological symptoms of dementia such as aggressiveness, agitation and psychosis are common and very distressful for Alzheimer’s disease patients and their caregivers. Their occurrence leads to an increased use of psychotropic medications. Memantine treatment has demonstrated significant benefit on these symptoms in the experimental setting of randomized clinical trials. The objective of this study was to assess the impact of memantine treatment initiation on the use of psychotropic medications in real life practice. METHODS: A retrospective prescription claims analysis was conducted using the Quebec provincial public health plan (RAMQ) database. Data on medical and pharmaceutical services were obtained for the period from January 2004 to March 2009 for a random sample of patients who received at least one scripts of memantine. Trends in the proportion of patients using psychotropic drugs, antidepressants, neuroleptics, and anti-anxiety agents were estimated one year before and after the first prescription of memantine. RESULTS: Data were obtained from the RAMQ for a total of 2,007 patients. The study sample was 82.2 years old on average (SD = 7.6), with 67.6% of female. Proportion of patients using a psychotropic drug in the year preceding the initiation of memantine increased by 58.5%, from a proportion of 0.450 to 0.713 while this proportion only increased by 3.5% (0.713 to 0.738) in the year following the memantine initiation. The increase in proportion of users of antidepressants, neuroleptics, and anti-anxiety agents before and after initiation of memantine were 48.3% (0.239 to 0.354) vs. 2.8% (0.354 to 0.364), 112.1% (0.219 to 0.465) vs. 1.9% (0.465 to 0.474) and 41.3% (0.175 to 0.247) vs. 1.5% (0.247 to 0.250) respectively. CONCLUSIONS: Results of this prescription claims analysis indicate that the increasing rate of psychotropic drugs, antidepressants, neuroleptics, and anti-anxiety agents use significantly decrease after the initiation of memantine. PND44 HEALTH RELATED QUALITY OF LIFE DEFICITS IN MULTIPLE SCLEROSIS: A POPULATION CONTROL ANALYSIS Langley P1, Wagner JS2, Gupta S3 1 University of Minnesota, Tucson, AZ, USA; 2Kantar Health, New York, NY, USA; 3Kantar Health, Princeton, NJ, USA OBJECTIVES: Existing assessments of the health related quality of life (HRQoL) in patients with multiple sclerosis have focused on persons with the disease. What has not been assessed is the extent to which the presence of multiple sclerosis generates HRQoL deficits compared to the general population. METHODS: Data are from the internet-based 2008 National Health and Wellness Survey in the UK, France, Spain, Italy and Germany. a total of 53,524 respondents completed the survey form, of whom 308 indicated they had a diagnosis of multiple sclerosis. Respondents completed the SF-12 generic HRQoL instrument. The analysis is based on the summary mental component scores (MCS), physical component scores (PCS) and health state utilities from the SF-6D items derived from the SF-12 instrument. a multivariate analysis (using ordinary least squares) was undertaken with the MCS, PCS and SF-6D utilities as dependent variables. The independent variables included a binary variable for the presence/absence of multiple sclerosis, socio-demographic characteristics, health risk factors (e.g., body mass index), country dummy variables and the Charlson Comorbidity Index (CCI). RESULTS: Persons reporting multiple sclerosis had substantially lower PCS scores (32.19 vs. 48.66; t = 24.60); MCS scores (39.78 vs. 46.53; t = 10.18) and utility scores (0.57 vs. 0.72; t = 23.57). In the multivariate modeling, the deficit associated with the presence of multiple sclerosis was substantially greater than any other independent variable in the PCS (−12.07; 95% CI: −14.02 to −10.12) and utility models (−0.11; 95% CI: −0.13 to −0.09). The utility decrement in the SF-6D model attributed to multiple sclerosis far outweighed the utility deficits attributable to the other independent variables. CONCLUSIONS: Compared to the non-multiple sclero- 13th Euro Abstracts sis population, there are substantial HRQoL deficits associated with the diagnosis of multiple sclerosis. These deficits are substantial and far outweigh those attributable to socio-demographic characteristics, health risk factors and the presence of comorbidities. PND45 MANAGEMENT OF DRUG RESISTANT EPILEPSIES: HTA REPORT FOR THREE ITALIAN REGIONS Lopatriello S1, Berto P1, Canevini MP2, Colicchio G3, Rubboli G4, Spreafico R5, Tassi L6, Tinuper P7 1 Pbe Consulting, Verona, Italy; 2Università degli Studi- Azienda Ospedaliera San Paolo, Milano, Italy; 3Policlinico Gemelli, Roma, Italy; 4Ospedale Bellaria, Bologna, Italy; 5Fondazione I.R.C.C.S. Istituto Neurologico Carlo Besta, Milano, Italy; 6Azienda Ospedaliera Niguarda, Milano, Italy; 7Università di Bologna, Bologna, Italy OBJECTIVES: Management of drug-resistant epileptic patients includes different surgical alternatives. The target patient (i.e. resistant to at least 2 anti-epileptic drugs) may undergo neurosurgery or Vagal Nerve Stimulation (VNS). Scope of this work was to estimate hospital costs for neurosurgery and VNS and compare with Regional funding. METHODS: Investigated phases were: Non-Invasive Diagnostics-NID, Invasive-ID, Neurosurgery Intervention-NI or VNS-I, FollowUP-FUP, in 6 referral centres. Average/patient cost was estimated by microcosting; resource consumption by questionnaire; unit costs valued by full hospital costs (staff, operating-room, hospital-stay), outpatient-tariffs (diagnostics), market-prices (drugs, devices) at 2010 values. RESULTS: NID-phase: c5070/adult (c3516-c7356) and c4382/child. Ranges reflect variability of hospital-stay and staff time; 79% of total cost is due to diagnostics. DI-phase costs c15,900/adult (37%) and c16,261/child (44%). The cost range for adults (c14,098-c17,702) reflects variability of invasive video-EEG protocols. Regional funding is under-remunerative. Considering individual procedures, videoEEG costs c3,406/adult (cost driver: staff workload); invasive-video-EEG c34,790 (driver: electrodes). For children, video-EEG costs c3,063 and invasive-video-EEG c27,619; lower values are due to higher seizures frequency and shorter duration of recordings. Neurosurgery intervention cost is c16,230/adult and c18,894/child. For adults, variability (c14,730-c18,513) depends on the duration of intervention, number and cost of staffs. Intervention cost is similar in the three Regions and hospital imbalance is the result of under-funding by regional tariffs. VNS insertion cost is c24,543/ adult (c3,518 for the intervention). Under-remuneration by tariffs is confirmed, even considering regional extra-tariffs for the device. Global treatment path for an adult (NID + intervention + FUP 5°yrs) amounts to c23,571-NI and c33,373-VNSI; for a child c20,066-NI. Regional data: Lombardy, c23,571-NI and c32,244-VNSI, for Lazio c25,571-NI and c34,176-VNSI, for Emilia Romagna c22,886-NI and c34,871VNSI. CONCLUSIONS: Funding of the management of resistant epilepsies in the target patient appears insufficient to cover costs actually incurred by Italian hospitals, irrespective of organization and Regional funding. NEUROLOGICAL DISORDERS – Conceptual Papers & Research on Methods PND46 NON-LOCAL PATIENT-LEVEL DATA FOR COST-EFFECTIVENESS ADAPTATION TO THE SWEDISH ENVIRONMENT Chowdhury CA1, Cuche M2, Nilsson M3, de Bustamante MM1, Lukasik KD1, Castañeda EC1 1 Insight Strategy Advisors, New York, NY, USA; 2Merck Serono S.A., Geneva, Switzerland; 3 Merck AB, Stockholm, Sweden BACKGROUND: Currently, data-access barriers exist throughout key European health-technology assessment countries, including Sweden. Patient registries are of limited use due to frequent lapses in data integrity, completeness and limitations around access to these data. Additionally, data-protection legislation often prohibits the use of existing data from providers and insurance funds. These limitations hinder the development of longitudinal health-technology assessments required to evaluate the cost-effectiveness of new and existing therapeutics. OBJECTIVES: To explore the opportunities and limitations of applying non-local, longitudinal, patient-level treatment and outcomes data to the Swedish environment. METHODS: A large, longitudinal, patient-level treatment and outcomes data set (>55 million US patients from >90 health insurance plans) from the USA was used to create a sample of patients with multiple sclerosis who were undergoing treatment with disease-modifying drugs. Patient outcomes were evaluated with regard to the extent to which patients were adherent to therapy. Based on adherence levels, patient tendency to use incremental health care system resources was assessed. Adaptation to the local Swedish environment was performed by incorporating or acknowledging customary aspects of the Dental and Pharmaceutical Benefits Board (Tandvårds-och läkemedelsförmånsverket) cost-benefit analysis. Costs of outcomes, specifically interventions as a result of nonadherence, were based on Swedish cost-estimates from prior independent research. CONCLUSIONS: When performing a cost-effectiveness analysis, incorporation of patient-level data from another country invokes caveats on the results, including differences in access to care, cultural attitudes and social beliefs. However, in the absence of accessible local data, complete and sufficient data from another country (e.g., the USA) can serve as a surrogate. This may limit the broad acceptance of the study’s results; however, it provides quantitatively supported and customized results based upon a robust data set, upon which health care decision-makers can derive resource allocation and prescribing decisions. A397 13th Euro Abstracts PND47 APPLICATION OF THE KANO METHODOLOGY FOR EVALUATING MIGRAINE TREATMENT EXPECTATIONS AMONG PATIENTS TREATED BY NEUROLOGISTS IN SPAIN: THE MIGREXX STUDY Matias J1, Nocea G2, Caloto T2 1 Hospital Universitario San Carlos, Madrid, Spain; 2Merck Sharp & Dohme, SA, Madrid, Spain OBJECTIVES: To classify the characteristics and properties of the pharmacological treatments among migraine patients treated by neurologists in Spain. Patients are becoming increasingly implicated in treatment decisions. Expectations in relation to therapy largely condition satisfaction with the results obtained. METHODS: Multicenter, cross-sectional study in adult patients with at least one prescription of antimigraine drugs within the last year. The protocol was approved by the CREC of La Princesa University Hospital (Madrid). Sociodemographic, and clinical characteristics of patients, as well as questions regarding their expectations with regard to migraine treatments and evaluation of the importance of their attributes are documented. Using a treatment expectations questionnaire and applying Kano methodology, sixteen treatment attributes were classified as: Must-be, One-dimensional, Attractive, Indifferent, Reverse or Questionable. Patients were required to give informed consent. RESULTS: 68 neurologists included 174 patients diagnosed with migraine (mean age 39 years; 75% women). None of the attributes were considered “Must-be”. The attributes considered by most patients to be “One-dimensional” were the absence of long-term adverse effects (55%), to allow work/study activities (50%), to allow social and family relationship (50%), achievement of symptoms relief (47%) and pain relief (44.3%). The attributes considered “Attractive” by most patients were: achievement of rapid symptoms and pain relief (54%), and achievement of total disappearance of the pain (53.4%). The attributes that led to a greater dissatisfaction were the occurrence of long-term adverse effects, not to allow work activity or studies and not to achieve pain relief. The three attributes that led to a greater satisfaction were achievement of total disappearance of the symptoms and the pain, and rapid pain relief. CONCLUSIONS: The expectations regarding a medical treatment significantly influence the satisfaction reached with the outcome of such treatment. a better knowledge of patient’s expectations can lead to a greater satisfaction associated with treatment outcomes. PND48 MODIFIED LAST OBSERVATION CARRIED FORWARD IS A SUPERIOR METHOD OF IMPUTATION DeGryse R, Lei X, Fainaru I, Turkel C Allergan, Inc., Irvine, CA, USA OBJECTIVES: To compare imputation methods using PREEMPT (Phase III REsearch Evaluating Migraine Prophylaxis Therapy) clinical data. METHODS: Data were analyzed from two phase 3 studies of onabotulinumtoxinA in chronic migraine (ICHD-II migraine and ≥15 headache days/month). Each study included a 24-week, randomized, double-blind, parallel-group, placebo-controlled phase, followed by a 32-week, open-label onabotulinumtoxinA treatment phase. The prespecified method of imputation of missing data (ie, <20 days of diary data in a 28-day period) was modified last observation carried forward (mLOCF), which estimated missing data by iteratively multiplying the subject’s most recent count by the subsequent average change rate across treatments. We compared mLOCF with last observation carried forward (LOCF), baseline observation carried forward (BOCF), mLOCF withintreatment (mLOCFw), and observed data (without imputation). Simulations to predict known data were analyzed using mLOCF, LOCF, and BOCF. RESULTS: 1384 adults were randomized (PREEMPT1: onabotulinumtoxinA [n = 341], placebo [n = 338]; PREEMPT2: onabotulinumtoxinA [n = 347], placebo [n = 358]). 1026 (74%) of the subjects had no missing monthly counts during the double-blind phase (PREEMPT1: onabotulinumtoxinA [n = 242], placebo [n = 248]; PREEMPT2: onabotulinumtoxinA [n = 258], placebo [n = 278]). Each imputation method (mLOCF, LOCF, BOCF, mLOCFw) and observed data generated a similar statistically significant betweengroup difference for the primary efficacy endpoint of headache-day counts at Week 24 in PREEMPT2 (P < 0.001). Similar results were observed within PREEMPT1 (P < 0.02). Compared to LOCF and BOCF, mLOCF was superior at predicting known values of headache-day counts in simulations that randomly set observed scores to missing, as demonstrated by least-squared errors of imputed minus actual counts. CONCLUSIONS: PREEMPT treatment differences were similar (and statistically significant) using each imputation method, likely due to low dropout rates within each treatment group. Compared to LOCF and BOCF, mLOCF was superior at predicting known data. SENSORY SYSTEMS DISORDERS – Clinical Outcomes Studies PSS1 META-ANALYSIS OF PARTIALLY HYDROLYZED 100%-WHEY FORMULA VERSUS EXTENSIVELY HYDROLYZED FORMULAS IN THE PREVENTION OF ATOPIC DERMATITIS Iskedjian M1, Szajewska H2, Spieldenner J3, Farah B4, Berbari J4, Navarro V5 1 PharmIdeas Research and Consulting Inc., Oakville, ON, Canada; 2Medical University of Warsaw, Warsaw, Poland; 3Nestlé Nutrition Institute, Vevey, Switzerland; 4PharmIdeas Research and Consulting Inc., Ottawa, ON, Canada; 5PharmIdeas Europe SAS, Lyon, France OBJECTIVES: The incidence rates of atopic dermatitis (AD) and relative risks (RR) associated to a brand of partially hydrolyzed 100% whey formula manufactured by Nestlé (PHF-W) and to extensively hydrolyzed formulas (EHF-Whey or Casein) were reported for the prevention of AD in infants who are not exclusively breastfed. METHODS: Szajewska et al. had previously undertaken a meta-analysis determining the incidence rates and RRs of PHF-W compared to cow’s milk formula (CMF) and EHF but details were not for provided for PHF-W vs. EHF. This analysis sheds light on the latter comparisons and extends the analyses to 3-month cycles from birth to 36-months of age. Included were any relevant randomized controlled trials comparing the use of PHF-W with CMF or EHF for the prevention of allergies. The primary outcomes of interest were the incidence, cumulative incidence and period prevalence of any allergic manifestations and AD in particular. Of 84 retrieved citations, 15 studies were included for analysis of which 6 studies pertained to PHF-W vs. EHF. All efficacy data were converted into inputs for a spreadsheet decision-analytic economic model based on 3-month cycles by applying weights derived from the Szajewska et al. meta-analysis and calculating them at 3-month intervals for input into an economic model. RESULTS: The analysis sample included 557, 559 and 580 patients for PHF-W, EHF-Whey and EHF-Casein, respectively. a RR of 0.75 [0.54,1.05] at 0–12 months and 0.80 [0.63,1.02] at 0–36 months was obtained for PHF-W vs. EHF-Whey while a RR of 1.06 [0.74,1.53] at 0–12 months and 1.13 [0.87,1.47] at 0–36 months was determined for PHF-W vs. EHF-Casein. All efficacy data were adapted into 3-month indicators. CONCLUSIONS: It appears that the efficacy of PHF-W falls within the range of the other two extensively hydrolyzed formulas. SENSORY SYSTEMS DISORDERS – Cost Studies PSS2 USE OF INCREASED DOSAGES OF BIOLOGICS IN PLAQUE PSORIASIS AND POTENTIAL SAVINGS BY TREATING HIGH-DOSE PATIENTS WITH USTEKINUMAB INSTEAD—A BUDGET IMPACT MODEL Klesse M, Wolbring F Janssen-Cilag GmbH, Neuss, Germany OBJECTIVES: Evaluation of real-life dosing of biologics in plaque psoriasis in Germany and the potential savings by treating patients with ustekinumab instead of using higher doses of TNF-alpha-inhibitors. METHODS: Based on an online survey among 100 dermatologists (DocCheck Medical Services, December 2009) the use of biologics and their dosing distribution for the treatment of plaque psoriasis was evaluated. The proportion of patients receiving maintenance dosages according to label and those receiving a higher dose were evaluated for each biologic separately. a budget impact model was created estimating potential savings by treating patients with ustekinumab according to label (45 mg in patients ≤100 kg body weight and 90 mg in patients >100 kg body weight) instead of high dose TNF-alpha-inhibitors in terms of annual medication costs. In this model, costs of the current usage pattern of biologics were estimated using German pharmacy prices (source: Lauer-Taxe, version June 1, 2010, most economical pack size). RESULTS: In most cases biologic treatments are used according to the respective labels. In some instances, deviating dosages are used: 14% of adalimumab-treated patients receive an increased dosage of 40 mg every week or 80 mg eow, and more than 15% of etanercept-treated patients receive 50 mg twice weekly or another deviating dosage as maintenance therapy. 27% of infliximabtreated patients receive maintenance injections more frequently than every 8 weeks or receive an increased dosage per injection. In Germany, assuming 80% of those highdose-TNF-alpha-inhibitor-treated patients are treated with ustekinumab instead, this would result in savings of approximately c18 million per year. If induction costs of ustekinumab are considered additionally, still c13 million could be saved within one year. CONCLUSIONS: Using ustekinumab instead of high dosages of TNF-alphainhibitors for treatment of plaque psoriasis can generate significant savings in medication costs in Germany. PSS3 EPIDEMIOLOGY, COSTS AND QUALITY OF LIFE IN PATIENT WITH SEVERE CHRONIC HAND ECZEMA Cortesi PA1, Scalone L1, De Pità O2, Gallo R3, Angelini G4, Cristaudo A5, Girolomoni G6, Lotti T7, Ayala F8, Cannavò SP9, Cottoni F10, Lisi P11, Peserico A12, Pigatto P13, Mantovani LG14, Belisari A15, Giannetti A16 1 University of Milano—Bicocca, Monza, Italy; 2Istituto Dermopatico dell’Immacolata (IDI)-IRCCS, Roma, Italy; 3Di.S.E.M., University of Genoa, Genova, Italy; 4University of Bari, Bari, Italy; 5Istituto Dermatologico San Gallicano IRCCS, Roma, Italy; 6University of Verona, Verona, Italy; 7University of Florence, Firenze, Italy; 8University Federico II of Naples, Napoli, Italy; 9University of Messina, Messina, Italy; 10University of Sassari, Sassari, Italy; 11University of Perugia, Perugia, Italy; 12University of Padua, Padova, Italy; 13University of Milan, Milano, Italy; 14 CIRFF, Federico II University, Naples, Italy; 15CHARTA Foundation, Milano, Italy; 16University of Modena, Modena, Italy OBJECTIVES: Some research has shown that Hand Eczema is often work-related, widespread, potentially disabling and costly disease, but misdiagnosed and mistreated. Severe Chronic Hand Eczema (CHE) can be particularly burdensome. We aimed to assess prevalence, costs and HRQoL in patients with severe CHE. METHODS: A naturalistic, multicentre, epidemiologic and cost-of-illness study is being conducted in 14 Italian dermatological centers. HE patients aged ≥18 years are being enrolled through a 6-month period. The following data are being collected: socio-demographic, clinical (severity, chronicity and responsiveness to treatment with topical potent corticosteroids) and, among chronic, severe and refractory patients, data on HRQol A398 (EQ-5D and the condition specific Dermatitis-Life-Quality-Index (DLQI)), direct and indirect costs, from the societal perspective. RESULTS: A preliminary sample of 305 patients (mean age + SD = 41.4 + 14.9, 38.2% male) were enrolled. Two hundred fifty five patients (83.6%) had CHE and of these the 27.8% were severe. Overall, 20.0% of the enrolled patients were chronic, severe and refractory to therapy. DLQI mean + SD sum score was 10.38 + 5.98. With EQ-5D 92.7% of patients reported moderate or severe pain/discomfort, 70.9% problems with usual activities, 58.2% anxiety/depression and 40% problems with self-care. VAS mean + SD = 61.96 + 21.48. On average hospitalization cost c87.90/patient-month, travels due to CHE cost c69.54/patient-month, specialist visit costs c52.81/patient-month, other products (gloves, gauze bandage, vacuum cleaner, cosmetics) costs c35.70/patient-month, diagnostic exam costs c20.77/patient-month, non pharmacological therapy (emollients, galenic products, soap, UV-ray) cost c15.65/patient-month and pharmacological therapy cost c9.22/patient-month. Patients lost a mean + SD 4.35 + 8.68 workdays/ patient-month for reasons attributable to their condition. CONCLUSIONS: most of the patients in the Italian centers of dermatology have CHE and a fifth are severe and refractory to therapy. Patients with severe CHE and refractoriness to therapy have poor HRQoL and high costs. a correct diagnosis and treatment is necessary to efficiently manage this condition. PSS4 COST-OF-ILLNESS IN PATIENTS WITH CHRONIC HAND ECZEMA: RESULTS FROM A MULTI-CENTRE STUDY IN GERMANY Augustin M1, Purwins S1, Diepgen T2, Posthumus J3, Kuessner D3 1 University Clinics of Hamburg, Hamburg, Germany; 2University Heidelberg, Heidelberg, Germany; 3Basilea Pharmaceutica, Basel, Switzerland OBJECTIVES: It is assumed broadly that the costs caused by chronic hand eczema (CHE) are significant. However, there is a lack of cost-of-illness studies on CHE. The objective of this study is, therefore, to determine the direct and indirect costs of chronic hand eczema under routine conditions overall and in different treatment stages in Germany. METHODS: The survey was conducted in 24 outpatient practices and clinics across Germany. Patients with CHE refractory to potent topical treatments and insured by statutory health insurances were eligible. Patient characteristics and resource use were directly gathered from patients and physicians. Costs were evaluated from the societal perspective. Four treatment stages were defined: only topical treatments (stage I), additionally photo therapy (II), systemic therapy (III) and inpatient treatment (IV). Bivariate associations between costs and treatment stage were assessed. RESULTS: A total of 223 CHE patients enrolled in the study. The yearly direct and indirect costs per patient are c1742 (SE: c139) and c386 (c83), respectively. a total of 63.2% of patients were treated only with topical treatments; additionally 15.7% with photo therapy, 11.7% with systemic treatments. a total 9.4% of all patients were admitted to hospitals. The total costs increase with treatment stage I-IV (P < 0.001): c1044 (c85), c2307 (c145), c2697 (c461) and c8407 (c991), respectively. Accordingly, costs also correlated with clinical severity. CONCLUSIONS: CHE patients refractory to topical steroids incur marked costs to the society. The costs are increasing disproportionately with escalating treatment stages, especially in patients admitted to hospitals. Hence, new and innovative treatments may help to reduce the societal costs of CHE. PSS5 COST OF GLAUCOMA IN THE UNITED KINGDOM ACCORDING TO THE UK GPRD Salmon J1, Lafuma A2, Robert J2, Berdeaux G3 1 Oxford Eye Hospital, Oxford, UK; 2CEMKA-EVAL, Bourg la Reine, France; 3Alcon France, Rueil-Malmaison, France OBJECTIVES: The objective of this analysis was to estimate the total budget dedicated to glaucoma care according to the UKGPRD and to identify factors associated with high costs. METHODS: Data were extracted on patients treated on the National Health Service with a diagnosis of ocular hypertension or glaucoma, or treated with topical intraocular lowering treatment, surgery or laser for glaucoma. The budget was estimated from resources consumed in 2008 and included glaucoma drugs, laser, surgery, hospitalization, specialist and general practitioner (GP) visits. In-patient resources were estimated from the Hospital Episode Statistics. Results were expressed in GBP, 2008. Factors associated with high cost were identified using linear stepwise regression. National extrapolation was performed according to the relative size of the GPRD to the UK general populations. RESULTS: Details of 33,441 patients were extracted, which suggests that about 510,000 patients were treated in the NHS in UK in 2008. The Mean age was 74.2 years, and 47.3% were male. The initial diagnosis was made at 67.8 years. Older patients, longer time since diagnosis, a higher number of previous treatments, a higher number of treatment switches in the previous one year period and use of laser/surgery were associated with a higher annual cost. Spending varied little between regions. Annual drug spending was £91.2 million on inpatient care, £4.4 million on drug prescription renewal (not specific to the glaucoma drug). Visits to the GP cost £34.8 million and visits to the eye specialist was >£54.0 million although the latter figure is likely to be an under estimate (GPRD underreported eye doctor care). CONCLUSIONS: The 2008 expenditures to care for glaucoma were >£185 million with no regional differences. Three factors were strongly associated with high costs: time since diagnosis, treatment changes, and rescue treatment. This analysis suggests that longer treatment persistence is likely to be associated with cost saving. 13th Euro Abstracts PSS6 COST OF ILLNESS OF PSORIASIS—A 1-MONTH PROSPECTIVE STUDY IN SOUTHERN SWEDEN Ghatnekar O1, Ljungberg A2, Lundqvist T2, Svensson Å3, Wirestrand LE4 1 The Swedish Institute for Health Economics, Lund, Skåne, Sweden; 2Abbott Scandinavia AB, Solna, Sweden; 3University of Lund, Malmö, Sweden; 4Kristianstad Hospital, Kristianstad, Sweden OBJECTIVES: Published cost-of-illness studies of psoriasis in Sweden are not available. This study estimates the societal cost of psoriasis care in a defined Swedish patient population. METHODS: A prevalence-based prospective recruitment of patients visiting two dermatology clinics in Sweden between September and December 2009 was performed. Patients collected resource utilization of health care contacts, treatment, travelling distance and time, and productivity loss (human capital approach) during 1 month (Swedish unit prices, 2009). RESULTS: A total of 164 patients (49% males) were included; average age 52, 76% plaque psoriasis with PASI 5.7, DLQI 7.7 and EQ-5D utility weight 0.71. The mean total cost per patient-month was c994. Main cost drivers were outpatient visits (OP) and light therapy (49%), biological drugs (20%) and productivity loss (22%). When patients were stratified according to treatment strategy, total costs (fraction of patients) for topical treatment only (TT; 34%) was c369, light therapy (LT; 24%) c1274, traditional systemic treatment (TST; 26%) c1085 and biological systemic treatment (BST; 16%) c1709 per patient-month. Main cost drivers in each treatment strategy were: OP (56%) in TT, OP (78%) in LT, productivity loss (40%) in TST and biological drugs (71%) in BST. There was no clear relationship between clinical (PASI) or subjective (DLQI) severity estimations and costs. CONCLUSIONS: In this study the cost-of-illness for a psoriasis patient amounts to almost c1000/month, with great variation depending on treatment strategy. Despite the 1200 difference in drug cost for TST vs. BST, total cost per month differed by c600 because of offsets from improved productivity and reductions in OP and topical treatment. As expected, biologically treated patients had higher costs but lower severity probably due to the treatment effectiveness. The relationship between costs and severity is complex, probably due to the selected study period and differences in effect between strategies. PSS7 THE ECONOMIC COST OF TREATING PATIENTS WITH AGE-RELATED MACULAR DEGENERATION IN SPAIN Darba J1, Kaskens L2 1 Universitat de Barcelona, Barcelona, Spain; 2BCN Health, Barcelona, Spain OBJECTIVES: Wet macular degeneration associated with age (wet-AMD) is the leading cause of legal blindness in Spain in people over 55. The aim of this study is to determine health care resource utilization and mean costs per patient with wet-AMD in 2009. METHODS: A micro-costing analyses was performed to estimate direct medical costs of patients with wet-AMD. Patient level data was obtained from different public hospitals in Spain and ophthalmologists were surveyed with a semi-structured questionnaire to obtain treatment patterns. Inpatient costs were considered from the perspective of the public health care system. Treatments under study were pegaptanib, verteporfin, ramibizumab and bavacizumab. Although bevacizumab in Spain is not approved for wet-AMD, it was used off-label in the hospital. Direct medical costs considered were drug costs, administration cost, doctors’ visits, nurse time, ophthalmologist time, anaesthetics, ambulant hospital care, external consultation, surgery and treatment of adverse effects. All costs are referred to 2009. RESULTS: Mean cost per patient treated with wet-AMD represented the following cost for the public health care system: c7290 for pegaptanib, c5810c for verteporfin, c8650 for ramibizumab and c3110 for bevacizumab. We also estimated that in Spain 180,000 people over 50 years have wet-AMD in 2009. CONCLUSIONS: Pharmacological treatments for wet-AMD are photo dynamic therapy with verteporfin, pegaptanib, verteporfin and ramibizumab, with the last one having the highest medical costs. The aging of the population and development of new drugs will probably increase the future economic impact of AMD, which remains a major health care burden. PSS8 PREDICTIVE FACTORS OF GLAUCOMA TREATMENT COST IN GERMANY Lorenz K1, Wolfram C1, Claus V2, Plesnila-Frank C2, Verboven Y3, Pfeiffer N1 1 Universitätsmedizin Mainz, Mainz, Germany; 2IMS Health, München, Germany; 3Alcon Research Ltd., Puurs, Belgium OBJECTIVES: To describe total costs and factors predicting cost in Germany for glaucoma disease states: ocular hypertension (OHT), and early (EARLY), moderate (MOD) and advanced (ADV) glaucoma. METHODS: A 5-year retrospective analysis collected health care utilization, clinical parameters, treatment(s) used and reasons for treatment change. Disease states defined by the European Glaucoma Society were applied. Costs for health care resources were based upon the German EBM/OPS code for ambulatory visits/procedures, diagnosis-related groups for hospital procedures and the Rote Liste for medication. Factors predicting cost were identified using stepwise backward multiple linear regression, entry criterion a = 0.2. RESULTS: A total of 154 patients (27 OHT, 43 EARLY, 35 MOD, 49 ADV) were enrolled from 15 centers across 5 German regions. Average age was 67 ± 11 and 57% were female. Number of years since diagnosis was 9.0 ± 5.7, 8.7 ± 4.6, 8.7 ± 4.1 and 13.2 ± 8.3 years for OHT, EARLY, MOD and ADV, respectively. Total costs, for patients with OHT, EARLY, MOD or ADV, were c226 ± 117, c423 ± 647, c493 ± 385, and c808 ± 877, respectively. Most costs were due to medication (c121 ± 99, c217 ± 150, c245 ± 161, c340 ± 193) and hospital interventions (c32 ± 101, c115 ± 538, c154 ± 285, c367 ± 13th Euro Abstracts 811). Other costs included consultations (c54, c63, c64, c58) and examinations (c19, c26, c26, c26). Factors predicting high cost (R2 = 0.6426) were the number of hospital interventions, treatment switches, disease state changes, current disease state and issues with ocular burning, stinging or aching affecting daily activities. For glaucoma patients having 0, 1, 2 or ≥3 treatment switches this gave total costs of c273 ± 108, c320 ± 231, c511 ± 462 and c932 ± 950, respectively. Medication costs were c177, c187, c268 and c368, and interventional procedure costs were c0, c50, c166 and c478, respectively. CONCLUSIONS: Glaucoma treatments (medication and interventional procedures) are the key cost drivers in management of glaucoma in Germany. Avoiding treatment switches, disease progression and interventional procedures should have an impact on the cost of glaucoma care. PSS9 COSTS OF GLAUCOMA IN SWEDEN—A PILOT STUDY 1 2 3 4 1 Svensson J , Berdeaux G , Bergstrom A , Forsby M , Ghatnekar O 1 IHE, Lund, Sweden; 2Alcon France, Rueil-Malmaison, France; 3Skåne University Hospital, Lund, Sweden; 4Pygargus AB, Sävedalen, Sweden OBJECTIVES: To estimate the cost of treatment of glaucoma in Sweden and cost by treatment change. METHODS: Two private ophthalmologic clinics in southern Sweden were chosen for a retrospective chart review. Longitudinal data from patients older than 18 years was extracted from digitalized medical records. a glaucoma diagnosis (ICD-10: H40) and prescription of glaucoma treatment (ATC: S01E) were used as search terms to identify patients. Data on outpatient visits, glaucoma related pharmaceuticals and laser surgery was collected for the period 2004-01-01 to 200906-30. Unit prices from official sources in year 2008. Health care provider costs incurred during year 2008 were stratified according to the number of treatment changes. RESULTS: A total of 815 patients with pharmacological treatment were included in the analysis. Mean age was 77.6, 63% were female, 39% had ocular hypertension (OHT), 55% primary open angle glaucoma (POAG) and 6% other glaucoma diagnosis. Mean direct medical cost per patient during 2008 was c393. The cost per patient varied depending on diagnosis and number of treatment changes since 2004. Costs for patients diagnosed with OHT, POAG and other glaucoma was c341, c423 and c477, respectively. Patients diagnosed with POAG ranged from c350 (no change in treatment) to c540 (3 or more changes). Patients with OHT ranged from c301 (no changes in treatment) to c613 (3 or more changes). Costs were approximately equally split between outpatient visits and drugs regardless of diagnosis or treatment changes. Regression analysis also revealed that costs were significantly influenced by years since diagnosis and clinic of enrolment. CONCLUSIONS: The annual cost of glaucoma treatment in Sweden is estimated to c393 in 2008. Costs increase with number of treatment changes but also depend on diagnosis, years since diagnosis and clinic of enrolment. This is in line with other studies. PSS10 CLINICAL AND COST-EFFECTIVENESS EVALUATION OF A LOW FRICTION AND SHEAR PRODUCT IN THE MANAGEMENT OF PATIENTS AT RISK OF SKIN BREAKDOWN WITHIN A HOSPITAL SETTING Ingram A1, Smith G2 1 NHS Innovations London, London, UK; 2St Mary’s Hospital, Isle of Wight, Newport, UK OBJECTIVES: To determine the clinical and cost-effectiveness of the low friction ParafrictaTM products to reduce pressure ulceration on pre-existing skin breakdown within a hospital setting. METHODS: Patients admitted to the participating hospital wards who had a Waterlow score ≥15 and were unable to reposition independently, were offered the Parafricta products. Assessments of their level of ulceration and outcome were made over a 3-month period prior to the use of Parafricta products and then for a further three months during which Parafricta products were used. RESULTS: A total of 650 patient cases were assessed. Of these, 204 patient cases met the inclusion criteria in the three months prior to Parafricta use and 165 patient cases during Parafricta use. The results demonstrated that, in patients at risk of skin breakdown, there was a statistically significant reduction in the number of patients who developed ulceration with the use of Parafricta products (16% reduction; p = 0.0286). In addition, the number of patients who were ulcer-free on admission but who developed ulceration and then went on to improve or completely heal before discharge was also statistically significant (41% increase; p = 0.0065). Fewer patients admitted with ulceration deteriorated on the Parafricta products (21% reduction; p = 0.0012). The results were used to build a cost-effectiveness model and locally derived costs for length of stay, wound dressings and mattresses, as well as the additional cost of the Prafricta products were applied to determine the overall costs savings that would result from the reduction in ulceration on Parafricta products. The base-case model indicated a saving of over £63,000 per 100 incidences. CONCLUSIONS: The results of this evaluation support the conclusion that Parafricta products in an acute inpatient environment have a significant role to play in the avoidance of skin breakdown, and that they constitute an intervention that is both clinically effective and cost-effective. A399 PSS11 THE COST-EFFECTIVENESS OF A NEW GEL FORMULATION OF CALCIPOTRIOL/BETAMETHASONE DIPROPIONATE FOR THE TREATMENT OF SCALP PSORIASIS IN NORWAY Holm MV1, Ekman M2, Ryttov J3 1 LEO Pharma, Ballerup, Denmark; 2i3 Innovus, Stockholm, Sweden; 3LEO Pharma, Marlow, UK OBJECTIVES: The study examines the cost-effectiveness of calcipotriol 50 mcg/g plus betamethasone dipropionate 0.5 mg/g in combination in a new convenient gel formulation for the treatment of scalp psoriasis compared to existing alternatives in Norway. METHODS: A decision analytic model with a time horizon of 16 weeks was developed. The perspective chosen is that of the Norwegian health care system. Responders to treatment were defined as cleared or almost cleared and efficacy was derived from two clinical trials. Relapse rates were estimated from data in the literature. The alternative treatments were betamethasone dipropionate gel and calcipotriol gel, comparators in the clinical trials, matched to marketed formulations in Norway. Costs of medical therapy and follow-up management in scalp psoriasis in 2009 were taken from the Norwegian Medicines Agency and hospital price lists. The outcomes were measured in QALYs derived from SF-36 collected in one of the clinical trials. RESULTS: The overall cost for calcipotriol liniment treatment added up to 2340 NOK over 16 weeks and for calcipotriol/betamethasone dipropionate gel 1497 NOK and betamethasone dipropionate liniment 1407 NOK (1 NOK = 0.127 EUR). The treatment associated with the best outcome was calcipotriol/betamethasone dipropionate gel, generating a gain in change from baseline over 16 weeks of 0.0111 QALYs while calcipotriol generated 0.0091 QALYs and betamethasone dipropionate 0.0106 QALYs. The analysis showed that calcipotriol/betamethasone dipropionate gel is a dominant treatment compared to calcipotriol. The ICER was 180,000 NOK per QALY gained compared to betamethasone dipropionate. The probabilistic sensitivity analysis showed a probability of calcipotriol/betamethasone dipropionate gel being cost-effective of 82% at a WTP of 400,000 NOK compared to betamethasone dipropionate. CONCLUSIONS: The new gel formulation of calcipotriol/betamethasone dipropionate indicated for the treatment of scalp psoriasis is a cost-effective alternative to both calcipotriol liniment and to betamethasone dipropionate liniment in a Norwegian health care setting. PSS12 ECONOMIC EVALUATION OF RANIBIZUMAB IN THE TREATMENT OF AGE-RELATED MACULAR DEGENERATION IN GREECE Maniadakis N, Athanasakis K, Fragoulakis V, Tsiantou V, Kyriopoulos J National School of Public Health, Athens, Greece OBJECTIVES: To investigate the cost-effectiveness of ranibizumab in relation to verteporfin photodynamic therapy (PDT), pegaptanib sodium and best supportive care (BSC) for the treatment of Age-Related Macular Degeneration (AMD) at varying disease states in the Greek health care setting. METHODS: A six-state Markov model was constructed according to patient visual acuity in the better seeing eye. Data on effectiveness were derived from randomized controlled trials comparing the outcomes of ranibizumab 0.5 mg administered over a 2year period (8 injections in the first year of treatment, 6 in the second) versus other alternative comparators for the treatment of AMD patients with predominantly classic (PC) lesions and versus BSC and pegaptanib for those with minimally classic (MC) or occult lesions. Resource utilization reflected the Greek health care setting and was defined via a panel of experts. Economic and clinical outcomes were estimated over a 10year timeframe from the perspective of a third-party payer (social insurance fund), discounted at 3.5% per annum. RESULTS: the estimated mean 10-year direct treatment cost in the ranibizumab arm ranged from c24,844 to c32,931 with a projected benefit of 4.50–4.74 Quality Adjusted Life-years, depending on type of lesion. For PC lesions, the cost per QALY gained with ranibizumab was estimated at c10,037, c19,152 and c3,759 relative to PDT, BSC and pegaptanib, respectively. The corresponding ratios for patients with MC lesions were c28,201/QALY and c19,018/QALY for ranibizumab relative to BSC and pegaptanib, whereas for patients with occult lesions were estimated at c23,976/ QALY and c39,696/QALY respectively. The probability of ranibizumab being costeffective at the c30,000/QALY threshold was 92.6%, 83.0% and 100% (PC lesions), 67% and 87% (MC) and 75% and 42% (occult) for the above presented ICERs. CONCLUSIONS: ranibizumab may be a cost-effective option for the treatment of AMD compared to selected alternatives in the Greek health care setting. PSS13 COSTS AND EFFECTS OF FIXED COMBINATION THERAPIES IN OPEN ANGLE GLAUCOMA Wickstrøm J1, Centofanti M2, Johnson M3 1 Wickstrøm & Langkilde, Vejle, Denmark; 2Fondazione G.B Bietti-IRCCS, Università di Roma Tor Vergata, Roma, Italy; 3Allergan, Inc., Marlow, UK OBJECTIVES: Using data from a recently published head to head clinical trial comparing the efficacy and safety of the fixed combination bimatoprost/timolol (BTFC, Ganfort®) versus fixed combination latanoprost/timolol (LTFC, Xalacom®), the objective was to investigate the cost-effectiveness of BTFC versus LTFC in 10 European countries. METHODS: A model was developed to evaluate the cost-effectiveness of BTFC versus LTFC taking a health care perspective including only direct health care costs. Efficacy data originated from a recent head to head trial of BTFC and LTFC. Outcomes where measured as percentage reduction in intraocular pressure (IOP) from baseline and percentage of patients achieving >15% and >20% reduction in IOP from baseline, respectively. Safety was not included in the base-case analyses as no significant A400 differences were found in the tolerability of BTFC and LTFC in the clinical study. National unit costs were applied. The time horizon of the model was 3 months. Sensitivity analyses were performed to test the results responsiveness to changes in key input parameters. RESULTS: Significantly more BTFC patients experienced >15% and >20% reduction in IOP compared to LTFC (p = 0.003, P < 0.001). Furthermore, 3-month health care costs for patients treated with BTFC were lower or comparable to those of LTFC in the 10 studied countries. Results were largely insensitive to changes in key parameters. The cost-effectiveness analyses revealed that BTFC was less costly and more effective than LTFC in 8 out of the 10 studied countries (Spain, Italy, Germany, UK, The Netherlands, Norway, Sweden and Denmark) and more effective at equal health care costs in France and Finland. Therefore, BTFC was a dominating treatment strategy in all countries. CONCLUSIONS: BTFC is an effective treatment strategy in terms of lowering IOP and is a cost-effective treatment strategy for patients with glaucoma. PSS14 PHARMACOECONOMICS OF INNOVATIVE MEDICINES FOR TREATMENT OF PSORIASIS IN UKRAINE Zalis’ka O, Mandrik O Danylo Halytsky Lviv National Medical University, Lviv, Ukraine OBJECTIVES: The objective of this study is to determine the daily and yearly cost per treatment of innovative biological preparations. In Ukraine live 46 millions inhabitans and the appearance of psoriasis is approximately 2%, 920,000 psoriasis patients. It is supposed that at 15–20% appearance of middle and severe of psoriasis there are 138,000–184,000 patients in Ukraine with degree of the illness. Adult patients with moderate to severe plaque form of psoriasis who have not clinical responce for other systematic therapies, including cyclosporine, metotrexate and PUVA, or when patients are contraindicated for these therapies or are intolerant of them. METHODS: Only direct medical cost were calculated from Ukrainian database “Compendium” on 01.05.2010. Annual treatment costs were calculated based on recommended dosesd as per Ukrainian Guidelines from 2007. The costs are presented using third-party payer’s perspective, i.e. direct cost to the health system is only considered. In the analysis there were three strategies of treatment compared: infliximab, adalimumab and ustekinumab. We used the clinical data of infliximab, adalimumab, ustekinumab from published clinical trials (IMPACT, ADEPT, PHOENIX 1). The cost-effectiveness analysis from the payer perspective was conducted. RESULTS: The costs for first year of therapy of infliximab per patient with PASI 75 response are 221998,4 UAH (1 USD = 7.92 UAH), of adalimumab—815443,2 UAH, and of ustekinumab −235224,0 UAH. In model the cost-effectiveness ratios (average cost per one unit PASI 75response) amounted to US$454 for infliximab, US$1745—adalimumab, US$362—ustekinumab. CONCLUSIONS: The cost-effectiveness analysis shows that ustekinumab is more cost-effective vs. other innovative biologics for severe psoriasis treatment in Ukraine. PSS15 COST-EFFECTIVENESS OF USTEKINUMAB VS. INFLIXIMAB FOR SEVERE PSORIASIS Omelyanovsky VV1, Avksentieva MV1, Sura MV2, Zorin N2, Hailov P3, Hailov N3 1 Research Center for Clinical and Economic Evaluation and Pharmacoeconomics, Moscow, Russia; 2Institute of Clinico-Economic Expertise and Pharmacoeconomics, RSMU, Moscow, Russia; 3NICEE, Moscow, Russia OBJECTIVES: To evaluate cost-effectiveness of ustekinumab vs. infliximab for severe psoriasis in Russia. METHODS: Cost-effectiveness analysis was performed. The evidence of efficacy and safety of biologic agents was analyzed. Incremental cost-effectiveness ratio (ICER) was calculated for both biologic agent vs. placebo. Drug costs were taken into account. Achievement of PASI 75 was considered to be expected outcome, data about efficacy was extracted from clinical trials. RESULTS: There were no trials with direct comparison of ustekinumab and infiximab. Data from separate trials showed that efficacy of ustekinumab was a little lower than infliximab at 10–12th and 24–28th weeks of treatment. On the contrary at 50–52th week of therapy ustekinumab becomes a little more effective than infliximab. Both biologic agents were generally well tolerated in most patients. Calculation of expected costs showed that ustekinumab was cheaper than infliximab if similar periods of follow-up are analyzed. At 10–12th weeks of therapy ICER for ustekinumab vs. placebo was a little higher than ICER for infliximab vs. placebo. At 24–28th and 50–52th weeks of therapy ICER vs. placebo was lower for ustekinumab. CONCLUSIONS: Ustekinumab is an appropriate alternative to infliximab for patients with severe psoriasis. PSS16 ECONOMIC EVALUATION OF A 100% WHEY-BASED, PARTIALLY HYDROLYZED INFANT FORMULA IN THE PREVENTION OF ATOPIC DERMATITIS AMONG FRENCH CHILDREN Iskedjian M1, Dupont C2, Spieldenner J3, Kanny G4, Raynaud F5, Farah B6, Haschke F3 1 PharmIdeas Research and Consulting Inc., Oakville, ON, Canada; 2Hôpital Saint Vincent de Paul, Paris, France; 3Nestlé Nutrition Institute, Vevey, Switzerland; 4Centre Hospitalier Universitaire de Nancy, Nancy, France; 5Université Paris-Descartes, Paris, France; 6PharmIdeas Research and Consulting Inc., Ottawa, ON, Canada OBJECTIVES: A pharmacoeconomic analysis was performed to determine costs, consequences and cost-effectiveness of a specific brand of partially hydrolyzed 100% whey formula manufactured by Nestlé (PHF-W), in the prevention of atopic dermatitis (AD) in “at risk” children when compared to standard cow milk formula (CMF) in France. METHODS: An economic model depicting treatment pathways, resource 13th Euro Abstracts utilization and costs associated with the treatment of AD in healthy “at risk” French newborns who were not exclusively breastfed was constructed for a 12-month time horizon, including an initial six months of formula consumption. Model inputs were based on the literature, official formularies and expert opinion, including outcomes from a meta-analysis. The treatment pathways included a dietary management approach, a medical treatment approach and a combination thereof. The final outcome was the expected cost per avoided case of AD, yielding an incremental cost per avoided case (ICER) of AD for PHF-W vs. CMF. Outcomes were presented from three perspectives: the French Ministry of Health (MOH), the subject’s family and society (SOC). a secondary analysis compared PHF-W to extensively hydrolyzed formula (EHF) in prevention. RESULTS: A total of 11,291 AD cases were expected to be avoided by selecting PHF-W over CMF in a birth cohort of 156,649 at risk infants. The base-case analyses yielded expected ICERs of c2684, -c1474 (savings) and c1210 from the MOH, family and SOC perspectives, respectively. Cost drivers were the formula from the MOH perspectives, time loss from the family perspective, and formula but also to a lesser extent time loss from the SOC perspective. PHF-W yielded approximately 81Mc savings against EHF in the secondary analysis. One-way and probabilistic sensitivity analyses confirmed the robustness of the model. CONCLUSIONS: Under a range of assumptions, this analysis has established the cost-effectiveness of PHF-W in the prevention of AD among French infants. PSS17 ECONOMIC EVALUATION OF A 100%-WHEY BASED PARTIALLY HYDROLYZED FORMULA IN THE PREVENTION OF ATOPIC DERMATITIS AMONG DANISH CHILDREN: PRELIMINARY ANALYSES Iskedjian M1, Spieldenner J2, Jarvi A3, Farah B4, Navarro V5 1 PharmIdeas Research and Consulting Inc., Oakville, ON, Canada; 2Nestlé Nutrition Institute, Vevey, Switzerland; 3Nestlé Sverige AB, Helsingborg, Sweden; 4PharmIdeas Research and Consulting Inc., Ottawa, ON, Canada; 5PharmIdeas Europe SAS, Lyon, France OBJECTIVES: A pharmacoeconomic analysis was undertaken to determine costs, consequences and cost-effectiveness of a brand of partially hydrolyzed 100% whey formula manufactured by Nestlé (PHF-W), in the prevention of atopic dermatitis (AD) in “at risk” children when compared to extensively hydrolyzed formula (EHF-Whey or Casein) in Denmark. METHODS: Based on a 6-month time horizon for formula consumption, an economic model was developed synthesizing treatment pathways, resource utilization and costs associated with the treatment of AD in healthy “at risk” Danish newborns who were not exclusively breastfed. The cost of formula was retrieved from market surveys while other model inputs were obtained from the literature. a meta-analysis of 6 studies that compared the efficacy of PHF-W (557 patients) and EHF (559 and 580 patients for EHF-Whey and EHF-Casein) yielded RR of 0.75[0.54,1.05] at 0–12 months and 0.80[0.63,1.02] at 0–36 months for PHF-W vs. EHF-Whey and RR of 1.06[0.74,1.53] at 0–12 months and 1.13[0.87,1.47] at 0–36 months for PHF-W vs. EHF-Casein. Given the evidence for non-significant differences between PHF-W and EHF, the analytic approach amounted to a cost-minimization analysis reporting the difference in formula acquisition costs. In the base case, it was assumed that infants consumed the formula of choice for the full 6 months. In a sensitivity analysis, subjects consuming PHF, EHF-Whey or EHF-Casein who developed AD symptoms were switched to EHF-Whey, EHF-Casein or EHF-Whey, respectively. RESULTS: Savings per child receiving formula of DKK 17,033 were generated for PHF-W vs. EHF-Whey while savings of DKK 16,974 were observed for PHF-W vs. EHF-Casein. The sensitivity analysis yielded a cost saving of DKK 16,800 with PHF-W. CONCLUSIONS: Under a range of assumptions, this analysis demonstrated the cost-saving nature of PHF-W vs. both types of EHF in the prevention of AD among Danish infants. Further sensitivity analyses, including multivariate, are planned for confirmation of results. PSS18 ECONOMIC EVALUATION OF A 100% WHEY-BASED, PARTIALLY HYDROLYZED INFANT FORMULA IN THE PREVENTION OF ATOPIC DERMATITIS AMONG SPANISH CHILDREN Iskedjian M1, Navot Falcó S2, Spieldenner J3, Casas Ramisa R4, Claver Monzón A2, González Enseñat MA2, Farah B5, Berbari J5 1 PharmIdeas Research and Consulting Inc., Oakville, ON, Canada; 2Hospital Sant Joan de Déu, Manresa, Spain; 3Nestlé Nutrition Institute, Vevey, Switzerland; 4Clínica Pediátrica Nevot-Casas, El Prat de Llobregat, Barcelona, Spain; 5PharmIdeas Research and Consulting Inc., Ottawa, ON, Canada OBJECTIVES: A pharmacoeconomic analysis was undertaken to determine costs, consequences and cost-effectiveness of a specific brand of partially hydrolyzed 100% whey formula manufactured by Nestlé (PHF-W), in the prevention of atopic dermatitis (AD) in “at risk” children when compared to standard cow milk formula (CMF) in Spain. METHODS: Based on a 12-month time horizon (including 6 months of formula consumption), an economic model was developed synthesizing treatment pathways, resource utilization and costs associated with the treatment of AD in healthy “at risk” Spanish newborns who were not exclusively breastfed. Model inputs were retrieved from the literature, official formularies and expert opinion, including outcomes from a meta-analysis. The treatment pathways considered a dietary management approach, a medical treatment approach and a combination thereof. The final outcome was the expected cost per avoided case of AD, yielding an incremental cost per avoided case (ICER) of AD for PHF-W vs. CMF. Outcomes were presented from three perspectives: the Spanish Ministry of Health (MOH), the subject’s family and society (SOC). a secondary analysis compared PHF-W to extensively hydrolyzed formula (EHF) in 13th Euro Abstracts prevention. RESULTS: The expected number of avoided AD cases by selecting PHF-W over CMF was 2,787 cases in a birth cohort of 38,661 at risk infants. The base-case analyses generated expected ICERs of c1921, -c1102 (savings) and c785 from the MOH, family and SOC perspectives, respectively. Cost drivers were formula from the MOH perspectives and time loss from the family perspective, with formula and to a lesser extent time loss for the SOC perspective. PHF-W yielded approximately c10.7 million savings against EHF in the secondary analysis. One-way and probabilistic sensitivity analyses confirmed the robustness of the model. CONCLUSIONS: Under a range of assumptions, this analysis has established the cost-effectiveness of PHF-W in the prevention of AD among Spanish infants. PSS19 COST-EFFECTIVENESS OF AGE-RELATED MACULAR DEGENERATION: A MODEL Visser MS1, Amarakoon S2, Dorrestijn N2, Missotten T2, Busschbach J1 1 Erasmus University Medical Center, Rotterdam, The Netherlands; 2Rotterdam Ophthalmic Institute, Rotterdam, Zuid-Holland, The Netherlands OBJECTIVES: Bevacizumab (Avastin) is a promising and low cost treatment of agerelated macular degeneration. Effectiveness and cost-effectiveness might be related to the frequency of injections of bevacizumab in the treatment. The standard frequency is 4 weeks, while it is uncertain whether lower frequencies lead to different cost and effects. METHODS: 170 patients were randomized to 3 treatment frequencies; 4, 6, and 8 week intervals between injections. Follow up was 1 year. Vision was measured with the Snellen chart, which health states have been valued for QALY analysis. We developed a 6 states Markov model with a 12 weeks cycle and a 6-year time-horizon. The model included one death state, and five states defined by visual acuity(VA) in the better seeing eye: VA >20/40, ≤20/40 to >20/80, ≤20/80 to >20/200, ≤20/200 to >20/400 and ≤20/400. RESULTS: At time of the (interim) analyses, 72 patients had completed the full follow up. The ‘6 weeks’ frequency compared with ‘4 weeks’ shows a negative ICER of c6.406, with c1,024 less costs and 0.16 more QALY’s. The ‘8 weeks’ frequency compared to the 4 weeks frequency shows also a negative ICER of c28,032, with c1.543 less costs and 0,06 more QALY’s. When looking at 6 and 8 weeks, the ‘8 weeks’ has lower costs but is also less effective, with an ICER of c4.946. Most uncertainty related to utilities and transition probabilities, while cost contributes relatively less to the uncertainty of the outcomes. CONCLUSIONS: Compared with 4 weeks frequency, the 6 and 8 weeks frequency were dominant, whereas the 8 weeks frequency has an ICER in the south west quadrant. PSS20 COST-UTILITY ANALYSIS OF A PRN (AS NEEDED) TREATMENT SCHEDULE WITH RANIBIZUMAB (LUCENTIS®) IN WET AMD BASED ON CLINICAL EVIDENCE Moeremans K1, Gerlier L1, Mitchell P2, Gallagher M3, Vincze G3 1 IMS Health Consulting, Brussels, Belgium; 2University of Sidney, Westmead, NSW, Australia; 3 Novartis Pharma AG, Basel, Switzerland OBJECTIVES: To assess whether evidence-based cost-utility of PRN-dosing with ranibizumab (Lucentis®, RBZ-PRN) in age-related macular degeneration (AMD) is comparable to RBZ-PRN cost-utility estimated previously based on clinical data from monthly and quarterly regimens. METHODS: A 10-year MS-Excel Markov model with 5 visual acuity (VA) levels and 1 death state predicts VA in patients treated with RBZ-PRN, RBZ quarterly (RBZ-Q), RBZ monthly (RBZ-M), Visudyne® photodynamic therapy (PDT) or best supportive care (BSC). Transition probabilities, adverse events and treatment frequencies were provided by newly available PRN-trials (SUSTAIN, MONT BLANC) + available trials covering other comparators (ANCHOR, MARINA, PIER, EXCITE, TAP). Comparability of populations and treatment effects (linear regressions) determined trial data pooling. Final analyses included pooled data sets (RBZ-PRN and RBZ-Q) or multiple single-trial data sets (other comparators). Secondary analysis included indirect comparison versus pegaptanib sodium (PGB). Two-year treatment duration was followed by BSC. Costs (2009, UK health care payer, 3.5% discount) were obtained from literature and expert opinion; utilities (3.5% discount) from a time-trade-off study. One-way and probabilistic sensitivity analyses (SA) covered variability in efficacy, costs, treatment frequency and utilities. RESULTS: Due to lower than predicted injection frequency, clinical trial-based costutility of RBZ-PRN was better than predicted cost-utility. Evidence-based cost-utility ranged from £4,414/QALY to £20,489/QALY versus BSC and from dominance to £2,383/QALY versus PDT. RBZ-PRN was dominant versus RBZ-Q and versus PGB (secondary indirect analysis). RBZ-M was slightly more effective but not cost-effective versus RBZ-PRN. The result was most sensitive to time horizon (2–10 y), cost of blindness and treatment duration (1–3 y) but conclusions remained throughout 1-way SA. Assuming a threshold of £30,000/QALY, the probability that RBZ-PRN is costeffective ranged from 68% to 97% versus different comparators. CONCLUSIONS: RBZ-PRN using the SUSTAIN and MONT-BLANC re-treatment criteria is costeffective compared to other therapies for wet-AMD and represents the most costeffective use of RBZ. A401 PSS21 AN ECONOMIC ANALYSIS OF TACROLIMUS OINTMENT MAINTENANCE USE (TWICE WEEKLY) VERSUS STANDARD USE IN PATIENTS WITH MODERATE TO SEVERE ATOPIC DERMATITIS IN NORWAY AND FINLAND Chambers C1, Bentley A2, Wickstrøm J3 1 Astellas Pharma Europe Ltd, Staines, UK; 2Abacus International, Bicester, Oxfordshire, UK; 3 Wickstrøm & Langkilde ApS, Vejl, Syddanmark, Denmark OBJECTIVES: The objective of this study was to determine the cost-effectiveness of tacrolimus ointment used twice weekly as a maintenance treatment regimen compared to a standard treatment regimen using tacrolimus twice daily in both adults and children with moderate to severe atopic dermatitis (AD) in Norway and Finland. METHODS: A decision analytic approach was used to compare costs and outcomes of two alternative management strategies for AD over a 12 month treatment period for the UK, which was adapted for Norway and Finland. Efficacy data were used from two randomised controlled trials where maintenance use of tacrolimus ointment (0.1% in adults and 0.03% in children applied twice weekly) was applied to previously affected areas compared to standard use (twice daily application) to treat disease exacerbations. Utility data were derived from a published source and cost data were taken from public list prices and tariffs in the two countries, Norway and Finland to generate cost/QALY. Sensitivity analyses were performed to test the degree of uncertainty around the results. RESULTS: The twice weekly maintenance use of tacrolimus ointment resulted in fewer days in disease flare and improved quality of life versus standard treatment with tacrolimus ointment in both adults and children with moderate to severe AD. The twice weekly regimen was cost-saving compared with the standard regimen in both Norway and Finland. Sensitivity analyses demonstrated that results were largely insensitive to change. CONCLUSIONS: Twice weekly maintenance treatment of moderate and severe AD in both adults and children using tacrolimus ointment improves health outcomes at a lower cost when compared with the reactive treatment strategy using tacrolimus ointment. PSS22 INDIRECT COSTS OF INFORMAL CARE FOR ONE EPISODE OF ACUTE OTITIS MEDIA IN GERMANY Jochum D, Knoll S GlaxoSmithKline GmbH & Co. KG, Munich, Germany Indirect costs are not only generated by sick people themselves, but often occur due to sick children needing informal care by parents or other informal care givers. Informal care givers are relatives or friends, not being paid for their services and thereby experiencing work, leisure time and productivity loss. Indirect costs induced by informal care givers are rarely included in CEA, although being part of societal cost. OBJECTIVES: This study quantified the indirect costs of an informal care giver caused by a sick child suffering from an episode of acute otitis media (AOM) in Germany. METHODS: Using a German access panel questionnaires were sent out to households with children <5years. The following indirect costs referring to the most recent AOM episode of any child were covered: hours of work absence and related costs for mothers, fathers and grandparents; loss of leisure time; productivity loss caused by solicitousness; work loss compensation by colleagues. RESULTS: A total of 68.3% of all parents experienced 7–29 hours of work loss (59.3/9.3% mothers/fathers) per episode, grandparents 14 hours (2%). While 40% of parents experienced paid and 28% unpaid work loss, grandparents being retired caused only unpaid productivity losses. Leisure time loss was experienced by all informal caregivers and an average was calculated across the total population accounting for c47.39. Indirect costs per AOM episode across the total population were in average c411.40 (Sum of: c152.51 paid, c161.89 unpaid work loss, c209.07 productivity, c47.39 leisure time loss; Subtraction of: c45.72 paid, c113.74 unpaid work compensation by a colleague). CONCLUSIONS: Indirect costs are mainly included in health economic (HE) analyses as paid work loss. For Germany paid work loss only accounted for 38% of total indirect costs referring to an AOM episode. Further detailed studies are needed to better understand the influence of different indirect cost categories across disease areas. SENSORY SYSTEMS DISORDERS – Patient-Reported Outcomes Studies PSS23 THE USE AND MISUSE OF ANTIBIOTICS FOR UPPER RESPIRATORY TRACT INFECTION Alawadhi FK Ministry of Health, UAE, Sharjah, United Arab Emirates OBJECTIVES: To measure the influence of introducing guidelines to doctors and educational leaflets to patients on reducing the level of prescribed antibiotics. To investigate the effect of factors such as socio-demographic characteristics, signs, symptoms and patient self management. METHODS: Research was conducted in two busy clinics, one in Dubai and one in Sharjah. a set of guidelines for the treatment of sore throat were extracted from the SIGN guidelines (Appendix 3). The set of guidelines was explained to the General Practitioners individually. The guidelines were accompanied by a covering paper which explained why and where the guidelines were produced. The researcher was based in the nursing room and approached patients suffering from a sore throat. These patients were given the educational leaflet and a brief explanation of what the leaflet included. The length of explanation depended on whether the patient could read or not. They were then asked a set of questions prior A402 to their visit to General Practitioner. Study population I: Before the implementation of guidelines and the educational leaflet. Study population II: After the implementation of guidelines and the educational leaflet. Cross-sectional analysis and descriptive analyses was performed using the Statistical Package for Social Sciences (SPSS). RESULTS: The total number of antibiotic prescriptions for patients suffering from U.R.T.I. including sore throat was significantly reduced in the intervention group (67% reduction). CONCLUSIONS: A multi-dimensional interventional approach for reducing antibiotic prescription in U.A.E. clinics resulted in a significant positive outcome. The significant reduction in antibiotic prescriptions indicates the willingness of physicians to follow guidelines and the willingness of patients to respond to educational information. PSS24 QUALITY OF LIFE IN OCULAR HYPERTENSION AND PRIMARY OPEN ANGLE GLAUCOMA Wolfram C1, Lorenz K1, Shlaen R2, Verboven Y3, Pfeiffer N1 1 Universitätsmedizin Mainz, Mainz, Germany; 2IMS Health, Munich, Germany; 3ALCON Couvreur NV, Puurs, Belgium OBJECTIVES: To estimate the impact of ocular hypertension (OHT)/primary openangle glaucoma (POAG) on health status and quality-of-life. METHODS: Classification of disease state followed European Glaucoma Society guidelines. Health status was based upon the Health Utility Index Mark 3 (HUI3) The National Eye Institute 25-Item Visual Function Questionnaire (NEI-VFQ-25) was self-administered. Utility scores were compared to a normal population matched by age and gender. Differences in health impact and quality-of-life between the different disease states were assessed. RESULTS: 154 patients were enrolled (27 OHT, 43 early, 35 moderate, 49 advanced POAG) from 15 centers in Germany, 137 were diagnosed 35 years ago. Average age was 67 ± 11 and 57% were female. 23% of patients had cardiovascular co-morbidity, 45% history of cataract, 45% hypertension, 18% diabetes, and 10% hypotension. Differences in baseline characteristics were seen for age (60, 63, 69, 72 years), history of cataract (24%, 25%, 54%, 62%), employment status (43%, 24%), and hypotension (14%, 0%, 14%, 15%). The HUI3 score for OHT, early, moderate and advanced POAG was 0.87 ± 0.09, 0.85 ± 0.15, 0.75 ± 0.23 and 0.58 ± 0.32, respectively. There was no difference in the health utility score for patients with OHT, early POAG and the normal population. Patients with moderate and advanced POAG were lower by 0.06 ± 0.24 and 0.19 ± 0.28, significantly different from OHT and early POAG (P < 0.01). The NEI-VFQ-25 for OHT and early POAG gave ocular symptoms and mental health the lowest scores. For moderate POAG the lowest scores were for driving, ocular symptoms, mental health, role limitation and peripheral vision. For advanced POAG, all domains, except color vision, were affected. CONCLUSIONS: Disease progression in glaucoma affects not only vision, but also quality-of-life. Whereas OHT and early POAG have little effect on quality-of-life, moderate and advanced POAG do. These findings can improve doctor-patient relationships, addressing quality-of-life issues for different glaucoma disease states. PSS25 MAPPING THE IMPACT OF DRY EYE ON EVERYDAY LIFE (IDEEL) QUESTIONNAIRE TO A PREFERENCE BASED UTILITY INDEX Acaster S1, Verboven Y2, Berdeaux G3, Lloyd A1 1 Oxford Outcomes Ltd, Oxford, UK; 2ALCON Couvreur NV, Puurs, Belgium; 3Alcon France, Rueil-Malmaison, France OBJECTIVES: The aim of the current study was to develop an algorithm to map the symptom domain of the Impact of Dry Eye on Everyday Life (IDEEL) questionnaire to a preference based utility index. METHODS: Data from an IDEEL psychometric validation study including 210 participants (130 dry eye patients, 32 Sjogren’s patients and 48 controls) were used to estimate the algorithm. Participants completed the IDEEL, EQ-5D and SF-36 at 2 time points; the first time point was used to estimate the algorithm and the second to validate the algorithm. The mapping work was preceded by determining bivariate correlations between the IDEEL items and each utility index (EQ-5D and SF-6D), and then examining the sensitivity of each index to variability in dry eye severity. Initial items were selected based on levels of missing data, floor and ceiling effects and correlations with the utility index. Items were then included in an OLS regression model with age and gender interaction terms. Following the item analysis the same procedures were applied to a domain level analysis. RESULTS: Based on the criteria outlined above, the SF-6D was selected as the utility index for the mapping algorithm. The final OLS regression model contained 2 IDEEL symptoms items and age, and explained 28% of the variance in SF-6D utility values; root mean squared error (RMSE) = 0.105. As the SF-6D data included few bounded or censored estimates Tobit and CLAD models were not estimated. The validation data set demonstrated a significant correlation between the predicted and observed SF-6D utility values (r = 0.53, p < 0.001). CONCLUSIONS: This algorithm forms a good basis to estimate utility values from the IDEEL for inclusion in cost-effectiveness analysis. PSS26 REVIEW OF UTILITIES IN ATOPIC DERMATITIS Mladsi D RTI Health Solutions, Research Triangle Park, NC, USA OBJECTIVES: To identify and review published utility estimates in atopic dermatitis (AD), and to catalogue the methods of utility assessment, patient populations studied, and economic evaluations incorporating the utility estimates. METHODS: A system- 13th Euro Abstracts atic search and review of the published literature, including health technology assessments, in AD was performed. Utility search terms were those recommended by NICE in the UK Publications were limited to English language only, from 1999 through 2010. RESULTS: Fourteen studies presenting 15 different sets of utility data in AD (one study used two separate methods to generate estimates) were identified. These 14 studies are summarized in 11 separate publications (one health technology assessment describes three otherwise unpublished utility studies). All studies but one present utility estimates (vs changes in utility). Two studies present a single utility estimate for AD. One study presents utility estimates for controlled vs. uncontrolled AD. All other studies present utility estimates by AD severity (e.g., clear, mild, moderate, severe), although only two studies link AD severity directly to IGA scores. Two studies present utility estimates for children. AD utilities have been collected or applied in economic evaluations in Canada, Germany, Sweden, the UK, and the US. Utilities in AD have been collected directly using SG, TTO, VAS (with the VAS results being converted into utilities for use in economic evaluations using an algorithm that reflects attitudes towards risk), and using the EQ5D. Three studies have generated utility estimates based on applying two separate published algorithms to SF-12 or SF6D data. Five sets of utility data have been used in economic evaluations. CONCLUSIONS: There are several published studies presenting utility estimates in AD; however, they vary greatly in terms of methods employed. Economic evaluations in AD, the results of which are sensitive to uncertainty in utility inputs, have relied on various estimates. PSS27 COSMETI QOL: A TOOL FOR ASSESSING QUALITY OF LIFE IN COSMETIC DERMATOLOGY Taieb C1, Sibaud V2 1 PFSA, Boulogne, France; 2Eau Thermale Avene, Lavaur, France OBJECTIVES: The assessment of quality of life (QoL) in dermatology is becoming increasingly popular as demonstrated by the creation and development of numerous questionnaires for the principal diseases of the skin. Paradoxically, although cosmetic dermatology is rapidly developing, there is no questionnaire to assess the impact of these products on the QoL of the women that use them. There was therefore a need for the creation of the Cosmeti QoL. METHODS: The questionnaire was developed using rigorous methodology in accordance with international standards in terms of quality of life. a literary review and face-to-face interviews were conducted to identify the concepts that preoccupied women over 25 years of age. Twenty-two items were identified after the first transcription; this was reduced to 12 items after an initial analysis making it easier to use. a representative population of 1002 French women aged 25 years and over, was put together by the CSA Santé institute using the quota method. They were given the Cosmeti QoL; the lower the score the better the Qol. RESULTS: The questionnaire is easy to use, good comprehension of the questions was observed. The Cosmeti QoL score is correlated to age. An improved QoL is seen in women who frequently use a moisturizing cream (13.7 vs. 14.23, P < 0.001). Sensitive skin resulted in poorer QoL (14.77 vs. 13.34 P < 0.001), the frequency of episodes of sunburn during childhood also reduced the QoL (14.96 vs. 13.86 P < 0.001). In the population over 65 years of age, the QoL was superior in women who claimed to use cosmetic, anti-wrinkle, or moisturizing products on a regular basis. CONCLUSIONS: The Cosmeti QoL scale, which is essentially based on the women’s point of view, is a valid, pertinent, and well accepted tool enabling the assessment of quality of life perceived through the skin. PSS28 DEVELOPMENT OF AN INSTRUMENT MIRRORING PATIENT AND PHYSICIAN PERCEPTION OF PSORIASIS SEVERITY AND TREATMENT EFFECT Roborel de Climens A1, Marant C1, Arnould B1, Bachelez H2, Bagot M2, Beaulieu P3, Joly P4, Jullien D5, Le Maitre M6, Ortonne JP7, Paul C8, Contreras L9, Thibout E9 1 Mapi Values, Lyon, France; 2Saint Louis Hospital, Paris, France; 3Private Practice, Pontoise, France; 4Charles Nicolle’s Hospital, Rouen, France; 5Hospices civils de Lyon, Lyon University, Lyon, France; 6Private Practice, Caen, France; 7University of Nice Sophia Antipolis, Nice, France; 8Paul Sabatier University, Toulouse, France; 9Abbott France, Rungis, France OBJECTIVES: No consensus on definition of plaque psoriasis severity currently exists. Although standard measures of psoriasis severity are commonly used in clinical practice, they are not consistent and rarely based on patient assessment. The objective was to develop an instrument assessing patient and physician perceptions of psoriasis severity and treatment effect. METHODS: Semi-directive exploratory interviews were conducted with 20 patients with mild to severe plaque psoriasis, and with 20 dermatologists. Interviews’ transcripts were analyzed to extract and organise into models the criteria used by patients and physicians to evaluate psoriasis severity and treatment benefit. Items were generated using patient words for each concept considered relevant by both patients and dermatologists. The instrument was developed in parallel for patients and for physicians, tested for relevance and comprehension on 5 patients and 5 physicians, and revised accordingly. The new version was tested on 5 new patients and 5 new physicians and revised to create a pilot version. a dermatologist advisory board was involved at each step of the instrument development. RESULTS: The test instrument consisted in 31 items including area involvement, lesion location, signs and symptoms (frequency, duration of lesions, joint involvement), treatment history, quality of life impact, rapidity and duration of treatment benefit, and patient satisfaction. The instrument was globally well-accepted by patients and physicians; few modifications were made. a 32-item pilot version resulted from the comprehension 13th Euro Abstracts tests. CONCLUSIONS: The instrument constitutes a single tool to assess both patient and physician perceptions of psoriasis severity and treatment effect. The availability of a shared instrument may improve treatment decision-making, reconciliating patient and physician perceptions. An observational study with 100 dermatologists and 561 patients is planned to assess agreement between patient and clinician perceptions; scoring and psychometric properties will also be validated. PSS29 INTERNATIONAL CO-VALIDATION OF A NEW INTERNATIONAL QUALITY OF LIFE INSTRUMENT SPECIFIC TO PHYSICAL APPEARANCE: BEAUTYQOL Beresniak A1, de Linares Y2, Auquier P3, Krueger GG4, Talarico S5, Tsutani K6, Walkiewicz B7, Berger G8 1 Data Mining International, Geneva, Switzerland; 2L’Oréal Research International, Asnieres sur Seine, France; 3University of Medicine, Marseille, France; 4University of Utah, Salt Lake City, UT, USA; 5Federal University of Sao Paulo, Sao Paulo, Brazil; 6University of Tokyo, Tokyo, Japan; 7Stowarzyszenie Lekarzy Dermatologów Estetycznych, Warszawa, Poland; 8University Pierre & Marie Curie, Paris, France OBJECTIVES: This research has been driven by the need for a quality of life (QoL) instrument that specifically assesses physical appearance. The BeautyQol instrument is a multi-dimensional, self administered questionnaire, which has been in development for over three years in 16 languages. METHODS: In the item generation phase, semi directive interviews were conducted in 309 subjects. In the second phase an acceptability study was conducted on 874 subjects in France, UK, Germany, Spain, Sweden, Italy, Russia, USA, Brazil, Japan, India (Hindi and English) China and South Africa (Zulu, Sotho and English). In the third phase, a total of 3231 subjects were recruited. to complete the BeautyQoL questionnaire, a skin clinical checklist, SF-36 and a socio-demographic questionnaire. a re-test has been carried out at 8 days on a subgroup of 652 subjects. The database was randomly divided into two subgroups and analyzed using a Rash analysis. Psychometric properties, construct validity, reproducibility, internal and external consistency were tested. RESULTS: From the item generation phase, 62 questions were selected. General acceptability was very good in the 16 cultures, with a very low rate of no answers. The validation phase reduced the questionnaire in 44 questions structured in five dimensions explaining 76.7% of the total variance: Social Life, Self confidence, Psychological life, Vitality and Seduction. Internal consistency was high (Cronbach alpha coefficients between 0.932 and 0.978). Reproducibility at 8 days was satisfactory in all dimensions. External validity testing revealed that BeautyQol scores correlated significantly with all SF-36 scores except for Physical Function. Mean completion time was 7 minutes (median:5 minutes). CONCLUSIONS: These results demonstrate the validity and reliability of the BeautyQol questionnaire as the very first international instrument specific to physical appearance. It is expected that BeautyQoL will be an instrument that will measure QoL affected by cosmetic products, techniques and agents that alter physical appearance. PSS30 DISCREPANCY IN PATIENT AND PHYSICIAN GLOBAL ASSESSMENTS OF DERMATOLOGIC DISEASES Tabolli S, Spagnoli A, Sampogna F, Pagliarello C, Abeni D, Paradisi A IDI IRCCS Rome, Rome, Italy OBJECTIVES: To investigate discrepancy in the perception of dermatologic diseases (DD) severity between patients and physicians. METHODS: A descriptive study was performed: 2459 patients with DD rated their level of disease severity on a five level scale: very mild, mild, moderate, severe, very severe (PtGA). Physician global assessment (PhGA) was performed on the same scale. Fifty three physicians were involved in an out-patient setting for three weeks (March 2010) in a dermatologic research hospital, Rome, Italy. RESULTS: Patients were predominantly females (59%), with an high education and the majority were employed; mean age was 45.9 ± 18.5 for females and 44.5 ± 18 for males. No discrepancy between PhGA and PtGA was observed in 37% of cases; PtGA under-rated compared to the physician in 35%; and PtGA over-rated relative to the physicians in 28%. Statistically significant differences were observed between PtGA and PhGA in each of the five levels of judgement (P < 0.001). Higher percentages of patients, in respect to physicians, reported very mild, severe and very severe evaluations. Physicians tended to overestimate for mild and moderate levels. Differences were observed between male and female physicians in the severity judgement, reaching a statistically significant difference for the very mild level (P < 0.001) where females were more represented. CONCLUSIONS: The perceived severity disease in DD was different between patients and physicians and it was different in patients in respect to sex. Only for very mild DD there was a difference in PhGA between males and females, with males underestimating the severity. PSS31 VISION-RELATED QUALITY OF LIFE INSTRUMENTS (QOL) AFTER REFRACTIVE CATARACT SURGERY Tugaut B1, Meunier J1, Viala-Danten M1, Arnould B1, Berdeaux G2 1 Mapi Values, Lyon, France; 2Alcon France, Rueil-Malmaison, France OBJECTIVES: To review the available vision-related QoL instruments that could be used to investigate the consequences of refractive cataract surgery, in particular the benefit of spectacle independence. METHODS: A literature review was undertaken on PubMed and Embase databases using keywords “Refractive Surgical Procedures”, A403 “Refractive Errors”, “Refractive”, “Questionnaire”, and “QoL”. Questionnaires were selected if they were developed for cataract or refractive surgery, based on the reading of the manuscript abstract. a further search was performed on PubMed, Embase and ProQolid databases to obtain information on development and psychometric validation of the questionnaires. Authors were contacted by email if missing data were identified from the published literature. Main characteristics of the questionnaires were described including number of items, targeted population, mode of administration, response scale, languages, and number of publications. Development methodology was reviewed (literature review, clinician input, patient input and comprehension test). Psychometric properties were examined (e.g. domain description, scoring algorithm, internal consistency, clinical validity, reproducibility, responsiveness). The above characteristics were then examined in light of the US FDA’s “Guidance for Industry Patient-Reported Outcome Measures: Use in Medical Product Development to Support Labeling Claims”. RESULTS: A total of 141 abstracts were reviewed and 14 questionnaires were identified. Four instruments had both a solid development methodology and good psychometric properties: the CatQuest (Cataract Questionnaire), the NEI-RQL-42 (US National Eye Institute Refractive Error QoL instrument-42), the NEI-VFQ-25 (US National Eye Institute Visual Function Questionnaire-25) and the RSVP (Refractive Status and Vision Profile). When including the ability to assess vision-related QoL with the benefit of not wearing glasses, it appeared that the NEI-RQL-42 was one of the best candidates, although the benefits of spectacle independence could be more deeply explored. CONCLUSIONS: According to this literature review, the NEI-RQL-42 could be considered as one of the best instruments to capture refractive vision-related QoL consequences after cataract surgery. PSS32 THE EXPERIENCE OF EXTERNAL GENITAL WARTS AND GENITAL HERPES ON QUALITY OF LIFE Langley PC1, Freedman D2, Wagner JS3, Gupta S2 1 University of Minnesota, Minneapolis, MN, USA; 2Kantar Health, Princeton, NJ, USA; 3Kantar Health, New York, NY, USA OBJECTIVES: Estimates of the lifetime prevalence of external genital warts (EGW) and genital herpes in the European Union range from 0.47% to 1.52% and 0.59% to 1.43% respectively. The aim here is to assess, for the first time, the impact of the experience on current health related quality of life at the general population level. METHODS: Data are from the 2008 National Health and Wellness Survey. This is an internet-based survey carried out in the UK, France, Spain, Italy and Germany. From a total of 53,524 respondents, 521 indicated they had experienced EGW and 520 genital herpes. Only 63 had experienced both conditions. The regression analysis is based on health state utilities (score 0–100) from the SF-6D. The independent variables included binary variables for the presence/absence of EGW and genital herpes, socio-demographic characteristics, health risk factors (e.g., body mass index) and the Charlson Comorbidity Index (CCI). RESULTS: The experience of EGW and genital herpes had a substantial negative impact on utility scores. The impact was significant at conventional decision levels: EGW—2.47 (95% CI: −3.58–−1.36), genital herpes −3.52 (95%CI: −4.63–−2.71) and EGW and genital herpes −5.00 (95%CI: 1.76–8.25). The impact of EGW and genital herpes experience was similar to the negative impact of BMI for persons who were underweight, obese and morbidly obese and the CCI (−2.53;95%CI: −2.65–−2.41). Age, education and income all had a positive and significant impact on HRQoL. CONCLUSIONS: This is the first time the lifetime experience of two of the most prevalence sexually transmitted infections (STIs) on current HRQoL has been assessed. The results point to the continuing impact of this experience, with herpes having a marginally greater impact than EGWs. The HRQoL deficit is most apparent for those who have experienced both STIs. SENSORY SYSTEMS DISORDERS – Health Care Use & Policy Studies PSS33 REAL-LIFE DOSING OF BIOLOGICS IN PLAQUE PSORIASIS—A GERMAN SURVEY Klesse M, Wolbring F Janssen-Cilag GmbH, Neuss, Germany OBJECTIVES: Evaluation of use, dose distribution, and dosing rationale of biologic treatments in plaque psoriasis within private practices and hospitals in Germany. METHODS: Fully structured Online Questionnaire using Umfragecenter® software. Panel participants were selected by DocCheck Medical Services using their MediAccess Pool. Survey was done in December 2009. 100 dermatologists (60 in private practices, 40 in hospitals) were included in the survey. Inclusion criterion: currently treating at least two psoriasis patients with biologics, at least one patient on adalimumab, etanercept or infliximab. RESULTS: Each surveyed dermatologist treated approximately 100 psoriasis patients per quarter. In private practice about 10% of these patients were treated with a biologic, while in hospitals about 23% received biologic treatment. About 40% of the patients receiving a biologic suffered from psoriatic arthritis as well. Distribution of the different biologics used was as follows: etanercept 37%, adalimumab 33%, infliximab 20%, and ustekinumab 10%. Only minor differences in those proportions were observed between private practices and hospitals. In about 80% of all cases, used dosing for each biologic conformed to the respective label. In other cases, increased dosages were observed, for example: 12% of adalimumab patients received 40 mg weekly, 17% of etanercept patients being treated longer than A404 24 weeks received a 50 mg BIW maintenance dose, and 23% of infliximab patients received infusions every 6 weeks. Reasons for dose increase included higher severity of disease, longer disease duration, loss of efficacy, overweight, joint involvement, higher number of previous systemic therapies received. CONCLUSIONS: Real-life dosing of psoriasis biologics does not always conform to the posology recommended in the product label, especially for high need patients. About 20% of patients being treated with a TNF-alpha-inhibitor seem to need a higher non-standard dose. PSS34 THE IMPACT OF POTENTIAL PILL SPLITTING ON GENERIC AND BRANDED FINASTERIDE UPTAKE Nichols E, Wiederkehr DP, Doyle J Quintiles Global Consulting, Hawthorne, NY, USA OBJECTIVES: Finasteride is indicated for the treatment of male pattern hair loss at 1 mg strength and for male symptomatic benign prostatic hyperplasia (BPH) at 5 mg strength. While the male pattern hair loss indication exclusivity does not expire until 2013, the BPH indication expired in 2006. Patients have been known to purchase generic finasteride in 5 mg strength and split the pills into quarters as a less expensive alternative to the branded 1 mg dose. The objective was to examine the uptake of generic finasteride in relation to the prescribing volume of branded finasteride for the treatment of male pattern hair loss to determine the impact of potential pill splitting on generic uptake and branded prescribing volume. METHODS: Generic and branded finasteride were selected as case products. From 1992 to 2010 the volume of prescriptions (TRx) were collected monthly using SDI’s VONA databases and grouped according to branded and generic sales. RESULTS: Sales of generic finasteride significantly accelerated immediately after launch. The compound average growth rate (CAGR) of TRx of branded 5 mg finasteride between 1992 and 2006 was 23.29%. Meanwhile, the CAGR of generic 5 mg finasteride from 2006 to 2010 was 96.17%, signaling an aggressive growth trend. Interestingly, the sales of branded 1 mg finasteride were not impacted as they maintained steady growth with a CAGR of 2.49% over the four years since generic launch. CONCLUSIONS: One hypothesis as to the immediate growth of generic finasteride after launch can be attributed to pill splitting; much of the volume increase was due to patients purchasing the much less expensive generic 5 mg finasteride and quartering the pills. However, if pill splitting did occur, it did not appear to impact sales of branded 1 mg finasteride as demonstrated by a maintained steady growth in TRx. This suggests that patients who pill split may not have been prior users of branded 1 mg finasteride. PSS35 COST COMPARISON OF PATIENTS WITH CHRONIC HAND ECZEMA COVERED BY STATUTORY VS. OCCUPATIONAL HEALTH INSURANCES IN GERMANY Diepgen T1, Augustin M2, Purwins S2, Posthumus J3, Kuessner D3 1 University Heidelberg, Heidelberg, Germany; 2University Clinics of Hamburg, Hamburg, Germany; 3Basilea Pharmaceutica, Basel, Switzerland OBJECTIVES: In Germany, patients with chronic hand eczema (CHE) can either be treated within the statutory health insurance system, or if CHE is suspected to be work-related CHE in the system of occupational health insurances. The study objective was to analyze the cost of CHE between both systems. METHODS: The survey was conducted in 24 practices in the statutory health insurance system and 2 specialized centers of the occupational health insurances. Patients with CHE refractory to potent topical treatments were included. Patient characteristics/status and resource use were elicited from patients and physicians. Costs were evaluated from the societal perspective (using insurance specific tariffs). Regression models were employed to compare the direct and indirect costs across both data sets. RESULTS: A total of 223 patients in the statutory health system and 87 patients in the occupational health insurance system were included in this study. The patients’ severity was similar across both samples. The yearly direct and indirect costs per patient are c1742 and c386 in the statutory health insurance system as well as c3309 (c2534 according to statutory health insurance tariff) and c3422 in the occupational health system. The indirect costs are more pronounced among patients with work-related CHE. a comparison of direct and indirect costs reveal higher costs for patients covered by occupational health insurances (P < 0.01); however, no cost differences between both systems can be detected after controlling for treatment stage and tariff differences. CONCLUSIONS: Differences in the costs between both systems can be explained by different treatment mixes, as direct costs are similar across treatment stages. As a result of longer absences from work, the indirect costs of patients with work-related CHE are higher. This may lead to the use of more effective and costly treatment strategies in this patient group. PSS36 ECONOMIC ANALYSIS OF CATARACT SURGERY IN EUROPE: AN ANALYSIS OF HOSPITAL DATABASES AVAILABLE IN 11 COUNTRIES Doz M1, Lafuma A1, Berdeaux G2 1 CEMKA-EVAL, Bourg la Reine, France; 2Alcon France, Rueil-Malmaison, France OBJECTIVES: Cataract surgery is one of the most efficacious medical strategies and one of the most frequently performed operations in developed countries. Organization for Economic Co-operation and Development (OECD) collects information on this topic but their data are not updated often. The aim of this survey was to estimate the number and costs of cataract surgeries performed in 14 European countries and the potential costs associated with astigmatism. METHODS: Cataract surgery numbers were estimated from available databases. Costs associated with cataract surgery were 13th Euro Abstracts based on official tariffs of local health care systems. The number of surgeries and the costs per 100 000 people were estimated for each country to allow comparisons. Astigmatism related costs were also explored. This survey was carried out in 14 European countries: Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, the The Netherlands, Portugal, Spain, Sweden, and the UK. RESULTS: Information was fully available in 10 countries and partly available in OECD statistics. Cataract surgery rates were in the range of 444 to 1,006 operations per 100,000 people in the countries with complete information. All the countries were using a DRG system for costs and tariffs. Costs of operations were highly variable, depending on complications, type of surgery, and surgical setting (outpatient or inpatient). Average cost of surgery ranged across the countries from c875 to c2000. Average cost per inhabitant was estimated at between c5 and c15 per year. Astigmatism associated costs are never taken into account. CONCLUSIONS: Cataract surgery is performed in a large part of the European population with variations across countries. Further research needs to be conducted to explain differences between countries. PSS37 CATARACT SURGERY GUIDELINES IN EUROPE: A LITERATURE REVIEW AND DOCUMENT SEARCH IN 15 COUNTRIES Jeanbat V1, Lafuma A1, Berdeaux G2 1 CEMKA-EVAL, Bourg la Reine, France; 2Alcon France, Rueil-Malmaison, France OBJECTIVES: Cataract surgery is one of the most efficacious medical strategies and one of the most frequently performed operations in developed countries. The aim of this survey was to establish an assessment of the available guidelines for cataract surgery in Europe. METHODS: The search was performed with a two steps design. a classical literature search was performed using the method elaborated by the French Haute Autorité de Santé directly through the Internet. a second step was to interview local experts to identify grey literature in local languages that could be available through local health authorities or medical societies. This survey was carried out in 15 European countries: Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, the The Netherlands, Portugal, Spain, Sweden, Switzerland, and the UK. RESULTS: The classical literature search identified 15 documents that could be considered as guidelines on cataract surgery. They were mainly written in English; only one originated in Europe. General guidelines (n = 9) considered the initial checkup and indications for cataract surgery, surgical techniques and follow-up. Specific guidelines were found covering the type of implants (n = 1), surgical technique (n = 1) and the anaesthesia technique (n = 4). This latter group was mainly extracted from the Cochrane database. The second step of the research identified 10 local guidelines in 7 European countries. These guidelines were rarely complete (n = 2) and envisaged some specific aspects of the operation (n = 8) or the implants (n = 2). CONCLUSIONS: Guidelines on cataract surgery are available at an international level but are rarely provided at a country level in Europe. SENSORY SYSTEMS DISORDERS – Conceptual Papers & Research on Methods PSS38 USE OF CALL CENTRE METHODOLOGY TO COLLECT QUALITY OF LIFE (QOL) DATA IN A CLINICAL TRIAL: A CASE STUDY OF PATIENTS WITH DIABETIC MACULAR EDEMA (DME) Loftus J1, Pleil A2, Li J2, Wallace S3 1 Pfizer Ltd., Walton, UK; 2Pfizer Inc, San Diego, CA, USA; 3Quintiles Ltd, UK, Livingston, Scotland, UK OBJECTIVES: QoL data collected in clinical trials traditionally involves self-report or face to face interviewer-administered questionnaires. The former may result in missing data/errors; the latter requires trained interviewers, possibly adding variability to the data collected (when multiple sites are involved) or bias in terms of social desirability. To improve the quality of data collected in a clinical trial for DME, we employed a centralised call centre methodology. METHODS: QoL data were collected in a large multicenter trial, NCT00605280, evaluating the safety and efficacy of pegaptanib sodium in patients with DME. An interface between the trial database and the independent call centre was established. Once subjects entered the trial, subject identifier, status and visit history were automatically generated and the call centre agent received subject contact and visit details directly from the trial database. The agent contacted the subjects to schedule interviews at the appropriate time points. Five call backs were allowed per call to ensure questionnaire completion. All QoL responses were recorded directly into the call centre database. The status of the QoL visit information and completions were sent to the trial database weekly. RESULTS: The two QoL measures (National Eye Institute Visual Function Questionnaire 25 and EuroQol Group’s EQ-5D questionnaire) had high completion rates; 99.8% of completed interviews included both questionnaires (57% successfully completed on initial attempt, 23% on the second attempt of contacting the subject). Only 4% of data were missing; 26 minutes was the average completion time. CONCLUSIONS: This call centre methodology resulted in high questionnaire completion and little missing data. This approach demonstrates robustness and offers a feasible alternative for questionnaire administration, particularly in visually impaired people. Administration in the subject’s home maybe advantageous and centralised interviewers (instead of site interviewers) potentially reduced the variability in questionnaire response. Further research comparing methods is indicated. 13th Euro Abstracts POSTER SESSION III HEALTH CARE USE & POLICY STUDIES – Consumer Role in Health Care PHP1 ORPHAN DRUGS: DOES SOCIETY VALUE RARITY? Desser A, Kristiansen IS University of Oslo, Oslo, Norway OBJECTIVES: A general societal preference for prioritizing treatment of rare diseases over common ones could provide a justification for accepting higher cost-effectiveness thresholds for orphan drugs. We attempt to determine whether such a preference exists. METHODS: We surveyed a random sample of 1547 Norwegians aged 40–67. Respondents chose between funding treatment for a rare versus common disease and completed a person trade-off (PTO) exercise between the diseases for each of two scenarios: 1) identical per person costs, and 2) higher costs for the rare disease. Diseases were described identically with the exception of prevalence. Respondents were randomized to either no information or different amounts of information about disease severity (severe vs. moderate) and expected benefits of treatment (high vs. low). All respondents rated five statements concerning equity attitudes on a Likert-scale. RESULTS: A total of 68% of respondents agreed completely with the statement “rare disease patients should have equal right to treatment regardless of costs.” Faced with trade-offs, 11.3% of respondents favored treating the rare disease, 24.9% the common disease and 64.8% expressed indifference. When the rare disease entailed a higher opportunity cost, results were 7.4%, 45.3% and 47.3%, respectively. Framing (“extra funding” vs. “replace patients”) and amount of information about severity and treatment effectiveness had a small impact on preferences. CONCLUSIONS: Although there is strong support for general statements expressing a desire for equal treatment rights for rare disease patients, that support evaporates when individuals are faced with opportunity costs. PHP2 THE HTA PUZZLE: VALUES, PRIORITIES, TECHNOLOGY AFFORDABILITY, AND PATIENT AND COMMUNITY PREFERENCES. HOW CAN WE MAKE IT LL FIT TOGETHER? Grainger D1, Kelly M2, Skilbeck M2, Srikanthan S2 1 Eli Lilly & Company, Sydney, Australia; 2Deloitte Touche Tohmatsu, Melbourne, Victoria, Australia OBJECTIVES: This research considers how mature health technology assessment (HTA) systems incorporate patient and community preferences and priorities into decision-making processes. This presentation examines the policy rationale for increased patient and community involvement, as well as providing case studies to illustrate different approaches to community focused HTA practices. Effective options for patient and community engagement will also be covered. METHODS: Qualitative semi-structured interviews with key stakeholders in six developed countries: Australia, Canada, England, France, Germany, and Scotland. a literature review was also undertaken to assess current HTA systems across the six countries. RESULTS: There was significant variation across countries in terms of implementation, organization, and prioritization. This study revealed two different, but parallel, considerations that are important in HTAs: The ‘consumer perspective’ offers insight into variations in health outcomes of technologies being assessed, as well as providing detail beyond what is captured by QALY assessments in relation to community values, preferences, and priorities. Secondly, the ‘community perspective’ takes a broader societal approach to examine values, access, use, affordability of new technologies. CONCLUSIONS: An understanding of these two ‘HTA puzzle pieces’ is required for the establishment of governance and participation processes that will enable well defined consumer and community roles in HTA, and will lead to more systematic approaches to the integration of patient perspectives. Findings from this research suggest there is a strong need to consider a wider perspective of patient engagement in health policy and delivery, of which HTA is but one part. PHP3 THE VALUE OF PATIENT PERSPECTIVES IN THE DRUG DEVELOPMENT LIFECYCLE Versnel J1, Watch J2, Jack O2, Sparrowhawk K2 1 Cambridge, London, UK; 2PriceSpective LLC, London, UK OBJECTIVES: Research was undertaken to determine the impact and value of soliciting patient viewpoints at key stages within the drug development lifecycle. METHODS: Interviews were held with representatives from the biopharmaceutical, regulatory and pricing and reimbursement sectors to identify the value and impact attributed to patient perspectives. Additionally, surveys of European patient organizations and patients with airways disease provided insight into their attempts to influence treatment decisions, which could affect market uptake. RESULTS: Industry ngagement with patients was often reactive and linked to the identification of factors which may impact on the success of a product. Earlier consultation may help the biopharmaceutical industry to optimise product development or assist key go/no-go decisions. Whilst there is a harmonised regulatory process that engages patient organizations through the European Medicines Agency, patient viewpoints or quality of life data seldom influence access or pricing and reimbursement decisions in the main European markets. Numerous examples were cited where this has led to political lobbying and emotive decision making and a more transparent process involving patients may help A405 to ensure decisions are based on clinical and cost-effectiveness. Involving patients in the drug development lifecycle has the potential to improve health outcomes and reduce health expenditure associated with poor management, often related to noncompliance. The earlier patients are involved, the greater the potential to avert or minimise factors that could affect commercial success and return on investment. CONCLUSIONS: For the biopharmaceutical industry, the value of patient perspectives lies as part of a risk-mitigation strategy to provide greater insight and control over factors that affect market success. For regulators and payers, the value lies in the potential to improve health outcomes and reduce health care budgets. PHP4 POST HEALTH CARE REFORM PERCEPTIONS IN THE US: CONSUMER PULSE SURVEY Mooney P1, Janssens S2 1 Deloitte Consulting, Boston, MA, USA; 2Deloitte Consulting, Diegem, Belgium OBJECTIVES: In March 2010, President Barak Obama signed the Patient Protection and Affordable Care Act (PPACA), which radically changed the health care landscape in the United States. We carried out a survey to gauge consumers’ opinion about the U.S. health care system after the PPACA was signed into the U.S. law. METHODS: We commissioned Harris Interactive to conduct a nationally representative telephone survey of 1,019 adults from the U.S. between May 21–24. Respondents were questioned on their current insurance coverage status, their knowledge about the PPACA, and their perception of the reform. Data were weighted to be representative of the total U.S. adult population in terms of age, sex, geographic region and race. The survey has a sampling error of ±3% at 95% confidence level. RESULTS: 84% of all adults surveyed have an health insurance; 96% of adults 65 years old and above have health insurance. Overall, while 36% of all adults surveyed think they will be “better off” as a result of the reform; 43% think they will become “worse off”. 61% of respondents said they are “very knowledgeable” or “somewhat knowledgeable” about the PPACA; 54% of adults surveyed who rated themselves as “very knowledgeable” think they will be “worse off” as a result of the bill. Adults surveyed anticipate an increase in tax (76%), hospital and physician prices (65%), health insurance cost (65%) and cost of medications (54%). CONCLUSIONS: According to the survey, while the majority of consumers are satisfied with their current health care coverage, many have concerns about potential changes from PPACA. Concerns include health insurance coverage, access to quality health care and potential cost increase. In addition, the majority of consumers said that they are at least somewhat knowledgeable about PPACA. HEALTH CARE USE & POLICY STUDIES – Diagnosis Related Group PHP5 PATIENT DISCHARGE FROM INTENSIVE CARE UNITS WITHIN AND BETWEEN HOSPITALS IN HUNGARY Varga S1, Gresz M2, Kriszbacher I1, Oláh A1, Betlehem J1, Sebestyén A3, Ágoston I1, Boncz I1 1 University of Pécs, Pécs, Hungary; 2National Health Insurance Fund Administration, Budapest, Hungary; 3National Health Insurance Fund Administration, Pécs, Hungary OBJECTIVES: In our earlier study we found that measured by the number of cases (patients), the market share of intensive therapy was analyzed in Hungary and was found between 0.84% and 1.80% during a 14-year period (1995–2008). The discharge of patient not needing intensive therapy is obligated. The aim of the study was to analyze patient discharge. METHODS: Data were derived from the National Health Insurance Fund Administration. The two types of patient discharge were analyzed during a 9-year period, the discharges from intensive care unit within the same hospital and between different hospitals from 2000 to 2008. RESULTS: Compared to the total discharges the within and between hospitals discharges changed between 70.6% (2000) and 77.3% (2005) during the 9 years (2000–2008). Within hospital discharge started from 65.2% (2000) and the highest value was 73.0% (2006). Between hospitals discharged changed between 2.9% (2006) and 5.4% (2000). The sum of the two types discharge steadily increased from 70.6% (2000) to 77.3% (2005) in the first 6 years and after a linear descending it finished on 74.7% in 2008. Within hospital discharge increased from 65.2% (2000) to 73.0% (2006) and decrease to 69.6% until 2008. Between hospitals discharges decreased from 5.4% (2000) to 4.3% (2001), but it were steady from 2002 to 2005 (4.5%–4.8%). In 2006 a sudden decrease happend to 2.9% and after it the values were 3.3%–3.2% in the last 2 years. The mean of the within hospital discharges were 70.5% while the between hospitals discharges were 4.2%. CONCLUSIONS: In Hungary between 2000 and 2008 the discharges of the patients from the intensive care units to other hospital departments were 70.5%. The majority (94.38%) of the discharged patients was treated in the same hospital and only a minor proportion (5.62%) were moved to other hospitals. A406 HEALTH CARE USE & POLICY STUDIES – Disease Management PHP6 ROUTINE OF PATIENTS’ REDIRECTION IN THE HUNGARIAN PRIMARY CARE Lipp S1, Boncz I2, Gresz M3, Varga S2, Oláh A2, Marada G2, Sebestyén A1 1 National Health Insurance Fund Administration, Pécs, Hungary; 2University of Pécs, Pécs, Hungary; 3National Health Insurance Fund Administration, Budapest, Hungary OBJECTIVES: The aim of this study is to analyze the GPs’ routine of redirection. We reveals which diagnoses induce the most frequent redirection cases to specialists, which specialties are the most frequently targeted and how the distance from the nearest outpatient unit influences the GPs’ practice. METHODS: The analysis is based on the monthly reports (B300 form of the Hungarian Naional Health Insurance Fund Administration (OEP) of 701 general practitioners. Study covers the years 2008 and 2009, 14 million visits of 952 thousand inhabitants of South-Transdanubian Region of Hungary. GPs’ redirection routine is measured by the redirection rate, which shows the number of redirections related to the number of GP visits (%). RESULTS: The population of the region decreased by 0.8% over the years 2008–2009, however the number of GP visits increased by 12%. The surplus is mainly due to the H1N1 vaccination campaign and the rise of cardiovascular diseases and respiratory diseases of allergic origin. The average redirection rate was 10%. 76% of the GPs’ surgeries are located not farther than 15 km from the nearest outpatient unit, and 62% not farther than 15 km from a hospital. The distance from the nearest outpatient unit influences the redirection rate, but this impact is not determining. The most frequently needed specialties are rheumatology, cardiology, pulmonology and orthopedics. GPs direct their patients of 0–14 years to specialists more rarely than patients of 15 years or above and also the referred diagnoses vary between the two age groups. CONCLUSIONS: The redirection rate is lower on case of young patients and higher on case of adult, and elderly patients. While young patients get to specialists mainly due to acute and chronic respiratory diseases (many of them have allergic origin), orthopaedic malformations or simply health monitoring, adult patients have cardiovascular, rheumatic, and endocrine problems. PHP7 AN EXPLORATORY STUDY EVALUATING THE PREPAREDNESS OF PRACTICING COMMUNITY PHARMACISTS IN MALAYSIA FOR MANAGEMENT ROLES Al-Haddad M, Hassali MA Universiti Sains Malaysia, Penang, Malaysia OBJECTIVES: To evaluate the preparedness of the community pharmacists in the State of Penang in six main management functions: Accounting, finance, human resource management, marketing management, planning, and directing. METHODS: A cross-sectional study design was carried out with a total of 100 community pharmacists practicing in the State of Penang, Malaysia. a pre-validated questionnaire, containing 3 sections was used as data collection tool. All data was analyzed using SPSS for Windows version 13.0. Inferential statistics were used whenever appropriate at alpha value of 0.05 or less considered significant. RESULTS: A total of 50 pharmacists successfully responded to the survey. When respondents were asked about the percentage of time spent in actual management functions, 42% of them responded that their time is being spent on actual management functions. More than 60% of community pharmacists said that most of the time, they undertake tasks in accounting, finance, human resource management, marketing management, planning, and directing. When respondents were compared in terms of their demographics, the time overseas graduates showed better preparation and performance of management functions compared to local graduates. CONCLUSIONS: The findings of this study showed that most of the community pharmacists understand and undertake the management functions in running their community pharmacy business. Furthermore the findings suggested that local pharmacy graduates were less prepared to undertake various management functions compared to their overseas graduated counterparts. PHP8 INTERVENTIONS TO REDUCE HOSPITAL READMISSIONS IN THE ELDERLY Linertová R1, García-Pérez L1, Vázquez-Díaz JR2, Lorenzo-Riera A3, Sarría-Santamera A4 1 Fundación Canaria de Investigación y Salud (FUNCIS), Las Palmas de Gran Canaria, Spain; 2 University Hospital of Canary Islands, Santa Cruz de Tenerife, Spain; 3Primary Care Services of Gran Canaria, Las Palmas de Gran Canaria, Spain; 4Agency for Health Technology Assessment, Carlos III Institute of Health, Madrid, Spain OBJECTIVES: Unplanned hospital readmissions of elderly people present an increasing burden for health systems. This could be, theoretically, reduced by adequate preventive interventions. However, there is uncertainty about the effectiveness of different types of interventions. The objective of this systematic review was to summarise available evidence on the effectiveness of interventions to reduce the risk of unplanned readmissions in patients of 75 years and older and to determine the role of home care components. METHODS: We searched studies in MEDLINE, CINAHL, CENTRAL and seven other electronic databases up to October 2007 and updated the search in MEDLINE up to October 2009. Clinical trials (randomized or controlled) evaluating the effectiveness of an intervention to reduce readmissions in elderly patients compared to a control group were selected. Quality was assessed by the SIGN tool. The extracted information was presented in text and tables. RESULTS: Thirty-two clinical trials were included and divided into two groups: in-hospital interventions (17 studies) and 13th Euro Abstracts interventions with home follow-up (15 studies). Three studies from the first group and seven from the second group found positive effects of the evaluated intervention on readmission outcome. CONCLUSIONS: Most of the evaluated interventions did not have any effect on readmissions of elderly patients. However, those interventions that comprised some kind of home care seem to be more likely to reduce readmissions in the elderly. HEALTH CARE USE & POLICY STUDIES – Drug/Device/Diagnostic Use & Policy PHP9 BUDGET IMPACT ANALYSIS OF ORPHAN DRUGS IN BELGIUM: ESTIMATES FROM 2008 TO 2013 Denis A1, Mergaert L1, Fostier C1, Cleemput I2, Simoens S3 1 Yellow Window Management Consultants, Antwerp, Belgium; 2Belgian Health Care Knowledge Centre, Brussels, Belgium; 3K.U. Leuven, Leuven, Belgium OBJECTIVES: This study aims to calculate the impact of orphan drugs on the Belgian drug budget in 2008 and to forecast its impact over the following five years. METHODS: The 2008 budget impact was calculated by triangulating information derived from multiple Belgian data sources, including Ministerial Decrees, figures published by the National Institute for Health and Disability Insurance, and reimbursement files submitted to the Drug Reimbursement Committee. The 2008–2013 budget impact analysis was based on three scenarios reflecting different levels of growth in the number of registered orphan drugs in the European Union, the number of drugs reimbursed in Belgium, and the average annual cost per patient per drug in Belgium. The price year was 2008. RESULTS: At the end of 2008, 31 different orphan drugs were approved for reimbursement in Belgium for 35 different indications. The annual number of Belgian patients treated varied between 1 and 1080 patients per orphan drug. The cost per patient per year ranged from c4600 to c411,000 between orphan drugs. The orphan drug budget impact amounted to c66.2 million (or 5% of the Belgian hospital drug budget) in 2008. The impact would increase to c130–c204 million in 2013, depending on the scenario. CONCLUSIONS: The analysis showed that the budget impact of orphan drugs in Belgium is substantial and rising, thereby putting pressure on total drug expenditure. This static analysis measured orphan drug costs only, assuming that other components of health expenditure do not change over time. Policy options to address the rising budget impact include pricing linked to return on investment, risk-sharing arrangements and re-appraisal of orphan drug status if additional indications are approved. PHP10 DOES THE MARKET SHARE OF GENERIC MEDICINES INFLUENCE THE PRICE LEVEL? A EUROPEAN ANALYSIS Dylst P, Simoens S K.U. Leuven, Leuven, Belgium OBJECTIVES: This study aims to investigate the relationship between the market share of generic medicines and the medicine price level in European off-patent pharmaceutical markets. METHODS: Data on medicine volumes and on medicine values for a basket of 35 active substances were derived from IMS Health. Ex-manufacturer medicine prices were calculated by dividing market value by market volume. The analysis was limited to medicines in immediate-release, oral, solid dosage forms. The included countries were Austria, Belgium, Denmark, Germany, France, Italy, the The Netherlands, Spain, Sweden and the United Kingdom, which constitute a mix of countries with low and high generic medicines market shares. Data were available from June 2002 until March 2007. RESULTS: Market value has seen a substantial decrease in high generic market share markets (−26.6%), while the decrease in low market share markets was marginal (−0.06%). Market volume has risen in both high generic market share markets (+29.27%) and low market share markets (+27.40%) but the cause of rise is different for the two markets. In low generic market share markets, the rise was caused by the increased use of generic medicines while in high market share markets the rise was driven by the increased use of generic medicines and a shift of use from originator to generic medicines. In high generic market share markets, medicines prices have dropped by −43.18% and in low market share markets by −21.56%. CONCLUSIONS: The extent to which price competition from generic medicines leads to price reductions appears to vary according to the market share of generic medicines. High market share markets of generic medicines have seen a larger decrease in medicine prices than low market share markets. Countries need thus to create an environment which stimulates the generic medicines use and so increases the market share. PHP11 NEW AND IMPROVED: DEFINING INNOVATION FOR HEALTH TECHNOLOGY POLICY Davis C, Menon D University of Alberta, Edmonton, AB, Canada OBJECTIVES: Both individuals and society stand to benefit from encouraging health technological innovation—individual patients through quicker access to promising technologies, and society through the development of a competitive life sciences sector. The “Innovation Pass” (UK) and the “Critical Path Initiative” (US) are 2 examples of current health policies aimed at promoting innovation. However, there does not seem to be a clear agreement on what constitutes innovation. The objective of this project 13th Euro Abstracts was to identify existing definitions of innovation, values inherent in innovation initiatives, and important considerations for the development of policies to promote innovation of health technologies. METHODS: A literature search of bibliographic databases including PubMed, EMBASE, Web of Science, Centre for Reviews and Dissemination databases, National Library of Medicine Gateway, EconLit and the New York Academy of Medicine Grey Literature collection was conducted for the period January 2005 to April 2010. The search terms were intended to capture concepts of “innovation” and “policy” in the health technology (drugs, devices, etc,) sector. Two researchers reviewed titles and abstracts of over 4500 references identified; 200 papers were retrieved for full review. Key components of innovation were extracted and summarized in tabular form to identify trends and emerging themes. RESULTS: System disrupting, development of relationships and improvement on current practice are examples of components of, or criteria for defining, health technology innovation. Thematic concepts that emerged during the review include innovative health technology as a novelty and as a mechanism for achieving some benefit or good (broadlydefined) at various levels of the health system. CONCLUSIONS: Based on the variety of definitions in the literature, lack of a common understanding of innovation may result in policy incoherence. The use of a consistent and unambiguous definition provides a solid framework from which to develop policy that is measurable, meaningful and, therefore, has a greater chance of being effective. PHP12 PHYSICIANS VIEWS ON BIOMEDICAL TECHNOLOGY IN GREECE 1 1 1 1 2 Tsiantou V , Skroumpelos A , Mylona K , Athanasakis K , Konstantinopoulos O , Kyriopoulos J1 1 National School of Public Health, Athens, Greece; 2Sotiria Hospital, Athens, Greece OBJECTIVES: Given the significant impact of biomedical technology on health, the present study aims at identifying the accessibility to certain biomedical technologies and factors affecting its use and diffusion, in Greece. METHODS: A strictly structured questionnaire was designed and sent to a sample of 388 internists and GPs over 50 years old, stratified by geographical area and employment sector. Participants were asked a) to evaluate on a 1 to 10 point scale patient access to selected biomedical technologies and the degree to which selected factors affect their decision to use the above interventions and b) to rank certain factors effect on the diffusion of biomedical technologies. RESULTS: The response rate was 76%. The statistical analysis revealed that the most accessible biomedical technologies were ultrasonography (9.4), PSA (9.38), cardiac enzymes (8.99), MRI and CT (8.86), and mammography (8.83). The most important factors affecting participants’ decision to use a technology were the treatment outcome (9.23), the disease severity (9.11) and the appropriateness of the technology for each condition (8.27) while factors such as health system and patient cost were proved less influential. 68.1% of participants claimed delays in the diffusion of biomedical technology in Greece, identifying as major barriers economic and specialized human resources deficiencies. CONCLUSIONS: Based on our results, higher access was observed to technologies related to neoplasms and cardiovascular diseases, which represent the main causes of morbidity and mortality in Greece. Furthermore, our findings support the view that when it comes to use a technology physicians are mostly concerned with the clinical effectiveness of an intervention and less with its impact on health care expenditures. Finally, the major diffusion barriers identified in this study show a suboptimal resource allocation practice, stressing the need for measures to be taken in this direction in order to enhance diffusion of biomedical technologies in Greece. PHP13 IMPLEMENTATION OF THE TRANSPARENCY DIRECTIVE IN HUNGARY 1 2 Kalo Z , Nagyjanosi L 1 Eötvös Loránd University, Budapest, Hungary; 2Syreon Research Institute, Budapest, Hungary OBJECTIVES: Transparency Directive (TD) of the European Union aims to ensure the transparency of procedures for the pricing and reimbursement of medicinal products by Member States. TD proposes strict timelines for the pricing and reimbursement process and indicates the necessity of objective and verifiable criteria for decisions and the availability of remedies for negative decisions. Our objective was to compare the routine process of pharmaceutical pricing and reimbursement with the TD in Hungary. METHODS: We analyzed official resolutions of 29 pricing and reimbursement submissions by the National Health Insurance Fund (NHIF) between January and June 2008. In 14 cases the NHIF granted reimbursement, in 15 cases the reimbursement claim was rejected. We calculated the time period between the submission of the reimbursement dossier and the official decision. We assessed the consistency of applying objective and verifiable criteria in positive or negative decisions. RESULTS: The average time period for pricing and reimbursement procedure was 172 days (min: 43 days; max: 534 days). We could not justify the consistency of employing objective and verifiable criteria in the pricing and reimbursement resolutions of innovative pharmaceuticals. CONCLUSIONS: The pharmaceutical pricing and reimbursement process in Hungary is neither transparent nor predictable. There are several open pricing and reimbursement submissions without resolution for long period. Although we could analyze only cases with resolution, the time period for pricing and reimbursment decision was still longer in several cases than 90 + 90 days recommended by TD. The appropriate use of scarce public health care resources could not be justified in case of positive decisions and there is no remedy for negative reimbursement decisions. TD has been implemented only partially in Hungary. A407 PHP14 THE JOINT COMMITTEE FOR NEW DRUGS EVALUATION IN SPAIN: 6 YEARS OF EXPERIENCE Collar J Pharmacoeconomics & Outcomes Research Iberia, Madrid, Spain OBJECTIVES: The Joint Committee for New Drugs Evaluation (JCNDE) was established in 2003 to improve efficiency in drug evaluation in Spain. Five Regional Drug Evaluation Centres are part of it and have common Standard Operating Procedures (SOPs) which are regularly updated and improved. The objective of this study was to analyze the drugs innovation degree scores assigned by the JCNDE and timing between the new drug commercialization and the JCNDE evaluation. METHODS: The JCNDE SOPs define a stepwise procedure with 4 key criteria for new drug innovation ratings: efficacy, safety, convenience and drug cost. The drug innovation scores range from 0 (insufficient experience with the drug) to 4 (relevant therapeutic improvement). The drug evaluation results were gathered from JCNDE reports and from the Regional Drug evaluation centre reports. The time period analyzed was from 2004 to 2009. RESULTS: Ninety drug evaluations were held, considering 86 different drugs and 11 evaluations for a new drug indication for the same drug. Seventy-eight (87%) of the evaluations were negative (scores 0–1), not finding any 0 in the last 2 years of the study. Ten and 2 evaluations were scored as 2 and 3 respectively. None of the drugs assessed were considered a relevant therapeutic improvement compared to the existing options. Five drugs not reimbursed were evaluated. Median time since commercialization to evaluation was 6 months (IQR: 2–11 months) and 32 drug evaluations were held before up to a maximum of 3 months after commercialization. CONCLUSIONS: The JCNDE has been an efficient instrument to develop new drug assessments in the Primary Care setting for the Regional Health Systems in Spain. Most of the assessments held have been negative. At present, health-economics arguments are basically focused on the daily treatment cost comparisons. About 1/3 of the evaluations are started before drug commercialization. PHP15 DRUG COMPARATOR DIFFERENCES IN THE THERAPEUTIC BULLETINS IN SPAIN: THE JCNDE EXPERIENCE Collar J Pharmacoeconomics & Outcomes Research Iberia, Madrid, Spain OBJECTIVES: The Spanish Joint Committee for New Drugs Evaluation (JCNDE) was created in 2003 and is formed by 5 Regional Drug Evaluation Centres. JCNDE has common Standard Operating Procedures (SOPs) to unify drug assessments and where the comparators availability is a key issue. Each individual Drug Evaluation Centre has its own new drug therapeutic bulletin to spread information between their health professionals. The objective of this study was to quantify the uniformity degree through the drug bulletins and the degree of heterogeneity reflected in the drug costs. METHODS: New drug therapeutic bulletins of the 5 members of the JCNDE were reviewed between 2006 and 2009. Each drug assessed by the JCNDE was tracked to identify which members reported it in their bulletins. Comparative drugs included in the bulletins were also registered to find potential heterogeneity amongst the JCNDE members. Daily/monthly treatment costs described were included in the study database. RESULTS: Fifty-seven drug evaluations were recorded, forty (70%) were published by at least 3 members of the JCNDE, and only 12 were published by all of them. Andalusia, Basque Country and Catalonia are the most active members, with 12–14 new drugs published per year. In all the bulletins were identified some differences in the comparators used for each new drug assessed. Andalusia and Catalonia tend to include more comparators. The highest level of heterogeneity is observed in drugs of diabetes treatment. Most of the drug costs reported were very similar. Only one disagrement between the JCNDE was identified in the drug innovation rating. CONCLUSIONS: JCNDE has made advances to unify drug assessment in Spain. Nevertheless, health-economics arguments are still focused on drug treatment cost comparisons. The comparators found in the drug bulletins are slightly different amongst the regions. Future SOPs version ought to improve these two weak points. PHP17 ACCESSIBILITY TO ORPHAN DRUGS IN JAPAN—HAS THE ORPHAN DESIGNATION SYSTEM CONTRIBUTED? Tomita N1, Kodama T2, Inagaki A1 1 Keio University, Kanagawa, Japan; 2National Institute of Public Health, Saitama, Japan OBJECTIVES: To promote the research and development of drugs for rare diseases, like other countries, orphan designations have been granted to pharmaceuticals in Japan since 1993. We investigated the accessibility of orphan drugs in Japan by comparing the accessibility of orphan designated and marketing authorised drugs in the EU and the US. METHODS: The present study used the data available until the November 30, 2009 from the European Medicines Agency, US Food and Drug Administration and National Institute of Biomedical Innovation. The International Nonproprietary Names (INNs) were used for comparing authorised orphan designated drugs in Japan, the EU and the US. RESULTS: A total of 228 products had been granted orphan designation, of which 142 (62%) obtained marketing authorization in Japan, which is equivalent to 122 in INNs. Meanwhile, the number of authorised orphan-designated medicines in INNs in the EU and the US was 57 and 198, respectively. Of the total 287 authorised orphan-designated pharmaceuticals in INNs in these 3 regions, 165 were inaccessible in Japan through the orphan designation system. Among such drugs, 25 (15%) were authorised orphan designated in both the EU and the US, 15 (9%) were authorised orphan designated in the EU alone and 125 (76%) were authorised orphan designated in the US alone. CONCLUSIONS: We A408 observed that the orphan designation system in Japan has achieved certain results for increasing the accessibility of necessary drugs to patients suffering with rare diseases. However, several drugs are still not available in Japan, partly because of the difference in definitions of orphan disease among the 3 regions. To increase the accessibility to orphan drugs, further policy interventions should be considered. PHP18 SCIENTIFIC EVIDENCE FOR THE RELATIONSHIP BETWEEN PHARMACEUTICAL BUDGET OF HEALTH INSURANCE FUND AND THE POLITICAL ELECTION CIRCLES IN HUNGARY Boncz I1, Sebestyén A2 1 University of Pécs, Pécs, Hungary; 2National Health Insurance Fund Administration, Pécs, Hungary OBJECTIVES: There is a continuous problem in Hungary—and assumable in many other countries—with planning the health insurance pharmaceutical budget. In Hungary there is a substantial gap between the planned and the actual budget resulting in an overspending of the planned pharmaceutical budget. The aim of our study is to analyze the gap between the planned and the actual health insurance pharmaceutical budget. METHODS: Data were derived from the nationwide administrative data set of the National Health Insurance Fund Administration (OEP), the only health care financing agency in Hungary. We analyzed the difference between the planned and the actual pharmaceutical budget of OEP between 1994 and 2006. Outcome (overspending rate) is measured with the following formula: the difference between the planned and the actual budget is divided with the planned budget. RESULTS: During the period of 1994–2006 we found significant overspending of pharmaceutical budget of OEP which varied between 3.4–36.6% of actual pharmaceutical expenditures. The peak of overspending showed a 4 years circle with the highest figures in 1994 (21,5%), 1998 (32.1%), 2002 (36.6%) and 2006 (30.4%). These 4 calendar years correspond with the time of national political (parliamentary) elections. CONCLUSIONS: We found the highest overspending of the Hungarian pharmaceutical budget in the years of national political elections. It is a scientific evidence for the political influence of the health insurance pharmaceutical budget. The overspending does not relate to any specific political parties or governments because it was a general phenomenon in Hungary between 1994–2006. PHP19 ANALYSIS OF THE GREEK PHARMACEUTICAL MARKET: THE FRAMEWORK, THE FACTS AND THE TRENDS (1998–2008) Vitsou E1, Souliotis K2, Caporis X3 1 Foundation of Economic and Industrial Research, Athens, Greece; 2Cival Ervants Security Fund, Athens, Greece; 3MSD Greece, Athens, Greece BACKGROUND: The Greek economy is being called upon to manage one of the highest proportions of public debt versus GDP in the European Union and the Euro zone. In turn, the health care sector is subject to the distortions of the Greek public sector, as there is no official method of measuring and evaluating the services being provided. OBJECTIVES: The objective of this analysis was both to interpret the development of pharmaceutical expenditure, based on the most significant changes in the pricing and reimbursement system, as well as to evaluate the institutional changes based on their effectiveness. METHODS: The analysis reflects the evolution of pharmaceutical expenditure and interprets the changes in the regulatory framework that took place in Greece from 1998–2008. It should be noted that there is significant confusion around the actual level of pharmaceutical expenditure, as the data provided for 2003–2007 (provisional data) from the National Statistical Service of Greece are not in agreement with OECD’s data. Following a more careful analysis of the data, numerous questions arise with respect to their reliability, as there is no explanation for the apparent rate of change with regard to the factors that could influence expenditure (e.g. changes in the pricing or reimbursement system). Therefore the actual level of pharmaceutical expenditure is also in question. RESULTS: After analyzing the price changes in the top 100 selling pharmaceutical products in Greece, throughout the indicated period, it became obvious that the cause of increasing pharmaceutical expenditure cannot be attributed to increasing prices, but to other factors mainly associated with over-consumption. CONCLUSIONS: Finally, taking into account the intense pressures on public financing of the health care system, it is important for any future measures to constrain costs to be based on reliable and accurate data and to target the entirety of the system. PHP20 QUALITY ASSURANCE OF FOURTH HURDLE CONCERNING TO MEDICAL DEVICES Tesar T1, Babela R2 1 UNION health insurance fund, Bratislava, Slovak Republic; 2St.Elizabeth University of Health and Social Sciences, Bratislava, Slovak Republic OBJECTIVES: The Slovak guidelines for economic evaluation, although appearing to be generic, have been written focusing on pharmaceuticals. The objective of this study was to analyze the quality of submitted economic studies and related critical appraisals process and to develop a policy-relevant, publicly available Slovak critical appraisal checklist for improving the quality of economic evaluation for reimbursement submission of dossiers concerning to medical devices. METHODS: We created a working group to review previously submitted economic evaluations and related critical appraisals in order to identify potential technical and methodological problems. The working group consisted of two independent academic experts who scrutinized previous submissions and critical appraisals concerning to medical devices between 2007– 13th Euro Abstracts 2010. RESULTS: Evidence suggests that the methods of economic evaluation can be equally applied to drugs and devices in general. However, there are several specific methodological issues that require more attention if reliable and informative evaluations of devices are to be conducted. These issues are underestimated within the Slovak Republic. It is well known that expenditures for medical devices do not result in the most cost-effective outcomes. Economic evaluations of medical devices are mandatory but the quality of evaluations and critical appraisals are poor. Our analysis shows that the simplified questionnaire, which is currently used for the critical appraisal process within Slovakia should be replaced by a new Slovak critical appraisal checklist, which will be detailed enough to address the specific problems in the local economic evaluations process. CONCLUSIONS: The transparent method of technology assessment can improve the consistency of reimbursement decisions making related to medical devices in Slovakia. Therefore in addition to the available Slovak health economic evaluation guidelines a detailed checklist for appraisal processes specific for medical devices have to be prepared. The economic evaluation of devices raises additional challenges and the current Slovak guidelines overlook several issues. PHP21 EARLY ACCESS: ANALYSIS OF THE FRENCH ATU SYSTEM 1 1 1 2 Hensen M , Heemstra HE , Meijboom MJ , Doreau C 1 Pharmerit International, Rotterdam, The Netherlands; 2Consultant, Antibes, France OBJECTIVES: The French cohort ATU provides early access to medicines for serious or rare diseases prior to marketing authorization (MA) for groups of patients. This study determines the length of time of patient access before MA for different types and classes of products subject to a cohort ATU. METHODS: All medicines subject to a cohort ATU between 1994 and 2009 were obtained from the website of Afssaps. Products which were never authorized were excluded. Time from initiation of a cohort ATU to marketing authorization was determined. Subgroup analyses were performed for orphan products, products developed by small and medium enterprises (SMEs), and products in major indication classes. RESULTS: 87 products matched the inclusion criteria. Products were available on average 23 (range 0–121) months prior to MA. Orphan drugs (N = 13;15%) and non-orphan drugs were available 19 months (2–74) and 24 (0–121) months before MA respectively. Drugs developed by SMEs (N = 5;6%) were available on average 34 (9–59) months before MA, while drugs developed by larger companies were available 22 (0–121) months before MA. Blood products (ATC = B;N = 10;11%) were available on average 53 (2–121) months before MA, and anti-infectives (ATC = J;N = 31;36%) 15 (0–70) months, followed by 13 (1–74) months for oncology products (ATC = L;N = 19;21%). The average length of time before MA decreased from 21.4 between 1994 and 2003 to 3.7 between 2004 and 2008. CONCLUSIONS: This study demonstrates that the cohort ATU provides access to medicines several months before MA and has been used primarily by larger pharmaceutical companies. The length of time of access before MA differs by indication and type of product and has been shorter in recent years. Price level and indication should reflect future authorized indications and price. The lower than expected number of orphan drugs subject to a cohort ATU may be the result of more frequent use via the nominative ATU or in clinical trials. PHP23 ECONOMIC EVALUATION OF TELEHEALTH/TELEMEDICINE AND COST-EFFECTIVENESS ARGUMENT AMONG KEY DECISION-MAKERS Shepelev J, Guhl A, Vekaria R, Gbenedio T GfK HealthCare, London, UK OBJECTIVES: To gather insights into global (5EU, US, emerging markets) national policy & key decision-makers’ beliefs & perceptions on current and future landscape as well as impact and value of telehealth/telemedicine initiatives within health care programmes from health economic perspectives. METHODS: Both primary and secondary research was used for this study. Literature reviews with compilation of all relevant and up-to-date information including data in local languages. Primary research focused on market specifics where clarification and market insights were required (such as MOH, budget-holder’s organizations, Public Health Institutions, senior health care professionals). RESULTS: Benefits for improving both accessibility and quality of health care through telehealth/telemedicine have vast potential. These initiatives require substantial investment, whereas human and financial resources are limited. It is important to demonstrate economic viability, in addition to technical and clinical evaluations, in order to understand the optimal conditions under which new telehealth/telemedicine programmes should be unfolded. Traditionally, cost-effectiveness analysis is used to evaluate telehealth/telemedicine programmes. Such an approach is represented only by monetary differentials (fixed, variable, marginal, and direct and indirect costs) and has numerous limitations due to the lack of available data as well as constant changes in technological progress. As discovered, it is imperative to add a range of different analytical techniques to traditional cost analyses. Other factors— such as perception and investment choices, quality of health care provision, consumer choice and political environment—should be taken into consideration. CONCLUSIONS: To understand the criteria national leaders use, and the supporting evidence they require, to allocate monetary resources for innovative telehealth/telemedicine programmes—traditional monetary analysis is not enough and we recommend two supplementary approaches: societal and perceptual analysis. This will allow further insights on views and value that telehealth/telemedicine initiatives deliver—and the priority within health care programmes—in order to reveal the willingness to pay among key decision-makers. 13th Euro Abstracts PHP24 CRITICAL ASSESSMENT OF BELGIAN REIMBURSEMENT DOSSIERS OF ORPHAN DRUGS Denis A1, Mergaert L1, Fostier C1, Cleemput I2, Hulstaert F2, Simoens S3 1 Yellow Window Management Consultants, Antwerp, Belgium; 2Belgian Health Care Knowledge Centre, Brussels, Belgium; 3K.U. Leuven, Leuven, Belgium OBJECTIVES: This study aims to conduct an assessment of reimbursement dossiers of orphan drugs in Belgium. First, a qualitative analysis reviews the evidence submitted in reimbursement dossiers of all orphan drugs. Second, an in-depth analysis compares the clinical evidence submitted to the European Medicines Agency with the Belgian reimbursement dossiers for selected orphan drugs. METHODS: A qualitative analysis examined all reimbursement dossiers of orphan drugs that have been submitted in Belgium between January 2002 and June 2008. The following information was extracted from each dossier: description of the orphan drug; indication; reimbursement status; therapeutic value and needs; budget impact; and number of registered indications. For selected orphan drugs, an in-depth analysis extracted and compared information about the clinical trials, their primary endpoints and results from the following data sources: European Medicines Agency documents (i.e. the marketing authorization application file, the European Public Assessment Report and the Summary of Product Characteristics); and the Belgian reimbursement dossiers. RESULTS: Reimbursement was awarded to the majority of orphan drugs. In addition to the official criteria, other arguments such as the price, employment, patient population, funding of diagnostic tests by the company seemed to play a role in the reimbursement decision. Despite the low number of patients, randomized controlled trials were conducted for many orphan drugs. Budget impact analyses were simplistic and did not consider the impact across multiple indications. Also, some differences were observed in the clinical evidence submitted to the European Medicines Agency and the Belgian reimbursement authorities. CONCLUSIONS: There is substantial variation in the evidence that is submitted to the Belgian authorities in the context of an orphan drug reimbursement dossier. Also, some differences have been noted in the clinical evidence reported in European Medicines Agency documents and the evidence included in Belgian reimbursement dossiers of orphan drugs. PHP25 DOES HUMAN MEDICINES DEVELOPMENT IN THE EUROPEAN UNION ADDRESS GLOBAL AND REGIONAL HEALTH CONCERNS? Catalá-López F1, Garcia-Altés A2, Álvarez-Martín E3, Gènova-Maleras R4, Morant-Ginestar C4 1 Spanish Medicines and Healthcare Products Agency (AEMPS), Madrid, Spain; 2King’s College London, London, UK; 3Rey Juan Carlos University, Madrid, Spain; 4Regional Health Council of Madrid, Madrid, Spain OBJECTIVES: Our aim was to investigate whether efforts to develop innovative medicines in the European Union (EU) are focusing on the most relevant conditions from a global public health perspective. METHODS: We reviewed the information on new medicinal products approved by centralised procedure (from 1995 to 2009), available for the public in the European Commission Register of medicinal products and the European Public Assessment Reports (EPAR) from the European Medicines Agency. We included morbidity-mortality data for each disease group, according to the World Health Organization Global Burden of Disease project. We evaluated the association between authorised medicinal products and burden of disease measures (disability-adjusted life-years [DALYs]) in the EU and worldwide. RESULTS: We considered 520 marketing authorizations for medicinal products and 338 active ingredients. New authorizations were seen to increase over the period analyzed. There was a positive, high correlation between DALYs and new medicinal products development (ρ = 0.619, P = 0.005) in the EU, that was moderate for middle-low income countries (ρ = 0.497, P = 0.030) and worldwide (ρ = 0.490, P = 0.033). In the EU, the most neglected conditions (related to their attributable health losses) were neuropsychiatric diseases, cardiovascular diseases, respiratory diseases, sense organ conditions or digestive diseases, while globally they were perinatal conditions, respiratory infections, sense organ conditions, respiratory diseases, or digestive diseases. CONCLUSIONS: Our findings suggest that the development of new medicinal products is higher in some diseases than in others. Pharmaceutical industry and decision-makers are invited to consider the implications of this imbalance establishing work plans that allow for setting future priorities from a public health perspective. PHP26 DEVELOPMENT OF SYSTEMATIC PROCESS MAPS TO ENABLE COMPARISON BETWEEN DIFFERENT HTA AND DECISION MAKING SYSTEMS: THE FIRST STEP TO BENCHMARKING Pichler FB1, Allen N2, Salek S2, Liberti L1 1 CMR International Institute for Regulatory Science, London, UK; 2Cardiff University, Cardiff, UK OBJECTIVES: The current variability in European health technology assessment and decision-making leads to uncertainty about how to measure efficiency and hampers adoption of new best-practice initiatives, such as the EUnetHTA Joint Action. In addition, such variability in organization makes comparison of the activities and outcomes of these systems problematic. The objective of this study was to develop a systematic process-mapping methodology to both clarify how these systems were organised and to enable comparison between the systems. METHODS: The process maps were designed from the point of view of the path that a new medicine takes from market authorization to market access. Key decision and evaluation gateways A409 occurring within the process and key activities, such as provision of scientific advice, were overlaid onto the process maps to further illuminate and contrast the different systems. RESULTS: The systems of eleven countries were mapped and common processes identified. Considerable differences in the systems were detected, including the extent of independence of appraisal, pricing and decision making bodies. However, despite differences in organization and sequence, the general path that a new medicine undertakes from registration to reimbursement was broadly similar. CONCLUSIONS: The process maps are a useful tool for the comparative visualization of the process of registration to reimbursement for new medicines. This methodology enables the identification of milestones that can be used to establish a benchmarking programme for the purpose of comparative performance evaluation between different countries. It is envisaged that benchmarking of HTA and decision-making agencies will lead to both process improvement and increased understanding of the costs and benefits of making changes in the systems under evaluation—as has been the case for our previous activities benchmarking the drug regulatory authorities. PHP27 COMPARISON OF THE USE OF PRESCRIBED MEDICATIONS BETWEEN QUEBEC RESIDENTS COVERED BY PUBLIC AND PRIVATE DRUG INSURANCE: THE REMED REGISTRY Blais L1, Cyr MC1, Perreault S1, Moisan J2, Bérard A1, Lalonde L1, Rahme E3 1 Université de Montréal, Montréal, QC, Canada; 2Laval University, Québec, QC, Canada; 3 McGill University Health Centre, Montreal, QC, Canada OBJECTIVES: To describe the use of medications among Quebec (Canada) residents covered by private drug insurance and to compare with Quebec residents covered by the public drug insurance plan. METHODS: Persons aged less than 65 years with private drug insurance were recruited in community pharmacies and medical clinics between 2007 and 2010 and were included in a computerized registry called reMed. Patient’s identification, private drug plan information and socio-demographic variables were gathered at recruitment. Data related to prescriptions dispensed in community pharmacies are transferred to reMed bi-monthly from the community pharmacies’ computer services providers. We compared the 10 most prevalent medication classes dispensed (using the AHFS classification) and the average drug cost between reMed participants and Quebec residents covered by the public drug insurance plan. RESULTS: In 2008, 2 211 reMed participants filled at least one prescription in a community pharmacy. Participants were 43 years old on average, 31% were male, 19% were current smokers and 46% were overweight or obese. In 2008, analgesics/ antipyretics, psychotherapeutic agents and antilipemic agents were the three most dispensed classes of medications among reMed participants and Quebec residents with public drug insurance. Despite that 8 of the 10 most prevalent classes of medications were in common between patients with public and private drug insurance, the average cost of medications was 22% lower among patients with public drug insurance ($37.92 versus $46.42 per prescription). CONCLUSIONS: Patients with private and public drug insurance were found to have a similar distribution of medications dispensed in community pharmacies. However, we found that patients covered by the public drug insurance plan paid on average their medications 22% less than patients covered by a private drug insurance plan. PHP28 THE COLLECTION AND USE OF “REAL WORLD” DATA: INDUSTRY VIEWS Devlin N1, Peperell K2, Gillen D3, Radway-Bright EL4 1 Office of Health Economics, London, UK; 2pH Associates, Marlow, UK; 3Pfizer, Surrey, UK; 4 Association of the British Pharmaceutical Industry, London, UK OBJECTIVES: ‘Real world’ data can improve understanding of the cost-effectiveness of health care technologies and are likely to become increasingly important in decisions affecting patient access and in drug development. However little is known about the investment industry currently makes in these studies, or their impact on HTA. The aim of this study was to examine industry views on the relevance and impact of ‘real world’ data on HTA and on the relative advantages and disadvantages of conducting these studies in the UK. METHODS: Survey of British pharmaceutical companies. RESULTS: Responses were received from 31 British pharmaceutical companies, covering most of the larger companies in terms of market share. Three quarters of these reported they were engaged in collecting real life data of some kind in the UK. Data collection efforts were more common for health service costs, resource use and treatment pathways; clinical outcomes; burden of disease; and patient outcomes data. The collection of data on patients’ ability to work; patient costs; and impacts on patients’ caregivers/family was less common. The availability of local expertise on HTA was a leading factor in choosing to undertake such studies in the UK; other important factors were the perceived importance of local data to NICE; and the influence of NICE on HTA decisions in other markets, particularly elsewhere in Europe. The main barrier to conducting studies in the UK was cost and perceived difficulty in obtaining ethics consent and NHS approval. Views on whether ‘real world’ evidence influenced NICE decisions were somewhat equivocal. However, participants felt that real world data were likely to be more important in providing evidence on health outcomes in chronic conditions, to track outcomes over longer time periods than is feasible in clinical trials, and for cancer and ‘orphan’ medicines. CONCLUSIONS: While industry recognises the value of ‘real world’ evidence, there are barriers to conducting such studies in the UK. A410 PHP29 THE EFFECTS OF NICE APPRAISALS ON PRESCRIBING AND COSTSHARING BEHAVIOR IN THE US Sepulveda B, Doyle J Quintiles Consulting, Hawthorne, NY, USA OBJECTIVES: As US health care increasingly looks towards proven clinical effectiveness for reimbursement decisions, we hypothesized that HTAs published by NICE would influence drug prescribing and patient cost-sharing expenditures in the US according to the nature of the published guidance. The primary objective was to determine whether trends in prescription volume (TRx) and the proportion of drug costs paid out-of-pocket (%OPC) of 10 drugs appraised by 7 HTAs since 2007. METHODS: Seven NICE HTAs since 2007, evaluating a total of 10 drugs, were analyzed. Six drugs were freely encouraged for utilization (lenalidomide, entecavir, tenofovir, donezepil, galantamine, rivastigmine), two were approved for restricted populations (cinacalcet, naltrexone) and the use of two of the drugs was restricted for all populations (telbivudine, memantine). TRx, and OPC as a percent of average retail price were collected quarterly for the same duration pre- and post-HTA until April 2010 using SDI’s VONA and VOPA databases. Statistical analyses were performed using one-way ANOVA; statistically significant results had P < 0.05. RESULTS: Comparing the periods before and after HTA publication, three drugs with positive guidance showed significant increases in prescription volume (P < 0.05). However, two drugs with advised use for restricted populations, and one with a negative guidance also showed significant increases. Interestingly, two drugs with positive guidance showed significant decreases in TRx. Only entecavir showed a significant decrease in %OPC, while all others failed to show a significant difference. CONCLUSIONS: NICE HTAs decisions appear to be associated with mixed effects on prescription utilization and expenditures in the United States. Though prescribing behavior was changed in the periods analyzed, further research is warranted to determine the true nature of that change. With increased cost-sharing in the health care environment in the US, it will be interesting to monitor the forces that might precipitate changes in OPC, whether related to HTA publications or not. PHP30 VANCOMYCIN UTILIZATION EVALUATION IN A TEACHING HOSPITAL BETWEEN FEBRUARY 2007 AND MAY 2008 IN IRAN Soleymani F Tehran University of Medical Science, Tehran, Iran OBJECTIVES: Increasing antimicrobial resistance is now a major problem in the world. Especially wide spectrum antibiotics resistance germs like vancomycin-resistant enterococci (VRE) should be dealt as soon as possible as an emergency conflict. Our study tries to reveal the amount of inappropriate use of vancomycin in a teaching hospital. METHODS: The study was conducted in a university hospital between February 2007 and May 2008. The hospital has 15 specialty and 5 subspecialty wards. a comprehensive questionnaire was designed. We random selected inpatients who received vancomycin. RESULTS: Forty four out of 45 patients had inappropriate indication and dose of vancomycin (97.7%). The most usage of vancomycin was recorded in hematology—oncology ward (71.11%) and then Intensive Care Unit (8.16%). Leukemia’s including Acute Myelogenous Leukemia (AML) and Acute Lymphoblastic Leukemia (ALL) were the most common reason of admission among patients. Cultures were negative (8.88%) despite great clinical evidence of infection. CONCLUSIONS: Vancomycin inappropriate use was high compared to other countries and it could be concerned as an area for implement educational and regulatory strategies by health policy makers to deal. However more detailed researches are needed to reveal the other aspects of this problem. Implementation of antibiotic protocols and standard treatment guidelines are recommended. PHP31 THE COMMERCIAL EFFECTS OF REFORMULATIONS OF EXISTING DRUGS Sepulveda B, Vincent L, White C Quintiles, Hawthorne, NY, USA OBJECTIVES: Given the high cost of creating a complete NME, it is not surprising that sixty percent of NDAs submitted to the FDA during the 1990s were for drugs based on an existing molecule (FDA, 2004). These products generally use three strategies for life cycle management (LCM): chemical reformulation, new drug combinations, or delivery reformulation, often resulting in an enhanced efficacy, safety, and compliance profile due to improved potency, tolerability, and duration of action. The objective of this study is to investigate the commercial effect of these LCM strategies. METHODS: Retail sales and prescription volume (TRx) of well-known reformulations (venlafaxine to desvenlafaxine, citalopram to escitalopram, trazodone extendedrelease, and paroxetine controlled-release) and reformulated combinations (esomeprazole/naproxen and fluticasone/salmeterol) were analyzed. Annual sales and TRx data were obtained since 2002 using SDI VONA. RESULTS: Interestingly, the two branded combination products analyzed had significant increases in TRx and in sales (P < 0.05) annually after the patent expiration of their original components, which showed significant annual decreases in sales and TRx (P < 0.05). The four reformulated products showed smaller changes than the combination products sales and TRx increases in the new branded products. Similarly, the utilization and sales decreases of the original components were smaller in the reformulated products than in the combinations. CONCLUSIONS: Given the higher cost of branded drugs, their unfavorable tier placement by payers with respect to their generic counterparts and the increased pressure to reduce drug spending, it may not be surprising that reformulated 13th Euro Abstracts generic products still enjoy sustained utilization. However, the premium for combination products that may increase patient adherence has not deterred utilization according to the present findings. PHP32 EFFECTIVENESS OF ANTIBACTERIAL THERAPY OPTIMIZATION BY CLINICAL PHARMACOLOGISTS Eliseeva E1, Lozinskaya L2, Solodovnikov V3, Gainullina Y3 1 Vladivostok State Medical University, Vladivostok, Russia; 2City Hospital #2, Vladivostok, Russia; 3Territory Fund of Compulsory Medical Insurance, Vladivostok, Russia OBJECTIVES: antimicrobials are vital medicines, although their excessive use leads to resistance. Impact of clinical pharmacologists on the rational use of antibiotics is worth scientific interest. The aim of the research: to analyze the dependence of the antibiotic therapy rationality on clinical pharmacologist’s administrative authority. METHODS: Segment of Territory Fund of Compulsory Medical Insurance database was analyzed including 51,128 informational units. Primarily we analyzed indications for separate antibiotis administration. Then rationality of antibiotic therapy according to nosologic classification was analyzed: pneumonia, chronic obstructive lung disease, acute cystitis, chronic pyelonephritis. The next step was a randomized analysis of antibacterial therapy rationality in three polyclinic groups. A clinical pharmacologist participated in antibiotic administrations in the 1st group of polyclinics. There was no clinical pharmacologist in the 2nd group. Clinical pharmacologist with administrative authority controlled antimicrobials administration in the 3rd group. RESULTS: Low rationality of inhibitor-protected aminopenicillins, ciprofloxacin, cefazolin use (64.82%, 23.39% and 1.77%) was revealed. At the same time inhibitor-protected penicillin’s administration frequency has increased in 1.5 times, and amoxicillin administration frequency has decreased in 5 times. It was found that the difference had been reliable only in the analysis of 2 groups: the first group of polyclinics (without clinical pharmacologist) and the 3rd group (participation of authorized clinical pharmacologist) (P = 0.046). The least rational antibiotic’s administration was noted in patients with erysipelas (54.5%) and most rational antibacterial therapy was given to patients with acute cystitis (94.1%). Difference in physician’s preferences is reliable (P = 0.037) only comparing the 1st and the 3rd groups of policlinics. CONCLUSIONS: The highest rate of the rational antibacterial therapy was registered in polyclinics and nosologies if clinical pharmacologist with administrative authorities had participated in antibiotic therapy. Our research makes it possible to substantiate additional administrative authorities for a clinical pharmacologist. Authorized clinical pharmacologists will promote to enhance rationality of antibacterial therapy. HEALTH CARE USE & POLICY STUDIES – Equity and Access PHP33 RATIONING IN PRACTICE- EQUITY IN WAITING TIMES FOR ELECTIVE SURGERY IN SWEDEN Tinghög G, Andersson D Linköping University, Linköping, Östergötland, Sweden OBJECTIVES: Health care can be rationed by various mechanisms. In publicly funded health care systems referring patients to waiting lists is a common administrative way to ration care. However, there is little systematic knowledge on how rationing through waiting time actually affects access to care among different socioeconomic groups. And since rationing through waiting lists most often is done implicitly rather than explicitly there is an obvious risk that less powerful groups of patients get discriminated, resulting in a longer waiting time before being treated. The overall objective of this study was to explore the association between patients’ socioeconomic status, ethnical background and waiting time within non-acute surgical specialities. METHODS: Days on waiting list was used as dependant variable in a multiple regression model. This data was collected from the county council of Östergötland in Sweden (N = 4634). Data on disposable income, ethnicity and workforce activity were retrieved from national registers. RESULTS: Examining disposable income as a potential predictor, we found that lower disposable income was significantly associated with longer waiting time in orthopedics (P = 0.05) and general surgery (P = 0.05). In orthopedics, the lowest income group waited on average 28% longer than the highest income group. Examining ethnicity as a potential predictor, the only significant association found was in gynecology where patients with foreign origin surprisingly waited on average 40% shorter than patients with Swedish origin. For workforce activity, we found that patients excluded from the workforce were significantly associated with longer wait in ophthalmology. CONCLUSIONS: Our results reveal horizontal inequalities in access within several non-acute surgical specialties. However, the mere association between socioeconomic factors and inequalities in waiting times cannot alone support judgements about inequity. Hence, there is both a descriptive and normative question that needs to be assessed when investigating the reasonableness of rationing by waiting time. PHP34 PATIENT ACCESS TO INNOVATIVE MEDICINES IN HUNGARY 1 1 1 2 Botlik O , Kalotai Z , Dessewffy Z , Kalo Z 1 Novartis Hungary, Budapest, Hungary; 2Eötvös Loránd University, Budapest, Hungary OBJECTIVES: Patient access to innovative medicines has critical importance from the societal perspective, pharmaceutical innovation contributes to increase in health capital. Delay in the pricing and reimbursement process limits the access of patients 13th Euro Abstracts to new medicines. Our goal was to determine the time period between the registration and reimbursement date in Hungary. METHODS: We selected all newly reimbursed pharmacy drugs between January 2004 and April 2010 and looked for the date of registration and reimbursement in public websites and Bulletins of EMEA/EMA, National Institute of Pharmacy and National Health Insurance Fund. We excluded hospital only medicines and drugs with special reimbursement budget from the analysis due to the lack of transparency of reimbursement dates in publicly available data sources. RESULTS: 106 newly reimbursed innovative medicines between January 2004–April 2010 were included into the analysis. The average time period between registration and reimbursement was 677 days. CONCLUSIONS: Hungary joined the European Union in May 2004 and implemented the EU Transparency Directive. Time to reimbursement of innovative medicines in Hungary is significantly longer than the recommended 90 + 90 days for pricing and reimbursement process set by Transparency Directive. The pricing and reimbursement process in Hungary takes more time than in 15 European countries included in the EFPIA Patients W.A.I.T. indicator database (from 101 to 403 days). Acceleration of patient access to innovative medicines is highly recommended in Hungary. PHP35 THE IMPACT OF THE HOSPITAL FUNDING SYSTEM ON THE RANGE OF THE EXPENSIVE DRUGS AVAILABLE IN FRENCH AND ENGLISH HOSPITALS Gridchyna I, Aulois-Griot M, Maurain C, Bégaud B Université Victor Segalen Bordeaux 2, Bordeaux, France OBJECTIVES: In French and English hospitals, there are a lists of drugs financed out of scope of casemix-based payment system that are Payment by Result (PbR) and “Tarification à l’activité” (T2A). We examined a difference in the range of these drugs in both countries. METHODS: In the study were included the drugs registered on the list “en sus” in French system, the drugs from the high cost drugs list (HCD) and from the oncology regimens list in English system. The information is available in official sources. The number and overlap of entities excluded from the casemix-based payment system in two countries were determined, as well as similarity rate. RESULTS: 210 entities are financed out of scope of casemix-based payment system in England and 101 in France. 69% (145/210) of entities excluded from PbR are not on the list “en sus”. Around 36% (36/101) of entities excluded from T2A are not on the English lists. There are 65 entities common for both lists; 51% (33/65) are from ATC class L (antineoplastic and immunomodulating agents). Four ATC classes have none common drugs. The aim of the list in two systems is fair reimbursement of the expensive drugs within the casemix-based payment system. In French system this list is used also to improve the access to the expensive and innovative drugs. So, 50% (73/145) of the entities excluded from PbR and not included on the list “en sus” are on another list in French system, the retrocession list. CONCLUSIONS: There is a difference in the range of drugs financed out of scope of casemix-based payment system in French and English hospitals. More drugs are excluded from the casemix-based payment system in England, but it does not facilitate access to new drugs. PHP36 THE AVAILABILITY AND FUNDING OF ORPHAN DRUGS IN BOSNIA AND HERZEGOVINA IN COMPARISON WITH NEIGHBORING COUNTRIES Catic T, Bajramovic A Society for Pharmacoeconomics and Outcomes Research in Bosnia and Herzegovina, Sarajevo, Bosnia OBJECTIVES: The aim was to examine the current availability and funding of orphan drugs in both entities of B&H and compare the obtained data with neighboring countries (Croatia, Serbia and Montenegro) and the EU. METHODS: We have analyzed the current published list of medicines in B&H and neighboring countries. We have compared the drugs that have the status of orphans according to the Orphanet report. RESULTS: In BiH there are no lists of orphan drugs while some of them are included in the list of chemotherapeutic agents and drugs for specific diseases (RS). Only Croatia has made a special list of expensive medicines containing drugs for treatment of hereditary enzyme deficiency. All countries have imatinib reimbursed. Only in the RS and Serbia thalidomide is reimbursed, and Serbia has listed sildenafil, zinc acetate and busulfan. Present practice in all countries is that patients apply individualy for orphan drugs reimbursement approval to HIFs. CONCLUSIONS: In order to improve access to orphan drugs it is necessary to adopt a national policy which will be harmonized with the EU. Decisions on the reimbursement must be based on real possibilities and it is necessary to implement appropriate registries for future resource allocation decissions. PHP37 THE IMPACT OF UNIVERSAL COVERAGE ON EQUITY IN HEALTH CARE FINANCE AND FINANCIAL RISK PROTECTION IN THAILAND Prakongsai P International Health Policy Program (IHPP), Nonthaburi Province, Thailand OBJECTIVES: To assess the impact of achieving universal coverage (UC) on equity in health care finance and on financial risk protection from expensive medical care costs for Thai households. METHODS: Secondary data analyses using nationally representative household surveys conducted by National Statistical Office, the Socioeconomic Survey 2000 (prior to UC) and 2002–2006 (after UC) to analyze changes in progressivity of overall health care finance and different health financing sources. A411 The share of households facing catastrophic health expenditure in the poorest and richest income quintiles prior to and after achieving UC was also assessed. RESULTS: The financing of the Thai health care system became more equitable after the UC policy was implemented. Improved financial risk protection after achieving UC was observed due to the comprehensive benefit package and literally free at point of services. The Kakwani index value for overall health care finance changed from −0.0038 (regressive) in 2000 to positive (progressive) values of 0.0014, 0.0342 and 0.0406 in 2002, 2004 and 2006, respectively. The share of households facing catastrophic spending on health decreased from 5.4% in 2000 to 2.0% in 2006. The 1st (poorest) quintile experienced a 77.5% reduction in the proportion of households facing catastrophic health expenditure, while there was a 41% reduction in the share of households in the 5th (richest) quintile. CONCLUSIONS: Factors contributing to equitable health finance are: the increasing share of progressive financing sources in particular direct tax; the decreasing share of the regressive out-of-pocket payments for health. Using general taxation to finance the poor and the informal sector not only helps reach universal coverage, it is also the most progressive financing source. Various factors contribute to the low incidence of catastrophic heath expenditure: comprehensive benefit package covering almost all health services which are free at point of use, and well-functioning primary care providers. PHP38 HOW EQUITABLE OF HEALTH SERVICE USE AND GOVERNMENT SUBSIDIES IN THAILAND AFTER ACHIEVING UNIVERSAL COVERAGE? Prakongsai P International Health Policy Program (IHPP), Nonthaburi Province, Thailand OBJECTIVES: To assess trends of equity in health service use and distribution of government subsidies to ambulatory services and hospitalization across income gradients of the Thai population prior to and after implementation of the universal coverage (UC) policy in 2002 and explained how such equity has been achieved. METHODS: Secondary data analyses using nationally representative household surveys conducted by National Statistical Office, the Health and Welfare Survey (HWS) 2001 (prior to UC) and 2003–2007 (after UC) to analyze equity in health service use at different health care levels and the distribution of government subsidies for health. The analytical approach employed a standard method for health equity analysis of the large scale household surveys proposed by O’Donnell et al. RESULTS: Outpatient and inpatient service use of the Thai health systems were both pro-poor before achieving UC in 2002 due to various government interventions in extending health insurance coverage and countrywide distribution of health service infrastructure, and the significant increase in human resource production. After the UC policy implementation, the propoor service utilization was further progressed. Overall, public subsidies for health were found to be pro-poor for both outpatient and inpatient services with the concentration indexes of −0.226, −0.186 and −0.180, in 2003, 2006 and 2007, respectively. District health provider networks, in particular health centres, district and provincial hospitals are the major determinants of the pro-poor distribution of service utilization and public subsidies, due to their geographical proximity and better access by the poor. A comprehensive benefit package and the provision of services that are free at the point of use resulted in the pro-poor benefit incidence. CONCLUSIONS: The pro-poor outcome is the result of an availability of functional primary care at the district level, and implementation of the UC policy which focuses on contracting primary care networks at the district level. PHP39 A COMPARISON OF GENERIC AND ORIGINATOR BRAND DRUG PRICES BETWEEN JORDAN AND THE UNITED KINGDOM El-Dahiyat F1, Kayyali R1, Alabbadi I2 1 Kingston University, London, UK; 2University of Jordan, Amman, Jordan OBJECTIVES: The study aimed to compare the retail prices of generics and originator brand for five drugs between Jordan and United Kingdom and to investigate the relation between number of generics available, retail price of originator and generic (s) and the effect of time in the market on these prices. METHODS: Prices of originators and generics and the number of generics available in each market were obtained from the Jordanian Food and Drug Administration, Royal Pharmaceutical Society of Great Britain, British National Formulary and Chemist & Druggist generics list. The prices were converted to British Pounds expressed per one dose unit. All data was tabulated in spreadsheets; prices were compared between the two countries at different preset times. RESULTS: The generics of all drugs investigated appeared in the Jordanian market before patent expiry of their originator worldwide due to lack of patency regulations in Jordan at the launch time of drugs under investigation (before 2004). Unlike the UK, the prices of originator drugs in Jordan did not change when the first generic was introduced to the market. The price of generic drugs have dropped dramatically in the UK at time of first generic launch approximately by 90% compared to 15% in Jordan. There was no apparent correlation between the numbers of generics available or the number of years of first generic being in the market and the prices of the drugs investigated in both countries. The current prices of all investigated drugs in Jordan are higher than the UK particularly for the generics. CONCLUSIONS: Although much lower income per capita in Jordan, generic drugs are more expensive than the equivalent prices of same drugs in the UK A412 PHP40 REIMBURSEMENT RECOMMENDATION PROCESS IN BRAZIL: CITEC SUBMISSIONS PROFILE Araújo GTBD, Fonseca M AxiaBio, São Paulo, Brazil OBJECTIVES: To access and understand the main characteristics of the Reimbursement submission process and HTA recommendations in Brazil. METHODS: This is a crossed analysis of secondary data displayed by the Brazilian new technology incorporation committee (CITEC) regarding the quantity and characteristics (disease, submitter, type of health technology, and year of submission) of the submitted protocols versus quantity and characteristics of the recommended health technologies. RESULTS: In 3 years of existence, 216 Health technologies (mean of 48 submissions per year) were appraised by the commission resulting in 22 positive recommendations and XX Negative recommendations. For the reconvened Health technologies, the median time from submission and recommendation was 1 year. Oncology and Rheumatology are the main submissions groups by disease (12% and 11%, respectively) and Hepatology and Rheumatology are the most recommended groups by disease (18% and 14% respectively). Manufacturers are the main submitters (74%), followed by the own federal government. 5% of the submissions was made by mixed submitters that can be chacterized by research partnerships between manufacturers, academia and government. CONCLUSIONS: This finds suggests that the conduction of HTA recommendation and adoption process is already a reality in Brazil and the decisions are based not anymore only in political issues but are strongly based on clinical and economical issues too. As a new process in the country, modifications as time to response and establishment of a Brazilian ICER during the next years can improve the results for CITEC and submitters. HEALTH CARE USE & POLICY STUDIES – Formulary Development PHP42 CURRENT AND FUTURE USE OF PHARMACOECONOMIC AND OUTCOMES RESEARCH DATA IN DECISION-MAKING IN THE USA Holtorf AP1, Keskinaslan A2, Biskupiak J3, Brixner D4 1 Health Outcomes Strategies GmbH, Basel, Switzerland; 2Novartis Pharma AG, Basel, Switzerland; 3University of Utah College of Pharmacy, Salt Lake City, UT, USA; 4University of Utah, Salt Lake City, UT, USA OBJECTIVES: To understand how PEOR data are used in decision making today and how this may change in future. The methodology and quantity of Pharmacoeconomics and Outcomes Research (PEOR) has advanced considerably over the last decade. With increasing utilization of health care, prioritization becomes unavoidable and PEOR evidence should be increasingly valuable to supplement efficacy and safety data for reimbursement and access decisions. METHODS: An internet based survey with 30 items was conducted among decision makers from the USA in April 2010. RESULTS: The 76 respondents represented organizations with national (54%), regional (32%) or local coverage (14%). Membership varied from 50,000 to 100million with a median of 500,000. While 70% of the respondents claimed that PEOR data are used regularly, only 5% indicated coverage by the bylaws of their organizations, and 66% of the users indicated that there is no quality standard for PEOR being used by the organization. The majority of the respondents expected increasing use of PEOR in the future (77%). Organizational requirements to such adoption included ‘more HEOR expertise’, better definitions and standards of methods, in-house data analysis expertise, and regular reevaluation. Outcomes based evidence is increasingly used in contracting (37%) and 73% expected an increased use in future with the perceived value of reduced financial and clinical risk. CONCLUSIONS: PEOR is used often in decision making in the USA. However, few organizations defined quality criteria for PEOR. There is a consistent expectation that there will be an increasing use of this data, along with a concern regarding a lack of expertise to evaluate this information in their organizations. Improved standards for quality control of such evidence are needed. Another survey will be targeted to the systems and decision makers in selected European countries to compare the utilization of PEOR in the different decision making environments. PHP43 ASSESSING THE ADDED VALUE OF NEW DRUGS: A MULTIDIMENSIONAL APPROACH Paladio N, Sunyer B, Almazan C, Solà-Morales O Catalan Agency for Health Technology and Research, Barcelona, Spain OBJECTIVES: In November 2008, the Program for the Assessment, Monitoring and Reimbursement of High Complexity Drugs (PASFTAC) took effect in Catalonia and the Committee for the Assessment of Hospital Use Drugs (CAMUH) was consequently appointed to give recommendations on medicines use. Our aim is to develop a tool that embraces the multidimensional aspects of added value in drug assessment. METHODS: A qualitative approach (adapted expert panel) has been used. Participants included decision makers, health economists, experts on bioethics, clinicians and pharmacists. A previous detailed review of literature was conducted for the topic under discussion. RESULTS: Several dimensions and items were identified and discussed in a single meeting. Main identified dimensions included efficacy, safety and cost-effectiveness as well as clinical relevance of outcomes, unmet therapeutic needs, ethics, and cost-oportunity. Individual interviews were further conducted. Discussion and rating of dimensions and items within these will be performed in a second meeting. A pilot test will be conducted after a minimum of 10 drugs are assessed by the CAMUH (expected by September 2010). CONCLUSIONS: In a context of limited evidence 13th Euro Abstracts dimensions other than classical efficacy, safety and cost-effectiveness need to be considered at the time of assessing added value. This new tool will contribute to that task and standardize the process at a local level. HEALTH CARE USE & POLICY STUDIES – Health Care Costs & Management PHP44 APPLICATION OF PERATO’S PRINCIPLE TO UNDERSTAND COST DRIVERS IN HEALTH CARE Long S1, Misra A2 1 Thomson Reuters, Cambridge, MA, USA; 2Thomson Reuters, Washington, DC, USA OBJECTIVES: To determine if Pareto’s principle (also known as the 80–20 rule) applies to health care expenditures in the US, and to describe the patients and treatments considered to be “cost drivers.” METHODS: The study population consisted of persons <65 years that were continuously enrolled in the Thomson Reuters MarketScan Commercial Database during the years 2004–2008. Individuals were rank ordered highest to lowest according to their 2008 health care expenditures and assigned to the Cost Driver (CD) Cohort if their costs were within the 80% cumulative sum of the overall total expenditures. Clinical and cost characteristics were compared between 2004 and 2008 using Compound Annual Growth Rates (CAGR). RESULTS: 2008 total expenditures were $27.5M among 4.2M covered lives; the CD cohort represented 27% of the overall study population. Mean annual expenditures in the CD Cohort were $15,581 (16.1% CAGR). Outpatient services, inpatient services, and outpatient pharmaceutical prescriptions represented 42%, 29% and 28% of total expenditures in 2008. From 2004 to 2008, pharmaceutical expenditures grew 23% CAGR, while outpatient and inpatient expenditure growth was 15% and 13% respectively. Drivers of medical service expenditures in the CD group were visits related to osteoarthritis, renal failure, vaginal deliveries, bone and joint disorders, and low back pain. Drivers of outpatient pharmaceutical expenditures in the CD group were antihyperlipidemics, gastrointestinal drugs, antidepressants, antidiabetic agents, and anticonvulsants. CONCLUSIONS: In a large US database of commercially insured persons, 27% of individuals were accountable for 80% of health care costs incurred during the year 2008. Payers should consider targeted interventions for their cost driver populations; for example, intense targeted programs aimed at reducing obesity could yield large cost offsets since much of the incurred costs were for treatment of conditions that are sequelae of obesity. PHP45 COSTS OF CARE UNDER A HEALTH CARE PLAN IN THE LAST 5 YEARS OF LIFE OF BENEFICIARIES—DATA FROM THE REAL WORLD Reis Neto J, Stefani S CAPESESP/CAPESAUDE, Rio de Janeiro, RJ, Brazil OBJECTIVES: To examine, in the ‘real-world’ scenario of a Brazilian health care plan, the effect of proximity to death in beneficiaries on health expenditure during the last 5 years of life. METHODS: Retrospective analysis from data stored on the computerised system of the health care plan of 1897 beneficiaries who died between January 2007 and June 2009. The outpatient and hospital health services in the year of death and in the 4 preceding years were examined. The analysis was restricted to total direct costs with medical-hospital care. The variables (demographic, clinical and costs) were subjected to statistical treatment considering a confidence interval of 95%. RESULTS: Out of a total of 1897 deaths analyzed, the majority were male (60.4%) and were aged 60 or over (77.0%). The overall mean age of the deceased was 70.6 years (CI95% 69.9–71.4). The cost over the 60 months of the study was US$64,427,852. This total, 66.8% was reimbursed in the year of death and the clinical admissions corresponded to 89.2% of the total. Among the main causes of death, neoplasia and chronic diseases had a heavier influence on expenditure. CONCLUSIONS: The rise in spending on health appears to be directly associated with the effect of increased age together with a second factor considered to be important, the proximity to death. It is estimated that in 2050 Brazil will have a population of 64 million individuals aged 60 or over, representing one of the main challenges to the adequate allocation of resources. This study suggests that the costs in the last year of life should be considered in the projections for spending on health care, in the same way as the ageing factor. One of the alternatives that appears to be important at the end of life is palliative care. PHP46 COSTS OF TREATING TERMINAL PATIENTS IN DIFFERENT BELGIAN HEALTH CARE SETTINGS Simoens S1, Kutten B2, Menten J2 1 K.U. Leuven, Leuven, Belgium; 2University Hospitals Leuven, Leuven, Belgium OBJECTIVES: This study aims to measure the costs of treating terminal patients in Belgian hospitals and in nursing homes from the health care payer perspective. Also, this study compares the costs of palliative care with those of usual care. METHODS: A multi-centre, retrospective cohort study enrolled patients in acute wards and in palliative care units from a representative sample of hospitals and from a representative sample of nursing homes. Health care costs included fixed costs and charges relating to medical fees, pharmacy and other charges. Data sources consisted of hospital and nursing home accountancy data and patient invoice data. The price year was 2007/2008. RESULTS: Six hospitals and nineteen nursing homes participated in the study, generating a total of 327 patients. Mean hospital costs per patient per day amounted to c391 (± c156). Mean hospital costs of patients receiving usual care in an acute ward (c340 ± 143) exceeded costs of palliative care in an acute ward (c283 13th Euro Abstracts ± 109) (P = 0.025). Mean nursing home costs amounted to c3243 (±2700) per patient during the final month of life. Mean nursing home costs per patient of c3822 (±3232) for patients receiving usual care were higher than costs of c2456 (±2117) for patients receiving palliative care (P = 0.068). CONCLUSIONS: This study suggests that palliative care models in acute hospital wards and in nursing homes need to be supported because such care models appear to be less expensive than usual care and because such care models are likely to better reflect the needs of terminal patients. PHP47 EFFICIENCY IN DRUG PRESCRIPTION MEASURED BY THE APPLICATION OF ADJUSTED CLINICAL GROUPS IN THIRTEEN SPANISH PRIMARY CARE CENTRES Sicras-Mainar A1, Velasco-Velasco S2, Llopart-López JR2, Navarro-Artieda R3, Blanca-Tamayo M2 1 Directorate of Planning, Badalona Serveis Assistencials, Badalona, Barcelona, Spain; 2 Badalona Serveis Assistencials SA, Barcelona, Spain; 3Hospital Universitari Germans Trias i Pujol, Barcelona, Spain OBJECTIVES: To measure the efficiency of pharmacy resources utilization in 13 Primary Care centres by the retrospective application of Adjusted Clinical Groups (ACG) in a usual clinical practice setting. METHODS: Retrospective study carried out on the basis of the clinical records from all the attended patients along the year 2008. Main variables: age, sex, case-mix/episodes, visits, pharmacy costs, 13 centres, physician and service (Family Medicine or Paediatrics). ACG grouper (Starfield and Weiner, Johns Hopkins University, n = 106) classifies each patient in a unique category of similar resource consumption. A Receiver Operating Characteristics (ROC-area under the curve) analysis was done to assess the predictive value of the model. The Efficiency Index (EI) was obtained as the quotient between the observed and the expected pharmacy costs according to ACG distribution (indirect standardization). The statistical package SPSS was used (P < 0.05). RESULTS: A total of 227,235 patients were included. Average number of episodes 4.5 ± 3.2 and visits 8.1 ± 8.2; mean age: 44.1 ± 23.7 years; and 55.6% of males. Intensity of utilization: 77.4%. Costs of drug prescription: c70.6 millions (47.6% of the total costs). Mean cost was c310.8 ± 681.2. ROC curve analysis for episodes: 0.588 (p = 0.001); sensibility: 37.3%, specificity: 73.1% and intra-class correlation coefficient C: 0.732 (P < 0.001). The EI for each centre was analyzed, P < 0.0001. Furthermore, differences between family physicians and paediatricians were observed (range: 0.55–1.46), P < 0.001. CONCLUSIONS: Results show a wide variability in the costs of pharmacy within centres and physicians. ACG provide an adjusted approximation to efficiency in pharmacy costs. Efficiency must not be considered as an isolated dimension of quality. The determination of the El could lead to a better knowledge of the prescription profile from individual physicians and/or primary care teams. PHP48 ECONOMIC EVALUATION OF HEALTH CARE INTERVENTIONS DURING MORE THAN 25 YEARS IN SPAIN (1983–2008) Catalá-López F1, Garcia-Altés A2 1 Spanish Medicines and Healthcare Products Agency (AEMPS), Madrid, Spain; 2King’s College London, London, UK OBJECTIVES: Economic evaluation has been promoted as a tool to guide decisionmaking processes regarding health care resources’ allocation and in the adoption of health care technologies. We analyzed the evolution and the main characteristics of economic evaluations of health care interventions done during the period 1983–2008 in Spain. METHODS: Observational descriptive study. We performed a systematic review in the main bibliographic databases (PubMed/MEDLINE, SCOPUS, ISI Web of Knowledge, CRD, IME, IBECS) and manually through Internet in journals and public reports. There were predefined inclusion and exclusion criteria, and a set of variables to analyze the characteristics of the selected reports. RESULTS: In total, 477 studies fulfilled inclusion criteria. Some of the studies characteristics were: cost-effectivenes analysis (62.5%), decision analysis techniques (34.0%), heath care system perspective (42.1%), therapeutic interventions (70.0%) and non explicit financing (44.0%). The geographical distribution for Spanish regions of the first authors was: Catalonia (29.3%), Community of Madrid (23.7%), Andalusia (6.7%) and Region of Valencia (6.3%). a total of 50.9% of the principal authors were employed at hospital centers. The most commontly disease conditions were: cardiovascular diseases (15.7%), infectious and parasitic diseases (15.3%) and malignant neoplasms (13.2%). a total of 82.2% of the reports provided recommendations guide for decision making. CONCLUSIONS: An increasing number of studies was observed. Identified reports combined heterogeneity in the quality of the information brought with regard to analysis methods, data sources, type of interventions, or disease conditions. It is suggested to do more efforts for improving the quantity and quality of reports in public health interventions. PHP49 INFLUENCE OF THE CHRONIC DISEASE IN HEALTH RESOURCES USE: ECONOMICS CONSECUENCES AND RISK STRATIFICATION Sicras-Mainar A1, Navarro-Artieda R2, Baulenas-Parellada D3, Frías-Garrido X3, Perez S3 1 Directorate of Planning, Badalona Serveis Assistencials, Badalona, Barcelona, Spain; 2Hospital Universitari Germans Trias i Pujol, Barcelona, Spain; 3Directorate of Primary Care. Badalona Serveis Assistencials SA, Barcelona, Spain OBJECTIVES: The Chronic disease are ones of high prevalence, duration and slow progression, its cure could not be predicted clearly or will never be. In general, are related to lifestyles and population that is older. To determine the direct sanitary cost depending on chronic comorbidity grade (risk) in patients attended in a primary care A413 setting (PC). METHODS: Retrospective multicentre design. Patients over 14 years were included, pertaining at 6 PC teams that demand assistance during year 2008. Main measures: sociodemographics, risk/casuistic/comorbidity, Charlson index (severity) and direct cost models. The chronic comorbidity was classified beginning from Adjusted Clinical Groups. It was obtained the Resources Utilization Bands (RUB) per patient (rank: 1- healthy user to 5- high morbidity). The complexity/comorbidity was grouped in 10 categories (6 expert forum). Fixed/semi fixed cost were considered (functioning: salary, services, purchases) and variable (tests, referrals, drugs). Explanation power calculation: determination coefficient (R2). It was made an ANOVA analysis for the correction (age, sex, comorbidity) of the models (procedure: Bonferroni). SPSSWIN program; P < 0.05. RESULTS: A total of 69,653 patients, age-average: 47.6 ± 18.8 years; women: 54.1%; high morbidity: 4.8%; chronic disease: 3.8 ± 2.3 and total cost: c51.3 million (fixed cost: 12.3%). The 35.4% (n = 24,670; CI: 34.5–36.3%) showed ≥5 chronic disease. Binary correlations: comorbidity-BUR: 0.716; cost-comorbidity: 0.596; age-comorbidity: 0.429; P < 0,001. Osseo-muscular illness (38.1%), mental (31.6%) and cardiovascular (30.4%) were the same frequencies, P < 0.001. The unitary average of the cost corrected was of c736.74 ± c921.97, with comorbidity ranks between: 1 = 309.62; 5 = 842.99; 10 = c2,354.05, respectively, P < 0.001. Predictive model (R2): age = 25.7%; age-sex = 26.5%; age-sex-comorbidity = 60.3%, of the cost. Women showed more comorbidity. CONCLUSIONS: Chronic comorbidity is associated with a sanitary cost increase. The number of comorbidities explains the major part of the cost. Patient knowledge of risk/complexity enable us doing different strategies of preventive/cure intervention. KEY WORDS: comorbidity, cost, resource use, risk. PHP50 FRAGMENTED HEALTH CARE SYSTEM—SOLVING THE JIGSAW PUZZLE 1 1 2 3 3 Miksch F , Urach C , Popper N , Weisser A , Endel G 1 Vienna University of Technology, Vienna, Austria; 2Dwh Simulation Services, Vienna, Austria; 3 Main Association of Austrian Social Security Institutions, Vienna, Austria OBJECTIVES: Austria’s health insurance system is divided into 13 sickness funds. Each fund has its own fee structure with individual names and codes for every procedure performed in an outpatient setting. Outpatient setting includes general practitioners, specialists, ambulatories and institutes. Each sickness fund has general contracts for GPs and specialists while each institute has its own contract. These contracts partly include procedures that are not comparable with any procedures of other sickness funds. Some sickness funds pay physicians and institutes different fees for same procedures. Moreover, the payment depends on the number of procedures performed by a physician or institute within a certain period of time. In order to evaluate procedure data, the so called meta-fee-structure, a set of pre-defined, standardized procedures that are mapped by individual fee-structures, was developed. METHODS: Given data with frequencies and costs of procedures for every sickness fund we are going to propose methods for examining and compare these data: Comparing the cumulative costs, frequencies, the average rates for procedures and the rates for different institutes and physicians paid by different sickness funds. All methods are then applied on real data in the field of radiology. RESULTS: By using this process with real world data it is possible to show the potential of these methods, lacks in data quality and the limitations of the meta-fee structure. Furthermore it is possible to point out proposals where procedure costs should be examined more closely and maybe health care costs could be reduced. CONCLUSIONS: Due to the heterogeneous health care system of Austria there is a wide variety of issues to be addressed when analyzing data. Although the proposed methods are very general they have to be adapted to the actual problems. Knowledge about data origin is crucial when choosing methods to get high quality results. PHP51 ROUTINE REPLACEMENT OF PERIPHERAL INTRAVENOUS CATHETER VERSUS CLINICALLY INDICATED REPLACEMENT: A COST COMPARISON STUDY FROM THE PUBLIC PAYER PERSPECTIVE Tolentino ACM, Takemoto ML, Fernandes RA, Passos RB, Cukier FN ANOVA—Knowledge Translation, Rio de Janeiro, RJ, Brazil OBJECTIVES: IV catheterization is the most common invasive procedure among hospitalized patients and the widest practiced to prevent complications is routine replacement of the catheter (RCC) in fixed intervals. a meta-analysis (Webster 2008) showed no significant differences in terms of efficacy from RCC and replacement only when clinically indicated (RCI). This study aims to compare costs and consequences of RCC versus RCI from the public payer perspective. METHODS: Efficacy data was obtained from Webster 2008 which showed no clinical benefits of RCC over RCI. Data from the Brazilian Hospital Information System (SIH/DATASUS) from January 1–December 31, 2009 was used to define the annual number of admissions of adult patients in public hospitals wards, assuming RCC as the current practice (no ICU patients included). The mean length of stay (LOS) and the mean time to replacement (TTR) were used to calculate the number of replacements in each scenario. Resource utilization was estimated through published data and unit costs were obtained from Brazilian official price lists. RESULTS: A total of 8,985,758 hospitalizations were identified in the database in 2009 with mean LOS of 5.53 days (132.66 hours). According to a published randomized clinical trial included in Webster 2008, mean TTR for RCC and RCI was 66.5 and 90.6 hours, respectively, resulting in average 1.99 and 1.46 replacements per hospitalization in each setting. The estimated cost of replacement was 2.54BRL (nursing time and medical supplies). For all admissions, the total cost for RCC and RCI was 45,556,510BRL and 33,438,277BRL. The estimated A414 savings were 12,118,232BRL/year for the public health care system or 1,348BRL/year per 1,000 admissions. CONCLUSIONS: RCI has shown similar efficacy when compared to RCC with fewer costs. The cost difference was mild in magnitude but when extrapolated to a large-scale perspective these results reinforce the need of evidencebased decision making and rational resource allocation. PHP52 HIGH AND INTENSIVE UTILIZERS IN HEALTH CARE—A STRATEGIC CHALLENGE FOR MEDICAL SUPPLY IN THE COMPULSORY HEALTH INSURANCE Schoenermark MP1, Beindorff N2, Kielhorn H1 1 Hannover Medical School, Hannover, Germany; 2Schönermark.Kielhorn and Collegen, Hannover, Germany OBJECTIVES: In Germany, the financing mechanism aggravates the cost issue of so called high-utilizers. So far, there were no care management concepts as the mostly multimorbid, complex medical conditions were judged as individual, non-influenceable cases. We developed a method to reveal issues in treatment and care of this special insurance population to increase efficiency. METHODS: The most critical challenge for insurance companies lies in the identification of relevant insured, in order to specifically target the high effort of control and management. In this project we analyzed the secondary data of the members of a German insurance company in the course of four years. Based on this we first differentiated high-utilizers from average-utilizers. In the following we deflected by which means existing care management concepts could be complemented. Based on the longterm data observation we developed a prognostic model to predict future high-utilizers. RESULTS: High-utilizers were defined as the 5% most expensive insurance-members, who generated 50% of total spending. We distinguished high-utilizers who caused the main costs in one specific care sector (pharmaceuticals and hospital) and patients who caused costs in multiple sectors (transsectoral). In addition, ultra high-utilizers were considered separately due to their extreme cost provocation. While it is possible to manage the first two groups by a combination or extension of existing cost reduction measures and care approaches such as disease management programs, ultra high-utilizers should be addressed by an individual and specifically developed case management. CONCLUSIONS: This segment of insured demands a stringent, integrated approach in order to efficiently employ the available financial resources. This study aims to explain a practical system for the controlling and management of expenses caused by high-utilizers based on our conclusion that the early identification and the specific management of high-utilizers in health care holds high financial potential and targeted programs are promising for care optimization. PHP53 COST ANALYSIS OF ANKARA UNIVERSITY SCHOOL OF MEDICINE HOSPITALS Esatoglu AE1, Agirbas I1, Doganay P1, Goktas B1, Akbulut Y1, Ozatkan Y2, Toruner M2, Gok H2, Atasoy KC2, Ugurluoglu E1, Cakir SU2, Okten I2 1 Ankara University Faculty of Health Sciences, Ankara, Turkey; 2Ankara University Faculty of Medicine, Ankara, Turkey OBJECTIVES: The purpose of this study was to determine the unit cost of the main production centers of Ankara University Faculty of Medicine Hospitals in 2008. METHODS: Expenses and costs of the two hospitals of Ankara University (Ibni Sina and Cebeci Hospitals) were obtained from the Hospital Information System and Revolving Fund distributed to the main production centers by using a step-down allocation method with five allocation steps. Unit costs were calculated in the last step. To determine the cost for outpatient and inpatient units, the data from the fifth allocation including the number of outpatients and inpatients as well as the number of inpatient days was used. RESULTS: The total cost of Ibni Sina Hospital (a 931-bed facility) and Cebeci Hospital (a 1153-bed facility) were calculated as c55,137,708.41 and c49,709,800.84 respectively. In Ibni Sina Hospital, among outpatient units, Dept. of Hematology had the highest, while the Aphaeresis Unit had the lowest cost. As to the inpatient unit costs, Dept. of Internal Diseases had the highest, and Dept. of Ear, Nose and Throat had the lowest cost. In Cebeci Hospital, Dept. of Algology had the highest outpatient unit cost, and Consultation-Liaison Psychiatry had the lowest outpatient unit cost. In terms of number of inpatients, Dept. of Pediatrics had the highest, while the Aphaeresis Unit had the lowest unit cost. Nuclear Medicine Department had the highest cost, while the Aphaeresis Unit had the lowest clinic unit cost with respect to inpatient hospital days. CONCLUSIONS: The results of this study show that the unit costs of outpatient clinics are higher in the departments of internal medicine compared to the departments of surgery whereas the reverse is true for inpatient clinics. Being aware of unit costs in a large-size hospital would improve strategic decision-making process including effective financial management, health care service planning and human resources management. PHP54 BUDGET IMPACT OF ORPHAN DRUGS IN DENMARK COMPARED TO OTHER EUROPEAN COUNTRIES Heemstra HE, Hensen M, Meijboom MJ Pharmerit International, Rotterdam, The Netherlands OBJECTIVES: Budget impact (BI) of orphan drugs (ODs) has increased over the past 10 years as more ODs entered the market since the introduction of the EU Regulation on Orphan Medicinal Products. The aim of this study is to determine the BI of ODs in a selection of countries where a large number of authorized ODs are reimbursed. METHODS: Public data on OD expenditure in Denmark was collected. The BI was 13th Euro Abstracts calculated per product for the period 2005–2009. Subgroup analyses were performed for different types of treatment setting and ATC classes. The results of Denmark were compared to data collected for France (total OD costs for 2002–2009 specified per treatment setting) and for Belgium (budget estimates per product for 2008). RESULTS: Total BI of ODs in Denmark increased from 20.3Mc for 13 reimbursed ODs out of 22 (59%) authorized ODs in 2005 to c58.3M for 41 out of 60 (68%) authorized ODs in 2009. Highest costs were for oncology drugs (57%–67% of total BI in 2005–2009) and for metabolism drugs (19%–30% in 2005–2009). Outpatient drugs accounted for 0.01% (2005) to 1.55% (2009) of total BI of ODs. In France, 34 ODs were reimbursed in 2008 and the BI increased from c71M in 2002 to c496M in 2008. In 2008, the BI represented 1.8% of the total value of drugs sold. The situation in Belgium is comparable as the total BI of ODs was c66.2M in 2008 representing 2% of total reimbursed drug costs. CONCLUSIONS: Although cost per patient is relatively high, total BI for a country is still modest as a result of lower volumes used. The results show a consistent picture for OD expenditure across different health care systems. Budget restrictions are not widely used for ODs but this might change when the anticipated increase in BI of ODs becomes more apparent. PHP55 INVESTIGATING THE IMPACT OF R&D INVESTMENT AND POLICY ON INNOVATIVE PERFORMANCE IN EUROPE Trevor NC, Tang M, Samuels ER Heron Evidence Development Ltd, London, UK OBJECTIVES: At the 2000 Lisbon Summit, the European Council set the quantitative target to increase R&D investment in all EU countries to 3% GDP by 2010. Today, there is growing emphasis on innovation in R&D, particularly in the development of pharmaceuticals, with the European Innovation Scoreboard (EIS) publishing annual ranking of the 27 EU member states to track and benchmark innovation performance. This work aims to explore the potential relationship between R&D investment and EIS innovation performance. In addition, the impact of national or regional innovation plans/policy will be considered. METHODS: The EIS innovation performance results for 2009 (based on data from 2005) were identified as the primary measure of innovation performance. The EIS includes 7 dimensions to accommodate the diversity of innovation processes and models that occur in varying national contexts. Eurostat data were used to identify the R&D investment in 2005. In order to identify the importance of R&D investment (%GDP) to EIS position, a simple quantitative linear regression was conducted. Supplementary qualitative literature searches were conducted to identify national and regional innovation plans and policies. RESULTS: The simple linear regression revealed a significant (P < 0.00005) relationship between R&D investment and position within the EIS innovation performance, with R&D investment explaining 72% of the scoreboard results. However, other dimensions have a noteworthy effect on innovation performance, since although the UK was identified as an innovation leader (along with Denmark, Finland, Germany, and Sweden), the UK ranks only 8th in terms of %GDP R&D investment. In this case, the presence of a national plan for innovation, alongside other factors, has led to high innovative performance. CONCLUSIONS: Although R&D has a significant effect on innovation performance, other dimensions also have a noteworthy effect. For example, the presence of national plans for innovation may aid a country in gaining innovation leader status. PHP56 SERVICES FOR WHICH PHARMACISTS MAY LEVY A FEE: PHARMACIST INITIATED THERAPY (PIT) Truter I1, Lubbe MS2, Butler N3, Nazer DP4, Gous AG5, Bayever DN6, Naidoo P7, Naidoo M8, Tlala V9, Putter S10 1 Nelson Mandela Metropolitan University (NMMU), Port Elizabeth, Eastern Cape, South Africa; 2North-West University, Potchefstroom, South Africa; 3University of the Western Cape, Bellville, Western Cape, South Africa; 4Tswane University of Technology, Pretoria, Gauteng, South Africa; 5University of Limpopo, Medunsa, Limpopo, South Africa; 6University of the Witwatersrand, Johannesburg, Gauteng, South Africa; 7University of KwaZulu-Natal, Durban, KwaZulu-Natal, South Africa; 8Rhodes University, Grahamstown, Eastern Cape, South Africa; 9South African Pharmacy Council, Pretoria, Gauteng, South Africa; 10 Management Sciences for Health, Pretoria, Gauteng, South Africa OBJECTIVES: The primary aim was to determine the extent of provision of Pharmacist Initiated Therapy (PIT) services in pharmacies in South Africa and the time it takes to provide this service. METHODS: A national research project was undertaken during 2008 by the South African Pharmacy Council on the services for which a pharmacist may levy a fee. The focus of this study is on one component of the larger study, namely the PIT service. RESULTS: A total of 369 pharmacies provided PIT services, and 3133 PIT services (cases) were measured. The majority were delivered by community (retail) pharmacies (95.79%). The PIT service was divided into three phases: Phase I (pre-administration procedure), Phase II (preparation and labelling of the prescribed medicine) and Phase III (provision of information and instructions to the patient to ensure the safe and effective use of medicine). Phase I was performed in 98.21% of cases, Phase II in 97.19% of cases and Phase III in 91.67% of cases. Pharmacists mostly delivered all three phases themselves (over 70% of cases). The weighted average time it took for a PIT service to be delivered was 199.02 seconds (just under 3.5 minutes) (SEM = 5.57 seconds). The weighted average time in community pharmacies was slightly less (192.82 seconds) compared to 312.15 seconds in private institutional pharmacies. The time taken was dependent on the number of items dispensed. The weighted average time taken was 160.76 seconds if there was 1 item dispensed, 220.31 seconds for more than 1 and equal to 2 items dispensed, and 327.19 seconds if more than 2 items were dispensed. a pharmacist may currently 13th Euro Abstracts charge a fee of R35.00 for this service. CONCLUSIONS: PIT is an important service that pharmacists deliver where the need exists. It is recommended that pharmacists be encouraged to counsel patients thoroughly when delivering a PIT service. PHP57 AN ANALYSIS OF DRUG COST CONTAINMENT POLICY AT A HOSPITAL IN SOUTHERN THAILAND Sae Wong AK, Kunthavaporn S, Junchareon N Songkhla Hospital, Muang, Songkhla, Thailand OBJECTIVES: To examine drug cost containment policy implemented at a hospital in southern Thailand. METHODS: This study was a retrospective, pre-post policy intervention descriptive design. During the fiscal years of 2005 and 2009, various drug cost containment strategies, including generic substitution for any drug group and a successful treatment guideline for orthopedic drugs, were adopted at a hospital in southern Thailand. Drug expenditures across those fiscal years were examined. The expenditure proportions between drugs listed and unlisted in National Essential Drug List were calculated. Cost-saving analysis of all generic substitution was conducted. Since the treatment guideline for orthopedic drugs was available in the hospital, their expenditures were also examined. RESULTS: Total drug expenditures had increased with decreasing rate across the study years. It increased by 47.15% from year 2005 to 2006, 43.19% from year 2006 to 2007, 21.17% from year 2007 to 2008 and 2.17% from year 2008 to 2009. The expenditures of essential drugs in the National Drug List were accounted for 61.64%, 56.62%, 54.38%, 48.67% and 50.94% across those study periods, respectively. Results showed that generic drug substitution policy reduced overall drug expenditures by 34.33%, or 7.66 million baths from year 2008. In 2009, only 11 items of generic drug substitution for branded-name drugs could reduce drug expenditures by 13.33%, or 4.73 million bahts which reflected annual cost-saving about 25.95 million bahts. In the same year, a result showed that the implementation of orthopedic drug guideline reduced drug expenditures by 5.53% or 2.10 million bahts. CONCLUSIONS: The study indicated that treatment guideline and generic drug substitution policies could control relatively high amount of drug expenditures at a hospital in southern Thailand. Hospital administrators should consider to continue these policies. PHP58 ANNUAL HEALTH INSURANCE REIMBURSEMENT OF DENTAL CARE IN HUNGARY Marada G1, Nagy Á1, Sebestyén A2, Benke B1, Kriszbacher I1, Boncz I1 1 University of Pécs, Pécs, Hungary; 2National Health Insurance Fund Administration, Pécs, Hungary OBJECTIVES: The aim of this study was to asess the annual health insurance reimbursement of of dental health service in Hungary. METHODS: The assessment base of the study was the annual reports of National Health Insurance Fund Administration (OEP). Only the data collected from the services in conctractual relationship with the OEP and delivered in 2008 were evaluated. Dental care services are organised in different levels: general dental service, specialist dental care, special dental care on university level and inpatient departments. Our study covers primary, outpatient and hospital dental care. RESULTS: Dental care was supplied by 3378 general and specialist dental care services until the end of 2008. For the hospital treatment of more serious cases 17 inpatient department is available with 154 patient beds. Within the period of examination (2008) 7.6 million cases or rather 23.6 million interventions were carried out. The health insurance expenditures of the OEP for outpatient dental care was 23.9 billion forints (c85,18 million). The total health insurance reimbursement of dental care (including primary, outpatient and hospital care) was 24.92 billion Hungarian forints (c88.82 million) in 2008. CONCLUSIONS: The health insurance reimbursement of dental care services in Hungary is approximately 2% of the total health insurance expenditure of OEP. PHP59 EVOLUTION OF PUBLIC EXPENDITURE WITH PHARMACEUTICAL CARE IN BRAZIL DURING THE PERIOD 2005–2008 Aurea AP, Garcia LP, de Magalhaes LCG, de Almeida RF, dos Santos CF Institute for Applied Economic Research, Brasilia, DF, Brazil OBJECTIVES: There is a known concern of health researchers and public managers in Brazil with the population’s access to medicines. We quantified the public expenditure on medicines in Brazil, during the period of 2005 to 2008. METHODS: The expenditure on medicines comes from a data warehouse of the Ministry of Planning, Budget and Management that stores the information concerning any purchase made by the Brazilian Federal Government. We also computed the amounts transferred to official laboratories to produce medicines. Information on the states, Federal District and municipalities came from the Information System on Public Health Budget (SIOPS). RESULTS: In the period 2005 to 2008, the public spending with drugs rose from US$ 1.8 billion to US$ 2.0 billion in real terms, with an average annual growth equivalent to 3%. The average spending in this period was US$ 1.8 billion. Most of the spending on medicines is attributed to the Federal Government, with values exceeding US$ 835 million per year. Just under half of the expenditure is given to states and municipalities. Considering only the federal spending, the “strategic component” represents the largest share, with participation from 56%–64% in the period. This result is expected, since the Ministry of Health is responsible for funding all the medicines from the “strategic component” of pharmaceutical care which includes, among others, the antiretroviral drugs and blood products. The amounts of transfers to official laboratories ranged between 20–25% of drug costs. CONCLUSIONS: The expenditure for those pharmaceutical care programs A415 whose purchases are centralized at the federal Ministry of Health, didn’t show a significant increase in the period of 2005–2008. Rather, the evidence suggests relative stability of procurement of medicines from pharmaceutical care programs under the Federal Government’s responsibility in this period. PHP60 HEALTH INSURANCE SUBSIDY OF SPA TREATMENT IN HUNGARY 1 1 2 2 Domján P , Zsigmond E , Ágoston I , Boncz I 1 University of Pécs, Zalaegerszeg, Hungary; 2University of Pécs, Pécs, Hungary OBJECTIVES: The aim of our study is to calculate the average health insurance reimbursement of spa treatment according to counties in Hungary. METHODS: Data were derived from the Hungarian National Health Insurance Fund Administration (OEP) and covers the fiscal year of 2007. These data was analyzed in the light of different value of its average point. The Hungarian spa financing method is based on relative system, which is depending on the treatment’s price. We calculated the average health insurance subsidy per of spa treatment (HIS/STN) as an indicator of average health insurance reimbursement of a single spa treatment. RESULTS: In 2007 the number of spa treatment was 8,160,438 and the full treatment expenditure of subsidy was 4.34 billion HUF (US$25,632 million). The average value of (HIS/STN) was 540,000 HUF (US$3124). Two outlier points are the region of North Hungarian Plan (different from average HIS/ STN value was −6.76%) and region of North Transdanubia (different from average HIS/STN value was 12.47%). The other regions performed similar outcomes ratio between 1.33% and 2.95%. CONCLUSIONS: The main cause of the two outlier regions is the inadequat structure of Spa services. Less people visit to North Transdanubian Region, because the number of spa facilities isn’t significant, but these thermal baths are significant, which price is higher. The North Hungarian Plan attracts a lot of patients with lower price. The result if the price is lower, the subsidy will be lower because of the financing system is based on relative method. PHP61 MARKET ACCESS AGREEMENTS IN EUROPE: TYPOLOGY AND RATIONALE Toumi M1, Jaroslawski S2, Lamure M3 1 University Claude Bernard Lyon 1, Rhone Alpes, France; 2Creativ Ceutical, Paris, Ile de France, France; 3University Claude Bernard Lyon 2, Rhone Alpes, France OBJECTIVES: Achieving Market Access for new products has become complex for pharmaceutical companies. Faced with growing expenditure, health care authorities accept or propose various Market Access Agreements (MAA) (risk-sharing/performance-based/commercial schemes) but often with little experience and knowledge. We performed in-depth analysis of their design and we formulate recommendations to stakeholders. METHODS: MAA is a formalized compromise between payers and industry to achieve: Price and Reimbursement, HTA recommendation and Formulary listing. We reviewed published and grey literature from major health insurers in France, Italy, Germany and UK. We conceptualize MAA typology according to the nature of uncertainty perceived by stakeholders and their motivations. RESULTS: We identified above 30 MAAs and classified them as follows: 1) Value for money not questioned: a) Conditional Market Access Agreement: Evidence development agreement→Aim: address actual uncertainty; b) Health Outcome Boostering Agreement: Disease Management Initiative→Aim: improve competitive advantage; 2) Value for money questioned: a) Cost Containment Agreement: Basic commercial agreement→Aim: reduce/control drug bill; b) Health Outcomes Agreement: Value based agreement→Aim: link payment to performance. Motivations of public payers: Main: Buy health production; Other: Control expenditure; Improve ICER of expensive products; Prevent media coverage of negative decision; Provide patient access; Expand benefits basket. Motivations of the industry: Main: Achieve Market Access for a product at high price in all markets; Other: Mitigate development failure; Reassure share holders; Improve company publicity; Fulfil requirements of authorities. In UK the design of MAA was a direct consequence of formalized HTA, in Italy there was no apperent rationale. CONCLUSIONS: Commonly used nomenclature needs to be revisited. Applying our typology framework should allow health care payers and the industry to design and implement MAAs rationally and with transparence. MAAs in UK are a direct repercussion of a not favorable primary HTA. PHP62 VALUE OF CONGRESS ABSTRACTS OF COST-EFFECTIVENESS STUDIES FOR DECISION MAKERS Karray S1, Jaroslawski S1, Dzbek J2, Altin S3, Gerber A3, Toumi M4 1 Creativ Ceutical, Paris, Ile de France, France; 2Jagiellonian University, Kraków, Małopolskie, Poland; 3Institute for Quality and Efficiency in Health Care (IQWiG), Cologne, Germany; 4 University of Lyon, Lyon, France OBJECTIVES: ISPOR, iHEA, and HTAi regularly organize congresses in the field of health economics. Given the number of abstracts accepted each year it is crucial to assess their credibility and how results of cost-effectiveness analyses differ across meetings. METHODS: We collected all abstracts published 2007–2009 at ISPOR (International and Europe), HTAi and iHEA meetings. Abstracts on cost comparison, cost of treatment, cost benefit, cost consequences, cost-effectiveness, cost minimization and cost utility analyses were reviewed in depth according to a reading grid which allowed extraction of essential information that could enable evidence-based decision-making in health policy. This included e.g. availability of key methodological parameters, involvement of the industry in authorship and details of conclusions. RESULTS: We analyzed 5488 abstracts from ISPOR, 1410 from HTAi and 1969 from iHEA. Our preliminary A416 results show that cost-effectiveness studies constituted 15%, 12% and 7% of all abstracts presented at ISPOR, iHEA and HTAi respectively. Non-drug technologies ranged from 11% at ISPOR to nearly 30% for HTAi and were excluded. 32% of analyses used best standard of care as comparator and 10% did not specify the comparator. Approx. Twenty percent of abstracts did not report discount rates, 28% the nature of costs included in analysis and 10% the time horizon. a total of 52% of analyses reported results as a point estimate of cost per QALY. 15% of abstracts submitted to ISPOR were not co-authored by the industry, 50% at HTAi and above that at iHEA. Analyses which judged the assessed drug to be cost-effective, cost-saving or dominant made up 82%, 70% and below 50% at ISPOR, HTAi and iHEA respectively. CONCLUSIONS: ISPOR is a congress preferred by the industry and a high proportion of abstracts reported favourable conclusions. This trend diminished for HTAi and further for iHEA. The quality of abstracts is not satisfactory for informed decision-making. PHP63 THE ECONOMIC IMPACT OF THE INITIATION OF PRESCRIPTION CONTROL IN THE GREEK SOCIAL SECURITY FUNDS Kousoulou F1, Argyri S1, Karapanos N2 1 Ministry of Employment and Social Security, Athens, Greece, 2Ministry of Health and Social Solidarity, Athens, Greece OBJECTIVES: Due to the financial crisis Greece was forced to implement hard cost containment measures almost in all fiscal sectors. The objective of the study is to investigate the economic impact emerging from the initiation of controls in prescriptions, implemented in the Greek social security funds as of 1st January 2010 to 30th April 2010. METHODS: The data derive from the drug reimbursement database of the three biggest social security funds of Greece from January to April 2010 comparing with the same period of the previous year. The three security funds of the analysis cover about 90% of the greek population with almost 10 million fully insured members. The security funds in scope were the following: IKA which covers the private sector with 6.3 million insured members; OPAD covering the public sector with 1.5 million insured members; OGA for agriculture with 2 million insured patients. RESULTS: In the first four-month period of 2010 form the initiation of the prescription control system the pharmaceutical expenditure was the following: for IKA c747 million in comparison to 716 million the same period in 2009, difference of 4.25%, for OPAD 172 million for 2010 whilst in 2009 the expenditure was 203 million, with savings of 15% and for OGA in 2010 was 310 million and the same period in 2009 the amount reimbursed for medicines was c288 million with 7.64% growth. It should be highlighted that although for IKA and OGA the pharmaceutical expenditure is higher in 2010 in comparison to 2009, still the growth of expenditure follows a downward slope, 2008–2009 14.82% for IKA and 11.64% for OGA respectively. CONCLUSIONS: The new cost containment measures implemented in the greek health care sector started presenting results. Other cost containment implemented measures were price cuts for all medicinal products in May 2010 and reduced supply prices for sanitary products. PHP64 VALUE BASED PRICING IN THE UK: A PRICE-QUANTITY MODEL ASSESSMENT Pisa P University of Edinburgh, Edinburgh, Midlothian, UK OBJECTIVES: In the wake of the 2007 Office of Fair Trading (OFT) Pharmaceutical Price and Regulation Scheme (PPRS) market study there is a debate whether the UK should switch to a value-based pricing (VBP) scheme. The OFT VBP system has its aim to price pharmaceuticals in line with their clinical effectiveness. METHODS: The switch from the traditional PPRS system to a VBP scheme in the UK was investigated with regards to the two main PPRS objectives: cost containment and value for money. This was carried out by modifying and applying a price quantity setting model (Das, 1980) to fit the UK pharmaceutical market and investigate the capital labour ratio of a firm. The model uses the assumptions that Von Neumann-Morgenstern 4 utility axioms are satisfied, the PPRS profit cap is binding and that the price set in the VBP scheme is efficient. RESULTS: This paper finds that a traditional PPRS, compared to a VBP system, might cause overinvestment in capital in proportion to labour resulting in higher than necessary production cost, also called the Averch-Johnson effect. The incentive to overinvest in capital is found to lie in the link between capital and the profit cap. We show that manipulating the profit cap by adding non-producing capital, i.e. rate-basepadding, is only incentivised when marginal productivity of capital is smaller than the cost of capital, net the capital allowance. The VBP effect on cost-containment is found to be ambiguous; if current pharmaceuticals are priced below what is warranted by their clinical effectiveness, cost may actually increase. CONCLUSIONS: This paper finds that, based on a price-quantity model assessment, a VBP system should be introduced. It would increase social welfare by pricing pharmaceuticals equal to the benefit they provide, and thus allocate NHS resources to their best possible use. PHP65 VALUE BASED PRICING IN THE UK: A SURVEY-BASED APPROACH Pisa P University of Edinburgh, Edinburgh, Midlothian, UK OBJECTIVES: A survey was carried out to investigate the economical impact of a switch from the traditional Pharmaceutical Price Regulation Scheme (PPRS) system to a value-based pricing scheme (VBP), as proposed by the Office of Fair Trading (OFT). The OFT VBP system has its aim to price pharmaceuticals in line with their clinical effectiveness. METHODS: Interviews were carried out with experts from the industry, 13th Euro Abstracts academia and the government comparing the traditional PPRS with the proposed VBP system. The interviews focused on the regulatory effectiveness, competition, launch delays, pharmaceutical pricing, risk-sharing agreements and uncertainty premium of the two systems. a systematic literature review was also carried out for all the above mentioned topics. RESULTS: In the interviews the current PPRS system was seen as very beneficial with high transparency and stability, nevertheless lacking mechanisms promoting price competition when compared to the VBP system. The main concern with a switch to a VBP system was the risk of a global price lock-in. Since the UK is directly or indirectly influencing pricing decisions within about 25% of global pharmaceutical consumption, the industry might delay drug launch in the UK, to maintain global pricing flexibility (e.g. in the adjuvant setting). Risk-sharing agreements were found to be one possible solution to maintain global price flexibility for the industry, while ensuring the NHS pays a fair price. The interviewed were unanimous about establishing an organization separate from any political influences needs to handle the pricing decision to avoid conflicting incentives. It was suggested to offer a price premium for pharmaceuticals with well documented cost-effectiveness. a premium would incentivise the industry and reduce reimbursement decision uncertainty. CONCLUSIONS: This survey indicates that a transparent and stable pricing process with proper risk-sharing agreements would increase the probability of a successful implementation of a VBP system in the UK. PHP66 AN ASSESSMENT OF THE VARIATION IN ACCEPTED ICERS BY DISEASE TYPE: RESULTS FROM FOUR HTAS Ternouth AM, Chapman M, Modha R Heron Evidence Development Ltd, Luton, UK OBJECTIVES: HTAs are frequently required to assess different treatment regimes for different disease types. Frequently, technology appraisal decisions are based on the ICERs estimated from economic models. The aim of this study was to identify any trends in accepted ICER thresholds by disease type. METHODS: All published technology appraisals since April 2005 were downloaded from PBAC, SMC, CADTH, and NICE websites. Appraisals were categorised by disease type according to BNF categories. The manufacturer’s base-case ICERs were extracted and compared across accepted submissions by disease type. RESULTS: Eighteen CADTH, 122 SMC, 81 PBAC and 122 NICE appraisals were identified. For PBAC and SMC, the accepted ICER level (<$75,000 and <£30,000, respectively) was consistent across >90% of the disease categories. However, accepted ICERs for malignant disease cluster at a higher level (up to $200,000 for PBAC and up to £63,000 for the SMC). Additionally, age related macular degeneration for PBAC, and severe osteoporosis, Lennox-Gastaut syndrome, plaque psoriasis, and hyperammonaemia for the SMC have treatments with accepted ICERs outside the typical range. Data is limited for CADTH, although Crohn’s disease and hepatitis B are exceptional in having treatments with accepted ICERs >$80,000. Complex data from NICE is qualitatively assessed in light of data from the other HTAs. Whilst most accepted ICERs were below £30,000, selected submissions for malignant disease were accepted above this commonly assumed threshold. CONCLUSIONS: Across disease types, accepted ICERs tend to cluster beneath a common threshold. However, submissions for treatments of malignant disease and immunosuppression have a greater chance of acceptance at higher ICERs than submissions for other disease categories. Rare diseases may also have a higher limit for ICER acceptance. PHP67 HAS THE QUALITY AND OUTCOMES FRAMEWORK INFLUENCED PRIMARY CARE DATA RECORDING? Blak BT, Lee J, Hards M, Sedani T CSD EPIC, London, UK OBJECTIVES: The Quality and Outcomes Framework (QOF) was introduced in the UK in April 2004. The scheme financially rewards practices for providing quality care and this is evaluated based on electronic medical records. This study therefore evaluated whether data recording changed after QOF was introduced. METHODS: Patients were selected from The Health Improvement Network (THIN) database, which holds longitudinal anonymised primary care records from >450 UK practices. Patients were grouped according to whether they ever had ≥1 of 15 chronic QOF diseases. Percentages of patients with ≥1 general practice (GP) visit, smoking status, blood pressure (BP) and weight record were estimated throughout nine 12-month time periods (January 4, 2000-January 4, 2009). T-tests compared mean percentages before and after QOF introduction (January 4, 2004). RESULTS: Percentage of QOF patients ranged from 26.6% to 32.9% over time and non-QOF patients from 67.1% to 73.4%. The average percentage of QOF patients with a GP visit was 80.5% (standard deviation (SD):3.2) before QOF and 84.5% (SD:0.9) after QOF (p = 0.086). These percentages were 57.5% (SD:3.5) and 62.0% (SD:0.6) (p = 0.082) for non-QOF patients. The average percentages for smoking recording were 26.8% (SD:12.9) versus 55.9% (SD:3.0) (p = 0.018) for QOF patients and 10.9% (SD:4.8%) versus 22.3% (SD:2.9) (p = 0.010) for non-QOF patients. For BP recording, 53.6% (SD:6.7) versus 68.1% (SD:4.8) (p = 0.013) for QOF patients and 20.5% (SD:2.9) versus 24.2% (SD:0.9) (p = 0.084) for non-QOF patients. For weight recording, 25.5% (SD:5.4) versus 40.4% (SD:3.2) (p = 0.006) for QOF patients and 9.8% (SD:1.7) versus 14.8% (SD:1.7) (p = 0.004) for non-QOF patients. CONCLUSIONS: Overall, the proportion of GP visits and clinical recording increased after QOF was introduced, although there was no evidence of a difference for GP visits or BP in non-QOF patients. This suggests that QOF influenced recording, especially the recording of the evaluated clinical measures for patients with chronic QOF diseases. 13th Euro Abstracts PHP68 THE USE OF HEALTH ECONOMICS TERMINOLOGY IN CLINICAL PUBLICATIONS: BRIDGING THE GAP FROM CLINICAL EFFECTIVENESS TO COMPARATIVE EFFECTIVENESS Reich S PAREXEL, Hackensack, NJ, USA OBJECTIVES: As the demands of the current regulatory climate and US Health Care Reform call for greater outcomes-based evidence in health care, the scientific literature is increasingly incorporating pharmacoeconomic data into peer-reviewed clinical publications in order to demonstrate value to payer audiences. Data from clinical outcomes and cost studies are increasingly finding their way into traditional clinical papers and review articles. It has been observed that variations may exist in the use of health outcomes data and terminology. The purpose of this presentation will be to demonstrate how clearer and more consistent terminology can be integrated into scientific publications and other vehicles in order to more effectively communicate economic and clinical outcomes information. METHODS: Using specific examples from clinical publications, the presentation will: Identify and analyze common terms used for communicating health economic and outcomes research information to determine; If they have multiple and/or unclear meanings; How they are being used to convey information; Define the specific meaning of these terms, using language that is understandable to all stakeholder audiences; Provide examples/case studies demonstrating how more uniform and consistent definitions can be integrated into payerfocused clinical publications RESULTS: An overview of the findings regarding how to address specific areas of confusion and inconsistency will be provided. CONCLUSIONS: The results will reiterate the need for clear and consistent terminology in communicating value in clinical publications. PHP69 MOBILE MEDICAL RECORD—A LIFE SAVING TOOL Friedman N, Goldberg A Ben Gurion University of the Negev, Beer-Sheva, Israel BACKGROUND: An emergency services team is launched to treat a man who collapsed in the street. The team takes John Doe’s mobile phone, and within seconds retrieves the required clinical parameters from his Mobile Medical Record (MMR) thus providing a life-saving treatment suited to his personal health condition. Have the necessary clinical parameters, required at emergency situations, ever been examined in order to best match both emergency situations and cellular technology? OBJECTIVES: Characterization of the clinical parameters which assemble an MMR in the context of saving life and propose a model for an MMR in emergency medicine. METHODS: Characterization of the essential emergency medicine clinical parameters in the context of life-saving treatments, through interviews with prehospital and hospital experts in emergency medicine. Supported by a Cellular multimedia expert, analysis of the results in order to incorporate the clinical parameters into the cellular world as an MMR. RESULTS: Emergency medicine teams chose individual and specific clinical parameters in a certain order of appearance from the general medical record which should assemble, in their opinion, emergency medicine MMR. MMR was chosen by the emergency medicine treatment teams as one of their preferred communication method among the possible communication methods presenting a patient medical record in the context of life-saving treatment. CONCLUSIONS: The MMR model, if applied correctly, will provide the emergency medicine treatment teams an available, reliable, homogeneous database of real time clinical parameters adapted to life-saving conditions. The MMR model represents a conceptual revolution of taking out the medical record from the caregiver and transferring it to the patient, which can be constantly at hand at any given time or place. By doing so, the MMR contributes and becomes integrated with the leading approaches in the world of medicine supporting a patient-centered care policy. PHP70 CURRENT TRENDS IN HEALTH TECHNOLOGY INCORPORATION (HTI) IN BRAZIL: INSIDE THE NATIONAL HEALTH SYSTEM’S (NHS) BLACK BOX Teich V1, Costa ALMA1, Cunha TMB1, Clark LGO2 1 MedInsight-Evidências, Rio de Janeiro, Brazil; 2MedInsight-Evidências, Campinas, Brazil OBJECTIVES: We aimed to evaluate the trends of submissions and recommendations received and emitted by the Brazilian Commission on Health Technology Incorporation (CITEC) during the process of adoption/exclusion of technologies on the country’s Public Health System and to critically appraise this process and to analyze the challenges this commission faces in responding to the increased demand for incorporation of new technologies. METHODS: All submissions on health technologies made to CITEC and the respective responses obtained between 2003 and 2010 were reviewed. They were classified by field of interest, origin of request and type of response. RESULTS: During the selected period CITEC received 222 proposals for technology incorporation and produced 83 recommendations. We identified an increase in submissions after 2007. From 2003 to 2006, only 39 requests were submitted on total. After 2007, an average of 49 submissions per year was made. The main fields were endocrine and metabolic disorders (12%), Oncology (11%) and Rheumatology (11%), but the percentage of responses in these fields was low (34%, 8% and 28% respectively). More than 70% of the submissions were commissioned by the Industry (162), but only 27% of these were answered. However, almost all submissions sent directly by the Ministry of Health were answered (94%). Among recommendations from CITEC, 53% were favorable to the incorporation of the new A417 technology and 30% were contrary to it, 17% were cancelled submissions. Adopted technologies were mainly for hepatitis (14%), endocrine and metabolic disorders (11%) and genetic disorders (11%). CONCLUSIONS: Although there isn’t a clear definition of priorities for the incorporation of new technologies by the Brazilian Ministry of Health, it is possible to map trends in the recommendations issued by CITEC. It is of the utmost importance to achieve greater disclosure of the criteria for selection of technologies for incorporation by the Commission. PHP71 PREDICTING OUTCOMES AMONG PATIENTS WITH PROLONGED MECHANIC VENTILATION: A DISCRIMINATION MODEL BASED ON LONGITUDINAL HEALTH INSURANCE REGISTRATION AND CLAIMS DATA Lu HM1, Chen L1, Hung MC2, Wang JD2, Lin MS3, Yan YH3, Chen CR3, Fan PS3, Kuo KN1 1 National Health Research Institutes, Zhunan, Taiwan; 2National Taiwan University, Taipei, Taiwan; 3Chia-Yi Christian Hospital, Chiayi, Taiwan OBJECTIVES: To identify patients with low survival probability among those with prolonged mechanical ventilation (PMV; >=21 days of use) under the Taiwan National Health Insurance (NHI), and explore the length of their time free of inpatient mechanical ventilation (MV) and MV-related hospital expenses during the four years following PMV. METHODS: This is a retrospective cohort study using death certificates data and longitudinal NHI data for a national representative sample of 25,482 new PMV patients in 1998–2003. Logit regression was used to determine factors associated with 3-month, 6-month, 1-year and 2-year survival. Explanatory factors included hospital characteristics, and individual demographics, socioeconomic conditions, diseases at the PMV onset, and diseases causing hospital care use within the year before PMV. The probability cutoff was set at 10% for identifying patients with low survival likelihood and suitable as targets of advocating less intensive care. RESULTS: Five disease types had high prevalence, and were significantly associated with lower survival: septicemia, neoplasm, shock, acute and unspecific renal failure, and chronic renal failure. Each had an odds ratio <0.65 (p < 0.001). Non-alcohol-related liver disease was also a significant problem. Each survival model had a C-statistic >= 0.7. At the 10% probability cutoff, 1.5%, 5.1%, 11.9% and 21.8% of PMV patients were identified as negative cases for 3-month, 6-month, 1-year and 2-year survival. The negative predictive values were 86.5%, 86.1%, 89.4%, and 91.8%, respectively. During the four years after PMV, 75% of these patients had 25 or fewer days free of MV-related hospital care, and more than 50% of them spent at least 10,800 dollars (2010 USD) on MV-related hospital care. CONCLUSIONS: This discrimination model has acceptable performance. It helps in selecting potential targets for advocating less intensive care, and also generates more empirical evidence on prognosis that can facilitate patient-clinician and family-clinician communication. PHP72 ANALYSIS THE AVERAGE SOJOURN TIME AND BOTTLENECK AFTER A DISASTER IN HOSPITALS Sakata K1, Kobayashi D2, Matsui M1 1 The University of Electro-Communications, Chofu, Tokyo, Japan; 2The University of Electro-Communications, Minato-ku, Tokyo, Japan OBJECTIVES: The purpose of this study is to consider the average sojourn time of patients figure out hospital’s function after a disaster striking. METHODS: This study explores the average sojourn time of patients and bottleneck in hospitals in the region defined by this study. Especially, we focus on three days after a disaster striking the region. We defined three scenarios. • Scenario I:Hospital No.II ceases to function two days from disaster strikes. • ScenarioII:Hospital No.II ceases to function one day from disaster strikes. • ScenarioIII:ScenarioIII is the same scenario as the scenario I, but considering transport from break down hospital (hospital No.2) to hospitals in operation. Patient’s arrival interval λand treatment time µ are equal to the data of Hanshin earthquake. With the use of those scenarios, we analyze the average sojourn time of patients and bottleneck with a queuing approach, critical path method and little’s law. In addition to that, we make Cumulative flow graph to analyze bottleneck. RESULTS: Comparing scenario I and scenario III, the average sojourn time of general hospital of scenario III is longer than scenario I, although scenario III has more choices. It reveals considering transport from break down hospitals to hospitals in operation extends the average sojourn time. We observed bottleneck in case of changing patient’s arrival interval λ by making cumulative flow graph. This study shows cumulative flow graph reveals status of hospital’s function after a disaster. CONCLUSIONS: The results confirm that the average sojourn time of patients increase in case of changing hospital, Braess’s paradox and cumulative flow graph reveals status of hospital’s function after a disaster. PHP73 CONTENT OF HOME PHARMACIES IN SERBIA 1 2 2 2 Paut Kusturica M , Sabo A , Tomic Z , Stilinovic N 1 A.D. Sanitarija, Novi Sad, Serbia; 2Faculty of Medicine, Novi Sad, Serbia OBJECTIVES: To get insight into the amount, structure and financial aspects of medicines in home pharmacies as well as learning about the population’s attitudes, habits and knowledge concerning self-medication. METHODS: This survey was conducted in the first half of 2010. on a sample of 50 families on the territory of Novi Sad, Serbia. The families were chosen randomly and the data were collected by using the standardized questionnaire and checking the inventory of medicines. RESULTS: The average number of packets in home pharmacies per a family is considerable (13 A418 packets). The most widely kept medicines are medicines for the central nervous system (21%), the disorders in the digestive system and metabolism (16.7%) and medicines for the diseases of the respiratory system (14.5%). Nearly three-fourths of opened packets were not entirely consumed. a large number of medicines in home pharmacies (79%) was obtained self initiatedly and many of them were beyond their expiry date (9.5%). a total of 25% of the financial means for those medicines was provided from health insurance funds. The total cost of medicines in home pharmacies per a family in Serbia was c25.13. a total of 72% of that amount was paid by the patients themselves. Pharmacoeconomic indicators in certain groups of medicines show that the largest percentage of the financial means for home pharmacies in Serbia is spent on the medicines for gastrointestinal problems and metabolism disorders (17%). CONCLUSIONS: Based on the analyzed data, self-medication, that is, unreasonable use of medicines, is a big problem in Serbia. The use of these medicines involves the risk of inadequate treatment of the illness, frequent occurrence of side-effects which sometimes requires additional treatment, which only increases the expenses for both the patients and health funds. Therefore, it is necessary to closely monitor the issuing of prescriptions as well as educate the population. PHP74 PRIMARY HEALTH CARE AND POTENTIALLY AVOIDABLE ADMISSIONS: A 10-YEAR ASSESSMENT FROM THE PERSPECTIVE OF A HEALTH CARE PLAN Reis Neto J, Tovar C CAPESESP/CAPESAUDE, Rio de Janeiro, RJ, Brazil OBJECTIVES: To examine the evolution of the rate of admissions for ambulatory care sensitive conditions in beneficiaries of a health care plan, monitoring of access to the health system and its performance and identifying strategies that will enable the improvement of the health situation of the population and the appropriate allocation of resources. METHODS: Retrospective analysis of 212,360 hospital admissions, occurring between 1999 and 2008, among beneficiaries of a health care plan, resident in Brazil. The outcome of the study was the proportion of admissions for primary care sensitive conditions, identified by ICD-10. The proportion of these admissions in relation to the total hospital admissions was established, as well as the quotient for the number of hospital admissions and the population, either total or by age group and specific causes, besides hospital direct costs. The significance was tested and estimates of the rates were defined based on their 95% confidence intervals. RESULTS: Admissions for primary care sensitive conditions accounted for 55,307 of 212,360 admissions reimbursed by the health plan (26.0%). The rate of admission for primary care sensitive conditions fell by 28.0% during the period (from 347.6 to 250.3 per 10 thousand). The most frequent cause was gastroenteritis (30.9%), followed by arterial hypertension (13.6%), cerebrovascular diseases (9.6%), heart failure (8.1%), kidney and urinary tract infections (6.8%), diabetes mellitus (5.8%) and angina pectoris (5.5%). The mean annual expenditure on admissions was R$15,232,494 against R$51,440,680 for other conditions (22.8%). CONCLUSIONS: The reduction observed in the rate of admissions for primary care sensitive conditions suggests possible improvements in primary health care. Further actions and strategies could reduce the number of avoidable admissions and contribute to the financial balance of the health care plan. PHP75 EVALUATION OF THE IMPACT OF INPATIENT CLINICAL PHARMACY SERVICES ON THE QUALITY AND COST OF PHARMACOTHERAPY IN INTERNAL MEDICINE WARDS Pitakthum S Songkhla Hospital, Songkhla, Thailand OBJECTIVES: The purpose of this longitudinal descriptive study was to evaluate the influence of interventions by clinical pharmacists on processes, outcomes of care and economical outcomes. METHODS: The patients who were hospitalized at internal medicine wards, Songkhla hospital during October 1, 2008-September 30, 2009 were recruited to this study. The core set of clinical pharmacy services were admission drug histories, drug protocol management, adverse drug reaction management, drug information, medical rounds and patient discharge counseling. The clinical pharmacists collected patients’ data, pharmacists’ interventions, direct drug cost-saving and activities on a specific designated form. The data was processed on a computerized database. Clinical pharmacist interventions were counted and classified and analyzed to determine the influence on the quality and cost of pharmacotherapy. Descriptive statistics were used in data analysis. RESULTS: At the end of the study period the following were found: a total of 3796 patients were recruited. Clinical pharmacy services on internal medicine wards contribute to maximizing the pharmacotherapeutic effect and rationalization of drug therapy in 609 patients (16.0%), increase medication safety in 338 patients (8.9%) and reducing drug expenditures in 575 patients (15.1%). Costsaving from reconciling process was US$12,682 and cost-saving from pharmacists’ interventions were US$6,298. The frequent type of interventions was maximizing the pharmacotherapeutic effect (40.8%), rationalization of drug therapy (30.3%) and minimizing the risk for developing adverse drug effects (28.9%), respectively. The top two of drug related problems found in cardiovascular drugs (30.5%) and antibiotic drugs (24.8%), respectively. CONCLUSIONS: The results of this study indicate that inpatient clinical pharmacy services on internal medicine wards contribute to rationalization of drug therapy, increase medication safety and reduced the direct expenditures on medications. Therefore, future studies should include multiple sites such as in a general surgical department and clinical pharmacy services should be continued. 13th Euro Abstracts PHP76 ALLOCATION OF A SINGLE BUDGET FOR HEALTH CARE, PROFESSIONAL EDUCATION, AND RESEARCH Gandjour A Pennington Biomedical Research Center/Louisiana State University, Baton Rouge, LA, USA OBJECTIVES: Funding for health care programs competes with funding for professional education (ie, implementation of health care programs) and funding for research to reduce uncertainty around program cost-effectiveness. No study has yet shown how to allocate a combined budget for health care, professional education, and research. Previous work did not consider the fundamental idea that interventions with an acceptable incremental cost-effectiveness ratio (ICER) for an adoption decision may not be acceptable any more once implementation decisions are made simultaneously. Furthermore, previous work did not consider that cost-effectiveness of research is conditional upon the programs chosen for adoption and implementation. The purpose of this work is to present a sequential process to allocate a combined budget with the goal to maximize health. METHODS: Development of a 3-step allocation process to maximize health based on a single budget for health care, professional education, and research RESULTS: As a first step, a league table approach determines which programs would be funded if the combined budget were spent just on adoption. The second step analytically determines a new (lower) threshold ICER which maximizes health by discontinuing programs above the new threshold and investing in implementing programs below the threshold. The third step considers cost-effectiveness of research to reduce uncertainty around program cost-effectiveness. In order to fund this research, the threshold ICER for adoption plus implementation has to be reduced further. In order to maximize health, ICERs of research studies need to equal the newly defined threshold ICER for adoption plus implementation. CONCLUSIONS: A 3-step allocation process is able to maximize health based on a single budget for health care, professional education, and research. Compared to a programming approach the proposed approach yields a closed-form analytical solution and is able to consider that research decisions are conditional upon adoption and implementation choices. PHP77 DIRECT CONTRACTS ON PATENT-PROTECTED PHARMACEUTICALS IN GERMANY: MOTIVES OF SICK FUNDS AND PHARMACEUTICAL COMPANIES Neurohr C, Huebner S, Ernst F, Welte R GlaxoSmithKline GmbH & Co. KG, Munich, Germany OBJECTIVES: To investigate the motives of both contracting parties, sick funds and pharmaceutical companies, to enter into direct contracts on patent-protected pharmaceuticals in the outpatient sector in Germany. METHODS: A systematic literature review was performed in March 2010 to identify contracts on patent-protected pharmaceuticals in Germany. Standard literature databases such as Pubmed and EMBASE, different market research databases and the ‘Lauer-Taxe’, the official German price registry for pharmaceuticals were utilised. We included only officially reported contracts. a detailed market analysis was conducted to identify the motives of the identified contracts. RESULTS: The number of direct contracts has been recently increasing. a total of 116 contracts were found. Formally, they are all rebate contracts due to German law. However, they include risk-sharing (n = 6), cost-sharing (2), added value (5) and simple rebate contracts (103). For sick funds, major motives were cost cutting, positioning within the market, attracting the “right” members, improving the medical treatment and right coding. Pharmaceutical companies show several motives that can be clustered into proactive and reactive motives: Proactive motives were: gaining access to reimbursement beyond standard coverage, funding of managed care concepts, bridging a time gap until market entry of a successor, maintaining the attractiveness after patent expiration as well as shaping the market access environment. Reactive motives were: rebate contracts of competitors, limitations and exclusions from reimbursement by the federal joint committee as well as reacting to attempts by sick funds and Associations of Statutory Health Insurance Physicians to control drug expenditures. CONCLUSIONS: Direct contracting on patent-protected pharmaceuticals seems to play a increasing role in the strategic considerations of pharmaceutical companies and sick funds. PHP78 A TALE OF FOUR COUNTRIES: COMPARING REIMBURSEMENT SUBMISSION REQUIREMENTS IN IRELAND, ENGLAND, WALES AND SCOTLAND Rycroft C1, Pearson I1, Khan S2, Heyes A1 1 RTI Health Solutions, Manchester, UK; 2RTI Health Solutions, Research Triangle Park, NC, USA OBJECTIVES: Requirements for a Health Technology Assessment (HTA) submission vary within the UK (England and Wales [National Institute for Health and Clinical Excellence (NICE)], Scotland [Scottish Medicines Consortium (SMC)], and Wales [All Wales Medicines Strategy Group (AWMSG)]) and the Republic of Ireland (National Centre for Pharmacoeconomics [NCPE]). The objective of this study is to determine whether the likelihood of reimbursement in these markets is linked to HTA submission requirements. METHODS: Dossier requirements issued by NCPE, NICE, SMC and AWMSG were compared, and a checklist of requirements compiled. We investigated 20 interventions most recently reviewed by the NCPE, and recorded the recommendations for these compared with those issued by NICE, SMC and AWMSG. RESULTS: Economic analysis is key for an NCPE submission, although there are no specific requirements or template for the clinical data. Requirements for NICE are the most stringent; SMC and AWMSG have similar requirements. Of 20 interventions reviewed by NCPE, 11 (55%) were reimbursed 13th Euro Abstracts under the Community Drugs Scheme. Of these, 2 treatments were reimbursed after a price reduction, 2 were borderline, and 1 treatment was cost-effective under the General Medical Services scheme but not the Drugs Payment Scheme. Of the 11 products reimbursed by NCPE, 5 were recommended by NICE (45%), 6 by SMC (55%) and 1 by AWMSG (9%). Not all of the 11 treatments were reviewed by NICE, SMC and/or AWMSG. Furthermore, 2 of the 11 treatments have been withdrawn in Scotland, England and Wales. Limited clinical/economic data was the reason for the one non-recommendation by the SMC. a more detailed analysis will be presented. CONCLUSIONS: Although requirements for HTA submissions vary between countries, demonstrating cost-effectiveness is a key factor for reimbursement in all markets investigated. PHP79 REFINING INPATIENT PSYCHIATRIC FACILITY REIMBURSEMENT: LESSONS FROM OTHER MEDICARE PROSPECTIVE PAYMENT SYSTEMS Pierce CA, Jones NA, Brocketti JJ The Resource Group, Richfield, OH, USA OBJECTIVES: The Inpatient Psychiatric Facility Prospective Payment System (IPF PPS) is the youngest of the Medicare PPSs. Inpatient Psychiatric Facilities (IPFs) were excluded from the acute care inpatient diagnosis-related groups (DRG) system as it was thought to be a poor predictor of resource use for psychiatric patients. Although the IPF PPS was to include a patient classification system to reflect the differences in resource use and costs among IPFs, it was implemented without an associated assessment tool. This project reviews Medicare PPSs in which diagnosis is not an adequate predictor of resource use; compares data collected via the related assessment instruments, and discusses applicability to the IPF PPS. METHODS: The Medicare prospective payment systems were evaluated and the relevant systems isolated. The systems’ defining criteria were defined and the related assessment instruments explored for identifying measures. RESULTS: Under the current IPF PPS, Medicare pays for the per diem costs associated with furnishing covered inpatient psychiatric services via a nationally established base rate adjusted by geographic, facility and patient factors. Patient characteristics are limited to seventeen Medicare severity-diagnosis related groups (MS-DRGs), age, length of stay and certain co-morbidities but, unlike home health, skilled nursing and rehabilitation facility systems, do not account for functional status or other contributors to resource consumption. CONCLUSIONS: Patients in IPFs may vary in care needs despite common MS-DRG classification, thus require different resources. With limited ability to adjust beyond primary diagnosis, the current IPF PPS may misalign facility payment and expenses, ultimately impacting access to care. By analyzing other Medicare PPSs policy makers may positively contribute to the refinement of this newest system. A419 prescription is allowed in 20 countries of all 27 European countries and in all 50 US states. The generic substitution by pharmacists is allowed in 21 European countries (mandatory in 8) and in all US states (mandatory in 14). Despite the high increase from 2002, generics market share in Portugal is still below the European average, furthermore with a slowdown in its growth since 2005. The doctors prescribed 39.1% of drugs by brand name with no permission for substitution in 2009. The simulations generated savings of c120.4 million (50.8 for the NHS and c69.6 million for patients), less 11.1% of the medicines expenditure covered by the RPS. CONCLUSIONS: The market shares observed seem to reflect the different legislative frameworks. In Portugal, despite the growing share of generics, there is still a large potential market with significant opportunities of additional savings for the health system and patients, and may contribute to a higher pharmaceuticals accessibility as well to maximize therapeutic adherence. PHP82 ASSESSMENT OF PHARMACEUTICAL EXPENDITURE TRENDS IN PORTUGAL—PRICING AND REIMBURSEMENT POLICY Teixeira I, Mendes Z CEFAR—Center for Health Evaluation Studies, Lisboa, Portugal OBJECTIVES: The Stability and Growth Pact approved by the Portuguese Government in 2010 limits the annual growth of public expenditure in 1% for drugs reimbursed in outpatient sector. Considering the 10.4% increased in pharmaceutical expenditure from the National Health Service (NHS) in the first 4 months of 2010, this study aims to identify the main factors of expenditure growth and implications of legislative modifications in Portugal. METHODS: We have considered the policy changes concerning pricing and reimbursement and analyzed the database that includes sales from Portuguese community pharmacies and drugs prescription data. The statistical analysis of monthly data by product was performed with Statistical Analysis System (SAS), version 8.2. RESULTS: The new legislation approved in June 2009, that establishes the generics reimbursement at 100% for pensioners whose income is below the national minimum wage was responsible for more c20.4 million of NHS spending in 2010. The special reimbursement regimes had also contributed positively for this trend: for example, the reimbursement rate raise for some antiasthmatics (from 37% to 69%) was responsible for c5.1 million (10.3%) of NHS expenditure increase. Nevertheless c10.7 million (21.5% of NHS expenditure growth) are explained by the top 10 branded drugs recently reimbursed (5 oral antidiabetics and 1 insulin). The reform in reimbursement system approved for July 2010 also predicts major changes in the expenditures trends. CONCLUSIONS: The legal framework has a high impact in pharmaceuticals expenditures trends, and consequently in public financial burden and patient co-payments. The different policy measures adopted should be assessed on a periodic basis in order to monitor the market dynamics and to identify the strategies that support the sustainable expenditure growth. PHP80 EFFECTS OF REFERENCE PRICING AND TENDER CONTRACTS IN GERMANY Volmer T1, Fieke H2 1 TV Healthcare Consulting, Hamburg, Germany; 2University Muenster, Muenster, Germany OBJECTIVES: The therapeutic reference price (RP) system in Germany is regarded as a successful mechanism to reduce pharmaceutical spending. However, later introduced price directed measures such as the so-called payer contracts (pharmaceutical tender contracts (TC) interact with the reference price mechanism. The aim of this paper is to compare and analyze the effects of both instruments on pharmaceutical price, volume and expenditure for substances at the period around loss of exclusivity. METHODS: a group of frequently prescribed drugs with patent expiry between the years of 2001 and 2008 are investigated and the sales and unit data in the statutory health insurance system are analyzed for a period one year before and two year post patent expiry. RESULTS: The data show for the group with reference price and no tender contract (RP+/TC-) and with no reference price but tender contracts (RP-/TC+) similar level for the sales but significantly higher volume data for the first group. In the first year post patent expiry there seams to be a significant drop of sales and volume 9–12 month post patent expiry— a period, when usually the reference price is set. The figures for the groups with tender contracts (TC+) show a more steadily decline post patent expiry. The group RP+/TC+ is in terms of cost containment (sales decline) the most prominent one. If no regulation takes place, sales and volume increases even post patent expiry. CONCLUSIONS: Reference pricing and tender contracts are two cost containment measures, which are equally effective if applied alone, and which have additional impact, when applied together. PHP83 EFFECTS OF COST-CONTAINMENT METHODS TO DRUG PRICE LEVEL IN EUROPE Ligeti I, Komáromi T Healthware Ltd, Budapest, Hungary OBJECTIVES: Due to the dynamic increase of pharmaceutical expenditure in the last decade different cost-containment method has been introduced by the European public insurance companies. The cost-effective drug reimbursement system might influence the price level either in a direct or an indirect way, so the pricing strategy should be based on demand drivers and on factors of the regulatory environment as well. The goal of this analysis was to identify the effect the certain reimbursement solutions and local market characteristic on average price level of the predefined submarket. METHODS: Before the statistical analysis we have built a comprehensive, standardized and continuously updated international drug database based on official sources which supported the analysis trough getting wider access to data’s. Because of the expected domestic introduce of international reference price system, we handled it with an extra attention among the explanatory variables. RESULTS: The variance of international pharmaceutical price level by the examined submarket found to be remarkable which can be explained by the difference product supply and the different apply of cost-containment methods of price regulation. CONCLUSIONS: The most effective way to curb the increased expenditure by the patient and the public insurance companies seems to be international reference price system; however the effect of this method differs across the countries. PHP81 EFFECTS OF THE POTENCIAL GENERIC SUBSTITUTION IN PORTUGAL. DOES REGULATION MATTERS? Teixeira I1, Guerreiro J1, Mendes Z1, Montenegro C2 1 CEFAR—Center for Health Evaluation Studies, Lisboa, Portugal; 2Associação Nacional das Farmácias, Lisboa, Estremadura, Portugal OBJECTIVES: Considering the pharmaceutical expenditure trend and the current economic and social context in Portugal, this study aims to: 1) analyze the national and international policy environments in Europe and US, concerning INN prescription and generic substitution, and 2) calculate the effect of generic substitution in Portugal, in terms of savings for the NHS and patients. METHODS: We conducted an international legislation review and generics market data. We estimated annual savings under the assumption of gradual increase in generics market share, and considering the difference between branded and generic medicines prices for each group of the Reference Price System (RPS). The database includes sales from Portuguese community pharmacies. RESULTS: The INN PHP84 A PROPOSAL FOR CLASSIFICATION OF PUBLIC FUNDING RESTRICTIONS Wilk NM Arcana Institute, Krakow, Malopolskie, Poland OBJECTIVES: Increasingly difficult situation of health care systems forces decision makers to limit access to publicly funded drugs compared to registration conditions. To keep the transparency, the decisions to deny health intervention to some group of patients have to be publicly justified—simple intuition is not enough. The objective is to present initial classification and details of methods to generate public funding restrictions. METHODS: A pool of public funding decisions has been identified through search of internet websites of the institutions that recommend or actually make public funding decisions. The specific conditions restricting the access were identified, grouped and further analyzed together with their justification if available. RESULTS: All restrictions generally aim to optimize the performance of the decision A420 in regard to public funding criteria. The following main types of restrictions were identified either because of the target public funding parameter affected or because of the reasoning: a) “because of lack of evidence”; b) “because it is the only effective treatment” (rule of rescue); c) “to improve efficacy”; d) “to improve efficacy-safety relation”; e) “to improve effectiveness”; f) “to improve cost-effectiveness”; and g) “to limit budget impact”. CONCLUSIONS: The limiting conditions should be perceived as tools to enable positive public funding decision when the current scope of financing is just behind the hypothetical threshold. Exploring and further analyzing methods and aspects concerning generating public funding restrictions is important for: 1) decision makers, so they be more aware of the consequences and impact of their decisions on the people/patients they serve, and could make more transparent decisions; 2) HTA analysts, to focus their interest on the subsequent use of HTAs to help decision makers identify all potential options to rationally limit funding; 3) Market Access managers, so they used the identified mechanisms and methods to better foresee the public funding decisions concerning their drugs. PHP85 ASSESSING PUBLIC ACCOUNTABILITY OF KOREAN DRUG REIMBURSEMENT DECISION PROCESS Cho E, Kang M Yonsei University, Seoul, South Korea OBJECTIVES: Since economic evaluation has been enforced to be considered for new drug reimbursement decisions in 2007, the structure and constituents of decision body, “Drug Reimbursement Evaluation Committee”, as well as the pharmacoeconomics report guideline promulgated by Health Insurance Review Assessment have been modified in Korea. These changes reflect deficiency in systematic adoption of economic evaluation and discontent among stakeholders. Recently, the fair and reasonable process and criteria have been highly emphasized at every level of policy administration in developed countries with concerns of increasing public demands for sustainable public practices and political acceptance of the importance of public accountability. Now, imminent practical task is how to connect the conceptual framework and feasible practice for publicly accountable drug reimbursement decisions. METHODS: Using theoretical structure of public accountability published in the European Governance Paper, we analyzed qualifications of drug reimbursement policy as ‘accountability’ and appraised the public accountability. We also performed interviews with ten key stakeholders from democratic, constitutional and learning perspectives. And then, recent reimbursement decision papers on two new drugs were analyzed to examine concrete shape of the accountability forums. RESULTS: Following scope of improvement would be suggested: (1) Clear and reasonable standards for coverage decision; (2) relevance of the standards to population’s health needs and health equity impacts; (3) disclosure of data used for decision, procedure and results to public; and (4) the procurement of due process of challenging decisions. CONCLUSIONS: Given value pluralism in democratic liberalism, it is matter of course that formal or procedural justice is given prominence. Conclusively, an explicit discussion for formal criteria and procedure is the essential component of the ongoing policy process. Accountable drug policy administration is impossible without accountable policy process which is impossible without transparent criteria for all decision stages. PHP86 SUCCESSFUL IMPLEMENTATION OF COVERAGE WITH EVIDENCE DEVELOPMENT SCHEMES: PRACTICAL EXPERIENCES IN SEVERAL WESTERN JURISDICTIONS Theunissen KA1, Delwel G2, Evers SM1, Goettsch W2, Severens H3, Hoomans T1 1 Maastricht University, Maastricht, Limburg, The Netherlands; 2Health Care Insurance Board (CVZ), Diemen, The Netherlands; 3University Maastricht & Erasmus University Rotterdam, Maastricht, The Netherlands BACKGROUND: In Western jurisdictions, coverage with evidence development (CED) is seen as a promising scheme for linking coverage decisions on innovative health technologies with the generation of additional evidence. Nonetheless, the implementation of such CED schemes is not guaranteed to be successful, with issues arising around both evidence generation and (dis)continued coverage of technologies. OBJECTIVES: This qualitative study aims to explore the practical experiences with CED schemes for technology coverage. METHODS: Semi-structured interviews were conducted with key stakeholders involved in the decision making process about reimbursement of health technologies in Australia, Belgium, Ontario (Canada), France, Germany, Galicia (Spain), Sweden, UK and USA. The questionnaire for the interview was developed on the basis of comprehensive literature review and expert opinion, addressing issues related to the aim of CED, initial assessment/appraisal, evidence development, re-assessment/reappraisal and coverage for technologies. RESULTS: CED schemes vary widely in the evidence generation, time frame, the regulation and funding of research, and the involvement of stakeholders in the CED process. This variation is caused by the structure and financing of health services delivery, the aim of CED, and the type of health technology. The CED process is often non-transparent, particularly relating to the selection of technologies and the re-assessment/re-appraisal for coverage (dis)continuation. (Dis) continued coverage of technologies is often primarily driven by the pressures from patients, health professional and health institutions to provide access to innovative technologies or public opinion. CONCLUSIONS: The successful implementation of CED schemes requires: a) a transparent (analytic) framework for the selection of health technologies and the generation of additional evidence; b) a clear legal authority to regulate evidence generation, time frame and research budget; c) a structural involvement of stakeholders in the process of CED; and d) a priori clear end points for the re-assessment/ re-appraisal and technology coverage. 13th Euro Abstracts HEALTH CARE USE & POLICY STUDIES – Health Care Research & Education PHP87 THE ROLE OF VALUE OF INFORMATION ANALYSIS IN HEALTH CARE RESEARCH PRIORITY SETTING: A THEORETICAL CASE STUDY Corro Ramos I, Rutten-van Mölken MP, Al MJ Erasmus University, Rotterdam, The Netherlands BACKGROUND: The Dutch reimbursement procedure for expensive hospital drugs requires the submission of a baseline cost-effectiveness analysis together with a research plan for the period of temporary reimbursement in order to estimate the real-life cost-effectiveness after 4 years. In this situation, a Value-of-Information (VOI) analysis might identify the critical parameters that need to be studied in such outcome study. OBJECTIVES: To identify when a VOI analysis alongside sensitivity analyses is warranted, and when such VOI analysis will not impact the decision making process. METHODS: We used a hypothetical Markov model with three groups of parameters: costs, utilities and transition probabilities. We studied different configurations of input parameters, forcing the outcomes into different directions on the CE-plane. For each input configuration we performed a multivariate sensitivity analysis (MSA) and a probabilistic sensitivity analysis (PSA). In the MSA, sensitivity was measured as percentage change from baseline INMB. Additionally, we analyzed the expected value of perfect information (EVPI) and the expected value of partial perfect information (EVPPI). Analyses were done for a range of threshold ICERs. RESULTS: For each situation it was possible to predict the shape (but not the absolute value) of the EVPI curve based on the PSA findings. When the PSA plot covered both northern quadrants, MSA and EVPPI came to the same ranking of the groups of parameters. When the outcomes were in the northeast quadrant the ranking differed: MSA indicated costs as most important parameters, for EVPPI this was utilities. When outcomes where in the southwest quadrant, costs were most important in MSA and EVPPI. For both other quadrants, MSA and EVPPI were close to zero for all groups. CONCLUSIONS: Whether MSA and EVPPI come to a different priority setting for future research depends both on the threshold ICER and on the location of results on the CE-plane. PHP88 STAKEHOLDER PERCEPTIONS OF CLINICAL DRUG TRIALS 1 2 Murphy LM , Maguire W 1 University of Tasmania/Manukau Institute of Technology, Auckland, New Zealand; 2 University of Tasmania, Hobart, Australia OBJECTIVES: To identify stakeholder perceptions of sponsored clinical trials in a publically funded New Zealand hospital, and then to identify the similarities and differences in perceptions across these stakeholder groups. The stakeholders are: 1) participants involved in clinical trials; 2) management and the multidisciplinary team; 3) the larger South Auckland community; 4) government and decision makers; and 5) the pharmaceutical industry. METHODS: We use purposive sampling to select representatives of the stakeholder groups, which provides 109 respondents. We gather data using focus groups, in-depth interviews, telephone interviews and surveys. RESULTS: Many of the respondents represent more than one stakeholder group. While there is consensus across the stakeholders on some costs and benefits such as developing safe medicines and collecting useful data there are marked differences in perceptions in other areas, such as those indicated below. Most stakeholders perceive the risk of adverse reactions as the greatest cost to trial participants but the participants themselves do not regard this as significant. Pharmaceutical representatives, management and the multidisciplinary team feel that gaining access to new medicines motivates people to participate in a trial. Trial participants feel that the support is more important to them than the medication. Most researchers and staff believe trial involvement increases their job satisfaction, motivation, knowledge and skills while a few have concerns that sponsor control leads to the loss of their flexibility and independence Generally there is a perception that New Zealand based clinical trials assist in the process of obtaining registration and subsidization of new drugs in New Zealand. However, this perception may be erroneous as location of trials is apparently not considered in the drug registration process. CONCLUSIONS: We find that most stakeholders are satisfied with the conduct of clinical trials in New Zealand and they believe the benefits outweigh the costs. PHP89 PERCEPTION OF PHARMACO-EPIDEMIOLOGICAL STUDIES BY PHYSICIANS AND MOTIVATION TO PARTICIPATE: FRENCH SITUATION Schmidely N1, Bonhomme C1, Veysseyre H2, Nechadi S2, Longin J2 1 Bristol-Myers Squibb, Rueil Malmaison, France; 2REGISTRAT-MAPI, Lyon, France OBJECTIVES: Recent development of Health Technology Assessment worldwide increased requirement for real world data (e,g, risk management plans, drugs utilization). Primary non-interventional research (NIR) involving physicians able to enrol patients is a means to generate such data. One key challenge of NIRs is the capacity to reflect real life conditions by providing unbiased estimations of physicians’ behaviours and patients outcomes in large representative samples. Participation of physicians is often an issue when conducting NIRs. Lack of understanding of scientific value of NIRs, compared to clinical trials, is often assumed to be the major reason for reluctance to participate. Our objective was to better understand perception of, motivation to participate in and expectation from participation in NIRs by physicians. 13th Euro Abstracts METHODS: A survey was conducted in France on representative samples of 196 psychiatrists, 221 cardiologists and 250 general practitioners. Demographic characteristics were used to check representativeness. Perception of NIRs’ scientific interest, motivation criteria and expectations were assed and analyzed using descriptive statistics. RESULTS: >70% of physicians are convinced that NIRs are of scientific interest and provide reliable information on drug utilization in real life conditions. However, 10–15% are not convinced, 10–20% don’t have any opinion. Motivation relates to interest for scientific objective and studied disease (>90%), scientific rigour and credibility of partners (IRB review, scientific committee, sponsor) (>70%). NIRs requested by Health Authorities are not incentive. Physicians appreciate contributing to improving knowledge about drugs (>80%). >70% consider NIRs a challenging opportunity to compare their behaviour to their peers’. All (>90%) are interested in receiving results. 49% and 89% think results should be available within 6 and 12 months respectively. CONCLUSIONS: NIRs are recognized as a rigorous scientific tool by a majority of physicians. Motivation to participate is based on all scientific components of NIR. Results are expected in short term timeframe. PHP90 THE EFFECT OF THE 2007 HOSPITAL REFORM ON THE NUMBER OF PSYCHIATRIC BEDS IN HUNGARY Oberfrank F1, Donka-Verebes É2, Boncz I3 1 Institute of Experimental Medicine (IEM)., Budapest, Hungary; 2Integra Consulting zRt., Budapest, Hungary, 3University of Pécs, Pécs, Hungary OBJECTIVES: The new act on developing of the Hungarian health care system—came into effect on the 1st of April 2007 cut the number of hospital beds significantly in Hungary and the National Institute of Psychiatry and Neurology (OPNI) was closed. The aim of our paper is to analyze the effect of the 2007 Hungarian health care reform on the number of psychiatric hospital beds. METHODS: Data were derived from the nationwide administrative data set of the National Health Insurance Fund Administration (OEP), the only health care financing agency in Hungary. The number of both acute and chronic care hospital beds was evaluated before and after 1st of April 2007. RESULTS: The 2007 act on developing of the Hungarian health care system cut the total number of acute care hospital beds by 15.631 (26 %) from 59.846 to 44.215 (from 59,2 to 43,7 beds per 10000 inhabitants) and increased the total number of chronic care hospital beds by 7.069 (35 %) from 20.100 to 27.169. The number of acute psychiatric beds was reduced from 3.644 beds to 2.750 (−894 or −24,5% decrease). At the same time, the number of chronic care psychiatric beds was increased from 4.898 beds to 5.091 (193 or +3,9% increase). Altogether, 701 psychiatric beds were eliminated from the Hungarian health care system. CONCLUSIONS: Although the total number of chronic care hospital beds was increased by 35 % within the Hungarian health care reform in 2007, the Hungarian health care system lost 701 psychiatric care hospital beds without any reasonable consideration or explanation. PHP91 REGIONAL DIFFERENCES IN NUMBER OF PSYCHIATRIC BEDS IN HUNGARY FOLLOWING THE 2007 HEALTH CARE REFORM IN HUNGARY Oberfrank F1, Donka-Verebes É2, Boncz I3 1 Institute of Experimental Medicine (IEM), Budapest, Hungary; 2Integra Consulting zRt., Budapest, Hungary, 3University of Pécs, Pécs, Hungary OBJECTIVES: The new act on developing of the Hungarian health care system—came into effect on the 1st of April 2007 cut the number of hospital beds significantly in Hungary and the National Institute of Psychiatry and Neurology (OPNI) was closed. The aim of our paper is to analyze the effect of 2007 Hungarian health care reform on the number of psychiatric care hospital beds at regional level. METHODS: Data were derived from the nationwide administrative data set of the National Health Insurance Fund Administration (OEP), the only health care financing agency in Hungary. We carefully review the formal legislation and informal background papers related to this issue. The number of hospital beds was evaluated before and after 1st of April 2007. RESULTS: We found significant regional differences in the number of acute care psychiatric beds according to counties in Hungary. The number of acute psychiatric hospital beds per 10,000 inhabitants was the highest in the following counties: Békés 4.56; Szabolcs-Sz-B 4.18; Komárom-E 3.67; Budapest 3.56. The number of acute psychiatric hospital beds per 10,000 inhabitants was the lowest in the following counties: Zala 1.69; Somogy 1.50; Heves 1.48; J-N-Szolnok 1.36; Vas 1.13. CONCLUSIONS: Following the 2007 hospital bed reform in Hungary, there is a 4 times differences in the number of acute care psychiatric beds among the Hungarian counties. We did not find any reasonable consideration or explanation behind that fact. PHP92 COST-EFFECTIVENESS OF KNOWLEDGE TRANSLATION METHODS 1 2 3 Worbes-Cerezo M , Linertová R , Serrano-Aguilar P 1 CIBER de Epidemiología y Salud Pública (CIBERESP), Santa Cruz de Tenerife, Spain; 2 Fundación Canaria de Investigación y Salud (FUNCIS), Las Palmas de Gran Canaria, Spain; 3 Servicio Canario de la Salud, Santa Cruz de Tenerife, Spain OBJECTIVES: Given the increased use of interventions of knowledge translation to change clinical practice, it is necessary to evaluate their effectiveness and cost-effectiveness. In this systematic review we synthesized scientific evidence on the costeffectiveness of knowledge translation methods in health professionals. METHODS: We searched studies in MEDLINE and EconLit databases up to March 2010. Com- A421 plete economic evaluations based on clinical trials of knowledge translation in primary and specialized health care were selected, if published in English, Spanish, German or Italian. The primary outcome was the incremental cost-effectiveness ratio. Quality of the economic evaluations was assessed by the check-list by Drummond et al. The extracted information was presented in text and tables. RESULTS: Eighteen economic evaluations met the inclusion criteria. The knowledge translation interventions were aimed at different areas, like prescription, diagnostics, treatment, counselling and clinical practice guidelines. Thirteen studies were cost-effectiveness analyses, and five were cost-utility analyses. The effectiveness measures used in the cost-effectiveness studies are mainly intermediate results, while the cost-utility studies use mainly the QALYs (quality-adjusted life-years). The perspective mostly used was the one of health care system. Training/education is the intervention most frequently assessed followed by feedback, academic detailing, software support and distribution of educational materials; eight studies combined two or more methods. Most of the interventions are effective in knowledge translation, i.e. fulfil the function for which they have been designed. CONCLUSIONS: Although the scientific evidence suggests that the knowledge translation interventions are effective, most of them are relatively costly, and therefore less cost-effective. Those interventions that require less resources or can be applied in different areas at low cost, should be developed and implemented. PHP93 DOES THE PACKAGE DESIGN MATTER? THE INFLUENCE OF OTC PACKAGE DESIGN ON PATIENTS’ DRUG KNOWLEDGE Wilke T1, Neumann K2, Mueller S2, Loder T3 1 Institute for Pharmacoeconomics and Drug Logistics, University of Wismar, Wismar, Germany; 2University of Wismar, Wismar, Germany; 3Pharmacy Lux 99, Hürth, Germany OBJECTIVES: Objective of this study is to discover whether OTC package designs differ in their ability to transfer drug safety related information to pharmacy customers. The research was intended to answer two questions: 1) how well (in terms of speed and accuracy) do customers comprehend the relevant medical information on packages? 2) do alternative package designs cause significantly different levels of comprehension?. METHODS: Face-to-face interviews with 452 customers of 17 German pharmacies were conducted. In each of three sequential experiments (one OTC brand per experiment), each of the participants was shown two packages of the same brand (Paracetamol/Bisacodyl) and asked three drug-related questions per package. Multivariate linear regression analysis was used to identify the factors capable of explaining the accuracy of the answers (first outcome; quality) and the speed with which the answers were made (second outcome; time). The different abilities of the three package designs to transfer information correctly and rapidly were compared by descriptive statistics. The significance of the differences was tested by Wilcoxon tests. RESULTS: Older people, those with physiological problems, people not in a good mood, and those with a poor doctor-patient relationship, or a lack of trust in medicines in general answered the questions with a significantly higher number of mistakes and/or needed significantly longer to do so. When compared to two other package designs, one of the tested designs proved itself to be superior in quality and time. CONCLUSIONS: 1) There exist significant differences between patient groups in their ability to comprehend medical information; specific patients do not understand medical information on packages in a sufficient way. 2) The design of a medication package can measurably influence the quality and speed of information transfer to pharmacy customers. An optimized package design is a source of added value from a drug safety point of view. PHP94 IS IT COST-EFFECTIVE TO CHANGE THE BEHAVIOR OF HEALTH CARE PROFESSIONALS? Kusel J, Costello S, Haynes S, Brooks-Rooney C Costello Medical Consulting Ltd., Cambridge, UK BACKGROUND: Despite adequate literature on the effectiveness of behaviour change strategies for physicians, there is little information on whether these strategies are cost-effective. OBJECTIVES: To review the current literature on the cost-effectiveness of behaviour change policies for health care professionals. METHODS: A search of the MEDLINE database and of reference lists was performed up to May 2010. RESULTS: Twelve economic evaluations, assessing sixteen interventions in total, were identified. These were performed in either Europe or Australia. All but one evaluation had a control group, which received either no intervention (5) or a standard, less labour intensive intervention (7) (One evaluation had two phases and therefore two control groups). The type of control group had no association with whether the strategy was deemed cost-effective or not. Seven of the policies were deemed costeffective by the authors (remuneration, telemarketing, outreach, training sessions (3), distance learning), six were deemed not (outreach (4), training session, courier delivery of guidelines) and three reported increased benefits with increased costs but did not conclude whether this was cost-effective (outreach (2), training session). Three studies actually reported decreased costs with the test intervention over control, two of which were due to decreased hospital referrals. Of those behaviour change policies not deemed cost-effective, 50% had minimal or no increase in effect and the other 50% had a very significant increase in costs. CONCLUSIONS: Not all strategies were found to be cost-effective, which highlights the necessity of this type of analysis before major behaviour change policies are implemented. Multiple studies found that there were increased benefits with increased costs. With no defined threshold as to what makes a behaviour change policy cost-effective, the discussion is open to debate as to whether these benefits are worth the increased costs. A422 PHP95 THE DOS AND DON’TS IN PAYER COMMUNICATION. A QUALITATIVE RESEARCH ACROSS FIVE EUROPEAN COUNTRIES Guhl AK1, Gbenedio T1, Vekaria R1, Shepelev J1, Kolominsky-Rabas P2 1 GfK HealthCare, London, UK; 2Centre for Health Technology Assessment (HTA) and Public Health (IZPH), University of Erlangen-Nurnberg, Erlangen, Germany OBJECTIVES: National authorities across Europe implement both controls and incentives to influence the supply and demand of pharmaceuticals and medical devices. Today, greater emphasis is placed on increasing efficiency in the resources spent on health care via tighter controls on pricing and reimbursement regulations. Companies increasingly face the challenge to communicate their value propositions towards payers adequately. We investigated in 5 European countries (D,F,UK,NL,I) if companies meet payers expectations and also asked for suggestions how companies could improve. METHODS: A questionnaire was developed, validated and translated into the local language. Some questions were specific to the national health care system. In total 20 face—to- face interviews were conducted with decision makers at various levels. All interviews were recorded. RESULTS: The expectations among payers across Europe with regard to industry communication vary widely. The emphasis lies in the presentation of scientific data with more focus on clinical rather than economic data. Some of the interviewees revealed some poor examples and mistakes made by companies, others gave us insights into what an “ideal” communication platform with the industry could look like on a national level. CONCLUSIONS: The results showed that even in Europe there is a huge variety how to communicate with payers from an industry perspective. Cultural and health care system specifics need to be taken into account to develop successful payer communication strategies. PHP96 SENSE OF COHERENCE—AS MEDIATING FACTOR IN REMAINING EMPLOYED IN HEALTH CARE PROFESSION Roznár J, Oláh A, Tóth Á, Betlehem J, Boncz I, Müller Á, Jeges S University of Pécs, Pécs, Hungary OBJECTIVES: To prove the hypothesis that sense of coherence has a significant mediating role in keeping career in health care. METHODS: Our database: students in the programme of nursing (n = 203) versus randomly selected sample of students studying in other programmes (n = 196); and health care professionals (n = 96) vs. a randomly selected sample of people employed elsewhere but not in health care (n = 741). Anonym self-completed questionnaires were used: Antonovsky’s SOC scale validated on Hungarian sample; Hennenhofer’Heil’s vegetative lability test and a health self-estimation scale. Used statistical methods: variance analysis and multinomial logistic regression. RESULTS: According to the ANOVA and post hoc tests, global sense of coherence of students studying in health care programmes was significantly lower (p < 0,001) than that of those students studying in other programmes; the results of health care professionals did not significantly differ from the control sample. The highest SOC were scored by those health care professionals who were studying besides working. According to the indicators of state of health the health self-estimation of health care students was significantly lower than that of those participating in other programmes (based on the post hoc t-tests, P < 0,05) however, between the two samples of employees there was no significant difference. We have proved with multinominal logistic regression models that SOC and the state of health are mostly influenced by physical activity. CONCLUSIONS: Our hypothesis has been confirmed by the results. Relatively more health care students with a high SOC are employed in the health care. This could be the explanation for the phenomenon that health care students have lower SOC compared to those studying in other programmes but in both groups of employees there is no difference concerning SOC. This proves the importance of enhancement of SOC among health care students, which can be primarily achieved by augmenting physical activity. PHP97 TRAINING OF PHARMACEUTICAL SALES REPRESENTATIVES IN REGARD TO CHANGES IN HEALTH CARE POLICY IN GERMANY Benkert D1, Saile C1, Schauer S1, Kossmann B1, Wasem J2, Aidelsburger P1 1 CAREM GmbH, Sauerlach, Germany; 2University of Duisburg-Essen, Essen, Germany OBJECTIVES: Due to rapid changes in health policy in Germany there is a need for actors in health care system for constant training about the system. Aim of the present study is to assess different teaching methods using the example of training pharmaceutical sales representatives concerning effectiveness assessment as performed by several institutions in Germany. METHODS: Two learning approaches are compared. “Health Consultant (HC)” is a training program teaching pharmaceutical sales representatives in health policy topics using traditional training methods like talks and discussions. Second approach is an e-learning program. Participants were interviewed by questionnaires about their learning preferences. RESULTS: Evaluation of the HC training program shows, that the traditional teaching method is well accepted. Learning success was confirmed by subjective valuation. Most participants report a demand for a more flexible and individual learning approach that is compatible to their everyday work. E-learning programmes could meet these criteria. CONCLUSIONS: Further studies are necessary to evaluate the learning success of both methods. Evaluation instruments are already developed. The dimensions that should be included are motivation, workloads and structure of learning unit. 13th Euro Abstracts PHP98 COST-EFFECTIVENESS METHODOLOGIES OF STRATEGIES THAT AIM TO CHANGE THE BEHAVIOUR OF HEALTH CARE PROFESSIONALS Kusel J, Costello S, Brooks-Rooney C, Hamer N Costello Medical Consulting Ltd, Cambridge, UK BACKGROUND: As the cost-effectiveness of strategies that aim to change the behaviour of physicians has not been widely studied, it is unclear whether a standard methodology is being utilised for measuring cost-effectiveness in this situation. OBJECTIVES: To review the current literature on the methodologies used in costeffectiveness evaluations of behaviour change policies for health care professionals. METHODS: A search of the MEDLINE database and of reference lists was performed up to May 2010. RESULTS: Twelve relevant evaluations, from ten publications, were identified; two discussed costs and benefits separately and were therefore not classed as true cost-effectiveness evaluations. Of the remaining ten, seven were based on randomised controlled trials, two used a model without empirical data and one was a prospective cohort study. Eight of the ten evaluations utilised health outcomes as the main effectiveness measure, five of which used life-years (LY) or quality-adjusted LYs gained. The ten cost-effectiveness evaluations measured: implementation cost per LY or health outcome gained (6); total policy cost per LY gained (2); implementation cost per percent change in prescribing (1) and percent change in health outcome per pound (1). Nine took a health care provider perspective; one took a societal perspective that included patient expenditure. There was no agreement between the ten evaluations as to whether physician time should be included in the cost analysis or not (yes (3), no (4), not stated (3)). Only two evaluations considered a time frame beyond 1 year and took into account that future implementations would be required to maintain the behaviour change. Discounting of values was utilised by these evaluations (3% per year), and by 1 other study (percent not stated). CONCLUSIONS: More robust evaluations are required, with health outcomes assessed as LYs for comparison purposes. Future evaluations must consider timeframe, discounting, societal costs and costs due to physician time. PHP99 FIRST-IN-THERAPY PRODUCTS AND REQUIREMENTS FOR SUCCESSFUL HTA ASSESSMENT Xia AD, Oraro T Heron Evidence Development Ltd., London, UK OBJECTIVES: To determine the necessary requirements of a first-in-therapy product to support positive HTA review. METHODS: An initial broad search was conducted to identify first-in-therapy products launching in recent years into disease areas with no alternative treatments. The products identified were eculizumab for paroxysmal nocturnal hemoglobinuria (PNH), pregabalin for fibromyalgia, vigabatrin for infantile spasms, tetrabenazine for tardive dyskinesia (TD) and chorea of Huntington’s disease, pirfenidone for idiopathic pulmonary fibrosis (IPF), and amifampridine for LambertEaton myasthenic syndrome (LEMS). These products were then examined within NICE, SMC, NCPE, PBAC and CADTH websites, and information collated on clinical endpoints, HTA comparators and assessment outcomes to provide an understanding of the requirements of first-in-therapy products for positive review. RESULTS: Of the drugs selected, 4 were assessed by HTA bodies. Due to a lack of alternative therapies, all but one product assessed carried out trials against placebo. Clinical endpoints were based either on metrics agreed upon by KOLs, quality of life measures, or both. Out of the seven drugs studied, only vigabatrin received positive recommendation due to strong trial data and inclusion of an active symptomatic comparator. a number of first-in-therapy products were actually rejected by HTAs in select countries. Pregabalin was rejected due to unconvincing trial data despite a favourable cost per QALY and eculizumab was rejected due to too high an incremental cost per QALY. CONCLUSIONS: We can conclude that based on the products assessed, first-in-therapy products do not typically receive special consideration when assessed by HTA bodies. An active comparator, use of recognized endpoints and the quality of data remain as important requirements for positive review. Furthermore, demonstrated cost-effectiveness is required even in a novel disease area, thus, it will be necessary to evaluate this earlier in development. More focused studies need to be carried out to ascertain this trend on a country-by-country basis. HEALTH CARE USE & POLICY STUDIES – Health Technology Assessment Programs PHP100 EVALUATION OF ACCEPTANCE AND REJECTION RATES OF ORPHAN DRUGS ACROSS SIX HTA AGENCIES Jäkel A1, Alnwick K2 1 Heron Evidence Development Ltd, Stopsley, UK; 2Heron Evidence Development Ltd, London, UK OBJECTIVES: Many orphan drugs receive market authorization, however an assessment of cost-effectiveness is usually required before these drugs can be reimbursed. Recent research has explored the differences between HTA bodies in their rates of acceptance/rejection of orphan drugs. The objective of this study was to consider the relative weight of economic and clinical reasons for rejecting orphan drug submissions to HTA agencies. METHODS: Six HTA websites (NICE, SMC, NCPE, CEDAC, PBAC, AWMSG) were searched for summary guidance on 71 licensed orphan drugs identified from a search of the Orphanet website. RESULTS: Of 53 total non-approved 13th Euro Abstracts submissions, in 53% of cases a high ICER was reported in the summary of guidance as a reason for rejection. In about 30% of these cases, the high drug cost was specified as the driver of the high ICER. The lack of a robust economic case was mentioned in 45% of rejections. Limited evidence of clinical benefit was shown in 43% of cases. Other reasons included inadequate type or quality of clinical data (21%) and nonacceptance of clinical positioning (11%). In 45% of cases the rejections were largely due to economic reasons; 6% of cases were not accepted due primarily to clinical reasons and in 49% of rejections the criticisms related to both the economic and clinical evidence. Uncertainty in the evidence was reported as a problem in most negative recommendations. CONCLUSIONS: Only slightly over half of the orphan drug HTA submissions to these agencies are explicitly rejected primarily on the basis of a high ICER. Most HTA rejections are due to a combined lack of robust economic and clinical evidence. This suggests that collecting the right kind of data and presenting a solid case that accounts adequately for any uncertainty is at least as important as meeting trial endpoints and choosing an optimal price. PHP101 EVALUATION OF ACCEPTANCE AND REJECTION RATES OF ORPHAN DRUGS ACROSS SIX HTA BODIES Oraro T, Alnwick K Heron Evidence Development Ltd., London, UK OBJECTIVES: Orphan drugs (ODs) face numerous difficulties in demonstrating their value through rigorous HTA processes. HTA bodies may therefore choose to take into consideration the special circumstances of treatments for orphan indications, either formally or informally. The objective of this study was to examine the decisions made on ODs by six English-speaking HTA bodies, and assess potential trends between agencies. METHODS: AWMSG, CEDAC, NCPE, NICE, PBAC and SMC HTA websites were searched for completed OD assessments (identified via Orphanet website) as of April 2010, and data extracted on the recommendations. Recommendations with restrictions were categorised as approvals. Comparisons were made to published approval rates for drug submissions as a whole. RESULTS: Of the 71 ODs selected, 55 were assessed by at least one HTA body. The proportion of positive recommendations for orphan treatments was lower than published approval rates for general (orphan and non-orphan) HTA drug submissions in most bodies. NICE approved 67% completed OD submissions, versus approximately 87% of drugs overall. However, PBAC recommended 60% of ODs compared to 54% of drugs as a whole. Decisions also varied substantially between agencies. CEDAC had the highest proportion of rejections (73%), compared to NICE, which rejected 33%. There was also variation in decisions made on specific treatments. Sutent, for example, with an orphan designation for renal cell carcinoma, was accepted by AWMSG, CEDAC and SMC, while it received a negative recommendation from NCPE, NICE and PBAC. CONCLUSIONS: The willingness to assess ODs varies widely by agency and drug, as do resultant approval rates. HTA agencies are far more likely to reject OD submissions than non-orphan drugs as a whole. However given the differences in their remit to assess ODs, direct comparisons should be interpreted with caution. Further research is needed to explore the reasons behind these differences in HTA agency decisions. PHP102 A SURVEY OF HTA RESEARCH METHODS AND TRENDS IN EUROPE ISPOR HTA SIG Research Methods/Principles Working Group ISPOR, Lawrenceville, NJ, USA OBJECTIVES: To describe research methods and key issues in the HTA process in Europe. METHODS: Representatives from HTA bodies globally were recruited by members of the ISPOR HTA SIG Research Methods/Principles Working Group to complete a 45-minute on-line survey consisting of 48 items within 4 topics related to 1) organizational information and process; 2) primary HTA methodologies and importance of attributes; 3) HTA application and dissemination; and 4) quality of HTA including key issues. Data were reported for Europe. RESULTS: The survey was completed by 11 European countries including Austria, Denmark, France, Germany, Hungary, Italy, The Netherlands, Portugal, Spain, Sweden, and Switzerland. Top reasons technologies were evaluated included perceived impact on patient outcomes, potential cost, and prevalence of the condition. The most common methodologies used were cost/economic analyses, systematic reviews & meta-analyses, clinical trials, modeling, and comparative analyses. The most important attributes (in order) were effectiveness, efficacy, safety, cost-effectiveness, and budget impact. While quality of life was frequently assessed by >74% of European respondents, it was not listed as an attribute of top importance. Only 24% repeat/update the assessment at regular intervals. For 82% a different organization makes the final decision on coverage, only partially relying on the report. The most common educational background for decision makers was physician-specialist. Stakeholders are allowed to review the report and are involved in assessments >50% of the time, and in the final decisions ~35% of the time. Key issues/trends included early assessment of technologies with mechanism for conditional coverage, increasing regional interest in HTA, reassessment/horizon scanning, and link between theory and practice in HTA. CONCLUSIONS: This survey of representatives within HTA and reimbursement bodies provides current insight into the state of HTA research methods in Europe. Future research could expand the results to specifically address Eastern European countries, Asia, and other emerging markets. A423 PHP103 NICE GUIDANCE: AN ANALYSIS OF LEVELS OF RESTRICTION BY DISEASE AREA Mesa OA1, Venus A2, Lebmeier M1, Davis M1, Jones C1 1 Bristol-Myers Squibb Pharmaceuticals Ltd, Uxbridge, Middlesex, UK; 2Bristol-Myers Squibb Pharmaceuticals Ltd, Princeton, NJ, USA OBJECTIVES: To assess the outcomes of NICE’s guidance in totality and different disease areas. METHODS: A list of NICE Guidance published between 2007 to the end of 2009 was identified using HTAinSite™. We classified these recommendations into; ‘recommended’, ‘restricted’ and ‘not recommended’, and calculated the percentages. We then analyzed these recommendations according to disease areas; ‘cardiovascular/metabolics’, ‘mental health’, ‘infectious diseases’, muscoskeletal conditions’, ‘oncology’, and ‘others’. RESULTS: In 2007, NICE assessed 25 drugs, 31 in 2008 and 18 in 2009. Of these, in 2007 NICE recommended 8 drugs (31%) for all eligible patients, restricted 13 (53%), and did not recommend 4 (16%). In 2008, 3 (10%) were recommended, 21 (68%) were restricted, and 7 (22%) were not recommended. Finally, in 2009, 2 (11%) treatments were recommended, while 11 (61%) received restricted recommendations and 5 (28%) were not recommended. Between 2007 and 2009 NICE completed 7 appraisals in ‘cardiovascular/metabolics’ of which 4 received a full recommendation, while in ‘mental health’ 2 out of 2 were fully recommended. In contrast, in ‘infectious diseases’, 1 out of 5 was fully recommended. In ‘muscoskeletal conditions’ only 1 out of 21 were recommended (17 restricted and 3 not recommended) while in ‘oncology’ only 1 out of 23 received a full recommendation (13 restricted, 9 not recommended). In the ‘others’ group, 4 out of 12 received a recommendation (6 restricted, 2 not recommended). If manufacturers had not proposed Patient Access Schemes (PAS) the proportion of guidance not recommended in 2009 would be 44%. CONCLUSIONS: Appraisal outcomes have become more restrictive over time. Furthermore, low cost primary care therapeutics are more likely to receive a positive NICE recommendation than high cost speciality care interventions. PHP104 DESIGNING EUROPEAN GUIDELINES FOR HEALTH OUTCOMES AND COST-EFFECTIVENESS ASSESSMENTS: THE ECHOUTCOME EUROPEAN COMMISSION PROJECT Beresniak A1, Auray J2, Duru G3, Medina-Lara A4, Praet J5, Sambuc R6, Tarricone R4, Torbica A4, De Wever A5, Lamure M7 1 Data Mining International, Geneva, Switzerland; 2Cyklad Group, Rilleux la Pape, France; 3 Cyklad Group, rilleux la Pape, France; 4Bocconi University, Milano, Italy; 5Université Libre de Bruxelles, Brussels, Belgium; 6Université de la Mediterrannee, Marseille, France; 7University Claude Bernard Lyon 1, Paris, France OBJECTIVES: Over the last decade the National Institute of Clinical Excellence in the UK has published guidelines for health technology assessments (HTA) that includes recommendations on health outcomes and cost-effectiveness assessments. In Europe, this has opened the opportunity for countries to either propose their own guidelines or use the British ones. The ECHOUTCOME project is an interdisciplinary European research platform funded by the seventh Framework Program of the European Commission with the aim of designing new European guidelines in Health Outcomes and Cost-Effectiveness assessments. METHODS: This three years project is structured in three phases. Phase 1 aims to conduct a pan-European survey of HTA organizations and health outcomes use in the 27 European countries. Multiple correspondence and cluster analyses will be carried out to study the potential similarities and divergences across Europe. The objective of Phase 2 is to test the robustness and underlying assumptions such as reproducibility, neutrality to risk, constancy of time-trade-off rate, utility independence, etc. on QALYs, DALYs and HYEs measures. This testing will be conducted in the general population (n = 300 per country) in Belgium, France, Italy and UK. Phase 3 aims to propose new approaches in Health Outcomes and Cost-Effectiveness analyses. RESULTS: The main deliverable of the ECHOUTCOME project will be new European Guidelines for assessing Health Outcomes and conducting Cost-Effectiveness assessments. Of particular interest will be the recommendations on the practical usefulness of QALYs, DALYs and HYEs based on the experimental validation of their underlying assumptions. CONCLUSIONS: The ECHOUTCOME project is the first European validation study of health outcomes measures. This work will produce guidelines for public health decision-making in the 27 European countries. The ECHOUTCOME outcomes will enhance the debate and increase the understanding that will improve the knowledge of existing Health Outcomes and Cost-Effectiveness techniques and will promote new approaches for decision-making. PHP105 HOW CAN THE USE OF PREDICTIVE BIOMARKERS LEAD TO POSITIVE HTA RECOMMENDATIONS? Trevor NC1, Alnwick K2 1 Heron Evidence Development Ltd, London, UK; 2Heron Evidence Development Ltd, Luton, UK OBJECTIVES: The popularity and availability of biomarkers has rapidly increased in recent years, thanks to innovative advances in pharmacogenomics. Predictive biomarkers have high potential value in HTA as they may increase the observed efficacy and costeffectiveness of treatments. This study reviews, in selected major markets worldwide, the impact of currently available predictive biomarkers on HTA in the context of the agency’s evidence requirements. METHODS: A broad review of biomarker tests used in HTA submissions in Europe, Australia, Canada, and the US was conducted; the A424 outcome of the HTA recorded and the impact of the biomarker test on the submission outcome was graded as high, medium or low according to its influence on the final decision. These findings were summarised, and 6 drugs were selected as case studies in order to identify key lessons relating to the risks, consequences, and ethical considerations of Diagnostic/Treatment partnering. RESULTS: The review identified five biomarkers in the five treatment areas of: HIV, Gastrointestinal stromal tumour (GIST), Non-small cell lung cancer (NSCLC), Colorectal cancer (CRC), and Breast cancer. Markers Her2 and K-RAS had a high impact in all included submissions, with 100% and 63% of these submissions resulting in a positive recommendation. In contrast, marker EGFR had a lower impact (not mentioned in 4 out of 10 submissions), with 60% of these submissions being approved, and 40% rejected. The agencies most likely to reject a surrogate-outcome submission were PBAC (Australia) and SMC (Scotland) with rejection rates of 57% and 66% respectively, whereas CADTH accepted 100% of included submissions. CONCLUSIONS: Findings indicate firstly that substantially different evidence requirements exist between HTA bodies in the markets considered (e.g. differing accuracy acceptability thresholds, prospective/retrospective analysis and the importance of cost-effectiveness), and secondly there are several ethical considerations to the selection or deselection of patients for treatment. PHP106 HTAINSITE: A DATABASE OF NICE SUBMISSIONS AND DECISIONS Howard S, Harper C, Poppe P Abacus International, Bicester, Oxon, UK OBJECTIVES: HTAinSite is an on-line, subscription-based database of all NICE Technology Appraisals (TAs). It includes information relating to submitted evidence, appraisal process and final decision and enables assessment of associated trends. METHODS: An academic steering group designed and agreed the data extraction protocol. a team of reviewers conducted the initial data extraction, which was validated by a second reviewer. Historical extraction is complete, with on-going TAs extracted on a monthly basis. RESULTS: A total of 181 TAs have been extracted to date, 5 of which were terminated, resulting in 176 complete TAs relating to 326 technologies. Of these, 53% were awarded a restricted decision, 32% were fully recommended and 16% rejected. Of the three most commonly appraised disease areas (cancer, cardiovascular (CVS), central nervous system (CNS); CVS technologies were least commonly rejected by NICE (3% vs. 21% for cancer and 17% for CNS). Of the four most commonly used assessment groups (Sheffield, York, Southampton and Birmingham); technologies assessed by Sheffield were most commonly rejected (28%) and those by York were least commonly rejected (4%). Technologies supported by a patient submission were less commonly rejected than those without (15% vs. 29%). a total of 35 TAs (20%) resulted in an appeal, no appeals were upheld entirely, all appeals were upheld partially. HTAinSite also allows detailed analysis of individual TAs and cross-comparison between TAs as well as identifying trends between submitted evidence (acquisition costs, budget impact, cost-effectiveness and clinical effectiveness) and final outcome. These additional analyses will be further explored in the poster and preliminary results will also be updated. CONCLUSIONS: HTAinSite is a useful tool for anyone interested in understanding the relationship between submitted evidence and ultimate NICE decision. The HTAinSite format may be useful for other HTA bodies, depending on the public availability of relevant information. PHP107 ECONOMIC EVIDENCE REQUIREMENTS: COMPARISON BETWEEN HTA AGENCIES AND IMPLICATIONS FOR MANUFACTURERS Balvanyos J, Alnwick K, Proudfoot C Heron Evidence Development Ltd., London, UK OBJECTIVES: Mathematical models are required by decision makers to provide insight into pharmcoeconomic benefits associated with a product. It is therefore essential that manufacturers understand economic evidence requirements when submitting an application to a Health Technology Assessment (HTA) agency. METHODS: A literature search of economic recommendations from the following HTA agencies was conducted: CADTH (Canada), HAS (France), IQWiG (Germany), NICE (England), PBAC (Australia), PHARMAC (New Zealand) and SMC (Scotland). RESULTS: Cost-effectiveness analysis is considered the most relevant analytical technique across the English-speaking agencies with a preference for QALY-based analysis, in contrast to IQWiG which does not consider QALYs and utilities as central to their methods. Unlike other HTA agencies assessed, the French agency HAS does not currently require cost-effectiveness modelling in its decision-making process. EQ-5D is the most commonly used utility instrument, NICE being the most prescriptive agency in this regard. However, utilities mapped from disease-specific quality of life measures may be accepted and agencies such as PBAC and IQWiG express no formal preference between instruments. The third-party payer is the most commonly required perspective adopted across the English-speaking agencies while IQWiG and PBAC recommend a societal perspective in addition to the payer’s perspective. This trend towards a broader, societal perspective may, however, be limited by uncertainties around measurement of wider costs. Sensitivity analyses are required by all agencies to explore uncertainty in the model. NICE and CADTH both favor a probabilistic approach while PBAC prefers univariate and multivariate analyses. Other agencies expect the manufacturer to justify their approach and choice. CONCLUSIONS: The recommendations of IQWiG and PBAC differ from the other selected agencies. Uniquely, in France the pharmaeconomic case is considered separately from the HTA process by the French Health Economists Association. These differences between agencies should be considered when planning evidence generation activities to support economic model development. 13th Euro Abstracts PHP108 CLINICAL EVIDENCE REQUIREMENTS: COMPARISON BETWEEN SEVEN HTA AGENCIES AND IMPLICATIONS FOR DRUG MANUFACTURERS Balvanyos J, Alnwick K, Proudfoot C Heron Evidence Development Ltd., London, UK OBJECTIVES: Health Technology Assessment (HTA) agencies require various types and qualities of evidence for clinical effectiveness evaluations due to differences in health care systems and policies. It is essential for manufacturers to understand these requirements when submitting an application to each individual HTA agency. METHODS: A literature search of clinical recommendations from the following HTA agencies was conducted for comparison: CADTH (Canada), HAS (France), IQWiG (Germany), NICE (England), PBAC (Australia), PHARMAC (New Zealand) and SMC (Scotland). RESULTS: The choice of the optimal comparator is crucial to the outcome of the HTA. Almost all agencies prefer comparison versus the most frequently used interventions except for PBAC which requires comparison to the interventions most likely to be displaced. All HTA agencies are cautious in their interpretation of surrogate outcomes (SO) and require manufacturers to provide evidence linking the SO to final patient-relevant outcomes. PBAC has notably developed a framework for assessing SOs and the impact of these on uncertainty in HTA submissions. Most agencies except for NICE clearly state their position on the definition and the use of SOs. All agencies recognize the value of observational studies in reflecting real-world situations and providing long-term data although RCTs provide the key evidence on comparative effectiveness. Systematic reviews (SR) of clinical evidence are essential to present comparative effectiveness relative to all comparators. Contrary to most agencies, HAS prefers SRs but does not require them and bases its assessments mainly on pivotal clinical trials provided by the manufacturer. NICE and IQWiG also differ from the other agencies as they perform in-house SR in addition to the manufacturer’s. CONCLUSIONS: The differences between agency requirements are subtle and mean that manufacturers need to put together a solid clinical evidence package needing very little adaptation to meet the seven country requirements. PHP109 RELATIVE EFFECTIVENESS ASSESSMENT OF PHARMACEUTICALS Kleijnen S1, George E2, Goettsch W1, d’Andon A3, Osi ska B4, Corbacho B5 1 College voor Zorgverzekeringen, Diemen, The Netherlands; 2National Institute for Health and Clinical Excellence, London, UK; 3Haute Autorité de Santé, Saint-Denis La Plaine Cedex, France; 4Agency of Health Technology Assessment in Poland, Warsaw, Poland; 5Andalusian Agency for Health Technology Assessment, Seville, Spain OBJECTIVES: The EUnetHTA Joint Action (2010–2012) is a joint initiative by EU Member States and the European Commission to advance European collaboration on health technology assessment (HTA). Part of the EUnetHTA Joint Action aims at reviewing methods used for the relative effectiveness assessment (REA) of pharmaceuticals and to develop, apply and field-test tools and methods. As a first step towards this goal, objectives, processes and methodologies used for REA by HTA organizations across Europe and other countries are summarised. This overview is crucial for the development of shared or common methodologies to be used in future REA across Europe. METHODS: Data were captured with a standardised data abstraction form. Data were initially abstracted from different types of literature (peer reviewed, grey literature, EU and national reports etc.). Where there were gaps in the data, telephone interviews were arranged with a relevant person at the respective HTA or reimbursement agency. RESULTS: Most European countries carry out some form of REA, however the definitions of the assessments in general are not consistent with the definition by the Pharmaceutical Forum. In addition the scope (inclusions of benefit-risk analysis and/or a cost-effectiveness analysis), the process (timing, involvement of stakeholders and separation of assessment and appraisal phase) and the purpose (inform clinical decision-making, reimbursement and/or for pricing decision-making) of the assessments vary as well as the methods (such as inclusion of surrogate endpoints, composite endpoints and quality of life data and extrapolation of efficacy data). Most agencies publish a guideline for the methodology used for REA however few publish them in English. CONCLUSIONS: A considerable number of European countries carry out REAs. However, the purpose and the methodology used vary across Europe. The reasons behind these differences need to be considered in the development of a common European methodology for REA. HEALTH CARE USE & POLICY STUDIES – Population Health PHP110 PHYSICIANS’ VIEWS OF THE RELATIVE IMPORTANCE OF SELECTED MEDICAL INNOVATIONS ON THE GREEK POPULATION HEALTH STATUS Athanasakis K, Karampli E, Kyriopoulos J National School of Public Health, Athens, Greece OBJECTIVES: To identify the pharmaceutical and medical innovations that contributed mostly to the improvement of Greek population health status during the last three decades, according to physicians’ views. METHODS: Building on the methodology by Fuchs and Sox, a questionnaire based survey was conducted on a representative sample of 500 Greek internists and general practitioners aged ≥50 years old. The study questionnaire was formulated by a panel of experts, with the use of the Delphi method and included one list of 22 pharmaceutical and a second list of 20 medical innovations. Physicians were asked to identify the seven more important and seven least important 13th Euro Abstracts innovations, with regard to their impact on patients’ health status. RESULTS: Response rate was 78%. The seven most important pharmaceutical innovations reported were Angiotensin Converting Enzyme (ACE) inhibitors and angiotensin II antagonists, inhaled steroids and β2-agonists, statins, proton pump inhibitors and H2-antagonists, novel antibiotics, antiviral drugs for hepatitis and C and calcium channel blockers. The seven most important technological innovations were: magnetic resonance imaging (MRI) and computed tomography scanning (CT), balloon angioplasty with stents, Coronary Artery Bypass Graft (CABG), gastrointestinal endoscopy, Human Immunodeficiency Virus testing, mammography and Prostate-Specific Antigen (PSA) testing. In both cases, innovations that ranked higher were used to treat high-prevalence diseases with a significant contribution on the burden of disease for the Greek population. In addition, innovations that contributed to improvements in disease diagnoses and management were also ranked higher. CONCLUSIONS: the epidemiological profile of the population, the effectiveness of each innovation in terms of clinical effectiveness as well as health-related quality of life, and the utility of the innovation in everyday clinical practice constitute important determinants of the physicians’ opinions regarding the relative importance of medical innovations on the health status of the Greek population. PHP111 PREDICTING PREVENTIVE CARE SERVICE UTILIZATION IN A UNITED STATES POPULATION Partha G1, Vaidya V1, Howe J2, Ferrell M1 1 University of Toledo, Toledo, OH, USA; 2The University of Toledo, Toledo, OH, USA OBJECTIVES: To analyze and predict trends in utilization of preventive care services in a United States population using a national database. Several national reports and evidence in literature indicate disparities in the utilization of preventive care services but only few have predicted utilization patterns of preventive care services. METHODS: The 2007 Medical Expenditure Panel Survey (MEPS), a national probability sample survey of the ambulatory civilian US population, was analyzed to determine demographic patterns of utilization. Utilization of blood pressure screening, cancer screening (mammography, colonoscopy/sigmoidoscopy, pap smear and breast), cholesterol and dental checkup and flu vaccination were used as the dependent variables while age, gender, race/ ethnicity, annual income insurance status and perceived health status were used as independent variables. Descriptive statistics were used to describe the population; univariate chi-squared analysis was used to determine group differences for the categorical variables. Multivariate logistic regression model was built to predict odds of utilization of services. All analyses were carried out using SAS v9.1. RESULTS: Utilization of preventive care services was found to be high for blood pressure (n = 17,959, 89.0%) and cholesterol (n = 14,956, 94.7%) checkup and low for flu vaccination (n = 30,964, 21.2%). Significant chi square differences in utilization of preventive care services were found for most of the dependent variables (P < 0.001), save for colonoscopy/sigmoidoscopy screening. Regression model showed that young women (20–30 years) with higher income were more likely to obtain breast exams (OR = 10.584; CI = 4.312–25.980) compared with young women earning lower income (<$20,000). Odds of utilization were nearly similar for all races. However, Hispanics were more likely to obtain a colonoscopy/sigmoidoscopy (OR = 3.069; CI = 2.216– 4.250) compared with non-Hispanics. The uninsured generally had lower odds of utilization across almost all preventive care services. CONCLUSIONS: This study was successful in identifying age, race, income and insurance status-related disparities and in predicting preventive care service utilization in a US population. HEALTH CARE USE & POLICY STUDIES – Prescribing Behavior & Treatment Guidelines PHP112 ANALYSIS OF PHARMACISTS’ INTERVENTIONS OF ELECTRONIC PRESCRIPTIONS AT SULTAN QABOOS UNIVERSITY HOSPITAL IN OMAN Al-Rashdi IS, Victoria M, Susan S, Al-Zakwani IS Sultan Qaboos University, Al-Khoudh, Oman OBJECTIVES: To evaluate the number and types of pharmacists’ interventions of electronic prescriptions at Sultan Qaboos University (SQU) hospital in Oman. METHODS: This was a prospective study where interventions on electronic prescriptions over one-year (2009) at SQU hospital were evaluated. a standard data collection form was used to capture the data related to drug choice and drug regimen. Clinical relevance was defined as to whether efficacy or toxicity was either improved or reduced. Clinical relevance was based on the judgments of at least two pharmacists. Analyses were performed using descriptive statistics. RESULTS: A total of 1,123 interventions were recorded out of 186,353 prescriptions (intervention rate/prescription = 0.6%) and 454,654 items (intervention rate/item = 0.25%) dispensed. During the year, only 3% of the interventions were administrative (absence of doctor’s signature/wrong patient’s card) while 97% were clinical. The clinical interventions were categorized into drug regimen (n = 886; 74%) and drug choice (n = 313; 26%). Almost 62% (n = 547) of the problems associated with drug regimen were related to wrong doses. The three most common drug choice issues included wrong formulations (n = 107; 34%), wrong drug (n = 62; 20%) and deletion (n = 62; 20%). Efficacy improved in 52% (n = 588) of the cases, avoided toxicity in 29% (n = 324) and avoided unnecessary exposure in 19% (n = 211). Interventions prevented organ damage in 1.7% (n = 20) of the cases, major in 17% (n = 188), moderate in 49% (n = 550) and minor in 32% (n = 364). The mean time (±SD) spent per prescription was 11 ± 9 minutes. A425 CONCLUSIONS: The data show that out-patient pharmacists play an important role in drug safety and optimization of pharmaceutical care. PHP113 LEVELS OF MEDICATION USE AMONG GERIATRIC PATIENTS IN ACUTE GERIATRIC CARE SETTINGS IN AUSTRIA Koenig C1, Tschapeller B2, Perner P2, Pils K3, Sommeregger U4, Fruehwald T5, Dovjak P6, Pinter G7, Boehmer F3, Huf J3, Siber H3, Scala M8, Stoiser E8, Kleindienst R9, Hartweger A10, Haid M11, Krippl P12, Hohl V13, Wehrmann A14, Reisinger W15, Reif-Gintl T16, Mueller W17, Gaugeler R18, Macho A19, Habacher W1, Beck P1, Mrak P20 1 Joanneum Research, Graz, Austria; 2JOANNEUM RESEARCH, Graz, Styria, Austria; 3SMZ Sophienspital, Vienna, Austria; 4KH Hietzing mit NZ Rosenhügel, Vienna, Austria; 5KH— Hietzing mit NZ Rosenhügel, Vienna, Austria; 6LKH Gmunden, Gmunden, Austria; 7LKH Klagenfurt, Klagenfurt, Austria; 8Albert Schweitzer Klinik, Graz, Austria; 9LKH Laas, Koetschach, Austria; 10LKH Rottenmann, Rottenmann, Austria; 11Marienkrankenhaus Vorau, Vorau, Austria; 12LKH Fuerstenfeld, Fuerstenfeld, Austria; 13LKH Voitsberg, Voitsberg, Austria; 14 KH Goettlicher Heiland, Vienna, Austria; 15Herz-Jesu-Krankenhaus, Vienna, Austria; 16KH Barmherzige Brueder, Vienna, Austria; 17KH der Elisabethinen, Klagenfurt, Austria; 18 Oeffentliches KH Waiern/Feldkirchen, Feldkirchen, Austria; 19Hartmannspital, Vienna, Austria; 20LKH Hoergas, Gratwein, Austria OBJECTIVES: Adequate levels of medication are critical among elderly patients given the existence of co-morbidities and impaired levels of functioning. Using patient data collected via a web-based information system, we investigated the average number of items prescribed to geriatric patients treated in Acute Geriatric and Remobilisation Units (AG/R) in Austria between January 2008 and December 2009. We investigated the association of age with medication levels and assessed whether treatment in an AG/R led to changes in levels of medication. METHODS: The standardised data set developed by QiGG was made available by AG/Rs using the geriatric module of the online information system “Healthgate BARS”. We assessed patients’ medication data including number of prescribed items at admission and discharge. Baseline characteristics (including age) were assessed. Additionally, changes in medication prescriptions during the stay were documented. RESULTS: Data from 14,923 stays were analyzed. Patients were prescribed a mean of 8.4 agents (±3.6) at admission which was significantly reduced to a mean of 8.0 (±3.35) at discharge after a mean length of stay of 18.9 (±14.1) days. a reduction in levels of medication was observed among all age groups (young-old: −74, middle-old: 75–84, oldest-old: 85+).The oldest old had the lowest mean number of prescriptions at both admission (8.1 ± 3.4) and discharge (7.6 ± 3.1). There was an overall increase in documented changes to prescriptions (either in number or type of medications) during this period from 61.1% to 80.1%. CONCLUSIONS: The observed reductions in levels of medication across all analyzed groups during (in-patient) stays in AG/Rs, as well as the documented changes in prescriptions, may suggest increased awareness both of over-prescription and the necessity of adequate medication levels in special geriatric care. PHP114 DERIVING DOCTORS’ PRESCRIBING PATTERNS FROM HEALTH CARE CLAIMS: AN INSTRUMENTAL VARIABLE ANALYSIS Baser O1, Wang L2, Xie L2, Dysinger A2, Gust C2, Yuce H3, Baser E2 1 STATinMED Research/University of Michigan, Ann Arbor, MI, USA; 2STATinMED Research, Ann Arbor, MI, USA; 3New York City College of Technology-CUNY/STATinMED Research, New York, NY, USA OBJECTIVES: Both observed and unobserved bias in observational claims data are controlled with instrumental variable approach. However, it is difficult to find a valid instrument that is strongly correlated with treatment choice but not directly correlated with outcomes. In this paper, we derived doctors’ prescribing patterns from the claims data and examined their validity. METHODS: Based on U.S. claims data, we assigned patients a unique doctor based on the greatest number of office visits during the 6 months pre-index period. For each physician, and for each 6-month calendar block of time, we estimated the physician preference as the time-varying proportion of patients treated with selective serotonin reuptake inhibitors (SSRIs) and serotonin– norepinephrine reuptake inhibitors (SNRIs) during that interval. We applied Shea’s partial R-square method, the Anderson canonical correlation, and Cragg-Donald tests to check if doctors’ prescribing patterns were weak instruments. RESULTS: Tests showed that doctors’ prescribing patterns are a valid and strong instrument for outcomes research studies. We showed that patients were more likely to be prescribed generic SSRIs relative to non-generic SSRIs if doctors’ prescribing patterns favored generic prescription (p = 0.000). Similarly, patients were less likely to be in the SSRIs group if doctors’ prescribing patterns favored SNRI prescriptions (p = 0.000). We showed that the treatment coefficient on the adherence model that does not control for the unobserved bias was 10% lower than the one that uses the instrumental variable method. CONCLUSIONS: Doctors’ prescribing patterns are important factors for prescription decisions. Any outcomes research models such as compliance, adherence or treatment effect studies should incorporate these patterns. Models that fail to control for these variables might contain omitted variable bias. PHP115 THE INCORPORATION OF ECONOMIC EVIDENCE IN THE DUTCH CLINICAL PRACTICE GUIDELINES Tan SS, Hakkaart-van Roijen L Erasmus Universiteit Rotterdam, Rotterdam, The Netherlands OBJECTIVES: Because economic evaluations aim to decrease practice variations and advocate (cost)-effective interventions, their use in the formulation of practice A426 guidelines for prevention and treatment has gained attention in many developed countries. The objective of our study was to investigate to what extent clinical practice guidelines consider cost-effectiveness and budget impact according to the most recent economic evidence. METHODS: We carried out systematic literature reviews of economic evaluations on the five most important medications by means of expenditures in the The Netherlands in 2007 (cholesterol-lowering drugs, antihypertensives, proton pump inhibitors, long-acting bronchodilators/ inhaled corticosteroids and antidepressants). Consequently, we compared the economic evidence to the recommendations of the relevant clinical practice guidelines. RESULTS: Eleven clinical practice guidelines were determined to be relevant for the medications under consideration. Although the recommendations of each of these guidelines are largely in agreement with the most recent economic evidence, 9/11 guidelines hardly considered the costeffectiveness of medications. The guidelines ‘Cardiovascular Risk Management’ (2006) and ‘Anxiety Disorders’ (2003) systematically regarded cost-effectiveness, but their recommendations are not based on the most recent economic evidence. Only the guideline ‘Cardiovascular Risk Management’ (2006) considered budget impact to take accessibility and affordability constraints into account when considering cost-effectiveness. CONCLUSIONS: Limited or no attention to economic evidence does not necessarily lead to ‘wrong’ recommendations. However, the consideration of costeffectiveness and budget impact in clinical practice guideline development is needed to increase clinician compliance, which in turn could ensure accessibility, affordability and quality of care in national health care systems. Furthermore, their consideration could harmonise national guidelines with reimbursement decisions. Engaging an economic expert in the guideline development process could contribute to the integration of the most recent economic evidence in clinical practice guidelines. PHP116 ANALYSIS OF ANTICOAGULATION BRIDGING THERAPY IN ORTHOPEDIC PATIENTS: REAL WORLD ANALYSIS Baser O1, Wang L2, Dysinger A2 1 STATinMED Research/University of Michigan, Ann Arbor, MI, USA; 2STATinMED Research, Ann Arbor, MI, USA OBJECTIVES: We assessed the real-world anticoagulation thromboprophylaxis patterns in patients undergoing major orthopedic surgery. METHODS: A retrospective study (January 1, 2005–December 31, 2007) was conducted using a large hospital database linked with outpatient claims. Patients’ demographics, and clinical and provider characteristics were compared using Chi-square testing and standardized differences. Risk-adjusted event rates were estimated using the Poisson regression model. RESULTS: The linked database is comprised of 2280 enrollees, of whom 1769 met the eligibility criteria. A total of 1552 patients received anticoagulant venous thromboembolism (VTE) prophylaxis; 264 of these patients received a combination of low molecular-weight heparin and warfarin. Of these patients, 105 (40%) were switched between the two monotherapies, whereas 159 (60%) received bridge (overlapping) prophylaxis. VTE and major bleeding events were significantly lower for bridged patients. However, duration of bridging varied widely. CONCLUSIONS: Although there is a benefit with longer duration of therapy, duration of anticoagulation bridging therapy varies widely and does not appear to be consistent with published guidelines. HEALTH CARE USE & POLICY STUDIES – Quality of Care PHP117 EMERGENCY NURSES AND PHYSICAL EXAMINATIONS—WHOSE DUTY? Fuller N, Marton-Simora J, Betlehem J, Oláh A, Boncz I, Nagy G University of Pécs, Pécs, Hungary OBJECTIVES: In the Hungarian emergency departments (EDs) it became necessary to hire nurses with wider competencies but there is no a nationally unified policy and training for them. Our aim was to determine the knowledge about the physical examinations of nurses working at EDs with the different scope of practices (triage nurses, departmental heads of nurses, general nurses). METHODS: Twenty-three Hungarian emergency departments was examined, a questionnaire was completed by the departments’ nurses (n = 301) and physicians (n = 159). SPSS 16.0 software was used for statistical analysis, khi-square test was used to determine correlations. RESULTS: A total of 71.5% of nurses have learned in some way the physical examinations wich did not present significant difference between the groups determined by nurses’ scope of practice (p = 0,228). Triage nurses and the departmental heads of nurses examine significantly more times for example the cardiovascular system (P < 0.001) and touch the stomach (p = 0.032) than departments’ general nurses. Emergency physicians would make the nurses done more diagnostical duties than they presently do in everyday practice for example examining the cardiovascular system (P < 0.001) and the chest (p = 0.001). CONCLUSIONS: Even majority of emergency nurses have learned in some way the physical examinations, they admitted a low rate of usage in the everyday practice. In contrary, physicians would place the physical examinations among the tasks of nurses possessing adequate qualification, thus extension of the competencies of nurses is necessary. 13th Euro Abstracts PHP118 PERCEPTION OF PHARMACISTS SALES REPRESENTATIVES BY PHYSICIANS Jocic DD1, Krajnovic DM2, ZIvanovic DM3, Kernican LA2 1 Pharmacy “Kumodraz II”, Belgrade, Serbia; 2Faculty of Pharmacy, Belgrade, Serbia; 3Institution of Health Care Facilities for Serbian Railway Workers, Belgrade, Serbia OBJECTIVES: Pharmacists as sales representatives influence on prescribing physicians. This research is focused on their characteristics valued by physicians to improve their communication and the effectiveness of cooperation. METHODS: The sample includes 203 physicians from Central Serbia: 78 general practitioners, 125 specialists, 72 males and 131 females. It was used the scale of attitudes on pharmacists sale representatives regarding the provision of drug information. The instrument includes 20 items and the five-point Likert type scale. RESULTS: The results show correlations between the following variables: the impression of responsible pharmacist when visiting physicians is in a positive correlation with proper terminology used by pharmacist (r = 0.650, p < 0.01), the argued presentation of preparations (r = 0.652, p < 0.01), the systematic exposure of drug materials (r = 0.626, p < 0.01), concise information (r = 0.583, p < 0.01), and visual eye contact during presentation (r = 0.648, p < 0.01); the impression of incompetence was positively correlated to fearful pharmacist performance (r = 0.654, p < 0.01) and uncertainty in an interaction with physicians (r = 0.792, p < 0.01), and negatively correlated to an argued information on preparation (r = −0.523, p < 0.01). CONCLUSIONS: The results show a correlation between pharmacists’ characteristics as sales representatives observed by physicians. Pharmacists as sales representatives proved competence and responsability only in assertive communication, as well as in concise and argued presentation with a live visual eye contact as essential for nonverbal communication. PHP119 ROLE OF QUALITY IMPROVEMENT FOR MANAGEMENT OF SURGICAL BLEED COMPLICATIONS (SBCS) Mohandas A1, Foley K2, Rupnow MF3, Nash D4, Doria C1 1 Thomas Jefferson University, Philadelphia, PA, USA; 2Thomas Jefferson University, Newtown, PA, USA; 3Ethicon, Inc, Somerville, NJ, USA; 4Jefferson School of Population Health, Philadelphia, PA, USA OBJECTIVES: Many different methods exist for managing surgical bleeding complications (SBCs) and reducing transfusions during procedures. The techniques utilized are dependent on institutional policies, resulting in highly variable clinical and economic outcomes between different institutions. Our objective was to review the current literature on the quality and costs of SBC interventions to prevent and manage transfusions during surgeries and provide recommendations on future directions for quality improvement of SBCs. METHODS: A comprehensive review was conducted using Ovid, Pubmed and Scopus databases with the following keywords: quality improvement (QI), blood loss, transfusion, hemostasis and costs. Inclusion criteria included English language, publication between 1999 and 2010, and studies where the key words were the primary endpoints. a total of 1331 abstracts were reviewed. RESULTS: A variety of blood loss prevention techniques were identified including autologous transfusion, pharmacological and non-pharmacological interventions. Studies suggest that bleed prevention (BP) algorithms incorporating a combination of interventions in the pre, peri and post-operative periods have the greatest potential to minimize transfusions. Most studies assessing the economic impact of BP interventions did not include the cost of staff time, other resources beyond blood acquisition cost and longer term complications. The exclusion of these components may underestimate the actual costs of transfusions and may have implications in assessing true cost-effectiveness of BP interventions. Consensus exists that transfusions can and should be prevented during surgery, yet there’s a lack of agreement on the optimal use of interventions for blood loss management. CONCLUSIONS: Wide agreement on the necessity of avoiding transfusion exists, yet proven methods for minimizing SBCs are underused. Multifaceted algorithms for minimizing SBCs show promising results where they have been used. Future QI projects should focus on reducing variation in practices through the development of evidence based standards and guidelines for the effective use of BP interventions. HEALTH CARE USE & POLICY STUDIES – Regulation of Health Care Sector PHP120 DESCRIPTION AND EVALUATION OF THE KNOWLEDGE OF THE BRASILIA POPULATION FOR THE POLITICS OF ECONOMIC MEDICINE REGULATION IN BRAZIL Araújo MAM, Rebelo FM, Alaver RT, Freitas PDC ANVISA, Brasilia, Brazil OBJECTIVES: The primary objective was to evaluate the knowledge of the drugstore consumers from the Pilot Plan Plano Piloto and workers of ANVISA, for the regulation of prices of medicines. The specific objectives were to assess the knowledge of the populations on PMC, different price of generic medicines, estimate monthly expenses of the two samples and a preliminary comparative analysis between the two samples. METHODS: The study developed is a cross descriptive and analytical study. Based in the application of different questionnaires for to assess the knowledge about regulation of prices of medicines and consumption habits of consumers interviewed in the door of the drugstore and workers of ANVISA. The research was conducted by sampling 13th Euro Abstracts according to the location of the drugstore and in accordance with the area of capacity in the workers of ANVISA. RESULTS: There were interviewed 105 consumers in drugstores and 140 workers of ANVISA. When asked about the existence of market regulation of drugs, 51.43% and 87.14% of drug users and employees of ANVISA, respectively, said there is regulation of these products. About the PMC, the knowledge was 20.00% of respondents in pharmacies and 57.86% of the employees of ANVISA. Despite of the relative knowledge of PMC, only 01 of the respondents knew where consumers search the PMC, and the employees of ANVISA, 29 knew the sources of research. Although the poor knowledge on the regulation of prices, 84.76% of consumers interviewed considered this activity exerted by the government as important. CONCLUSIONS: Although the population consider important that the prices of medicines are regulated by the government, the knowledge of the regulation is small, even on the tools of consumer protection. PHP121 INSPECTION OF THE PHARMACEUTICAL COMPANIES IN IRAN BY INSPECTION SOFTWARE Radmanesh R1, Nikfar S2 1 Tehran University of Medical Sciences, Faculty of Pharmacy, Tehran, Iran; 2Tehran University of Medical Sciences, Faculty of Pharmacy and Food & Drug Laboratory Research Center, Iran MOH, Tehran, Iran OBJECTIVES: This study was investigating of costs and consequences of two manual and computerized systems for management of information during 2008–2009 for inspection of pharmaceutical industries in Iranian Drug Regulatory Affairs. METHODS: To compute costs of processes following items had been considered: Cost of filling and archiving, data collecting (person-hour), reporting (person-hour), transport, software, hardware (main server computer, pocket computer), stationeries. To evaluate the efficacy following outputs and outcomes was considered: Time of information recovering, ability of ranking, preventing of data missing, capacity building and tracking and monitoring. RESULTS: The cost of running the new system is 35,000 US Dollars. Cost of education in new method and conventional are 5000 and 1000 US Dollars respectively. Cost of inspection in computerized management of information system (MIS) is decreased to 250 US Dollars from 600 US Dollars for each inspection process. Alongside capacity of system is increased by arming to fast processing method, time of information recovering diminished to 1 working day instead of 3 days, preventing of data missing from 80% to 95% in new one, tracking and monitoring took 7 working days to applied, but new MIS changed that to 1 working day. Ranking of pharmaceutical industries is now available for Iranian Drug Regulatory Affairs after establishment of new inspection system by computer-based MIS. CONCLUSIONS: It seems that beside overhead cost of new computerized system that is more than conventional method; considering capacity building; due to decreasing the cost of inspection and increasing of outputs and outcomes indicators, the new system is more efficient. PHP122 COMPETITIVENESS OF HUNGARY IN INTERNATIONAL CLINICAL TRIALS Kalo Z1, Kovacs G2, Nagyjanosi L2, Nagyistok S2 1 Eötvös Loránd University, Budapest, Hungary, 2Syreon Research Institute, Budapest, Hungary OBJECTIVES: Patients, health service providers, payers and the society also gain from intensive clinical trial participation, however the majority of benefits are intangible. According to a recent survey Hungary generates 0.15% of the GDP from clinical trials and related activities, therefore the economic importance of this area is acknowledged by the Hungarian government. The clinical trial activity is traditionally strong in the country, however the growth rate of clinical trials is lower than in other CentralEastern European countries. Our objective was to explore how Hungary can improve its competitiveness in attracting clinical trials. METHODS: We conducted a literature review, searched for publicly available documents and interviewed key stakeholders in Hungary to explore potential fields for intervention. RESULTS: We identified seven key target areas for intervention to improve the competitiveness of Hungary in clinical trials: the simplification of legal framework for clinical trial related activities, development of infrastructure at main potential sites, organizational development with special focus on SMOs, the simplification of rules and processes for financing clinical trials, investment into developing databases to support the set-up of clinical trials, and finally the marketing promotion of Hungary and its sites to sponsors of clinical trials. CONCLUSIONS: The area of international clinical trials is a very competitive market. Hungary can strengthen its market position, if legislators, competent authorities and management teams of investigational sites—by acknowledging the professional and financial benefits of these studies—support the successful implementation of clinical trials in coordinated actions. PHP123 PHARMACY NETWORK DEVELOPMENT DURING ECONOMIC TRANSITION IN POLAND Tuszyska M, Hermanowski TR, Feliksiak H Medical University of Warsaw, Warsaw, Poland OBJECTIVES: To analyze the effects of liberalization of regulations, concerning establishment of pharmacies and pharmacy points in Poland after 1989. METHODS: The data on the number of pharmacies and pharmacy points, used for the analysis were collected by mail survey completed by all regional pharmaceutical supervision authorities. Population data were sourced from the statistical yearbook of Poland. RESULTS: A427 During economic transition Poland was one of few European Union countries, where the number of pharmacies was not regulated within a given area. Therefore, their numbers increased rapidly, reaching 12.153 private pharmacies and 1.397 private pharmacy points in 2009. This means that in 2009 one pharmacy served 3.127 people. Before privatization process started in 1989, all retail pharmacies were managed by a state-owned enterprise “Cefarm”. Since then, the regulations have changed and the number of pharmacies started to grow. The main reasons of this trend was a regulation stipulating that permission to establish a pharmacy can be granted to any private or legal person, disregarding professional education of that person. Pharmacy owner was only required to employ a qualified pharmacist, responsible for managing the pharmacy. Another major institutional change was introduced by the Pharmaceutical Law, voted October 10, 1991, which authorized establishment of pharmacy points. In spite of the general trend, in some poor provinces of Poland the number of pharmacies decreased. June 1, 2010, European Court of Justice ruled that the right to own and operate a pharmacy may be reserved exclusively for pharmacists and that demographic or geographic criteria may be used in the process of issuing permits to operate pharmacies. CONCLUSIONS: Currently, the Polish Ministry of Health has to propose amendments to the Pharmaceutical Law, implementing the above mentioned criteria. PHP124 DESCRIPTION OF THE PRICING AND REIMBURSEMENT SYSTEM IN THE CZECH REPUBLIC Petrikova A1, Dolezal T2, Lamka J3, Klimes J4 1 VFU Brno, Brno, Czech Republic; 2Institute for Health Economics and Technology Assessment, Prague, Czech Republic; 3Charles University, Prague, Czech Republic; 4Charles University, Hradec Kralove, Czech Republic OBJECTIVES: To investigate changes realized in the Czech pricing and reimbursement system since January 2008 and to describe the current development of this system. METHODS: We described the legal framework defining the Czech system (Act No. 48/1997 Coll. as amended by the Act No. 261/2007 Coll. on Public Health Insurance) as well as evaluation of the current system from the literature and reports provided by the State Institute for Drug Control (SUKL). RESULTS: In order to increase transparency according to EU Transparency Directive and to set fairly-defined deadlines the competencies for the pricing and reimbursement system were merged from Ministries of Finances and Ministry of Health (MoH) under one responsible institution— SUKL. The pricing rules were fully changed—the maximum price is set based as the mean value of all available ex-factory prices in the reference countries (Estonia, France, Italy, Lithuania, Hungary, Portugal, Greece and Spain). The reference reimbursement system contains 251 reference groups of therapeutically interchangeable products with similar clinical efficacy and safety (should be updated annually by MoH). Medicinal products included in one reference group have the main common therapeutic indication in the same reimbursement level which is calculated on the basis of retail prices in all EU countries. The cheapest price for equipotent dose is chosen and re-counted according to local pharmacy and wholesaler margins and value added taxes. In compliance with new legislation the pharmacoeconomic criteria (cost-effectiveness evaluation and budget impact analysis) should be taken into account. There is possibility of extra bonus of basic reimbursement for better efficacy, safety, dosing schedule, compliance, etc. CONCLUSIONS: The system has gone through dramatic changes in last two years and some aspects are still facing challenges. Although the new system should reassessed all medicines covered in the country till 2008, currently there are only 20% revised (April 2010). HEALTH CARE USE & POLICY STUDIES – Risk-Sharing/Performance-Based Scheme/Agreements PHP125 TRENDS IN UK-BASED PATIENT ACCESS SCHEMES: FINANCIAL-BASED VERSUS OUTCOMES-BASED AGREEMENTS Woods RM, Johnson KI Complete Market Access, Macclesfield, Cheshire, UK OBJECTIVES: Whilst patient access schemes (PAS) are not a new concept, they are clearly receiving increasing attention. The first notable PAS, devised to improve access to beta interferons for multiple sclerosis, was an outcomes-based scheme designed to overcome uncertainty in long-term clinical and cost-effectiveness. It is the nature of the uncertainty that drives the design of PAS, but is there a trend towards which schemes are more popular? We analyzed the design of published PAS employed in the UK in order to determine if outcomes-based or financial-based schemes predominate. METHODS: Published PAS were identified from health technology assessment websites (e.g. the National Institute for Health and Clinical Excellence and Scottish Medicines Consortium), literature searching of ISPOR conference abstracts, and searching of ‘patient access scheme’ or ‘risk-sharing scheme,’ using internet search engines. PAS identified were categorised as financial-based (price- or volume-based agreements) or outcomes-based schemes. Desk research was the performed to identify the preference for each type of scheme, in terms of uptake by UK Primary Care Trusts. RESULTS: Seventeen published PAS were identified from the literature search, from 2002 until 2010. Categorization of PAS as financial-based versus outcomes-based showed that schemes were balanced, but favoured financial-based schemes (59% versus 41%, respectively). The uptake of financial-based schemes was found to be higher than outcomes-based schemes due to administrative burden posed by schemes which rely A428 on the measurement and reporting of clinical outcomes. CONCLUSIONS: PAS were pioneered by the beta interferon outcomes-based scheme for multiple sclerosis, 2002. Since then, published UK-based schemes favour price or volume-based schemes, a trend illustrated by two of the three newest schemes published in 2010. The simplicity of financial-based agreements, combined with poor health care system uptake of outcomes-based schemes and the demand for value-based pricing, suggests that this trend is set to continue. PHP126 PATIENT ACCESS SCHEMES (PAS) IN THE UK COMING OF AGE: WHAT IMPACT WILL THEY HAVE ON OTHER EU COUNTRIES LIKE ITALY? Kirpekar S, Mallinson M, White R, Gannedahl A Double Helix Consulting Group, London, UK OBJECTIVES: Although patient access schemes (PAS) have historically been implemented for high-cost oncology drugs, recent schemes for chronic diseases like rheumatoid arthritis (RA) have been seen in the UK. This study tries to understand how the shift in PAS from being accepted by payers only for short-term oncology drugs to those for chronic diseases like RA in the UK will influence the situation in Italy, where PAS are known to be widespread. METHODS: This study used qualitative telephone interviews to analyze trends in the UK (n = 7) and Italy (n = 7). Interviews were conducted in tertiary hospitals as well as local, regional and national level reimbursement authorities involving financial and clinical stakeholders, and key individuals in the implementation of the scheme. Importance of a number of variables affecting new PAS was ranked. RESULTS: Budget-holders in both markets were seen to be sceptical about the impact on long-term budgets due to the move towards PAS for chronic conditions. Of the 14 stakeholders interviewed, 9 said that such schemes in chronic conditions might help companies access markets with lesser clinical evidence on the basis of the class-effect of the drugs and the risk-sharing nature of the schemes. Monitoring of the outcomes on implementation was thought to be crucial. CONCLUSIONS: In Italy, increase in PAS for chronic diseases will help drug companies bring drugs to the market earlier. On the other hand, the payers will see this as an increased burden on their budgets as it will mean funding longer term treatment. Also, due to recent issues with monitoring of outcomes in PAS in the UK, their future needs reconsideration. This move of PAS towards chronic conditions is expected to increase the impact that post-marketing monitoring will have on market access for expensive ‘me-too’ drugs in the EU. PHP127 ADOPTING A FINANCIAL RISK-SHARING SCHEME FOR NEW TECHNOLOGIES ADDED TO THE NATIONAL LIST OF HEALTH SERVICES IN ISRAEL: STAKEHOLDERS’ STATED INCENTIVES AND DISINCENTIVES Hammerman A, Feder-Bubis P, Greenberg D Ben Gurion University of the Negev, Beer-Sheva, Israel OBJECTIVES: To explore major stakeholders’ incentives and disincentives to adopt a financial risk-sharing mechanism regarding budget-impact estimates of adopting new technologies in the Israeli National List of Health Services (NLHS). According to the proposed scheme, HMOs will be partially compensated by the pharmaceutical and medical device industry if actual use of a technology is substantially higher than what was projected and allocated. On the other hand, HMOs will partially refund the government for budgets allocated to specific technologies that were not fully used. These unused budgets will be used for adopting other technologies in subsequent years. METHODS: Using a semi-structured protocol, we interviewed major stakeholders involved in the process of updating the NLHS (N = 31). Interviewees included government officials, senior managers in the country’s four HMOs, pharmaceutical industry executives, and health economists. We inquired into the interviewee’s view towards our proposed risk-sharing mechanism, and their opinion on the other stakeholders’ incentives to accept or object the proposed scheme. RESULTS: Our interviews revealed a wide range of incentives, disincentives, and barriers for adopting the risk-sharing mechanism. There was no consensus on what would be the different stakeholders’ incentives and disincentives for adopting the proposed mechanism, even within the various stakeholders groups themselves. Most interviewees from the HMOs and the pharmaceutical industry supported the proposed risk-sharing agreement. Among government officials, the Ministry of Finance decision-makers tended to object to the proposed mechanism, while Ministry of Health executives usually supported the scheme, but believed that the pharmaceutical industry will not support this risksharing agreement. CONCLUSIONS: Since the success of implementing a risk-sharing mechanism depends mainly on its perception as a win-win situation for all stakeholders, we recommend that decision-makers consider the different incentives and disincentives exposed in our interviews, when implementing such a mechanism. HEALTH CARE USE & POLICY STUDIES – Beyond Drug Interventions PHP128 THE USE OF HEALTH PROMOTION INITIATIVES BY MUNICIPALITIES IN BELGIUM. A RETROSPECTIVE OBSERVATIONAL STUDY Verhaeghe N, Barbe T, Annemans L Ghent University, Ghent, Belgium OBJECTIVES: Growing attention is currently given to preventive health care. Prevention can decline the appearance of several diseases and as a consequence lead to a 13th Euro Abstracts decrease of health care expenses. The aim of this study was to evaluate if municipalities in Belgium currently are using intervention strategies for optimizing the health behavior of citizens. Special attention was given to what extent certain high-risk groups are reached. METHODS: The design consisted of a retrospective observational design. The general services or—if existing—the sports services of 570 Belgian cities (308 in the Flanders and 262 in the Walloon region) were invited by e-mail to participate in the study. After the first mailing, another two mailings were performed. The data were collected using an online questionnaire with closed questions. PASW Statistics 18 was used for statistical analysis. RESULTS: The response rate in the Flemish region and Walloon region was respectively 57.47% and 16.41%. In Flanders, 94.9% of respondents reported that they organize health promotion initiatives, while in Wallonia this percentage was 65.1%. Most common organized initiatives in the Flanders region were sport (74.4%) and social welfare (46.4%) initiatives. In Wallonia most common initiatives were prevention of disease/vaccination (45.3%) and sport (44%). Both in Flanders and Wallonia, initiatives concerning mental health were little organized. In general, the initiatives aimed at certain high-risk groups such as persons with mental health problems, older and disabled persons were scarce. On the level of municipal policy lack of appropriate financing was reported as the most common barrier for not organizing preventive strategies. CONCLUSIONS: On the level of municipal policy reinforcement of the value of preventive strategies concerning health promotion with appropriate financing is required. When organizing preventive initiatives, special attention to reach certain high-risk groups will be needed. PHP129 ATTITUDES OF HUNGARIAN POPULATION TOWARD CO-PAYMENTS 1 2 1 2 Baji P , Pavlova M , Gulacsi L , Groot W 1 Corvinus University of Budapest, Budapest, Hungary; 2University of Maastricht, Maastricht, The Netherlands OBJECTIVES: The issue of the introduction of co-payments is a great policy challenge in most of the Central-Eastern European countries. This is also the case in Hungary, where visit fee was introduced for health care services in 2007, and abolished one year later as a result of a referendum. The aim of our study is to identify different types of attitudes towards patient payments, and answer why visit fee was so unpopular among Hungarian population. METHODS: 8 focus-group discussions with health care consumers and physicians and 7 in-depth interviews with policy makers were conducted in Hungary during the summer 2009 on the attitude of patient payments in health care. RESULTS: Based on the transcripts and questionnaires filled in by all respondents during the focus group discussions and interviews three different groups of attitude were identified. The group of “Supporters” support the introduction of patient payments with the aim of controlling the unnecessary use of services. The group of “Undecided” concern patient payments as an opportunity to provide additional resource for health care system by paying for “extra-better-quality services”. “Sceptics” strongly refuse the idea of patient payments mainly referring to ethical issues. Consumers mainly belong to the group of “Undecided”, while one part of the physicians belongs to “Supporters”, the other part to “Sceptics”. Policy makers are all belong to the group of “Supporters”. CONCLUSIONS: Before the implementation of patient payments, mapping of population’s attitude is inevitable. In Hungary the failure of the introduction of visit fee can be explained by different expectations of health care consumers. They are not against to pay for health care services, but expecting better quality of provided services in return. PHP130 TRENDS IN COST CONTAINMENT MEASURES Mukku S, Klein Lankhorst E, Edathodu AS Double Helix Consulting Group, London, UK OBJECTIVES: In the current economic climate health authorities are finding ways to control spending. a popular measure is the introduction of cost containment strategies for pharmaceuticals to lessen the burden of cost of drugs. The objective of this research was to analyze the importance of different cost containment measures that payers use in selected markets. The research extends to analyze the gaps between the policies on paper and the extent these are interpreted and executed in practice. METHODS: Interviews were conducted with payers at the national, regional and local level in selected European and non-European markets. RESULTS: All countries utilise different measures to control spending on pharmaceuticals. The measures that are used by markets include: internal therapeutic referencing, international price referencing, generic substitution, risk-sharing agreements, budget caps, profit caps, index pricing, price cuts, rebates and price volume agreements. The critical finding from the research was that although combinations of measures are included in the national and regional policies, in reality their interpretation and execution varies substantially. For example, Spain uses internal referencing and price referencing at the national level, regions apply price discounts and rebates. Similarly, in the The Netherlands a preferred drug policy is introduced, which means that health insurers can now choose a preferred drug for reimbursement, which is usually the cheapest option of the reference basket. If the patient does not want this product, they have to pay the full price of the other product. Another European country which has become the recent focus is Germany; where price regulation will become dominant in the near future. There are uncertainties on achievable prices of drugs, one reason being there is lack of sophisticated systems in place that can monitor these measures. CONCLUSIONS: There are currently many developments in the area of cost containment of pharmaceuticals which will have a profound effect on the pharmaceutical industry. 13th Euro Abstracts PHP131 THE USE OF ECONOMIC EVALUATION IN A MEDICAL DEVICE COMMISION OF AN ITALIAN TEACHING HOSPITAL Alello A, Bassotto F, Filippi C, Fratucello A, Ghirlanda G, Marini P, Scroccaro G, Fantelli V Azienda Ospedaliera Universitaria Integrata di Verona, Verona, Italy OBJECTIVES: In order to control Medical Devices (MDs) purchasing process and to rationalize the MDs evaluation process, the Verona Hospitals (2 accounting for 1.700 beds), have set up a Medical Devices Committee (MDC). MDC members, who have expertise in a broad range of disciplines, evaluate MDs according to Health Technology Assessment (HTA) criteria. This study is concerned with the use of economic evaluations in policy decision making of MDC. METHODS: We developed a DataBase to record all economic evidences considered by MDC from January 2009 to April 2010 for each MD evaluated: health economics analysis in literature (cost-effectiveness, cost-to-cost etc), cost information (price, costs impact etc) and Disease Related Group (DRG) simulation. We verified also the presence and costs of current alternative treatments. RESULTS: Among 37 MDs evaluated 30 technologies had at least a current alternative treatment in Hospital formulary (81.1%); MDC always considered at least the price of MDs, costs impact of introducing a new MD and costs when compared with alternative/current treatments (100%). For 23 MDs the management office individuated the DRG value (62.2%) and for 12 of them (32.4%) verified whether DRG covered direct costs. Only for 4 MDs (10.8%) was possible finding at least an economic study in literature (2 cost-effectiveness analysis and 2 cost-to-cost analysis). CONCLUSIONS: At the moment the decisions taken by the MDC mainly focused on the clinical benefits, on the comparison of costs vs. alternative treatments and on the incidence of the DM cost on the DRG tariff. The economic evidences seldom influence the opinion of MDC, most of all because the clinical value of DM is not always supported by economic evidences. Recently, technology application form has been modified including questions about the potential savings derived by the new technology, economic studies (also not published) and Budget Impact analysis. PHP132 VALUE OF MULTI-CRITERIA DECISION ANALYSIS IN EARLY ASSESSMENT OF MEDICAL DIAGNOSTIC DEVICES Ijzerman MJ, Hummel J University Twente, Enschede, The Netherlands OBJECTIVES: Multicriteria decision analytic (MCDA) techniques are a powerful tool in evaluating health care interventions where multiple, often competing, factors need to be considered. The analytic hierarchy process (AHP) is one such technique. We have applied AHP to evaluate medical diagnostic technologies as to support product development and market access. The main objective of this study is to show feasibility of AHP in terms of applicability and outcome. We therefore have studied the expected performance of the Photoacoustic Mammoscope (PAM), a new imaging device for detection of breast cancer based on photoacoustic imaging. METHODS: Before starting the study we defined different diagnostic and screening scenarios for the PAM including criteria that are important in the acceptance of the device. We finally defined four strategies including mammography, ultrasound and MRI imaging. The Analytic Hierarchy Process (AHP) analysis was then used to identify the relative importance of evaluation criteria, and to identify the relative performance of four different breast cancer imaging techniques. RESULTS: According to the expert panel the most important criterion in the design of a diagnostic breast imaging device is sensitivity (overall weight was 54%). Factors that mainly determine the performance on sensitivity are the visualization of mass margins, mass shape and vascularization. Second most important criterion was safety. CONCLUSIONS: From an early assessment of photoacoustic imaging it may be concluded that the potential clinical performance justifies further development and implementation of PAM. However, further research on the use of AHP in medical product development is required. We present and discuss some ideas to further integrate clinical evidence into MCDA using Bayesian techniques as well as sensitivity analyses on the model inputs. PHP133 PERSONALIZED PHARMACOTHERAPY AND ANALYSIS OF EXTEMPORANEOUS DOSAGE FORMS FOR CHILDREN IN UKRAINE Maynych Y, Zalis’ka O Danylo Halytsky Lviv National Medical University, Lviv, Ukraine OBJECTIVES: Pharmacotherapy of children requires the use of drugs for individual preparation of personalized treatment. Medications extemporal manufacturing approach allows for Personalized Medicine. The use of extemporaneous preparations to precisely match the appropriate concentration of drug forms, volume and available dosage forms for komplayensu, extemporaneous medicines do not contain stabilizers, preservatives, it is important for children with allergic diseases. METHODS: The ABC—analysis was conducted of 3525 prescription extemporaneous preparations which were made in 8 pharmacies specialized of Western Ukraine regions for 2007– 2009. RESULTS: The total list of pediatric formulations consisted of 106 items. Determined that the practical range of medicines includes 7 extemporaneous dosage forms, including solutions of 45% solutions, ointments 23% and 12% suspensions et al. Revealed that the production of extemporaneous preparations seasonal influences. Increasing production of liquids, powders at the time of respiratory exacerbation in children. Distinguished set of extemporaneous preparations, which are often assigned practical pediatrician in Ukraine: protargol drops—when rhinitis, cough mixture, powder eufilinu with glucose—with bronchitis, Foenuculi water—with flatulence, pepsin solution with hydrochloric acid—in violation of appetite, anestezynum suspen- A429 sion; dexamethasone ointment, betamethasone ointment—for allergies vitamin powder. Determined extemporaneous medicines for children, which have no analogues in industrial production—a solution of pepsin with hydrochloric acid, dexamethasone ointment, paste Lassara with oil Plum; diazolinum powder with calcium gluconate and glucose—from allergies, suppositories with cocoa butter and levomitsetinum—with diarrhea, Foenuculi water—with flatulence in newborns and infants. CONCLUSIONS: Using drugs extemporal medicines, namely those that have no industrial counterparts, to ensure individualized pharmacotherapy children pick up the dose, volume of product, and prevent undesirable side reactions, especially for treatment of chronic diseases in children with allergic manifestations and provide a personalized therapy in pediatric practice. PHP134 IMPACT OF BLEEDING-RELATED CONSEQUENCES ON LENGTH OF STAY (LOS) AND HOSPITAL COSTS IN PATIENTS UNDERGOING SURGERY IN FRANCE Lafuma A1, Ye X2, Torreton E1, Bastide P3, Arnaud A3 1 CEMKA-EVAL, Bourg la Reine, France; 2Ethicon, a Johnson & Johnson Company, Somerville, NJ, USA; 3Ethicon, Inc, Issy les Moulineaux, France OBJECTIVES: The objectives of this study were to estimate the incidence, costs and length of stay (LOS) of bleeding-related consequences in various surgical procedures in France. METHODS: A retrospective analysis was conducted using the national Diagnosis Related Group (DRG) database (PMSI). Patients with surgical procedures in 2008 were identified based on DRGs. For each DRG, the rate of bleeding-related consequences was estimated. For DRGs with an incidence of bleeding-related consequences over 10%, age and gender-adjusted hospital LOS and costs were examined among patients with (WB) and without bleeding related consequences (WoB). The rates of hospitalizations exceeding the expected LOS (derived from PMSI) between WB and WoB surgeries were compared using multivariate logistic regression. RESULTS: Of the 88 different DRGs in French database, 24 (a total of 321,657 hospitalizations) had an incidence of bleeding-related consequences over 10%, ranging from 10.3% to 25.3%. DRGs with higher rates were transplantations, cardiac and major orthopedic surgery, vascular and solid organ surgery. Age and gender adjusted mean LOS for WB patients was 3.38 days longer (+26.5%) compared with WoB patients (P < 0.001). Average adjusted costs were estimated at c12,087 per WB stays versus c10,086 per WoB stays, corresponding to a 19.9% increase (P < 0.001). The rate of hospitalizations exceeding the expected LOS was respectively 42.3 % and 37.0% for WB and WoB hospitalizations (P < 0.001). WB hospitalizations were approximately 10% more likely to exceed expected LOS compared to WoB (OR = 1.09, 95% CI: 1.07–1.11, P < 0.001), after adjusting for patient characteristics. CONCLUSIONS: This study characterizes the increased hospital LOS and cost among patients with bleeding-related consequences in France. In particular, the excess LOS than expected DRG LOS presents significant financial burden related to bleeding consequences for hospitals. Awareness of potential clinical and economic impacts of bleeding-related consequences provides important framework for understanding the value of blood-conservation strategies. PHP135 THE EXAMINATION OF THE CHRONIC STRESS 1 1 1 1 2 1 1 Müller Á , Józsa R , Gál N , Betlehem J , Sándor J , Roznár J , Boncz I , Kriszbacher I1, Sebestyén A3, Oláh A1 1 University of Pécs, Pécs, Hungary; 2University of Debrecen, Debrecen, Hungary; 3National Health Insurance Fund Administration, Pécs, Hungary OBJECTIVES: We examined the effects of chronic stress with modelling different shift work schedules in animal experimental model. METHODS: Internationally most frequently used night shift schedules were applied in three groups of animals (4DL/3LD; 8DL/6LD; 2DL/2LD) beside a group kept in normal LD cycle (12 hour light-L/12 hour dark-D = control group). Groups were divided in two parts, one of them were exposed to CMS. Degree of anxiety was evaluated in light-dark box. Differences between groups according to variables (sex, lighting regimens and CMS) and time spent in light, number of changing compartments and latency of changing dark to the light section were examined with multiple one-way analysis of variance. Plasma concentrations of corticosterone and testosterone were measured after 4 weeks of exposure to stress procedures, concentrations were determined by radioimmunoassay. RESULTS: In groups kept under different lighting schedules compared to control group, significant differences were found: animals kept in 4DL/3LD (p = 0.05, p = 0.079 /nearly significant/ p = 0.011) and 2DL/2LD (p = 0.025, p = 0.001, p = 0.045) schedules spent less time in light, latency increased, while no statistically significant differences were found in 8DL/6LD group in any of the variables (p = 0.113, p = 0.118, p = 0.45). We found that groups kept in different lighting schedules and exposed to chronic stress kept their circadian rhythm (corticosterone: p = 0,095— border line statistical significance; testosterone: p = 0.004), while groups kept in different lighting schedules but not exposed to chronic stress lost circadian rhythms (corticosterone: p = 0.071 for 12-hour component; testosterone: p > 0.20). CONCLUSIONS: According to light-dark box test and hormonal examinations it seems, that shifts different than normal light/dark regimen lead to anxiety also without CMS, and may have been harder load than the CMS procedure. Group 8DL/6LD shows the closest approximations of parameters observed in LD group suggesting that this schedule applied in shift work is less harmful to health, and provides the best way of regeneration. A430 INFECTION – Clinical Outcomes Studies PIN1 COST AND QUALITY OF LIFE ISSUES ASSOCIATED WITH PROTEASE INHIBITOR-BASED COMBINATION THERAPY TO TREAT HEPATITIS C Carter J1, Gao X1, Stephens JM1, Rustgi V2, Haider S3 1 Pharmerit North America, LLC, Bethesda, MD, USA; 2Georgetown University Medical Center, Fairfax, VA, USA; 3Pfizer Inc., New London, CT, USA OBJECTIVES: To systematically analyze the literature to assess the economic and health-related quality of life (HRQoL) impact of adverse events (AEs) related to the addition of protease inhibitors (PI) to standard of care (SOC) for the treatment of hepatitis C (HCV). METHODS: A literature search (2000-Present) was conducted to identify and analyze clinical trials for PI triple therapy (PITT = PI + SOC) and SOC (PEGIFN/RBV for 48 weeks). HRQoL and safety data were synthesized by study design, sample characteristics, and AEs. Economic and resource use data were synthesized in an economic analysis of AEs in PITT vs. SOC. Costs (2009) were derived from published literature. RESULTS: Twenty-three SOC and 7 PITT trials were identified. Statistically significant (p = or <0.05) changes from baseline were most often seen in trials of SOC in the following domains: vitality, depression, physical limitations, and fatigue. The following 4 PITT-related AEs could be linked to HRQoL domains: anemia and depression were linked to fatigue and vitality, and headache and rash were linked to physical limitations. In terms of economic impacts, the costs to manage a PITT-related episode of anemia, depression, diarrhea, and rash were $4825, $2837, $566, and $633, respectively. The average AE cost in PITT ranged from $1732 to $3578. Corresponding costs in SOC ranged from $1608 to $2229. The treatment costs of PITT-related AEs were 30% (range 8–60%) higher than the cost of treating SOC-related AEs. CONCLUSIONS: The costs to manage PITT-related AEs appear higher than the costs of SOC-related AEs. Since PITT is associated with higher AE rates, it can also be expected to result in worse HRQoL. Gaps in symptom burden assessment with existing instruments also exist. Future studies should incorporate the economic burden of AEs and the appropriate use of HCV-validated instruments to capture potential HRQoL differences among treatment strategies. PIN2 ESTIMATED IMPACT OF SUSTAINED VIROLOGICAL RESPONSE (SVR) ON LIFE EXPECTANCY, QUALITY-ADJUSTED LIFE-YEARS (QALYS) AND LIFETIME COSTS IN CHRONIC HEPATITIS C (CHC) PATIENTS Cure S1, Bianic F1, Cawston H1, Dartois L1, Zhang H2 1 i3 Innovus, Uxbridge, Middlesex, UK; 2Johnson & Johnson Pharmaceutical Services, LLC, Raritan, NJ, USA OBJECTIVES: The World Health Organization considers hepatitis C a global burden. CHC is a curable disease since patients achieving SVR can remain virus-free. We evaluated CHC treatment for efficacy (SVR) and long-term effects on reduction of disease progression to advanced stages, improvement in quality of life, and reduction of health care costs. The objective was to estimate the burden of disease and impact of SVR in CHC patients on lifetime life-years (LY), QALYs and costs using a Markov model and published literature. METHODS: A systematic literature review identified publications reporting the cost-effectiveness of antiviral therapies in CHC (November 2009). PubMed, CRD and Health Technology Assessment reports were searched. We selected all cost-effectiveness (CE) studies with a similar Markov structure including eight health states (SVR, mild/moderate CHC, compensated and decompensated cirrhosis, hepatocellular carcinoma [HCC], liver transplant and death). CE inputs were used to populate a Markov model to estimate average lifetime benefits and costs associated with SVR and non-SVR. Age of entry in the model was 40 years. Costs were updated to 2008 Great Britain Pounds (GBP). Discount rates for costs and outcomes were 3.5%. RESULTS: 893 unique references were retrieved and 14 articles met the inclusion criteria. Compared to non-SVR patients, the SVR group was consistently associated with more LYs (18.33, range [18.05;18.53] versus 17.49, range [16.17;18.51]) and more QALYs (15.81, range [14.8;17.87] versus 13.25, range [11.9;14.03]). Lifetime costs associated with non-SVR patients (£20,406, range [£7,186;£41,383]) were also consistently higher than with SVR patients (£11,156, range [£5,449;£14,582]). CONCLUSIONS: Using this model, the SVR group was consistently associated with longer life expectancy and more QALYs. Complications avoided by reaching SVR were associated with reduced lifetime CHC costs. It will be important to account for these correlations with lifetime benefits and costs when the values of an antiviral treatment in CHC are evaluated. PIN3 PRESCRIBER INFLUENCE ON THE PREVALENCE OF POTENTIAL ANTIRETROVIRAL DRUG-DRUG INTERACTIONS ON PRESCRIPTIONS IN SOUTH AFRICA Katende-Kyenda NL1, Lubbe MS2, Serfontein JHP2, Truter I3 1 Walter Sisulu University, Mthatha, Eastern Cape, South Africa; 2North-West University, Potchefstroom, South Africa; 3Nelson Mandela Metropolitan University (NMMU), Port Elizabeth, Eastern Cape, South Africa OBJECTIVES: The purpose was to determine the prevalence of potential drug-drug interactions (DDIs) between antiretroviral (ARV) drugs on prescriptions prescribed by general practitioners and specialists in South Africa and the evaluation of the prescribed daily doses (PDDs) of the interacting drugs. METHODS: A quantitative, retrospective drug utilization study was performed on 49,995, 81,096 and 88988 ARV prescriptions claimed through a South African pharmacy benefit management 13th Euro Abstracts company during 2005 to 2007. Potential DDIs between ARVs were identified and classified according to a clinical significance rating. The clinical significance ratings of potential DDIs are described in three degrees of severity, identified as major, moderate and minor as described by Tatro. (RSA Rand(R)/$US = 6.8595 on 31 December 2007). RESULTS: Antiretroviral drugs represented 0.91% of all drugs (N = 59 971 226) claimed during the three years at a total cost of N = R 5 758 783 544 (1.92%). The average cost of antiretroviral prescriptions decreased with 4.19% from 2005 (R524.40 ± R178.79) to 2007 (R502.41 ± R161.19). ARV prescriptions prescribed by general practitioners with potential DDIs and PDDs not according to recommended ARV dosing increased dramatically from 12.33% in 2005 to 24.26% in 2007. Those prescribed by specialists increased from 15.46% in 2005 to 35.30% in 2006 and decreased to 33.16% in 2007. The highest percentage of ARV prescriptions with potential DDIs and PDDs not according to the recommended ARV dosing guidelines were identified in ARV regimens between lopinavir/ritonavir at PDD 1066.4 mg/264 mg and efavirenz at PDD 600 mg prescribed to patients 19 to 45 years. These regimens were mostly prescribed by general practitioners as compared to specialists. CONCLUSIONS: There is need for more education of prescribers to be aware of the potential medication-prescribing errors associated with highly active antiretroviral therapy which could lead to treatment failures, development of resistance and DDIs. PIN4 IMPACT OF MMRV MASS VACCINATION WITH OR WITHOUT A CATCH UP PROGRAM ON THE INCIDENCE OF VARICELLA COMPLICATIONS IN FRANCE Ouwens M1, Littlewood K1, Sauboin C2, Tehard B3, Alain S4, Denis F5, Boelle PY6 1 Mapi Values, Houten, The Netherlands; 2GlaxoSmithKline Biologicals, Rixensart, Belgium; 3 Laboratoire GlaxoSmithKline, Marly le Roi, France; 4French National Cytomegalovirus Reference Center, Limoges, France; 5CHU Dupuytren, Limoges, France; 6Université Pierre et Marie Curie, Paris, France OBJECTIVES: Varicella complications place a large burden on health care resources, however, varicella is preventable by mass vaccination with MMRV. The impact on complications of three MMRV programs were explored. METHODS: An age-structured dynamic model compared natural and breakthrough varicella following mass vaccination, to current cases. Age-specific complication rates (neurologic, cutaneous, pulmonary, other) were applied per case. MMRV replaced 80% of MMR over 1 year (basecase) compared to 100% of MMR with a catch-up in 11–13 year-olds (‘catch up’). In an ‘optimal’ scenario, MMR coverage increased from 90% to 95% (1st dose) and 60% to 90% (2nd dose), MMRV replaced 100% MMR within 1 year with a catch-up in 10 year-olds. RESULTS: MMRV decreased varicella incidence and caused an age shift, however, many post-vaccination cases were breakthrough cases, believed to be milder and require less resource use. Prior to MMRV: highest varicella incidence (per million-person-years of total population) was among 1–4 year-olds (7,482 cases), with 355 complications (age-specific complication rate: 4.75%). Basecase: highest incidence among 10–14 year-olds (907 natural, 633 breakthrough), with 54 complications (age-specific rate: 3.53%); 85% lower versus 1–4 year-olds. ‘Catch-up’: highest incidence among 15–24 year-olds (161 natural, 398 breakthrough), with 49 complications (age-specific rate: 8.82%); 86% lower versus 1–4 year-olds. ‘Optimal’: highest incidence among 15–24 year-olds (22 natural, 80 breakthrough), with 9 complications; 97.5% lower versus 1–4 year-olds. Assuming breakthrough cases have 10% of complications of natural cases, the incidence in the basecase, catch-up and optimal scenarios becomes: 34, 18 and 3 complications respectively. Despite greater risks of complications in older age groups, the reduction in varicella cases significantly reduced the incidence of complications. The ICER remained below c16,000 (direct costs). CONCLUSIONS: Mass varicella vaccination is predicted to significantly reduce varicella and complication cases. As many vaccine era cases will be breakthrough; burden to patients and health care systems may be reduced further. PIN5 HOW LONGITUDINAL PATIENT RECORDS CAN HELP PUBLIC HEALTH AUTHORITIES IN THE MANAGEMENT OF RAPIDLY GROWING EPIDEMICS? THE EXPERIENCE OF FLU A/H1N1 IN FRANCE Toussi M, Pertus D, Antonio M, Robain M CSD, Boulogne, Billancourt, France OBJECTIVES: In the context of an epidemic, public health authorities need to have real time information of the disease propagation to make appropriate and timely decisions. We present our experience in using centralized electronic patient records in the early detection and follow-up of Flu A/H1N1 epidemic in France. METHODS: We used Longitudinal Patient Data (LPD), which is a dynamic centralized database of more than 1.3 million anonymous patient records uploaded regularly by a network of 1300 representative general practitioners and pediatricians in France. The trends of seasonal influenza, Flu A/H1N1 and flu-like syndrome diagnoses are traced and compared with the reports published by Sentinel Network, a group of trained physicians set up by public authorities. RESULTS: The trends of seasonal influenza, Flu A/ H1N1, and flu-like syndrome obtained from LPD database show a very close similarity with those published by Sentinel Network (Pearson’s correlation coefficient = 0.97). The seasonal pattern of the incidence of influenza and flu-like syndrome were less similar between 2009 and 2008 (Pearson’s correlation coefficient = 0.60), which can be explained by an over declaration of all kinds of flu by the doctors during the second half of 2009. There was also a close similarity between seasonal Flu, approved Flu A/ H1N1, and flu-like syndrome trends with a peak incidence in late November 2009. CONCLUSIONS: The LPD data matched very closely the results published by the 13th Euro Abstracts official Sentinel Network. The advantage of the LPD data is that its trends are available almost in real time, whereas the active reporting of the doctors of Sentinel Network is done on a weekly basis. However, the LPD trends lack of geographical precision because of the censorship of the post codes of patient during the de-identification process. We recommend a combined use of LPD and Sentinel Networks in the follow up of future epidemics. PIN6 A MULTIVARIATE MODEL: PREDICTORS OF DISEASE PROGRESSION IN HIV/AIDS PATIENTS IN WESTERN NEW YORK Voltz C1, Castleman C2, Purdy C2, Magar R3 1 Aids Community Services of Western New York, Buffalo, NY, USA; 2AHRM Inc., Buffalo, NY, USA; 3AHRM Inc., Raleigh, NC, USA OBJECTIVES: The medical care of HIV/AIDS patients is becoming increasingly complex involving a host of issues including medication management, co-infections, cardiovascular risk factors and patient characteristics. This analysis used a set of patient characteristics and treatment patterns to explore potential predictors of HIV progression. METHODS: Aids Community Services of Western New York provides primary medical care services to approximately 600 HIV/AIDS patients yearly in WNY. A total of 1128 patients, receiving care from August 2006 through June 2010 were included in the analysis. The patient characteristics, health indicators, medication information and risk factors were used as potential predictors of disease progression. Disease progression was measured by the surrogate endpoints of CD4 counts and CD4/CD8 ratio values. a mixed model with a repeated structure was fit to the data set (with a significance level of 0.05). RESULTS: The patient population was comprised of 757 men and 371 women; the average age of the population was 44 years. The population sampled was predominantly Caucasian, African American and Hispanic (39%, 39%, and 19%). The most common HIV medications administered to these patients were ritonavir, a tanofovir/emtricitabine combination, atazanavir and a lopinavir/ritonavir combination. The most common risk factors in this population were male with male sex, heterosexual sex and IV Drug use (27%, 15%, and 9%). For the multivariate model utilizing the CD4 counts as the outcome variable, only medications and cardiovascular risk factors were significant predictors. For the multivariate model utilizing the CD4/CD8 ratio as the outcome variable, only the HIV risk factors and medications were significant predictors. Both models achieved global statistical significance. CONCLUSIONS: The management of HIV/AIDS is an increasingly complex problem which requires constant advances in both research and practice. This exploratory analysis was able to identify significant relationships between patient characteristics, treatment patterns and measures of disease progression. INFECTION – Cost Studies PIN7 BUDGET IMPACT MODEL FOR CATCH-UP PROGRAM WITH 13 VALENT PNEUMOCOCCAL CONJUGATE VACCINE IN CHILDREN UNDER 5 YEARS OLD IN THE AUTONOMOUS REGION OF MADRID (RM) Picazo J1, Gil de Miguel A2, Mendez C3, Guijarro P3, Garcia L3 1 University Complutense of Madrid, Madrid, Spain; 2Rey Juan Carlos University, Alcorcon, Madrid, Spain; 3Pfizer Spain, Alcobendas, Madrid, Spain OBJECTIVES: Madrid Health Authorities decided the introduction of Pneumococcal Conjugate Vaccine (PCV) as systematic use for infants in November 2006 (3 + 1 pattern; 2, 4, 6 + 18 months). Recently, a flawless transition to PCV13 has been recommended. This study was aimed to assess the budget impact of an additional catch-up vaccination strategy from the age of 19 to 60 months. METHODS: A oneyear budget impact model has been developed stratifying population by age groups (19 to 24, 25 to 36, 37 to 48 and 49 to 60 months) and diseases (IPD and Non-IPD). Clinical data, PCV13 serotype coverage and disease related costs were based on published data. Model was built up under regional health care system perspective and assumed 80% of coverage for PCV13. Indirect effect was not considered. All costs were expressed in c2010. RESULTS: The model predicts that the implementation of a catch-up vaccination program with PCV13 in the RM would be a cost saving measure in infant groups from 19 to 36 months due to disease burden reduction caused by the 6 PCV13 additional serotypes. Globally, 49.6% of IPD, 58.9% of OMA and 47.1% of out-patient pneumonia cases would be prevented in these groups. CONCLUSIONS: Based on this health economic evaluation, the inclusion of a catchup program with PCV13 in the RM would be an efficient measure. Model results showed that a PCV13 catch-up dose would have a high impact on pneumococcal disease prevention, avoiding its related costs. PIN8 COST-EFFECTIVENESS ANALYSIS OF PEGINTERFERON (ALFA-2B) WITH RIBAVIRIN COMPARED WITH PEGINTERFERON (ALFA-2A) WITH RIBAVIRIN FOR THE TREATMENT OF CHRONIC HEPATITIS C Omelyanovsky VV, Avksentieva MV, Krysanov I, Ivakhnenko O Research Center for Clinical and Economic Evaluation and Pharmacoeconomics, Moscow, Russia OBJECTIVES: Combination therapy with pegylated interferon with ribavirin is the standard of care for the treatment of chronic hepatitis C infection. This analysis compares the cost efficacy of of combined antiviral therapy with peginterferon alfa—2b with ribavirin compared peginterferon alfa—2Ü with ribavirin for treatment A431 at patients with a hepatitis C infection counting for 1 patient. METHODS: A decision analysis model was constructed from the viewpoint of a managed care organization to compare Peg-2b plus RBV (1.5 mcg per kilogram per week plus RBV 1000 mg per day) and Peg-2a plus RBV (180 mcg per week plus RBV 1000 mg per day). For purposes of this analysis, distribution on genotypes and treatment duration and efficacy data were obtained from the published literature. The positive predictive value was calculated for each treatment group for genotype 1, which is determined from the values for early virologic response and sustained viral response. Genotype 2 and genotype 3 were assumed to be treated for 24 weeks. RESULTS: Antiviral therapy of combined antiviral therapy with peginterferon alfa—2b with ribavirin is economically more favourable in comparison with therapy in a combination peginterferon alfa—2Ü with ribavirin and taking into account efficiency of treatment the difference indicator makes c3293 thousand on 1 patient. CONCLUSIONS: Antiviral therapy of combined antiviral therapy with peginterferon alfa—2b with ribavirin is economically more favourable in comparison with therapy in a combination peginterferon alfa—2Ü with ribavirin and taking into account efficiency of treatment the difference indicator makes c3293 thousand on 1 patient. PIN9 ESTIMATED HEALTH AND ECONOMIC IMPACT OF QUADRIVALENT HPV TYPES 6, 11, 16, 18 VACCINE IN JAPAN USING A TRANSMISSION DYNAMIC MODEL Yamabe K1, Abe M1, Singhal PK2, Kamae I3 1 Banyu Pharmaceutical Co.,Ltd, Tokyo, Japan; 2Merck & Co., Inc., West Point, PA, USA; 3Keio University Graduate School of Health Management, Fujisawa, Japan BACKGROUND: The quadrivalent (6,11,16,18) HPV vaccine has been approved in many countries for prevention of cervical cancer, vulvar/vaginal pre-cancers, and genital warts in women age 9 to 26 years. OBJECTIVES: To assess the health and economic impact of the quadrivalent (6,11,16,18) HPV vaccine in Japan. METHODS: A published mathematical model of the transmission dynamics of HPV infection and disease was adapted for Japan. Model inputs were used from Japan or the Asia/Pacific region when available; otherwise, the default values in the original model were used. Maintaining current cervical cancer screening practices in Japan, we evaluated two strategies: routine vaccination of females by age 12 (S1), and S1 combined with a temporary (5 years) female catch-up program for age 12–24 years (S2). The vaccine coverage rates were 80% for the routine and 50% for the catch-up vaccination programs. RESULTS: The most effective strategy was S2. Using this strategy over 100 years in the population of Japan, the estimated cumulative percent reduction in incident HPV 6/11/16/18-related genital warts-female, genital warts-male, cervical intraepithelial neoplasia (CIN) grade 1, CIN 2/3, and cervical cancer cases was 90% (2,113,723 cases), 86% (2,082,637 cases), 72% (263,406 cases), 71% (1,328,366 cases), and 58% (323,145 cases), respectively. The cost-effectiveness ratios were US$ 12,434 (weekly dominated), and US$ 12,058 per quality-adjusted life-years (QALY) gained for S1 and S2 compared with no vaccination, respectively. CONCLUSIONS: In Japan, vaccination of females age 12–24 years with a quadrivalent (6,11,16,18) HPV vaccine can reduce the incidence of cervical cancer, CIN, and genital warts at a cost per QALY ratio within the range typically regarded as cost-effective. PIN10 COST VERSUS DRG REVENUE IMPLICATIONS OF TREATING A PATIENT WITH VENTILATOR-ASSOCIATED PNEUMONIA (VAP) WITH DORIPENEM VERSUS IMIPENEM IN GERMANY De Cock E1, Gast C2, Berndt K3, Kubitz N3 1 United BioSource Corporation, Barcelona, Spain; 2Private Consultant, Seattle, WA, USA; 3 Janssen-Cilag GmbH, Neuss, Germany OBJECTIVES: A Phase III study of doripenem versus imipenem in VAP (DORI-10) has shown a statistically significant reduction in length of stay (LOS) and mechanical ventilation (MV) duration with doripenem. This study estimated expected hospital cost versus G-DRG payment for a patient treated with doripenem versus imipenem. METHODS: From both DORI-10 trial arms, and for nine categories of post-randomization MV duration, we obtained percentages of patients per group and median LOS by type of ward (ICU with vs. w/out MV, general ward). One clinical expert provided ICD and OPS codes for four likely patient profiles, and another expert on intensive care severity (TISS/SAPS). Using the Muenster Webgrouper, we determined G-DRG weights when simultaneously varying MV duration and intensive care severity (8–890 code). Per treatment arm and for each MV subgroup, the appropriate G-DRG weight was selected based on the selected profile, total MV duration, 8–980 code, and total LOS. Using the distribution of patients by category, a weighted G-DRG was calculated for each arm, and G-DRG weight was multiplied by the 2010 base value (c2936) to yield expected G-DRG payment. Within each arm, we calculated cost per MV category (hospital per diem plus antibiotic therapy) as well as a weighted cost. RESULTS: Expected cost for a patient receiving doripenem was c30,183 vs. c32,549 for imipenem. Expected revenue reduction ranged from c2084 to c2428 across 4 scenarios. Net budget impact of the introduction of doripenem ranged from −c85 to c282, showing that revenue reduction may be more than offset by cost reduction. CONCLUSIONS: For a patient receiving doripenem instead of imipenem, reduced hospitalization costs more than offset reduced G-DRG revenue in a wide range of cases. In addition, less time on the ventilator can improve patient outcomes (including quality-of-life), and frees up ICU bed-days, allowing more patients to be treated. A432 PIN11 COSTS OF QUANTIFERON TB-GOLD VERSUS TUBERCULIN SKIN TEST IN SPANISH HEALTH CARE WORKERS Linertová R1, Alvarez-León EE2, García-Pérez L1, Serrano-Aguilar P3 1 Fundación Canaria de Investigación y Salud (FUNCIS), Las Palmas de Gran Canaria, Spain; 2 Complejo Hospitalario Universitario Insular Materno-Infantil, Las Palmas de Gran Canaria, Spain; 3Servicio Canario de la Salud, Santa Cruz de Tenerife, Spain OBJECTIVES: Health care workers are a population at risk of Latent Tuberculosis Infection (LTI), which is usually detected by the Tuberculin Skin Test (TST). In vitro immunological tests such as QuantiFERON-TB Gold® (QFT-G) have been recently recommended for LTI screening in health care workers. We compared direct and indirect costs of two LTI screening strategies among health care workers in Spain: TST and QFT-G. METHODS: This was a comparative cost study conducted from a societal perspective, using data on costs and results from a prospective observational study carried out in a Spanish public hospital, where the 2 screening alternatives were concurrently applied to a cohort of 134 health care workers. RESULTS: In a base-case analysis, the costs of the QFT-G test amounted c42.5 per screened health care worker and those of the TST c39.5. Both tests varied in their cost structure: in the case of TST, most of the total costs (70%) were indirect costs, basically time spent by the participants, whereas QFT-G was more expensive in terms of fungible material, which meant 50% of the total costs. The results are sensitive to the hourly wages of the participants and to the estimation of the time spent by them in the tests. CONCLUSIONS: This cost study showed that, in the conditions of Spanish health care system, the societal costs of the new QFT-G are comparable to those of the TST; however, their cost structures vary considerably. Therefore, these results could change if applied in other countries with different relation between salaries and prices. PIN12 PSEUDOMONAS AERUGINOSA RELATED BURDEN ON CYSTIC FIBROSIS PATIENTS: COMPARING HEALTH CARE COSTS AND RESOURCE UTILIZATION ACROSS AGE GROUPS Sansgiry S1, Joish V2, Boklage S2, Goyal RK1, Seal B2, Sethi S3 1 University of Houston, Houston, TX, USA; 2Bayer, Wayne, NJ, USA; 3University at Buffalo, SUNY, Buffalo, NY, USA OBJECTIVES: To determine if the average cost of medical care among cystic fibrosis (CF) patients with Pseudomonas aeruginosa (PA) infection is different across agegroups. METHODS: Data were derived from MarketScan claims database, which captures person-specific direct medical utilization, expenditures, and enrollment from approximately 150 payers. a retrospective cross-sectional study design was used. CF subjects with an initial claim for a PA infection were identified using international classification of diseases diagnosis codes of 277.0 and 482.1, respectively. Demographic information from administrative claims and health care utilization and costs from medical and pharmacy claims were extracted for 12 months pre and post initial PA claim. All resource use and costs were annualized and compared across 7 agegroups with parametric (ANOVA) and Duncan’s post-hoc tests using SAS version 9.2. RESULTS: A total of 347 CF subjects with PA infection met the study criteria with mean age 19.9 (SD: 15.4) years and 47.8% females. a monotonic trend of increasing (P < 0.05) overall post period costs was observed across the 7 age-groups. Example, children 0–4 years had the lowest (P < 0.05) overall post costs of $31,569 (median = $22,887) vs. $95,024 (median = $36,783) for adults 45–64 years. a similar trend with the exception of age group 30–44, was observed with PA-related costs as well. Mean and median per patient per year PA-related costs were lowest in children 0–4 years ($12,472, $3,572) compared to the oldest age-group of 45–64 years ($26,673, $3,311). In general, total prescription claims and outpatient visits, and PA-related inpatient visits increased with increasing age groups. Among children, PA-related prescription claims were statistically higher. For example, mean PA-related prescription claims were 1.4 vs. 3.8 in the 0–4 vs. 15–19 groups (P < 0.05). CONCLUSIONS: Overall and PA-related health care resource use and costs tended to vary across age groups. Future research needs to explore the underlying reasons for this trend. PIN13 LONG-TERM ECONOMIC AND CLINICAL BURDEN OF COMPLICATED INVASIVE MENINGOCOCCAL DISEASE: EVIDENCE FROM A UNITED STATES MANAGED CARE POPULATION Karve S1, Misurski D2, Miller J3, Davis KL1 1 RTI Health Solutions, Research Triangle Park, NC, USA; 2GlaxoSmithKline, Philadelphia, PA, USA; 3GlaxoSmithKline Biologicals, King of Prussia, PA, USA OBJECTIVES: There is a paucity of data on the long-term (i.e., post-hospital discharge) economic and clinical burden of invasive meningococcal disease (IMD) and its related complications among IMD survivors. The objective of this study was to compare health care utilization and costs between IMD survivors with and without related complications. METHODS: We conducted a retrospective cohort analysis of the Ingenix Impact database (1997–2009). Patients with an inpatient admission for IMD (ICD-9-CM: 036.x) and continuous health plan enrollment for ≥6 months before and ≥12 months after the initial IMD hospitalization were selected. Patients were further classified based on the presence (complicated IMD) or absence (uncomplicated IMD) of a diagnosis code for relevant clinical sequelae (identified based on literature review and clinical expertise) during the 12-month follow-up period. Health care utilization and costs (in 2009 US$) incurred during the 12-month follow-up period were compared between patients in the complicated and uncomplicated IMD groups using univariate and multivariable regression analyses. RESULTS: Among 343 IMD 13th Euro Abstracts patients identified, stroke (14.3%), seizure (11.7%) and hearing loss (10.5%) were the most commonly observed complications, with 34.1% experiencing ≥1 complication. Significant differences in health care utilization and costs were observed, with the largest between-group difference in follow-up costs being for inpatient services (mean [95% CI]: $72,512 [$6,439] for complicated cases vs. $24,679 [$1,234] for uncomplicated IMD; P < 0.001). Large differences were also observed for rehabilitative services ($24,405 [$21,418] vs. $159 [$116]; P < 0.05) and total health care costs ($97,854 [$9,248] vs. $32,239 [$1,611]; P < 0.001). Risk of re-hospitalization following discharge from the initial IMD admission was higher among complicated IMD patients (hazard ratio = 1.69 [95% CI] = 1.04–2.74) vs. uncomplicated cases. CONCLUSIONS: Driven mainly be the need for repeat hospitalization, the presence of serious complications in cases of IMD increased health care utilization and costs by almost 3 fold compared to uncomplicated cases during 12 months post-diagnosis. PIN14 COST-OF-ILLNESS OF CANDIDEMIA IN KOREA 1 2 3 4 5 5 Peck KR , Han SH , Chang HH , Jung SI , Oh JJ , Ko SK 1 Samsung Medical Center, Seoul, South Korea; 2Yonsei University College of Medicine, Seoul, South Korea; 3Kyungpook National University Hospital, Daegu, South Korea; 4 Chonnam National University Hospital, Gwangju, South Korea; 5Pfizer Pharmaceuticals Korea Ltd, Seoul, South Korea OBJECTIVES: This study sought to estimate the direct medical costs associated treating candidemia in Korea. METHODS: This incidence based, cost-of-illness study collected retrospective data of 200 patients in 4 tertiary hospitals. The cost data targeted for adult patients (≥19 years) diagnosed as candidemia through blood culture. We assessed the costs attributable to candidemia by examining resources utilization during candidemia treatment period. The medical costs were calculated by multiplying quantity of resource utilization by unit cost of according resource. RESULTS: The enrolled patients were 54.0% male, average 65.3 yrs old. The average length of stay attributable to candidemia was 16.3 days. After 6 weeks follow up, only 2 patients were relapsed, otherwise all-cause mortality rate was 66.0%. The estimated average direct medical costs of candidemia were KRW 4,723,160. The resources consisted of hospitalization (KRW 1,308,521, 27.7%), medication (KRW 1,310,739, 27.8%), lab test (KRW 489,818, 10.4%), imaging test (KRW 157,633, 3.3%), procedure/surgery (KRW 113,774, 2.4%) and other medical treatment (KRW 1,342,675, 28.4%). The main occupied costs were those of hospitalization, medication and when other medical treatment resources (e.g, hemodialysis, blood transfusion) were utilized, the costs tended to sensitively increase. According to admission type, costs for ICU (SICU: n = 24, KRW 8,837,825, MICU: n = 51, KRW 6,914,280) were higher than those for general ward (n = 125), KRW 3,039,167. Analysis by baseline disease/condition revealed that the costs for transplant (n = 8, KRW 10,070,472), HIV/AIDS (n = 2, KRW 25,426,018) were higher than others (cancer: n = 103, KRW 3,658,142, central catheterization: n = 101, KRW 5,554,520, surgery: n = 68, KRW 5,050,941). The costs for C.albicans (n = 90) were KRW 3,878,166 and for non C.albicans (n = 110) were KRW 5,414,518. CONCLUSIONS: This study is significant in that it estimated cost-of-illness of candidemia by examining the health resources consumption and assessing the costs attributable to candidemia. PIN15 THE ECONOMIC BURDEN OF SURGICAL SITE INFECTION USING THERAPEUTIC ANTIBIOTIC UTILIZATION MEASURE—COMPARISON OF TWO TIME PERIODS Patkar AD1, Magee G2, Vaughn B1, Edmiston CE3, Vardireddy N2 1 Ethicon, Inc, Somerville, NJ, USA; 2Premier, Inc, Charlotte, NC, USA; 3Froedtert Hospital, Milwaukee, WI, USA OBJECTIVES: Significant attention is being focused on reducing surgical site infections (SSI) in the US and numerous national initiatives have been put into place to achieve measurable reductions. The purpose of the study was to examine the economic impact of therapeutic antibiotic (TA) utilization as an indicator of postoperative SSI between two time periods. METHODS: Premier inpatient database was utilized for assessing length of stay (LOS) and costs. Two time periods identified, 2005–2008 (period 1) and 2009 (period 2) with 1,138,989 patients discharged in 2005–2008 and 305,073 discharged in 2009. The patients with non-SSI nosocomial infections were excluded. TA usage was determined by the antibiotic administration after day 4 of surgery (TA utilization rate: 0.61% in 2005–2008; 0.75% in 2009). Multivariate analysis used to assess the effects of using TA on LOS and total costs outcomes. RESULTS: Patients receiving TA had significantly higher LOS and costs for both time periods (P < 0.001). Average LOS for patients receiving TA was 12.2 and 12.6 for periods 1 and 2 respectively. Mean post-surgical LOS was 9.8 and 10.0. Patients not receiving TA had average LOS of 4.6 and 4.6 in periods 1 and 2 and mean postsurgical LOS was 3.9 and 3.8 for periods 1 and 2. Mean (SD) total costs for TA patients were significantly higher, $28,601 ($31,892) in period 1 and $32,751 ($38,194) in period 2, compared to $15,336 ($33,406) and $15,412 ($20,311) for patients not receiving TA. Predictors for significantly higher risk of TA use included General Surgery, Non-cardiac Thoracic procedures, bed-size under 500 or Rural, and Staples or Non-absorbable sutures usage. CONCLUSIONS: These findings suggest minimal impact on the SSI economic burden indicated by similar trends in antimicrobial utilization, costs and LOS over time. Further studies are warranted assessing the role of innovative technology to improve patient outcomes while reducing antibiotic utilization and LOS. 13th Euro Abstracts PIN16 HEALTH CARE UTILIZATION AND COSTS ASSOCIATED WITH HEPATITIS A IN A US COMMERCIALLY INSURED POPULATION Eisenberg D1, Changolkar A2, Misurski D2 1 HealthCore, Wilmington, DE, USA; 2GSK, Philadelphia, PA, USA OBJECTIVES: While the economic burden associated with acute hepatitis a outbreaks studied, the impact on health care utilization and associated cost has not been examined over one year in a commercially insured population. METHODS: An observational, retrospective study using HealthCore Integrated Research Database (HIRDSM) medical and pharmacy claims (January 1, 2004–December 31, 2009) was conducted. Index date was assigned based on the first medical claim for hepatitis a (intake period January 1, 2005—December 31, 2008). The control group did not have hepatitis. The index date assigned was of the hepatitis a match. Subjects had continuous plan eligibility for ≥12 months pre- and post-index. Descriptive statistics, including t-tests and chi-square analyses were used. RESULTS: Comparisons (n = 2331 for each cohort) were significant (p < 0.01). Post-index inpatient hospitalization occurred for 482 (20.7%) hepatitis a patients and 193 (8.3%) controls. Associated mean costs (+SD) were $5,401 (+/−$30,567) and $1,929 (+/−$23,282), respectively. Emergency room (ER) services occurred in 382 (16.4%) hepatitis a subjects versus 277 (11.9%) controls. Mean costs were $231 (+/−$816) and $139 (+/−$604), respectively. The mean number of outpatient services for the hepatitis a group was 21.5 (+/−22.1) versus 14.7 (+/−18.2) for controls with associated mean costs of $4,985 (+/−$14,936) and $2,948 (+/−$14,624), respectively. The mean number of physician visits were 9.1 (+/−9.9) and 5.5 (+/−7.4) for each group, respectively [mean costs = $1185 (+/−$3712) and $675 (+/−$2183)]. There were 20.13 (+/−24.1) and 19.37 (+/−24.5) pharmacy claims with mean costs of $2255 (+/−$5828) and $1420 (+/−$2656), in the hepatitis a and controls. Mean total associated costs were $14,056 (+/−$42,666) and $7,111 (+/− $30,519) in the hepatitis a and controls. CONCLUSIONS: In the 12-month post-index period hepatitis a patients had greater total health care utilization and costs compared to controls suggesting hepatitis a is associated with an increase in health care utilization and costs. PIN17 IMPACT OF YELLOW FEVER OUTBREAKS IN BRAZIL, 2000–2009 Ramos CG, Ferreira Da Silva AL Universidade Federal do Rio Grande do Sol, Porto Alegre, Brazil OBJECTIVES: To evaluate the health and economic impact of the expansion of YF virus in Brazil in the period of 2000–2009. METHODS: The perspective of study was the public health care system. We evaluated the direct costs of hospitalization and YF vaccine. Data source was the public official national information on YF cases and YF vaccination from Brazilian Ministry of Health. We assumed that all cases required hospitalization and that all fatalities required intensive care unit (ICU) hospitalization and dialysis. All values were referent to December of 2009. The values of hospitalization by yellow fever, five days ICU hospitalization and five days dialysis paid by public health system were, respectively, R$174.45 (c79.3); R$2393.60 (c1088) and R$ 557.1 (c253.23). RESULTS: During the evaluated period, YF virus circulation was detected in eight states previously YF free. There were 320 cases of YF with lethality rate of 46.5%. The cumulative population vaccinated was 90,906,675 individuals. The YF related hospitalization cost was R$ 495,478.30 (c225,217.41) and YF vaccine cost was R$89,997,608 (c40,908,003.75) within the 10 year period evaluated. CONCLUSIONS: Even in areas with susceptive population there were few cases of yellow fever. However, the significant clinical impact of the disease was due to its high lethality. Despite the severity of yellow fever, hospitalization costs were modest in the study period. The main cost was with the vaccine, the only form of prevention and control of disease. Thereby, it is necessary to evaluate the financial impact of vaccinating the areas that still have no evidence of the yellow fever virus, since there seems to be an expansion tendency of YF virus circulation to the country as a whole and the vaccination could prevent YF cases and deaths. PIN18 A PILOT COST-OF-ILLNESS STUDY ON LONG TERM COMPLICATIONS/ SEQUELAE FOLLOWING ACUTE OTITIS MEDIA (AOM) IN BELGIUM Strens D1, Dhooge I2, Knerer G3, Van Vlaenderen I1 1 Deloitte, Diegem, Belgium; 2Ghent University Hospital, Ghent, Belgium; 3GlaxoSmithKline Biologicals, Wavre, Belgium OBJECTIVES: A study was carried out to assess the economic burden associated with complications and/or long -term sequelae of AOM. This pilot study was done to assess feasibility to extend to a larger sample. METHODS: A retrospective patient chart review was performed in one Belgian center to obtain data on medical consumption related to the management of complications/sequelae of AOM. Data were collected for 10–15 years from the first AOM event onwards. Patients were eligible if they were between 10 and 20 years old with minimum 9 years of follow up data available and experienced a complication of AOM. Data on demographics, occurrence and type of complication, management of AOM, complications and sequelae were collected per year. Direct costs were calculated by multiplying each item of resource use with its unit cost (2010 in c) using the Belgian public health care payer’s perspective. Average (SD) yearly cost per patient, average yearly cost pre-complication, average cost during the year of complication, average yearly cost post-complication were calculated. RESULTS: Twenty-five patients were recruited in this pilot study. Complications were chronic suppurative OM with cholesteatoma, chronic perforation of the tympanic membrane and sensorineural hearing loss. The average yearly cost per patient was A433 c966(SD:1579) with following components; hearing aid devices (c381(SD:1356)), hospitalizations (c298(SD:250)), outpatient procedures (c121(SD:107)), visits (c91 (SD:183)), tests (c48(SD:46)) and drugs (c25(SD:89)). a complication occurred on average 5.3 years (SD: 2.9) after the first AOM event. Average annual cost pre-complication was c232(SD:218), with outpatient procedures as main cost drivers. Average cost of complication per annum was c2808(SD:4827), average yearly cost postcomplication was c1593(SD:3366). As from occurrence of complication onwards costs are mainly driven by hospitalizations and hearing aid devices. CONCLUSIONS: Long term complications/sequelae of AOM result in considerable costs for the health care payer, mainly driven by hospitalizations, and hearing aid devices. PIN19 HEPATITIS C BURDEN IN RUSSIA Omelyanovsky VV, Avksentieva MV, Krysanov I, Ivakhnenko O Research Center for Clinical and Economic Evaluation and Pharmacoeconomics, Moscow, Russia OBJECTIVES: To assess social and economic burden of a hepatitis Ñ infection in Russian Federation. METHODS: Cost of illness analysis was performed. Available data on epidemiology of hepatitis C infection in Russian Federation has been analyzed. Experts were surveyed to describe common practice of treating patients with acute and chronic hepatitis C. Costs of medical care were calculated from the health care system point of view. RESULTS: State registration of hepatitis C infection has been launched in 1994. The incidence of acute hepatitis C is decreasing while there is increase in chronic hepatitis C new cases. Number of people with asymptomatic virus carriage was increasing till 2001 and has been slowly decreasing since then. There is no data about hepatitis C prevalence in Russia. Expected economic burden was calculated on the basis of several assumptions: 1) only new cases of chronic hepatitis C are treated; 2) all new cases of acute hepatitis C are hospitalized, all new cases of chronic hepatitis C get hospital treatment first, then out-patient treatment; 3) and only 15 % of all patients with chronic hepatitis C receive treatment with pegylated interferons and ribavirin reimbursed from national health care system. CONCLUSIONS: Based on these assumptions expected economic burden of hepatitis C infection in Russian Federation is about c0,16 million. PIN20 PSEUDOMONAS AERUGINOSA RELATED ECONOMIC BURDEN STRATIFIED BY COMORBIDITY IN CYSTIC FIBROSIS PATIENTS Sansgiry S1, Joish V2, Boklage S2, Goyal RK1, Seal B2, Sethi S3 1 University of Houston, Houston, TX, USA; 2Bayer, Wayne, NJ, USA, 3University at Buffalo, SUNY, Buffalo, NY, USA OBJECTIVES: To determine if the average cost of medical care among cystic fibrosis (CF) patients post Pseudomonas aeruginosa (PA) infection is different for those with underlying comorbid conditions. METHODS: MarketScan claims database which captures person-specific direct medical utilization, expenditures, and enrollment from approximately 150 payers was used to conduct a retrospective cross-sectional study. CF subjects with an initial claim for a PA infection were identified using international classification of diseases diagnosis codes of 277.0 and 482.1, respectively. Demographic information from administrative claims and health care utilization and costs from medical and pharmacy claims were extracted for 12 months post initial PA claim. All resource use and costs were annualized and compared among subjects with comorbidity and the overall sample using SAS version 9.2. RESULTS: Of the 347 CF subjects with PA infection that met the study criteria, 79% had multiple co-morbidities, 67% had pancreatic insufficiency, followed by chronic sinusitis (19%), CF-related diabetes (13%), bronchiectasis (7%), and osteoporosis (5%). Subjects had a mean age of 20 years (SD 15.4) and 47.8% were females. As compared to an average CF subject, those with multiple co-morbidities had significantly higher (P < .05) overall pharmacy claims (mean 39 vs. 44) and cost (mean $20,673 vs. $24,121, median $17,121 vs. $22,344). Overall cost in subjects with CF-related diabetes (mean = $66,364 and median = $60,759) and osteoporosis (mean = $88,703 and median = $55,686) was significantly (P < 0.05) greater as compared to an average CF subject (mean = $51,821 and median = $36,046). Among CF-related diabetic subjects and osteoporosis subjects, significantly (P < 0.05) higher outpatient visits and pharmacy claims and higher inpatient costs contributed to overall increase in health care costs compared to the average CF subject. CONCLUSIONS: Overall post-PA infection medical care costs were higher in comorbid cohorts compared to the overall CF sample. Future research needs to determine cost-effectiveness of PA infection treatment for among these comorbid conditions. PIN21 EMPIRIC ANTIBIOTIC PRRESCRIBING IN SUBJECTS WITH INFLUENZA IN A UNITED STATES COMMERCIALLY INSURED POPULATION Misurski DA1, Lipson D2, Changolkar AK1 1 GlaxoSmithKline, Philadelphia, PA, USA; 2GlaxoSmithKline, King of Prussia, PA, USA OBJECTIVES: To evaluate empiric antibiotic prescribing and associated costs in commercially insured members with seasonal influenza. METHODS: This retrospective analysis was conducted using Impact National Benchmark Database pharmacy, medical, and confinement claim data (01/01/2004 through 12/31/2009). The index date was defined as the first influenza (ICD-9:487.xx) claim, within the intake period (01/01/2005 through 11/30/2009). Subjects had continuous eligibility for ≥12 months prior to and one month after the index date. We used the first influenza event only. Subjects with a pre-index diagnosis requiring antibiotic treatment were excluded. A434 Influenza subjects with an antiviral claim within one day pre- and one day post-index and/or an antibiotic claim within three days pre- and three days post-index were identified. Subjects with a secondary respiratory infection within the two week postindex period were identified via ICD-9 codes. RESULTS: The study cohort included 270,057 members with influenza (51.8% male; mean age 31.6 years). Antibiotics were prescribed in 63,358 (23.5%) of subjects (42,374 [15.7%] received an antibiotic only; 20,984 [7.8%] received both an antiviral and an antibiotic). The majority of subjects with influenza (157,447 [58.3%]) did not receive an antiviral. Among subjects receiving antibiotic treatment, 98.9% did not have a follow-up coded respiratory bacterial infection within two weeks. The mean antibiotic costs were $66.15 and $42.71 for subjects with and without a secondary bacterial respiratory infection, respectively. Antibiotic usage in influenza subjects without a secondary bacterial respiratory infection accounted for approximately $2.7M (2009 US$). CONCLUSIONS: Almost all subjects with influenza receiving antibiotic treatment did not have a secondary bacterial respiratory condition. The cost associated with the empiric prescribing of antibiotics was considerable. Furthermore, over half the subjects with influenza did not receive antiviral therapy. As our study only allowed for one case per subject the total cost associated with empiric antibiotic prescribing in subjects with influenza is likely an underestimate. PIN22 A NOVEL APPROACH FOR ESTIMATING THE VALUE OF ENTECAVIR THERAPY FOR NUCLEOSIDE-NAÏVE CHRONIC HEPATITIS B PATIENTS IN SPAIN Ramirez de Arellano A1, Fernandez-Rodriguez C2, Andrade R3, Clemente G4, Sola R5, Wells J6, Rentero P1 1 Bristol-Myers Squibb Iberia, Madrid, Spain; 2Hospital Fundación Alcorcón, Madrid, Spain; 3 Hospital Virgen de la Victoria, Malaga, Spain; 4General Hospital Gregorio Marañon, Madrid, Spain; 5Hospital del Mar (IMIM), Barcelona, Spain; 6Monitor Group, London, UK OBJECTIVES: Prescribers, payers and health care decision-makers operating in diverse health care settings are increasingly examining the value of innovative treatments. The analysis aims at estimating the relative value of the two preferred first-line oral antiviral treatment options for chronic hepatitis B (CHB) according to the EASL guidelines: entecavir (ETV) and tenofovir (TDF) from a Spanish payer’s perspective. METHODS: Hepatitis B disease was simulated using a health-state transition model with disease-states defined as mild disease (Ishak F0/F1), fibrosis (F2–F4) and advanced fibrosis/cirrhosis (>F4), and complicated disease-states (decompensated cirrhosis, hepatocellular carcinoma, liver transplant and death) based on available natural history data. The value of treatment-specific attributes on disease progression/regression was estimated from clinical practice data in real world settings in terms of lifetime events and costs avoided. Treatment attributes included: additional testing requirements at treatment initiation and required renal function monitoring; adverse events per product labelling; reversal of fibrosis; and treatment intensification in case of suboptimal viral suppression to avoid potential resistance. a 5-year treatment duration was assumed. Primary model output is the estimated cost avoided per-patient per day of treatment, associated with the clinical attributes of ETV compared to TDF in the treatment of nucleoside-naïve HBeAg-positive and -negative CHB patients. RESULTS: The daily cost of antiviral therapy for a CHB patient was estimated at c13.2 for ETV compared to c9.6 with TDF. Lifetime treatment cost avoided due to ETV’s clinical attributes was quantified at c3.6 [95% confidence: interval: 2, 5.2] per day of therapy. CONCLUSIONS: Estimating total treatment cost-avoidance associated with specific clinical attributes is a novel approach in assessing the value of treatment options. ETV exhibits the characteristics of a favourable CHB treatment option, which directly translates into economic and therapeutic added value. PIN23 DIFFERENTIAL PRICING POLICY AND ITS IMPACT ON TREAMENT COSTS OF VISCERAL LEISHMANIASIS IN THE BRAZILIAN PUBLIC HEALTH SYSTEM Marinho DS1, Romero CNP1, Cechinel MP2, Martins EV1, Morel CM1 1 Oswaldo Cruz Foundation, Rio de Janeiro, Brazil; 2Brazilian Ministry of Health, Brasília, Brazil BACKGROUND: Visceral leishmaniasis (VL) (“Kalazar”) is an important public health issue in Brazil. The guidelines of Brazilian Ministry of Health set up pentavalent antimony as the first line treatment, and amphotericin (deoxicolate and liposomal) as the second line treatment of VL. There are evidences that liposomal amphotericin (LA) and amphotericin deoxicolate (AD) present similiar efficacy, but LA has a better safety profile. Unfortunately, its price has been the principal barrier to its usage by public health system. OBJECTIVES: To evaluate the impact of differential pricing policy of LA in the Brazilian public health system. METHODS: The global amount spent on medicines from 2007 to 2009 by the Ministry of Health on the different types of treatment was retrieved from the Brazilian sanitary surveillance system. The following associated costs were also analyzed: hospital care and professional services; laboratorial tests (potassium, urea, creatinine and transaminases); and loss of working days (average daily income). This information allowed us to perform the cost analysis of the different therapeutic options. RESULTS: The differential pricing policy for LA implemented in 2008 allowed a cost reduction of 76.8% (from US$5.881 to US$590 per treatment) enabling the Ministry of Health to double the number of patients treated with LA. The associated costs of hospitalization and loss of productivity by patients are higher with AD than with LA, but the total cost of LA treatment (medication + associated costs) is still 11.3% to 27% higher. CONCLUSIONS: In a societal perspective it was shown that a substantial price reduction of a safer drug enhanced 13th Euro Abstracts its use and distribution by the Ministry of Health. It became also evident that the policy of differential pricing is an effective way to promote access to an efficient and safer treatment of VL, a disease that mainly affects the poor. PIN24 IMPACT OF APPROPRIATE PRESCRIBING OF ANTIBIOTICS ON TREATMENT OUTCOMES, DAYS OF PATIENT HOSPITALIZATION AND COSTS OF ANTIBIOTIC TREATMENT IN HOSPITALS IN LESOTHO Adorka MKB1, Serfontein JHP2, Lubbe MS2, Gous AGS3 1 National University of Lesotho, Roma, Lesotho; 2North-West University, Potchefstroom, South Africa; 3University of Limpopo, Pretoria, South Africa OBJECTIVES: The primary aim of this study was to determine the possible impact of appropriate prescribing of antibiotics on treatment outcomes, days of patient hospitalization and costs related to antibiotic treatment. METHODS: Data from case notes of all inpatients on antibiotic treatment and all antibiotic prescriptions from outpatient departments were collected from June 15 to July 15, 2006 at 5 government and mission hospitals in Lesotho. All prescriptions were classified into categories of appropriateness based on their conformities to criteria developed from principles of antibiotic prescribing. Analyses were further conducted to determine the possible impact that appropriate prescribing may have on treatment outcomes, days of patient hospitalization and costs of antibiotic treatments. RESULTS: A total of 307 inpatients and 867 outpatient prescriptions were assessed. Total frequencies of prescribed antibiotics were 584 and 1073 for inpatients and outpatients respectively. For inpatients antibiotics were most often prescribed for skin and soft tissue (31.7%) and respiratory tract (28.9%) infections. Outpatients received most antibiotics for respiratory tract (42.0%) and skin and soft tissue infection (21.9%) respectively. Of the 307 inpatient prescriptions 41.6% were appropriately written for either empiric treatment or prophylaxis of infections while 57.1% did not conform to antibiotic prescribing principles. a majority (81.0%) of outpatient prescriptions were in line with the prescribing principles. Inpatient antibiotic prescribing patterns based on principles of antibiotic prescribing have had a positive impact on treatment outcomes, days of hospitalization and costs of antibiotic treatment. In outpatients prescribing of antibiotics based on principles of antibiotic prescribing failed to show any significant impact on costs of antibiotics. Due to geographical challenges the outcome for treatment could not be determined for outpatients. CONCLUSIONS: Rational prescribing had a positive impact on treatment outcomes. PIN25 SEASONAL INFLUENZA VACCINATION OF HEALTHY, WORKING-AGE ADULTS: A SYSTEMATIC REVIEW OF ECONOMIC INVESTIGATIONS Gatwood J1, Meltzer M2, Messonnier M2, Prosser LA1 1 University of Michigan, Ann Arbor, MI, USA; 2Centers for Disease Control and Prevention, Atlanta, GA, USA OBJECTIVES: To assess cost-benefit and cost-effectiveness studies examining seasonal influenza vaccination in the healthy, working-age population of the United States (US). METHODS: A systematic review of economic evaluations investigating the costbenefit or cost-effectiveness of vaccinating healthy adults was conducted on articles published between January 1990 and January 2010. Studies considered for inclusion in this review were identified using Medline, Embase, Econlit, and Google Scholar (n = 11). The source and appropriateness of parameters (epidemiological data, probabilities, and costs), the designs employed, and the sufficiency of sensitivity analysis were considered during the review. RESULTS: Seven studies were identified as appropriate for this review. Key inputs of the selected studies included influenza or ILI attack rates, outpatient visits averted, total vaccination costs, and lost workdays. All studies were conducted in the US and from the societal perspective; three were randomized, clinical trials and the remaining four constructed economic simulation models to compare vaccination to antivirals or no intervention. Results ranged from net savings of $69.19 to net costs of $85.89 per vaccination (4 studies) and from $103.95 to $1004.85 per episode of influenza averted (1 study). Only 2 studies reported cost-effectiveness ratios; these ranged from $6,930 to $50,510 per quality-adjusted life-year gained. Nearly all of the studies conducted sensitivity analysis; results were most sensitive to variation in wage rates, levels of worker productivity, the cost and effectiveness of vaccination, and the rate of illness. CONCLUSIONS: This systematic review found that seasonal influenza vaccination of healthy, working-age adults is generally not cost saving, requiring an investment to generate health benefits. The decision to vaccinate such a group will depend upon the societal and payer valuation of those benefits. PIN26 EARLY VS. LATE DETECTION OF HIV IN UK: COST CONSEQUENCE STUDY Mesa OA1, Mason N1, Zah V2 1 Bristol-Myers Squibb Pharmaceuticals Ltd, Uxbridge, Middlesex, UK; 2ZRx Outcomes Research Inc, Toronto, ON, Canada OBJECTIVES: Identifying budgetary implications and opportunities for earlier detection versus late detection of HIV (CD4 cell count less than 200/µl) in the 1 through 5 year term following diagnosis. In 2009, there were 90,468 HIV positive individuals not accounting for unreported HIV positive cases (in 2008, estimated additional 24%). Late detection HIV patients account for 32% from the total. METHODS: A literature review was conducted using MEDLINE, Cochrane Library, EMBASE and CRD for guidelines and clinical trials published 1995–2010 in UK, England, Scotland, Wales, Northern Ireland, USA, Canada, Australia, Norway and EU-26. We extrapolated UK 13th Euro Abstracts population breakdown, survival rates, average costs of triple and quadruple HAART (Highly Active Antiretroviral Therapy) therapy, inpatient, outpatient care costs—HIV related HRG costs and excess cost ratio due to late presentation. Study follows standard good modeling practices by discounting, adjusting for survival, and considering multiple time horizons for analyses. UK data sources include: BHIVA, NHS, HPA and SOPHID. RESULTS: From 1995–2010 literature review yielded 126 publications related to late detection, of which 18 were included in our study design. The number of publications has rapidly increased over the last 12 months. Preliminary results suggest that 10% of HIV detection shift from late to early detection (2655 patients or 3.2% of Late HIV detection) would result in £20.3 million conservative savings just in the first year (keeping index multiplier scalar for HIV spread rate at zero). Over 5-year term, such annual shift would result in 708.25 life-years saved or £14.2 million in additional savings. CONCLUSIONS: Enhanced testing to achieve earlier detection and initiation on HAART could potentially reduce overall costs of HIV medical care. Further research is required to determine additional budgetary implications of HIV comorbidity patterns. PIN27 COST-EFFECTIVENESS ANALYSIS OF HPV VACCINATION AGAINST CERVICAL CANCER IN YOUNG ADULT WOMEN IN ITALY La Torre G1, Gabutti G2, Cristoforoni P3, Bonanni P4, Amunni G4, Costa S5, Capri S6 1 Sapienza University of Rome, Rome, Italy; 2University of Ferrara, Ferrara, Italy; 3National Institute for Cancer Research, IST Genova, Genova, Italy; 4University of Florence, Florence, Italy; 5St. Orsola-Malpighi University Hospital, Bologna, Italy; 6Cattaneo-LIUC University, Castellanza, Varese, Italy OBJECTIVES: Human papillomavirus (HPV) has been implicated as a causal factor in cervical cancer (CC), the second most common cancer among women worldwide. Prophylactic vaccination against HPV-16/18 has been shown to be highly effective in preventing HPV related infections and pre-cancerous lesions. The aim of this study was to determine the potential clinical and economic impact of combining CC screening programme with HPV vaccination programme with the bivalent HPV-16/18 vaccine in women aged 25 (when CC screening starts) compared to screening alone. METHODS: A Markov cohort model was adapted to the Italian setting. The model replicates the HPV infection natural history leading to CC and includes the effect of screening and vaccination. Based on the latest results of the PATRICIA clinical trial for the bivalent vaccine, the vaccine efficacy includes cross-protection against nonvaccine HPV types 31/33/35/39/45/51/52/56/58/59. The vaccine efficacy in naïve girls and HPV-exposed women (>17-years old) was differentiated. Lifetime protection and 90% vaccination coverage were assumed. Costing was analyzed from the perspective of the Italian health care system. Main outcomes are lifetime costs, QALYs, CC cases, deaths and ICER. Both costs and effects were discounted at 3% annually to calculate the ICER. RESULTS: The model shows that, compared to screening alone, vaccinating a single cohort of 330,000 women aged 25 would prevent over a lifetime 696 CC cases and 316 CC deaths (undiscounted) of which 131 and 59 respectively are due to cross-protection, with a discounted ICER of c33,918 per QALY gained. The ICER remains under the cost-effective threshold defined by the WHO (between 1 and 3xGDP/capita). CONCLUSIONS: Under the assumptions of the model, extending vaccination to young adult women post-HPV exposure could lead to a substantial reduction in CC and remains cost-effective in Italy compared with screening alone. Cross-protection would play an important role in this reduction. PIN28 IMPACT OF METHODOLOGICAL CHOICES AND ASSUMPTIONS IN ECONOMIC EVALUATIONS OF ROTAVIRUS VACCINATION Millier A1, Aballea S1, Petrou S2, Quilici S3 1 Creativ Ceutical, Paris, France; 2University of Oxford, Oxford, England; 3Sanofi Pasteur MSD, Lyon, France OBJECTIVES: Rotavirus is the leading cause of severe gastroenteritis in children under 5 years. Two vaccines are currently available: RotaTeq® (Merck/Sanofi Pasteur MSD) and Rotarix® (GSK). Published economic models on rotavirus vaccination have produced contradicting results. We reviewed and critically appraised existing economic models. METHODS: The literature search covered worldwide cost-effectiveness models of RotaTeq® and Rotarix® published or presented at conferences until October 2009. We extracted information on model structures, input data (epidemiology, vaccine efficacy, utilities, vaccination costs) and results. RESULTS: We identified 44 publications referring to distinct cost-effectiveness analyses.18 used cost per QALY as primary outcome, providing 22 incremental cost-utility ratios (ICURs) for highincome countries. Thirteen ICURs came from health authorities and 9 from manufacturers. While some European studies predicted cost-savings from societal perspective, the ICUR was estimated at up c160,000 per QALY for the UK, from NHS perspective. 5 of 13 analyses from health authorities reported ICURs below c30,000 per QALY gained, compared to 7 of 9 studies from manufacturers. There were many differences between models, such as perspective, epidemiological inputs, vaccine efficacy and vaccination costs. However differences in results were mainly driven by assumptions related to estimation of QALYs: utility values, duration of symptoms, inclusion of quality of life burden for caregivers and for cases without medical attention. CONCLUSIONS: Results of cost-effectiveness analyses are highly variable due to uncertainty surrounding epidemiological inputs, and most importantly to assumptions for deriving QALYs. Different choices were attributable to different value judgments, and to the difficulty of measuring and valuing quality of life in children, which lead to a situation where neither manufacturers nor health authorities obtained valid utilities. A435 There is currently no valid reference case for cost-utility analyses in paediatrics. More open communication and expertise sharing between authorities and manufacturers could lead to more reliable analyses. PIN29 A DYNAMIC MODEL TO EVALUATE THE COST-EFFECTIVENESS OF 10-VALENT PNEUMOCOCCAL CONJUGATE VACCINE IN TAIWAN Chang CJ1, Wu BS2, Wu CL1, Lin YJ1, Fann SJ3 1 Chang Gung University, Taoyuan, Taiwan; 2National Yang-Ming University, Taipei, Taiwan; 3 Academia Sinica, Taipei, Taiwan OBJECTIVES: Streptococcus pneumonia is associated with two invasive diseases (meningitis and bacteremia) and non-invasive disease (pneumococcal pneumonia), which has caused high morbidity and mortality in infants and the elderly in Taiwan. a cost-effectiveness analysis of pneumococcal conjugate vaccine (PCV 10) in Taiwan was conducted using transmission-dynamic model. METHODS: As static pharmacoeconomic model fails to account for vaccination-induced herd-immunity, we developed an age-structured multi-compartment dynamic model to estimate both economic and clinical impact of universal PCV10 vaccination over 10-year time horizon under health care system’s perspective in Taiwan. Model parameters such as economic and clinical burden of pneumococcal diseases in Taiwanese population, vaccine efficacy, utilities and demographic parameters were captured by retrospective population-based National Health Insurance Reimbursement Database (NHIRD), published sources, unpublished data, and assumptions made in consultation with clinical experts. Univariate sensitivity analyses were conducted to test the robustness of model parameters. RESULTS: Assuming a four-dose schedule and 90% of vaccination coverage for both vaccines, universal infant vaccination with PCV10 would prevent 133 cases of IPD, 122,476 cases of pneumococcal pneumonia and 3,857 deaths, an equivalent of 93,393 life-years and leads to a net medical cost savings of NT$ 6,047.4 million, compared to no vaccination. At the current proposed price of NT$2,700 for PCV10, the incremental cost-effectiveness ratio (ICER) is expected to be cost-effective with NT$106,362 (US$3,324) per life-year gained. The program’s cost-effectiveness results are highly sensitive to the vaccine price and number of doses while not sensitive to uncertainty in disease incidence and costs of treatment. CONCLUSIONS: Universal pediatric PCV10 vaccination in Taiwan is estimated to have considerable impact on reducing the burden of pneumococcal diseases and expected to be cost-effective in payer’s perspective compared to no vaccination. PIN30 ECONOMIC EVALUATION OF DAPTOMYCIN AS FIRST-LINE THERAPY VERSUS DAPTOMYCIN AS RESCUE THERAPY AFTER VANCOMYCIN OR LINEZOLID FAILURE IN GRAM-POSITIVE BACTEREMIA TREATMENT Lahoz R1, Galera J1, Font B1, Gil Parrado S1, Soengas C1, Grau S2 1 Novartis Pharma, Barcelona, Spain; 2Hospital del Mar (IMIM), Barcelona, Spain OBJECTIVES: To assess the efficiency of daptomycin as first-line therapy (D) versus daptomycin as rescue therapy after failure of vancomycin (V+D) or linezolid (L+D) in gram-positive bacteremia treatment. METHODS: A cost-effectiveness analysis comparing the three therapeutic alternatives (D, V+D and L+D) was performed using data from a previous observational study (EUCORE). In the present sub-study, data on 19 (D), 33 (V+D) and 19 (L+D) bacteremic patients were analyzed. Effectiveness was measured in terms of cure or clinical improvement. Costs were gathered from “BOT” and “e-salud” Spanish databases. Direct costs (medication and hospitalization) due to bacteremia were included. Costs are expressed in 2009 Euros. Patients were observed until either end of daptomycin therapy or exitus. a probabilistic multivariate sensitivity analysis was carried out. Dispersion was estimated using bootstrap (three alternative scenarios were considered in the simulations: 50%, 25% and 10% of the estimated dispersion) and Monte Carlo simulations were performed for both costs (log-normally distributed) and effectiveness (normally distributed) measures. RESULTS: Effectiveness figures: D: 84.2% (95%CI: 60.4%–96.6%); V+D: 87.9% (95%CI: 71.8%– 96.6%); L+D: 68.4% (95% CI: 43.5%–87.4%); pvalue = 0.206 (Chi2). Average costs per treated patient: D: c6672.8 (95%CI: 4076.8–9268.8); V+D: c9786.6 (95%CI: 7,124.7–12,448.5); L+D: c12,190.4 (95%CI: 8,693.2–15,687.7); pvalue < 0.001 (Kruskal-Wallis). Cost-Effectiveness Incremental Ratios: c53,478.8 (D versus V+D) and −c6,639.5 (D versus L+D) per additional cured patient. Sensitivity analysis results: D versus V+D: D is efficient with lower costs (in 41.2%–77.9% of the simulations) and dominant (22.1%–38.0%); D versus L+D: D is dominant (44.6%–99.6%) and efficient with lower costs (0.4%–30.0%). CONCLUSIONS: Daptomycin as first-line therapy is mainly a dominant alternative when compared to daptomycin as rescue therapy after linezolid failure, showing better clinical results with lower associated costs. Daptomycin as first-line therapy does not show significant differences in effectiveness when compared to daptomycin as rescue therapy after vancomycin failure, but it shows significantly lower costs. PIN31 PHARMACOECONOMIC MODELING USING THE 13-VALENT PNEUMOCOCCAL CONJUGATE VACCINE IN GERMAN ADULTS Claes C1, Pletz MW2, van der Linden M3, Welte T2, Graf von der Schulenburg J-M1 1 Leibniz University of Hannover, Hannover, Germany; 2Medical School Hannover (MHH), Hannover, Germany; 3Universitaetsklinikum Aachen, Aachen, Germany OBJECTIVES: Vaccination with the 23-valent pneumococcal polysaccharide vaccine (PPV23) is the current standard in Germany for diagnosis defined risk groups (age 5–59 years) and for seniors (≥60 years). a redefinition of this recommendation A436 following market authorization of the 13-valent pneumococcal conjugate vaccine (PCV13) for adults might be considered. Thus, this study analyzes potential costeffectiveness of an adult vaccination with PCV13 from the point of view of the German social health insurance. METHODS: In a cross-sectional steady state Markov model, efficacies of the vaccines on invasive pneumococcal diseases (IPD), and communityacquired pneumonia (CAP, treated in either a hospital or a nonhospital setting) were calculated. a steady state is defined as a setting in which the whole population (including new entrants) is vaccinated as recommended. The modeling of PPV23 and PCV13 distinguished between risk groups and both vaccines were compared to non-vaccination. Data on PCV13 were derived from published results on the 7-valent pneumococcal conjugate vaccine (PCV7). The effectiveness of individual pneumococcal vaccination of adults was adjusted for expected herd immunity effects of a pediatric vaccination with PCV. Utilization of health care services and unit costs were taken from publicly accessible data bases. RESULTS: Compared to PPV23, PCV13 revealed the potential to avoid a greater number of yearly cases and deaths due to IPD and CAP in Germany. For PCV13 it can be expected that monetary savings, resulting e.g. from less hospitalization, compensate the costs of the vaccination program. The preliminary version of the model concludes that the cost-benefit ratio is 1.16, i. e. c1 spend on vaccination saves c1.16 treatment cost. CONCLUSIONS: Our model shows that the health economic benefit of an immunization of adults with PCV13 can be expected to be higher than that of PPV23. PIN32 ECONOMIC IMPACT OF THE 10-VALENT PNEUMOCOCCAL NONTYPEABLE HEMOPHILUS INFLUENZAE PROTEIN D CONJUGATE VACCINE (PHID-CV) IN A COHORT OF NEWBORN IN HONG KONG Lee KK1, Hon EK2, Lee VW2, So TM3, Chow DP2 1 Monash University Kuala Lumpur Sunway Campus, Selangor Darul Ehsan, Malaysia; 2Chinese University of Hong Kong, Hong Kong, China; 3Princess Margaret Hospital, Hong Kong, China OBJECTIVES: To examine the health and economic impact of PHiD-CV in the public sector of Hong Kong compared to no vaccination. METHODS: A Markov analytical model modified from the recent Cost-effectiveness Cohort Model for Synflorix (COSY Model by HE-group, GCRD, GSK bio, Wavre, Belgium, 2010) was used for the analysis of the outcomes of vaccination. The cohort model simulates in a birth cohort of 82,100 newborns the disease process of invasive disease (ID) (meningitis and bacteremia), community acquired pneumonia (CAP), and acute otitis media (AOM) over life-time caused by S pneumoniae and Non-typeable H influenzae (NTHi) within ‘monthly’ cycles. Superimposed on the disease process is integrated in the model its disease management expressed in hospitalization rates, medical visits, and specific interventions such as myringotomies. The epidemiology and disease management is Hong Kong specific. The study was performed from a health care payer’s perspective. RESULTS: The study shows for a 4-dose regimen of PHiD-CV compared to no vaccination over a period of 100 years: a gain of 234 quality-adjusted-life-years (QALY); an avoidance of 16 cases of IPD, 873 cases of pneumonia, 26,746 cases of AOM and 25,885 sequelae. Using a 5% discount for cost and benefits, cost/QALY gained was estimated to be HKD261,492. CONCLUSIONS: With GDP per capita of Hong Kong being HKD233,060 in 2009, our study results suggest PHiD-CV vaccination is costeffective in providing protection to the people of Hong Kong based on the WHO recommended thresholds for cost-effectiveness. PIN33 THE POTENTIAL PUBLIC HEALTH BENEFIT OF PNEUMOCOCCAL CONJUGATE VACCINES: EXAMPLE OF THE CZECH REPUBLIC Dolezal T1, Skoupá J2, Mrkvan T3, Topachevskyi O4 1 Institute for Health Economics and Technology Assessment, Prague, Czech Republic; 2 Pharma Projects s.r.o., Prague, Czech Republic; 3GlaxoSmithKline, Prague, Czech Republic; 4 GlaxoSmithKline Biologicals, Wavre, Belgium OBJECTIVES: To evaluate cost-effectiveness of routine pneumococcal vaccination with 10-valent pneumococcal non-typeable Haemophilus influenzae protein-D vaccine (PHiD-CV) compared with 13-valent pneumococcal conjugate vaccine (PCV-13) and no vaccination. METHODS: A Markov cohort model with a 100-year time horizon was developed to project the impact of vaccination on the incidence of pneumococcal and non-typeable Haemophilus Influenzae (NTHi) infections in children and adults. Data Sources: Czech Republic-specific epidemiological and demographic data and data from other country sources. Base-case assumptions include estimates of pneumococcal and NTHi infection rates as well as vaccine efficacy based on published literature. Model inputs: payer perspective, 80% vaccine coverage, no herd protection and a (3 + 1) vaccination schedule. One-way sensitivity analyses performed to assess the impact of changes in key model assumptions. RESULTS: PHiD-CV and PCV-13 are projected to prevent more cases of invasive disease (89 and 92 respectively) and pneumonia hospitalizations (738 and 760 respectively) compared with no vaccination. PHiD-CV and PCV-13 are projected to prevent additional GPP/ENT visits due to acute otitis media (AOM) (7851 and 3838 respectively) compared with no vaccination strategy. The total number of cases of invasive disease was comparable for both vaccines but PHiD-CV is estimated to prevent more than 14 000 cases of AOM in comparison with PCV-13. Vaccinating a birth cohort with PHiD-CV is expected to generate 75.5 more QALYs and 5.9 LYG compared to PCV-13. Under vaccine price parity assumptions, estimated total savings for health care system are 10.2 mil CZK or 395.6 k EUR for PHiD-CV compared to PCV-13 respectively. Sensitivity analyses indicate that AOM efficacy and incidence of AOM related GPP/ENT visits have biggest impact on results. CONCLUSIONS: Overall, PHiD-CV is expected to have better quality of life 13th Euro Abstracts impact than PCV-13. Under price parity assumptions, PHiD-CV dominates PCV-13 because it also has a larger cost offsets. PIN34 A CONTINUOUS-TIME ECONOMIC MODEL TO EVALUATE RALTEGRAVIR USE STRATEGIES IN TREATMENT-NAIVE HIV-1 PATIENTS IN PORTUGAL Chaudhary MA1, Elbasha EH1, Pereira R2, Kumar RN3 1 Merck & Co., Inc., North Wales, PA, USA; 2Merck Sharp & Dohme, Lda., Lisbon, Portugal; 3 Merck & Co., Inc., Whitehouse Station, NJ, USA OBJECTIVES: In contrast to commonly used discrete-time Markov models, we developed a more realistic continuous-time multi-stage Markov model to evaluate long-term clinical and economic outcomes of raltegravir in treatment naïve HIV-1 patients. METHODS: The multi-stage cost-effectiveness model incorporating 3 lines of therapy was developed using differential equations and was solved in Mathematica® 6.0. The analysis was conducted from the perspective of the National payer in Portugal. a typical patient enters the model in a given health state, transitions to another health state, can develop acquired immunodeficiency syndrome (AIDS)/ coronary heart disease (CHD)/other adverse events or die. Eighteen health states were defined based on CD4 and HIV RNA levels. We used the maximum likelihood method to estimate matrices of instantaneous transition rates corresponding to the efficacy of the included therapies. Six multi-stage treatment strategies depicting clinical practice in Portugal were evaluated. The model outputs included projected number of AIDS and CHD events, life expectancy and incremental cost-effectiveness ratios (ICER). The model was evaluated for internal and external validity and extensive sensitivity analyses were conducted. RESULTS: The predicted prevalence of patients in different health states over time provided a good fit to the clinical trial data. Strategies where raltegravir was included in the initiating therapy followed by an NNRTI or PI based regimen against efavirenz or PI based initiating therapies followed by raltegravir resulted in longer undiscounted life expectancy [20.07 Yrs. vs. 18.88 Yrs] and demonstrated costeffectiveness [ICERs: c1,078–c33,406/QALY]. CONCLUSIONS: Continuous-time Markov modeling based on transition matrices estimated using robust statistical methodology is a more realistic and sound approach. Results suggest initiating therapy on raltegravir generated lower costs and higher survival versus saving the drug for later lines of therapy in Portugal. PIN36 ECONOMIC EVALUATION OF AZITHROMYCIN COMPARED WITH OTHER ANTIBIOTICS FOR COMMUNITY-ACQUIRED PNEUMONIA AND SINUSITIS TREATMENT Sveshnikova ND1, Omelyanovsky VV2, Khachatryan G1, Ilkovitch J3 1 Institute of Clinico-Economic Expertise and Pharmacoeconomics, RSMU, Moscow, Russia; 2 Research Center for Clinical and Economic Evaluation and Pharmacoeconomics, Moscow, Russia; 3Research Institute for Pulmonology of Pavlov’ State Medical University, Saint-Petersburg, Russia OBJECTIVES: To assess clinical efficacy and economic effectiveness of Azithromycin compared with other antibiotics for community-acquired pneumonia (CAP) and sinusitis treatment. METHODS: Decision tree was used to calculate costs of antibiotic treatment. The chance nodes included treatment failures and introduction of secondline therapy or hospitalization. Transition probabilities were obtained from clinical trials. Preferred antibiotics for comparison with azithromycin(extended release, 2.0 g.) and for second-line treatment were chosen by experts in a survey. Medical care costs were derived from Moscow mandatory medical insurance system. Costs of the medications were obtained from consumer prices database. Sensitivity analysis was carried out. RESULTS: Clinical efficacy was equal. Costs of CAP treatment were 4390 rub (142.14$) for Amoxycillin and clavulanic acid, 5386 Rub ($174.39) for Cefuroxime axetil and 4053 Rub ($131.23) for Azithromycin. Costs of sinusitis treatment were 3637(117.76), 3728 (120.71) and 3570 rub ($115,59) for Amoxycillin and clavulanic acid, Cefuroxime axetil and Azithromycin respectively. CONCLUSIONS: Azithromycin is more efficient option for CAP and sinusitis treatment than Cefuroxime axetil and Amoxycillin and clavulanic acid in Russia. PIN37 HEALTH ECONOMIC EVALUATION OF CONJUGATE PNEUMOCOCCAL VACCINES IN LATIN AMERICAN COUNTRIES Gomez JA1, Naranjo LT2, Knerer GX3 1 GlaxoSmithKline, Victoria, Buenos Aires, Argentina; 2GlaxoSmithKline, Panama City, Panama; 3 GlaxoSmithKline, Wavre, Belgium OBJECTIVES: To simulate the potential benefits of implementing infant vaccination with the 10-valent pneumococcal & non-typeable Haemophilus influenzae (NTHi)protein D conjugate vaccine (PHiD-CV) or the 13-valent pneumococcal conjugate vaccine (PCV13) in Latin American countries. METHODS: Vaccine impact was assessed using a Markov cohort model for Mexico, Brazil, Chile, and Colombia. The model simulates the burden of pneumococcal- and NTHi-related diseases (Invasive Disease (ID), Community Acquired Pneumonia (CAP) and Acute Otitis Media (AOM)) in a birth cohort followed over a lifetime. Epidemiology, disease management and costs are country-specific. Vaccination schemes (3 + 1) at 90% coverage & price parity were compared to no intervention. Future QALYs and costs discounted at 3.5%, using the health care payer perspective are presented. RESULTS: Mortality impact on ID and CAP for the two vaccines is projected to be comparable under base-case conditions which include minimum assumptions of NTHi infection rates. Vaccines are predicted 13th Euro Abstracts to reduce 12.8 to 39.0 deaths (PCV13) and 12.4 to 37.3 deaths (PHiD-CV) per 100,000 vaccinated children. The model predicts that PHiD-CV will prevent 93 to 494 additional Myringotomies and 651 to 8,314 additional AOM cases per 100,000 vaccinated children, when compared with PCV13. Medical costs averted are estimated similar for ID and CAP. The model predicts that PHiD-CV will prevent 48 to 116% more AOM associated costs to the health care system over lifetime than PCV13. In the scenarios analyzed, both vaccines are cost-effective but PHiD-CV generates more QALYs gains (range 0.04%–14.5%) and is cost saving (range 1.88–12.54%) compared to PCV13. CONCLUSIONS: The model shows both vaccines would significantly reduce the clinical & economic burden of Pneumococcal disease and are cost-effective for Latin American countries. Nevertheless, due to its greater impact on AOM-related cases and costs, PHiD-CV would generate more QALY gains and is cost-saving to the health care system compared to PCV13. PIN38 AN ECONOMIC EVALUATION OF THE PEDIATRIC VACCINATION SCHEDULE IN THE UNITED STATES Clements KM1, Misurski DA2, Miller J3, Skornicki ME1, Hill GJ1, McGarry L1 1 i3 Innovus, Medford, MA, USA; 2GlaxoSmithKline, Philadelphia, PA, USA; 3GlaxoSmithKline Biologicals, King of Prussia, PA, USA OBJECTIVES: To estimate the cost-effectiveness of the recommended US pediatric vaccination schedule from a public health perspective. METHODS: An Excel-based cost-effectiveness calculator was constructed for the current pediatric vaccine schedule including: diphtheria, tetanus, and pertussis (DTaP), measles, mumps, and rubella (MMR), polio, hepatitis A, hepatitis B, haemophilus influenza B (Hib), varicella, pneumococcal, adolescent meningococcal, influenza, human papillomavirus (HPV) and rotavirus vaccines. Estimates of the incremental direct medical costs and qualityadjusted life-years (QALYs) for vaccination versus no vaccination were obtained from published literature. Where estimates were not available, a decision tree was constructed to model QALYs gained per vaccinated child. The tree includes branches for disease incidence pre- and post-vaccine introduction, case-fatality, and permanent, serious sequelae. 2008–2009 vaccination coverage and 2009 prices were used to estimate vaccine costs. Estimates of lifetime costs and QALYs per vaccinated child, discounted at 3% annually, were applied to a US birth cohort, assuming direct effects only. Costs are expressed in 2009 US$. Incremental costs and QALYs for individual vaccines and the whole schedule were evaluated. The model assessed schedule completion with single disease vaccines as well as completion with two different pentavalent combination vaccines (DTaP, polio, Hib or DTap, polio, hepatitis B) plus single disease vaccines. RESULTS: Regardless of how the current pediatric vaccine schedule is completed, annual estimated cost savings range between $13.8 billion to $14.3 billion. DTaP, MMR, polio, Hib and HPV are cost-saving, as are both pentavalent vaccines. Other single-disease vaccines add cost but contribute to a total of 1.4M QALYs gained per year. CONCLUSIONS: The current US pediatric vaccine schedule is estimated to be cost-saving and to provide substantial benefits in quality-adjusted survival. Use of combination vaccines increases the savings. Neither herd immunity nor indirect costs were considered in the model; their inclusion likely would increase the estimated cost savings. PIN39 COST-EFFECTIVENESS OF NUCLEIC ACID TEST SCREENING IN BLOOD DONATION FOR HIV IN BRAZIL Araújo MAM1, Soares DP1, Garcia GC1, Martins ACM2 1 ANVISA, Brasilia, Brazil; 2ANS, Brasilia, Brazil OBJECTIVES: To build a Markov Model that is able to assess the cost-effectiveness of adding NAT to the HIV screening strategy at the Brazilian public health system. METHODS: A mathematical model was made of the transfusion chain from donors to recipients of blood in Brazil. The annual number of avoided HIV transmissions was estimated with the window-period incidence model. The natural history of the whole blood receptors is described by a Markov model. RESULTS: The incremental costeffectiveness ratio—ICER of using NAT instead of ELISA concurrently to another ELISA in Brazilian public health system is R$666.493,56 per QALY. The Brazilian gross domestic product per capita is R$18.315,50. CONCLUSIONS: The Markov Model built is consistent and shows that, despite the narrower window period, at current costs, NAT is not cost-effective for HIV screening of donated blood in Brazil. Given that the NAT kit is already being produced in-house, the price per kit can be adjusted to achieve a better ICER. PIN40 COST-EFFECTIVENESS OF PNEUMOCOCCAL VACCINATION AMONG INFANTS IN RUSSIA: ECONOMIC EVALUATION OF THE PNEUMOCOCCAL 7-VALENT CONJUGATE VACCINE Omelyanovsky VV, Krysanov I, Ivakhnenko O Research Center for Clinical and Economic Evaluation and Pharmacoeconomics, Moscow, Russia OBJECTIVES: Cost-effectiveness analysis of 4 dose (3 + 1) schedule of the conjugate pneumococcal 7-valent vaccine (PCV-7) in infants in the Russian Federation. METHODS: Costs associated with Streptococcus pneumoniae infection were calculated in a modeled cohort of children 0–5 years old with and without vaccination with PCV-7. Key parameters in the model included: number of children in the age of 0–5 years in the country; incidence of diseases caused by S. pneumoniae; the data about efficacy of PVC-7. Costs of vaccination, medical care costs and economic losses of a A437 society were taken into account from the societal point of view. Time horizon was 5 years in the model. RESULTS: Expected cost of vaccination program is c0.469 million. Implementation of vaccination with PCV-7 will decrease direct medical costs of care for S. pneumoniae infection by c0.516 million. Societal economic losses will decrease by c0.975 million. Thus benefit amount of the PCV-7 vaccination program in a cohort of 0–5 years old children will be c1022 million. CONCLUSIONS: Vaccination with PCV-7 is an efficient program in Russia. PIN41 ADAPTATION OF A TRANSMISSION DYNAMIC MODEL FOR THE QUADRIVALENT HPV VACCINE TO GERMANY Schobert D1, Schmitter S2, Remy V3, Schöffski O4 1 Universität Erlangen-Nürnberg, Nürnberg, Germany; 2Sanofi Pasteur MSD GmbH, Leimen, Germany; 3Sanofi Pasteur MSD, Lyon, France; 4Friedrich-Alexander-Universität ErlangenNürnberg, Nuremberg, Germany OBJECTIVES: The impact of the HPV(6,11,16,18) vaccine to 12-year-old girls has been assessed in Germany by using a static model. However this kind of model cannot consider a change in HPV infection rate over time, as observed with high coverage rates. Further, this model does not adequately reflect the recommendation in Germany to vaccinate 12 to 17-year-old-girls. a transmission dynamic model was developed for the USA to include these features. Our objective was the adaptation of this model to Germany to precise the assessment of HPV-vaccination impact. METHODS: In a first step we assessed the transferability of the model structure to Germany. In a second step we checked input parameters for transferability and identified parameters for adaptation. For the identified parameters, we performed a comprehensive literature research, supplemented by expert opinions to determine German-specific values. The model was manually calibrated to fit observed data in Germany. Calibration parameters were number of annual cases of genital warts and cervical cancer as well as deaths due to cervical cancer to reflect the entire time span of occurrence of HPVrelated diseases. RESULTS: The US model structure is also applicable to the German context. Components of the model that had to be adapted include demographics (e.g. mortality), screening participation, treatment and vaccination strategies, sexual behaviour, health utility and economic input parameters. In case no German-specific data could be found, we used data from the UK and the US. Annual number of incidental genital warts was calibrated easily and fits well with observed data. However, changes in a variety of parameters were necessary for calibration of cervical cancer cases and related deaths. CONCLUSIONS: After its successful adaptation, this transmission dynamic model can be used for a far more realistic estimation of the clinical and economic impact of HPV vaccination in the German context. PIN42 THE COST-EFFECTIVENESS OF 13-VALENT PNEUMOCOCCAL CONJUGATE VACCINE (PCV13) COMPARED WITH 10-VALENT PNEUMOCOCCAL CONJUGATE VACCINE (PCV10) IN TAIWAN Chang CJ1, Wu DBS2, Wu CL1, Strutton D3, Hwang S3, Huang VWH4, Rubin J5, Gilmore K5 1 Chang Gung University, Taoyuan, Taiwan; 2National Yang-Ming University, Taipei, Taiwan; 3 Pfizer Limited, Collegeville, PA, USA; 4Pfizer Limited, Taipei, Taiwan; 5i3 Innovus, Medford, MA, USA BACKGROUND: Streptococcus pneumonia causes invasive diseases as meningitis and bacteremia and non-invasive diseases as pneumonia and acute otitis media (AOM), leading to high morbidity and mortality in infants and the elderly worldwide. OBJECTIVES: To evaluate the cost-effectiveness of universal infant vaccination with 13-valent pneumococcal conjugate vaccine (PCV13) compared with PCV10 in Taiwan. METHODS: A Markov model was developed to evaluate the potential public health and economic impact of PCV13 versus PCV10 when used as routine vaccination of infants in Taiwan with 4 doses at 2, 4, 6, and 12–15 months of age over a 10-year time horizon. We included both direct and potential indirect benefits of the vaccine from societal perspective. Direct effectiveness of PCV13 and PCV10 is estimated from clinical trial data while indirect effectiveness is estimated from U.S. surveillance data. Epidemiology, serotype, medical, and non-medical cost are from Taiwan CDC report, and retrospective Taiwan-population-based insurance database. Other model parameters were captured by published sources, unpublished data, and assumptions made in consultation with clinical experts. Probabilistic sensitivity analyses was performed to test the robustness of model assumptions. RESULTS: At vaccination price (PCV13 cost used current PCV7 price of NT$3,200 (US$ 98) and PCV10 NT$2,700 (US$ 83), our model predicts that, compared to PCV10, universal infant PCV13 vaccination would avoid 2,215 cases of IPD, approximately 12,473 and 14,018 cases of hospitalized and non-hospitalized pneumonia, 246,578 cases of AOM; prevent 207 deaths from IPD and 85 deaths from hospitalized pneumonia; resulting in 4,596 life-years saved, and 3,359 QALYs gained. Comparing PCV13 to PCV10 results in NT$ 187,462 (US$ 5,768) per life-year saved and cost of NT$ 253,307 (US$ 7,794) per QALY gained from the societal perspective. CONCLUSIONS: Universal pediatric PCV13 vaccination in Taiwan is estimated to reduce the burden of pneumococcal disease and expected to be cost-effective from the societal perspective compared with PCV10. A438 PIN43 A COST-EFFECTIVENESS ANALYSIS OF ANTIMICROBIAL THERAPY OF BLOODSTREAM INFECTIONS TREATMENT IN INTENSIVE CARE UNIT Omelyanovsky VV, Protsenko DN, Derkach EV, Sveshnikova ND Russian State Medical University, Moscow, Russia OBJECTIVES: According to the recent international data, the incidence of bloodstream catheter—associated infections is 3–8% in the structure of hospital- acquired infections in the intensive care units (ICU) and the leading cause of them are Gram positive bacteria. In this study we conducted comparative analysis of clinico—economic effectiveness of daptomycin (dapto) vs. vancomycin (vanco) usage in treatment of patients with MRSA catheter—associated infection in the ICU. METHODS: “Decision Tree” pharmacoeconomic model was built based on results of international clinical studies and data of routine practice treatment of bloodstream infections in Moscow clinics. Two variants of antibacterial treatment of patients with catheter-associated infections differing on starting products (dapto or vanco) were assessed. If the first line of therapy was ineffective, patients switched over to the second line therapy covering resistant strains—meropenem and fluconazole. Direct and indirect medical costs were assessed: cost of antibiotics and additional medical treatment, antibacterial diagnostic, laboratory and cost of bed-days in ICU and therapeutic department. Costs were based on official data on hospital medical service in municipal Moscow clinics and purchasing price on medical products from price list of the biggest Russian pharmaceutical distributors. Clinical recovery was considered as efficacy with the goal to evaluate cost-effectiveness ratio (CER) of two groups (CERdapto and CERvan accordingly). RESULTS: Better clinical efficacy in daptomycin group resulted in lower— needs to change antibiotic in catheter—associated infection treatment in the ICU in comparison to vancomycin group. In spite of the higher price of drug, average cost of successfully treated patient by daptomycin (CERdapto) was 227,887 RUR/pt in compare with CERvan 235,032 RUR/pt. Exchange rate is 1USD = 30 RUR. CONCLUSIONS: Good cost-effectiveness ratio in comparison to vancomycin supports use of daptomycin as the first line antibacterial therapy in bloodstream catheter—associated infections. PIN44 COST-EFFECTIVENESS ANALYSIS OF 1 YEAR PEGINTERFERON ALFA-2A VERSUS 4 YEARS ENTECAVIR FOR THE TREATMENT OF HBEAG-POSITIVE CHRONIC HEPATITIS B IN CHINA Chen W Fudan University, Shanghai, China OBJECTIVES: The objective of the study was to evaluate the cost-effectiveness of 1 year peginterferon alfa-2a compared to 4 years entecavir for the treatment of HBeAgpositive chronic hepatitis B in China. METHODS: A Markov health-state model was designed to evaluate the direct medical costs and outcomes (life-years and QALYs gained) of treating HBeAg-positive chronic hepatitis B in China. The model included 10 health states—Chronic hepatitis B (CHB), HBeAg seroconversion, HBsAg loss, CHB with resistance, Compensated cirrhosis, Decompensated cirrhosis, Hepatocellular carcinoma, Liver transplant, Post-liver transplant and death. The model incorporates a maximum analysis time horizon of 80 years with yearly cycles. The clinical and quality of life data were obtained from published literature and re-confirmed based on a questionnaire survey from a clinical expert panel of 20 hepatitis B specialists. From the perspective of China’s health insurance system, cost data was calculated based on the published literature about economic burden of chronic hepatitis B. a discounting rate at 3% was used to discount utilities and medical costs happened at different years. a univariate sensitivity analysis was performed to understand the key drivers and general sensitivity of the model. RESULTS: The model results showed that the utilization of 1 year peginterferon alfa-2a treatment for HBeAg-positive CHB can prolong 0.32 QALYs, compared to the 4 years entecavir treatment (16,394 QALYs vs. 16,074 QALYs). The total cost per patient treated with peginterferon alfa-2a was US$22,221, and US$19,989 for patient treated with entecavir. The discounted incremental cost per QALY gained for pegainterferon alfa-2a was US$6965. CONCLUSIONS: The results of the model suggest that 1 year pegainterferon alfa-2a improves health outcomes in a cost-effective manner compared with 4 years entecavir in the treatment of HBeAg-positive chronic hepatitis B in China. PIN45 COST-EFFECTIVENESS ANALYSIS OF TREATMENT OF CHRONIC HEPATITIS C PATIENTS WITH PEGINTERFERON ALFA-2A OR PEGINTERFERON ALFA-2B BOTH PLUS RIBAVIRIN IN SPAIN Ventayol-Bosch P1, Rubio-Terrés C2, García-Samaniego J3, Planas R4, Solà-Lamoglia R5, Romero-Gómez M6, Diago-Madrid M7, Crespo-García J8, Calleja-Panero JL9, Turnes-Vázquez J10 1 Hospital Universitario Son Dureta, Palma de Mallorca, Spain; 2HealthValue, Madrid, Spain; 3 Hospital Carlos III, CIBEREHD, Madrid, Spain; 4Hospital Germans Trias i Pujol, CIBEREHD, Barcelona, Spain; 5Hospital del Mar (IMIM), Barcelona, Spain; 6Hospital Universitario de Valme, CIBEREHD, Sevilla, Spain; 7Hospital General Universitario de Valencia, Valencia, Spain; 8 Hospital Universitario Marqués de Valdecilla, Santander, Spain; 9Hospital Universitario Puerta de Hierro, Madrid, Spain; 10Complejo Hospitalario de Pontevedra, Pontevedra, Spain OBJECTIVES: To estimate long-term cost-effectiveness of treatment of chronic hepatitis C (CHC) patients with peginterferon alfa-2a (180 mcg/week) versus peginterferon alfa-2b (1.5 mcg/kg/week) both in combination with ribavirin (800–1400 mg/day) from the Spanish National Healthcare System perspective. METHODS: A metaanalysis of head-to-head randomized trials of peginterferon alfa-2a and peginterferon 13th Euro Abstracts alfa-2b both plus ribavirin, evaluating sustained virological response (SVR) has been recently published. It showed RR = 1.11 (95%CI 1.04–1.19) for all genotypes, RR = 1.21 (95%CI 1.03–1.42) for G-1/4, and RR = 1.11 (95%CI 1.02–1.22) for G-2/3. a Markov model with 7 health states was developed to simulate the disease progression of adult patients with CHC for a lifetime horizon. Efficacy, in terms of SVR, was calculated from the meta-analysis, showing a better SVR rate for peginterferon alfa-2a than for peginterferon alfa-2b: the absolute differences were of 6.0%; 7.6%; and 8.7% for all genotypes, G-1/4 and G-2/3 respectively. Transition probabilities and health states utilities were obtained from published literature. Health direct costs of peginterferon + ribavirin (48 weeks for G-1/4 and 24 weeks for G-2/3) and of disease complications were collected from Spanish databases and studies ([c] 2010). The annual discount rate was 3.5% for costs and outcomes. RESULTS: Each patient gained 0.469, 0.600 and 0.685 life-years (LY) and 0.155, 0.198 and 0.227 qualityadjusted life-years (QALY) with peginterferon alfa-2a in comparison with peginterferon alfa-2b, for all genotypes, G-1/4 and G-2/3 respectively. The savings per patient treated with peginterferon alfa-2a were c705, c672 and c1900, for all genotypes, G-1/4 and G-2/3, respectively. Peginterferon alfa-2a was the dominant treatment strategy (lower costs and higher efficacy than peginterferon alfa-2b treatment). CONCLUSIONS: Treatment of patients with chronic hepatitis C with peginterferon alfa-2a + ribavirin is a cost-effective strategy in comparison with peginterferon alfa-2b + ribavirin for all genotypes, G-1/4 and G-2/3. PIN46 KIOVIG FOR PRIMARY IMMUNODEFICIENCY: REDUCED INFUSION AND DECREASED COSTS PER INFUSION Connolly M1, Simoens S2 1 Global Market Access Solutions, St Prex, Switzerland; 2K.U. Leuven, Leuven, Belgium OBJECTIVES: Kiovig is a new, ready-to-use 10% liquid immunoglobulin preparation that is medically indicated for the treatment of primary immunodeficiency. This study aims to conduct an economic evaluation which compares the intravenous immunoglobulin (IVIg) preparations Kiovig, Multigam, and Sandoglobulin from the Belgian societal perspective. METHODS: Given that three prospective studies have observed no difference in outcomes, a cost-minimization analysis considered the differences in costs that can arise from these immunoglobulin products. a decision-analytic model simulated immunoglobulin treatment over a one-year time horizon. Cost items included immunoglobulin costs, pharmacy administration and nursing costs, miniforfait paid for hospital infusion, costs of adverse events, and lost productivity. Cost data were identified from published sources and Belgian hospital administrators. a probabilistic sensitivity analysis explored the impact of parameter uncertainty. The price year was 2009. RESULTS: Costs per infusion cycle in adult primary immunodeficiency patients were c1.046 (95% confidence interval: c1.006–1.093) with Kiovig; c1.102 (c1.064–1.147) with Multigam; and c1.147 (c1.108–1.193) with Sandoglobulin. The average cost savings per infusion with Kiovig as compared to Multigam and Sandoglobulin amounted to c56 and c101 per infusion. CONCLUSIONS: Treatment costs with Kiovig were shown to be lower as compared to other IVIgs in Belgium. Reduced costs per infusion were attributed to lower costs associated with treating adverse events and the opportunity cost of nursing time and time off work for working adults. PIN47 COST-EFFECTIVENESS ANALYSIS OF PNEUMOCOCCAL VACCINES IN TAIWAN Chang CJ1, Wang PC2, Huang YC3, Wu BS4, Wu CL1, Topachevskyi O5, Jung W6 1 Chang Gung University, Taoyuan, Taiwan; 2Cathay General Hospital, Taipei, Taiwan; 3Chang Gung Memorial Hospital, Taoyuan, Taiwan; 4National Yang-Ming University, Taipei, Taiwan; 5 GlaxoSmithKline Biologicals, Wavre, Belgium; 6GlaxoSmithKline, Singapore, Singapore OBJECTIVES: To evaluate cost-effectiveness of pneumococcal non-typeable Haemophilus Influenza protein D conjugate vaccine(PHiD-CV), 7-valent pneumococcal conjugate vaccine(PCV7), and PCV13 in Taiwan. METHODS: An age-compartmental, deterministic, static cohort model simulated in a 198,733 birth cohort the disease process of invasive disease(ID, meningitis and bacteremia), community acquired pneumonia(CAP), and acute otitis media(AOM) over life-time. The model was developed by GlaxoSmithKline and adapted with local data. For base-case analysis for all vaccines, a 4-dose(3 + 1) schedule was assumed with 95% vaccine coverage. Cost and outcomes were analyzed from health care payer perspective with 3% discount rate. Herd protection on ID, limited cross protection against 6A and 19A, minimal estimation of NTHi(non-typeable Haemophilis Influenzae) infection rates in ID and CAP were assumed for base case. Costs for PHiD-CV, PCV7 and PCV13 were assumed parity at 3200 New Taiwan Dollar(NTD). Epidemiological and cost data were obtained from local disease burden study. Vaccine efficacy data were obtained from published sources. One way and probabilistic sensitivity analyses were conducted. RESULTS: Compared with PCV7, PHiD-CV is expected to prevent additional cases of 1 ID, 2,291 CAP and 63,597 AOM, with 548 additional quality-adjusted lifeyears(QALY) gained with total saving of NTD 69,347,211. Compared with PCV13, PHiD-CV is expected to prevent less cases of 4 ID and 264 CAP, but expected to prevent additional 45,111 AOM cases. PHiD-CV is expected to provide additional 181 QALY gained with total saving of NTD 14,532,124 compared to PCV13. Sensitivity analyses show the results are most sensitive to the changes of AOM related parameters, but when the AOM related parameters were changed up to +/− 20%, PHiD-CV is still cost-saving to PCV7 and PCV 13. CONCLUSIONS: PHiD-CV is expected to provide more QALYs with potential saving of total health care cost 13th Euro Abstracts compared with PCV7 and PCV13. At price parity, PHiD-CV is cost-saving to PCV7 and PCV13 in Taiwan. PIN48 PHARMACO-ECONOMICS OF ANTIBIOTICS Simoens S K.U. Leuven, Leuven, Belgium OBJECTIVES: Antibiotics have made a significant contribution to improving patient health, but policy makers and health care payers are concerned about the costs of antibiotics in addition to their effectiveness. This study aims to assess the value of antibiotics by examining published incremental cost-utility ratios of antibiotics. METHODS: Evidence was derived from cost-utility analyses of antibiotics included in the Tufts-New England Center Cost-Effectiveness Analysis Registry through September 2009. For each cost-utility analysis, the following variables were examined: publication year, target population, intervention type, country of patient sample, disease classification, prevention stage, funding source, study perspective, discounting, sensitivity analysis, incremental cost-utility ratio, and methodological quality. Evidence of the value of antibiotics was summarized by calculating median incremental cost-utility ratios and frequency distributions. Associations between incremental costutility ratios on the one hand and the prevention stage, study perspective and methodological quality were examined by means of the Mann-Whitney U-test for ordinal variables. RESULTS: The analysis included 85 incremental cost-utility ratios from 23 cost-utility analyses. The findings showed that 38.8% of incremental cost-utility ratios related to dominant antibiotics; 45.9% referred to antibiotics that improved effectiveness, but also increased costs; and 15.3% related to dominated antibiotics. The median ratio was c748 per quality-adjusted life-year. Using threshold values of c20,000 per quality-adjusted life-year and c50,000 per quality-adjusted life-year, the probability that an antibiotic provides value for money was 64% and 67%, respectively. No statistically significant association was observed between incremental cost-utility ratios and the prevention stage (p = 0.119), study perspective (p = 0.285) or methodological quality (p = 0.146). CONCLUSIONS: The current evidence base suggests that the majority of antibiotics provide value for money. PIN49 COST-EFFECTIVENESS OF PEGINTERFERON ALPHA-2A PLUS RIBAVIRIN FOR TREATING CRONIC HEPATITIS C VIRUS INFECTION COMPARED WITH NO TREATMENT IN MEXICO Carlos F1, DeHesa2 1 R a C Salud Consultores S.A. de C.V., México City, D.F., Mexico; 2Hospital de Especialidades Centro Médico Nacional Siglo XXI, Instituto Mexicano del Seguro Social, Mexico City, D.F., Mexico OBJECTIVES: Prevalence of hepatitis C virus infection is approximately 2.2–3.0% worldwide (130–170 million people). a persistent infection develops in up to 85% of these patients, leading to chronic hepatitis C (CHC), a condition associated with serious liver-related complications. We aimed to perform an economic evaluation of peginterferon alpha-2a (PEG-IFN alpha-2a) plus ribavirin (RBV). METHODS: We developed a Markov model with 40 annual cycles to project cumulative cost and quality-adjusted life-years (QALY) for two identical cohorts of patients aged 40 years. One cohort received PEG-IFN alpha-2a 180 mcg per week plus daily doses of 1200 mg of RBV during 48 weeks for genotype 1/4 or 24 weeks for genotype 2/3. The other cohort did not receive any antiviral treatment. The analysis was performed under the perspective of national public health care system. Only direct medical costs were accounted for; these included acquisition cost of antiviral drugs and medical attention for health states incorporated into the model. Costs (expressed in 2010 Euros) and QALY were discounted at an annual rate of 5%. Transition probabilities and utility scores were gathered from published literature and cost data was based on local sources and experts’ opinion. RESULTS: Average discounted costs were estimated at c15,626 for PEG-IFN alpha-2a plus RBV and at c17,350 when no antiviral treatment is given to CHC patients, leading to overall savings of c1724 per patient. Without antiviral treatment, 9.9 QALY per patient are expected. There is a gain of 2.2 QALY for patients who are treated. Results are robust to variations in model parameters. PEG-IFN alpha-2a plus RBV was both more effective and less costly than no treatment in more than half of the simulations performed in a probabilistic sensitivity analysis. CONCLUSIONS: PEG-IFN alpha-2a plus RBV is a dominant strategy compared to given no antiviral treatment to CHC patients in Mexico. PIN50 COST-EFFECTIVENESS OF PEGINTERFERON ALPHA-2A VERSUS PEGINTERFERON ALPHA-2B FOR TREATMENT OF CHRONIC HEPATTIS C INFECTION IN MEXICO Carlos F1, DeHesa2 1 R a C Salud Consultores S.A. de C.V., México City, D.F., Mexico; 2Hospital de Especialidades Centro Médico Nacional Siglo XXI, Instituto Mexicano del Seguro Social, Mexico City, D.F., Mexico OBJECTIVES: Prevalence of hepatitis C virus infection is approximately 2.2–3.0% worldwide (130–170 million people). A persistent infection develops in up to 85% of these patients, leading to chronic hepatitis C (CHC), a condition associated with serious liver-related complications. Current standard of treatment includes Peginterferon (PEG-IFN) alpha plus ribavirin (RBV). We aimed to compare two different options of PEG-IFN alpha. METHODS: We developed a Markov model with 40 annual cycles to project cumulative cost and quality-adjusted life-years (QALY) for A439 two identical cohorts of moderate CHC patients without cirrhosis aged 40 years. One cohort received PEG-IFN alpha-2a 180 mcg per week and the other received PEG-IFN alpha-2b 1.5 mcg/Kg weekly, both combined with daily doses of 1,200 mg of RBV during 48 weeks for genotype 1/4 or 24 weeks for genotype 2/3. The analysis was performed under the perspective of national public health care system. Only direct medical costs were accounted for; these included acquisition cost of antiviral drugs and medical attention for health states incorporated into the model. Costs (expressed in 2010 Euros) and QALY were discounted at an annual rate of 5%. Transition probabilities and utility scores were gathered from published literature and cost data was based on local sources and experts’ opinion. RESULTS: Average discounted costs were estimated at c16,854 for PEG-IFN alpha-2a plus RBV and at c18,247 for PEG-IFN alpha-2b plus RBV, leading to overall savings of c1,393 per patient when PEG-IFN alpha-2a is used. Discounted QALY were 12.29 for PEG-IFN alpha-2a and 12.17 for PEG-IFN alpha-2b. Results are robust to variations in model parameters. CONCLUSIONS: PEG-IFN alpha-2a plus RBV is a dominant strategy compared to given PEG-IFN alpha-2b treatment to CHC patients in Mexico. PIN51 UPDATING THE COST-EFFECTIVENESS OF ROTAVIRUS VACCINATION IN THE NETHERLANDS Rozenbaum MH1, Hak E1, Wilschut JC1, Postma MJ2 1 University of Groningen, Groningen, Groningen, The Netherlands; 2University of Groningen, Groningen, The Netherlands OBJECTIVES: To investigated the most important factors responsible for the large differences in previously estimated cost-effectiveness ratios and to update the costeffectiveness of rotavirus vaccination in the The Netherlands applying assumptions resulting from 2 consensus meetings with national and international experts in the field, from academia, clinical environments, industry and health policy. METHODS: We constructed a decision analytic model to compare the expected net costs and health benefits over a period of 5 years in two hypothetical cohorts of 180,000 children (approximating the Dutch birth cohort), one being vaccinated and one unvaccinated. The base-case analysis reflected the most likely—but also overall conservative—estimate of cost-effectiveness from the societal perspective. Robustness of the base-case result was investigated in sensitivity and scenario analyses. RESULTS: In the base-case analysis, it was estimated that approximately 59,495 RVGE cases would occur, resulting in 11,453 GP visits and 3,238 hospitalizations. With vaccination, approximately 34,000 cases of RVGE cases are averted corresponding to a total QALY gain of 167. Assuming a total cost of vaccination of c75, vaccination would result in cost-effectiveness of c30,540 per QALY gained (c152 per case averted). Results were sensitive to the number of deaths due to RVGE, inclusion of potential herd protection, inclusion of QALY decrements of care givers, further potential tender price reduction and the exact discount rate used. CONCLUSIONS: Our economic analysis indicates that a potential national immunization programme against rotavirus can be considered costeffective if applying a threshold of c50,000 per QALY for the The Netherlands. PIN52 AN ECONOMIC EVALUATION OF RALTEGRAVIR FOR THE TREATMENT OF ANTIRETROVIRAL-NAIVE HIV-1 INFECTED PATIENTS IN HUNGARY Erdesz D1, Nagy L1, Brandtmuller A1, Kiss Z1, Chaudhary MA2, Kumar RN3, Elbasha EH2 1 MSD Hungary Ltd., Budapest, Hungary; 2Merck Research Laboratories, North Wales, PA, USA; 3Merck & Co., Inc., Whitehouse Station, NJ, USA OBJECTIVES: Raltegravir (Merck & Co., Inc.) is an inhibitor of HIV type 1 (HIV-1) integrase, and is the first drug in a new class of therapy known as the integrase inhibitors. Raltegravir has already shown to be cost-effective in HIV treatment experienced patients in Hungary. The objective of the current study is to assess if the first line use of raltegravir is cost-effective compared to its use in rescue therapy from the Hungarian public payer perspective. METHODS: First line use of raltegravir was evaluated versus a protease inhibitor (PI) using a Markov model. Raltegravir was also included as a 3rd line therapy within the model arm that initiates on a PI. The Markov process comprised a three stage continuous-time model representing successive HIV therapies over a patient’s life-time. Patients moved between 18 health states—based on CD4 and HIV RNA levels. Patients progressed to the next stage after they either failed current therapy or discontinued for toxicity reasons. At any time, patients could develop acquired immunodeficiency syndrome (AIDS), suffer from a coronary heart disease (CHD) event and/or experience other adverse events. Mortality was also captured in the model. RESULTS: The incremental cost-effectiveness ratio (ICER) for initiating therapy with raltegravir versus using it as a rescue therapy was 4,075 million HUF per quality adjusted life-year gained (QALY), equivalent to $16,830/QALY. The model predicted lower cumulative incidence of CHD in the raltegravir arm versus the PI arm. (15.1% versus 16.1%). The model predicted that patients initiating on raltegravir therapy have longer life expectancy than patients starting with PI treatment over a 50-year time horizon (18.74 versus 17.17 years). CONCLUSIONS: Our long term economic model suggests that it is cost-effective to use raltegravir early in HIV therapy versus in patients who have experienced multiple failures. A440 PIN53 COST-UTILITY ANALYSIS OF NEW PNEUMOCOCCAL CONJUGATE VACCINES IN THE REGIONAL IMMUNIZATION PROGRAM OF THE AUTONOMOUS REGION OF MADRID—IMPACT ON INVASIVE PNEUMOCOCCAL DISEASE Picazo J1, Mendez C2, Oyagüez I3, Casado MA3, Guijarro P2 1 University Complutense of Madrid, Madrid, Spain; 2Pfizer Spain, Alcobendas, Madrid, Spain; 3 Pharmacoeconomics & Outcomes Research Iberia, Pozuelo de Alarcón, Madrid, Spain OBJECTIVES: The inclusion of conjugate pneumococcal vaccines to the Regional Immunization Program avoids cases, among others, of invasive pneumococcal diseases caused by serotypes of Streptococcus pneumoniae contained in the vaccines. Research objective was to assess efficiency of conjugate pneumococcal vaccine- 13 valent (PCV13) versus 10 valent (PCV10), under National Health System perspective. METHODS: A cost-utility model was developed to estimate cost per quality adjusted life-year (QALY) associated to primary bacteriemic, empyema, meningitis and bacteriemic pneumonia with vaccine administration in the Autonomous Region of Madrid. The estimation of cases to avoid and serotype coverage was based on epidemiologic information available at regional level. Additionally to direct protection, the model considers the potential indirect effect on the general population. One-way sensitivity analyses were performed, including parameters with most uncertainty. RESULTS: PCV13 yields 224 QALYs versus 70 QALYs with PCV10. The incremental ratio is c16,908/QALY gained. Deterministic sensitivity analysis showed that model results are robust, parameter that most influence on the results was vaccine indirect effect. Reducing PCV13 vaccine schedule from 4 to 3 doses was a dominant strategy. CONCLUSIONS: The inclusion of PCV13 is a cost-effective strategy versus PCV10. a 3 doses schedule of PCV13 is a dominant option, being a cost-saving measure for the National Healthcare System. PIN54 COST-EFFECTIVENESS ANALYSES (CEA) OF LOPINAVIR/RITONAVIR (LPV/R) AND ATAZANAVIR PLUS RITONAVIR (ATV + RTV) REGIMENS FOR ANTIRETROVIRAL (ARV) NAÏVE HIV-1 INFECTED PATIENTS BASED ON CASTLE 48-WEEK STUDY: APPLICATION TO SWEDEN Simpson K1, Baran R2, Woodward C3, Dietz B4 1 Medical University of South Carolina, Charleston, SC, USA; 2Abbott Laboratories, Abbott Park, IL, USA; 3Abbott Laboratories, Sinking Springs, PA, USA; 4Abbott GmbH & Co. KG, Ludwigshafen, Germany, Germany No differences in viral load (VL) or CD4+ count at 48 weeks were reported for the CASTLE study. However, total cholesterol (TC) levels were elevated in 7% and 18% of subjects receiving ATV + RTV and LPV/r, respectively. These measures can predict outcomes which affect the future cost of HIV in the Swedish health system. OBJECTIVES: To compare the incremental cost-effectiveness (ICER) and budget impacts for a population similar to those enrolled in the CASTLE study for Sweden. METHODS: Using a previously published Markov model of HIV disease and 2009 cost data from Sweden, we compared the cost/QALY and budget impact of the two ARV regimens. Daily drug costs were 160,52 SEK for ATV + RTV and 147,87SEK for LPV/r. Costing for other health care resources used a health systems perspective with 2009 inputs from www.fass.se and published literature. Costs and QALYs were discounted by 3% when calculating ICERs. RESULTS: The CHD risk favored ATV + RTV, resulting in a life expectancy increase of 0.031 QALYs (11 days). The cost-effectiveness ratio for ATV + RTV for Sweden was 1.251.545 SEK /QALY gained. Three times the Swedish GDP in 2008 was 886,670 SEK. Thus the modeled ICER exceeds the WHO criteria for cost-effectiveness by 40%. Sensitivity analysis showed the model was mainly sensitive to ARV price. Five and 10 years per-patient savings for subjects on the LPV/r regimen were estimated to be 21,314 SEK and 32,564 SEK, respectively. CONCLUSIONS: Selection of an ATV + RTV based regimen in an ARV-naïve population with a CHD risk similar to subjects in the CASTLE study does not appear to be a cost-effective use of scarce resources for the cost structure seen in Sweden. The costs associated with the very small added CHD risk incurred by LPV/r treatment are more than offset by its short and long term cost savings. PIN55 RISK SCORES PREDICTING RESOURCE USE OF HEPATITIS C PATIENTS IN GERMANY—AN ANALYSIS OF SICK FUND CLAIMS DATA Tomeczkowski J1, Slawik L1, Guthoff-Hagen S2, Schreder H1, Fleischmann J1 1 Janssen-Cilag GmbH, Neuss, Germany; 2Sgh-Consulting, Hamburg, Germany OBJECTIVES: To compare costs for hepatitis C patients by analyzing retrospective sick fund claims data. METHODS: A representative panel of 1,193,464 patients from several national governmental sick funds were analyzed for 2008. These claims data include all direct medical costs from hospitals, materials in the outpatient sector and medication (covering 60% of sick fund full costs). Selection criteria for patients were at least one confirmed out- or inpatient diagnosis for acute or chronic hepatitis C (ICD-10 GM B17.1 and B18.2). The sample was divided into two groups, those treated with pegylated Interferon/Ribavirin ((peg)INF/Rib) (standard of care for chronic hepatitis C patients according to German guidelines) and those without (identified by ATC Code L03AB and J05AB04). Of these patients all ICD-10 diagnoses were grouped by the DxCG software to hierarchical condition categories in order to calculate risk scores predicting resource use. The expected cost at the average for all insured persons analyzed in this German sick fund panel is represented by a risk score of 2.2. Risk scores and co-morbidities were compared with the sick funds average means. RESULTS: A total fo 2628 hepatitis C patients were identified (0.2% of the 13th Euro Abstracts panel), 245 with and 2383 without (peg)INF/Rib therapy (treatment rate 9.3%). Sixty-two percent of patients with hepatitis C suffered from at least 11 co-morbidities (risk score 8.5), while only 27% of all other patients from the panel suffered from at least 11 co-morbidities. Further analyses revealed additional pharmacy costs of c12,200,- for patients treated with (peg)INF/Rib compared to those without. The risk score for the second year after treatment was 5.7 for (peg)INF/Rib-treated versus 8.7 for (peg)INF/Rib-untreated patients. CONCLUSIONS: A diagnosis of hepatitis C is associated with high numbers of co-morbidities and high risk scores. Results suggest that treatment lowers risk scores but keeping them substantially above mean. INFECTION – Patient-Reported Outcomes Studies PIN56 REFILL-BASED ADHERENCE RATES OF ANTIRETROVIRAL MEDICATION USING RETROSPECTIVE MEDICINE CLAIMS DATA: A COST ANALYSIS Paterson T, Lubbe MS, Serfontein JHP, Du Plessis JM North-West University, Potchefstroom, South Africa OBJECTIVES: To determine the refill-based adherence rates and to compare the costs associated with an under- or over supply of antiretroviral medication. METHODS: A non-experimental, quantitative retrospective drug utilization review was performed on medicine claims data from one pharmacy benefit management company. The study population were selected from all patients (N = 15 901) who received more than one prescription for antiretroviral medication during a 36 consecutive month period (1 January 2005 to December 2007). An overall refill-based adherence rate was calculated by using the following equation: Refill-based adherence rate = (total days of antiretroviral items supplied—days supplied at the last refill)/(date last claimed—date first claimed). (RSA Rand(R)/$US = 6.8595 on 31 December 2007). RESULTS: Refillbased adherence rates were calculated for 41 967 antiretroviral drugs. Less than 50% (n = 17 267; 41.15%) of all antiretroviral drugs had acceptable adherence rates (between 90% and 110%). Antiretroviral drugs with adherence rates below 90% (possibly under-supplied) (n = 7243; 17.26%), accounted for 14.30% (n = R15 829 876.53) of the total cost of all antiretroviral drugs (N = R110 728, 214.00) while those that were possibly over-supplied (n = 17,454; 41.59%) accounted for 25.60% (R28 347 266.48). CONCLUSIONS: The calculated refill adherence rates indicated that most antiretroviral drugs were either possibly over- or under-supplied for the specific treatment period. PIN57 EVALUATION OF MEDICATION ADHERENCE IN PATIENTS RECEIVING ANTIRETROVIRAL THERAPY (ART) IN MAITAMA DISTRICT HOSPITAL, ABUJA NIGERIA Agu KA1, Okojie O2, Omonaiye O1, Oqua D1, King RC3, Onuoha C1, Muhammadu I1, Iyaji PG1 1 Howard University PACE Center, Abuja, FCT, Nigeria; 2University of Benin, Benin, Edo, Nigeria; 3Howard University PACE Center, Washington, DC, WA, USA OBJECTIVES: ART has dramatically increased the life expectancy of HIV-infected patients and its adherence has been strongly correlated with viral suppression, reduced rates of resistance, an increase in survival, and improved quality of life. The study evaluated medication adherence, the association of medication adherence with occupation and educational status and identified possible causes of non-adherence in HIVinfected patients after 6 months on ART. METHODS: In a cross-sectional survey, medication adherence of 118 HIV-infected patients was evaluated using a self-administered study-specific 16-items questionnaire after 6 months on ART. Patients’ selfreport adherence assessment method was used. Chi square statistics was used to test the association of adherence with occupation and education at 95% CI. RESULTS: The mean age of the 118 HIV-infected ART patients was 33.89 (95%CI, 29.63–38.15) years; and majority (82.2%) were between the ages of 26–45 years; 60.2% were females, 80.5% attained secondary education at the least, while 77.1% employed. All participants reported being counselled on the benefits of ART and medication adherence at ART initiation. On assessment of participants’ knowledge of the benefits of ART and medication adherence, 92.2% reported correctly, 2.9% reported wrongly that ART is a cure for HIV while 4.9% did not respond. Medication adherence level reported among participants was 79.1%. Educational status was not associated with adherence (p > 0.05) unlike the occupational status which was associated with adherence (P < 0.05). The major reasons reported for non-adherence were busy at work or school (33.1%), forgetfulness (15.5%), fasting (12.0%), and travelled or moved away from home (10.6%). CONCLUSIONS: The medication adherence level among participants was somewhat poor compared to the desired value of >95%; occupational status was associated with adherence unlike the educational status. Consequently, busy at work or school was a major reason for non-adherence to medication. Routine evaluation of medication adherence and intervention in clinical practice is recommended. 13th Euro Abstracts PIN58 REPORT ON A PILOT STUDY OF THE SIMPILL REALTIME MEDICATION ADHERENCE SYSTEM IN TUBERCULOSIS PATIENTS IN CAPE TOWN, SOUTH AFRICA Green D1, Tsodzo C2 1 Liberty Health Holdings, Hout Bay, South Africa; 2Ruhr-Universität Bochum, Bochum, Germany OBJECTIVES: SIMpill is a medication adherence management system using cellular networks to provide real-time information about, and support to, patients on long term medication. The SIMpill system is described in detail. METHODS: A pilot was carried out from July 1, 2006 to April 12, 2007 in Cape Town, South Africa. Patients diagnosed with Tuberculosis were taken onto the pilot as they arrived at the clinics until the desired number of patients was achieved. Contrary to the standard DOTS protocol, patients were allowed to self medicate either at their homes or their places of work, supported by the SIMpill system. RESULTS: Of the 155 TB patients enrolled in the pilot study the following results were recorded: 12 patients (8%) did not have adequate documentation of outcomes at the end of the pilot and are excluded from the data. One patient (0.8%) was recorded as a sputum positive treatment failure. Seven patients (3.3%) interrupted their treatment whilst four patients transferred to other clinics or died during the pilot study. a total of 131 patients (91.6% including unrelated interruptions, 94.2% excluding unrelated interruptions) were recorded as successful treatments. This is contrasted with the background treatment success rate under the DOTS regimen of 76.4% at the same time in the same communities. CONCLUSIONS: It is concluded that the SIMpill system is practical to implement, could dramatically improve treatment success in TB patients and warrants further study, both in TB and other conditions. PIN59 DEVELOPMENT OF A SPECIFIC QUESTIONNAIRE TO ASSESS TREATMENT ADHERENCE OF PATIENTS WITH HCV INFECTION: ADHEPTA STUDY Martín-Suárez JM1, Martín-Escudero V2, Pérez-Álvarez R3, Alcántara-Benítez R4, Castellano G5, de Cuenca B6, Turnes-Vázquez J7, Moreno D8, Delgado-Fernández M9, ÁlvarezGonzález MA10, Testillano-Tarrero M11, Castro-Iglesias MA12, Tural C13, Planas R13, Lara N14, Solà-Lamoglia R15 1 Hospital de Gran Canaria Dr. Negrín, Las Palmas de Gran Canaria, Spain; 2Roche Farma, Madrid, Spain; 3Hospital Universitario Central de Asturias, Oviedo, Spain; 4Hospital Clínico Virgen de la Victoria, Málaga, Spain; 5Hospital Universitario 12 de Octubre, Madrid, Spain; 6 Hospital Infanta Cristina, Madrid, Spain; 7Complejo Hospitalario de Pontevedra, Pontevedra, Spain; 8Hospital de Móstoles, Madrid, Spain; 9Hospital Carlos Haya, Málaga, Spain; 10Hospital Sant Joan de Deu Manresa, Barcelona, Spain; 11Hospital de Cruces, Barakaldo, Spain; 12 Hospital Universitario de a Coruña, a Coruña, Spain; 13Hospital Germans Trias i Pujol, CIBEREHD, Barcelona, Spain; 14Health Economics Outcomes Research, IMS Health S.A., Barcelona, Spain,; 15Hospital del Mar (IMIM), Barcelona, Spain OBJECTIVES: Treatment adherence is a key issue in chronic infectious diseases such as Chronic Hepatitis C (CHC). The aim of the study was to develop two valid and feasible disease specific questionnaires to assess adherence to treatment of CHC monoinfected patients and HCV/HIV co-infected patients. METHODS: Questionnaires development was carried out in two phases. In the first phase, a literature review was performed in order to identify factors closely related to treatment adherence. By means of a focus group, disease experts selected the items considered most likely to affect adherence from the initial list, then the chosen items underwent a qualitative analysis and were rephrased to make them suitable for inclusion in a questionnaire. The first version of the questionnaires was completed by 18 patients (8 monoinfected and 10 co-infected) in order to evaluate the comprehension of questions, response categories, questionnaires applicability and the patients’ opinion on the proposed questionnaires design. The questionnaires are currently being validated in the second phase, an ongoing prospective study including 1120 patients. RESULTS: The two questionnaires developed for monoinfected and for co-infected patients contained 11 and 13 questions, respectively. Both questionnaires included two questions related to treatment adherence and questions about reasons of non-adherence. To date 843 patients have been included in the prospective study; 71.9% are male, 77.1% are monoinfected patients, 22.9% co-infected and 63.0% of patients have baseline RNA-HCV > 800.000 UI/ml (62.1% monoinfected and 66.1% co-infected). The HCV genotype distribution is 74.8% G-1/4 and 25.7% G-2/3. The HIV positive patients receiving ART Therapy are 91.2%. CONCLUSIONS: The two disease specific questionnaires developed to assess adherence to treatment showed feasibility and good comprehension by patients with HCV monoinfection and with HCV/HIV co-infection, respectively. Further validation is ongoing in a prospective study. PIN60 CLINICAL PROFILE AND ADHERENCE PATTERN OF ANTI-RETROVIRAL THERAPY OF PEOPLE LIVING WITH HIV/AIDS REGISTERED AT DISTRICT ANTI RETROVIRAL THERAPY CENTRE, UDUPI DISTRICT Singh UP, Nayak GC, Singh S, Rath S, Bihan A, Kumar M Manipal College of Pharmaceutical Sciences, Manipal, India OBJECTIVES: To find out the number of people infected with AIDS, the age group that was badly hit by this disease, the number of deaths and most important of all the adherence shown by the patient to the treatment regimen. METHODS: Study subjects: All the people living with HIV/AIDS who were started on Anti Retroviral Therapy at the District ART Centre, Udupi District from 1st July, 2008 to 31st June, A441 2009. Case files were reviewed from medical records section and required data was collected and analyzed using SPSS 11.5 version. RESULTS: Study included 946 cases of which majority were males 534 and 815 were in the age-group of 15–49. 836 were residents of Udupi District. CONCLUSIONS: All of them were started on ART and have been on treatment for atleast 3 months. More than 90% of the patients had adherence of more than 95%. PIN62 AN ECONOMIC EVALUATION OF HIBMENCY VACCINATION FOR THE PREVENTION OF HEMOPHILUS INFLUENZAE TYPE B AND NEISSERIA MENINGITIDIS TYPES C AND Y Taylor MJ1, Saxby RC1, Abu-Elyazeed RR2, Misurski DA3 1 University of York, York, UK; 2GlaxoSmithKline, King of Prussia, PA, USA; 3GlaxoSmithKline, Philadelphia, PA, USA OBJECTIVES: Since the introduction of conjugate vaccines against Haemophilus influenza type b (Hib) and Streptococcus pneumonia, Neisseria meningitis is the leading cause of infant bacterial meningitis. HibMenCY is a new combination vaccine being developed for protection against meningococcal groups C and Y (MenCY) and Hib. This analysis compared the cost-effectiveness of a four dose HibMenCY vaccination strategy (2, 4, 6 and 12 months) to Hib vaccination (2, 4, 6 and 12 months) and to no Hib or MenCY vaccination from a public health perspective. METHODS: A Markov model was used to simulate lifetime events, predicting the likelihood of disease with and without specific vaccinations. Costs and health consequences were built into the model to allow the prediction of lifetime costs and quality-adjusted life-years (QALYs). The model was populated with published data throughout, and one-way sensitivity analyses were undertaken when there was uncertainty about key parameter values. a discount rate of 3% annually was applied. RESULTS: For HibMenCY vaccination relative to Hib alone, 13 additional cases per 100,000 treated patients would be prevented, resulting in an additional 104 QALYS saved. The additional HibMenCY vaccine program cost over Hib alone was $6.61M resulting in a cost/QALY of $63,701. For HibMenCY vaccination relative to no Hib or MenCY vaccination, 61 cases of meningitis were prevented per 100,000 treated patients, resulting in an additional 467 QALYS saved. The additional HibMenCY vaccine program cost was $3.14M resulting in a cost/QALY of $6738. CONCLUSIONS: When accounting for the ability of the HibMenCY vaccine to prevent infection from both pathogens this intervention is estimated to be highly cost-effective. If only the incremental benefit of MenCY is considered, the intervention would result in a better cost/QALY than that reported for several widely implemented vaccine strategies. Thus infant HibMenCY vaccination is expected to be a cost-effective strategy that provides substantial health benefits. PIN63 PATIENT PREFERENCES FOR BENEFITS, RISKS, AND WEIGHT OF EVIDENCE ASSOCIATED WITH HEPATITIS B TREATMENTS Mohamed A1, Johnson FR1, Hauber AB1, Lescrauwaet B2, Masterson A2 1 RTI Health Solutions, Research Triangle Park, NC, USA; 2Bristol-Myers Squibb, 1420 Braine—L’Alleud, Belgium OBJECTIVES: To elicit patient preferences for benefits, risks, and weight of evidence associated with hepatitis B treatments. METHODS: Adults in 5 countries (France, Germany, Spain, Italy, and Turkey) with a self-reported physician diagnosis of hepatitis B completed a web-enabled, choice-format, conjoint-analysis survey. The survey presented patients with a series of 12 trade-off questions, each including a pair of hypothetical hepatitis B medication profiles. Each profile was defined by how long the medicine has been studied (weight of evidence), chance that the medicine will work well for 5 years (long-term efficacy), 5-year fracture risk, 5-year risk of kidney disease, and cost. Trade-off questions were based on predetermined experimental design with known statistical properties. Random-parameters logit was used to estimate a preference weight for each attribute level and the mean relative importance of each attribute. RESULTS: 560 subjects completed the survey. Patients actively traded between efficacy, side-effect risks, and weight of evidence among the hypothetical hepatitis B medication profiles. On an importance-weight scale where 0 is the least important outcome and 10 is the most important outcome, the 5-year risk of kidney disease was the most important attribute to patients over the range of levels included in the survey [95% CI: 7.8–12.2]. The remaining attributes were ranked in decreasing order of importance as: chance that the medicine will work well for 5 years (8.9 [95%CI: 7.1–10.7]), how long the medicine has been studied (8.2 [95%CI: 7.6–8.8]), and 5-year fracture risk (5.8 [95%CI: 3.6–8.0]). CONCLUSIONS: Although long-term efficacy is important to hepatitis B patients, the 5-year risk of kidney disease,in particular, also influence patients’ treatment choices. Patients are willing to make tradeoffs between efficacy, side-effect risks, and weight of evidence among hepatitis B treatments. PIN64 HERPES ZOSTER AND POST-HERPETIC NEURALGIA: CONSIDERATIONS FOR A PROSPECTIVE, CROSS-SECTIONAL, PATIENT REPORTED OUTCOMES STUDY Carroll S1, Gater A2, Abetz L2, Demuth D3, Smith F3, Mannan A1 1 Sanofi Pasteur MSD, Maidenhead, Berkshire, UK; 2Mapi Values Ltd, Bollington, Cheshire, UK; 3 Adelphi Real World, Bollington, Cheshire, UK OBJECTIVES: The pain and discomfort associated with Herpes Zoster (HZ) and Post-Herpetic Neuralgia (PHN) can have a substantial negative impact on patients’ A442 Health-Related Quality of Life (HRQoL). To inform the design of a large-scale prospective study of the patient-reported impact of HZ and PHN in the UK, a review of published prospective studies was conducted. METHODS: A targeted literature review was conducted in MEDLINE® to identify prospective studies investigating the impact of HZ and PHN from the patient’s perspective. Searches were limited to studies published in English between April 2000 and April 2010. The adequacy of these studies in quantifying the burden of HZ and PHN was critically assessed. RESULTS: Only a small number of published prospective studies (n = 6) designed to investigate the impact of HZ or PHN from patients’ perspectives were identified; one of which was conducted in the UK. The HRQoL burden associated with HZ and PHN was highlighted in the studies reviewed albeit a number of key points were noted. First, the burden of HZ and PHN has not been compared to HRQoL estimates for the general population or to those with comparable acute or chronic conditions, thereby making interpretation difficult. Furthermore, studies conducted to date have not accounted for the variations in pain and HRQOL burden at different stages of HZ presentation. Finally, from the studies reviewed, the pain experienced by patients with HZ and PHN is managed with a wide variety of products. However, until now no formal assessment of patient satisfaction with these products has been conducted. CONCLUSIONS: Although it is known that HZ and PHN can have a negative impact on HRQOL, limitations in previous research can be noted. In order to address these concerns, a large-scale UK-based prospective study is currently being conducted. PIN65 IMPACT OF HERPES ZOSTER AND POSTHERPETIC NEURALGIA ON PATIENTS’ QUALITY OF LIFE Chassany O1, Gaillat J2, Hanslik T3, Bourhis Y4, Launay O5, Mann C6, Rabaud C7, Rogeaux O8, Strady C9, Bouhassira D10 1 Assistance Publique-Hopitaux de Paris, Paris, France; 2CHR Annecy, Pringy, France; 3Hôpital Ambroise Paré, Boulogne-Billancourt, France; 4REGISTRAT-MAPI, Lyon, France; 5Groupe hospitalier Cochin—Saint-Vincent de Paul, Paris, France; 6Hôpital Saint-Eloi, CHU de Montpellier, Montpellier, France; 7CHU de Nancy—Hôpitaux de Brabois, Vandoeuvre les Nancy, France; 8CH de Chambéry, Chambéry, France; 9CHU de Reims, Reims, France; 10 Hôpital Ambroise Paré, Boulogne-Billancourt, France OBJECTIVES: Incidence and severity of herpes zoster (HZ) increase with age. Postherpetic neuralgia (PHN) is its most common complication. The study objective was to assess the impact of HZ and PHN on daily life and quality of life (QoL) as perceived by the patient. METHODS: A 12-month, longitudinal, observational study conducted on patients aged ≥50 years with acute HZ (eruptive phase), included by a representative sample of 644 French general practitioners (GP) between June 2007 and June 2008. Demographic and disease characteristics and treatment procedures were collected by the GP at the inclusion and at 3, 6 and 12 months. PHN and QoL assessments were performed by telephone interview of patients, at regular time points over 12 months, using comprehensive and validated questionnaires. RESULTS: a total of 1358 patients met the selection criteria and 1032 completed the follow-up. Mean age was 67.7 ± 10.7 years (62.2% female). Mean time between rash onset and HZ diagnosis was 2.6 ± 3.0 days. Most patients (94.1%) were treated with antiviral drugs. PHN prevalence at inclusion, 3, 6, and 12 months was 79.6%, 11.6%, 8.5%, and 6.6% respectively. Mean Zoster Brief Pain Inventory (ZBPI) interference score (range = 0–10 worse) at inclusion, 3, 6 and 12 months was 3.1 ± 2.4, 2.9 ± 2.3, 2.5 ± 2.1, 2.7 ± 2.3 respectively, with main impact on general activity, sleep and mood. Throughout the 12-month study period, patients with persistent PHN presented lower scores on the physical (e.g. at 12 month: 39.7 ± 13.7) and mental (43.6 ± 11.3) component summaries of the 12-item short-form health survey (SF-12) than the patients without PHN, respectively 46.0 ± 10.6 and 47.8 ± 10.5. Moreover, univariate analysis found that among predictive factors at inclusion for PHN were a ZBPI score ≥5 (P < 0.001) and a low SF-12 physical component summary (P < 0.001). CONCLUSIONS: This large cohort study confirms the HZ and PHN impact on daily life and quality of life, in spite of prompt diagnosis and antiviral treatment. PIN66 EVALUATION OF HEALTH-RELATED QUALITY OF LIFE (HRQOL) IN PATIENTS RECEIVING ANTIRETROVIRAL THERAPY IN MAITAMA DISTRICT HOSPITAL, ABUJA NIGERIA Agu KA1, Okojie O2, Omonaiye O1, Oqua D1, King RC3, Onuoha C1, Muhammadu I1, Iyaji PG1 1 Howard University PACE Center, Abuja, FCT, Nigeria; 2University of Benin, Benin, Edo, Nigeria; 3Howard University PACE Center, NW, WA, USA OBJECTIVES: HIV-infected patients demonstrate a health-related quality of life (HRQoL) lower than that of the general population. The goals of antiretroviral therapy (ART) include the improvement of HRQOL in addition to the reduction of symptoms, suppression of the virus, and enhancement of survival. The study evaluated HRQoL of HIV-infected patients at months 0 and 6 of receiving ART; and compared the HRQoL index of male and female participants. METHODS: The HRQoL of a cohort of 150 HIV-infected patients was evaluated at months 0 and 6 of receiving ART using SF-36 health survey form. Wilcoxon’s signed ranks test was used to compare the HRQoL index between two groups of data. All reported P values were 2-sided, and P < 0.05 was used to determine statistical significance. RESULTS: The mean age of the 150 patients at initiation of antiretroviral treatment was 34.30 (95% CI, 32.96–35.64) years; and 59.3% were females. At 6 months follow-up, 118 (78.7%) ART patients who met the 6th month medication refill appointment were evaluated for HRQoL. Before commencing ART, the mean HRQoL of participants 13th Euro Abstracts was 71.22 (95%CI, 68.00–74.44) which increased significantly to 90.83 (95%CI, 88.49–93.17) at 6 months follow up on ART (P < 0.05). The change across all the HRQoL domains in the SF-36 instrument was also statistically significant (P < 0.05) except for the domains of role limitation due to emotional problems, social functioning and pain (p > 0.05). Difference in the HRQoL index of male and female gender at months 0 and 6 was not statistically significant (p > 0.05). CONCLUSIONS: There was a significant improvement in the mean HRQoL of the participants after 6 months of ART; the difference between HRQoL index of male and female participants was not significant. Routine evaluation of HRQoL in clinical practice is highly recommended as long-term outcome may be different. INFECTION – Health Care Use & Policy Studies PIN67 MODELING G-DRG REVENUE WHEN HOSPITALIZATION DURATION FROM CLINICAL TRIALS IS AVAILABLE: THE CASE OF DORIPENEM VERSUS IMIPENEM IN TREATING VENTILATOR-ASSOCIATED PNEUMONIA (VAP) IN GERMANY De Cock E1, Gast C2, Berndt K3, Kubitz N3 1 United BioSource Corporation, Barcelona, Spain; 2Axio Research Acquisition Co. LLC, Seattle, WA, USA, 3Janssen-Cilag GmbH, Neuss, Germany OBJECTIVES: Since the introduction of G-DRGs, hospitals have become increasingly concerned about impact of changes in clinical practice on hospital costs and revenue. We developed an algorithm to estimate expected G-DRG payment by drug, when length-of-stay (LOS) data from clinical trials are available. We compared results with a scenario where only median values by trial arm are available. METHODS: G-DRG attribution is driven by MV duration and intensive care severity score. From a Phase III trial comparing doripenem and imipenem (DORI-10), we determined nine categories of MV duration (ranges based on MV hours resulting in higher-paying G-DRGs). Using the Muenster Webgrouper, we obtained G-DRG weights when simultaneously varying MV duration and intensive care severity (15 levels) for four clinical profiles. For the nine categories and by drug, we obtained percentage patients per group and median LOS by type of ward using. Per category, the appropriate G-DRG weight was selected based on: chosen profile, total MV duration, intensive care severity, and total LOS. Using distribution of patients by category derived from DORI-10, a weighted G-DRG was calculated for each arm, and multiplied by the 2010 base value to yield expected G-DRG payment. RESULTS: G-DRG revenue ranged from c2,578 to c8,575 (across 4 scenarios) to c136,164 for MV ≥ 1800 hours. Using trial-based median LOS values, expected G-DRG revenue is c30,397 (all scenarios), with no difference in G-DRG revenue in both trial arms. When using the categories approach, expected revenue reduction when using doripenem instead of imipenem ranged from c2,084 to c2,451. CONCLUSIONS: A G-DRG revenue analysis helps to better understand net impact of introducing a new drug on a hospital’s budget. When LOS data (including MV) vary by drug and data from clinical studies are available, a classification approach enables more refined G-DRG revenue estimates as opposed to using median values by drug. PIN68 EFFECT OF THE ANTIBIOTICS USE MONITORING PROGRAM FOR THE ACUTE UPPER-RESPIRATORY INFECTIONS DISEASE IN KOREAN CLINICS Choi SE1, Park KH1, Lim EA1, Kim SK2 1 Seoul National University, Seoul, South Korea; 2Health Insurance Review & Assessment Agency, Seoul, South Korea OBJECTIVES: Overuse of antibiotics has been focused for the public health and pharmaceutical expenditure in Korea. National Health Insurance introduced the antibiotics use monitoring program in 2002. This program includes a monitoring the antibiotics prescription rates of medical institutes, a feedback to prescribers, and a report to the public via website. This study aims to evaluate the impact of this program on the antibiotics prescription rates for the acute upper-respiratory infections (URI) in Korea. METHODS: We assessed the antibiotics prescription rates for the acute URI (J00-J06) of individual medical institutes from 2002 to 2008. Data was extracted from NHI outpatient’s claims database and the first quarter data of each year were selected to avoid seasonal variation. To evaluate the effect of program, we assumed the early change of prescription rate during 2002–2003 as a baseline, and compared it with every year variations of the prescription rate for 2004–2008. Generalized Estimation Equation (GEE) model was constructed to investigate the changes in antibiotics prescription. RESULTS: Data included about 49 million claims from 13,211 clinics (94.5% of all acute URI claims and 70.5% of all clinics in 2008). Prescription rate declined from 72.8% to 55.1% during 2002–2008. The analysis showed that the antibiotics use monitoring program reduced 6.11% of antibiotics prescription rate per clinic annually (OR: 0.62, 95% CI: 0.61, 0.64, P < 0.0001). The reduced antibiotics prescription estimate was average 4.5 million a year. CONCLUSIONS: After monitoring system introduced, the antibiotics prescription rates for acute URI patients have continuously decreased in the clinics. 13th Euro Abstracts PIN69 OUTPATIENT ANTIBIOTIC USE IN PRIMARY HEALTH CARE IN NIS REGION Kodela B1, Velickovic-Radovanovic R2, Petrovic J2, Mitic R2, Aleksic G1 1 Pharmaceutical Institution of Nis, Nis, Serbia; 2Clinical Centre, Nis, Serbia OBJECTIVES: Monitoring of antibiotic prescribing promotes rational use of drugs, reduces costs and slows down the development of resistance. The purpose of this study was to evaluate the prescribing of antibiotics in PHC of Niš region, during 2005–2009 and to identify the practice of drug overuse, under-use or inappropriate use. METHODS: The data on outpatient use of antibacterials for systemic use was obtained by retrospective study and expressed as the number of defined daily doses per 1000 inhabitants per day (DDD), according to WHO anatomic therapeutic chemical classification and DDD measurement methodology. The City Pharmacy Department provided automatic reports on antibiotics prescribed by physicians for certain diagnoses in the Nis region over the 2005–2009 period. RESULTS: In the same period, outpatient antibiotic use in Nis region increased by 9.02% (22.83/ 25.96 DID). The most commonly prescribed drugs were semi-synthetics penicillins (9.67/10.00 DDD) and macrolides (3.05 /4.90 DDD). The greatest increase in antibiotics prescriptions was noted in azithromycin (0.26/0.70 DID, by 164%). Out of the total number of antibiotics, 70.5% prescriptions were made for treatment of URIs. Pharingitis ac. (J02) was the most common indication for prescription of antimicrobial medicines (45.26%). Amoxicillin (33%) and cephalexin (35%) were most commonly prescribed antibiotics for this indication. The total number of antibiotics prescriptions was increased by 14% while the cost was increased by 28% compared to 2005. CONCLUSIONS: Outpatient antibiotics use in Niš is high compared to the majority of European countries, which indicates the need for additional educational programs and monitoring of microbiological sensitivity in future. The received results will be the basis for further evaluation of the rationality of use of antibiotics in primary health care. PIN70 COST EQUITY ANALYSIS OF MALARIA MANAGEMENT: A PILOT STUDY IN THE AFRICAN SETTING Espinoza MA1, Griffin S1, Cookson R2 1 Centre for Health Economics, York, UK; 2Department of Social Policy and Social Work, York, UK OBJECTIVES: To study the equity efficiency trade-off between decisions made based on maximization of health benefits averaged for the whole population and decisions made considering the distribution of benefits across socio economic groups of the population. METHODS: Cost-equity analysis nested in a CEA where ten different strategies for the management of patients with suspected non-complicated malaria were compared, in the African setting. Access to health clinics, access to treatment and adherence to the treatment were considered to vary between socio economic groups. a further hypothetical therapeutic alternative associated with 100% adherence for all groups was incorporated. Probabilistic sensitivity analysis was performed to reflect the uncertainty around parameter values. Inequality in the distribution of benefits was estimated for each strategy using Concentration Indices (CI). The opportunity cost was estimated comparing the most cost-effective alternative against the most cost-equitable strategy in terms of health benefits forgone to achieve a certain number of equality units. RESULTS: The most equitable strategy was also the most efficient for thresholds between US$137 and US$375 per DALY averted. Differences in equity of DALY burden between strategies were small (range CI units: 0.061–0.096) meaning the opportunity cost of choosing the most equitable strategy for threshold values outside of that range were relatively high, e.g. at US$400/DALY averted the opportunity cost was 0.11 DALYs per percentage point reduction in the CI. CONCLUSIONS: Equity concerns can be incorporated in the economic evaluation offering an estimation of the trade-off between efficiency and equity. Collection of information about access, adherence, effectiveness and costs according to socioeconomic group is required, but also information on prevalence for appropriate calculation of CI. PIN71 CLINICAL AND ECONOMIC BURDEN OF PEDIATRIC INFLUENZA IN NINE EUROPEAN COUNTRIES Rycroft C1, Leeuwenkamp O2, Heikkinen T3, Principi N4, Herz J5, Moren S6, Beard S7 1 RTI Health Solutions, Manchester, UK; 2Formerly MedImmune Ltd, Cambridge, UK; 3Turku University Hospital, Turku, Finland; 4University of Milan, Milan, Italy; 5AstraZeneca, Zaventem, Belgium; 6MedImmune, LLC, Gaithersburg, MD, USA; 7RTI Health Solutions, Sheffield, UK OBJECTIVES: Seasonal influenza incurs a substantial clinical and economic burden in infants, children, and adolescents. The objectives of this literature review were to assess the reported incidence and complications of virologically-confirmed paediatric influenza, and to evaluate the extent of health care utilization, absenteeism, direct health care costs, and societal costs in nine European countries: the United Kingdom, Germany, Italy, Spain, France, Sweden, the The Netherlands, Finland, and Austria. METHODS: A structured literature search (January 1970-March 2009) of PubMed, EMBASE, and the Cochrane Library was conducted. Data on influenza burden in children (aged ≤15 years) were extracted from 43 publications. Of these, 13 articles presented data on virologically-confirmed influenza. Pooled data from these 13 studies are presented. RESULTS: Data varied widely between countries. Median incidence of confirmed influenza was 15.4% but ranged from 1.1 to 33.0% (n = 6 studies) in children aged ≤15 years. This wide range in incidence resulted from differences between studies in influenza confirmation methods, age groups, and influenza seasons. Complications due to influenza were frequently reported (e.g., acute otitis media A443 [range 0–40% [n = 7 studies]], pharyngitis [range 31–48% [n = 3 studies]]). In one study, a 3-fold increase in direct medical costs (c189–c206 vs. c59 [2004 costs]) was observed. In another study of children aged ≤1 year, hospitalization was reported to be the main driver of disease-related direct medical costs with hospitalization rates of about 20%. Of children with confirmed influenza, 28–55% used antibiotics (n = 6 studies) and 76–86% required antipyretics or symptomatic treatment (n = 5 studies). Confirmed influenza in children was associated with 3–12 days’ absence from school/ daycare (n = 5 studies), and 1.3–6.3 days’ parental absence from work (n = 6 studies). CONCLUSIONS: The clinical and economic burden of virologically-confirmed paediatric influenza in Europe appears to be significant as revealed by incidence and associated morbidity and costs. This study was funded by MedImmune, LLC. PIN72 BRAZILIAN STD/AIDS PROGRAMME: EXPENDITURE ON ANTIRETROVIRAL DRUGS AND EPIDEMIC INDICATORS DURING THE PERIOD 2004–2008 Aurea AP, Garcia LP, de Magalhaes LCG, dos Santos CF, de Almeida RF Institute for Applied Economic Research, Brasilia, DF, Brazil OBJECTIVES: The Brazilian STD/AIDS Programme stands out for ensuring universal and free access to antiretroviral drugs. We estimated the Federal expenditure on antiretroviral drugs of the Brazilian STD/AIDS Programme, during the period 2004– 2008, and compared it to epidemic indicators of the Brazilian population. METHODS: The expenditure on antiretroviral drugs comes from a data warehouse of the Ministry of Planning, Budget and Management that stores the information concerning any purchase made by the Brazilian Federal Government. The epidemiological indicators of AIDS are derived from the Brazilian STD/AIDS Programme, which combines data from different national information systems. RESULTS: The expenditure on antiretroviral drugs rose from US$144.416.113,57 in 2004, to US$333.050.295,64, in 2008. In Brazil, from January 2004 to June 2008, there were 148.852 cases of AIDS reported, which corresponds to an incidence rate of approximately 20 cases per 100 thousand inhabitants. In the period from 2004 to 2007 there were 44.381 deaths due to AIDS in Brazil, representing a crude mortality ratio of about 6 deaths per 100 thousand inhabitants. CONCLUSIONS: The use of antiretroviral drugs increases the survival of patients with HIV/AIDS and reduces the mortality rate. Despite the stability of AIDS incidence and mortality coefficients over the analyzed period, the expenditure increased, which can be due to an increase in coverage and range of drugs offered. As the resources are limited, it is important to understand the magnitude of the consumption as well as the mechanisms of purchase in order to improve the supply of antiretroviral drugs. VA3 FACTORS INFLUENCING THE DECISION TO BE VACCINATED AGAINST H1N1 IN GREECE Athanasakis K, Zavras D, Kyriopoulos J National School of Public Health, Athens, Greece OBJECTIVES: In view of the 2009 emergence of Influenza a (H1N1) and the declaration of a phase 6 pandemic by WHO, Greece established a voluntary vaccination program for the entire country population. The aim of the present study was to investigate factors affecting individuals’ decision to participate. METHODS: We conducted repeated weekly telephone surveys until a random sample of 12,639 households stratified according to place of residence (geographical area and level of urbanization) was formulated. The interviews were based on a strictly structured questionnaire designed for the purpose of the study. a logistic regression analysis was conducted in order to identify the factors that influence individuals’ decision to be vaccinated against H1N1. The Hosmer–Lemeshow criterion was used to check the model’s goodness of fit. RESULTS: Higher levels of personal education (Odds Ratio: 0.61) and income (OR: 0.42) were associated with a negative decision to participate to the vaccination program. However, when an interaction term between the two variables was used, individuals of both higher education and income were more likely to have participated in the program (OR: 1.26). On the contrary, increased levels of concern/fear for the H1N1 and increased confidence to the personal/family physician, both measured in a 5-point Likert scale, showed a positive relationship. The Hosmer– Lemeshow goodness of fit test statistic was 0.98 suggesting satisfactory fit of the model. CONCLUSIONS: In Greece, a country where less than 5% of the total population participated to the voluntary vaccination program against H1N1, socioeconomic status, as expressed through income and education, the role of the family physician and the personal attitude towards H1N1 were important factors for a positive decision. In the case of similar future efforts, these variables should be taken into account, in order to facilitate focused interventions. PIN73 FACTORS AFFECTING THE DECISION TO RECEIVE VACCINATION FOR INFLUENZA VIRUS H1N1 Zavras D1, Kyriopoulos J2 1 National School of Public Health, Athens, Attiki, Greece; 2National School of Public Health, Athens, Greece OBJECTIVES: The objective of this study was to determine factors affecting the population’s decision to get vaccinated against the influenza H1N1 virus. METHODS: Data for this study was obtained from a Panhellenic survey, organized by the National School of Public Health of Greece. The sample (n = 12,639) was drawn from strata of various degrees of urbanization and geographical region. The primary outcome of the survey was the intent to get vaccinated for H1N1; logistic regression was used to A444 identify significant potential determinants. RESULTS: The decision of vaccination for influenza virus H1N1 was associated with factors related to education, income, interaction between education and income, gender, trust to Public Health Organizations and perceptions about the health effects of influenza virus H1N1. Individuals of higher education and income, do not intend to get vaccinated. This also holds for individuals who express low degree of trust to or believe that the H1N1 virus is not a serious threat to human health. CONCLUSIONS: We identified several socioeconomic and demographic factors affecting individual intent to get vaccinated for H1N1. Personal beliefs about the health effects of H1N1 virus and trust to Public Health Organizations were also significant predictors of vaccination intent. PIN74 PARENTS’ DILEMMA WHETHER OR NOT TO VACCINATE THEIR CHILDREN AGAINST INFLUENZA A (H1N1) Tsiantou V1, Zavras D2, Kyriopoulos J1 1 National School of Public Health, Athens, Greece; 2National School of Public Health, Athens, Attiki, Greece OBJECTIVES: In April 2009, WHO announced the emergence of Influenza a (H1N1) and a phase 6 pandemic was declared a few months later. Vaccination was identified as an effective measure in order to eliminate the pandemic and ensure public health. The aim of the present study was to investigate the factors affecting parents’ decision to vaccinate their children. METHODS: A telephone survey was conducted in a random sample of 12,639 households stratified by geographical area using a strictly structured questionnaire designed for the purpose of the study. The survey lasted from October 2009 to January 2010. a logistic regression analysis was conducted in order to identify the factors that influence parents’ decision to vaccinate their children. The Hosmer–Lemeshow criterion was used to check the model’s goodness of fit. RESULTS: A total of 3585 households were eligible and were included in the analysis. Education, income, concern about H1N1 and trust to the CDC were statistically significant. According to the analysis the higher the levels of education (OR 0.64) and income (OR 0.80) of the parents the lower the probability to vaccinate their children. However, the interaction term between education and income showed a positive relationship between these and the dependent variable. The greater the concern about H1N1 and the trust to the CDC the higher the probability for the parents to vaccinate their children. The Hosmer–Lemeshow goodness of fit test statistic was 0.52 suggesting satisfactory fit of the model. CONCLUSIONS: Based on our results concerns about the H1N1 and trust on the CDC were influencing positively the decision for vaccination. The negative relationship between education and income and parents’ decision for vaccination was interesting and need further investigation. Results can explain the low vaccination rates against H1N1 in Greece and can be used by policy makers to improve future vaccination campaigns. PIN75 ESTIMATION OF PATIENTS WITH ANTIRETROVIRAL THERAPY POTENTIALLY USED FOR HIV PREVENTION (POST-EXPOSURE PROPHYLAXIS, MOTHER-TO-CHILD TRANSMISSION PROPHYLAXIS) IN GERMANY Tomeczkowski J1, Guthoff-Hagen S2, Biteeva I1, Kruppert S3, Stoll M4 1 Janssen-Cilag GmbH, Neuss, Germany; 2Sgh-Consulting, Hamburg, Germany; 3IMS Health GmbH & Co. OHG, Frankfurt/Main, Germany; 4Medical University of Hanover (MHH), Hannover, Germany OBJECTIVES: Determine the number of patients who received antiretroviral therapy (ART) potentially prescribed for prevention of HIV infection in Germany in 2008. METHODS: A representative panel of 1,193,464 patients from different nationwide operating governmental sick funds were analyzed for 2008. Patients with a confirmed diagnosis of HIV (ICD-10: B20–24: 64,8%; R75: 0,9%; U85: 0,6%; U60–61: 14,9%; Z21: 18,8%) were included. In addition, an IMS pharmacy panel (LRx) with 26,815 patients who had received ART identified by ATC Code J05C (Tenofovir was excluded because of use in Hepatitis) in October 2008 to September 2009 were analyzed (54% of all patients with ART-prescriptions). RESULTS: Sick funds: 927 patients with HIV diagnosis were identified (0.08% of the panel), 548 received an ART (treatment rate 59.1%). Pharmacies: 46% of patients received an ART prescription in 4 quarters, 16.9% in 3 quarters, 12.1% in 2 quarters and 24.9% in 1 quarter. 17.5% received prescriptions only on 1 day while of these patients 12% did not have any other ART prescription at least 4 months before or after the analysis period. Patients who received prescriptions on 1 day only, were younger, more often female and received more often Lopinavir/Ritonavir compared to patients who received prescriptions on more days. When extrapolating the numbers to the German governmentally insured population and comparing the numbers from sick funds of patients with a confirmed diagnosis receiving ART with patients who received an ART from pharmacies, about 30% of patients with ART prescription could not be matched to a confirmed diagnosis. CONCLUSIONS: Thirty percent of patients received ART prescriptions without confirmed HIV diagnosis. Twelve percent of patients received ART prescriptions on one day only. Since these patients were younger, more often female and received agents more often used in prophylactic regimens, we assume that prescription of ART for prevention or post-exposure prophylaxis is likely in this group. PIN76 THE USE OF ECONOMIC ANALYSIS IN VACCINE EXPERT REVIEWS Jacobs P Institute of Health Economics, Edmonton, AB, Canada OBJECTIVES: Since the beginning of the new millennium, prices of vaccines have been increasing significantly, reflecting decades of research and development. As they 13th Euro Abstracts have done with pharmaceuticals, governments have begun to incorporate economic considerations into scientific reviews. Vaccines fall into the public health category, whose ethic differs historically from that in the pharmaceutical market. The purpose of this paper is to review how countries have incorporated economics into the scientific vaccine review process. METHODS: We contacted experts in countries which had scientific review committees according to the VENICE website, and the USA, Australia, New Zealand, and Hong Kong. We asked a series of questions regarding the structure of the review process, the role of economics, and the economic criteria used. RESULTS: We obtained information from ten countries which used economics in their vaccine scientific reviews. In terms of structure, reviews fell into two groups—those which incorporated economics into the scientific review (GB, FI, FR, HK, NL, NZ, US) and those which separated economic and clinical considerations (SW, AU). The criteria used in economic studies followed the pharmaceutical guidelines in all instances. CONCLUSIONS: Pharmaco-economics has grown up in the world of pharmaceuticals, not in the public health arena. The ethics in these two arenas differ. Certain elements—epidemiological considerations, herd immunity, long effect times—are relevant to vaccines. Using pharmacoeconomic rules for vaccines may not create a level playing field. INFECTION – Conceptual Papers & Research on Methods PIN77 DO WE ADEQUATELY MODEL THE BENEFIT OF ROTAVIRUS VACCINATION OVER TIME? Standaert B1, Gomez J2, Acosta-Rodriguez C3, Debrus S1 1 GlaxoSmithKline Biologicals, Wavre, Belgium; 2GlaxoSmithKline Biologicals, Buenos Aires, Argentina; 3GlaxoSmithKline, Philadelphia, PA, USA OBJECTIVES: Models estimating the impact of rotavirus vaccines over time use vaccine efficacy (VE) results from clinical trials measured at different time points. The formula to calculate VE measures diarrhea events observed in the vaccinated arm divided by events in the non-vaccinated arm during certain periods. Two problems occur: 1) the control arm builds up its own natural immunity over time: VE measured by the formula is therefore the net vaccine effect (NVE) that decreases as soon as the rate of infection increases; 2) natural immunity should be considered in the vaccinated arm as well as a booster phenomenon whereas this is generally omitted. How much will the cost-effectiveness result be impacted if natural immunity in the vaccinated arm is considered by improving the NVE-value over time? METHODS: A markov cohort model is used to measure the cost-effectiveness of vaccinating children at 2 and 3 months against rotavirus infection in a country such as Panama as an example. The HE-model considers the health authority perspective, a life time horizon, and an annual discount rate of 3% on cost and effect. In sensitivity analysis the NVE decrease is varied over time post-vaccination from an annual linear 15% decrease to equal value post-2 doses. The outcome measure is the change in cost-effectiveness result in function of the variation in NVE-decrease. RESULTS: With the 15% decrease in NVE the QALY gain is 0.0124/person, the extra cost $2.46, and the ICER = $199/QALY gained. Improving the NVE to no decrease post-2 doses results in a QALY gain of 0.0132 (+6%), an extra-cost of $0.98 (−99%), and an ICER reduction to $74/QALY gained. CONCLUSIONS: Assuming sustained VE over time because of natural immunity in the vaccinated arm improves the economic results -especially the costs data. The gain will mainly occur post-disease peak after 2 years. PIN78 ECONOMIC IMPACT OF INFLUENZA EPIDEMICS: MODELLING INTERACTIONS OF INFECTIONS, TREATMENT PATHWAYS AND REIMBURSEMENT Einzinger P1, Zauner G2, Gyimesi M1, Schiller-Frühwirth I3, Pfeffer N3 1 Vienna University of Technology, Vienna, Austria; 2Dwh Simulation Services, Vienna, Austria; 3 Main Association of Austrian Social Security Institutions, Vienna, Austria OBJECTIVES: Infectious diseases spread through social contacts and affect people of all age groups. Traditional epidemic models consider these effects with the use of differential equations, contact networks or explicit modelling of households and workplaces. However past modelling studies did not implement structures of the health service system, like service providers (e.g. physicians) and their reimbursement. Therefore we investigate the integration of an epidemic contact model for influenza into a framework for modelling treatment pathways and reimbursement of service providers. METHODS: The model framework is of object-oriented and agent based type. It incorporates patients and providers as spatially distributed agents. When patients develop diseases they search for service providers and treatment through a central health market, which in general returns one of the nearest providers of requested type. Each disease is connected to the possible treatment pathways and health services that a patient with the specific disease will go through. In the study we incorporate influenza epidemics occurring at specified time intervals. These epidemics spread through contacts of the agents modelled by a small-world network. Patients get immune after recovering and therefore one epidemic cannot infect them twice. The model calculates reimbursement from consumed health provider services and drugs. RESULTS: Modeled epidemics show the well-known behaviour of SIR-type models. Reduced capacity of service providers provokes that many patients do not consult them due to queuing effects. The model maps the development of costs over time plausibly. Different prescription probabilities of neuraminidase inhibitors influence the course of the epidemic only marginally. CONCLUSIONS: The integration of epidemic models with models of health service processes and reimbursement can lead toward 13th Euro Abstracts a broader understanding of the influence of epidemics on health care systems. It is possible to investigate feedback of health service structure changes and reimbursement decisions on prevalence and effects of infectious diseases. PIN79 IS IT POSSIBLE TO OBTAIN WILLINGNESS-TO-PAY ESTIMATES IN EUROPE? A VALIDITY TEST OF STATED PREFERENCES FOR HEPATITIS-B TREATMENTS Johnson FR1, Mohamed A1, Hauber AB1, Lescrauwaet B2 1 RTI Health Solutions, Research Triangle Park, NC, USA; 2Bristol-Myers Squibb, Braine—L’Alleud, Belgium BACKGROUND: Because health care is highly insured in Europe, many discretechoice experiment (DCE) researchers have concerns about the feasibility of obtaining valid willingness-to-pay (WTP) estimates. OBJECTIVES: To test the validity of WTP estimates obtained from patient DCE data in 5 countries. METHODS: Adults with a self-reported physician diagnosis of hepatitis B in 4 European countries (France, Germany, Spain, and Italy) and Turkey completed a web-enabled, DCE. The survey presented patients with a series of 12 trade-off questions, each including a pair of hypothetical hepatitis-B treatments described by efficacy, two side-effect risks, weight of evidence, and cost. All the subjects saw cost levels of c0, c10, and c25. Half the subjects evaluated an additional cost level of c75 and half the subjects evaluated an additional cost level of c150 per month. RESULTS: 664 subjects completed the survey. About 15% of subjects refused to accept any tradeoffs between costs and outcome attributes, while the remainder perceived no significant difference between costs of c0 and c10. The difference in importance weights between the highest cost levels in each treatment arm was significantly different (P < 0.000). The importance weight of one additional euro in cost of c25 or greater was negative, highly significant, and equal in both arms (p = <0.000). Compared to a treatment with 70% probability of achieving no detectable virus after 5 years and no side-effect risk, the increase in value to patients of a hypothetical treatment that has 95% effectiveness, 1% 5-year fracture risk, and 1% 5-year risk of kidney disease is an additional c34 (c17–c51) per month. CONCLUSIONS: We obtained DCE responses in a split-sample test of cost sensitivity that were consistent with theoretical requirements. Results suggest that it is possible to obtain valid WTP estimates in a properly motivated European DCE. MENTAL HEALTH – Clinical Outcomes Studies PMH1 THE EFFICACY OF DONEPEZIL AND MEMANTINE FOR TREATING BEHAVIOURAL AND PSYCHOLOGICAL SYMPTOMS OF DEMENTIA (BPSD) IN PATIENTS WITH ALZHEIMER’S DISEASE: SYSTEMATIC REVIEW AND META-ANALYSIS Orme M1, Mitchell S2, Lockhart I3, Collins S2 1 ICERA consulting Ltd, Swindon, Wiltshire, UK; 2Abacus International, Bicester, Oxfordshire, UK; 3Pfizer Limited, Tadworth, Surrey, UK OBJECTIVES: Behavioral and psychological symptoms of dementia (BPSD) in Alzheimer’s disease (AD) greatly increase caregiver burden and often trigger nursing home placement. a systematic review of double-blind randomized controlled trials (RCTs) was conducted to compare the ability of donepezil and memantine to manage BPSD in AD. METHODS: MEDLINE, EMBASE, Cochrane Library, and hand searches identified 4739 citations, of which 16 studies had Neuropsychiatric Inventory (NPI) data suitable for meta-analysis (6 memantine and ten donepezil trials). All trials were placebo-controlled, and no head-to-head comparisons of the two drugs were identified. a random-effects meta-analysis was conducted using AD severity subgroups to investigate heterogeneity between trials. Thereafter meta-regression was conducted using study level covariates as potential predictors of between-treatment weighted mean difference (WMD) in NPI score. RESULTS: Unadjusted random effects metaanalysis of all 16 RCTs found significant between-study heterogeneity (all studies I2 = 64.4%; donepezil vs. placebo WMD −1.84, 95%CI −3.57, −0.10, I2 = 61%; memantine monotherapy vs. placebo WMD −1.19, 95% CI −3.70, 1.32, I2 = 68.7%; memantine + AChEI combination therapy vs. placebo WMD −1.68 95% CI −5.70, 2.33, I2 = 80.4%). Meta-regression that stratified studies into four AD severity groups measured by Mini-Mental State Exam (MMSE) (severe 0–9, moderately-severe 10–14, moderate 15–20 and mild 21–26) and controlling for age at baseline, found these covariates accounted for most (59.85%) of the between-study variance. Using this meta-regression model, the pooled NPI score for donepezil showed significant improvement compared to placebo (MD −1.50; 95% CI −2.63, −0.36) whereas this was not the case for memantine vs. placebo (MD −1.25; 95% CI −2.63, 0.13). CONCLUSIONS: Donepezil was associated with significant improvement in the management of BPSD in AD patients compared to placebo, whereas memantine failed to show a significant improvement versus placebo in the management of these symptoms. A445 PMH2 COMPARING ALL-CAUSE MEDICATION DISCONTINUATION WITH DEPOT AND ORAL ANTIPSYCHOTICS IN MATCHED COHORTS OF PATIENTS WITH SCHIZOPHRENIA: A 12-MONTH OBSERVATIONAL STUDY Brnabic AJ1, Kelin K1, Ascher-Svanum H2, Kadziola Z3, Montgomery W4 1 Eli Lilly & Company, Sydney, NSW, Australia; 2Eli Lilly & Company, Indianapolis, IN, USA; 3 Eli Lilly Regional Operations Ges.m.b.H, Vienna, Austria; 4Eli Lilly & Company, West Ryde, NSW, Australia OBJECTIVES: To assess the all-cause medication discontinuation rate in matched cohorts of patients with schizophrenia at risk of nonadherence who were initiated on depot or oral antipsychotics and followed over 12 months. METHODS: At study entry, outpatients with schizophrenia from Australia, Mexico, Romania, and Taiwan were switched, due to clinician-perceived medication non-adherence risk, from their current oral antipsychotic to either a depot or different oral antipsychotic in this 12-month, prospective observational study. Patients were compared on all-cause medication discontinuation rates, defined as a switch from the initiated medication or its augmentation with another antipsychotic. Patients initiated on depot were matched with those initiated on any oral antipsychotic, using full optimal and nearest neighbour (greedy) matching algorithms. The Rank-based Mahalanobis metric was chosen as the distance based on propensity score plus other relevant covariates, with country and antipsychotic class used for exact matching. RESULTS: Based on the full optimal 1:1 matching, only 40 of the 43 original depot initiators were matched to a corresponding oral initiator. After matching, there were no statistically significant differences between the depot and oral cohorts on any of the study entry covariates examined. During the 12-month study, 20% of depot patients discontinued their initial medication, compared with 40% of oral patients (survival curve comparison, p = 0.025; Hazard Ratio = 0.33 [0.12, 0.92], p = 0.033, NNT = 6 [3,-34]). No statistically significant differences were found between the two groups on other assessed outcomes, including hospitalization, length of stay, and quality of life and disease severity measures. CONCLUSIONS: Systematic matching of patients initiated on depot with those initiated on oral antipsychotics showed that the oral-initiated patients were statistically significantly more likely to discontinue their medication. Findings highlight the importance of systematic matching of patient cohorts when comparing treatment outcomes in observational studies. PMH3 DETERMINANTS OF PSYCHIATRIC HOSPITAL ADMISSION IN SCHIZOPHRENIA Olfson M1, Marcus S2, Ascher-Svanum H3, Faries D3 1 Columbia University Medical Center, New York, NY, USA; 2Penn Social Policy & Practice, Philadelphia, PA, USA; 3Eli Lilly & Company, Indianapolis, IN, USA OBJECTIVES: Hospital admission is a common and costly event in schizophrenia. An analysis of phase 1/1A CATIE clinical trial data assessed various patient socio-demographic and clinical characteristics in relation to risk of psychiatric hospital admission. METHODS: We followed 1460 study participants from baseline until first schizophrenia-related hospital admission, study medication discontinuation, or 18 months. Stepwise Cox regression models assessed the adjusted hazard ratio (AHR) of hospital admission by baseline patient socio-demographic and clinical characteristics. RESULTS: In 869 person-years of follow-up, 203 patients were hospitalized. The adjusted hazards of hospital admission were not significantly related to patient socio-demographic characteristics. Increased risk of admission was linked to early age (<17 years) of first antipsychotic treatment (AHR: 2.09; 95%CI: 1.45–3.02), psychiatric hospital admission in past year (AHR: 2.92; 95%CI: 2.18–3.90), and DSM-IV alcohol (AHR: 1.55; 95%CI: 1.15–2.08) and drug (AHR: 1.50; 95%CI: 1.13–2.00) use disorders in the past 5 years. Severe (5–7) as compared with mild (1–3) baseline global clinical severity (AHR: 1.51; 95%CI: 1.03–2.23) (CGI-I), high (>20) as compared with low (7–15) positive symptoms (AHR: 1.53; 95%CI: 1.08–2.16) (PANSS-positive subscale), and low (0–2.2) as compared with high (>3.1) social function (AHR: 1.47; 95%CI: 1.04–2.08) (Heinrichs-Carpenter QLI) were related to significantly increased risk of hospital admission. As compared with olanzapine treatment assignment, quetiapine (AHR: 2.12; 95%CI: 1.37–3.27), perphenazine (AHR: 1.64; 95%CI: 1.02–2.65), and ziprasidone (AHR: 2.67; 95%CI: 1.62–4.39), though not risperidone (AHR: 1.40; 95%CI: 0.89–2.21), were also associated with increased hospital admission risk. Self-rated physical health (SF-12 PCS) and drug attitudes (DAI) were not significantly related to risk. CONCLUSIONS: In the treatment of schizophrenia, efforts to lower hospital admission risk should focus on patients with early onset disorders, recent inpatient admissions, severe positive symptoms, high global clinical severity, poor social function, and comorbid substance use disorders and should select an appropriate antipsychotic medication. PMH4 A PHASE IV STUDY OF THE EFFECTIVENESS OF QUETIAPINE EXTENDED RELEASE 600 MG ONCE A DAY TO CONTROL THE SYMPTOMS OF MANIC PHASE OF BIPOLAR DISORDER: THE EMMY TRIAL Segovia M1, Polanco AC1, Anaya P1, López RJ1, Gutiérrez D2 1 AstraZeneca, Naucalpan, Mexico; 2National Institute of Psychiatry, D.F., Mexico OBJECTIVES: To assess the efficacy of a 600 mg/day dose of Quetiapine Extended Release administrated once a day at evening as monotherapy or in combination with lithium or valproate for 21 days. METHODS: A multi-center phase IV study was designed to assess the efficacy of quetiapine extended release 600 mg per day either as monotherapy or combined therapy with lithium and valproic acid in the treatment A446 of patients with mania associated to Bipolar Disorder assessed by change in the Young Mania Rating Scale (YMRS) score and Clinical Global Impression (CGI) score from inclusion to day 21. Quality of life and safety/tolerability were measured with the Euro Quality of Life 5D (EQ5D), Work Productivity and Activity Impairment, Barnes Akathisia Rating Scale, Simpson-Angus Scale, physical examinations, adverse events, change of weight and other adverse events of special interest. The efficacy analysis was based on the modified intention-to-treat population, that included all patients who received study medication and who had a YMRS assessment at inclusion and at least one YMRS valid assessment after inclusion resulting in a total of 88 patients. RESULTS: YMRS total score reduction in 21 days was 20.55 points (95% CI, 22.82–18.27); P < 0.0001. CGI total score decreased in 2.41 units, (95% CI, 2.69– 2.13); P < 0.0001. The EQ5D index increased 0.21 points (95% CI, 0.16–0.27) P < 0.0001. The EQ5D Visual Analog Scale increased 23.49% (95% CI, 16.36%– 30.61%); P < 0.0001. CONCLUSIONS: Overall results shown in this study demonstrate an improvement in the control of patients in maniac phase of bipolar disorder with an increase in Quality of Life. PMH5 ESTABLISHING THE COMPARATIVE EFFICACY OF ALZHEIMER’S DISEASE THERAPY THROUGH SYSTEMATIC REVIEW AND COMPARATIVE ANALYSIS Modha R1, Wieffer H2, Pietri G2, Pueschner F3, Gaudig M3 1 Heron Evidence Development Ltd, London, UK; 2Heron Evidence Development Ltd, Luton, UK; 3Janssen-Cilag GmbH, Neuss, Germany OBJECTIVES: For therapeutic augmentation of impaired cholinergic transmission in Alzheimer Disease (AD), Acetylcholine-Esterase-inhibitors (ACHE-I; galantamine, donepezil, rivastigmine) are approved therapies in mild to-moderate AD. The NMDA receptor partial antagonist (memantine) is licensed for therapy of moderate–severe AD. In order to inform clinical decision-making about efficacy, safety and broader non-cognitive outcomes, we identified evidence from comparative and non-comparative studies. METHODS: A comprehensive search was conducted on Medline, Embase, conference abstracts and the Cochrane Library aimed to identify all randomised, placebo controlled trials (RCTs) reporting efficacy and/or safety outcomes and randomised, controlled comparative trials. In a second step, trials with drug dosing outside the approved European Summary of Product Characteristics were excluded. Eligibility of trials was assessed by two blinded reviewers; quality of trials were assessed by CONSORT. Meta-analyses were performed, reporting fixed and random effects using comparative analytical techniques. RESULTS: Fifty—eight studies fulfilled the inclusion criteria. In 45 trials, ACHE-I were tested against placebo, in 9 trials against another ACHE-I. For memantine, 9 placebo-controlled trials were identified. In most trials, patients were treated between 12–30 weeks; 9 RCTs reported longer-term outcomes, up to 24 months. After critical appraisal, 33 studies were included in further analyses. Meta-analysis of effects on cognition (ADAS-cog, MMSE, SIB) showed superiority of at least one ACHE-I versus placebo at the 3 months, 6 months and longer-time timepoint. Results for memantine indicated non-significant improvement at any assessed timepoint. Behavioural outcomes were less well reported but showed superiority versus placebo for galantamine and memantine measured by the Neuropsychiatric Inventory Scale. CONCLUSIONS: There is abundant evidence for the efficacy, safety and tolerability of ACHE-Is in the treatment of patients with mild to moderate AD. Data for improved behavioural functioning is limited though available for galantamine and memantine. PMH6 INCREASING ADMINISTRATIVE PREVALENCE OF ATTENTION-DEFICIT/ HYPERACTIVITY DISORDER (ADHD) IN GERMANY: EVIDENCE FROM NORDBADEN, 2003 TO 2008 Schlander M1, Schwarz O1, Trott GE1, Banaschewski T2, Scheller W3, Viapiano M4, Bonauer N4 1 Institute for Innovation & Valuation in Health Care, Wiesbaden, Germany; 2University of Heidelberg, Mannheim, Germany; 3Verband der Ersatzkassen (vdek), Stuttgart, Germany; 4KV Baden-Wuerttemberg, Karlsruhe, Germany BACKGROUND: In our age and gender specific analyses for calendar year 2003, we observed an administrative prevalence rate of ADHD in Nordbaden (a region in the South-West of Germany with a population of >2.7 million) of 0.53%, with a peak among 9-year old boys at 8.43%. From 2003 to 2008, methylphenidate prescriptions (defined daily doses, DDDs) in Germany increased 2.65-fold, raising concern about potentially inappropriate use. OBJECTIVES: To establish a longitudinal ADHD patient database and to assess changes of ADHD administrative prevalence rates by age and gender during the period from 2003 to 2008, in order to lay the foundation for further analyses of treatment and prescribing patterns. METHODS: The complete claims database of the organization of physicians registered with statutory health insurance [SHI] (Kassenaerztliche Vereinigung, KV) in Nordbaden/Germany was available for analysis, covering the total regional population enrolled in SHI (>2.2 million). Age and gender specific 1-year prevalence rates of ADHD were determined for years 2003 through 2008. RESULTS: During the 6-year period under study, the overall one-year ADHD prevalence rate increased from 0.53% to 0.90%. ADHD (hyperkinetic disorder: ICD-10, F90.0, F90.1) prevalence rates were highest in the age group 6–12 years (peak [2008] among nine-year-old children, 9,55%; boys, 12,31%), increasing continuously during the observation period (a) age group 6–12: from 4.75% (boys, 6.91%; girls, 2.46%) to 7.62% (boys, 10.44%; girls, 4.66%); (b) age group 13–17: from 1.73% (boys, 2.66%; girls, 0.75%) to 3.78% (boys, 5.69%; girls, 13th Euro Abstracts 1.74%); adults (age >18 years): from 0.04% (males, 0.05%; females, 0.03%) to 0.14% (males, 0.18%; females, 0.10%). CONCLUSIONS: German methylphenidate prescription growth outpaced the increase in ADHD diagnoses from 2003 to 2008. Further research seems warranted and has been initiated with regard to the underlying dynamics of physician group involvement, coexisting conditions, utilization patterns (treatment duration, intensity, switches), and regarding economic implications. PMH7 USING DRUG DISPENSING DATA TO STUDY THE VALIDITY OF PARASKAVEDEKATRIAPHOBIA Pechlivanoglou P, Rozenbaum MH, Le HH, Postma MJ, Vegter S University of Groningen, Groningen, The Netherlands OBJECTIVES: From tales of Christian tradition to essays of 17th century physicians, Friday has been identified as a day of ill luck. Amongst all Fridays of the year however the one coinciding with the 13th day of the month is known to be associated with extreme misfortune. Therefore, the fate of persons born on Friday the 13th (F13) warrants investigation. We investigated the effect of being born on Friday the 13th on the prevalence of a) cancer; b) diabetes; c) mental health disorders; and d) sexual disorders. METHODS: Persons born on Friday the 13th were identified from a drug utilization database. To control for seasonal variation and time-trends, persons born on Friday the 6th and Friday the 20th were selected as controls. Differences in prescription prevalences were analyzed using multivariate logistic regression. RESULTS: A total of 4202 cases and 8987 controls were identified. The number of cases was less than expected based on the number of controls (expected number 4396, P < 0.01), suggesting a survival disadvantage for F13 cases. Age (32.8 ± 22.6 vs. 32.7 ± 23.0) and gender (53.1% vs. 53.3% female) were similar. Multivariable logistic regression showed no increase in prescription prevalences for F13 cases (P > 0.1 for all drug classes). Exploratory analyses showed that F13 cases being prescribed antipsychotics were younger than controls (P = 0.03) and diabetic cases were more often male (P = 0.04). CONCLUSIONS: Although attenuating effects of survivor selection bias cannot be ruled out, the current study found no scientific basis for paraskavedekatriaphobia in relation to drug utilization. Significant differences in age and gender for specific subgroups may be attributed to the multiple comparison issue. Results will be presented on interaction effects of Friday the 13th coinciding with full-moon (n = 154) and good Friday (n = 113). PMH8 EXCESS MORTALITY RISK IN PATIENTS WITH PSYCHOSIS HOSPITALIZED IN JAPANESE NATIONAL MENTAL HOSPITALS Inagaki A1, Nakagawa A2, Fuwano S3, Itoh T4, Tsukada K4, Urata J5 1 Keio University, Fujisawa, Kanagawa, Japan; 2National Center of Neurology and Psychiatry, Kodaira, Tokyo, Japan; 3Fuwano Clinic, Jouetsu, Niigata, Japan; 4Kohnodai Hospital, International Medical Center, Ichikawa, Chiba, Japan; 5Sakuragaoka Memorial Hospital, Tama, Tokyo, Japan OBJECTIVES: To examine the standardized mortality ratio (SMR) in inpatients with schizophrenia and related psychosis. METHODS: To compare the mortality risk in Japanese with psychosis with that of the general population, we utilized two large studies, called “the JESS2000” and “the JESS2000 Follow-up Study.” In the JESS2000 study, 2,309 psychotic patients who had been hospitalized in Japanese national mental hospitals as of September 1, 2000 were included. In the JESS2000 Follow-up Study, the JESS2000 patients were followed up and those who had died as of September 1, 2005 were identified retrospectively. The expected number of deaths was calculated by multiplying the number of patients in each gender- and age-specific subgroup by the mortality rate derived from the Japanese abridged life table and adding all of the figures in each of these subgroups. The SMR was calculated by the observed number of deaths, identified in the JESS2000 Follow-up Study, divided by the expected number. RESULTS: Of 2309 original samples, 56.3% were male. On September 1, 2000, the mean age (SD) was 52.0 (14.6) years. The mean duration of hospitalization was 10.3 (12.0) years. The mean age at onset of psychosis was 23.7 (7.9) years. The deaths of 204 patients were confirmed as of September 1, 2005. Of these 204 deceased patients, 13 committed suicide, 6 died from accidents and 148 died of natural causes. No information regarding cause of death for the remaining 16 patients could be obtained. The expected numbers of deaths in this cohort was 100.67. Therefore the SMR for all causes in this cohort was estimated at 2.03 or more. CONCLUSIONS: Our findings were similar to the results of the systematic review by Saha et al., which reported the SMR for schizophrenia as 2.58. MENTAL HEALTH – Cost Studies PMH9 BUDGET IMPACT ANALYSIS OF SERTINDOLE IN THE TREATMENT OF SCHIZOPHRENIA IN POLAND Walczak J, Augusty ska J, Sołtys E, Nogas G Arcana Institute, Cracow, Poland OBJECTIVES: To evaluate the financial consequences of sertindole reimbursement for the Polish National Health Fund (NHF) budget. METHODS: Budget impact analysis was performed in a 3-year time horizon from two perspectives of NHF and a patient. Costs of oral antipsychotic drugs and EKG test of patients treated with sertindole were included. Two future scenarios were estimated: 1) with reimbursement of sertindole, and 2) without sertindole reimbursement. Target population was estimated using 13th Euro Abstracts Polish epidemiological and statistical data, and a systematic review (proportion of patients treated with antipsychotic drugs who withdrew because of adverse events). Current and predicted market shares were assessed on the basis of present sales level data. RESULTS: In the first scenario, NHF expenditures on atypical antipsychotic drugs will increase by: 13 mln PLN in 2010, 20 mln PLN in 2011 and 26 mln PLN in 2012. In the second scenario of no sertindole reimbursement NHF expenditures on atypical antipsychotic drugs will increase by: 11 mln PLN in 2010, 18 mln PLN in 2011 and 23 mln PLN in 2012. From patient’s perspective, expenditures will increase in 2010–2012 by 0.9 mln PLN 1.3 mln PLN and 1.7 mln respectively in scenario with sertindole reimbursement, while in a second scenario costs will increase in 2010–2012 by 1.1 mln PLN, 1.6 mln PLN and 2.0 mln PLN respectively. The 2009 weighted average exchange rate of Polish National Bank was c1 = PLN 4.3273. CONCLUSIONS: Reimbursement of sertindole would result in a minor increase in Polish NHF expenditures. Simultaneously, it would enlarge the scope of accessible therapies for patients intolerant to at least one other antipsychotic agent. PMH10 COST SAVING POTENTIAL OF GENERIC SUBSTITUTION: THE CASE OF ANTIDEPRESSANTS van der Westhuizen E, Burger JR, Lubbe MS, Serfontein JHP North-West University, Potchefstroom, South Africa OBJECTIVES: Generic medicines are generally considerably less expensive than branded products and their endorsement can lead to substantial savings in costs. The main objective of the study was to calculate potential cost savings that can be generated by generic substitution of antidepressants within the private health care sector of South Africa. METHODS: Data on computerised medicine claims of patients receiving one or more antidepressants during three consecutive years (i.e. 2004, 2005 and 2006) were elicited from a South African pharmaceutical benefit management company. a non-experimental, quantitative, retrospective drug utilization review was conducted and data were analyzed using the Statistical Analysis System® programme. Potential cost savings defined as the collective amount that could be saved annually by substituting the average price of innovator active substances for that of generic equivalent(s) were computed for criteria-eligible substances in the study population. All costs are expressed in $U.S. RESULTS: A total of 292,071 items (N = 5,982,869) on 273,673 prescriptions (N = 5,213,765) at a total cost of $8,652,289.48 (N = $207,316,483.10) were included in the study. Generic products constituted 58.7% (n = 292,071) of all antidepressants claimed, at a total cost of 28.2% (N = $207,316,483.10) of all incurred costs. With total substitution of the average price of all criteria-eligible innovators, a potential saving of 9.3% (N = $8,652,289.48) of the actual antidepressant cost over the study period, was calculated. CONCLUSIONS: In developing countries with limited health care budgets, such as South Africa, generic substances can be cost-saving treatment alternatives. Health care professionals, third-party payers, and patients all have fundamental roles to play in order to encourage greater use of generics. Medicine expenditure can thereby be reduced and access to scarce resources increased, in order to meet the pressing health care needs within South Africa. PMH11 BUDGET IMPACT ANALYSIS OF AMISULPRIDE IN TREATMENT OF SCHIZOPHRENIA IN POLAND Wilk D1, Rutkowski J1, Dziewiatka M1, Lis J2, Glasek M2, Plisko R1 1 HTA Consulting, Krakow, Poland; 2Sanofi-Aventis Poland, Warszawa, Poland OBJECTIVES: To estimate the impact of amisulpride continued reimbursement in schizophrenia treatment on payer’s budget in Poland. METHODS: The analysis was performed in 5-year time horizon from the payer (National Health Fund, NHF) perspective and payer + patient perspective. Only costs of medicines were included. On the base of IMS Health Poland sale data for years 2005–2010 linear regression was conducted to predict consumption and prevalence of antipsychotics in Poland. Cost data of medicines were obtained from Ministry of Health and medicine portals in case of lack of reimbursement. One-way sensitivity analysis were performed for the key input parameters. RESULTS: From the payer perspective, cost of amisulpride is approximately c7.13 million in 2010 and c9.15 million in 2014 and it represents from 4.62% in 2010 to 4.28% in 2014 of the total cost of schizophrenia treatment estimated at approximately c154.22 million in 2010 and c213.87 million in 2014. From the payer + patient perspective, cost of amisulpride is approximately c7.26 million in 2010 and c9.32 million in 2014 and it represents from 4.01% in 2010 to 3.85% of the total cost of schizophrenia treatment estimated at approximately c181.11 million in 2010 and c241.81 million in 2014. Increase of NHF and patients expenses is related to an increase of antipsychotics sales over a span of the next five years caused by expanding awareness of schizophrenia and the importance of treatment. CONCLUSIONS: Our findings suggest that the cost of treatment with amisulpride are at a reasonable level and represent a small proportion of the total costs of schizophrenia treatment both from the payer perspective and common payer + patient perspective. The declining trend in the share of amisulpride cost in total cost of schizophrenia is noticeable. Amisulpride is an alternative therapeutic option of schizophrenia treatment in Poland and its reimbursement from public funds is justified. A447 PMH12 COST IMPACT OF INITIATING PREGABALIN TREATMENT IN SWEDISH PATIENTS WITH GENERALIZED ANXIETY DISORDER Myrén KJ1, Löfroth E1, Sandelin R2 1 IMS Health, Stockholm, Sweden; 2Pfizer, Sollentuna, Sweden OBJECTIVES: To compare the health care costs 6 months prior to and 6 months after initiation of pregabalin in generalized anxiety disorder (GAD) patients in Sweden. METHODS: This was a retrospective longitudinal database study of GAD patients from the South-West region of Sweden (1.5 million inhabitants). Individual patient data on health care visits (outpatient, inpatient, primary care), costs, mortality and diagnoses were included from year 2000. Data from the Swedish Prescribed Drug Register were included from July 1, 2005 until December 31, 2007. Patients with a GAD (ICD-10 F41.1) diagnosis and who initiated pregabalin treatment in 2006 were included. Health care utilization was measured six months before and six months after pregabalin initiation. Patients with 2 or more prescriptions of benzodiazepines six months prior to pregabalin initiation were categorized as benzodiazepine patients. Non-parametric statistical tests (Mann-Whitney) were used for the cost and resource use comparisons. RESULTS: A total of 149 patients met the inclusion criteria, of whom 99 used benzodiazepines prior to pregabalin treatment. The number of in-patient (P < 0.05) and primary care (P < 0.05) visits significantly decreased in the 6-month period following pregabalin initiation. There was also a statistically significant reduction (p = 0.0004) in overall health care costs from SEK 71,000 (c7,408) to SEK 45,000 (c4,695). Among the benzodiazepine users (n = 99), the number of in-patient visits (p = 0.0213), days in hospital (p = 0.0026) and primary care visits (p = 0.0121) were all statistically significantly reduced following pregabalin initiation. The reduction in total cost (from SEK 79,000 to SEK 43,000; c8,243 to c4,486) among the users of benzodiazepines was also statistically significant (p = 0.0001). The decrease in total cost, in all patients as well as benzodiazepine treated, was to a large extent explained by the decrease in in-patient costs. CONCLUSIONS: Initiating treatment with pregabalin in GAD patients significantly reduced health care utilization and costs during the following 6 months. PMH13 COSTS ASSOCIATED WITH ANTIPSYCHOTIC MEDICATIONS FOR PATIENTS WITH A BIPOLAR DIAGNOSIS AT CLINICALLY RECOMMENDED DOSES Stafkey-Mailey D1, Richards K2, Rascati K3, Ott CA4, Goddard A5, Alvir J6, Sanders K6, Mychaskiw MA6 1 University of South Carolina, Columbia, SC, USA; 2University of Texas at Austin, Austin, TX, USA; 3University of Texas, College of Pharmacy, Austin, TX, USA; 4Purdue University, Indianapolis, IN, USA; 5Indiana University, Indianapolis, IN, USA; 6Pfizer Inc, New York, NY, USA OBJECTIVES: There is accumulating evidence of sub-therapeutic second-generation antipsychotic (SGA) dosing for patients diagnosed with bipolar disorder, leading to suboptimal control of disease and higher overall treatment costs. The objectives of this study were to identify Medicaid bipolar patients receiving clinically effective doses of SGAs and compare their medical costs. METHODS: Patients with bipolar disorder taking an oral SGA (aripiprazole, olanzapine, quetiapine, risperidone or ziprasidone) were identified in Medicaid claims databases (2005–2008) from 8 US states. Patients were followed for 18 months (6-month pre-index period during which patients did not receive an SGA, followed by a 12-month post-index utilization period to determine total costs). For patients on recommended dosing, costs were compared using a generalized linear model with a gamma distribution and log-link function. Baseline covariates (age, gender, race, pre-index costs, Charlson co-morbidity score, and specific psychiatric co-morbidities) were adjusted for. Ziprasidone-treated patients comprised the reference group. RESULTS: A total of 2446 patients met inclusion criteria, with 45% (N = 1102) taking clinically effective doses by day 61 of their follow-up period. Patients on quetiapine had the lowest percentage of effective dosing at 26% (N = 280/1072). Other results were aripiprazole 77% (N = 336/448), olanzapine 52% (N = 118/226), risperidone 50% (N = 238/474), and ziprasidone 58% (N = 130/226). Regression analyses indicated that mental health-related prescription costs (P < 0.01) and all-prescription costs (P < 0.01) were statistically significantly lower for the risperidone group compared to the ziprasidone group. There were no significant differences between the groups for total mental health-related costs or total all-cause costs (includes prescription and medical services). CONCLUSIONS: Less than half of the patients in this sample were prescribed clinically recommended doses 2 months after their initial start. Among patients using recommended doses, while those on risperidone had lower prescription costs, there were no significant differences for total costs compared to patients taking ziprasidone. PMH14 COST ANALYSIS OF ADVERSE EVENTS ASSOCIATED WITH TREATMENT OF BIPOLAR DISORDER: A COMPARISON BETWEEN ARIPIPRAZOLE AND OLANZAPINE IN THE SPANISH HEALTH SYSTEM Ramirez de Arellano A1, Rubio-Terrés C2, Baca E3, Riesgo Y4, Luque J1, Mena E5 1 Bristol-Myers Squibb Iberia, Madrid, Spain; 2HealthValue, Madrid, Spain; 3Hospital Universitario Puerta de Hierro, Madrid, Spain; 4Bristol-Myers Squibb Iberia, Malaga, Spain; 5 Otsuka Pharmaceuticals, Barcelona, Spain OBJECTIVES: This study investigates the health care costs of adverse events (AE) associated with treatment of bipolar disorder with two atypical antipsychotics (AA): aripiprazole (ARI) and Olanzapine (OLA). METHODS: It was performed a cost analysis through a markov model considering the following health states: no existence A448 of adverse events (NAE); extrapyramidal symptoms (EPS); weight gain (WG); and sexual dysfunction (SD). The transition probabilities amongst health states were estimated from two different meta-analysis of clinical trials and from a retrospective Spanish study. The health care costs associated to each health state were obtained from a published Spanish study. It was used the minimum acquisition cost per mg. of the mean daily dose, for each AA, which is regarded as a relevant efficiency criterion in Hospital Pharmacy Departments. The time horizon applied in the analysis was 12 months. a probabilistic sensitivity analysis was performed for all the variables involved in the analysis via Monte Carlo simulations. All costs were inflated to 2009 costs using Spanish Health System pay and prices index. RESULTS: In comparison with OLA, the treatment with ARI generates annual average cost savings per patient of c−688.70 ± 21.69 (CI 95% −614.52; −729.18). In the most unfavourable scenario for ARI, that in which we assumed that ARI may have a similar rate of sexual dysfunction than that of quetiapine (i.e. the lowest rate amongst AA) the costs savings per patient would be c–270,94 ± 17,11 (CI 95% −237,20; −303,48). CONCLUSIONS: The results of this analysis show that patients treated with aripiprazole demonstrate lower adverse events costs in comparison to olanzapine. This difference may generate significant cost savings to the Spanish health system in the treatment of patients affected by bipolar disorders. The robustness of the results was tested via a probabilistic sensitivity analysis. PMH15 REAL-WORLD TREATMENT PATTERNS AND HEALTH CARE RESOURCE UTILIZATION IN GENERALISED ANXIETY DISORDER (GAD): A RETROSPECTIVE UNITED STATES DATABASE ANALYSIS Locklear J1, Wade S2, Palmer LA3, Järbrink K4, Toghanian S4 1 AstraZeneca Pharmaceuticals, Wilmington, DE, USA; 2Wade Outcomes Research and Consulting, Salt Lake City, UT, USA; 3Thomson Reuters, Washington, DC, USA; 4 AstraZeneca R&D, Mölndal, Sweden OBJECTIVES: Real-world treatment patterns and health care resource utilization for patients with GAD in the United States are not fully defined. METHODS: Data from the 2003–2007 Thomson Reuters MarketScan® Commercial Claims and Encounters and Medicare Supplemental and Coordination of Benefits databases were utilised for analysis. For the current analysis, participants with a diagnosis of GAD (ICD-9 CM 300.02) between January 1, 2004 and December 31, 2006 were included. The sample was divided into the following five subgroups: patients receiving non-pharmacological treatment, first-line therapy only, first-line + augmentation within 90 days of index prescription, first-line + switch within 90 days of index prescription and second-line therapy only. Additionally, GAD patients were compared with healthy controls without GAD or other mental health conditions. RESULTS: In total, 23,553 GAD patients (mean age range: 41.6–48.1 years; 56.1–68.6% female across the five groups) were included (non-pharmacological, n = 7055; first-line only, n = 6538; first-line + augmentation, n = 903; first-line + switch, n = 1953; second-line only, n = 7104). Paroxetine was the most commonly used first-line treatment at index (first-line cohorts: 48%, 42%, 41%, respectively). In the second-line only cohort, a benzodiazepine (37%) or second-line SSRI/SNRI (27%) were the most commonly used agents at index. Benzodiazepines were the most commonly prescribed agents for augmentation of, or switching from, first-line treatments (augmentation 52%, switch 44%). Overall, GAD patients had higher health care utilization and significantly higher total health care costs versus healthy controls (mean per patient: $8058 vs. $2938, P < 0.0001). On average, GAD patients incurred an additional $425/month in direct health care costs in the 1-year post-index period versus healthy controls. CONCLUSIONS: The realworld management of GAD is both complex and costly. Paroxetine was the mostwidely used first-line treatment for GAD. Benzodiazepines were the most widely used agents for augmentation of, or switching from, first-line treatments, and for secondline therapy. Total health care costs were 2.7 times higher for GAD patients compared with healthy controls. PMH16 THE ECONOMIC AND HUMANISTIC BURDEN OF ILLNESS IN GENERALISED ANXIETY DISORDER (GAD): A RETROSPECTIVE DATABASE ANALYSIS IN EUROPE Toghanian S1, DiBonaventura M2, Järbrink K1, Locklear J3 1 AstraZeneca R&D, Mölndal, Sweden; 2Kantar Health, New York, NY, USA; 3AstraZeneca Pharmaceuticals, Wilmington, DE, USA OBJECTIVES: There is a paucity of published evidence estimating the economic and humanistic burden of illness in GAD. We report results of a retrospective database analysis examining the burden of GAD in Europe. METHODS: Data were derived from the European (France, Germany, UK, Italy, Spain) National Health and Wellness Survey database for 2008. The database captures information from adults (≥18 years) and is representative of the adult population in each country. Respondents reporting a diagnosis of GAD were propensity-score matched 1:1 to non-GAD controls on country, age, gender and employment status. Data were extracted on GAD medication use, resource utilization (emergency room visits, hospitalizations and health care provider visits) and work productivity (using the Work Productivity and Activity Impairment questionnaire) to calculate direct and indirect costs. Health-related quality of life (HRQoL) was derived using SF-12 mental and physical summary scores. Utilities were derived from SF-6D preference scores and used to calculate cost per qualityadjusted life-year (QALY). RESULTS: Of 53,524 respondents, 3,669 were assigned to the GAD group. GAD respondents accrued considerably higher direct (medication and health care resources) plus indirect (work productivity loss) costs (per person/ 13th Euro Abstracts year) versus controls across pooled European countries (c5,308.80 vs. c2,441.10; P < 0.0001) and for each country (France c6,083.70 vs. c2,896.30; Germany c12,797.00 vs. c4,876.10; UK c4,021.70 vs. c2,011.20; Italy c3,514.50 vs. c1,869.60; Spain c5,051.70 vs. c1,954.80; P < 0.0001 vs. controls within each country). Direct costs were driven by hospitalizations and psychologist/psychiatrist visits. Total costs increased with GAD severity from c4,094.00 for respondents with mild GAD to c7,753.10 for those with severe GAD. HRQoL was significantly poorer for GAD versus non-GAD respondents (P < 0.0001). Costs/QALY increased with GAD severity from c6,795 for mild GAD to c15,286 for severe GAD. CONCLUSIONS: The economic and humanistic burden of illness of GAD across Europe is considerable and increases with the severity of disease. PMH17 COST OF ATTENTION DEFICIT/HYPERACTIVITY DISORDER IN GERMANY Braun S1, Zeidler J1, Linder R2, Ahrens S2, Manouguian MS2, Greiner W3 1 Leibniz University of Hannover, Hannover, Germany; 2Scientific Institute of Techniker Krankenkasse for Benefit and Efficiency in Health Care, Hamburg, Germany; 3University of Bielefeld, Bielefeld, Germany OBJECTIVES: Data concerning costs of patients with Attention deficit/hyperactivity disorder (ADHD) are scarce in Germany. Aim of this claims data analysis was to examine the costs of ADHD from the perspective of the statutory health insurance. METHODS: Nation-wide claims data of a major statutory health insurance fund was used to evaluate the overall and ADHD-related costs of ADHD-patients in 2008. All costs for outpatient care, inpatient care, pharmaceuticals, rehabilitation, occupational therapy as well as devices and aids, and sick leave payments were analyzed. To calculate ADHD-related costs the overall health care costs of the identified ADHDpatients were compared to an age and gender matched control group. RESULTS: Based on the used identification algorithm 30,264 ADHD-patients were identified. Mean overall costs of c3802 in the year 2008 were incurred from the health insurance perspective. c1704 (45%) were incurred by occupational therapy as well as devices and aids, and c779 (20%) were due to inpatient care; c751 (20%) resulted from outpatient care. Costs for pharmaceuticals were c483 (13%) and for rehabilitation were c35 (≤1%). Costs for sick leave payments came to c50 (1%). The matched control group contains 404,565 patients. Compared to this control group the incremental mean costs of ADHD-patients were EUR 2,744. EUR 1,214 of these resulted from occupational therapy as well as devices and aids, c586 from inpatient care and c517 from outpatient care. Mean additional costs for pharmaceuticals in the ADHDgroup were c367. CONCLUSIONS: The major cost driver in ADHD from a health insurance perspective in Germany is occupational therapy as well as devices and aids. PMH18 THE COSTS OF DEPRESSION IN SWITZERLAND 1 2 Tomonaga Y , Szucs TD 1 Institute of Social and Preventive Medicine, University of Zurich, Zurich, Switzerland; 2 Institute of Pharmaceutical Medicine/European Center of Pharmaceutical Medicine, University of Basel, Basel, Switzerland OBJECTIVES: To investigate the burden of depression in the Swiss population. The costs for the management of depressive patients will be analyzed for different severity classes of disease, based on the Hamilton depression rating scale, over a period of 12 months following diagnosis. METHODS: A prospective, multicentre, non-interventional study in psychiatrist practices was carried out. Patients who have been diagnosed with depression in the last three years were included. Patient’s characteristics and resource utilization in the first twelve months after diagnosis were collected. Costs analysis, subdivided in direct and indirect costs, was performed for three depression classes (mild, moderate, severe), according to the Hamilton depression score (HAMD17). Costs were also extrapolated to the national level. RESULTS: A total of 556 patients were included. Hospitalization and hospitalization days were directly correlated with disease severity (P < 0.001). Medical resource utilization linked to depression and antidepressant treatments were also correlated to disease status. Severe patients reported a significantly higher number of workdays lost and were significantly more often in disability insurance. Total direct costs per person, mainly due to hospitalization costs, were CHF 4,823 for mild, CHF 13,251 for moderate, and CHF 22,138 for severe depressions. Indirect costs, mainly due to workdays lost, resulted in CHF 11,892 for mild, CHF 17,267 for moderate, and CHF 22,710 for severe depressions. Extrapolation at national level resulted in a total burden of about CHF 11 billion. CONCLUSIONS: The burden of depression in Switzerland was estimated to be around CHF 11 billion. Costs of depression were directly related to disease severity. However, since many depressions remain unreported and since this analysis only included individuals between 18 and 65 years of age, it is reasonable to suppose that the total burden of depression may be even greater. PMH19 BURDEN OF ILLNESS OF TREATMENT RESISTANT DEPRESSION McInnis MM1, Halseth MJ2, Cyr P1, Howland RH3, Bockstedt Santiago L2, Goss T1 1 Boston Healthcare Associates, Inc., Boston, MA, USA; 2Medtronic, Inc., Minneapolis, MN, USA; 3University of Pittsburgh School of Medicine Western Psychiatric Institute and Clinic, Pittsburgh, PA, USA OBJECTIVES: Major depressive disorder (MDD) is a leading cause of disability, morbidity, and mortality worldwide. The lifetime prevalence in the US is 17%. Treatment resistant depression (TRD) is generally defined as failure to achieve remissions 13th Euro Abstracts despite adequate treatment. About 30% of patients do not achieve remission after 4 different antidepressant treatment trials (Rush et. Al., 2006). a few studies have examined the economic burden of TRD, but none have investigated the cost associated with more chronic and extensive forms of TRD characterized by non-response to four or more treatment trials. The objective of this study is to compare the direct medical expenditures of TRD patients to chronic MDD patients. METHODS: Patients with chronic MDD (defined as 2 or more years of continuous treatment) and patients with very severe TRD (defined as switching, augmentation, or addition of at least four qualifying therapies) were identified in the PharMetrics Patient-Centric Database. Medical resource utilization (MRU) and payer costs per patient per year enrolled in the plan were compared. Patients were matched on duration of enrollment, length of depression, medications taken, and age. Statistical significance was assessed using Student’s t-tests. RESULTS: The average medical expenditures per TRD (n = 31,803) patient per year enrolled was $12,260 compared to $7,619 for non-TRD chronic MDD patients (n = 50,092) (p < 0.001), representing 60.9% higher costs per year. The differences in annual medical expenditures per TRD patient were primarily driven by higher medical and pharmacy MRU (number of services per patient) relative to the chronic MDD group (70.3 vs. 107.6, p < 0.001). CONCLUSIONS: Our results demonstrate that TRD patients accrued significantly higher per patient medical costs than MDD patients due to higher medical resource use. These findings suggest that more effective non-pharmacological alternative interventions are needed. PMH20 IMPACT AND COSTS OF HOSPITALIZATION IN SCHIZOPHRENIA 1 2 2 3 Zeidler J , Slawik L , Fleischmann J , Greiner W 1 Leibniz University of Hannover, Hannover, Germany; 2Janssen-Cilag GmbH, Neuss, Germany; 3University of Bielefeld, Bielefeld, Germany OBJECTIVES: Data concerning impact and costs of hospitalization in patients with schizophrenia are scarce in Germany. Aim of this claims data analysis was to examine the impact and costs of hospitalization in schizophrenia from the perspective of a major statutory health insurance fund. METHODS: A nation-wide database was used to evaluate the impact and related costs of hospitalization in schizophrenia (ICD-10 F20.x) between 2004 and 2006. All hospitalised patients were identified based on claims data and schizophrenia related costs for outpatient care, inpatient care, medications, rehabilitation, occupational therapy, and sick leave payments were analyzed before, during and after the hospitalization. RESULTS: Data from 4126 hospitalised patients were available, with 46% being female and a mean age of 42 years. Mean length of the index hospitalization was 45.9 days and mean health insurance costs of c9366 incurred during the index hospitalization. In the 3 months before the index hospitalization, average costs of c297 and in the 3 months after the index hospitalization costs of c2322 incurred for outpatient care, inpatient care, medications, rehabilitation, occupational therapy, and sick leave payments. 42.5% of the patients had at least one rehospitalization after the index hospitalization and mean costs of c8710 incurred during the first rehospitalization. In the 4 weeks before the first rehospitalization average costs of c1310 and in the 3 months after the rehospitalization costs of c2834 incurred. With any further rehospitalization costs for the inpatient care itself are decreasing slightly but overall costs for schizophrenia related medical care before and after the event are increasing. CONCLUSIONS: Impact and costs of hospitalization in schizophrenia are significant, especially if the costs before and after hospitalization were taken into account. Rehospitalizations are a frequent event in schizophrenia care with high impact on costs. Further studies including the predictors of hospitalization are needed. PMH21 SOCIETAL COSTS OF BIPOLAR DISORDER—THE CASE OF SWEDEN 1 2 3 2 4 Ekman M , Granström O , Omerov S , Jacob J , Landén M 1 i3 Innovus, Stockholm, Sweden; 2AstraZeneca Nordic MC, Södertälje, Sweden; 3Northern Stockholm Psychiatry, Stockholm, Sweden, 4Gothenburg University, Göteborg, Sweden OBJECTIVES: To investigate the health care resource utilization and costs for Swedish bipolar disorder patients. METHODS: Registry data on socio-demographics and disease-related resource use during 2006–2008 was collected for 1846 patients with bipolar disorder. Health care visits and hospitalizations were obtained from the Northern Stockholm psychiatric clinic, while data on pharmaceuticals and sick leave were obtained from the national pharmaceutical registry and the Swedish social insurance agency, respectively. Community care data was obtained from the Swedish quality registry for bipolar disorder (BipoläR). Indirect costs were valued according to the human capital method. Costs across mood episodes (manic, depressive, and unspecified) as well as remission were calculated. a patient for which >70 days had passed since the last new episode or inpatient visit, was defined as being in remission. RESULTS: The mean annual cost per patient with bipolar disorder was estimated at c24,300 (SEK 242,810) in 2009 prices, of which 23% represented direct treatment costs. The costs per month for active bipolar disorder were c6800 during depression, c6300 during mania, EUR 6,100 during unspecified episodes, and c1500 for patients in remission. a patient who was not hospitalized during 2006–2008 generated a mean annual cost of c18,900, whereas a patient who was hospitalized at least once every year 2006–2008 generated mean annual costs of c67,600. CONCLUSIONS: Societal costs for bipolar disorder were more than 3.5 times higher for patients who were admitted at least once per year in 2006–2008 compared to those who were not hospitalized. Monthly costs during a depressive or manic episode were more than 4 times higher than during remission. Attempts to prevent mood episodes by means of effective treatment may not only decrease patient suffering, but also have the potential to dramatically reduce the societal cost of illness. A449 PMH22 RESULTS OF THE GERMAN IDA STUDY—ASSESSING THE FINANCIAL IMPACT OF INFORMAL CARE AMONGST COMMUNITY LIVING DEMENTIA PATIENTS Schwarzkopf L1, Kunz S1, Wunder S2, Marx P3, Lauterberg J4, Gräßel E5, Holle R1 1 Helmholtz Zentrum München—German Research Center for Environmental Health (GmbH), Neuherberg, Germany; 2AOK-Bayern /die Gesundheitskasse, Nürnberg, Germany; 3 Pfizer Deutschland GmbH, Berlin, Germany; 4Federal Association of the AOK, Berlin, Germany; 5University Hospital Erlangen, Erlangen, Germany OBJECTIVES: Rising life expectancy is associated with increasing prevalence of dementia in European countries. With progressing disease severity patients’ call on health care services and social support grows. Several studies stress the burden imposed on family members caring for dementia-patients. However, empirical data assessing the economic value of informal care compared to health insurance expenditures is scarce. Within the cluster-randomized IDA study health care service utilization and informal caregiving time were assessed for 383 community living individuals suffering from mild to moderate dementia. METHODS: To examine costs from a health insurance perspective administrative data on all types of services provided were collected and valued with corresponding administrative prices over a three-year period. Patients’ caregivers reported in yearly interviews the hours daily dedicated to the patient in care and supervision. The time provided for informal caregiving was valued by applying hourly wages of a nursing service and a domestic help respectively. Total costs and cost components were calculated from societal perspective and payers’ point of view and analyzed by relevant subgroups. RESULTS: Caring for a homedwelling dementia-patient amounts annually to around c47.000 from societal perspective with informal care covering nearly 80% of this sum. For a patient with moderate dementia values assigned to informal care are approximately 70% higher than for a person with mild dementia. Health insurance has to spend c10,000 for an average dementia-patient per year, with services for long-term care representing the most costly part. CONCLUSIONS: Informal care is the major cost component in care for dementia, and it is strongly rising with disease progression. Changes in family structures and traditional living arrangements thus pose an enormous challenge regarding the future organization of dementia care. To maintain today’s care-setting, concepts fostering community-based dementia care and support to family caregivers need to be further developed. PMH23 COSTS OF A COMMUNITY SUPPORT PROGRAM FOR DEPRESSION: RESULTS FROM THE BEAT THE BLUES (BTB) TRIAL Jutkowitz E1, Pizzi LT1, Gitlin L1, Foley K2, Chernett N1, McCoy M3, Dennis M1 1 Thomas Jefferson University, Philadelphia, PA, USA; 2Thomas Jefferson University, Newtown, PA, USA; 3Center in the Park, Philadelphia, PA, USA OBJECTIVES: Literature on the costs of community support programs is scant. Beat the Blues (BTB) is a home support program for depression which involves trained social workers who meet with participants to identify care management concerns, make referrals and linkages, provide depression education, develop tailored action plans to accomplish identified behavioral goals and enhance engagement in pleasurable activities, and teach stress reduction techniques for managing daily stressors (e.g., deep breathing). The primary objective of this study is to assess the cost of implementing BTB in African American elders with depressive symptoms. METHODS: The cost analysis is piggybacked onto the parent BTB trial, which employs randomized twogroup experimental design (treatment vs. control). Candidates are screened twice for depressive symptoms (PHQ-9 score > 5) to determine eligibility. Participants are randomly assigned to intervention or waitlist control. BTB is administered over 4 months with assessments in both groups at baseline and 4-months. Cost components comprise 4 main categories (Screening, Intervention Delivery, Supervision, and Other). Senior center management and social workers (interventionists) kept detailed logs of time spent conducting each component. The cost of conducting the intervention was calculated as the time spent performing each task multiplied by wage rate of the individual who performed the task. Non-time related cost (material and mileage) was also calculated. RESULTS: A total of 166 participants received the intervention. Average cost of BTB was $342 per person. The most costly aspect of BTB per person was cost of program delivery ($275), followed by materials ($48).The least costly aspect of the intervention was time spent supervising interventionists ($9). CONCLUSIONS: This study informs decision makers about the resources required for depression community support interventions, which are not typically reimbursable through private or public health care programs. Results can also inform decision making regarding the costs of such interventions in relation to pharmacological treatments. PMH24 HEALTH CARE UTILIZATION OF MAJOR DEPRESSIVE DISORDER PATIENTS Myrén KJ, Löfroth E, Dahlberg E IMS Health, Stockholm, Sweden OBJECTIVES: To analyze the health care utilization by patients diagnosed with major depressive disorder (MDD). METHODS: This was a retrospective longitudinal descriptive database study of the utilization of health care by patients with MDD from the South-West region of Sweden (1.5 million inhabitants). All patients who were diagnosed with MDD (ICD-10 F32 or F33) during 2007 were included in the study. All costs from health care visits when an MDD was the primary diagnose were included as well as all cost from all dispensed antidepressant (ATC N06A) drugs. RESULTS: A total of 37,445 patients had at least one health care visit with an MDD A450 diagnoses during 2007 in the South-West region of Sweden. Sixty-two percent of the patients had at least one visit to the primary care, 35% had been treated as out-patient, 9% had been treated by private physician, and 7% had been treated by in-patient care; 73% had at least one dispensed drug. The total health care cost, including drug cost, was 683 million SEK. 35% were from drugs and 23%, 26%, 14%, and 1% were from in-patient, out-patient, primary, and private care. During 2007 the average drug cost per patient, who utilize the current treatment, was 8847 SEK for drugs. The average cost for in-patient care was 64,576 SEK and 13,859 SEK for out-patient care. For primary and private care the average cost was 4268 SEK and 1342 SEK. CONCLUSIONS: The cost for antidepressant drugs was the largest part of the total health care cost for MDD, 35%. The in-patient care was the greatest cost per patient. The relevant question is what mix of drugs should be prescribed in order to minimize the total health care cost. Further research needs to be preformed. PMH25 COMPARISON OF ESCITALOPRAM VS. CITALOPRAM AND VENLAFAXINE IN THE TREATMENT OF MAJOR DEPRESSION IN SPAIN: CLINICAL AND ECONOMIC CONSEQUENCES Sicras-Mainar A1, Blanca-Tamayo M2, Gimeno-de la Fuente V3, Salvatella-Pasant J3, Navarro-Artieda R4 1 Directorate of Planning, Badalona Serveis Assistencials, Badalona, Barcelona, Spain; 2 Badalona Serveis Assistencials SA, Barcelona, Spain; 3Lundbeck España SA, Barcelona, Spain; 4 Hospital Germans Trias i Pujol, CIBEREHD, Barrcelona, Spain OBJECTIVES: Population based study to determine clinical consequences and economic impact of using escitalopram (ESC) vs. citalopram (CIT) and venlafaxine (VEN) in patients initiating treatment for a new episode of major depression (MD), in real-life conditions of outpatient practice. METHODS: Observational, multicenter, retrospective study conducted using computerized medical records (administrative databases) of patients treated in six primary care centers and two hospitals between January 2003 and March 2007. Study population: patients >20 years of age diagnosed with a new episode of MD who initiate treatment with ESC, CIT or VEN, and whithout any antidepressant treatment within the previous 6 months, who were followed for 18 months or more. Main variables: socio-demographic variables, remission, comorbidity, annual health care costs (medical visits, diagnostic and therapeutic tests, hospitalizations, emergency room and psychoactive drugs prescribed) and non-health care costs (productivity loss at work). Statistical analyses: logistic regression and ANCOVA models. RESULTS: A total of 965 patients (ESC = 131; CIT = 491; VEN = 343) were included. ESC-treated patients were younger, with a higher proportion of males, and a lower specific comorbidity (P < 0.01). ESC-treated patients achieved higher remission rates (58.0%) compared to CIT (38.3%) or VEN (32.4%) patients (P < 0.001) and had lower productivity work ESC (37.9 days) vs.CIT (32.7 days) or vs. VEN (43.8 days), p = 0.042. No differences in average/unit costs of psychoactive drugs were observed between the ESC and CIT groups (c294.7 vs. c265.2), with higher costs seen in the VEN group (c643.0), p = 0.003. In the corrected model, total (health care and non-health care) costs were lower with ESC (c2276.2) vs. CIT (c3093.8), p = 0.047 and VEN (c3801.2), p = 0.045. CONCLUSIONS: ESC appears to be dominant in the treatment of new MD episodes when compared to CIT and VEN, resulting in higher remission rates and lower total costs. PMH26 COST ANALYSIS OF METHADONE MAINTENANCE THERAPY (MMT) PROGRAM IN MALAYSIA Abdul Ghani N, Shafie AA, Hassali MAA Universiti Sains Malaysia, Penang, Malaysia OBJECTIVES: To analyze the cost of MMT treatment for patient with substance use disorder. METHODS: This was a retrospective study involving 185 opiate dependence patients who were receiving MMT treatment program in Malaysian’s government hospitals starting December 2005 until February 2010. The study was done from the provider perspective using activity based costing that includes both capital and variable cost associated with MMT program. RESULTS: Most of the respondents were male (99%) with mean age of 38.5 years. Most (72.5%) of them were Malay and 50.5% working in semi or unskilled manual job with basic salary less than MYR1000 (US$312). Most (99.5%) were smoker with history of heroin addiction of more than 10 years. The median length of MMT treatment was 14 months. Fifty-seven percent of the patients were suffering from chronic disease. The capital cost for MMT program was US$8013.16 and the variable cost of providing a month of treatment per patient was US$50.43. Approximately 47.19% of this was for methadone and personnel cost accounted for 31.94%. The variable cost was statistically higher (P < 0.05) in patients present of chronic disease. CONCLUSIONS: This study estimated the variable cost of MMT treatment per patient in Malaysia (US$50.43/month) is less expensive compare USA (US$220/month, Jones et al., 2009). These cost data may be useful to policymakers and researchers for further developing the program. PMH27 RECENT TRENDS IN PSYCHIATRIC PRESCRIPTION DRUG SPENDING 1 2 Mark T , Long S 1 Thomson Reuters, Washington, DC, USA; 2Thomson Reuters, Cambridge, MA, USA OBJECTIVES: To describe new data on recent trends in U.S. psychiatric prescription drug spending and utilization over the period 2003 to 2008 and to understand drivers of those new trends. METHODS: SAMSA Spending Estimates were coupled with MEPS data to examine trends in mental health drug spending and utilization through 2005. 1997–2008 Thomson Reuters MarketScan Commercial Claims data were used 13th Euro Abstracts to decompose spending where: Total Expenditures = Price X Quantity, and ΔCost/ Enrollee = ΔCost/Day X ΔDays/User X ΔUsers/Enrollees. The following drugs were included in the analysis: Antidepressants, Antipsychotics, Stimulants, and Anxiolytic/ Sedative/Hypnotics. RESULTS: The average annual growth rate in mental health prescription drug spending dropped from 28% in 1998 to 3% in 2008. The rate of growth has stayed below 10% since 2005, and was negative in 2007. For persons with private insurance, the average annual expenditure growth rate overall was 6% during the years 2001–2008, where 2% was attributable to days/user, 2% was due to users per population, and 2% was due to cost/day. In contrast for the years 1997–2001, the average growth rate overall was 18%, where 3% was attributable to days/user, 7% was due to users per population, and 8% was due to cost/day. CONCLUSIONS: Mental health prescription drug spending growth has slowed in recent years. This is due primarily to slower growth in additional users and slower price growth. The lower price growth is mainly due to generic entries starting in 2005, particularly within the antidepressant drug class. PMH28 EFFECT OF PREGABALIN ON COSTS AND CONSEQUENCES IN PATIENTS WITH REFRACTORY SEVERE GENERALIZED ANXIETY DISORDER AND CONCOMITANT SEVERE SYMPTOMS OF DEPRESSION IN DAILY MEDICAL PRACTICE Rejas J1, Carrasco JL2, Alvarez E3, Olivares JM4, Vilardaga I5, Pérez M6, Lopez Gomez V6 1 Pfizer España, Alcobendas/Madrid, Spain; 2Department of Psychiatry, Hospital Clínico San Carlos, Madrid, Spain; 33Department of Psychiatry, Hospital de la Santa Creu i San Pau, Barcelona, Spain; 4Hospital Meixoeiro, Complejo Hospitalario Universitario, Vigo, Spain; 5 Department of Biometric and Statistics, European Biometric Institute, Barcelona, Spain; 6 Department of Neuroscience, Medical Unit, Pfizer Spain, Alcobendas, Madrid, Spain OBJECTIVES: To analyze the clinical and economic effects of the use of Pregabalin (PGB) in patients with refractory severe Generalized Anxiety Disorder (GAD) and severe concomitant depressive symptoms in daily medical practice. METHODS: Data were obtained from a post-hoc analysis of a 6-month, prospective, non-interventional study conducted in outpatient psychiatric clinics to ascertain the impact of broadening GAD diagnostic criteria. This before-after study included PGB naïve patients above 18 years, with GAD (using DSM-IV criteria), refractory to anxiolytics for a minimum of 3 months, severe symptoms of anxiety (HAM-A > 24) and depression (MADRS > 35), who received flexible doses of PGB either as monotherapy or add-on, per psychiatrist judgment. Changes in HAM-A and MADRS and responders rates (reduction > 50% of baseline scoring) were the study end-points, together with health care resources utilization and corresponding costs. RESULTS: A total of 159 patients [69.2% women, mean age 45.9 (12.6) years] fulfilled criteria for analysis. Ninety percent or more patients were previously exposed to benzodiazepines and/or antidepressants [mean 2.7 (1.3) drugs]. Adding PGB [mean dose: 223.1 (126.3) mg/day] reduced both anxiety and depressive symptoms, respectively, in HAM-A and MADRS scales, by −57.9% (from 35.5 + 5.8 to 14.8 + 9.4; P < 0.001, effect size: 3.57) and −58.1% (from 39.4 + 4.3 to 16.5 + 10.3; P < 0.001, effect size: 5.33). As a result, responder rates were 63.1% and 62.9%, respectively. Costs reductions in medical visits [−c1022 (−1,376; −669), P < 0.001] and hospitalizations [−c144 (−245;−44), p = 0.005], offset the higher incremental drug cost of PGB treatment [c353 (292; 415), P < 0.001] showing statistically reduction of health care costs; −c619 (−1,040;−197), p = 0.004. CONCLUSIONS: The use of Pregabalin resulted in both clinical and economic benefits in patients with severe refractory GAD and concomitant severe depressive disorder. Patients experienced a significant improvement in their anxiety and depression symptoms, improvements which were associated with significant decreases in health care costs. PMH29 COST-CONSEQUENCE ANALYSIS OF ARIPIPRAZOLE IN SCHIZOPHRENIA IN SPAIN: DIABETES AND CORONARY HEART DISEASE PROJECTIONS (STAR STUDY) Ramirez de Arellano A1, Rubio-Terrés C2, Baca E3, Riesgo Y4, Luque J1, Mena E5 1 Bristol-Myers Squibb Iberia, Madrid, Spain; 2HealthValue, Madrid, Spain; 3Hospital Universitario Puerta de Hierro, Madrid, Spain; 4Bristol-Myers Squibb Iberia, Malaga, Spain; 5 Otsuka Pharmaceuticals, Barcelona, Spain OBJECTIVES: Patients with schizophrenia experience elevated rates of morbidity and mortality, largely due to an increased incidence of cardiovascular disease and diabetes. The STAR study showed that the metabolic side effects of aripiprazole treatment are less than that experienced by those receiving standard-of-care (SOC) antipsychotic treatment (olanzapine, queatipine or risperidone). This study estimates the difference in direct and indirect cost-consequences of diabetes and coronary heart disease (CHD) in schizophrenia patients treated with aripiprazole or SOC. METHODS: On the basis of the results provided by the STAR study, risks for diabetes and CHD were projected over a time horizon of 10 years in the Spanish setting using the Stern and Framingham models. The comparators were aripiprazole versus SOC. The prevalence and costs (direct and indirect) of diabetes and CHD were obtained from Spanish sources. All costs were inflated to 2009 costs using the Spanish Health System pay and prices index. RESULTS: The number of avoided diabetes cases (23.4 cases per 1,000 treated patients) in patients treated with aripiprazole compared to SOC was associated with estimated total (direct and indirect) cost saving of c27,798,018 over 10 years for the Spanish population. Similarly, with aripiprazole the number of avoided CHD events (3.7 events per 1,000 treated patients) yields an estimated total cost saving of c4,173,818 over 10 years. CONCLUSIONS: Compared with SOC, aripiprazole 13th Euro Abstracts treatment can significantly reduce the number of diabetes and CHD events. As a consequence, it can provide considerable reductions in the health and economic burden to schizophrenia patients and health care services in the Spanish setting due to its favourable metabolic profile. PMH30 ECONOMIC EVALUATION OF AGOMELATINE IN MAJOR DEPRESSIVE DISORDERS IN IRELAND Alegre P1, Lacey L2, McAuliffe A3 1 Science Union et cie (Servier), Suresnes, France; 2Larry Lacey Consultant, Dublin, Ireland; 3 Servier Ireland, Dublin, Ireland OBJECTIVES: A cost-utility analysis of a new antidepressant, agomelatine (Valdoxan®) compared to generic fluoxetine and sertraline in the treatment of adult Major Depressive Disorders (MDD) was performed from the societal perspective and that of the Irish Health Service Executive (HSE). For each of these two perspectives, a separate analysis was performed for two different drug schemes: General Medical Services scheme (GMS) and the Drug Payments scheme (DP). METHODS: A Markov model was developed with health states for well, depressive episode, remission and death. The model also incorporated sleep disorders, discontinuation rates, discontinuation symptoms and adverse drug reactions. The time horizon of the analysis was two years. Remission, relapse and discontinuation rates as well as frequencies of the different clinical parameters were obtained from head-to-head comparative trials. Utility/disutility estimates were obtained from the literature. Costs (euros 2009) and effects were discounted at 4% per annum after year 1. RESULTS: From the societal perspective, agomelatine was dominant over both generic comparators in both drug schemes. From the HSE perspective, agomelatine was cost-effective compared to the two comparators in both drug schemes. One-way sensitivity analysis showed that the results were robust to uncertainty in model parameters. Probabilistic sensitivity analysis demonstrated that for a willingness-to-pay threshold of c45,000/QALY, agomelatine was costeffective compared to the two comparators in more than 93% of cases for both drug schemes CONCLUSIONS: From the societal perspective, agomelatine was dominant over generic fluoxetine and sertraline for the treatment of adult MDD in Ireland. From the Irish HSE perspective, agomelatine was cost-effective compared to the two comparators with high probability. PMH31 THE COST-EFFECTIVENESS OF ARIPIPRAZOLE IN PATIENTS WITH BIPOLAR I DISORDER IN THE UK Lebmeier M1, Dudley E1, Pericleous L1, Treur M2 1 Bristol-Myers Squibb Pharmaceuticals Ltd, Uxbridge, Middlesex, UK; 2Pharmerit BV, Rotterdam, Amsterdam, The Netherlands OBJECTIVES: To explore the cost-effectiveness of aripiprazole in UK patients with bipolar I disorder (BPD). METHODS: A Markov state transition model was developed to estimate the cost-effectiveness of aripiprazole, post-olanzapine, compared with risperidone or quetiapine, the most commonly used atypicals in the UK after olanzapine. Modelling was undertaken from a UK NHS perspective using discount rate of 3.5%. The time horizon of the model was 5 years (NICE BPD Guideline). The model incorporated an acute manic phase and relapse prevention phase. Response to therapy (at least a 50% reduction in Young-Mania Rating Scale) and discontinuation in the acute phase were based on an indirect comparison of trials in acute manic treatment. Probabilities of experiencing a new manic or depressive episode or discontinuing treatment in the relapse prevention phase were informed by a network meta-analysis. Quetiapine (at time of analysis) and risperidone are not indicated for relapse prevention in the UK; those patients were switched to lithium. Outpatient, hospitalization and drug costs were included based on published sources. a probabilistic sensitivity analysis (PSA) was used to examine uncertainty. RESULTS: Key drivers in the model are cost and quality-of-life reduction associated with acute mania. Aripiprazole is more effective than quetiapine in the acute phase and than lithium in the relapse prevention phase. Therefore, used post-olanzapine, aripiprazole gains 0.025 quality adjusted life-years (QALYs) and saves £3,995 compared with quetiapine and gains 0.01 QALYs and saves £607 compared with risperidone. Results from the PSA demonstrate 92% confidence that aripirazole is cost-effective versus quetiapine, and 61% confidence versus risperidone when using a threshold of £30,000 per QALY. CONCLUSIONS: Aripiprazole used post-olanzapine in the treatment of BPD is a cost-effective use of NHS resources compared with quetiapine and risperidone. PMH32 ECONOMIC EVALUATION OF AGOMELATINE IN MAJOR DEPRESIVE DISORDERS IN HUNGARY Alegre P1, Nagy B2, Nagy J2 1 Science Union et cie (Servier), Suresnes Cedex, France; 2GKI-EKI Healthcare Research Institute Ltd, Budapest, Hungary OBJECTIVES: A cost-effectiveness analysis of a new antidepressant, agomelatine (Valdoxan®) compared to duloxetine in the treatment of Major Depressive Disorders (MDD) in adults was performed in a Hungarian setting from a societal perspective and the National Health Insurance Fund (NHIF) perspective. METHODS: A Markov model was adapted with health states for well, depressive episode, remission and death. The model also incorporated sleep disorders, discontinuation rates, discontinuation symptoms and adverse drug reactions (incl. constipation, diarrhoea, dyspepsia, headache, nausea, somnolence, and sexual dysfunction). The time horizon was set to two years. An indirect comparison was run based on a meta-analysis of duloxetine A451 from a literature review to document clinical parameters unavailable from completed trials. a prospective cost of illness study in Hungary implemented in 2009 documented direct and indirect costs of MDD. Utilities for each health state and disutilities for each clinical event were taken from the literature. The discount rate was 5% per year RESULTS: From the societal perspective, agomelatine was cost-saving and more effective than duloxetine. From the NHIF perspective, agomelatine resulted in 0.037 QALYs gained compared to duloxetine with 5073 HUF as additional direct costs. One way sensitivity analyses showed that the results were robust to most parameter changes. From a societal perspective, agomelatine dominates duloxetine in 75% of cases and is cost-effective in 91% of cases at a willingness-to-pay threshold of 7 million HUF/QALY. From the NHIF perspective, agomelatine is cost-effective versus duloxetine in more than 90% of cases. CONCLUSIONS: In a Hungarian setting, agomelatine is dominant from a societal perspective and cost-effective from a NHIF perspective versus duloxetine. These results are robust, confirmed by sensitivity analyses. PMH33 ADJUNCTIVE ANTIPSYCHOTICS IN PATIENTS WITH MAJOR DEPRESSIVE DISORDER IN TURKEY: A HEALTH ECONOMIC PERSPECTIVE Isik E1, Dilbaz N2, Savas H3, Gönül AS4, Saylan M5, Postema R6, Loze JY7, Drost P8, Treur M6 1 Gazi University Medical Faculty, Ankara, Turkey; 2Ankara Numune Research & Training Hospital, Ankara, Turkey; 3Gazientep University Medical Faculty, Gazientep, Turkey; 4Ege University Medical Faculty, Izmir, Turkey; 5Bristol-Myers Squibb, Istanbul, Turkey; 6Pharmerit Europe, Rotterdam, The Netherlands; 7Otsuka Pharmaceutical Co., Ltd., Rueil-Malmaison, France; 8Bristol-Myers Squibb, Braine-l’Alleud, Belgium OBJECTIVES: Major Depressive Disorder (MDD) is a chronic illness associated with major burden on Quality-of -Life (QoL) and health care resources. a recent study estimated annual cost being c282 per European inhabitant. Adjunctive treatments with aripiprazole, quetiapine and olanzapine have demonstrated efficacy in patients with MDD that respond insufficiently to antidepressant treatment. The objective is to estimate the cost-effectiveness of adjunctive therapies in depressive patients failing to respond to antidepressant therapy in Turkey. METHODS: An economic model was built simulating MDD patients between major depressive episodes (MDEs) and remission over lifetime. During MDEs, patients were treated with adjunctive aripiprazole, quetiapine or olanzapine. Patients who did not respond at 6 weeks switched to subsequent treatment lines. Comparative effectiveness between aripiprazole, quetiapine and olanzapine, was imputed using an indirect comparison combining 6-week published studies. Resource use data and unit costs were obtained from Turkish studies. RESULTS: Over life-time, aripiprazole patients spent less time in MDEs compared to quetiapine (7,7 weeks) and olanzapine (7.5 weeks). Compared to Quetiapine, patients with aripiprazole showed improvement in QoL (+0.045 QALY) at incremental direct cost of 421TL. Compared to olanzapine, patients with aripiprazole dominated, meaning improvement of QoL (+0.042 QALY) respectively, at lower direct cost (-32 TL), despite higher drug costs. Sensitivity analyses estimated a 87% likelihood that aripiprazole improved QoL at a comparable cost verus quetiapine and 88% versus olanzapine. CONCLUSIONS: This is the first lifetime health-economic model in Turkey taking patient heterogeneity into account when assessing QoL and costs of different adjunctive strategies in MDD. These results indicate that adjunctive treatment with aripiprazole provides health benefits at lower costs compared to quetiapine and olanzapine, in patients with MDD. PMH34 COST-EFFECTIVENESS ANALYSIS OF RISPERIDONE LONG-ACTING INJECTION IN SCHIZOPHRENIA: 24-MONTH DATA FROM CZECH REPUBLIC Skoupá J Pharma Projects s.r.o., Prague, Czech Republic OBJECTIVES: To evaluate 24 months cost-effectiveness of risperidone long-acting injection (RLAI) in patients with schizophrenia enrolled in the electronic-Schizophrenia Treatment Adherence Registry (e-STAR) from Czech Republic. This evaluation is a follow-up of 12 months data presented at the European ISPOR congress (Athens 2008). METHODS: e-STAR is an international 24-month, prospective, observational study in patients with schizophrenia who commenced RLAI treatment (based on SmPC indication). Analyzed data covered one year of retrospective (prior to RLAI initiation) vs. 24 months of prospective observation. a total of 607 patients have completed the 24 months study. Assessed direct costs were: hospitalization (duration and frequency), antipsychotic medication and co-medication all from the payer’s perspective in 2009 prices. Efficacy parameters included GAF (Global Assessment of Functioning) and CGI-S (Clinical Global Impression-Severity) scores. RESULTS: Mean annual costs/ patient increased form c2173 (1 Euro = 26 CZK) in the retrospective period to c4270 during the first and 4453 during the second years of observation. Mean cost drivers were hospitalization (63% of total retrospective costs) and antipsychotic medication (90% and 91% of total prospective costs). Costs of RLAI could not be fully offset by significant reductions in hospitalization (both frequency and duration) and co-medication. Improvements in GAF and CGI scores seen at the first treatment year were maintained during year 2 of assessment. Overall GAF changed from 49.6 (baseline) to 74.6 (at 24 month), CGI decreased from baseline 4.61 to 2.96 (24-month). Costeffectiveness (costs/clinically relevant change in parameters) could also be maintained. More than one-third of patients achieved disease remission, resulting in acceptable incremental c850/patient in remission. CONCLUSIONS: Switching to risperidone long-acting injection in patients with treatment failure, non-compliance or intolerance A452 of current antipsychotic medication is cost-effective despite higher costs of antipsychotic medication. Positive clinical and economic results persisted over 24 months. PMH35 COST-EFFECTIVENESS OF MEMANTINE IN THE TREATMENT OF MODERATE AND SEVERE ALZHEIMER’S DISEASE PATIENTS WITH AGITATION, AGGRESSION AND PSYCHOSIS—THE UK EXAMPLE Rive B1, Grishchenko M1, Guilhaume C1, Katona C2, Lamure M3, Livingston G2, Toumi M4, Francois C1 1 Lundbeck SAS, Issy-Les Moulineaux, France; 2University College London, London, UK; 3 University Claude Bernard Lyon 1, Paris, France; 4University Claude Bernard Lyon 1, Lyon, France OBJECTIVES: To assess the cost-effectiveness of memantine in moderate and severe AD patients who exhibit agitation/aggression and psychotic symptoms (APS) from the UK National Health Service and Personal Social Services perspective. METHODS: The cost-utility analysis was based on 5-year Markov cohort simulations. The model evaluated the impact of memantine on time to Full-Time-Care (FTC), QualityAdjusted-Life-Years (QALYs) and costs, in pre-FTC patients compared with standard care, i.e. no pharmacotherapy or background treatment with acetylcholinesterase inhibitors. FTC was defined based on locus of care and patient’s physical and functional dependency status. Transition probabilities, baseline characteristics, resource utilization volumes, health utility weights and mortality rates were derived from the 4.5-year London and South-East Region (LASER-AD) epidemiological study. Effectiveness estimates came from a meta-analysis of six large randomised clinical trials. Costs covered routine patient management, hospitalization, social community services, institutionalization, and medications. Results were reported in EUR (GBP), 2009. The model underwent extensive stochastic and one-way sensitivity analyses, testing the model assumptions and changes in input parameters. RESULTS: Over five years, patients receiving standard care spend on average 78.8 weeks in the pre-FTC state. Overall costs in this group were c117,960 (£98,810). QALYs were estimated at 1.49 (30% of full health). Memantine was associated with a longer time-to-FTC of 11.2 weeks, QALY gains of 0.07 and cost-savings of c5930 (£4970). Lower costs in the memantine group were due to prolonged pre-FTC period. Memantine was more effective and less costly strategy relative to standard care in 99.98% of simulations. The estimated benefits and cost savings were almost twice higher than those previously estimated in all moderate and severe AD patients, largely due to enhanced efficacy of memantine in APS patients, who, when left untreated, rapidly deteriorate. CONCLUSIONS: The model showed that memantine yielded higher benefits at no additional costs relative to its alternative. PMH36 ECONOMIC EVALUATION ANALYSIS IN THE TREATMENT OF BIPOLAR DISORDER WITH ARIPIPRAZOLE AND OTHER ATYPICAL ANTIPSYCHOTIC DRUGS IN SPAIN Ramirez de Arellano A1, Rubio-Terrés C2, Baca E3, Riesgo Y4, Luque J1, Mena E5 1 Bristol-Myers Squibb Iberia, Madrid, Spain; 2HealthValue, Madrid, Spain; 3Hospital Universitario Puerta de Hierro, Madrid, Spain; 4Bristol-Myers Squibb Iberia, Malaga, Spain; 5 Otsuka Pharmaceuticals, Barcelona, Spain OBJECTIVES: This analysis investigates the relative efficiency in the treatment of BD with the use atypical antipsychotics (AA): aripiprazole (ARI); olanzapine (OLA); quetiapine (QUE); risperidone (RIS); and ziprasidone (ZIP). METHODS: our analysis takes into consideration the treatment cost of AA and the impact on hospitalization costs associated with each AA. Mean daily dose of each AA, length of treatment and the probability of hospitalization for each AA were obtained from a retrospective study (90 days of follow-up) in 6,162 BD patients (Jing et al. 2009). The minimum acquisition cost per mg. of the mean daily dose, for each AA, is regarded as an efficiency criterion in Hospital Pharmacy Departments. The cost per day of hospitalization (c347.90) and the length of hospitalization associated to BD (18.1 days; 14.5–21.7) were obtained from a Spanish study. RESULTS: after 90 days of follow-up, hospitalization rates were higher with OLA (8.7%), QUE (8.5%), RIS (8.6%) and ZIP (10.2%) in comparison with ARI (5.9%; 5.8%; 5.7%; 6.5%, respectively). The treatment of BD with ARI gave rise to the following cost savings per patient, in relation to other AA: c149.31 ARI versus OLA; c33.42 ARI versus QUE; c19.45 ARI versus RIS; and c242.22 ARI versus ZIP. a sensitivity analysis tested the following variables: minimum daily dose for each AA; maximum daily dose for each AA; length of treatment with AA; and probability of hospitalization for each AA. The sensitivity analysis confirms the cost savings associated with aripiprazole, with the only exception of risperidone where the cost saving per patient is almost neutral (c−2.89). CONCLUSIONS: using the criterion of cost rationalization based upon the minimum acquisition cost per mg. aripiprazole may have economic benefits over other AA in terms of lower psychiatric treatment costs and lower total health care costs in the Spanish NHS. PMH37 COST-EFFECTIVENESS OF ANTIPSYCHOTICS FOR THE TREATMENT OF RELAPSE PREVENTION FOR SCHIZOPHRENIA: THE SPANISH PERSPECTIVE García-Ruiz A1, Perez Costillas L2, Montesinos Galvez A2, Alcalde de Hoyos J1, Garcia-Agua i Soler N1 1 Malaga University, Malaga, Spain; 2Carlos Haya Hospital, Malaga, Spain OBJECTIVES: To assess the efficiency of the drugs used to reduce relapses in schizophrenia, taking into account costs and effectiveness (measured as QALY). METHODS: 13th Euro Abstracts The National Health Care System perspective and a 1 year temporal horizon have been used. Taking into account the last NICE review on schizophrenia1, four fundamental aspects related with schizophrenia management were analyzed: relapse rates, treatment discontinuation due to intolerable side effects, treatment discontinuation due to other reasons and stable patients (where the probability of remission, EPS syndrome, weight gain, glucose intolerance and diabetes were analyzed). The health care direct costs corresponding to the drug acquisition costs have been analyzed together with the costs of the side effects considered, the cost of the complications (diabetes) and the costs of hospitalary relapses (stay and drug cost) updated with data from Spanish Ministry of Health (2008). a Monte Carlo simulation was performed to obtain the cost-effectiveness ratio (euros/QALY). RESULTS: Paliperidone ER presents the lower total costs (c3060) compared to the other strategies (risperidone = c3206; haloperidol = c3220; olanzapine = c3903; amisulpride = c4281; aripiprazole = c4719). Paliperidone ER presents the higher efficacy (QALY) compared to the other strategies (Paliperidone ER = 0.7573; risperidone = 0.7335; haloperidol = c0.7230; olanzapine = c0.7474; amisulpride = c0.7320; aripiprazole = c0.7379). With these results, we can conclude that all the strategies are dominated by Paliperidone ER with a C/E ratio of 4073 (risperidone = 4382; haloperidol = 4461; olanzapine = 5.235; amisulpride = 5.827; aripiprazole = 6.421) CONCLUSIONS: At a willingness-to-pay of c30,000 per QALY all the drugs considered are cost-effective in Spain. However, the most efficient (more net benefic, monetary (euros) and in health—QALYs) vs. Haloperidol are: paliperidone ER, followed by risperidone, olanzapine, amisulpride and aripiprazole. PMH38 A COST UTILITY ANALYSIS OF FIRST LINE ANTIPSYCHOTICS FOR THE PREVENTION OF SCHIZOPHRENIA RELAPSE IN THE UNITED KINGDOM Jones MA, Meier GC AstraZeneca UK Ltd, Luton, England OBJECTIVES: To evaluate the cost-effectiveness of eight first line antipsychotics in the prevention of schizophrenia relapse, from a UK National Health Service and Personal Social Services perspective. METHODS: A Markov model, similar to that used by the National Institute for Health and Clinical Excellence in the Schizophrenia (update) guideline (published March 2009), was developed to assess the cost per QALY gained for amisulpride, aripiprazole, haloperidol, olanzapine, paliperidone, quetiapine (XL), risperidone and zotepine. Clinical parameters were populated with data from mixed treatment comparisons. Patients entered the model in remission and could remain in remission, relapse, move to next line of therapy due to side effects, or discontinue for other reasons. Utilities for schizophrenia in remission and relapse were taken from a direct utility elicitation study conducted in a UK population. Disutilities for side effects were captured. Resource use and unit costs (reference year 2009) were taken from published sources and a 10 year time horizon was adopted. Probabilistic results were derived from 10,000 model simulations. RESULTS: The deterministic analysis showed quetiapine (XL) to yield the most QALYs and lowest overall treatment costs. Model outcomes were supported by the probabilistic analysis but results were characterised by uncertainty. In one scenario, drug costs for all comparators bar quetiapine (XL) were set to zero. Quetiapine (XL) generated 7.051 QALYs compared to a range of 6.930–6.972 for other medications. Overall treatment costs for quetiapine (XL) were £153,104 compared to £155,766-£167,329 for other medications. Thus it was still the dominant strategy. CONCLUSIONS: The additional benefit of Quetiapine (XL) in terms of schizophrenia relapse prevention results in it being the most cost-effective of the antipsychotics assessed. Clinical efficacy is the key driver of cost-effectiveness in relapse prevention, hence generic antipsychotics should not be recommended based upon drug costs alone. Further long term trials of antipsychotics are required to reduce uncertainty. PMH39 THE COST-UTILITY OF AGOMELATINE IN MAJOR DEPRESSIVE DISORDER IN POLAND Golicki D, Paja˛k K, Dbrowska A, Niewada M HealthQuest sp z o.o., Warsaw, Poland OBJECTIVES: a cost-utility analysis of agomelatine, a treatment of major depressive disorders in adults, was performed from the Polish public payer’s perspective (the National Health Fund). Agomalatine was compared with the most commonly used antidepressants in Poland, i.e. generic sertraline and generic venlafaxine. METHODS: The analysis with a time horizon of 2 years is based on a Markov model. The onemonth cycle model included the following health states: depression episode, remission, well and death. It also incorporated sleep disorders, discontinuation rates, discontinuation symptoms and adverse drug reactions. The clinical parameters for compared drugs were extracted from head-to-head clinical trials. Utility and disutility estimates were derived from a systematic literature review. Only direct costs have been considered in order to be consistent with Polish pharmacoeconomic guidelines. Costs and effects were discounted at 5% and 3.5% per annum at year 1 respectively. RESULTS: The benefit of agomelatine over sertraline or venlafaxine was estimated at 0.005 QALY. This effectiveness was associated with the additional costs of 491 PLN and 149 PLN for agomelatine compared to sertraline and venlafaxine, respectively. The corresponding agomelatine incremental cost-utility ratios (ICURs) were therefore 92,000 PLN/QALY and 28,000 PLN/QALY. Sensitivity analysis demonstrated that agomelatine ICUR remained below the three times GDP per capita threshold (33,347 PLN for 2008) in 79.7% and 88.4% of the cases compared with sertraline and venlafaxine, respectively. CONCLUSIONS: From the Polish public payer’s perspective, 13th Euro Abstracts agomelatine is a cost-effective treatment of major depressive disorders compared to generics of sertraline and venlafaxine. PMH40 COST-UTILITY ANALYSIS (CUA) OF SERTINDOLE IN THE TREATMENT OF SCHIZOPHRENIA IN POLAND IN GENERAL POPULATION OF SCHIZOPHRENIC PATIENTS AND POPULATION OF PATIENTS INTOLERANT TO AT LEAST ONE OTHER ANTIPSYCHOTIC AGENT: A FIVE-YEAR MARKOV MODEL Walczak J, Stelmachowski J, Obrzut G, Nogas G Arcana Institute, Cracow, Poland OBJECTIVES: To evaluate a cost-utility of sertindole compared with commonly used antipsychotic drugs in Poland: haloperidol and risperidone in the treatment of general population of schizophrenic patients and with risperidone in a group of patients intolerant to at least one other antipsychotic agent. METHODS: Cost-utility Markov model was constructed in five-year time horizon. Clinical effectiveness was assessed on the basis of RCTs, meta-analyses, observational studies and systematic reviews. The measures of effectiveness in the CUA were quality adjusted life-years (QALY) and lifeyears gained (LYG). Direct medical costs were calculated from perspective of two payers (Polish National Health Fund and patient). Cost of: oral antipsychotic treatment, inpatient care, primary health care, specialist care and adverse events treatment (extrapyramidal symptoms, sedation and weight gain) were included. Discount rates of 5% for costs and 3.5% for benefits were used. Utilities and disutilities associated with AEs and relapse were derived from published literature. RESULTS: In general population of schizophrenic patients cost per QALY/LYG gained, using sertindole instead of risperidone is 97,396/26,585 PLN (c22,507/6,144) respectively; cost per QALY/LYG gained, using sertindole instead of haloperidol, is 82,132/20,670 PLN (c18,980/4,777) respectively. In population intolerant to at least one other antipsychotic agent sertindole dominates risperidone (sertindole is less expensive and more effective considering both QALY/LYG gained). The sensitivity analysis indicated the robustness of the results. Obtained results are placed below the acceptability threshold (3xGDP per capita in Poland) which is about 105,500 PLN (c24,380). The 2009 weighted average exchange rate of Polish National Bank was c1 = PLN 4.3273. CONCLUSIONS: Treatment of schizophrenia with sertindole is a cost-effective strategy in comparison with risperidone and haloperidol therapy. In a group of patients intolerant to at least one other antipsychotic agent sertindole compared to risperidone is dominant strategy. PMH41 COST-UTILITY ANALYSIS OF ESCITALOPRAM VERSUS CITALOPRAM IN MAJOR DEPRESSIVE DISORDER IN ISRAEL Leshno M1, Ben-Amnon Y2, Brignone M3, Marteau F4, Hansen K3 1 Faculty of Management, Tel-Aviv, Israel; 2Lundbeck Israel Ltd., Petach-Tikva, Israel; 3 Lundbeck SAS, Issy-les-Moulineaux cedex, France; 4Lundbeck SAS, Issy-les-Moulineaux cedex, France OBJECTIVES: A recent meta-analysis has shown that escitalopram, a selective inhibitor of serotonin reuptake, is one of the best choices when starting a treatment of moderate to severe major depressive disorder (MDD), with a favourable balance between efficacy and tolerability. Our objective was to perform an economic evaluation of escitalopram versus citalopram in Israel. Citalopram is available as a generic drug and is included in the National Health Basket. METHODS: A decision tree model was used to assess the cost-utility of escitalopram versus citalopram in first-line treatment of MDD. The model also accounted for second-, third- and fourth-line treatments for patients not responding to therapy, with an overall time horizon of 12 months. The analysis adopted a Health Maintenance Organization perspective. The main outcomes were quality-adjusted life-years (QALYs) and costs associated with local health care resource use (including drug use, and hospitalizations due to depression or suicide attempt). One-way and two-way sensitivity analyses were performed with the following parameters: probabilities of remission after initial and subsequent treatment, probabilities of relapse, suicide attempt and suicide-related death, number of hospitalization days due to a suicide attempt, costs of initial and subsequent treatments and utility values. In addition, we used a Monte Carlo simulation. RESULTS: Compared to citalopram, escitalopram was more effective in terms of QALYs at a small incremental cost. The average QALYs of patients treated with escitalopram and citalopram were 0.728 and 0.720 respectively. The average cost of patients treated with escitalopram and citalopram were c105 and c63 respectively. The incremental cost-utility ratio was c5,150 (NIS23,875) per QALY. The variation of remission rates had the most influence on this ratio in the sensitivity analyses. CONCLUSIONS: This local economic evaluation shows that compared to citalopram, escitalopram should be the preferred option in depression in Israel with an acceptable cost per QALY. PMH42 UTILIZATION OF ANTIDEPRESSANTS IN SLOVAKIA 1 1 2 1 3 Bellová K , Gatialová K , Foltanova T , Petrová L , Bielik J 1 Comenius University, Bratislava, Slovak Republic; 2Faculty of Pharmacy, Bratislava, Slovak Republic; 3AD University, Tren ín, Slovak Republic OBJECTIVES: The main objective was to evaluate the consumption of antidepressants, it s utilization as the basic aspect of the quality of life and to analyze the using of the original drugs or generics and to show the pharmacoeconomical impact of this group of drugs considering to it s financial costingness. METHODS: We used the analysis of consumption trend of antidepressants and the analysis of the trend of prices of antidepressants. We compared the original and generic market with antidepressants A453 in Slovakia, we monitored and compared the prices for DDD and we evaluate the impact of new generics on the prices of originals as well. RESULTS: The consumption of antidepressants has increased because of increasing number depression illnesses which could be continually treated by modern antidepressants since successful treatment depends on relaps prevention. Such outcome could be reached by usage of the latest antidepressants with good compliance (escitalopram, venlafaxine, mirtazapin). Since 2005 the consumption (packages sold) of antidepressant increased by 29% whereas the sales value stagnated as result of declining price. The price for daily recommended dosage decreased by 52.7% (from c0.80 in 2001 to c0.38 in 2008). This was influenced mainly by the increasing number of generics medicaments which pushed the price down. In 2009 the price of original antidepressants fall by 26.4% as reason of generic molecules. Based on long-term surveys the similar trend could be expected also in the future. CONCLUSIONS: The utilization of antidepressants lead to better quality of life. From the pharmacoeconomical view, the development of new molecules goes hand in hand with the decrease of invalidation, decrease of the cost for hospitalization, as well as increase people ‘s ability to take care about themselves. On the other hand, the generic substitution leads to increase of the amount of treated patients. PMH43 HEALTH CARE UTILIZATION OF PATIENTS WITH DEPRESSION BEFORE AND AFTER INTIATING DULOXETINE Shi N1, Durden E2, Cao Z1, Torres A3, Happich M4 1 Thomson Reuters, Cambridge, MA, USA; 2Former Employee of Thomson Reuters, currently with Covance Market Access Services Inc., San Diego, CA, USA; 3Thomson Reuters, Washington, DC, USA; 4Eli Lilly & Company, Bad Homburg, Hessia, Germany OBJECTIVES: Evaluate health care utilization among depressed patients before and after initiating duloxetine. METHODS: Depressed adults treated with duloxetine during January 1, 2006-September 30, 2007 were identified from the General Practice Research Database. All-cause health care utilization in the 12-months before and after treatment initiation was compared for patients and for the subset of patients with pre-period unexplained pain (UPain). Generalized Estimating Equation models were used to identify factors associated with the pre-post differences in hospitalization rates. Covariates included age, gender, pre-period comorbid conditions and medication use. RESULTS: A total of 909 patients were identified (mean age = 49.6; female = 67.7%), and 45% experienced UPain. Three-quarters of patients were prescribed a SSRI and 11% were untreated with antidepressants before initiating duloxetine. Among patients, the rate of hospitalization was higher in the pre-period compared to the post-period (20.7% vs.16.5%, p = 0.006). There were no significant pre-post differences in the rates of accident & emergency visits, specialist referrals, or analgesic use. Among patients with UPain, fewer were hospitalized (26.2% vs. 19.1%, p = 0.003) or received analgesic in the post-period (77.0% vs. 71.7%, p = 0.015). Multivariate analysis confirmed fewer patients experienced hospitalization after duloxetine initiation. All depressed patients with pre-period anxiolytic use (OR = 0.70, 95% CI, 0.50–0.97, p = 0.033) or UPain patients with pre-period anticonvulsant use (OR = 0.44, 95% CI, 0.21–0.91, p = 0.026) experienced a decreased hospitalization rate after initiating duloxetine and the magnitude of change was larger than that for patients not receiving these medications. All depressed patients with pre-period alcohol/drug dependence (OR = 1.83, 95% CI, 1.01–3.30, p = 0.045) and sleep disorder (OR = 1.65, 95% CI, 1.05–2.59, p = 0.031) had an increase in the hospitalization rate from the pre- to post-period, while those without the conditions had a decreased hospitalization rate. CONCLUSIONS: Hospitalization rates were lower following duloxetine initiation among depressed patients. Pre-period alcohol/drug dependence, sleep disorders, anxiolytic and anticonvulsant use were associated with the pre-post difference in the hospitalization rate. PMH44 AN OBSERVATIONAL STUDY OF THE BURDEN OF INFORMAL CAREGIVERS TO OUT-PATIENTS WITH SCHIZOPHRENIA SPECTRUM DISORDERS Flyckt L1, Granström O2, Löthman A3, Jörgensen L2, Koernig T2 1 Norra Stockholms Psykiatri, Danderyd, Sweden; 2AstraZeneca Nordic MC, Södertälje, Sweden; 3Psykiatricentrum Väst, Bromma, Sweden OBJECTIVES: The lifetime emotional, social, and financial consequences experienced by individuals with schizophrenia spectrum disorders have significant effects on their families. Therefore, it is essential to increase the knowledge and awareness of the informal caregiver burden. Here, the purpose was to study objective burden (time and money spent), as well as subjective burden (care-related quality of life) of informal caregivers to patients with schizophrenia spectrum disorders. METHODS: Patients with schizophrenia spectrum disorders and their closest informal caregivers were recruited and patient- and caregiver-related characteristics and variables related to the provision of formal care were assessed. The subjective burden was assessed using the CarerQol. The objective burden was assessed prospectively during four weeks with daily recordings by the caregiver using diaries of the money and time spent on the patient. Data was analyzed using descriptive statistics. RESULTS: One-hundred and seven patients (53% females; mean age 43 ± 11 years) from nine centers in Sweden, and 118 of their closest informal care-givers (67% females; mean age 58 ± 15 years) entered the study. The mean (standard deviation, SD) GAF for the patients was 52 (11). Subjective burden (CarerQol VAS) correlated positively to GAF. The mean (SD) time spent on caring for relatives with schizophrenia spectrum disorders was 22 (36) hours per week. Half of this (11 hours) was “stand-by” time and another 5 hours were spent on household work. Caregivers’ mean (SD) monthly expenses in support A454 of their relatives with schizophrenia were c292 (424), which corresponds to approximately 14% of the mean gross income of caregivers. Large expenditure items were groceries, clothes, rent and travel. CONCLUSIONS: The objective burden adds halftime to an ordinary fulltime job (time spent on care) and additionally reduces gross incomes with approximately 14% (money spent on care) for informal caregivers to patients with schizophrenia spectrum disorders. MENTAL HEALTH – Patient-Reported Outcomes Studies PMH45 DISCONTINUATION OF DIFFERENT ANTIPSYCHOTIC MEDICATIONS AFTER DISCHARGE FROM ACUTE CARE FOR PERSONS WITH SCHIZOPHRENIA Boaz TL1, Becker M1, Andel R1, Van Dorn R1, Choi J2, Sikirica M3 1 University of South Florida, Tampa, FL, USA; 2Ortho-McNeil Janssen Scientific Affairs, LLC, Titusville, NJ, USA; 3Ortho-McNeil Janssen Scientific Affairs, LLC, Bethesda, MD, USA OBJECTIVES: To examine risk of discontinuation of antipsychotics following discharge from acute care settings in Florida for Medicaid-enrolled persons with schizophrenia. METHODS: Medicaid claims, enrollment data, and community services data from the state mental health authority from 2004–2008 were used. Demographics, diagnosis, service history during the year before admission to acute care (hospitals and crisis units), and post-discharge psychopharmacologic treatment were obtained. Persons who received antipsychotic monotherapy and had a discharge diagnosis of schizophrenia were included. Cox proportional hazards regression estimated risk of post-discharge discontinuation of antipsychotics (formulary restrictions were not considered). Second-generation antipsychotics (SGAs) were examined individually; oral first-generation antipsychotics (FGAs) and depot FGAs were each considered as a group. Demographics, pre-admission service history, length of stay, and pre-discharge length of time receiving medication were control variables. Data were right-censored at end of Medicaid enrollment, end of data availability, or death. RESULTS: Participants (N = 6365) were 42.7 ± 11.2 (mean ± SD) years old; 61% male; 34% white. Mean ± SD length of acute care stay was 11.0 ± 19.7 days. Median length of time receiving medication post-discharge was 45 days. With risperidone long-acting therapy (RLAT) as reference, depot and oral FGAs had highest risk of discontinuation (hazard ratio [HR] = 2.49, P < .0001 and HR = 1.68, P = .0005), respectively. Risk was also elevated for oral SGAs aripiprazole (HR = 1.65, P = .0012), ziprasidone (HR = 1.46, P = .0191), olanzapine (HR = 1.43, P = .0202), and quetiapine (HR = 1.35, P = .0434), except for oral risperidone and clozapine. Fewer months receiving pre-discharge medication (HR = 1.04, P < .0001), black race (HR = 1.39, P < .0001), history of involuntary commitment (HR = 1.28, P < .0001), and more pre-admission acute care services (HR = 1.13, P = .0053) were associated with higher risk of discontinuation. CONCLUSIONS: Persons with schizophrenia taking oral and depot FGAs were at highest risk for discontinuation of medication after discharge from acute care. Those taking oral SGAs (except risperidone and clozapine) also may have higher risk of discontinuation than patients taking RLAT. PMH46 ADHERENCE RATES OF INJECTABLE VS. ORAL RISPERIDONE FOR PATIENTS WITH SCHIZOPHRENIA Rascati KL, Kim DN The University of Texas at Austin, Austin, TX, USA OBJECTIVES: Non-adherence to antipsychotic therapy can lead to increased risks of relapse and rehospitalization. Long-acting injectable antipsychotics have been shown to be effective and well-tolerated, with increasing evidence of a potential benefit on adherence. At the time of this study, only one long-acting injectable atypical antipsychotic entity—risperidone—was approved for patients with a diagnosis of schizophrenia. The objective of this study was to determine the medication adherence rates of Medicaid patients treated with either 1) oral risperidone only, or 2) long-acting injectable risperidone +/− oral risperidone. METHODS: Texas Medicaid prescription claims data from January 2006 to December 2008 were analyzed for patients with schizophrenia. Records were assessed for a period of 18 months (6-month pre-index period with no antipsychotics plus 12-month post-index period). a medication possession (MPR) ratio of ≥80% was considered adherent. RESULTS: A total of 1313 patients met inclusion criteria, and less than half, 48% (N = 626) had an MPR of at least 80%. Of the patients on long-acting injectables +/− oral risperidone, 60% (N = 96/160) were adherent with their regimen versus 46% (530/1153) for patients taking only oral risperidone. a logistic regression analysis adjusting for demographic covariates (gender, race, age) showed that patients on injectable +/− oral medications were about 80 percent more likely to be adherent (OR = 1.82, 95% CI = 1.29–2.55; p = 0.0006) than patients on oral medication alone. There were no statistically significant differences in any of the demographic characteristics between the two groups. CONCLUSIONS: Overall, about half of the patients did not meet the definition for medication adherence (MPR ≥ 80%) one-year after their index date. Long-acting injectable risperidone had a higher rate of adherence compared to oral risperidone alone. With the release of newer atypical long-acting injections (paliperidone and olanzapine), future studies comparing adherence rates of these antipsychotic agents are needed. 13th Euro Abstracts PMH47 MEDICATION COMPLIANCE AND HEALTH CARE COSTS OF TYPE II DIABETICS WITH SCHIZOPHRENIA NEWLY STARTING HYPOGLYCEMIC THERAPY Cheng JS, Shih YT National Health Research Institutes, Zhunan Town, Miaoli County, Taiwan OBJECTIVES: Schizophrenia has been found to be associated with poor medication compliance, and a higher prevalence of diabetes. Once a hypoglycemic therapy is started, appropriate compliance to the therapy is crucial for good glycemic control. This study aimed to compare hypoglycemic therapy medication compliance, and health care costs after the initiation of the therapy, between type II diabetics with schizophrenia versus those without schizophrenia. METHODS: This study used the claims database of the National Health Insurance program. Enrollees who began oral hypoglycemic therapy in 2001–2002 (the index date), and had been diagnosed with schizophrenia and refilled at least one prescription of antipsychotic(s) in the year prior to the index date were included in the study (the case group). Enrollees without schizophrenia who began oral hypoglycemic therapy in 2001–2002 were selected from a randomly selected sample of the enrollees (the comparison group). Good medication compliance was defined as an medication possession ratio (MPR) ≥ 0.8. Ordered logistic models were adopted to assess associations between factors with medication compliance. RESULTS: There were 1196 and 3188 subjects in the case group and comparison group, respectively. Forty-one percent of the case group and 30% of the comparison group were good compliant. The cost of hypoglycemic therapy was significantly higher in the case group, while the total diabetes-related cost was not significantly different between the two groups. The regression results indicated that comorbid schizophrenia was correlated with a higher likelihood of good compliance. In a subgroup analysis of diabetics with schizophrenia, good compliance to antipsychotics in the previous year was significantly associated with a greater probability of being better compliant with hypoglycemic therapy. CONCLUSIONS: Diabetics with schizophrenia, compared with those without such a condition, were more compliant in their hypoglycemic therapy, and had higher cost of hypoglycemic therapy in the first year of the therapy. PMH48 IMPACT OF ONCE-DAILY DOSING ON DRUG DISCONTINUATION OF PATIENTS WITH SCHIZOPHRENIA Kalo Z1, Nemeth A2, Jozwiak-Hagymasy J3 1 Eötvös Loránd University, Budapest, Hungary; 2National Centre for Psychiatry, Budapest, Hungary; 3Syreon Research Institute, Budapest, Hungary OBJECTIVES: According to evidence from scientific literature the ease of administration, such as once-daily dosing, improves the adherence and persistance of patients with the pharmaceutical therapy. Adherence and persistance are key success factors in the treatment of schizophrenic patients, as drug discontinuation or non-adherence increase the risk of relapse, which eventually results in reduced quality adjusted lifeyears and increased cost of care by more acute hospitalization episodes. Our objective was to analyze the impact of once-daily dosing vs. twice daily dosing on the discontinuation rate of pharmaceutical therapy among patients with schizophrenia. METHODS: Our real world analysis is based on aggregated data from the Hungarian National Health Insurance Fund database. We selected adult patients (between 18–65 years) on quetiapine therapy with F20 schizoprenia ICD-10 code between January and March 2008. We compared the one-year discontinuation rate among once-daily vs. twice daily quetiapine users. Discontinuation was defined as no prescribed antipsychotic medication in the N05A ATC group between May—July 2009. RESULTS: After 1 year 23.4% in the once-daily group compared to 27.2% in the twice-daily group discontinued the pharmaceutical therapy. Once-daily dosing reduced the relative risk of discontinuation by 13.9%. (χ2 test: p = 0,017). The absoluate risk reduction was 3.8% (NNT = 26.5). CONCLUSIONS: Our analysis indicates that once-daily dosing reduces the risk of drug discontinuation among schizophrenic patients. Short time-horizon, potential selection bias and confounding factors may limit the generalizability of our conclusions. Long-term observational study with multiple regression analysis based on anonimized individual patient records can confirm the validity of our findings. PMH49 ANTIDEPRESSANT UTILIZATION, ADHERENCE AND HEALTH CARE SPENDING IN THE UNITED STATES: THE CASE OF MDD PATIENTS 2000–2007 Lin HC1, Erickson S2, Smith D1, Balkrishnan R2 1 University of Michigan, Ann Arbor, MI, USA; 2University of Michigan, College of Pharmacy, Ann Arbor, MI, USA OBJECTIVES: Innovative antidepressants such as SSRIs and SNRIs have been widely adopted. However, the differences in patient antidepressant adherence and associated health care spending across patient factors and antidepressant choice needed further research. This study was trying to understand how patient factors and antidepressant choice influenced medication adherence and associated health care expenditure. METHODS: A retrospective cross-sectional study was conducted using the 2000– 2007 Medical Expenditure Panel Survey (MEPS) database. a multiple OLS regression was used to examine MDD patient’s antidepressant adherence measured by proportional days covered (PDC). a two-part model was implemented to study the impact of MDD patient factors and antidepressant choice on associated health care expenditure. RESULTS: Linear regression models indicated that patient gender, ethnicity and 13th Euro Abstracts health insurance status were associated with differential levels of antidepressant adherence and associated health expenditure. Hispanic ethnicity was associated with decreased antidepressant adherence compared to non-Hispanic white (β = 12.53, P < 0.05) and other ethnicities (β = 28.27, P < 0.01). Patient who were covered by public insurance had better PDC compared to uninsured patients (β = 16.23, P < 0.05). Patient who were covered by private insurance spent more on MDD-specific drug compared to uninsured patients (β = 0.36, P < 0.05). Higher antidepressant adherence was associated with higher MDD-specific drug expenditure (β = 0.03, P < 0.01). Use of innovative antidepressants such as SSRIs and SNRIs was associated with an increase in MDD-specific drug expenditure. CONCLUSIONS: Differences in antidepressant adherence and health care spending across patient factors could have important policy implications for drug formularies and health disparities. Solutions for gaps between optimal and suboptimal health care for patient mental health caused by systematic differences in sociological factors need to be well tailored. We need policy makers to be engaged in designing effective policy interventions to improve patient medication adherence, and to fund cost-effectiveness studies to improve patient outcomes and in turn, reduce associated health expenditure. PMH50 TRENDS IN ANTIDEPRESSANT UTILIZATION, AND ASSOCIATED LABOR MARKET PARTICIPATION AND QUALITY OF LIFE OUTCOMES IN THE UNITED STATES: 2004–2007 Lin HC1, Erickson S2, Smith D1, Balkrishnan R2 1 University of Michigan, Ann Arbor, MI, USA; 2University of Michigan, College of Pharmacy, Ann Arbor, MI, USA OBJECTIVES: Innovative antidepressants have been widely adopted. However, the differences in patient factors and antidepressant use, and associated patient health and work outcomes were not jointly studied. This study was trying to understand how patient factors and antidepressant utilization were associated with patient employment durations and quality of life. METHODS: A retrospective cross-sectional study was conducted using the 2004–2007 Medical Expenditure Panel Survey (MEPS) database. Proportional hazard duration models were used to examine MDD patient’s employment duration. Linear regression models were implemented to study the impacts of MDD patient factors and antidepressant utilization on associated patient physical and mental health status. RESULTS: Differences in employment duration across several patient factors were found. Cox proportional hazard model showed that, compared to uninsured patients, MDD patients covered by private insurance had a lower level of hazard of job termination (hazard ratio = 0.15, P < 0.01). Patients who were in better physical health conditions had a lower level of hazard of job termination (hazard ratio = 0.96, P < 0.01). Results from OLS regressions showed that, compared to patient without antidepressant pharmacotherapy for MDD treatment, patients who took innovative antidepressants such as SSRIs/SNRIs and other newer ones had a huge increase in MCS (β = 11.35, P < 0.01). In addition, better antidepressant adherence was significantly associated with an increase of MCS (β = 0.10, P < 0.01). CONCLUSIONS: This study suggested that effective policy interventions were needed for improving medication adherence, and the design of prescription drug benefit within health insurance should be tailored considering its associations with patient factors and related improvement in health status according to the findings of this study. We need policy makers to be engaged in designing effective policy interventions to improve patient medication adherence, which may in turn improve patient health status and labor market participation. PMH51 HEALTH STATUS AND COST OF CARE IN PATIENTS WITH DEMENTIA IN GERMANY Reese JP1, Heßmann P1, Seeberg GM1, Baum E1, Rieke J2, Müller MJ3, Balzer-Geldsetzer M1, Dodel R1 1 Phillips University Marburg, Marburg, Germany; 2Office-based neurologist, Giessen, Germany; 3Vitos Clinic for Psychiatry and Psychotherapy Marburg, Marburg, Germany OBJECTIVES: To assess cost-of-illness and patient-relevant outcomes in patients with dementia in different settings in the German health system. METHODS: 600 patients with dementia were recruited at 1) a university hospital; 2) general practitioners; 3) office-based neurologists; 4) a regional psychiatric hospital; and 5) in long-term care facilities. Socio-demographic, economic and clinical parameters were assessed using a standardized questionnaire. Disease severity was measured by means of the Minimental Status Examination (MMSE) and the Alzheimer’s Disease Assessment Scale. Neuropsychiatric status was assessed with the Geriatric Depression Scale, the Neuropsychiatric Inventory and the Alzheimer’s Disease Cooperative-Study-Activities of Daily Living. Patient’s quality of life was reported by the patient and also by the caregiver (employing the EuroQol and the QoL-AD instruments). RESULTS: For an interim analysis, 278 patients (180 female, 98 male) were available. Mean age was 78.2 yrs and mean disease duration was 4.5 yrs (SD 3.9). On average, care was needed for 3.0 yrs (SD 3.1). Cognitive impairment was severest in institutionalized patients (MMSE 12.2 pts SD 8.2) compared to the mean of all patients (MMSE 16.9 Pts). Mean EQ VAS values were 60.5 pts (SD 20.2). Disease-specific QoL-AD health status mean was 30.2 (SD 5.6). Health status was rated highest in long-term care facilities (EQ VAS) and at office-based neurologists (QoL-AD). In all severity stages patients rated their own health status (Qol-AD) better than their relatives. The costs of antidementia drugs were c45,000 per 3 months. Memantine accounted for 61%. a total of 41% of the patients received Memantine whereas 45% of the patients received no anti-dementia medication. CONCLUSIONS: Health status is considerably impaired in patients with dementia and their caregivers. Interestingly, caregivers often appraise A455 patients’ Quality of life worse than the patients themselves. Altogether, our results indicate a considerable under-supply with anti-dementia drugs. PMH52 A SHORT 12-ITEM ZARIT BURDEN INVENTORY FOR THE ASSESSMENT OF DEMENTIA CAREGIVERS AS OBTAINED BY ITEM RESPONSE THEORY Ballesteros J1, Santos B1, González-Fraile E2, Muñoz-Hermoso P2, Dominguez-Pachón A3, Martín-Carrasco M4 1 University of the Basque Country, UPV/EHU, CIBERSAM, Leioa, Spain; 2Instituto de Investigaciones Psiquiátricas-IIP, Bilbao, Spain; 3Hospital Aita Menni, Arrasate-Mondragón, Spain; 4Clínica Psiquiátrica Padre Menni, Pamplona, Spain OBJECTIVES: The Zarit Burden Inventory (ZBI) is a 22-item self-report scale frequently used to asess patients’ caregiver burden on several dimensions. As a multidimensional instrument the interpretation of its total score is sometimes unclear. Our aim was to obtain a short-ZBI unidimensional scale based on Item Response Theory (IRT) approaches. METHODS: The validation sample comprised 246 caregivers of patients diagnosed with dementia and recruited for an ongoing multi-center randomized clinical trial on the efficacy of psychoeducational interventions (EDUCA-2 trial). The pre-randomization 22-item ZBI was analyzed according to the Samejima’s graded response model to select the more informative items. The dimensionality of the scale was further tested with Confirmatory Factor Analysis (CFA). Finally, discriminant validity was assessed by Receiving Operator Characteristic (ROC) analysis and the Area Under the Curve (AUC) contrasting the short scale total score against the psychological distress criterion evaluated with the General Health Questionnaire 28-item at 5/6 cut-off. RESULTS: A 12-item short-ZBI was selected. It covered 87% of the total 22-item ZBI information and showed appropriate item curve characteristics according to the Samejima’s model. The short-ZBI had an internal reliability of 0.89 (Cronbach’s alpha), and was compatible with a unidimensional latent structure for the burden construct (CFI = 0.99; RMSEA = 0.05). According to the GHQ-28 cut-off 131 caregivers (53% of the total sample) could be considered at high risk for developping psychological distress. The discriminant validity of the short-ZBI scale against that criterion was good (AUC = 0.84, 95% CI = 0.79 to 0.89) and not significantly different from the parental 22-item ZBI (p = 0.85). CONCLUSIONS: We have found good psychometric properties for the short-ZBI scale derived from IRT. Its unidimensionality might be important to enhance its interpretation. Further psychometric studies, mainly on its sensitivity to change are now warranted. PMH53 FACTOR STRUCTURE OF A SOCIAL SUPPORT SCALE FOR ADOLESCENTS TREATED FOR SUBSTANCE USE DISORDER Ciesla J, Yao P Northern Illinois University, DeKalb, IL, USA OBJECTIVES: The literature indicates that social support is protective of relapse for adolescents treated for substance use disorder (SUD). Unfortunately, no standard measure of social support exists. The objective of this research is to use factor analysis to elucidate the underlying factor structure of a 14-item social support scale for use in outcomes assessment in this treatment population. METHODS: Subjects are 517 adolescents discharged from primary substance abuse treatment from 2004–2008. The data is from research conducted between 6 and 12 months post discharge via a 234item questionnaire that included the 14-item social support scale. The scale has questions that assess the degree to which the adolescent’s social contacts conform to norms of positive behavior and therefore foster non-use and recovery. The response rate was 62 percent. RESULTS: The factorability of the scale was assessed by Keiser-MeyerOlkin statistic (it was 0.727, > the recommended 0.6) and by Bartlett’s test of sphericity which was significant (x2 = 1066.89, p = 0.00001). The scale was decomposed by principal component factor analysis and three factors emerged. Initial Eigen values explained 65.6, 23.1 and 11.3 percent of the variance, respectively. Final factor solutions were examined using varimax, oblimin and promax rotations with 3, 4, 5 & 6 solutions, respectively. a three-factor solution via promax explaining 99 percent of the variance emerged as the best solution although results were similar using the other rotations. Factor 1 yielded ten items that are attributes of a peers’ potential to be positive a or negative influence and thus supportive of recovery. The three items in Factor 2 related to emotional dimensions of social support. Factor 3 contained two items indicating a recovering adolescent’s ability to seek-out and establish positive social contacts. CONCLUSIONS: This scale is useful as a standard measure of social support which is an important aspect of treatment success. PMH54 USE OF THE ANALYTIC HIERARCHY PROCESS TO PRIORITIZE PATIENT-RELEVANT ENDPOINTS OF ANTIDEPRESSANT TREATMENT van Manen JG1, Hummel JM1, Ijzerman MJ1, Volz F2, Gerber A2, Danner M2 1 University Twente, Enschede, The Netherlands; 2Institute for Quality and Efficiency in Health Care (IQWiG), Cologne, Germany OBJECTIVES: In deciding about coverage of new medical technology, multiple clinical outcomes are used to support reimbursement claims. Neither the real world value nor the relevance of these outcome measures for patients is systematically assessed. Hence, there is growing interest in the use of patient-reported outcome measures. Multi-criteria decision analysis, like the analytic hierarchy process (AHP), is a technique to elicit patient preferences. In the present study we used AHP to prioritize patient relevant endpoints related to the use of antidepressants in major depression. A456 METHODS: Patient relevant endpoints of treatment (remission of depression, response to treatment, no relapse, serious adverse events, adverse events, social function, anxiety, pain, cognitive function) were prioritized using pairwise comparisons of these outcomes. In two separate groups, twelve patients and seven experts judged on a 9 point scale the relative importance of pairs of two outcome measures. The geometric mean of these judgments was used to derive weighting factors for the outcome measures (scale 0–1). RESULTS: Of all outcome measures, patients rated response to treatment highest (0.32), while experts rated remission of depression highest (0.48). Adverse events were all rated lowest by patients as well as by experts, and diseasespecific quality of life domains such as social function (0.11 & 0.09), anxiety (0.12 & 0.05) and cognitive function (0.13 & 0.06) were rated in between. CONCLUSIONS: The most important outcome measures according to the patients are, in order of decreasing importance: response, cognitive function, no anxiety, social function, no relapse, no adverse events, and remission. The AHP appears to be suitable in gaining an overview of the importance of patient relevant outcome measures. An additional advantage of AHP is that the group discussions offer insight in the question why the endpoints are important. PMH55 THE SUBJECTIVE WELL-BEING UNDER NEUROLEPTIC SCALE SHORT FORM (SWN-K20) AND THE SF-36 AS QUALITY OF LIFE MEASURES IN SCHIZOPHRENIC PATIENTS Sanjuán J1, Haro J2, Maurino J3, Diez T4, Ballesteros J5 1 University of Valencia, CIBERSAM, Valencia, Spain; 2Parc Sanitari Sant Joan de Deu, CIBERSAM, Sant Boi de Llobregat, Spain; 3AstraZeneca, Madrid, Spain; 4AstraZeneca, Zaventem, Brussels, Belgium; 5University of the Basque Country, UPV/EHU, CIBERSAM, Leioa, Spain OBJECTIVES: Outcomes research in patients with schizophrenia should take into account the subjective interpretation of the mood and physical changes accompanying medication. Those changes influence the behavioural response to treatment and ultimately the patient’s clinical outcome as mediated by his treatment compliance. Our aim was to assess the relationship between a specific well-being measure, the SWN-K20 that presents a general and 5 specific measurement subdomains (mental functioning, social integration, emotional regulation, physical functioning, and self-control), and the 8 domains of the SF-36 v1 as a general quality of life measure. METHODS: The validation sample for this study comprised 97 patients diagnosed with schizophrenia and who were rated as clinically stable at the moment of the study (1 week test-retest intraclass correlation coefficient for clinical symptoms = 0.96). The patients were recruited as part of a multicenter psychometric trial to validate the SWN-K20 in Spanish. The associations between the domains of the SWN-K20 and the SF-36 were evaluated by the Spearman’s rank correlation test. RESULTS: All correlations among domains were positive and most were statistically significant (p < 0.05). As expected the bodily pain domain of the SF-36 presented the lower correlations with the SWN-K20 (rho range of 0.10 to 0.25), whereas the other 7 domains correlated significantly with the total SWN-K20 score (rho range 0.49 to 0.60, all p < 0.001). Overall the largest correlations were obtained between the SWN-K20 and the SF-36 domains of general health (rho = 0.53), mental health (rho = 0.60), and vitality (rho = 0.54). CONCLUSIONS: The positive but nevertheless moderate correlations observed between a specific well-being scale, as the SWN-K20, and a general quality of life scale, as the SF-36, supports the inclusion of specific and diagnose-tailored instruments for outcome assessments of patients with schizophrenia. PMH56 INNOVATIONS IN COMBINING PATIENT REPORTED OUTCOMES WITH COGNITIVE TESTING DATA TO STREAMLINE AND LEVERAGE REAL-TIME DATA COLLECTION Curry C PHT Corporation, Boston, MA, USA OBJECTIVES: Understand features of an electronic device that allow a marked improvement in the quality of collected data; the importance of improved data quality leading to enhanced patient safety and drug labeling; populations best suited for paired PRO and cognitive measurement technologies; important practical considerations for implementation in clinical trials including training and compliance; the potential for using real-time parallel data for adverse event safety monitoring. METHODS: This session discusses using ePRO and biometric technology for parallel data capture emphasizing advantages, disadvantages, execution, and ways to leverage these data. The session will review PRO and cognitive testing technologies, including comparisons of devices that combine physiological measures with a patient interface with systems that use separate PRO input and biometric devices. RESULTS: Assessing a treatment’s ability to enhance or prohibit reduction of cognitive processing efficiency is an emerging study in the pharmaceutical industry. Case studies examine how the use of cognitive function tests in combination with ePRO can enhance the data collection so drug effects otherwise unidentified can be determined. The speaker will discuss the future of ePRO combined with biometric measurements as a standard of clinical research. CONCLUSIONS: Clinical trial endpoints can involve collection of physiologic and patient-reported outcome data; a combination of subjective and objective data. Electronic forms of information capture assure trial efficiencies including edit checks and shorter time to database lock. ePRO provides time-stamped, legible and complete data from subjects. Biometric devices capture the physiological measurements. Typically, cognition data have been collected from patients separately from PRO data during clinical trials, increasing respondent burden and risk of error such as transposing 13th Euro Abstracts manually entering data. The use of ePRO and biometric devices, evolution of data transmission technology, and greater technologic sophistication of consumers, provide an opportunity for parallel electronic data capture, simultaneously capturing and transmitting physiologic and PRO parameters in clinical studies. PMH57 REVIEW OF CO-MORBIDITY OF EATING AND BODY DYSMORPHIC DISORDERS Faragó E, Pakai A, Dér A, Oláh A, Boncz I, Németh K University of Pécs, Pécs, Hungary OBJECTIVES: Our intention was to investigate the prevalence of Eating Disorder and Body Dysmorphic Disorder in patients diaghnosed with depression, anxietay and borderline personality disorder by gender, using controll groups. METHODS: The reserach was 2009 at the Department of Phsyciatry of Szigetvár Hospital (Hungary). Eating Disorder Inventory was used. a self-made questionnaire aimed the body mass, body height and several demographic data. Inclusion criteria: ones between 18 and 50 years age, and according to BNO F32-F34 (depression), F41 (anxiety), F6030 (borderline personality disorder) diagnoses. Control group: participants with age between 18 and 50 years like that, who do not stand under a psychiatric treatment. The statistical analysis consisted of two sample T test, χ2-probe. RESULTS: The target group’s number was 82, the control group 85. In the “The feeling of the insufficiency” (P < 0.001), “Interpersonal distrust” (P < 0.005), “Interoceptive consciousness” (P < 0.001) scales, the members of the control group from all three psychiatric patient groups reached a significantly lower score away. In the “Bulimia” scale there was a significantly lower score in the control group as well than the borderline in a group (P < 0.005), and here I found a significant difference between the members of two psychiatric groups: the anxious group reached a lower score, compared with the borderline group (P < 0.001). CONCLUSIONS: The three psychiatric patient groups did not attain the threshold value onto one of the eating disturbances relevant scales neither There is not a direct, causal contact between the examined psychiatric clinical pictures. The men’s higher result achieved on the „Bulimia” scale relates rather onto the binge eating disorder. PMH58 MODELING PROGRESSION IN DEMENTIA: ASSESSING THE PERFORMANCE OF FIVE CLINICAL MEASURES IN SPANISH SUBJECTS AND CAREGIVERS Darba J1, Kaskens L2, Lacey L3, McLaughlin T4 1 Universitat de Barcelona, Barcelona, Spain; 2BCN Health, Barcelona, Spain; 3JANSSEN Alzheimer Immunotherapy, Dublin, Ireland; 4Alzheimer Immunotherapy Research & Development, Dublin, Dublin, Ireland OBJECTIVES: The primary objective of this analysis was to compare five different clinical measures and their impact on economic modelling. Clinical measures compared are the Mini-Mental State Examination (MMSE), the Cognitive Component Score (CCS), the Functional Component Score (FCS) the Behaviour Component Score (BCS), and the Dependence Scale (DS). METHODS: The MMSE, CCS, FCS, BCS and DS were compared in their ability to explain variation in clinical outcomes, economicand other utilized resources, caregiver burden (Zarit Scale) and caregiver QoL (EQ-5D) using univariate (Pearson correlations) and multivariate (linear regression) analyses. Data on subjects and caregivers was obtained from multiple centres in Spain. RESULTS: In total 394 subjects, males and females aged 50 to 93 years old with mild cognitive impairment to severe dementia were included in this study. CCS, FCS, BCS and DS were moderately correlated with MMSE, with Pearson correlations ranging from −0.26 for BCS to −0.56 for CCS. These four clinical measures were also moderately correlated with medical costs, Zarit Scale and EQ-5D while MMSE was not. These measures also performed better in explaining variation in medical costs, Zarit Scale and caregivers’ EQ-5D. MMSE performed better explaining variation in the number of concomitant conditions and caregiver time (hours per day). CONCLUSIONS: The CCS, FCS, BCS and DS are better predictors in modelling AD progression on a higher number of variables including medical costs, caregivers’ burden and caregivers QoL than the MMSE. PMH59 EVALUATION OF THE EFFECT OF ARIPIPRAZOLE ON QUALITY OF LIFE IN PATIENTS WITH SCHIZOPHRENIA IN A PROSPECTIVE, MULTICENTRE, OPEN-LABEL STUDY Bervoets C1, Constant E2, Sabbe B1, Morrens M1, Vansteelandt K3, De Patoul A4, Pitsi D5, Halkin V6, Loze JY7, Peuskens J8 1 University Antwerp, Antwerp, Belgium; 2UC Louvain, Bruxelles, Belgium; 3KU LEUVEN, Kortenberg, Belgium; 4BMS, Bruxelles, Belgium; 5BMS, Bruxxelles, Belgium; 6Bristol-Myers Squibb, Braine-l’Alleud, Belgium; 7Otsuka Pharmaceutical Co., Ltd., Rueil-Malmaison, France; 8 KULeuven, Leuven, Belgium OBJECTIVES: Aripiprazole has been claimed to have a beneficial effect on cognition with an emphasis on verbal functioning in schizophrenic patients. a prospective, multicenter, open-label study of Aripiprazole was set to evaluate the effect on quality of life, in relation to illness severity and cognitive functioning of a treatment with aripiprazole in schizophrenic patients. METHODS: A total of 363 schizophrenic patients from 18 to 65 years, treated with different typical and atypical antipsychotics or had no previous treatment, were switched to aripiprazole after a 2 week washout period. Quality of life was assessed by use of the Quality of Life Enjoyment and Satisfaction Questionnaire (QLESQ) at 3 separate test moments in a 12 weeks period. A457 13th Euro Abstracts At the same time, verbal cognitive function was assessed with the California Verbal Learning Test (CVLT) and the Verbal Fluency test (VF). The Clinical Global Impression-Severity (CGI-S) was assessed as well at the same visits. Statistical analysis was done using SAS 9.2 with the PROC MIXED module for mixed effects repeated measures analysis. RESULTS: A mixed models analyses on the QLESQ scores showed significant effects for CGI-S in all subscores but “ school “ (p = 0,1485) and for the verbal cognitive measures only in subscores “leisure “ for CVLT-LTFR (p = 0,0002). CONCLUSIONS: The observed difference over time in QLESQ values is dependent on the CGI-S score which contributes mostly to the prediction of the QLESQ outcome. Verbal cognitive outcome was not predictive . These findings are not in agreement with other publications reporting an independent contribution of both measures on quality of life. As the study was open-label, interpretation of the results should be approached with some caution. PMH60 QUALITY: A NON-INTERVENTIONAL STUDY EVALUATING QUALITY OF LIFE IN SCHIZOPHRENIC PATIENTS TREATED WITH ATYPICAL ANTIPSYCHOTICS, IN THE AMBULATORY SETTING. A 9-MONTH, OBSERVATIONAL, MULTICENTRIC PROSPECTIVE STUDY Peuskens J1, Tomas M2 1 University Psychiatric Centre Catholic University Leuven, Kortenberg, Belgium; 2AstraZeneca Belgium, Brussels, Belgium OBJECTIVES: The QUALITY study evaluated Quality-of-Life in schizophrenic patients treated with atypical antipsychotics (AAPs) in the ambulatory setting. METHODS: Patients (18–65 years-old) diagnosed with schizophrenia and had started treatment with one AAP before visit-1 (minimum: 4-weeks, maximum: 8-weeks) were enrolled into this Belgian study. At visit-1 patients’ demographics and medical history were recorded with follow-up visits after 3-, 6- and 9-months. At each visit, patients completed the Subjective Well-being under Neuropletic treatment short form (SWNK), while investigators assessed the Positive and Negative Symptom Scale (PANSS-8) and the Global Assessment of Functioning (GAF). RESULTS: A total of 121 patients (mean age 36.7 ± 10.8, 75.2% male) were enrolled with the main AAP treatments risperidone (31.4%), apripirazole (23.1%), quetiapine (20.7%). At baseline, the associations between the SWN-K subscales and age, gender, positive/negative PANSSscores, Schizophrenia-subtype or treatment-subgroups were small. Mental-functioning and physical-functioning subscales were significantly correlated with negative PANSSscores, as was the emotional-regulation subscale with gender. The majority of changes from baseline in the SWN-K subscale scores were between −0.5 and +0.5 (range: −1.8 to 1.6). Most mean changes were positive, suggesting that patients felt better compared to baseline. PANSS-8-score changes were slightly negative, suggesting decreases in symptom severity. 57.5% of patients showed symptom remission (PANSS-8-score ≤ 3 after 6 months-of-treatment).GAF scores increased (study end: 8.2 ± 16.0) indicating improvement in the patient’s global functioning. By the study end, the number of hospitalizations and addictive substance use had decreased. Patients with more severe negative symptoms considered their mental- and physical-functioning better and men were more satisfied than women with their emotional-regulation. Few patients had sufficient disease insight; most patients had a vague notion or did not see that they were ill. CONCLUSIONS: The patients’ well-being seemed to improve slightly during the study without differences between the treatments. Patients having more severe negative symptoms considered their mental and physical functioning better at the study end. PMH61 PRODUCTIVITY LOSS AMONG PATIENTS WITH MOOD DISORDERS 1 1 2 2 1 Bouwmans C , Rutten F , François C , Danchenko N , Hakkaart-van Roijen L 1 Erasmus University Rotterdam, Rotterdam, The Netherlands; 2Lundbeck SAS, Issy-lesMoulineaux, France OBJECTIVES: Mood disorders are generally associated with a high level of productivity loss and a substantial economic burden. The aim of this study was to measure the impact of mood disorders on productivity in terms of absence from work, at-work productivity loss, and job performance among patients in the The Netherlands. METHODS: Patients meeting DSM-IV criteria for depressive disorder, dysthymic disorder, panic disorder (with or without agoraphobia), social phobia and generalised anxiety disorder were included from a large naturalistic trial in 7 Dutch mental health care centres. The short form of the Health and Labour Questionnaire (SF-HLQ) was used to assess absenteeism (days absent from work), presenteeism (at-work productivity loss) and job performance (on a 3-point scale ranging from ‘not impeded’ to ‘very much impeded’). ANOVA analyses and t-tests were performed to explore differences between socio-demographic groups. RESULTS: Of the total 702 patients, two thirds were in paid employment. Over 3 months, 55% reported long-term absenteeism (>2 weeks), with a mean of 73 work days lost per patient, and 13% patients reported a shorter-term absenteeism (<2 weeks), with a mean of 33 work days lost per patient. Women were absent from work for a longer (99 days) than men (46 days) (P < 0.005). Presenteeism accounted for a 12% productivity loss in terms of working time lost. The number of days at work while impeded by health problems was numerically higher in men than in women (7.4 vs. 5.6, p > 0.05). Overall, 31% of the patients had difficulties in job performance when at work. CONCLUSIONS: Mood disorders are associated with impairment of patients’ productivity in terms of absence from work, presenteeism and difficulties in job performance The study demonstrated a tendency towards lower absenteeism and higher presenteeism in men compared to women. MENTAL HEALTH – Health Care Use & Policy Studies PMH62 TRENDS IN ANTIDEPRESSANTS AND ANXIOLYTICS DRUG USAGE IN CROATIA AND SLOVENIA Vitezic D1, Kucan M2, Vitezic M3, Mrsic Pelcic J4 1 University of Rijeka Medical School and University Hospital Centre Rijeka, Rijeka, Croatia; 2 JADRAN Galenski laboratorij, Rijeka, Croatia; 3University of Rijeka Medical School and Orthopedic Hospital Lovran, Lovran, Croatia; 4University of Rijeka, Rijeka, Croatia OBJECTIVES: The objective of this study is to analyze antidepressants and anxiolytics drugs usage (ATC subgroups: N06A and N05B) in two mid European, neighbouring countries, Croatia and Slovenia, for the 9-year period, from 2000 to 2008. Further, the aim was to identify the generic drugs usage in each country, the price for DDD for original and generic drugs, and to identify the most prescribed drugs in each drug group. METHODS: The data have been obtained from the International Medical Statistics database for Croatia and Slovenia. Drugs usage is presented in defined daily doses per 1000 inhabitants per day (DDD/1000) according to the WHO Methodology. Financial expenditure data are presented in Euros. An average cost per DDD was calculated for each drug group. RESULTS: In 2008, the total usage of antidepressants was higher in Slovenia (42.7 DDD/1000 inhabitants/day) than in Croatia (22 DDD/1000 inhabitants/day), and it increased in both countries during the investigated period. The total usage of anxiolytics is more than 3 times higher in Croatia (67.4 DDD/1000 inhabitants/day) than in Slovenia (22.2 DDD/1000/inhabitants/day). The total usage of anxiolytics decreased in Slovenia in 2008 in comparison with prescriptions in 2000, while it increased in Croatia for 44.3% during the same period. The rate of generic prescriptions among antidepressants during the investigated period was higher in Croatia, i.e. in Croatia 73.6% of all prescribed antidepressants were generics in 2008, while in Slovenia 33.5%. CONCLUSIONS: Drugs prescription patterns are different comparing Croatia and Slovenia. a possible reason for relatively higher usage of anxiolytics and lower usage of antidepressants in Croatia could be depression treatment with anxiolytics. This indicates the need for a more thorough analysis and the introduction of national drugs guidelines for rational prescribing, monitoring and evaluation especially anxiolytics. Although the generic drugs usage in the mentioned groups is relatively high, it should be further supported and promoted. PMH63 ANTIPSYCHOTIC PRESCRIBING TO THE ELDERLY: A TWO-YEAR COMPARATIVE ANALYSIS Truter I Nelson Mandela Metropolitan University (NMMU), Port Elizabeth, Eastern Cape, South Africa OBJECTIVES: Antipsychotic medicines are commonly prescribed to elderly patients. These patients are at an increased risk of adverse drug events because of age-related pharmacodynamic and pharmacokinetic changes. The primary aim of the study was to determine antipsychotic prescribing patterns and cost to patients aged 60 years and older in a private health care sector primary care patient population in 2008 and 2009. METHODS: A retrospective, exposure-cohort drug utilization study was conducted on prescription data of a pharmacy group in South Africa for 2008 and 2009. No diagnoses were available. All records for antipsychotics (MIMS category 1.5) for patients 60 years and older were extracted for analysis. RESULTS: A total of 1 800 patients in 2008 and 3 086 patients in 2009 received 7 877 and 14 538 antipsychotic products at a sales value of R2 785 896 and R4 928 127, respectively. The average cost per antipsychotic product was R353.67 in 2008 and R338.98 in 2009. Most patients were female (61.78% in 2008 and 57.87% in 2009). Tablets were the preferred dosage form. Risperidone (37.55% in 2008, and 38.20% in 2009), quetiapine (27.56% and 28.88%) and olanzapine (10.91% and 10.24%) together accounted for more than 75% of antipsychotics prescribed to the elderly. a high percentage of atypical antipsychotics (84.23% in 2008 and 85.14% in 2009) were prescribed. Differences were observed between 2008 and 2009 with respect to the prescribing frequency of the atypical antipsychotics (chi2 = 32.624; d.f. = 6; p < 0.0001). CONCLUSIONS: Prescribing patterns were relatively similar in both years. The dosages and duration of treatment of antipsychotics should be investigated in relation to the diagnoses for which they are prescribed. It is recommended that diagnoses be included in databases and also that qualitative studies be conducted to determine possible side effects experienced by patients. PMH64 HEALTH CARE ORGANIZATIONS/SERVICES CLUSTERING: EXAMPLE OF PCTS IN ENGLAND Millier A1, Toumi M2, Lamure M3, Aballea S1, Flostrand S1, Hansen K4 1 Creativ Ceutical, Paris, France; 2University Claude Bernard Lyon 1, Lyon, Rhone Alpes, France; 3University Claude Berard Lyon 1, Villeurbanne, Rhone Alpes, France; 4Lundbeck SAS, Issy-les-Moulineaux cedex, France OBJECTIVES: England has 152 Primary Care Trust (PCT) each having its own characteristics, and prescription patterns. As a consequence, PCT formulary lists differ, creating an additional hurdle for patient access to innovative medicines, often referred to as the post-code lottery. The objective of this study was to segment these PCTs in groups with homogeneous attitudes towards patient access for innovative drugs in Mental Health (MH) using a clustering analysis. METHODS: We collected all available information about the PCTs through public sources and the OneKey + database. We identified 62 variables per PCT and classified them in 5 groups: profile A458 of population, economic & financial information, organization, prescription pattern and patient management. We first ran a principal component analysis on quantitative variables and then a multiple correspondence analysis on qualitative variables, in order to get individuals coordinates of PCTs on several axes on which they were projected. We then merged the results, and ran an analysis to create PCTs clusters. RESULTS: We identified 5 clusters of PCTs with differentiated, homogeneous attitude towards innovative drugs. Cluster 1 focused on MH disorders with GP-level decision-making, cluster 2 focused on MH disorders and contained high-prescribing physicians, cluster 3 had low recognition of MH disorders, cluster 4 had low recognition but offered a high level of care while cluster 5 was focused on cost containment. The clusters also varied in population size from 1.2 M to 23.9 M. CONCLUSIONS: As this work was rather innovative it was validated by local pharmaceutical sales and marketing organization. The cluster approach proposed was endorsed by account managers’ pragmatic experience in the field and correlated well with their experience of obtaining access for their innovative MH drugs. This study helped to improve our understanding of the UK landscape in patient access for innovative drugs in MH. PMH65 EFFECTIVENESS OF PHARMACEUTICAL THERAPY OF ADHD (ATTENTION-DEFICIT/HYPERACTIVITY DISORDER) IN ADULTS—A HEALTH TECHNOLOGY ASSESSMENT Benkert D1, Krause KH2, Wasem J3, Aidelsburger P1 1 CAREM GmbH, Sauerlach, Germany; 2Ludwig Maximilian University, Munich, Germany; 3 University of Duisburg-Essen, Essen, Germany OBJECTIVES: Attention-Deficit/Hyperactivity Disorder (ADHD) is a neurobehavioral developmental disorder, that is characterized by hyperactivity and impulse control disorder. In Germany pharmaceutical therapy is approved for children and adolescents solely. Therefore, treating adult patients with ADHD complies with off-label use of stimulants. Aim of this Health Technology Assessment is to examine the clinical effectiveness, cost-effectiveness, ethical-social and legal aspects of pharmacotherapy (stimulants, antidepressants, norepinephrine reputake inhibitors) of adult ADHD patients. METHODS: A systematic literature research is conducted in relevant electronic literature databases. Studies matching the ex ante defined inclusion criteria are assessed systematically and qualitatively according to methodical standards by two reviewers. Due to heterogeneity of included studies no meta-analysis is performed. RESULTS: Nine randomised controlled trials (RCT), five systematic reviews, three economic and two studies concerning legal aspects are included. All RCT show improvement regarding ADHD symptoms (hyperactivity, impulsivity). Response rates vary between 7 % and 42 % in the control group and 17 % to 59,6 % in the intervention group. The studies use primarily investigator and self-rated questionnaires such as the ADHD Rating Scale, the Conners Scales and the Clinical Global Impression as outcome parameters for the core symptoms of ADHD. Some studies show a larger improvement of the ADHD symptoms by a flexible dose approach. The systematic reviews demonstrate statistically significant improvement in symptoms of ADHD compared to placebo and other medications. The effect sizes for stimulants are somewhat higher than for non-stimulants. Adult ADHD patients cause higher annual direct and indirect costs than matched controls. The average medical costs are reported with 1.262 US-Dollar in 1998 and 1.673 US-Dollar in 2001. CONCLUSIONS: Methodological limitations of the RCT are the short study duration and the high drop-out rates. Further research is needed to determine the cost-effectiveness of medical treatment of adult ADHD patients. PMH66 PREVALENCE OF AFFECTIVE DISORDERS IN SOUTH-WEST REGION OF SWEDEN Löfroth E, Myrén KJ, Bruce S, Dahlberg E IMS Health, Stockholm, Sweden OBJECTIVES: To determine the prevalence of affective disorders (section F3 in ICD-10) in South-West region of Sweden. METHODS: This was a retrospective longitudinal descriptive database study of the utilization of health care of patients from the South-West region of Sweden (1.5 million inhabitants). All patients who were diagnosed with affective disorders between 2001 and 2007 in the South-West region of Sweden were included in the study. The patients were diagnosed in primary-, in- and out-patient care. RESULTS: A total of 138,194 patients were diagnosed with at least one type of affective disorder. This corresponded to a prevalence of 9% in the population of South-West region in Sweden. 89 percent (123,524 patients) of the total sample were diagnosed with major depressive disorder (F32) followed by recurrent depressive disorder (F33) at 14% (19,944 patients). Bipolar affective disorder (F31) was diagnosed in 5% of the patients (6932) and unspecified mood (affective) disorder (F39) was diagnosed in 4% of the patients (5,431). The prevalence of the different diagnoses were 8 percent for depressive episode, 1 percent for recurrent depressive disorder, 0.5 percent for bipolar affective disorder, 0.4% for unspecified mood (affective) disorder. CONCLUSIONS: Depressive episode was by far the most common of the affective disorders diagnosed, with a prevalence of 8 percent. a question remains whether the prevalence for bipolar is underestimated as this study found it to be 0.5%. The expected prevalence for bipolar disorder is 1 percent of the population. One plausible explanation could be that bipolar patients are classified with other codes than F31 in the South-West region of Sweden. 13th Euro Abstracts PMH67 FACTORS ASSOCIATED WITH ANTIPSYCHOTICS USE AMONG COMMUNITY-DWELLING OLDER PERSONS WITH DEMENTIA Rhee Y1, Shega J2, Csernansky J1 1 Northwestern University, Chicago, IL, USA; 2University of Chicago, Chicago, IL, USA OBJECTIVES: Despite the use of antipsychotics to treat dementia patients, the efficacy of the treatment is not well established. We examine factors associated with antipsychotics use among community-dwelling older persons with dementia. METHODS: We used data from The Aging, Demographics, and Memory Study (ADAMS) to assess dementia severity and service use from 2002 to 2004. We used logistic regressions to identify factors associated with antipsychotic use; N = 307. RESULTS: Among older persons with dementia living in community (weighted sample = 207,544), 7.8% (weighted sample = 16,272) took any antipsychotic medications; 69.1% were female, 70.7% were white, and 21.8% were African American. The average age was 85 years. Physical functions were measured by the number of ADL (average. 2.9) and number of IADL (mean 3.6). The most frequent primary diagnoses were Alzheimer’s disease (74.3%), vascular dementia (16.0%) and other dementia (9.8%).The most frequently prescribed antipsychotics were: risperidone (58.0%), quetiapine (15.4%), and haloperidol (7.7%). Of those taking an antipsychotic, 86.0% were diagnosed with Alzheimer’s dementia. We used the Neuropsychiatry Inventory (NPI) for behavior problems (delusions, hallucinations, agitation/aggression, depression, apathy, elation, anxiety, disinhibition, irritability/lability, and aberrant motor behavior). We evaluated severity of dementia using the Clinical Dementia Rating Scale (CDR).Communitydwelling older persons with dementia are significantly more likely to receive antipsychotics if they were agitated (OR = 3.4, P < 0.05), had disinhibition (OR = 4.6, P < 0.05), or had greater dementia severity (OR = 1.9, P < 0.01). Also, Medicaid recipients were significantly more likely to receive antipsychotic medications (OR = 5.4, P < 0.01). Participants were significantly less likely to be medicated with antipsychotics if they had vascular dementia (OR = 0.09, P < 0.05) or the caregivers were clinically depressed (OR = 0.2, P < 0.05). CONCLUSIONS: Community-dwelling older persons with dementia are more likely to receive antipsychotics if caregivers report behavior problems, the dementia is more advanced, and the patient has Medicaid coverage. Persons with vascular dementia are less likely to be treated with antipsychotics. PMH68 TREATMENT DISCONTINUATION IN A LOCAL MENTAL HEALTH CARE SYSTEM IN GERMANY FROM THE PATIENTS’ AND HEALTH CARE PROVIDERS’ PERSPECTIVE Pueschner F1, Diewald V2, Schoul A3, Fleischmann J1 1 Janssen-Cilag Germany, Neuss, Germany; 2Rhein Kreis Neuss, Neuss, Germany; 3 YouGovPsychonomics, Cologne, Germany OBJECTIVES: Treatment discontinuation in psychiatric care is a major cause for the deterioration of mental health problems. This study assessed reasons for treatment discontinuation of patients suffering from schizophrenia or depression in a local mental health care system in Germany. METHODS: A retrospective survey was performed to assess how coordination of care through mental health care providers in the Rhein-Kreis-Neuss region (Germany) is organized. Within this survey reasons for treatment discontinuation were explored from the perspective of providers and patients. 88 providers of care (60 institutions) and 30 patients diagnosed with schizophrenia and 17 patients diagnosed with major depression were interviewed using indepth face-to-face interviews. Additionally providers were qualitatively interviewed on strategies to prevent discontinuations. RESULTS: Providers were asked to rank pre-specified reasons for treatment discontinuation according to their occurrence (often, occasionally, seldom, never). 38.8 % ranked non-adherence to medication as often, 38.1 % waiting times for next treatment. 25.6 % identified problems in the patient’s social environment as being often a reason for treatment discontinuation, 23.7 % mentioned revolving doors effects, 19.3 % resistance to therapy and 12.3 % lack of communication between providers. Patients diagnosed with schizophrenia (mean age 45 years; mean duration of illness 16 years) and depression (mean age 45 years; mean duration of illness 13 years) indicated that they consider discontinuation or actually discontinue treatment when there is a lack of social integration or when there is no reaction of providers to articulated problems with treatments. In response to an open question on suggestions for improvement, providers recommended better cross-sectoral coordination, earlier start of treatment and early return to normal daily living as appropriate measures. CONCLUSIONS: From the subjective perspective of providers and patients a variety of causes seem to influence treatment discontinuation. More research into specific interventions to improve treatment adherence is needed. PMH69 INFLUENCE OF WAITING TIME IN A LOCAL MENTAL HEALTH CARE SYSTEM IN GERMANY FOR CARE IN PATIENTS DIAGNOSED WITH SCHIZOPHRENIA OR DEPRESSION Fleischmann J1, Diewald V2, Schoul A3, Pueschner F1 1 Janssen-Cilag Germany, Neuss, Germany; 2Rhein Kreis Neuss, Neuss, Germany; 3 YouGovPsychonomics, Cologne, Germany OBJECTIVES: To assess if patients suffering from schizophrenia or depression are affected by waiting times for the next treatment in a mental health care system in Germany. METHODS: Aa retrospective survey was performed to assess coordination of care through providers engaged in the care for patients with schizophrenia or major depression in the Rhein-Kreis-Neuss region (Germany). One parameter for evaluating quality of coordination is the occurrence of waiting times, emerging when patients are A459 13th Euro Abstracts transferred from one point of care to the next with treatment continuation potentially being delayed. 88 providers of care (60 institutions), 30 patients diagnosed with schizophrenia and 17 patients diagnosed with depression were surveyed through indepth interviews. The total number of questions was 56 for providers, and 54 questions for patients. RESULTS: A total of 56% of patients diagnosed with schizophrenia (mean age 45 years, mean duration of illness 16 years) experience waiting times when transferred to an outpatient rehabilitation center, 50% while being transferred to an institution of assisted working or to a place of sheltered living (33%). Patients diagnosed with major depression (mean age 45 years; mean duration of illness 13 years) experience waiting times when transferred to an institution of assisted working (63%), to an outpatient rehabilitation center (57%) or to a psychotherapist (56%). Providers experience waiting times when they transfer patients to psychotherapists (55%). a total of 40% of patients diagnosed with schizophrenia and 58% of patients diagnosed with major depression perceive waiting times as heavily constraining, affecting their treatment and care, and potentially leading to a setback in their treatment progress. Providers see capacity constraints as main reason for waiting times. CONCLUSIONS: This study describes waiting times in coordination of mental health care between different providers as a problem. Potential negative effects on treatment success show the need for optimization. MENTAL HEALTH – Conceptual Papers & Research on Methods PMH70 TURKISH CULTURAL ADAPTATION AND VALIDATION OF THE ALCOHOL DEPENDENCE SCALE Malhan S, Oksuz E, Yilmaz F, Tekin R, Baytar S Baskent University, Ankara, Turkey OBJECTIVES: The Alcohol Dependence Scale (ADS) is a quantitative measure of the severity of alcohol dependence consistent with the concept of the alcohol dependence syndrome. The 25 items cover alcohol withdrawal symptoms, impaired control over drinking, awareness of a compulsion to drink, increased tolerance to alcohol, and salience of drink-seeking behavior. This study aims to adapt the ADS questionnaire into Turkish culture and check the reliability and validity of the questionnaire culturally. METHODS: The original instrument was translated and back translated by two independent translators. For psychometric measures, a small sample was used to check the initial comprehension and factibility. Cronbach’s Alfa was used to assess reliability and factor analysis to assess dimensionality. The EuroQol questionnaire and corresponding Visual Analogue Scales were used for concurrent validity. RESULTS: A total of 200 students between the ages of 19–28 were participated in the study. Mean age was 22. The internal consistency coefficient (Cronbach’s alpha) of ADS was 0.92. Factor analysis of the scale revealed that it was composed of six factors with Eigenvalues >1.0, accounting for 74 % of the total variance. Correlations were moderate with EuroQol and VAS. CONCLUSIONS: The culturally adapted Alcohol Dependence Scale has good validity and reliability. The instrument is likely to be suitable for the evaluation of alcohol dependence of people in Turkey. PMH71 THE APPLICABILITY OF THE QALY AS EFFECT MEASURE IN CRIMINAL YOUTHS Schawo S, Hakkaart L Erasmus University Rotterdam, Rotterdam, Zuid Holland, The Netherlands OBJECTIVES: In health economics, cost-effectiveness analyses are performed based on standardized established methods expressing outcomes in terms of costs per quality-adjusted life-year (QALY). In the field of care for criminal youth, interventions are increasingly drawn into the medical domain and towards reimbursement protocols so that a common cost-effectiveness measure becomes a necessity. However, cost-effectiveness of these interventions has so far not been expressed in terms of cost per QALY, but in terms of clinical measures. The aim of this study is to explore the applicability of the QALY as effect measure of interventions for criminal youth. METHODS: The 112 juvenile delinquents between 12 and 18 years included in this study were recruited at the start of Functional Family Therapy (FFT) from five mental health institutions in The Netherlands. Treatment efficacy was assessed using clinical outcome measures SDQ, YSR, CBCL, and PACS. Quality of life was measured using EuroQol EQ-5D. Data was analyzed to determine whether quality of life and clinical outcome measures show a significant correlation. Dutch population EQ-5D data was consulted to determine whether EQ-5D sufficiently captures differences between criminal and noncriminal groups. RESULTS: Comparison between groups of criminal and non-criminal youths yielded no significant differences for juvenile delinquents or their parents. Correlation between several clinical outcome measures and EQ-5D scores of delinquent youths has been found suggesting that EQ-5D captures some differences in criminal behaviour. CONCLUSIONS: The study confirms the expected correlation between several classical forensic effect measures and general health economics outcomes, but several aspects were not reflected by the QALY. Furthermore, the QALY measure insufficiently distinguishes between criminal and non criminal groups. We suggest further research before considering the application of the QALY as an effect measure for delinquent youths. PMH73 IMPACT OF PLACEBO RUN-IN PERIOD ON RELATIVE TREATMENT EFFECT IN GENERALIZED ANXIETY DISORDER (GAD) Sharma R, Arora M, Ubhadiya B Heron Health Private Ltd, Chandigarh, India BACKGROUND: High placebo response is observed in trials measuring subjective outcomes as in the case of GAD. This contributes to investigational drug failure owing to lack of differentiation between intervention and control group. Placebo run-in is frequently used to eliminate early responders however this has been criticised on the ground that it favours the active treatment. OBJECTIVES: To test the hypothesis that “The use of placebo run-in overestimates the treatment effect in placebo-controlled trials conducted in GAD”. METHODS: The Medline and EMBASE databases were searched to retrieve randomised, placebo-controlled trials conducted in adult GAD patients. Citations were screened on the basis of title and abstract and full text article of citations meeting eligibility criteria were sourced for detailed evaluation. Following this, citations reporting relevant outcomes were extracted using a pre-defined extraction grid. Clinically relevant dichotomous outcomes: CGI-Improvement response, HAM-A response, and HAM-A remission were extracted. Considering heterogeneity among studies included, the random effects meta-analysis was performed in STATA 9.0 using a standard meta-analysis approach. Results are presented as odds ratios with a 95% confidence interval. RESULTS: The literature search retrieved 1133 citations of which 41 were included following detailed evaluation (23 used placebo run-in phase). The point estimate for the relative treatment effect was similar (confidence interval overlap widely) irrespective of the use of placebo run-in phase (OR: 2.13 with run-in vs. 2.19 without run-in). Results were similar when individual outcomes were analyzed separately (HAM-A response OR: 2.17 vs. 2.14; CGI-I response OR: 2.20 vs. 2.13; HAM-A remission OR: 1.98 vs. 2.05). CONCLUSIONS: Irrespective of outcomes considered, there was no statistically significant difference in the effect size between trials involving placebo run-in phase compared to those that did not. Our results do not support the hypothesis that the inclusion of placebo run-in overestimates the relative treatment effect in GAD. SYSTEMIC DISORDERS/CONDITIONS – Clinical Outcomes Studies PSY1 DEVELOPMENT OF A CO-MORBIDITY SCALE IN PATIENTS WITH CHRONIC LYMPHOCYTIC LEUKEMIA Carbonell F1, De La Serna J2, Giraldo P3, Lopez A4, Rios E5, Perulero N6, Castro-Gomez AJ7 1 Consorci Hospital General Universitari Valencia, Valencia, Spain; 2Hospital Universitario 12 de Octubre, Madrid, Spain; 3Hospital Universitario Miguel Servet, Zaragoza, Spain; 4Hospital Universitari Vall d’Hebron, Barcelona, Spain; 5Hospital Virgen de Valme, Sevilla, Spain; 6IMS Health, Barcelona, Spain, 7Roche Spain, Madrid, Spain OBJECTIVES: The study aimed to develop a standardized scale for co-morbidity assessment in Chronic Lymphocytic Leukaemia (CLL) patients to support physicians for selecting the optimal treatment according to the patient co-morbidity profile. METHODS: The co-morbidity scale was developed in four steps: a literature review to assess the existence of other scales and define the scale content; consensus meetings with five CLL experts to determine the content, format and weighting factor for each co-morbidity; a pilot study of 10 CLL patients to assess the scale feasibility; a meeting to agree the final version RESULTS: The literature review did not identify any existing scale but six papers related to assessment of CLL co-morbidities were selected. These allowed the initial selection of 21 diseases. In the expert meeting, the scale was reduced to 13 co-morbidities, with major impact on CLL treatment selection. Some additional variables (age, patient dependence and ECOG performance status) were added. Subsequently, the final list of co-morbidities. variables and response options were agreed. The experts applied a weight factor to each co-morbidity, from 1 (minimal importance) to 3 (very important). a pilot study of 10 patients, using an electronic version of the scale, resulted in some changes but confirmed the scale’s feasibility. The final scale includes three sections. Part 1, the Functional Vital Scale, and a global assessment of the patient (not included in the global score). Part 2, the Co-morbidity scale, including 11 co-morbidities weighted from 0 to 2 and scored from 0 (absence) to 2 (severe co-morbidity). Part 3, CLL alerts, is descriptive and includes the presence of splenectomy and hypogammaglobulinemia. CONCLUSIONS: The co-morbidity scale is a tool to support the clinician in the selection of the optimal treatment for CLL patients. Further research is required to validate the scale and assess its benefits. PSY2 THE USE OF ETANERCEPT FOR TREATMENT OF PSORIASIS IN REAL CLINICAL PRACTICE—NON-INTERVENTION OBSERVATIONAL STUDY Dolezal T1, Klimes J2 1 Institute for Health Economics and Technology Assessment, Prague, Czech Republic; 2 Charles University, Faculty of Pharmacy, Hradec Kralove, Czech Republic OBJECTIVES: Biologic treatment for severe forms of psoriasis proved to be effective and safety treatment in clinical trials. However, these drugs comprise a great financial burden. Therefore, we examined etanercept for treatment of severe forms of psoriasis in real clinical practice. METHODS: This was a prospective cohort non-interventional study of phase IV in real clinical practice which included 149 patients enrolled for 6 months. Patients’ data were collected via electronic questionnaire. Clinical data (PASI score and BSA index), direct cost (inpatient and outpatient care, diagnostics, joint replacement etc.) and on QoL (expressed with EQ-5D and DLQI) were collected by A460 dermatologists. Results were evaluated in months 0, 3 and 6 of etanercept treatment. RESULTS: Average patients’ age was 46.74 years (21–75 years), average time from diagnosis was 24.5 years. Occurrence of psoriatic arthritis was 31.7%. 14.6% of patients were work disabled and 16.3% reported incapacity to work with average duration of 44.3 days in previous 6 months. Previous biologic treatment: 46.6% of patients were naive, 16.8% were after previous failure and 36.5% were after successfully finished treatment with biologics within 6 previous months. Within the observation in 0, 3 and 6 months BSA index decreased from 27.12 to 18.64 to 11.47% and PASI score from 13.69 to 8.09 to 4.97. Utility measured according to EQ-5D increased from 0.7674 to 0.8344 and 0.859 and DLQI index decreased from 10.74 to 6.33 and 4.24 in months 0, 3 and 6. Percentage of patients required inpatient care decreased from 25.8 to 5.0 to 0.66%, additionally average length of stay had shortened from 15.9 to 11 days. CONCLUSIONS: Significant clinical effect was observed within 3 and 6 months of etanercept treatment. QoL was increased in vast majority of patients. Treatment with etanercept decreased other costs associated with psoriasis—inpatient care (frequency and length). PSY3 CLINICAL AND ECONOMIC OUTCOMES OF ZICONOTIDE NEUROMODULATION IN ITALIAN CANCER PATIENTS Zaniolo O1, Carsi M1, Bellini R2 1 Adres srl, Torino, Italy; 2Ospedale Santo Spirito—Casale Monferrato, Casale Monferrato, Italy OBJECTIVES: To evaluate the cost/effectiveness of ziconotide for the spinal neuromodulation of cancer pain in real clinical practice, from the third payer point of view. METHODS: Clinical and resources consumption data related to intrathecal neuromodulation for severe complex pain in progressing cancer patients followed by one Italian centre were collected. The observation spanned from implantation to exitus or drop-out. Change in Numeric Rating Scale Pain Intensity (NRSPI) is the primary outcome of the analysis and the base of the evaluation of number of days with controlled pain (at least 30% reduction). Secondary outcome measures are Karnofsky index, Pain Management Index, Edmonton Symptom Assessment Scale score and Bowel Function Index. Collected consumption data include drugs, visits, port maintenance, and pump recharge and amortization. Current Italian prices, real practice acquisition and remuneration costs are applied. RESULTS: Between January 2006 and August 2009, 19 patients received intrathecal ziconotide (N: 8) or morphine (N: 11). Both groups showed a statistically significant reduction in NRSPI score after 1 week (Z:−33,9% vs. M:−10,6%), 2 weeks (Z:−47,9% vs. M:−22,1%), and 3 weeks (Z:− 52,1% vs. M:−26,1%) from implantation, with markedly higher reductions in ziconotide patients, although the difference was statistically significant only for the first time point. Patients receiving ziconotide lived significantly more days with controlled pain (74% vs. 40%; P < 0,05). Secondary endpoints showed similar trends. Average weekly cost is about c240 for ziconotide and c120 for morphine; the main cost driver is pharmaceutical cost (respectively 82% and 65% of the total). Higher ziconotide acquisition costs are partially offset by minor expenses for adjuvant therapies. The incremental cost for one further day with controlled pain resulted of c50. CONCLUSIONS: Ziconotide permits effective treatment of extremely difficult-tomanage pain, with improvement of the tolerability and a mild increment of cost, as compared to intrathecal therapy without ziconotide. PSY4 BIOLOGICAL AGENTS FOR THE TREATMENT OF NAIL PSORIASIS: A SYSTEMATIC REVIEW OF THE LITERATURE Naci H1, Green J2, Prasad M3, Fleurence R4 1 United BioSource Corporation, London, UK; 2United BioSource Corporation, Lexington, MA, USA; 3Merck & Co, Kenilworth, NJ, USA; 4United BioSource Corporation, Bethesda, MD, USA OBJECTIVES: To systematically review the efficacy of biologics for nail psoriasis in plaque psoriasis and psoriatic arthritis. METHODS: MEDLINE and EMBASE were searched using predefined terms: “psoriasis”, “psoriatic arthritis”, “parapsoriasis”, “psoriasis vulgaris”, “infliximab”, “adalimumab”, “alefacept”, “etanercept”, “golimumab”, and “ustekinumab”. Abstracts from ISPOR, AAD, BAD, EADV, WCD, and EULAR from 2006 to 2009 were reviewed. Studies reporting both comparative and non-comparative findings from at least one biologic of interest were included if they reported the nail psoriasis severity index score (NAPSI), resolution, or remission. RESULTS: Thirteen studies were included. Golimumab and infliximab were assessed in randomized controlled trials (RCTs). In a RCT of infliximab, the % reduction (indicating improvement) from baseline NAPSI was significantly greater with infliximab than with placebo at weeks 10 and 24, with the reduction maintained at 50 weeks. a similar response through 24 weeks was observed in a RCT of golimumab, in which reductions from baseline NAPSI increased over time, from 14 to 24 weeks. The third comparative study evaluated adalimumab, etanercept, and infliximab in an observational study, favoring infliximab. The remaining studies reported findings for single-arms and examined infliximab (3 studies), alefacept (3 studies), adalimumab (3 studies), and etanercept (1 study). At follow-up periods ranging from 12 to 72 weeks, these studies reported modest reductions in NAPSI scores for infliximab, etanercept, alefacept, and adalimumab. CONCLUSIONS: Comparative studies demonstrated that golimumab and infliximab appeared to be effective treatment options for nail psoriasis. Infliximab also has evidence supporting long-term efficacy at 50 weeks. Evidence from non-comparative studies supported the findings of the comparative evidence and suggested that there may be moderate nail improvement associated with adalimumab, 13th Euro Abstracts alefacept, and etanercept. No quantitative synthesis of the data could be performed, mainly due to the paucity of comparative evidence. Further studies are warranted to understand the comparative efficacy of biologic treatments for nail psoriasis. PSY5 A MARKOV MODEL TO ESTIMATE THE EFFECT OF CANAKINUMAB ON THE PROGRESSION OF AA AMYLOIDOSIS IN PATIENTS WITH CRYOPYRIN-ASSOCIATED PERIODIC-FEVER SYNDROMES (CAPS) Gauthier A1, Jiang Y1, Roberts D2, Warburton J2, Hawkins PN3, Lachmann HJ3 1 Amaris Consulting UK, London, UK; 2Novartis Pharmaceuticals UK Limited, Surrey, UK; 3 National Amyloidosis Centre, University College London Medical School, Royal Free Hospital, London, UK OBJECTIVES: CAPS is a spectrum of rare inherited autoinflammatory syndromes. It is an ultra-orphan disease affecting approximately 70 patients in the UK. CAPS patients are at increased risk of developing AA amyloidosis (AA) and subsequent end-stage renal failure (ESRF). This study aimed to model the impact of a new treatment, canakinumab, on the progression of AA in patients with CAPS. METHODS: A Markov model was developed which comprised four health states: CAPS, CAPS with AA, CAPS with ESRF and dead. Based on the literature, 25% of CAPS patients are thought to develop AA over their lifetime. The risks of ESRF and death in patients with AA were obtained from published observational studies and life expectancy in CAPS patients with ESRF was based on expert opinion. Excess mortality in CAPS for reasons other than AA was estimated from rheumatoid arthritis, another systemic inflammatory disease. Treatment efficacy for canakinumab was obtained from the results of a pivotal clinical trial. The model was run for a typical adult patient and a typical paediatric patient and assumed that AA was not present at treatment initiation. RESULTS: Assuming a UK CAPS population of 70 patients, if canakinumab was initiated at the age of 4 years, a total of 18 AA and 8 ESRF cases would be avoided, and patients would gain 13.7 years of life on average compared to those not treated with canakinumab. Assuming treatment initiation at the age of 38 years (estimated mean age at diagnosis), a total of 11 AA and 4 ESRF cases would be avoided and patients would gain 10.3 years of life on average. CONCLUSIONS: Based on a model estimating the progression of AA in CAPS, treatment with canakinumab is associated with a gain of 10 to 14 life-years due to a reduced incidence of AA and ESRF. PSY6 THE ICATIBANT OUTCOME SURVEY: A PROSPECTIVE STUDY OF SAFETY AND EFFECTIVENESS OF ICATIBANT IN A ‘REAL WORLD’ SETTING IN PATIENTS WITH HEREDITARY ANGIOEDEMA Maurer M1, Aberer W2, Amat LM3, Karon-Klefbom M4, Malmenäs M4, Bouillet L5, Kanny G6, Martin L7, Coppere B8, Longhurst H9, Fain O10, Launay D11, Gompel A12, Bork K13, Greve J14, Yann O15, Rubio-Sotes M16, Sala A17, Lleonart R18, Hernando de Larramendi C19, Hernandez D20 1 Charité—Universitätsmedizin Berlin, Germany, 2Medical University of Graz, Graz, Austria, 3 Hospital Santa Maria, Lleida, Spain, 4Shire Human Genetic Therapies, Inc, Danderyd, Sweden, 5CHU Grenoble, Grenoble, France, 6CHU Nancy, Nancy, France, 7CHU Angers, Angers, France, 8CHRU Lyon, Lyon, France, 9Barts and the London Hospital, London, UK, 10 Hopital Jean Verdier, Bondy, france, 11CHRU Lille, Lille, france, 12CHU Hotel-Dieu, Paris, France, 13Universitatsklinik, Mainz, Germany, 14Klinik und Poliklinik fur Hals-NasenOhrenheilkunde, Essen, Germany, 15CHU Hopital de la Cote de Nacre, Caen, France, 16 Hospital Universitario Gregorio Maranon, Madrid, Spain, 17Hospital Vall D’Hebron, Barcelona, Spain, 18Hospital Bellvitge, L’Hospitalet de Llobregat, Spain, 19Hospital Marina Baixa, Vila Joiosa, Spain, 20Hospital Le Fe, Valencia, Spain OBJECTIVES: Icatibant, a specific bradykinin-B2 receptor antagonist, is approved for symptomatic treatment of acute attacks of HAE in adults with C1-esterase-inhibitor deficiency. Although no treatment-related serious adverse events were reported during randomized clinical trials with icatibant in HAE, a study of outcomes in normal clinical practice is important since bradykinin-B2 receptor antagonism is a new therapeutic approach. The Icatibant Outcome Survey (IOS) is an international, multicentre, prospective observational study designed to provide further information on the safety and effectiveness of icatibant in an unselected patient population. METHODS: The study population is all patients receiving icatibant outside of the blinded controlled studies. After obtaining approval from local Institutional Review Board or Ethics Committees, treating physicians are invited to enrol consenting patients and to participate in the study. a series of survey forms are used to record baseline data relating to relevant medical history and demographic information at study entry; details of previous HAE attacks, icatibant administration and outcomes; hematological, liver function and plasma lipid status; and any adverse events. Routine data entry occurs electronically via a secure link and participating physicians are encouraged to update records for each patient at 6-monthly intervals. Any serious adverse events will be notified immediately in accordance with normal pharmacovigilance practice. Data harvested from the database will be summarised and scrutinised by an independent Steering Committee. RESULTS: Data collection commenced in July 2009 and the first analysis was done early 2010. CONCLUSIONS: The Icatibant Outcome Survey will provide further prospective data on safety and outcomes of use of this novel drug for acute treatment of HAE attacks in a ‘real world’ setting. 13th Euro Abstracts SYSTEMIC DISORDERS/CONDITIONS – Cost Studies PSY7 AN ECONOMIC EVALUATION OF RECOMBINANT ACTIVATED FACTOR VIIA ROOM TEMPERATURE STABLE IN THE MANAGEMENT OF HEMOPHILIA PATIENTS WITH INHIBITORS IN SERBIA Plun-Favreau J1, Dimitrov DG2, Cel M3, Lukic J4 1 Novo Nordisk, Zurich, Switzerland; 2Novo Nordisk s.r.o., Praha, Czech Republic; 3Novo Nordisk Pharma Sp z.o.o., Warsaw, Poland; 4Novo Nordisk Pharma d.o.o. Beograd, Belgrade, Serbia OBJECTIVES: Recombinant activated factor VIIa room temperature stable (rFVIIa RTS) enables immediate access to treatment, which may lead to more rapid bleeding control and require less product compared to original rFVIIa, leading to cost savings despite the greater cost of rFVIIa RTS. The total annual cost of managing mild/moderate bleeds in one average hemophilia patient with high titre, high responding inhibitors by original rFVIIa and rFVIIa RTS was examined. METHODS: Only main medications costs were compared from the public payer perspective. Resource utilization and clinical outcomes were based on a review of international literature. Excel based budget impact model (BIM) was developed to assess the financial consequences of treating bleeding episodes with rFVIIa compared to current treatment practices. Cohort of individuals in BIM can be followed sequentially from bleed initiation, taking into consideration first-line efficacy, switching to other products, re-bleeds and bleed cessation. RESULTS: Patients with rFVIIa RTS were treated on an outpatient or home basis. First-line and second-line efficacy was assumed to be 92% for original rFVIIa and rFVIIa RTS. An early treatment with rFVIIa was associated with a lower incidence of re-bleeds compared to delayed treatment 5.2% vs. 13.7% and therefore with less product usage 2.1 vs. 2.3 doses per treatment line. Total annual costs per patient from initiation to cessation in the current treatment environment was CSD 12.58 million (c0.12 million). One-way sensitivity analyses showed that at price of rFVIIa RTS from 0% to 16% premium introduction of this new form can deliver savings for the Serbian health care budget due to immediate patients’ access to the treatment. If not literature but current real life treatment patterns are considered savings can reach 56%. CONCLUSIONS: rFVIIa room temperature stable (RTS) in comparison to the original rFVIIa represents cost-saving first-line treatment option for the Serbian health care system. PSY8 BUDGET IMPACT OF THE USE OF HYDROMORPHONE ONCE DAILY IN CHRONIC PAIN PATIENTS IN THE GERMAN HEALTH CARE SYSTEM— AN UPDATE Fleischmann J, Wimmer A, Kubitz N Janssen-Cilag, Neuss, Germany OBJECTIVES: The budget impact of treating patients with severe chronic cancer and non-cancer pain with OROS®hydromorphone once-daily was determined in the German health care system. METHODS: An existing Excel® based hypothetical budget impact model (Fleischmann et al. 2008) calculating the cost consequences of using strong opioids (WHO step III) was updated using new market data (2009) and current prices (2010 public prices). The model has a one year time horizon adopting the perspective of the social health insurance accounting for costs of opioids, breakthrough pain and adverse events. Patient numbers are calculated using data from literature; adverse event rates are based on literature findings. Comparators included sustained-release morphine (twice-daily), controlled-release oxycodone (twice-daily), hydromorphone (twice-daily), transdermal fentanyl and transdermal buprenorphine. Initial prescription share of OROS® hydromorphone was 3.7% (2009). This share was hypothetically extended to 5%. It was assumed that this increase in prescription is gained by switching patients from their previous oral medication to OROS® hydromorphone. Titration and maintenance dosing schemes taken from previous analyses are used to model the switch. Morphine equivalences were chosen according to SmPC. RESULTS: The number of patients treated was estimated to be 810,608 per year. The model predicted that the expansion of OROS® HM would lower the per patient drug cost from c592.28 to c590.88. Due to generic entry drug cost per patient was considerably lower than in the previous version of the model (c679.52). The model also predicts that if the prescription share of OROS® HM increased to 5% the total budget for strong opioids would decrease by c1,135,086. CONCLUSIONS: Our analysis suggests that under current circumstances our finding from 2008 that extending the use of OROS® hydromorphone to treat patients with severe chronic pain will reduce the overall budget spent on strong opioids is still valid. PSY9 THE ADMINISTRATION OF LIQUID INTRAVENOUS IMMUNOGLOBULINS IN SPAIN: A BUDGET IMPACT ANALYSIS TO ESTIMATE COSTS DUE TO AN INCREASING PRIVIGEN PENETRATION Darba J1, Restovic G2, Kaskens L2, de Agustin T3 1 Universitat de Barcelona, Barcelona, Spain; 2BCN Health, Barcelona, Spain; 3CSL Behring, Esplugues de Llobregat, Spain OBJECTIVES: To assess the economic impact for the Spanish health care system of the substitution of Flebogamma by Privigen in the Spanish market setting for liquid intravenous immunoglobulins (IVIGs) in 2009. METHODS: A budget impact model was developed with analyses based on Spanish data on disease prevalence, population forecasts, drug use, health care resource utilization and market shares. Data was obtained by literature research and consulting a panel of local clinical experts. The A461 perspective of the Spanish health care system was considered and a 5-year time horizon was evaluated. All costs referred to 2009 and a discount rate of 3% was applied. Indications of IVIGs included in this study were replacement therapy, immunomodulation and allogeneic bone marrow transplantation. Sub-indications of these primary indications were also considered distinguishing between children and adolescents vs. adults. IGIV treatments considered were Kiovig, Octagamocta, Flebogamma and Privigen. Direct medical annual costs per patient for each (sub-)indication were estimated before and after the substitution by Privigen. RESULTS: The target population for IVIG treatment was estimated to be 5743 in 2009, increasing to 6926 in 2014. Total costs for the next 5 years with the actual market share were estimated at c565 million. a minimum increase in Privigen’s market share from 2% in the first year up to 12% in the fifth year- was considered. Under these circumstances, the estimated costs were c563c million which represents a saving of c2 million for the Spanish health care system. CONCLUSIONS: An increase of Privigen’s market share in Spain is likely to decrease the budget utilization of the health care system within the next 5 years. Savings are the result of less drug and administration costs per treated patient. PSY10 A BUDGET IMPACT MODEL TO INVESTIGATE POTENTIAL COST SAVINGS ASSOCIATED WITH IMPROVEMENTS IN THE SAFETY PROFILE OF STRONG OPIOID ANALGESICS Ruff L1, Plich A1, Liedgens H2 1 Medaxial Group, London, UK; 2Gruenenthal GmbH, Aachen, Germany OBJECTIVES: The use of strong opioid analgesics to treat severe chronic pain is associated with adverse events (AEs) and treatment discontinuations, which can increase the overall cost of treatment. New interventions with more favourable safety profiles could reduce these costs. An economic model was constructed to investigate the potential budget savings associated with the introduction of an intervention with increased tolerability to a formulary. METHODS: A prevalence-based, deterministic budget impact model with a five-year time horizon was developed from the perspective of a UK health care budget-holder. The model takes into account drug acquisition, AE and discontinuation costs associated with the five most frequently-used strong opioid analgesics (WHO step 3) in the UK. Cost and usage data were derived from the British National Formulary, market research studies and physician interviews. AE and discontinuation rates were derived from published studies. RESULTS: In a theoretical population of 100,000 individuals, the model estimated that 205 patients experienced severe chronic pain and subsequently received strong opioid analgesics. The overall annual cost of treating these patients with currently-available strong opioid analgesics was estimated to be £152,426, 62% of which was spent managing AEs and discontinuations. Over five years, the model showed that if 50% of patients received an alternative strong analgesic which reduces the number of AEs and discontinuations by 25% (compared to currently-used strong opioid analgesics), there would be 198 fewer AEs and 58 fewer discontinuations. This would result in a cost saving of £51,985. CONCLUSIONS: The cost of managing AEs and discontinuations is a significant contributor to the overall treatment cost associated with strong opioid analgesics. New treatments with improved safety profiles may reduce the economic burden of managing AEs and discontinuations associated with the use of strong opioid analgesics. This may partially or even completely offset any potential increase in acquisition costs. PSY11 COSTS OF SUPPORTIVE CARE (SC) FOR THE TREATMENT OF MYELODISPLASTIC SYNDROME (MDS) IN BRAZIL: AN ANALYSIS FROM THE PRIVATE PAYER’S PERSPECTIVE Clark O1, Teich V2, Pereira ML3, Faleiros E1, Clark LGO1 1 MedInsight-Evidências, Campinas, Brazil; 2MedInsight-Evidências, Rio de Janeiro, Brazil; 3 Janssen-Cilag, São José dos Campos, Brazil OBJECTIVES: MDS is a rare hematological disease, that affects the production of blood cells. One of the goals of the treatment is to maintain the blood-cell count in near-normal levels. The means to achieve this are done mainly with the use of hematopoietic- growth factors and transfusions. Our objective was to determine the costs of this supportive treatment under the private payers perspective in Brazil, for patients with intermediate risk (INT) MDS. METHODS: We adapted the National Comprehensive Cancer Network (NCCN) for INT MDS decision tree to the reality of the Health care system in Brazil. Then, we calculated the costs for each branch of the tree, according to the local prices. We also simulated the total costs for a cohort of 100 patients, distributed in the branches according to the expected epidemiology. We assumed an horizon of one year of treatment. RESULTS: The mean cost of SC for MDS was US$45,006/ patient/ year. This value can vary from US$33,368 to US$104,210, according to the patients’ characteristics and types of treatments used. Overall, patients requiring the use of immunotherapy, with anti-tymocyte globulins were associated with the highest costs. Those achieving disease stabilization solely with the use of erythropoietin were associated with the lowest costs. CONCLUSIONS: The SC of SMD is associated with a mean cost of US$45,000/patient/year in Brazil. New treatments have the potential to change this scenario if they can diminish the requirements for the costliest supportive procedures. A462 PSY12 REAL-LIFE BUDGET IMPACT (BI) OF PARENTERAL IRON TREATMENT OF IRON DEFICIENCY ANEMIA/SYNDROME (IDA/IDS) IN SWITZERLAND Brock E1, Braunhofer P2, Troxler J3, Schneider H1 1 HealthEcon AG, Basel, Switzerland; 2Vifor Pharma International, Glattbrugg, Switzerland; 3 Vifor AG, Villars-sur-Glane, Switzerland OBJECTIVES: IDA/IDS is common in pregnancy, postpartum, inflammatory bowel disease, chronic kidney disease, chronic heart failure, menorrhagia/hypermenorrhagia, cancer and following surgery. We estimate the BI associated with substituting iron sucrose (standard) with ferric carboxymaltose (new treatment), which allows for the application of higher dosages in a shorter time. The analysis adopted the perspective of the Swiss mandatory health insurance (MHI) covering the above indications and is based on real-life data to verify the hypothetical BI estimated prior to launch. METHODS: Resource use (no. of patients, dosage per application, no. of applications) was based on recent primary data (Polyquest Prescriber Analysis, Anemia Patient Record Study in Switzerland). Personnel costs were estimated using the Swiss Tarmed fee-for-service reimbursement system. Drug costs and costs of materials used were based on official tariffs (Spezialitätenliste, MiGeL). Real-life IMS sales data of both products were used to verify the BI model. RESULTS: Ferric carboxymaltose was associated with cost savings of 30–44% compared to iron sucrose based on costs of CHF 101/ 210/ 420 and CHF 144/ 375/ 721 per 200/ 500/ 1000 mg treatment cycle, respectively. This leads to cost savings of CHF 15–33 million per year to the Swiss MHI across all indications in the first 3 years post-launch, due to reductions in personnel costs. Ferric carboxymaltose was shown to be cost-saving in all indications except dialysis (due to flat-fee reimbursement). Sensitivity analyses showed the amount of cost savings to be sensitive to changes in the number of inpatients (10–20% of total) treated with intravenous iron (due to flat-fee reimbursement). CONCLUSIONS: Treating IDA/IDS involves substantial costs to the Swiss MHI. Substitution of iron sucrose with ferric carboxymaltose may help to reduce these due to reduced personnel costs. This novel type of real-life BIA will be of increasing interest as conditional reimbursement increases. PSY13 COST IMPACT FROM INITIATING PREGABALIN TREATMENT IN PATIENTS WITH REFRACTORY NEUROPATHIC PAIN Myrén KJ1, Löfroth E1, Sandelin R2 1 IMS Health, Stockholm, Sweden; 2Pfizer, Sollentuna, Sweden OBJECTIVES: To compare the health care costs 6 months prior to and 6 months after initiation of pregabalin in difficult-to-treat neuropathic pain (NeP) patients. METHODS: This was a retrospective longitudinal database study in NeP patients from the South-West region of Sweden (1.5 million inhabitants). Individual patient data from the 1st of January 2000 on health care visits (outpatient, inpatient, primary care), costs, mortality and diagnoses were included. Data from the Swedish Prescribed Drug register were included from July 1, 2005 until December 31, 2007. Difficult-totreat NeP was defined as patients with a NeP diagnosis in 2006, who had had 2 prescriptions of at least three pain medications during one year from the index diagnosis date. The patients should also have had two or more prescriptions of pregabalin preceded by at least a six months pregabalin naïve period. RESULTS: A total of 462 difficult-to-treat NeP patients met the above criteria and were included in the analyses. There was a statistically significant reduction in NeP related costs (visits registered with a NeP diagnosis) after initiation of pregabalin (p = 0.0042 Mann-Whitney). The mean per patient NeP related costs were SEK17,684 (c1,845) 6 months before and SEK10,642 (c1,110) 6 months after pregabalin initiation. The mean non-NeP related costs before treatment (SEK46,095; c4,809) did not differ significantly from the non-NeP costs after treatment (SEK 51,632; c5,387), p = 0.8016. The number of NeP-related in-patient visits, primary care visits and the number of days in hospital decreased significantly (p = 0.0475, p = 0.0129, p = 0.0179, respectively) after treatment with pregabalin. CONCLUSIONS: Initiation of pregabalin significantly reduced the NeP related health care costs in the 6-month period following initiation. Of note was the non significant difference in non-NeP related costs, probably reflecting the on-going costs associated with management of the patients’ concurrent conditions, eg, diabetes. PSY14 REAL-WORLD USE OF DULOXETINE FOR CHRONIC LOW BACK PAIN: TREATMENT PATTERN AND COSTS Ivanova JI1, Birnbaum HG2, Kantor E2, Schiller M2, Swindle R3 1 Analysis Group, Inc., New York, NY, USA; 2Analysis Group, Inc., Boston, MA, USA; 3Eli Lilly & Company, Indianapolis, IN, USA OBJECTIVES: Examine the real-world role of duloxetine versus other treatment for chronic low back pain (CLBP). METHODS: Study sample was selected from a U.S. privately-insured claims database (2004–2008). Selection criteria: ages 18–64 years, had a low back pain (LBP) diagnosis (per HEDIS specifications) with a subsequent CLBP-qualifying diagnosis recorded 90 days or more after the initial LBP diagnosis. Duloxetine-treated patients had ≥1 duloxetine prescription within 6 months after CLBP diagnosis, no prior duloxetine claim, and continuous eligibility ≥12 months before first LBP diagnosis and ≥6 months after index duloxetine prescription (study period). 553 duloxetine-treated patients were matched to a total of 553 control patients who initiated another non-surgical LBP treatment based on propensity score and time from first LBP diagnosis to treatment initiation. a subset (n = 103 each) of matched employees was also analyzed. McNemar tests were used to compare LBP 13th Euro Abstracts treatment rates. Bias-corrected bootstrapping was used to compare direct (medical and drug) costs from third-party payer perspective and employee indirect (workloss) costs. RESULTS: During the 6-month study period, matched duloxetine-treated patients had significantly lower rates of other pharmacological therapy than controls (e.g., 56.2% vs. 64.9% narcotic opioids, p = 0.002; 34.9% vs. 49.5% NSAIDs; P < 0.001) and non-invasive therapy (28.8% vs. 38.5% chiropractic therapy; 25.5% vs. 35.4% physical therapy; 17.5% vs. 28.4% exercise therapy; all P < 0.001). Duloxetine-treated patients versus controls had numerically lower back surgery rates (2.2% vs. 3.8%, p = 0.117) and similar direct costs ($7658 vs. $7439, p = 0.812). Among CLBP employees, duloxetine-treated employees versus controls had lower rates of other non-surgical therapy, numerically lower back surgery rates (0% vs. 3.9%, p = 0.125), lower total direct and indirect costs ($5227 vs. $7229, p = 0.042), and numerically lower indirect costs ($1806 vs. $2664, p = 0.053). CONCLUSIONS: Duloxetine treatment in CLBP patients/employees versus other non-surgical treatment was associated with reduced rates of non-surgical therapies and numerically lower surgery rates without increased costs. PSY15 CONSTIPATION RELATED TO OPIOID THERAPY: A COST-OF-ILLNESS AND PREVALENCE DATABASE STUDY IN THE BRAZILIAN PRIVATE SETTING Takemoto ML1, Fernandes RA1, Passos RB1, Colombini-Netto M2, Bertola-Neto A2, Almeida GR3, Monteiro R3 1 ANOVA—Knowledge Translation, Rio de Janeiro, RJ, Brazil; 2Axismed Gestão Preventiva da Saúde, São Paulo, SP, Brazil; 3Pfizer, São Paulo, SP, Brazil OBJECTIVES: To estimate the prevalence of constipation concomitant to opioid treatment and related resource utilization and costs from the private payer perspective. METHODS: Patients on opioid therapy were identified from a longitudinal insurance claims database consisting in 1,057,033 individuals for a period of 35 months. An algorithm was used to identify patients on opioid therapy with coincident constipation-related claims according to ICD-10 codes, targeted procedures and opioid use criteria. Resource utilization and costs were determined for these individuals and compared with patients on opioid therapy without constipation, without opioid therapy with constipation and without both conditions. Results were compared using ANOVA with a significance level of 0.05 and are presented per individual per month. RESULTS: A total of 23,313 patients were classified as opioid treated patients (2.21% of total population) and 6,678 had events related to constipation (29.03% of the opioid population). Compared with opioid treated patients without constipation, incremental mean total costs per month per patients with the condition were 261.18 BRL (P < 0.001). The average cost per month for opioid-related constipation patients was 787.84 BRL, significantly higher than patients on opioid therapy without constipation (526.66 BRL), with no opioid therapy but constipated (284.47 BRL) and without both (90.17 BRL) (P < 0.001 for all comparisons). Patients with claims related to both conditions had significantly more days in hospital per month (0.25 vs. 0.497, P < 0.001), outpatient office visits (1.04 vs. 1.59, P < 0.001), outpatient procedures (4.69 vs. 14.05, P < 0.001) and tests and therapies (31.95 vs. 36.66, P < 0.001) than did patients without opioid-related constipation claims. CONCLUSIONS: The economic burden of patients with constipation events coincident with opioid treatment is significantly higher when compared to all other groups. Constipated patients without opioid therapy had also higher costs than those free of both conditions. These results indicate that reducing opioid-induced constipation has potential cost savings for the health care system. PSY16 CHANGES IN SELF-REPORTED PAIN SCORES ASSOCIATED WITH DULOXETINE VERSUS OTHER ANTIDEPRESSANTS AMONG PATIENTS WITH MAJOR DEPRESSIVE DISORDER IN THE UNITED STATES VETERANS AFFAIRS HEALTH CARE NETWORK Liu J1, Zhao Y2, Happich M3, Shi L1 1 Tulane University, New Orleans, LA, USA; 2Eli Lilly & Company, Indianapolis, IN, USA; 3Eli Lilly & Company, Windlesham, UK OBJECTIVES: This study compared self-reported pain scores among patients with major depressive disorder (MDD) in the U.S. Veterans Affairs (VA) health system treated with duloxetine versus other antidepressants. METHODS: The electronic medical records between October 1, 2004 and October 31, 2008 were obtained from the VA Veterans Integrated Service Network 16 data warehouse. All patients treated with either duloxetine monotherapy or other antidepressants (non-duloxetine) were selected. The first dispense date of the index agent was defined as the index date. All patients must have: 1) 1 + prior MDD diagnosis (ICD-9-CM: 296.2 or 296.3); 2) no prior diabetes (ICD-9-CM: 250.xx) or bipolar (ICD-9-CM: 296.4x-296.8x) diagnosis; and 3) self-reported pain score measured within 60 days both before the index date (baseline pain score) and after the last dispense date of the index antidepressants during the 12-month post-index period. The non-duloxetine-treated patients were matched to the duloxetine-treated patients via propensity scoring (1:1 ratio), controlling for demographics, comorbidities, prior opioid use, prior health care utilization, and baseline pain scores. Opioid utilization and pain scores over the 12-month postindex period were examined between cohorts. RESULTS: The study sample included 210 duloxetine- and 210 non-duloxetine-treated patients. Significantly less duloxetinetreated patients than non-duloxetine-treated patients used opioids (20.1%; vs. 34.3%, p = .002) over the 12-month post-index period. Both cohorts had similar morphineequivalent opioid daily use and pain scores during the follow-up period. Controlling for baseline pain scores and medication duration, duloxetine-treated patients had 1.58 13th Euro Abstracts points (p = 0.007) lower pain score than non-duloxetine-treated patients among individuals who had any opioid use during the 12-month post-index period. CONCLUSIONS: Among similar VA patients with MDD, patients treated with duloxetine were associated with lower opioid use than those treated with other antidepressants. Among individuals with any opioid use over the 12-month post-index period, duloxetinetreated patients had better pain scores than non-duloxetine-treated patients. PSY17 ECONOMIC OUTCOMES ASSOCIATED WITH HYDROXYUREA ADHERENCE AMONG PEDIATRIC MEDICAID ENROLLEES WITH SICKLE CELL DISEASE Candrilli S1, O’Brien SH2, Balkrishnan R3 1 RTI Health Solutions, Research Triangle Park, NC, USA; 2The Research Institute at Nationwide Children’s Hospital, Columbus, OH, USA; 3University of Michigan, College of Pharmacy, Ann Arbor, MI, USA OBJECTIVES: Although not approved for use in pediatrics, hydroxyurea has been shown to decrease painful vaso-occlusive events and hospitalizations (HU) in children with sickle cell disease (SCD), a genetic hematologic disorder primarily affecting people of African descent. However, few data exist on the use of HU, and adherence to treatment, in real-world settings. This study assessed the extent to which children with SCD are adherent with prescribed HU therapy and the association between HU adherence and economic outcomes in this population in a real-world setting. METHODS: Insurance claims of North Carolina Medicaid program enrollees (June 1999-August 2008) were analyzed. Inclusion criteria were ≥1 claims with a diagnosis for SCD (i.e., ICD-9-CM 282.6, 282.6x), ≥2 HU prescriptions in the year following HU initiation, ≤18 years old at the time of HU initiation, and continuous Medicaid enrollment for ≥12 months prior to and following HU initiation. Adherence was measured using the medication possession ratio (MPR), with MPR > 0.8 considered adherent. Multivariable models were estimated to assess the association between HU adherence and economic outcomes (e.g., inpatient costs) in the first year of HU therapy. RESULTS: A totla of 159 subjects (55% female, mean age [SD] 11.3 [4.4] years) met all inclusion criteria. The mean MPR was 0.63, with 41% of subjects adherent. Multivariable models revealed that in the year following HU initiation, adherence was associated with a reduction in all-cause (-$3.695, p = 0.0003) and SCD-related (-$3.002, P < 0.0001) total, inpatient (-$2172, p = 0.001; -$1815, P < 0.0001, respectively), ER (-$314, P < 0.0001; -$202, p = 0.002, respectively), outpatient (-$951, P < 0.0001; -$541, p = 0.005, respectively), and vaso-occlusive event-related (-$3639, P < 0.0001) costs, and an increase in all-cause (+$239, p = 0.014) and SCD-related pharmacy (+$198, P < 0.0001) costs. CONCLUSIONS: Adherence to HU among pediatric SCD patients may be suboptimal and improving adherence to HU therapy among these patients may reduce the economic burden of the illness. PSY18 DIRECT MEDICAL COSTS OF LIQUID INTRAVENOUS IMMUNOGLOBULINS IN CHILDREN, ADOLESCENTS AND ADULTS IN SPAIN Darba J1, Restovic G2, Kaskens L2, de Agustin T3 1 Universitat de Barcelona, Barcelona, Spain; 2BCN Health, Barcelona, Spain; 3CSL Behring, Esplugues de Llobregat, Spain OBJECTIVES: The aim of this study is to determine health care resource utilization and direct medical costs in patients treated with intravenous immunoglobulins (IVIGs) in Spain in 2009. METHODS: A cost-of-illness analyses was performed to estimate direct medical annual costs of patients treated with IVIGs. Prevalence data was obtained from the Spanish Primary Immunodeficiency registry. a semi-structured questionnaire was used to collect data on health care resource utilization and target population distribution. Inpatient and outpatient costs were considered from the perspective of the public health care system. Hospital costs considered were ambulatory-, ward-, premedication-, drug- and administration costs. All costs referred to 2009. In medical practice, IVIG doses depend on a patient’s weight and age. Therefore separate analyses were conducted for children and adolescents vs. adults. RESULTS: Patients with replacement therapy were estimated to be 4192 (724 children and adolescents vs. 3450 adults) for 2009. Patients with immunomodulation were estimated to be 1378 (244 children and adolescents vs. 1134 adults) and patients with an allogeneic bone marrow transplantation were estimated at 172 (30 children and adolescents vs. 142 adults ). Mean annual costs for children and adolescents were c6283 with Privigen, c6292 with Kiovig, c6946 with Flebogamma and c6559 with Octagamocta. For adults estimations were c17,078 with Privigen, c17,103 with Kiovig, c18,095 with Flebogamma and c17,423 with Octagamocta. CONCLUSIONS: Direct medical costs for IVIGs were estimated to be c6.9 million for children and adolescents and c84.8 million for adult patients. Drug costs represented 94% of the total direct medical costs. In the context of limited health care resources, the choice for a certain IVIG treatment depends on individual patient characteristics and cost considerations. PSY19 PREDICTORS OF DULOXETINE TREATMENT INITIATION AMONG PATIENTS WITH DIABETIC PERIPHERAL NEUROPATHIC PAIN IN SOUTH CENTER VETERANS AFFAIRS HEALTH CARE NETWORK IN THE UNITED STATES Shi L1, Liu J1, Zhao Y1, Zhao Y2 1 Tulane University, New Orleans, LA, USA; 2Eli Lilly & Company, Indianapolis, IN, USA OBJECTIVES: To explore predictors of duloxetine initiation versus other standard of care (SOC) treatments for patients with diabetic peripheral neuropathic pain (DPNP) A463 in the Veterans Affairs (VA) health system. Duloxetine was not on the VA national drug formulary. METHODS: The electronic medical records from January 2004 to December 2008 were requested from the Veterans Integrated Service Network 16 data warehouse. All patients selected initiated either duloxetine or SOC (tricyclic antidepressants [TCAs], venlafaxine, gabapentin, pregabalin) over the study period, with the first dispense date of the index agent as the index date. All patients must have at least 1 prior DPNP diagnosis (ICD-9-CM: 250.6x or 357.2), but no prior depression (ICD9-CM: 296.2, 296.3, 300.4, 309.1, 311.0), fibromyalgia (ICD-9-CM: 729.1), or neuralgia (ICD-9-CM: 729.2) diagnosis. Logistic regressions were used to examine the predictors of initiating duloxetine versus SOC controlling for demographics, comorbidities, prior pain level, and prior SOC or opioid use. RESULTS: The analytical sample included 2675 (215 duloxetine and 2460 SOC) patients. Prior use of gabapentin (odds ratio [OR] = 13.66, 95% confidence interval [CI]: 9.70–19.24, TCAs (OR = 5.40, 95% CI: 3.73–7.82), or venlafaxine (OR = 3.67, 95% CI: 1.67–8.06) was strongly associated with duloxetine initiation. Patients comorbid with anxiety, cerebrovascular disease, or substance abuse were 1.08 (95% CI: 1.40–3.08), 0.44 (95% CI: 1.01–2.07), and 1.11 (95% CI: 1.10–4.03) times more likely to initiate duloxetine, respectively. Prior opioid users were 1.47 (95% CI: 1.02–2.12) times as likely to initiate duloxetine as those with no prior opioid use. Patients with self-reported severe pain were 1.66 (95% CI: 1.11–2.50) times as likely to initiate duloxetine as those with no reported pain. CONCLUSIONS: DPNP patients in the VA health care system who initiated duloxetine appeared to have prior SOC use, more comorbid conditions, prior substance abuse or opioid use, and higher pain level. PSY20 SYSTEMIC LUPUS ERYTHEMATOSUS (SLE) IS COSTLY FOR THE HEALTH CARE SYSTEM AND WORKPLACE: A SYSTEMATIC LITERATURE REVIEW Doria A1, Grandfils N2, Maurel F2, Rudge H3 1 University of Padova, Padova, Italy; 2IMS Health, Paris, France; 3GlaxoSmithKline, Uxbridge, Middlesex, UK OBJECTIVES: To summarise the state of knowledge of the economic impact of SLE and to gather information to implement a European study on costs of SLE care. METHODS: A systematic review of the literature was undertaken in the ECONLIT, EMBASE, Medline and EMBR databases for full papers relating to the objectives. Key words included SLE, cost, resource, economic, productivity, absenteeism, employment and work disability. Papers were excluded if they were not in English, French, Spanish, Italian or German language. RESULTS: Thirty relevant research papers (relating to 13 different studies) were included. These studies were conducted in Canada, China, Germany, UK and US and used retrospective and prospective methods. Data were mainly obtained from the hospital care setting over a one-year study period. The main resources reported were physician and emergency visits, as well as inpatient hospital stays (of which SLE flares and infections were the main causes). Annual direct costs (converted to December 2009 Euros) ranged from c2,879 (Canada) to c14,873 (US), with indirect costs of c1,214 (US/UK/Canada, friction method) to c45,668 (US, labour market activity method). Indirect costs accounted for at least 38% of the total costs when both market and non-market work was valued. Higher disease activity was reported as a predictor of higher direct and indirect costs. Greater damage, more flares and nephritis were also found to predict higher direct costs whilst worse physical and/ or mental status predicted higher indirect costs. Only one study focussed on the cost of flares. CONCLUSIONS: SLE is costly for the health care system and workplace. However, there is a lack of data on the cost of SLE flares and little recent European data relating to the cost of SLE care. HGS/GSK funded. PSY21 THE IMPACT OF COMORBIDITY ON INFLAMMATORY BOWEL DISEASE HEALTH COSTS Lepore V1, D’Ettorre A1, Graziano G1, Martinelli D2, Tafuri S3, Tognoni G1 1 Consorzio Mario Negri Sud, Santa Maria Imbaro, Chieti, Italy; 2Apulia Regional Epidemiological Observatory, Foggia, Italy; 3Apulia Regional Epidemiological Observatory, Bari, Italy OBJECTIVES: To quantify the direct costs (hospitalizations and prescription drugs) of the caring strategies for the Inflammatory Bowel Disease (IBD). METHODS: All cases of the major types of IBD, Crohn’s disease (CD) and ulcerative colitis (UC), occurred in Apulia between 2002 and 2006 were in the study. The patients were identified through a record linkage procedure among three different regional health databases. Multivariate Poisson regression accounting for over-dispersion was used to assess the average annual direct costs per person. Results were reported, separately for the two diseases, as Incidence Rate Ratios (IRR) with 95% confidence intervals. RESULTS: The cohort consisted of 2,700 cases of CD and 4902 cases of UC. We estimated the average annual costs per person in patients with CD to be equal to 7,153 Euros for hospitalizations and to c2107 for prescription drugs. In patients with UC the average annual costs per person was of c6317 and c2277 for hospitalizations and prescription drugs, respectively. Surgical to medical DRGs ratio for both diseases was 1:5 and the ATC codes relative to intestinal anti-inflammatory, proton pump inhibition, anti-inflammatory/anti-rheumatic and corticosteroids products, represented the 40% of all prescriptions. Overall, the analyses showed, both for DC and CU, that the Charlson Comorbidity Index (CCI) was the only variable associated with the direct costs i.e. a significant increase in the costs (p-value < 0.0001) was found according to the severity of the CCI. In particular, it was noted that a worse clinical condition before disease onset (CCI before index date) was predictive of larger costs especially A464 in the case of UC. CONCLUSIONS: These preliminary results show that this innovative cost estimation, obtained with this carefully controlled use of population wide databases, is a valuable alternative to traditional analyses obtained with ad hoc designed and less representative protocols. PSY22 INDIRECT COSTS OF OBESITY IN POLAND Warowny M, Macioch T, Hermanowski T Medical University of Warsaw, Warsaw, Poland OBJECTIVES: Obesity is becoming to be treated like a worldwide epidemic. In Poland about 10.6% employees are obese. The objective of the studies was to estimate the indirect costs of obesity among Polish society. METHODS: Human Capital Approach method was used in costs quantifying. Data were collected from obese Polish employees. Work Productivity and Activity Impairment General Health questionnaire was used to estimate absenteeism and presenteeism in obese population. Indirect cost for general population was calculated on the basis of gross value added per employee in 2008. Central Statistical Office (GUS) data were used to identify obese epidemiology in employed population. Values are presented in Euro (exchange rate: 1 EUR = 4.10 PLN). RESULTS: Data from 96 people were analyzed (mean age = 41.7 years, 34.4% men, average BMI = 34.2 kg/m2). Overall work impairment due to health problems in questioned population was estimated at 36.3%, with 11.8% of work time missed due to health problems. Taking into consideration that based on GUS data near 1.5 million employees are obese total indirect costs of obesity in Poland in the year 2008 were estimated at c10,5 billion representing 0.9% of gross domestic product. Absenteeism costs accounted for less than 1/3 of this costs (c3.4 billion) while pressenteism costs were estimated at amount of c7.0 billion. We didn’t find any correlation between BMI score and work impairment due to health problems (Pearson r = 0,15). CONCLUSIONS: Previously estimated direct medical costs of obesity (without obesity related diseases) covered by public payer were quantified at 4 million EUR in the year 2008. We’ve found that indirect costs of lost productivity due to obesity are substantial to polish economy. However we conclude that not obesity itself but obesity related diseases generate most of indirect costs. PSY23 COST OF POMPE DISEASE IN POLAND IN 2008 AND 2009 Baran A, Czech M, Hermanowski T Medical University of Warsaw, Warsaw, Poland OBJECTIVES: The objective of this review is to estimate the direct and indirect costs of Pompe disease. METHODS: Direct and indirect cots were estimated based on a questionnaire (consisting of 108 questions) measuring costs specific for Pompe disease and a survey concerning the enzyme replacement therapy created for the research. The direct costs of Pompe disease were estimated from the patient`s social and the public health care system payer (National Health Fund) perspectives. While estimating indirect costs human capital approach methodology was used taking into consideration absence from work due to the illness, lost time and salaries of the members of the family taking care of the patients. The analysis was done in a 2-year time span. RESULTS: In the research 80% of Pompe disease population were studied (N = 16). Indirect costs constituted 3% of the total cost (c17,959,67 per patient in the period of 2 years). The highest component of the amount was the cost of the income lost by the members of the families taking care of the patients; 96% of the total direct costs were the direct medical costs (c550,975.03 per patient in the period of 2 years). The main direct medical cost determinant is the enzyme therapy (c543,350.47)-constituting 99% of the total direct medical costs. CONCLUSIONS: Providing patients with the prompt access to the enzyme replacement therapy result in lowering future indirect and direct costs: smaller number of patients taking disability pension, lower cost of medical equipment used by patients and higher productivity of patients able to work. Additionally there is a need to introduce an unusual and untypical approach in Health Technology Assessment. PSY24 COST OF CARE FOR CHRONIC MYELOID LEUKEMIA (CML) IN PATIENTS EXPERIENCING RESISTANCE AND/OR INTOLERANCE TO IMATINIB FROM THE PUBLIC HEALTH SYSTEM PERSPECTIVE IN MEXICO Hernández-Rivera G1, Aguayo Á2, Cantu-Rodríguez O3, Cervera E4, Gomez-Almaguer D3, Gutiérrez H3, Lopez-Hernández M5, Martínez D2, Oropeza P6, Rico E7, Donato BM8, Juarez-Garcia A1, Vargas-Valencia J9 1 Bristol-Myers Squibb, D.F., Mexico; 2Instituto Nacional de Ciencias Médicas y Nutrición Salvador Zubirán, Mexico, D.F., Mexico; 3Hospital Universitario Dr. José Eleuterio González, Monterrey, NL, Mexico; 4Instituto Nacional de Cancerología, Mexico, D.F., Mexico; 5Centro Médico Nacional 20 de Noviembre ISSSTE, Mexico, D.F., Mexico; 6Hospital General de Zona 8 IMSS, Mexico, D.F., Mexico; 7Hospital General Regional 110 IMSS, Guadalajara, JAL, Mexico; 8Bristol-Myers Squibb, Wallingford, CT, USA; 9Econopharma Consulting S. A. de C. V., Mexico City, Mexico, Mexico OBJECTIVES: To estimate the cost of care for CML in patients with resistance and/ or intolerance to imatinib from the public perspective in Mexico METHODS: Retrospective study of 208 patients diagnosed with CML, ≥18 years of age, treated with imatinib and having at least 12 months follow-up with final contact in the past three months, whom were attended at 6 tertiary level public hospitals in Mexico. Cost of care data captured includes drug costs (chemotherapy and imatinib), inpatient visits, toxicity treatment, and medical monitoring (labs studies, outpatient visits and other 13th Euro Abstracts drugs costs). Costs are estimated on a monthly basis and are classified according to the presence of failure (toxicity or resistance to imatinib). Descriptive statistics on the use of resources are reported. RESULTS: At diagnosis of CML, 95% of patients were in chronic phase and 5% in accelerated phase. Fifty-five percent were female, with a mean age of 48.30. Defining treatment failure as any imatinib dose adjustment, increasing if resistance (lack of hematological and/or cytogenetic response) or reduction if intolerance (toxicity), 84.7% of patients failed initial treatment with imatinib. The median time to imatnib dose adjustment was 5.1 months. The average monthly cost of diagnosis and treatment prior to receiving imatinib was estimated US$2,210.54 ($2461– $2808); the average monthly cost during treatment of imatinib and prior to failure was estimated US$2793 ($1612–$3125). The average monthly cost after failure increased to US$4706.18 ($3642–$5770), which represents an increment of 68.5% (t-test P < 0.001). The cost drivers of the increase are primarily: 1) inpatient visits, which increase from US$180 pre-failure to USD$367 post-failure, and 2) resources used in medical monitoring and CML treatment, which increase from US$3678 pre-failure to US$4338 postfailure per month. CONCLUSIONS: Once patients are resistant or intolerant to imatinib, their cost of treatment increases through additional demand for medical resources at Mexican public health care institutions. PSY25 A PHARMACOECONOMIC EVALUATION OF ROMIPLOSTIM (NPLATE) FOR THE TREATMENT OF CHRONIC IMMUNE THROMBOCYTOPENIA (ITP) Brosa M1, Arocho R2, Gutierrez L2, Kutikova L3 1 Oblikue Consulting, Barcelona, Spain; 2Amgen, Barcelona, Spain; 3Amgen (Europe) GmbH, Zug, Switzerland OBJECTIVES: To assess treatment costs per overall platelet response with romiplostim+concurrent treatment vs. placebo+concurrent treatment in chronic adult ITP, from a national payer perspective in Spain. METHODS: Platelet response was defined as a platelet count of ≥50×109/L. Overall platelet response rates to romiplostim and placebo were used for the analysis and derived from two parallel clinical trials in splenectomized and non-splenectomized ITP patients [1]. All patients were allowed to enter on concurrent ITP medication (danazol, corticosteroids, azathioprine) and receive rescue medication (eg, intravenous immunoglogulins). Treatment costs were calculated over the 24-week clinical trial period and included intervention, rescue medication and management of bleedingrelated events. Costs were based on the 2009 national reimbursement list. Mean treatment cost per response was calculated for overall population, and for splenectomized and nonsplenectomized patients. RESULTS: Cost per response was substantially lower for romiplostim compared to placebo. Overall response rate was 83% for romiplostim and 7% for placebo (splenectomized patients 79% vs. 0%, non-splenectomized patients 88% vs. 14%, respectively). Mean treatment costs were c15,781 for romiplostim and c8,111 for placebo (splenectomized patients c15,436 vs. c10,263, non-splenectomized patients c16,125 vs. c5,958, respectively). Cost per response with romiplostim was c19,013 compared to c115,871 with placebo (splenectomized patients c19,539 vs. infinite cost/response, nonsplenectomized patients c18,324 vs. c42,557, respectively). The main cost-offsets were due to reduced immunoglobulin rescue use. CONCLUSIONS: Romiplostim represents an efficient use of health care resources in both splenectomized and non-splenectomized ITP patients for the Spanish health care system, leading to a significant improvement in managing a disease with a limited number of existing effective therapies. [1] Kuter et al. Lancet 2008;371:395–403. PSY26 A RETROSPECTIVE STUDY TO EVALUATE THE MORBIDITY PROFILES AND THERAPEUTIC COST OF PATIENTS WITH CHRONIC PAIN IN URBAN AREA Sicras-Mainar A1, Navarro-Artieda R2, Villoria-Morillo J3, Esquivias-Escobar A4 1 Directorate of Planning, Badalona, Barcelona, Spain; 2Hospital Germans Trias i Pujol, CIBEREHD, Barcelona, Spain; 3Medicxact, Alpedrete, Madrid, Spain; 4Grünenthal Pharma SA, Madrid, Spain OBJECTIVES: Chronic pain is gaining importance as a major cost factor in health care. To gather information about analgesic pharmacotherapy of patients with chronic pain and perform cost estimations to guide future cost-effectiveness research in the area. METHODS: Data from patients aged above 44 years suffering from any chronic condition and receiving regular analgesic pharmacotherapy (≥6 months) who attended health care facilities within the area of Badalona during 2008 were collected in a retrospective study. Morbidity profiles were defined according to treatment setting (pain unit, hospital), WHO analgesic stage (1–2 versus 3), and a raw cost model based on resource use and work absenteeism was applied. Patients attending the pain unit or the hospital were considered undertreated if they were on stage 1–2 analgesics. Multiple regression was used to compare costs between undertreated and non-undertreated patients among those attending the pain unit or the hospital. RESULTS: Just 410 out of 18,157 patients ascertained (2.3%) were on stage 3 analgesics. Direct health care costs were greater in patients on Step 3 analgesics (mean [SD]: 5,505.6 [5,046.4] ) than in patients on Steps 1 and/or 2 analgesics (c2,407.4 [2,436.2]), but not indirect costs (c258.5 [1,578.4] vs. c279.5 [1,423.6], respectively). Of patients attended in the pain unit and the hospital, 2.3% and 20.1%, respectively, were considered undertreated. Regression analyses revealed even greater costs in the subgroup of undertreated patients. CONCLUSIONS: Conclusion: Stage 3 analgesics are barely used. Up to one-fifth of patients may be undertreated, and prompted greater costs than those judged as properly treated. Regression analyses did not clarify what proportion of their cost excess was attributable to undertreatment. 13th Euro Abstracts PSY27 A HEALTH ECONOMIC EVALUATION OF THE SWITCH OF ERYTHROPOIESIS-STIMULATING AGENTS (ESAS) IN PATIENTS WITH RENAL FAILURE UNDERGOING HEMODIALYSIS Chevalier P1, Lamotte M1, Leenaerts P2 1 IMS Health Consulting, Brussels, Belgium; 2Ziekenhuis Oost-Limburg, Genk, Limburg, Belgium OBJECTIVES: Different erythropoiesis-stimulating agents (ESAs) are used to correct anaemia in hemodialyzed patients. There is conflicting evidence on whether the dose needed to obtain the same hemoglobin level differs between ESAs. This study assessed the changes in ESA dose and the differences in cost of treatment when patients were switched from epoetin beta (EB) to darbepoetin alfa (DA). METHODS: A monocentre retrospective chart review was performed. All hemodialysis patients switched from EB to DA during the first week of January 2008. The dose conversion from EB to used was 200 IU:1 µg based on the European DA label. In patients dialyzed for at least 9 months at conversion, ESA dose and hemoglobin level were collected 3 months prior (baseline) to 12 months after conversion. The ratio “average weekly ESA dose/ average weekly hemoglobin level”, the average weekly ESA-drug-related cost per patient (Belgian public payer’s perspective), the average weekly dose of ESA and the average weekly hemoglobin were compared for the baseline period vs. weeks 14–26 and 40–52 post-switch using the Wilcoxon signed rank test. RESULTS: Out of 144 screened patients, 96 were eligible and survived without renal transplantation until week 52. The ratio “average weekly ESA dose/average weekly hemoglobin level” decreased from 611 IU/g/dL to respectively 359 IU/g/dL and 348 IU/g/dL in the defined post-switch periods (P < 0.0001), without any change in average hemoglobin. a substantial reduction in drug cost was observed after switch (from c66.8 to c53.6 and c52.8/week). The average weekly ESA dose decreased from baseline to post-switch from 7230 IU to 4203 IU and 4137 IU (P < 0.0001). Both ESA dosing and hemoglobin were stabilised after a 13-weeks titration period post-switch. CONCLUSIONS: Switching patients from EB to DA resulted in a significant decrease of the ESA dose needed to obtain the same hemoglobin level. This significant decrease in dosage resulted in a substantial decrease in drug costs. PSY28 COST OF INHIBITOR DEVELOPMENT IN PATIENTS WITH SEVERE HEMOPHILIA A IN SPAIN Lucía JF1, Romero JA2, Febrer L3, Trabal I3, Sabater J4, Lindner L4 1 Hospital Miguel Servet, Zaragoza, Aragón, Spain; 2Hospital Universitario La Paz, Madrid, Madrid, Spain; 3Bayer Healthcare, Barcelona, Spain; 4IMS Health, Barcelona, Catalunya, Spain OBJECTIVES: Risk and consequences of inhibitor (antibodies) development in patients with hemophilia are the main safety issue in patients treated with recombinant drugs, due to its impact on health and quality of patient’s life. The objective of this study is to quantify the economical impact of treating A-hemophiliac patients developing inhibitors against factor VIII (FVIII) for the National Health System (NHS) in Spain. METHODS: An economical model was built as a decision tree allowing assigning the resource use to handle inhibitor development and its associated cost in different patient groups according to its age, treatment pattern and response profile. Data was obtained from a literature review and validated by an experts’ panel. a one-way sensitivity analysis was performed to check for results robustness. RESULTS: The mean annual cost per patient suffering from severe hemophilia a developing inhibitors against FVIII in Spain was c567,518 (EUR 2009), 99% due to pharmacological costs. Results show an important variability, from c166,845 to c2,408,486 depending on the type of patient: its age related to its weight and adequate treatment dosage, its inhibitors’ titer (low or high-titer) and its treatment (on demand or to eradicate inhibitors). The number of bleeding episodes was the variable with the greatest impact on results. According to population data and illness incidence, a 5% increase on the risk of inhibitor development implies a new case per year in Spanish population and, therefore, an expenditure of more than half a million Euro for the NHS. CONCLUSIONS: This study analyzes for the first time the cost of inhibitors’ development among patients with severe hemophilia a treated with recombinant FVIII from the perspective of the Spanish NHS. Results show that treatments to overcome this safety issue represent an important economical burden, so that strategies preventing from inhibitor development should be implemented in this population. PSY29 DARBEPOETIN ALFA VERSUS EPOETIN ALFA FOR TREATMENT OF CHEMOTHERAPY-INDUCED ANEMIA: A HEALTH ECONOMIC EVALUATION Finek J1, Holubec L1, Wiesnerova A1, Pav Z1, Dusek L2 1 Teaching Hospital and Medical Faculty of Charles University, Plzen, Czech Republic; 2 Masaryk University Brno, Brno, Czech Republic OBJECTIVES: Erythropoiesis-stimulating proteins (ESPs) are used in the treatment of chemotherapy-induced anemia (CIA) with the aim of improving quality of life and reducing the need for blood transfusions. a retrospective analysis based on data from a single institution compared costs and effectiveness of two ESPs, epoetin alfa and darbepoetin alfa, in consecutively treated patients over a 2-year period. METHODS: Data from all patients treated for CIA between January 1, 2007 and December 31, 2008 with one of the two ESPs—epoetin alfa- (40 000 IU once weekly) or darbepoetin alfa- (500 µg every 3 weeks)—were analyzed. Total and per patient costs and costs per clinical response (hemoglobin ≥ 11 g/dL) were calculated, based on drug acquisition costs for the ESPs. RESULTS: A total of 161 patients were treated with epoetin A465 alfa (799 doses; mean 4.96 doses per patient) and 52 with darbepoetin alfa (94 doses; mean 1.81 doses per patient). Total and per-patient costs were 8,682,226 CZK (c331,003.66) and 54,461 CZK (c2076.29) for epoetin alfa versus 2,578,984 CZK (c98,321.92) and 49,596 CZK (c1,890.81) for darbepoetin alfa. The response rate was higher (58% vs. 47%, not statistically significant), and the mean cost per treatment response was lower, with darbepoetin alfa, at 85,966 CZK (c3,277.39) versus 115,763 CZK (c4,413.38) for epoetin alfa. CONCLUSIONS: Our results indicate that darbepoetin alfa is associated with lower drug acquisition costs than epoetin alfa, as well as a lower cost per treatment response, for treatment of CIA. PSY30 ONE-YEAR-LONG TREATMENT OF FAILED BACK SURGERY SYNDROME WITH SPINAL CORD STIMULATION: COSTS AND BENEFITS IN THE ITALIAN CONTEXT (PRECISE STUDY) Beccagutti G1, Zucco F2, Lavano A3, De Rose M3, Poli P4, Fortini G5, Demartini L6, De Simone E7, Menardo V8, Cisotto P9, Meglio M10, Grifi M1, De Santo T11, Costantini A12 1 Medtronic Italia, Sesto San Giovanni, Italy; 2Azienda Ospedaliera Salvini, Garbagnate Milanese, Italy; 3Università Magna Grecia, Catanzaro, Italy; 4Azienda Ospedaliera Universitaria Pisana, Pisa, Italy; 5Azienda Ospedaliero Universitaria Ospedale di Circolo e Fondazione Macchi, Varese, Italy; 6IRCCS Fondazione Salvatore Maugeri, Pavia, Italy; 7A.O.R.N. “S.G. Moscati”, Avellino, Italy; 8A.S.O. “S. Croce e Carle, Cuneo, Italy; 9Ospedale “S. Maria di Cà Foncello”, Treviso, Italy; 10Policlinico Gemelli, Roma, Italy; 11Medtronic Italia, Roma, Italy; 12 Ospedale Clinicizzato “SS Annunziata”, Chieti, Italy OBJECTIVES: Spinal Cord Stimulation (SCS) is considered to be an effective therapy for patients affected by Failed Back Surgery Syndrome refractory to conventional medical management (CMM). PRECISE Study aims to evaluate the costs and clinical benefits of SCS (plus CMM) versus CMM alone on a 2-year horizon in Italy. Here 12-months results are reported. METHODS: PRECISE Study is an observational, pre-post, multi-centre study. Eighty patients were enrolled to be screened for SCS and, if responders, to be implanted. Health care and non-health care resources use and costs, clinical outcomes and HR-QoL data were collected before and after the implantation. The 12-months data was analyzed and compared to pre-implantation results. Mean monthly resources use, costed in euros 2008, was evaluated according to three perspectives: patient, National Healthcare System (NHS), Society. RESULTS: Seventytwo (90%) patients responded to screening and were implanted. Out of them, 62 (mean age 57; 61% female) completed the 12-months follow-up; ten patients discontinued the therapy for: therapy failure (10%), complications/technical issues (60%), other reasons (30%). SCS significantly reduced pain, and improved function and HR-QoL: mean pain intensity decreased from 7.4 ± 1.3 to 4.4 ± 2.5 (pain Numerical Rating Scale), 53 patients (85%) experienced an improvement in function measured with Oswestry Disability Index and EQ-VAS increased from 37 to 59. The mean utility score was 0.49 (baseline: 011). Patient’s monthly out-of-pocket expenditure halved from c161.45 to c80.04. Excluding upfront costs (screening and implantation), total NHS monthly per-patient expenditure diminished from c163.77 to c100.17, Society monthly per-patient expenditure from c461.89 to c253.27. Including them, NHS and Societal monthly costs per-patient were c1,318.30 and c1,471.40, respectively. CONCLUSIONS: SCS with CMM is more effective than CMM alone in controlling pain and improving HR-QoL. Excluding the upfront costs related to the implantation, SCS allows a reduction in resources consumption and productivity losses from all the possible perspectives of analysis (patient, NHS, Society). PSY31 POSACONAZOLE VS. FLUCONAZOLE OR ITRACONAZOLE FOR THE PREVENTION OF FUNGAL INFECTIONS IN CANCER PATIENTS WITH PROLONGED NEUTROPENIA: A COST-EFFECTIVENESS ANALYSIS Dranitsaris G Augmentium Pharma Consulting, Toronto, ON, Canada OBJECTIVES: Invasive fungal infections (IFI) remain a clinical concern in hematological cancer patients with prolonged neutropenia because they are major cause of morbidity and mortality. As a result, the guidelines of the International Society of Infectious Diseases recommend adequate prophylaxis while the patient remains neutropenic. In this study, a cost-effectiveness analysis was conducted to measure the economic value of posaconazole as an alternative to fluconazole or itraconazole when used to prevent IFI in this patent population. METHODS: A decision analysis model was developed using clinical and economic data from randomized comparative trials, the economic literature and from expert opinion. The data were then used to estimate the incremental cost per life-year saved with oral posaconazole prophylaxis relative to fluconazole or itraconazole from the Canadian provincial health care system perspective. The base-case results were then tested with an extensive sensitivity analysis which evaluated extremes in the incidence of IFI as well as variations in their cost of management. RESULTS: The base-case findings revealed that prophylaxis with posaconazole provides at least comparable efficacy and an overall cost savings of approximately $4260 per patient. Despite variations in the base-case parameters, the sensitivity analysis suggested stability in the primary findings. Posaconazole was associated with an overall cost savings (range = $1765 to $5118) and at least comparable benefit. Optimal cost-effectiveness was obtained when the drug was able to avoid more invasive Aspergillus infections. CONCLUSIONS: Prophylaxis with posaconazole in neutropenic hematological cancer patients is not only cost-effective but also cost saving. The economic benefits were due to the drug’s ability to reduce the incidence of high cost fungal infections, particularly Aspergillus species. A466 PSY32 THE COST-EFFECTIVENESS OF ELTROMBOPAG FOR THE TREATMENT OF CHRONIC ADULT IMMUNE THROMBOCYTOPENIC PURPURA (ITP) IN IRELAND Hanley RM1, Redmond S1, Thompson G2 1 GlaxoSmithKline, Dublin, Ireland; 2GlaxoSmithKline, Uxbridge, UK OBJECTIVES: Immune thrombocytopenia (ITP) is a rare autoimmune disorder characterised by low platelet counts resulting in symptoms that are usually minor, e.g. petechiae. However, patients with very low platelet counts face the risk of serious, but rare, bleeds, e.g. intracranial hemorrhage. The study objective was to estimate the cost-effectiveness of treating two groups of chronic adult ITP patients (refractory to splenectomy and where splenectomy is contraindicted) with oral eltrombopag versus romiplostim or rituximab from an Irish health care payer perspective. METHODS: A Markov model with a four week model cycle, two year time horizon and seven health states was used. In one arm, patients were treated with eltrombopag and in the other romiplostim or rituximab. Patients began in either the ‘controlled platelet’ or ‘uncontrolled platelet’ health states. Response rates from clinical trials (eltrombopag and romiplostim) and a published literature review (rituximab) determined whether patients were controlled or not. Controlled patients experienced a higher level of utility and lower risk of transitioning to bleed health states. Utilities (using the SF-36v2) for controlled/uncontrolled platelet levels were from the eltrombopag pivotal trial, RAISE. Utility for bleeds and costs were taken from the literature. Costs and benefits were discounted by 4%. RESULTS: The model estimated that eltrombopag dominates romiplostim, i.e. cost savings of c13,000 and c18,000 with an incremental QALY of 0.1 and 0.03 for patients who are refractory to splenectomy and where splenectomy is contraindicated respectively. This result was driven by the fact that eltrombopag is less expensive, does not incur wastage or administration costs and has slightly better durable response rates compared to romiplostim. Eltrombopag is not cost-effective compared to rituximab for both groups of patients because rituximab is significantly less costly than eltrombopag with comparable response rates. CONCLUSIONS: The economic model showed that eltrombopag dominates romiplostim but is not costeffective versus rituximab. PSY33 EARLY STAGE ECONOMIC EVALUATION OF PHARMACOGENOMICBASED DIAGNOSTICS IN CHRONIC MYELOID LEUKEMIA Gaultney J1, Sanhueza E2, Janssen JJWM3, Redekop WK1, Uyl-De Groot C1 1 Institute for Medical Techonology Assessment (iMTA), Rotterdam, The Netherlands; 2 Facultad de Medicina Universidad Católica de Chile, Santiago, Chile; 3VU University Medical Center, Amsterdam, The Netherlands OBJECTIVES: Uncertainty regarding the effectiveness of tyrosine kinase inhibitors (TKIs) for treatment of chronic myeloid leukemia (CML) has resulted in difficulty for some appraisal committees to recommend second-line TKIs. Implementation of pharmacogenomic-based diagnostics may facilitate individualized treatment with TKIs and improve overall treatment effectiveness. The objective of this economic evaluation was to estimate the potential cost-effectiveness of a companion diagnostic that predicts treatment response in CML patients eligible for dasatinib and nilotinib. METHODS: A decision analytic model was created to assess the cost-effectiveness of individualized treatment with dasatinib and nilotinib using a companion diagnostic test relative to current care (i.e., treat all patients with dasatinib until progression or no response, then treat with nilotinib). Transition probabilities and utility values were taken from published RCTs. The cost analysis included costs of treatment, response monitoring and the companion diagnostic. Scenario analyses were conducted to assess the impact of unavailable or highly uncertain values. RESULTS: Total costs and effects for the base-case analysis were c99,011 and 1.84 progression-free LYs (PF-LYs) and 1.63 QALYs with the companion diagnostic strategy after a 2-year time horizon. In the current care situation, total costs and effects were c100,904 and 1.74 PF-LYs and 1.61 QALYs. The companion diagnostic was the dominant strategy with a savings of approximately c4,000/PF-LY or c2,000/QALY. Scenario analyses revealed that the cost-effectiveness of the companion diagnostic is sensitive to time to response, time to progression and comparative effectiveness between dasatinib and nilotinib. CONCLUSIONS: Treatment of CML patients eligible for second-line TKIs using a companion diagnostic may be cost-effective. More precise estimates for uncertain or unavailable parameters should be incorporated when available. This case demonstrates the use of economic evaluation methods at early stages of technology development for internal decision-making and communicating the potential value of pharmacogenomic-based diagnostics to stakeholders. PSY34 COST-EFFECTIVENESS OF TAPENTADOL PROLONGED-RELEASE (PR) COMPARED TO OXYCODONE AND MORPHINE IN PATIENTS WITH CHRONIC MODERATE TO SEVERE NON-CANCER PAIN IN SWEDEN Hertel N1, Ikenberg R1, Fricke FU1, Liedgens H2 1 IMS Health GmbH & Co. OHG, Nuremberg, Germany; 2Gruenenthal GmbH, Aachen, Germany OBJECTIVES: To assess the cost-effectiveness of tapentadol compared to oxycodone and morphine as 1st line strong opioid therapy in chronic non-cancer pain (CNCP) patients in Sweden. METHODS: A Markov transition state model over one year with cycle times of one month was built. Four health states were defined: ‘occurrence of adverse events (AEs) with need for medical treatment’, ‘withdrawal due to AEs’, ‘response to therapy’ and ‘death’. If patients did not adequately respond to tapentadol 13th Euro Abstracts or comparators or withdrew, switching to alternative 2nd line opioid (oxycodone, morphine or transdermal fentanyl) was considered. After initiating 3rd line therapy, patients could stay on this therapy or die according to morbidity rate. Data regarding efficacy, tolerability and utility values (EQ-5D) were derived from clinical trials and published literature. Switch rates to subsequent opioid therapies and resource consumption were estimated by clinical experts. Costs were calculated from the societal perspective. Direct costs were calculated based on official Swedish prices/tariffs, indirect costs were calculated based on figures obtained from the literature and current wage rates. Costs and benefits were not discounted in base-case calculation. Impact of selected parameters on the results was evaluated in one-way sensitivity analyses (OWSA). RESULTS: Mean annual total costs per patient amount to 242,583 SEK (Swedish krona) for tapentadol vs. 247,813 SEK for oxycodone and 246,093 SEK for morphine. Tapentadol generates 0.4712 QALYs compared to 0.4518 QALYs for oxycodone and 0.4535 QALYs for morphine. More QALYs generated in the model reflect tapentadol’s better tolerability profile than comparators. Cost parameter comprising physician visits, co-medications and non-drug therapy costs revealed highest impact in OWSA. CONCLUSIONS: Tapentadol, once approved, will generate more QALYs at lower costs than oxycodone and morphine. Tapentadol appears to be the favourable treatment option in the 1st line therapy of CNCP patients in Sweden from a societal perspective. PSY35 THE COST-EFFECTIVENESS OF DULOXETINE IN THE TREATMENT OF FIBROMYALGIA IN THE UNITED STATES Beard S1, Le TK2, Zhao Y2, Roskell N3 1 RTI Health Solutions, Sheffield, UK; 2Eli Lilly & Company, Indianapolis, IN, USA; 3RTI Health Solutions, Manchester, UK OBJECTIVES: To evaluate the cost-effectiveness of duloxetine for the management of fibromyalgia assessed from the perspective of a health care payer in the United States. METHODS: A Markov model was used to evaluate the economic and clinical advantages in controlling fibromyalgia pain symptoms, considering duloxetine as an additional treatment option. The standard treatment sequence was defined based on clinical guidelines including tricyclic antidepressants (TCAs), second-generation antidepressants (based on SNRIs), anticonvulsants, and opioid therapies. The model included 2 levels of pain response (a ≥30% and ≥50% change from baseline, using a standard 11-point severity scale). Clinical efficacy and discontinuation data were taken from a systematic literature review and mixed treatment comparison, including both placebo and active controlled trials. Utility data were linked to pain severity using trial-based EQ5D data. Discounting was applied at 3.0% per year. RESULTS: The introduction of duloxetine resulted in additional symptom-controlled months (SCMs), defined as the amount of time at a ≥30% response level, and quality-adjusted life-years (QALYs), over a 2-year time horizon. First-line treatment resulted in an additional 665 SCMs and 12.3 QALYs, with a cost of $582,911 ($877 per SCM and $47,560 per QALY). Second-line treatment resulted in an additional 460 SCMs and 8.7 QALYs, with a cost of $143,752 ($312 per SMC and $16,565 per QALY). a costeffectiveness frontier analysis suggested that second-line duloxetine is likely to be the most cost-effective option, however, in sensitivity analyses, the cost-effectiveness of first-line duloxetine improved when assumptions around continued treatment switching for long-term drop-outs were relaxed. CONCLUSIONS: There is currently a significant unmet need for patients with poorly controlled fibromyalgia where pain is a predominant symptom. These analyses show that the introduction of duloxetine into the standard treatment sequence for fibromyalgia can provide additional patient benefits, which are cost-effective when compared to commonly adopted thresholds. PSY36 ASSESSING THE COST-EFFECTIVENESS OF PRIMARY PROPHYLAXIS FOR THE TREATMENT OF SEVERE HEMOPHILIA A Axelsen F1, Miners A2 1 Baxter Healthcare SA, Zurich, Switzerland; 2London School of Hygiene and Tropical Medicine, London, UK OBJECTIVES: The objective of this analysis was to understand the economic consequences of different products in primary prophylaxis in severe hemophilia A patients. Miners’ (2002, 2009) hemophilia economic Markov model was revised to compare two different factor VIII products—a full-length recombinant FVIII (Advate) and a B domain-deleted recombinant FVIII (BDDrFVIII) (Refacto AF). Based on documented differences in half life in a pharmacokinetic cross-over license study [1] with results submitted to EMA[2], this research evaluated the amount of clotting factor required to maintain the patient factor VIII trough levels above 1%. If a product has a shorter half-life, more clotting factor is required to prevent factor VIII levels falling below the minimum trough level. METHODS: The same basic model structure of Miners’ hemophilia economic Markov model was used to analyze the effects of differences in half-life in an Italian economic country setting, full-length recombinant FVIII (Advate) half-life 13.3 hours at a list price of c0.75 per IU versus BDDrFVIII (Refacto AF) half-life 11.2 hours at a list price of c0.69 per IU. The model considers the amount of factor required to prevent trough clotting factor levels from falling below 1%. RESULTS: The analysis estimated that the mean expected, discounted lifetime (70 years) cost of primary prophylaxis in a severe hemophilia A patient was c0.88 million less using full length recombinant FVIII (Advate) when compared to a BDDrFVIII (Refacto AF), assuming a 2.1 hour shorter half-life; i.e. equal to over c12,500 in cost savings per patient per year. CONCLUSIONS: This analysis shows that prophylaxis therapy using full-length recombinant FVIII (Advate) may be cost saving compared to 13th Euro Abstracts BDDrFVIII (Refacto AF) based on the difference in half-life reported, even if BDDrFVIII has a lower price per IU. [1] Recht et al. Hemophilia 2009:1–12; [2] EMA Assessment report Refacto AF. London March 2009. http://www.ema.europa.eu/ humandocs/PDFs/EPAR/Refacto/refactoaf-H-232-II-59-68-AR.pdf PSY37 COST-EFFECTIVENESS AND BUDGET IMPACT ANALYSIS OF OROS HYDROMORPHONE IN PATIENTS WITH CHRONIC CANCER PAIN FROM THE PUBLIC PAYER PERSPECTIVE IN BRAZIL Fernandes RA1, Takemoto ML1, Guerra FC1, Passos RB1, Pereira ML2, Morais AD2 1 ANOVA—Knowledge Translation, Rio de Janeiro, RJ, Brazil; 2Janssen-Cilag, São Paulo, SP, Brazil OBJECTIVES: To conduct cost-effectiveness and budget impact analysis (BIA) of OROS hydromorphone versus CR morphine and CR oxycodone for moderate to severe cancer pain from the public payer perspective in Brazil. METHODS: A decision tree followed by a Markov Model with a 12 month time horizon was developed with data from the Phase III trial Hanna 2008. The achievement of mild pain (worst pain scores < 4) was considered as outcome. Only direct medical costs were considered and unit costs were obtained from Brazilian official lists. For the BIA, 10% of currently used CR morphine daily doses was substituted for equivalent OROS hydromorphone doses. The same rational was adopted for CR oxycodone comparison. Univariate deterministic sensitivity analyses showed that the results remained consistent through model parameters variation. RESULTS: OROS hydromorphone showed 1.66 additional months in mild pain per patient per year when compared to both CR morphine and CR oxycodone. Annual treatment costs were 2.401 BRL, 1.256 BRL and 5.114 BRL for OROS hydromorphone, CR morphine and CR oxycodone respectively. The incremental cost-effectiveness ratio was 689 BRL per additional month in mild pain per patient per year, when OROS hydromorphone was compared to CR morphine. Versus CR oxycodone, OROS hydromorphone was more effective with fewer costs, being cost saving (ICER -1,634 BRL). BIA results showed that the substitution of 10% of current utilization of CR morphine for OROS hydromorphone and CR oxycodone would result in a budgetary impact of 118,722 BRL and 347,295 BRL, respectively. CONCLUSIONS: OROS hydromorphone is cost saving when compared to CR oxycodone and is more cost-effective than CR oxycodone when both are compared to the current scenario of chronic cancer pain treatment with CR morphine, with a lower budgetary impact. PSY38 COST-EFFECTIVENESS OF GOLIMUMAB IN ANKYLOSING SPONDYLITIS FROM THE UK PAYER PERSPECTIVE Muszbek N1, Punekar Y2, Buchanan J3, Prasad M4, Vioix H1, Brown R1 1 United BioSource Corporation, London, UK; 2Schering Plough, Welwyn Garden City, UK; 3 Johnson & Johnson Pharmaceutical Services, LLC, Malvern, PA, USA; 4Merck & Co, Kenilworth, NJ, USA OBJECTIVES: Golimumab is a novel TNF-α inhibitor for treatment of patients with severe active ankylosing spondylitis (AS). This study evaluated the cost-effectiveness of golimumab and its appropriate comparators in the treatment of AS from UK National Health Service perspective. METHODS: A Markov model with an initial decision tree was developed to simulate the progression of a hypothetical cohort of active AS patients for 20 years. The primary outcome measure was quality-adjusted life-years (QALYs) estimated using Bath Ankylosing Spondilytis Functional Index (BASFI) whereas the primary response measure was ≥50% improvement in Bath Ankylosing Spondilytis Disease Activity Index (BASDAI) at 12 weeks. Direct costs including medication costs and AS management costs were included. Golimumab was compared with conventional treatment and other TNF-α inhibitors. Costs and outcomes were discounted at 3.5%. RESULTS: All TNF-α inhibitors were superior to conventional treatment and comparable to each other on BASDAI response. The incremental cost-effectiveness ratio (ICER) for golimumab was £15,353 per QALY compared to conventional treatment. The probability of golimumab being cost-effective at a threshold of £30,000/QALY was 92%. Compared to etanercept and adalimumab, golimumab generated marginally more QALYs at marginally more costs. CONCLUSIONS: Golimumab may be considered as a cost-effective treatment alternative for patients with AS. With comparable costs and efficacy to other TNF-α inhibitors, golimumab’s position in the treatment pathway is likely to be driven by patient and physician choice. PSY39 REAL-WORLD COST-EFFECTIVENESS OF BORTEZOMIB IN RELAPSED OR REFRACTORY MULTIPLE MYELOMA IN THE NETHERLANDS Gaultney J1, Franken M1, Huijgens PC2, Sonneveld P3, Uyl-De Groot C1, Redekop WK1 1 Institute for Medical Techonology Assessment (iMTA), Rotterdam, The Netherlands; 2VU University Medical Center, Amsterdam, The Netherlands; 3Erasmus University Medical Center, Rotterdam, The Netherlands OBJECTIVES: The Dutch reimbursement policy for expensive inpatient medicines requires outcomes research after four years of temporary reimbursement. Based on a retrospective study, we explored the cost-effectiveness of bortezomib for relapsed/ refractory multiple myeloma in Dutch daily practice. METHODS: Detailed clinical data from a real-world cohort of 72 patients treated with bortezomib and 67 patients never treated with bortezomib were collected from medical records. Validity of the incremental cost-effectiveness was assessed by comparing baseline prognosis between bortezomib and non-bortezomib patients. Clinical effectiveness was evaluated by A467 comparing Kaplan-Meier survival estimates. Costs of resource use from a hospital perspective were based on patient-level data. RESULTS: Prognostic factors for bortezomib patients were significantly different compared to non-bortezomib patients. Incremental analyses for bortezomib versus non-bortezomib patients were therefore not performed. Total mean costs and median survival from start of relapsed/refractory treatment for bortezomib patients were c84,042 and 33.2 months. Bortezomib accounted for 21% of total costs among these patients. For non-bortezomib patients, total mean costs and median survival from start of relapsed/refractory treatment were c54,435 and 21.6 months. The proportion of patients still in follow-up at the end of data collection was slightly higher in bortezomib versus non-bortezomib patients (51% vs. 46%). Total mean costs for bortezomib patients did not differ significantly when excluding patients still in follow-up. For non-bortezomib patients, total mean costs differed significantly when excluding patients still in follow-up, mainly due to high costs of lenalidomide treatment, stem cell transplants and inpatient hospital stays. CONCLUSIONS: Our real-world data challenged the assessment of the incremental cost-effectiveness of bortezomib versus other treatments in the indication of relapsed/ refractory multiple myeloma. It was possible to estimate the cost and effects for bortezomib patients in daily practice to determine the real-world value. Data synthesis incorporating effectiveness for the relevant comparator might facilitate estimation of a valid ICER. PSY40 COST-EFFECTIVENESS OF PREGABALIN VERSUS USUAL CARE IN REFRACTORY OUT-PATIENTS WITH NEUROPATIC PAIN FOLLOWED IN PRIMARY CARE SETTINGS de Salas-Cansado M1, Pérez C2, Saldaña MT3, Navarro A4, Rejas J5 1 Trial Form Support Spain, Madrid, Spain; 2Pain Clinic, Hospital de la Princesa, Madrid, Spain; 3 Primary Care Health Centre Raíces, Castrillón, Asturias, Spain; 4Primary Care Health Centre Puerta del Ángel, Madrid, Spain; 5Pfizer España, Alcobendas/Madrid, Spain OBJECTIVES: Estimate the cost-effectiveness of Pregabalin (PGB) versus Usual Care (UC) in refractory out-patients with Neuropathic Pain (NeP) treated according to usual medical practice in Primary Care settings (PCS) in Spain. METHODS: Data extracted from a 12-week non-interventional prospective study conducted to ascertain the cost of NeP were used. PGB naïve patients treated with UC or PGB, matched by age (+5 years), sex and pain intensity (+5 pts), refractory (≥40VAS-McGill) to previous treatment during the prior 6 months, were selected in a 1:1 ratio. Patients could switch to PGB (monotherapy/add-on) or to UC (non-narcotics, opiates, antidepressants and/ or anticonvulsants). Time horizon was 12 weeks. Effectiveness was quality-adjusted life-years (QALY) gain. Perspectives of the National Health Service (NHS) and society (2006) were included, and expressed as an incremental cost-effectiveness ratio (ICER). Bootstrapping techniques (10,000 re-samples) were used to obtain the probabilistic ICER, its 95% percentile confidence interval (CI) and the cost-effectiveness acceptability curve. RESULTS: A total of 160 patient-pairs were extracted. Compared with UC, PGB was associated with significantly higher QALY gain; 0.0374 ± 0.0367 vs. 0.0224 ± 0.0313 (P < 0.001). Notably, although drug costs were higher for PGB (c251 ± 125 vs. c104 ± 121, P < 0.001), its QALY gain did not incur a higher overall total cost (c1335 ± 1302 vs. c1387 ± 1489; P = 0.587), nor higher health care costs (c529 ± 438 vs. c560 ± 672; p = 0.628). In fact, the ICER was dominant for total and health care costs, with CIs respectively, dominant-c17,268, and dominant-c6,508. ICER for drug costs was c10,672/QALY (dominant-c19,858). The 99% of the re-samples were below the threshold of c30,000/QALY. CONCLUSIONS: Versus usual care in the community medical setting, pregabalin is highly cost-effective in the treatment of refractory NeP patients. The high indirect costs and increased health care costs associated with the treatment of refractory patients, which offset higher cost of pregabalin, highlight the economic burden of the condition on society. PSY41 COST-EFFECTIVENESS OF BIOLOGICS IN PSORIASIS IN TWO LA COUNTRIES—COMPARISON WITH THE EUROPEAN EXPERIENCE Alandete JC Janssen-Cilag, Bogota, Colombia OBJECTIVES: To evaluate cost-effectiveness of biologics used in patients with psoriasis in Colombia and Peru. METHODS: We estimated direct costs of etanercept, adalimumab, ustekinumab and infliximab based on their labels for first/induction year and second/maintenance year (EUR1 = COL$2340 = SOL$357). For etanercept we considered two induction schemes: 50 mg weekly 52 weeks-D1- and 100 mg 12 weeks followed by 50 mg 40 weeks-D2-. Effectiveness was evaluated as 75% reduction in Psoriasis Area and Severity Index-PASI 75- from meta-analysis presented by Hawkins et al. in the 14th International ISPOR: infliximab = 80%; ustekinumab = 69%; adalimumab = 59%; etanerceptD2- = 52%; etanerceptD1 = 39%. Infliximab and ustekinumab effectiveness were not significantly different. However, both were significantly superior to etanercept (D1 and D2). RESULTS: In Colombia, Ustekinumab was dominant (c29.012 in 2 years) generating cost savings of −c4.416 vs. etanerceptD1; −c7.411 vs. adalimumab; −c8.119 vs. etanerceptD2 and −c25,340 vs. infliximab; with higher or same effectiveness than the other biologics used in that country. In Peru, all the options were more effective and more costly than the standard of care (etanerceptD1). The ICER per patient with PASI 75 of etanerceptD2, adalimumab, ustekinumab and infliximab, compared to etanerceptD1 were c21.654; c19,860, c13,036 and c29,008, respectively. Then, the efficiency frontier was formed by etanerceptD1, ustekinumab and infliximab. Given evidence shows the last two products do not have effectiveness significant differences, ustekinumab became the dominant A468 option. An additional analysis was done supposing that infliximab’s effectiveness was superior to ustekinumab’s. The ICER per patient with PASI 75 between these products was c220,352 in Colombia and c50,989 in Peru. These values were higher than Colombian and Peruvian health systems’ willingness to pay per PASI 75 (c40.334 and c29,242, respectively), calculated based on average effectiveness and costs of biologics. CONCLUSIONS: In the studied LA countries, Ustekinumab resulted the most costeffective biologic, even being cost-saving in Colombia. These results corroborate those observed in European countries. PSY42 COST-EFFECTIVENESS ANALYSIS OF BIOLOGICS IN THE TREATMENT OF MODERATE TO SEVERE PSORIASIS IN THE PUBLIC HEALTH CARE SYSTEM OF BRAZIL Morais AD, Ribeiro F Janssen-Cilag, São Paulo, SP, Brazil OBJECTIVES: To compare treatment costs and cost-effectiveness results for biologics approved for the treatment of moderate to severe psoriasis from the perspective of a public payer in Brazil. METHODS: Annual treatment costs were calculated based on the number of vials used, as defined in the label for each biologic, and from the price per vial. The prices of etanercept, adalimumab and infliximab were obtained from purchase disclosures by the Brazilian Ministry of Health. As ustekinumab, the latest biologic approved for psoriasis, has not yet been purchased by that Ministry, its list price after deducting the mandatory discount for government sales (22.85%) was considered. The effectiveness of each treatment was defined by a recent metanalysis as the PASI75 response at the primary endpoint for each drug. The average patient weight was assumed at 70 kg. RESULTS: Ustekinumab has the lowest treatment cost across the biologics in both induction and maintenance years with the least number of vials used. With a total of 4 applications during a maintenance year, ustekinumab has the lowest treatment cost, followed by infliximab (55% more expensive), adalimumab (62% more expensive) and etanercept (75% more expensive). Combining the cost and effectiveness results, ustekinumab has the lowest cost per response, followed by infliximab (44% less cost-effective), adalimumab (104% lesss cost-effective) and etanercept (152% less cost-effective). CONCLUSIONS: With the lowest treatment cost, ustekinumab is an important treatment option in moderate to severe psoriasis. When comparing the cost-per-response across biologics, ustekinumab is further differentiated from the remaining biologics. Ustekinumab represents a rare situation in Brazil combining the lowest treatment cost and a high effectiveness result. PSY43 ECONOMIC ANALYSIS OF RITUXIMAB IN COMBINATION WITH CYCLOPHOSPHAMIDE, VINCRISTINE AND PREDNISOLONE IN THE TREATMENT OF PATIENTS WITH ADVANCED FOLLICULAR LYMPHOMA IN PORTUGAL Pereira C1, Negreiro F2, Silva C2 1 Roche Farmacêutica Quimica, Portugal, Amadora, Portugal, 2Eurotrials—Consultores Científicos, Lisbon, Portugal OBJECTIVES: Evaluate costs and benefits of rituximab in combination with cyclophosphamide/vincristine/prednisolone chemotherapy regimen (R-CVP) versus CVP alone in previously untreated patients with indolent non-Hodgkin lymphoma (NHL). METHODS: Cost-effectiveness (Life-years Gained—LYG) and cost-utility analysis (Quality Adjusted Life-years—QALYs) were performed for a time horizon of 10 years according to a Markov economic model with three health states (“progression free survival”, “progression” and “death”) and monthly cycles. Data from a phase III clinical trial (Marcus R. et al 2007) was used and expanded in the model to include unpublished 53-month median follow-up data. Survival after first-line therapy was estimated from the Scotland and Newcastle Lymphoma Group registry data and utilities were derived from a UK study. Resource consumption was estimated by a Portuguese expert panel. Costs were calculated from the Portuguese Health System perspective through official data with prices updated to 2010. Only direct medical costs were considered. Costs and consequences were discounted at 5% per annum. Deterministic and probabilistic sensitivity analyses were performed around assumptions on the time horizon, costs, utilities and excess mortality rate due to progression applied in the base-case analysis. RESULTS: The 10-year base-case analysis showed a lower total cost per patient with CVP alone (c88,373) versus R-CVP (c89,899). Life expectancy and quality-adjusted life expectancy per patient were higher with R-CVP (6.361 and 4.166, respectively) versus CVP alone (5.557 and 3.438, respectively), representing increases of 0.804 in LYG and 0.728 in QALYs gained. The incremental cost per QALY gained was c2097 (base-case) and c6006 considering a lifetime horizon (25 years). The sensitivity analyses confirmed the robustness of the model. CONCLUSIONS: This study demonstrates that the combination R-CVP in previously untreated NHL patients improves life expectancy and is a cost-effective alternative to CVP in Portugal. PSY44 COST-EFFECTIVENESS OF A LIDOCAINE PLASTER RELATIVE TO GABAPENTIN AND PREGABALIN IN THE TREATMENT OF POSTHERPETIC NEURALGIA IN SPAIN Nuijten M1, Hidalgo A2, Obradovic M3, Liedgens H3 1 Ars Accessus Medica/Erasmus University Rotterdam, Jisp, The Netherlands; 2Castilla-La Mancha University, Toledo, Spain; 3Gruenenthal GmbH, Aachen, Germany OBJECTIVES: To assess the cost-effectiveness of using a lidocaine 5% medicated plaster in the treatment of post-herpetic neuralgia (PHN) compared with gabapentin 13th Euro Abstracts and pregabalin from the perspective of the payer in Spain. METHODS: A Markov model was constructed to calculate the cost-effectiveness of gabapentin, pregabalin and lidocaine plaster in terms of the cost per quality-adjusted life-year (QALY) gained when used over a six-month time horizon in patients with PHN. The model structure allowed for differences in costs, utilities and transition probabilities between the initial 30-day run-in period and maintenance therapy. Transition probabilities were based on the comparative and long-term clinical trials identified through a systematic literature review. Missing data, including resource utilization, were obtained from a Delphi panel, and cost data were obtained from the official price/tariffs lists. Utilities were derived from the literature and were supplemented and validated by the Delphi panel. RESULTS: The total cost of treatment with the lidocaine plaster was c1414 per patient at a daily consumption of 1.1 plasters, compared with c1100 for gabapentin (average dosage 2100 mg/day during maintenance phase), and c1348 for pregabalin (average dosage 488 mg during maintenance phase). Lidocaine plaster generated 0.428 QALYs, compared with 0.339 for gabapentin, and 0.399 for pregabalin. Lidocaine plaster therefore had an incremental cost-effectiveness ratio of c3525/QALY gained relative to gabapentin at generic price, and c742/QALY relative to pregabalin. Scenario analyses and extensive one-way sensitivity analyses on all parameters including the time horizon confirmed the robustness of the results. CONCLUSIONS: The lidocaine 5% plaster is a highly cost-effective treatment for PHN in Spain. PSY45 PREGABALIN IS A COST-EFFECTIVE MEDICINE FOR REFRACTORY NEUROPATHIC PAIN IN SWEDEN Sandelin R1, Prettyjohns M2, Lister SP2 1 Pfizer, Sollentuna, Sweden, Sweden; 2Cardiff Research Consortium Ltd, Cardiff, UK OBJECTIVES: Patients refractory to older therapies for neuropathic pain (NeP) have few remaining therapeutic options. To demonstrate pregabalin’s value to decision-makers, this study evaluates the cost-utility of pregabalin in the treatment of patients with refractory neuropathic pain in Sweden, from a health care and societal perspective. METHODS: A discrete event simulation (DES) model was constructed to compare pregabalin with usual care. Pain profiles were generated for the pregabalin and usual care cohorts using clinical data from a synthesis of five non-randomised pregabalin studies in treatment-refractory patients. Utility data were generated from a UK survey of patients with NeP, which were used in a recent successful Health Technology Assessment (HTA) of pregabalin in the UK [1]. Cost data were generated from the Swedish TLV’s product price database, a national NeP register, and a regional register study. Indirect costs were estimated from published sources. One-way, and probabilistic sensitivity analyses (PSA), evaluated uncertainty in the model’s output. RESULTS: The incremental cost-effectiveness ratio (ICER) for pregabalin compared to usual care was 51,616 SEK/c5,364 (123,993 SEK/c12,886 excluding indirect costs). One-way sensitivity analyses confirmed the clinical input data as the main driver of the model; even considerable changes to all other input parameters had only a modest effect on the ICER. The PSA generated an ICER of 41,634 SEK/c4,327 (with indirect costs included), suggesting that the model is relatively insensitive to the combined variation in all input parameters; this is evident in the ICER scatter, in which the costeffectiveness pairs are tightly grouped. CONCLUSIONS: Our study found pregabalin to be highly cost-effective compared to usual care in Swedish treatment-refractory patients. Moreover, the PSA showed pregabalin’s favourable cost-effectiveness to be robust in all modelled scenarios, with an ICER well below a conservative threshold of 347,495 SEK / c36,113/£30,000. [1] Scottish Medicines Consortium (SMC). Pregabalin re-submission, advice issued May 2009. PSY46 INDIRECT COSTS OF MODERATE AND SEVERE FORMS OF PSORIASIS IN CZECH REPUBLIC: CALCULATION BASED ON SELF REPORTED QUESTIONNAIRE Klimes J1, Dolezal T2 1 Charles University, Faculty of Pharmacy, Hradec Kralove, Czech Republic; 2Institute for Health Economics and Technology Assessment, Prague, Czech Republic OBJECTIVES: Patients with moderate to severe forms of psoriasis reveal lower quality of life. This QoL deterioration comes together with their productivity loss and disability. Therefore, we calculated indirect costs, from societal perspective, of psoriasis in Czech Republic. METHODS: This calculation was based on 6 months-retrospective self-reported questionnaire. We included 179 patients with psoriasis, aged 18–62 years either at working status, partly-disabled or fully-disabled. For calculation of indirect costs we included costs associated with absenteeism (number of missed workdays) and costs associated with disability pension. We used friction costs approach (FCA) with defined friction period of 6 months (130 workdays). The height of average monthly income in year 2009, c851.3 and height of average monthly partly-disability pension and fully-disability pension c229.44 and c365.61, respectively were used for calculation. Indirect costs were expressed as mean value per one patient with moderate to severe form of psoriasis per one year—2009. Clinical data (PASI score, BSA index and QoL) were collected by dermatologists. RESULTS: Average patients’ age was 45.34 years (23–62 years), average time from diagnosis was 23.9 years. Mean BSA index and PASI score were 26.78% and 13.32, respectively. Occurrence of psoriatic arthritis was 34.1%. Percentage of fully-disabled and partly-disabled patients was 9.5 and 6.1%, respectively. 18.4% of patients reported incapacity to work with average duration of 28.8 workdays in previous 6 months. Mean costs associated with absenteeism, partly-disability pension and fully-disability pension were c214, c168 and c414, respectively. CONCLUSIONS: By using friction cost approach total mean indirect costs of patients with moderate to severe forms of psoriasis were c796 per patient per year. The height of indirect costs was correlated with PASI score, BSA index and QoL. 13th Euro Abstracts PSY47 VARIATION IN PRIMARY CARE CONSULTATION COSTS AMONG NON-CANCER PATIENTS TREATED WITH CHRONIC OPIOID ANALGESIA IN THE UNITED KINGDOM Poole CD1, Conway P2, Baxter G2, Currie CJ3 1 Pharmatelligence, Cardiff, UK; 2Grunenthal Ltd., Stokenchurch, UK; 3Cardiff University, Cardiff, Wales, UK OBJECTIVES: Chronic pain affects around five million people in the UK. Noninjectable, strong opioid analgesia is increasingly being used to manage chronic pain but this requires careful management to avoid common adverse effects. This study aimed to characterise variation in primary care consultation costs as a function of analgesic stability among patients treated with long-term, non-injectable opioids. METHODS: Cases with at least 6-months’ observation prior to opioid analgesia and no history of cancer were extracted from The Health Improvement Network database (THIN), a source of anonymised patient-level data from UK general practice. The number of different, strong opioid drugs (British Pain Society defined) prescribed to each patient during their first year of chronic opioid analgesia was used a proxy measure of analgesia stability. Primary care consultations were costed using a standard tariff at £UK2007. RESULTS: A total of 21,261 cases were studied, of which 54% were female and the mean age at initiation of opioid analgesia was 62 years (sd 11). Prior to initiating strong opioid analgesia the annual, average, total cost of consultations with primary care health practitioners was £188 (95%CI £184 to £191). After initiation of opioid analgesia, patients treated with a single drug incurred annual consultation costs of £546 (£537 to £554) per year on average, and these costs increased sharply with each additional opioid: £706 (£686 to £726), £836 (£789 to £884), and £975 (£818 to £1132; pANOVA < 0.001) for two, three, and four or more opioids, respectively. The majority of consultations (84%) were with general practitioners in the surgery. CONCLUSIONS: The occurrence of persistent pain was associated with a considerable increase in the rate and the related financial cost of consultations with health professionals in primary care. Furthermore, a substantial cost gradient was evident among patients who failed to be managed with stable opioid monotherapy during their first year of opioid analgesia. PSY48 RESOURCE USE AND COSTS OF THE GASTRIC BYPASS SURGERY VERSUS CLINICAL TREATMENT FOR OBESE PATIENTS WITH COMORBIDITIES UNDER THE BRAZILIAN PRIVATE HEALTH CARE SYSTEM Nasciben V Johnson & Johnson, Sao Paulo, SP, Brazil OBJECTIVES: To assess use of resources and costs of the surgical versus clinical treatment for obese patients with comorbidities, under the Brazilian private health care system. METHODS: In order to estimate the differences of resource use and costs related to surgery versus clinical treatment one panel of specialists was conducted to raise up the resources necessaries for the treatment from pre-operatory evaluation to the last medical visit after 5 years. For the clinical treatment group we assumed a conservative scenario where the complications are not considered. a micro costing technique based on public price lists was applied to value the resources from the panel considering only direct medical costs. a discount rate of 5% was assumed. RESULTS: For the first year the total costs per patient for the surgically treated group were R$14,547, for the open approach and R$31,415 for the laparoscopic approach (preoperatory assessment R$2,009, inpatient costs R$2,728 versus R$2,493, procedure R$7,233 versus R$24,478, complications R$323 versus R$181 and the clinical follow up R$2,254 for both) and for the clinical treated group the total costs per patient was R$7,545 where R$5,062 occurred due to the obesity management (diet, consults, and drugs for weight management) and R$2,488 for comorbidities (drugs, consults and exams for Diabetes TII, hypertension and hypercholesterolemia). After the first year the surgically treated group had R$1690 of annually costs for the follow up versus R$3290 for the clinical treatment group. After 5 years the total costs were R$22,284 and R$39,153 for open and laparoscopic surgery and R$19,217 for the clinical treatment. CONCLUSIONS: Findings suggest that gastric bypass can reduce the use of resources and costs related to the management of obese patients, under the Brazilian private health care system. SYSTEMIC DISORDERS/CONDITIONS – Patient-Reported Outcomes Studies PSY49 PATIENTS PREFERENCES IN OVERWEIGHT AND OBESITY THERAPY WITH DISCRETE CHOICE EXPERIMENTS Mühlbacher AC, Bethge S Hochschule Neubrandenburg, Neubrandenburg, Germany OBJECTIVES: Overweight is associated with increased risk of morbidity, mortality and appears to adversely affect health-related quality of life (HRQOL). Though advances in obesity therapy can be observed, the long-lasting outcome is dissatisfying. This study aims to investigate patients’ preferences of overweight and obesity therapy in the setting of rehabilitation. METHODS: Qualitative and quantitative methods were applied to identify key attributes. Literature review, focus groups (N = 44) and expert interviews (N = 5) were conducted to elicit relevant attributes and endpoints. a total of 64 possible therapy items were surveyed (N = 201) using a classic rating A469 scale in 5-point Likert format. Psychometric tests (frequency, factor analysis, reliability tests) were used to verify the structure of relevant attributes. Discrete choice experiments were developed using a fractional factorial design in fold over method. For statistical data analysis, we used a random effect logit models for the DCE. Finally a self-administered survey, measuring preferences, was conducted in Germany in 2009 (n = 110). RESULTS: A total of 110 patients responded of which 51,82% were male, mean age 53,05 years, mean BMI 33,54 kg/m2 (SD 7,73) answered the questionnaire. The highest relevance for the respondents’ selection was the aspect of care coordination (Item 6: Coefficient 1.473; SE 0.185) and individual therapy (Item 4: Coefficient 1.446; SE 0.188). The aspect of infrastructure of care was less relevant in the discrete choice experiment. All included attributes lead to significant coefficients. CONCLUSIONS: Weight reduction therapy should enable patients to receive a structured, coordinated and predefined therapy that is individualized based on personal needs, behaviour and circumstances. Patients are willing to abandon infrastructural quality for getting better coordination, structure in there therapy. Further research is needed to distinguish the possible interpretations of the presented preference dimensions, the possibility of including them into weight loss programs and to take notice of heterogeneity within patient population in weight loss programs. PSY50 HEALTH-RELATED QUALITY OF LIFE, HEALTH UTILITIES, AND WORK PRODUCTIVITY ASSOCIATED WITH SPECIFIC OBESITY-RELATED COMORBIDITIES Goren A1, Gupta S2, DiBonaventura M1, Freedman D2 1 Kantar Health, New York, NY, USA; 2Kantar Health, Princeton, NJ, USA OBJECTIVES: The study assessed whether, beyond general risks from obesity-related comorbidities, specific comorbidity types were associated differentially with impairments to health-related quality of life and work productivity. METHODS: Data were from the European 2008 National Health and Wellness Survey of German, Spanish, Italian, UK, and French adults. Logistic regressions identified significant predictors of presence vs. absence of obesity (BMI ≥ 30) among 62 medical conditions, for 52,206 respondents with valid BMI data (of 53,524 total). Significant predictors were then submitted to a principal component analysis (PCA) to identify obesity-related components accounting for the most variance. Conditions loading most highly on the resulting components were used to predict the following outcomes for all 10,391 obese respondents, using multiple regressions: physical (PCS) and mental (MCS) component summary scores from the SF-12v2 quality of life instrument, health utilities (SF-6D), and percentage overall work impairment (WPAI). Comparisons were against obese respondents with none of the PCA-identified conditions, controlling for the other conditions. RESULTS: Logistic regressions identified 35 obesity predictors that reduced to 9 PCA components with one condition loading prominently on each: hypertension, high cholesterol, arthritis, angina, thyroid condition, diabetes, chronic bronchitis, psoriasis, and urinary incontinence. All conditions except psoriasis predicted PCS reductions ranging from 1.22 (with presence of angina) to 9.02 (arthritis), ps < 0.001. All except diabetes and angina, and arthritis and hypertension (associated with MCS increases), predicted MCS reductions ranging from 0.70 (cholesterol) to 4.11 (incontinence), ps < 0.05. All except angina predicted health utilities reductions ranging from 0.011 (hypertension) to 0.077 (incontinence), ps < 0.001. Among 5,242 employed respondents, all except angina and psoriasis predicted overall work impairments of 2.11% (hypertension) to 12.48% (bronchitis), ps < 0.05. CONCLUSIONS: Obese EU respondents differed significantly in quality of life and productivity impairments, based on both presence and specific types of comorbidities. These findings can help guide interventions among obesity-related conditions. PSY51 STATED PREFERENCES OF PHYSICIANS AND CHRONIC PAIN SUFFERERS IN THE USE OF CLASSICAL STRONG OPIOIDS Chancellor J1, Martin M2, Liedgens H3, Baker MG4, Müller-Schwefe GH5 1 Chancellor Health Economics Ltd, Beaconsfield, Buckinghamshire, UK; 2i3 Innovus, Uxbridge, Middlesex, UK; 3Gruenenthal GmbH, Aachen, Germany; 4European Federation of Neurological Associations, Helensburgh, UK, 5German Pain Association, Göppingen, Germany OBJECTIVES: Treating chronic pain involves difficult trade-offs between the goals of pain relief and avoidance of side-effects, but evidence is scarce on relative perspectives of physicians and pain sufferers. We studied each group’s preferences concerning classical strong opioids in France, Germany, Italy, Spain, Sweden and the UK. METHODS: In online, discrete choice experiments (DCE), chronic pain sufferers (n = 242) and physicians (n = 270) chose between hypothetical profiles or an opt-out in 15 choice tasks. Profile descriptions were based on attributes elicited in focus groups with 84 sufferers and semi-structured interviews with 11 physicians. Models were specified by multinomial logit and individual respondent part-worths were estimated by hierarchical Bayesian regression. RESULTS: All main-effects, but no interactions, were significant. Sufferers ranked nausea, pain impact, energy, alertness and constipation; and physicians ranked pain response, CNS effects, nausea, dosage form and constipation in descending order of importance. Sufferers were unwilling to incur severe side-effects to relieve pain, opting out in approximately half of the choice tasks, while physicians were always willing to trade between profiles. The models predicted physicians’ choices well but those of pain sufferers less so. No age, sex or country effects were seen but stronger preferences were expressed by the minority (15%) of physicians treating non-cancer pain, and by the 55% of sufferers who had ever discontinued chronic pain medication and the 41% who reported extreme pain and A470 discomfort in EQ-5D profile responses. Sufferers’ mean pain scores on an 11-point Likert scale were 4.0, 5.7 and 8.6 on their best, average and worst days, respectively. CONCLUSIONS: Online DCEs provide a practical means to compare preferences between physicians and patients, whose concerns overlap substantially but are expressed differently. Sufferers balance pain relief and side-effects. Their refusal to trade between undesirable profiles mirrors high reported rates of treatment discontinuation. PSY52 DEVELOPMENT AND CONTENT VALIDATION OF A MULTICENTRIC CASTLEMANS DISEASE SYMPTOM SCALE Teschendorf B1, Vernon M2, O’Quinn S2, van Rhee F3 1 Johnson & Johnson, Malvern, PA, USA; 2United BioSource Corporation, Bethesda, MD, USA; 3Myeloma Institute for Research and Therapy University of Arkansas for Medical Sciences, Little Rock, AR, USA OBJECTIVES: Multicentric Castleman’s Disease (MCD) is a rare lymphoproliferative disorder characterized by generalized lymphadenopathy and systemic symptoms such as fever, night sweats, fatigue, or loss of appetite. While many of these symptoms can best be assessed through patient-report there currently is no validated MCD PatientReported Outcome (PRO) instrument available to evaluate symptom severity and response to treatment. The purpose of this study was to develop an MCD Symptom Scale and assess its content validity through cognitive debriefing interviews. METHODS: A 16-item MCD PRO was developed through literature review, expert clinician input, and qualitative work with 12 patients diagnosed with MCD. Subsequent cognitive debriefing was used to evaluate the content validity of the PRO with additional MCD patients (N = 10) 6 male, 4 female, mean age 44 years. Study procedures were IRB approved, participants provided written informed consent, interviews were conducted using a discussion guide, audio-recordings were transcribed, and data were analyzed using qualitative analysis software (Atlas.ti). RESULTS: Emergent discussion of symptoms demonstrated that all MCD symptoms experienced by patients are evaluated in the PRO. Symptom experience was variable by patient, supporting the 24-hour recall period for capturing patient-reported symptom severity. All items were consistently understood by participants, who were able to select a response using a 5-point response scale to represent their symptom severity experience. CONCLUSIONS: The MCD Symptom Scale content was found to be comprehensive. Design elements, including response options and recall period were suitable, and content validity was confirmed. The instrument was deemed acceptable for inclusion in clinical trials. Psychometric testing is ongoing. PSY53 ORGANIZATION OF ACUTE PAIN SERVICE (APS) IN THE HUNGARIAN HOSPITAL SETTING Németh F1, Brünner S1, Oláh A2, Pakai A2, Dér A2, Boncz I2, Gerdesics V2, Németh K2 1 Kanizsai Dorottya Kórház, Nagykanizsa, Hungary; 2University of Pécs, Pécs, Hungary OBJECTIVES: In our hospital nearly 5000 operations are performed yearly. Based on document analysis, the most frequently mentioned complaint was postoperative pain in 48 hours following the surgery. There was no Acute Pain Service (APS) in our institution, neither existed a standardized professional protocol for executing postoperative analgesia. Our aim was to organize APS, wide-spread all over the world but hardly known in Hungary. METHODS: Our longitudinal, qualitative examination was carried out in May 2008. We investigated the data of patients regarding pain and analgesia at five operating departments via not self-filled questionnaires and document analysis. Patients’ actual pain was measured before and 4th and 24th hours after the surgical intervention by visual analogue scale. The analgetics received after surgery were documented and compared to application patterns of each department. During evaluation T- and Chi-square probes, and variance analysis were used. RESULTS: Mean age of the examined sample (n = 130) is 52 years. Mean of the pain scale of examined patients is 3,98, higher than the internationally recommended, thus expected value of 3. Significant correlation was found (P < 0.05) between operating departments and applying frequency of analgetics used, as well as with mean values of the pain scale. Most often used medications are: noraminophenazon, tramadol and nalbuphin products. There is significant difference between pain values of patients operated at departments of surgery and laryngology (p = 0.048), urology and laryngology (p = 0.036), and surgery and traumatology (p = 0.014). CONCLUSIONS: A perceptual form has been developed for regular documentation of pain values measured during resting and physical activity, beside the vital parameters, medications exhibited and the side-effects. Acute Pain Service (APS) has been introduced in our institution and standardized professional protocol has been elaborated for performing postoperative analgesia in accordance with preventive approach. PSY54 THE DETERMINANTS OF HRQOL FOR PERSONS TAKING PRESCRIPTION PAIN MEDICATIONS: EVIDENCE FOR THE EUROPEAN UNION Langley PC1, Liedgens H2 1 University of Minnesota, Minneapolis, MN, USA; 2Gruenenthal GmbH, Aachen, Germany OBJECTIVES: This study assesses the quantitative impact of the experience of pain on HRQoL for persons taking prescription pain medications in the UK, France, Spain, Germany and Italy. METHODS: The study is based on data from the internet based 2008 National Health and Wellness Survey undertaken in the big 5 EU countries. This study identified some 4,503 respondents out of 11,419 who had experienced pain in 13th Euro Abstracts the last month and who had reported taking prescription pain medications. The assessment of the quantitative impact of pain status on HRQoL for the sub-group of pain respondents is estimated via three single equation generalized linear (ordinary least squares) models which estimate the impact of pain on PCS, MCS and utility scores. The model includes a range of variables which have been shown in previous population studies to impact HRQoL. These include: socio-demographic factors, health risk behaviors, comorbidity status, medication utilization, duration of medication utilization and satisfaction with care. The experience of pain is captured by a combination of severity and frequency categorical variables. RESULTS: The impact of pain severity and frequency on HRQoL for those taking prescription pain medications is substantial. Compared to mild pain experienced less than weekly, the impact of severe daily pain is to reduce the PCS score by −14.65 points, the MCS score by −4.60 points and utilities by −0.15. Severe pain 4–6 times a week has an impact of −0.11 on utility and daily moderate pain −0.08. Overall, the impact of pain is attenuated by reduced severity and frequency. The impact of pain on HRQoL is substantially greater than the impact of comorbidities, duration of medication utilization, socio-demographic and health risk factors. CONCLUSIONS: Persons reporting utilizing prescription pain medications continue to experience substantial deficits in the impact of pain severity and frequency on HRQoL in the major EU countries. PSY55 QUALITY OF LIFE OF OVERWEIGHT AND OBESE PATIENTS SEEKING CARE AT A PHARMACEUTICAL CARE SERVICE IN VENEZUELA Bastardo Y1, Alfonzo N2 1 Central University of Venezuela, Caracas, Venezuela; 2Proveeduria Farmacéutica IPP UCV, Caracas, Venezuela OBJECTIVES: The purpose of this study was to describe health-related quality of life (HRQL) of overweight and obese patients seeking care at the Pharmaceutical Care Service of the Proveeduría Farmacéutica IPP-UCV. METHODS: A convenience sample of 49 patients, ranging in age from 18 to 83 years was surveyed using a written questionnaire from February 2010 to May 2010. HRQL was measured using EuroQol comprising the health states descriptive system (EQ-5D) and visual analogue scale (EQ-VAS) as a general instrument. RESULTS: The sample consisted of 42 females and 7 males. The sample had a mean age of 48.55 years (s.d. 15.25 years). The mean weight of patients was 82.88 kg (s.d. 18.24). The mean body mass index (BMI) of the patients was 31.99 kg/m2 (s.d 5.77). Thirty-eight patients (77.6%) reported no exercising regularly. Thirty-four patients (69.4%) reported some problems with mobility. Seven patients (14.3%) reported some problems with self-care. Thirty-eight patients (77.6%) reported some problems with usual activity. Twenty-five (51%) patients reported some problems and 18 (36.7%) reported extreme problems with pain. Nineteen subjects (38.8%) reported extreme problems and 17 reported some problems with anxiety. CONCLUSIONS: Study limitations include the sample size and the use of a convenient patient sample. Overall, this exploratory study demonstrates that nearly all aspects of HRQL are adversely affected in overweight and obese patients seeking care at a pharmaceutical care service in Venezuela. PSY56 SYSTEMIC LUPUS ERYTHEMATOSUS (SLE): UNDERSTANDING THE BURDEN Schneider M1, Schmeding A2, Carnarius H3, Ager M4, McWade V4 1 Heinrich-Heine-University Düsseldorf, Düsseldorf, Germany; 2GlaxoSmithKline, Munich, Germany; 3GlaxoSmithKline, Hamburg, Germany; 4GlaxoSmithKline, London, UK OBJECTIVES: To assess knowledge of the burden of SLE on patients. METHODS: A Medline search was conducted to identify relevant articles published between 2000 and 2010 in English or non-English language. Search criteria were SLE/lupus plus one or more of the following terms: QoL; patient perspective; patient burden of illness/ disease; family impact/burden; prognosis; self; employment/work impact; patient(s) and psychological impact; patient(s) and physical impact; psychological impact; patient(s) and daily living/functionality; patient(s) and fatigue; impact on social life; patients and functioning; compared with rheumatoid arthritis/RA. Prospective studies involving ≥100 patients diagnosed with SLE were incorporated into the analysis. To focus on the burden of SLE in adult patients, studies in juvenile patients were excluded, as were economic analyses, studies relating to the development of HRQoL tools and studies of non-pharmacological interventions. RESULTS: The search identified 4244 articles, 62 of which met the criteria for incorporation into the analysis. Studies involved a mean of 460 patients with SLE (range: 100–4,603). SLE was shown to affect all aspects of patients’ lives, including physical and mental health, happiness and relationships. Common symptoms include fatigue (50–92%), pain (71–89%), sleep disturbance (56–88%) and neuropsychiatric symptoms (28–80%), which all influence HRQoL and work ability. Unemployment is highly prevalent among patients with SLE (26–54%) and can impact further on patients’ HRQoL. While most patients with SLE (94–100%) report unmet needs primarily reflecting physical, daily living and psychological concerns, physicians appear to place more emphasis on clinical and laboratory features. However, clinical measures of disease activity and organ damage are poor indicators of HRQoL. Few studies examined the effect of SLE treatment on HRQoL. CONCLUSIONS: SLE has a considerable impact on HRQoL and ADL. To improve understanding and raise awareness of the burden of SLE, further research is needed. 13th Euro Abstracts PSY57 WILLINGNESS TO PAY FOR OBESITY PHARMACOLOGY Doyle S1, Lloyd A1, Birt J2, Curtis B2, Godbey K2, Ali S3 1 Oxford Outcomes, Oxford, UK; 2Eli Lilly & Company, Indianapolis, IN, USA; 3University of York, York, UK OBJECTIVES: Several different treatments for obesity have been licensed for use by FDA/EMA, but health insurers and other payers typically refuse to support access to them. Thus, patients are left to bear significant out of pocket costs for obesity treatment. This study aims to assess preferences and willingness to pay (WTP) for obesity medications amongst patients seeking weight loss in the United States and United Kingdom. METHODS: An online WTP survey was developed based on a literature review, expert clinician interviews, and profiles of available obesity therapies. Obese participants recorded their demographics, weight history, and indicated their preference for hypothetical, paired-choice weight loss treatments which varied across 7 attributes. RESULTS: A total of 502 participants (N = 251 US; N = 251 UK) completed the survey (73.5% female; 47.7 years of age). The mean Body Mass Index of the sample was 37.12 kg/m2 (±4.63). All treatment attributes were significantly important except “time to noticeable weight loss” (P < 0.001). Corrected for purchasing power parity, the combined country sample was willing to pay £6.40 per month per % of body weight reduction; £0.68 per % reduction in long-term health risk; £18.12 for a one-pill-per-day treatment vs. a daily injection (£14.79 for a three pill per day regimen); £7.34 for a treatment that required only small changes in lifestyle versus large changes in diet and exercise (£29.37 for one that required no changes). Participants were also willing to pay £15.05 per month for a treatment that avoided gastrointestinal upset or a 5 point increase in blood pressure. CONCLUSIONS: Participants placed a lot of value on weight loss and avoiding changes to their lifestyle (i.e. diet/ exercise), and much less value on reducing long terms risks to health. Avoiding side effects was also valued highly. SYSTEMIC DISORDERS/CONDITIONS – Health Care Use & Policy Studies PSY58 AVAILABILITY OF AND ACCESS TO ORPHAN DRUGS: AN INTERNATIONAL COMPARISON OF PHARMACEUTICAL TREATMENTS FOR PULMONARY ARTERIAL HYPERTENSION, FABRY DISEASE, HEREDITARY ANGIOEDEMA AND CHRONIC MYELOID LEUKEMIA Blankart CR1, Stargardt T2, Schreyögg J3 1 Helmholtz Zentrum München, Muenchen, Germany; 2Helmholtz Zentrum München, Neuherberg, Germany; 3Munich University, Munich, Germany OBJECTIVES: To identify differences in the availability of orphan drugs and in patient access to them in eleven pharmaceutical markets: Australia, Canada, England, France, Germany, Hungary, the The Netherlands, Poland, Slovakia, Switzerland and the US. METHODS: Four rare diseases were selected for analysis: pulmonary arterial hypertension, Fabry disease, hereditary angioedema and chronic myeloid leukaemia. Indicators for availability were defined as the indications for which orphan drugs had been authorized in the treatment of these diseases, the application date and the date upon which these drugs received market authorization in each country. Indicators of patient access were defined as the outcomes of technology appraisals, the extent of coverage provided by health care payers and the price of the drugs in each country. RESULTS: Most drugs included had received market authorization in all countries, but the range of indications for which they had been authorized differed by country. The broadest range of indications was found in Australia, and the largest variations in indications were found for PAH drugs. Authorization process speed, the time between application and market authorization, was fastest in the US with an average of 362 days followed by the EU (394 days). The highest prices for the included drugs were found in Germany and the US, and the lowest in Canada, Australia and England. Although the prices of all of the included drugs were high compared to those of most non-orphan drugs, most of the insurance plans in our country sample provided 100%coverage for authorized drugs after a certain threshold. CONCLUSIONS: Although the present study showed some variations in selected indicators of availability and access to orphan drugs, virtually all of the drugs in question were available and accessible in our sample. Substantial copayments in the US and Canada, however, represent barriers to patient access, especially in the case of expensive treatments. PSY59 ACCESSIBILITY AND UTILIZATION OF BORTEZOMIB IN DUTCH DAILY PRACTICE Blommestein H, Franken M, Uyl-De Groot C Erasmus University, Rotterdam, The Netherlands OBJECTIVES: The Dutch government introduced in 2002 policy regulations to provide additional financing for expensive inpatient pharmaceuticals. In 2006 the policy was revised after signals of ‘zipcode prescribing’ of trastuzumab (Herceptin®). The aim of this study is to examine national and regional utilization of the expensive drug bortezomib (Velcade®) and investigate whether accessibility issues exist in Dutch daily practice. In 2004 bortezomib received market approval and in 2006 it was added to the expensive drug list for the indication of multiple myeloma. METHODS: Our longitudinal study investigated sales data (excluding trial use) from 2004 to 2009 and compared these figures with an estimation of the prevalence of patients with multiple myeloma. Prevalence was estimated from incidence figures of the The Netherlands Comprehensive Cancer Centres. We calculated the percentage of patients treated with A471 bortezomib at both national and regional level. In addition, interviews with hematologists were conducted investigating potential prescription barriers for bortezomib. RESULTS: At the national level, utilization gradually increased from treating 2% of the patients to 17%, indicating a long adjustment period. Furthermore, underutilization might still be present since utilization is lower than 27%, which is the estimated percentage of eligible patients by the professional group. In 2009, regional utilization ranged from 12% to 24%. Even though interviews revealed that hematologists were aware of the high treatment costs, they did not experience absolute prescription barriers. CONCLUSIONS: Utilization of bortezomib is increasing and accessibility has improved in the The Netherlands. However, the trend indicates a long adjustment period. Although hematologists did not experience absolute barriers, the indication of underutilization and residual regional differences might point towards accessibility issues in the The Netherlands, despite the policy revisions. PSY60 RESULTS FROM A COMPREHENSIVE LITERATURE REVIEW OF HUTCHINSON-GILFORD PROGERIA SYNDROME (HGPS): MORE RESEARCH IS NEEDED? Guri S, Kumar J, Ghasolia D, Sharma Y, Saini P, Baghla G JRDS ePharma Health Pvt Ltd, Dehli, India OBJECTIVES: Results from a comprehensive literature review of hutchinson-gilford progeria syndrome (HGPS): more research is needed? METHODS: Information was retrieved from websites of World Health Organization, Pubmed, clinicaltrials.gov and relevant grey literature using key words like Progeria, HGPS etc. Data regarding prevalence rate and current treatment options were obtained and complied. RESULTS: Progeria is an extremely uncommon genetic condition (ICD-10.E34.8), wherein symptoms resembling aspects of aging are manifested at early age. To our knowledge, only four clinical trials are in-process and results are waiting. No randomized clinical trial has been reported till date. The prevalence rate of Progeria is approximately 1 per 4 million births. a total of 150 cases have been reported since 1886 and 40 cases are living currently. The p.G608G LMNA mutation is the most commonly reported mutation. Around 95% of patients are Caucasians and male to female ratio is anticipated as 1.5:1. Symptoms of the disease start appearing at an age of 6–12 months. The patients demonstrate a life expectancy of average 13 years (range: 7–27 years). Death largely occurs due to cardiovascular abnormalities. New treatment options include Farnesyltransferase inhibitor (FTI) (ABT-100) which is presently under clinical study conducted by Progeria Research Foundation (PRF). FTI has shown positive results in previous studies on animals. Clinical trials are in process evaluating Lonafarnib, Zoledronic Acid, and Pravastatin in retarding the process of aging in Progeria patients. Growth hormone treatment in a few patients has also shown growth in height by 10% and in weight by 50%. Improvement in joint stiffness and BMR has been observed with Physical therapy and occupational therapy. CONCLUSIONS: The pace of research in this area has increased in recent years. Extensive genetic research can play a vital role in the early treatment of this disease. PSY61 RISK OF SURGICAL PROCEDURES IN THE INFLAMMATORY BOWEL DISEASE. ADMINISTRATIVE DATABASES AS A RESOURCE OF KNOWLEDGE Lepore V1, D’Ettorre A1, Graziano G1, Scaccianoce G2, Vacca M1, Tognoni G1 1 Consorzio Mario Negri Sud, Santa Maria Imbaro, Chieti, Italy; 2“Umberto I” Hospital, Altamura, Bari, Italy OBJECTIVES: To provide a comprehensive epidemiological profile of patients affected by Inflammatory Bowel Disease (IBD), and to assess the risk of preventable intestinal surgical procedures. METHODS: All cases of IBD identified in validated administrative databases of the Apulia region (about 4 millions inhabitants) in the period 1998–2006 were included in the study, and qualified by a record linkage process in terms of index clinical conditions and surgical procedures. Crohn’s disease (CD) and Ulcerative Colitis (UC), were considered separately in the analysis. Similarly, the surgical procedures coded according to the ICD9-CM were analyzed as a whole group, as well as separately for small and large intestine and rectum. Multivariate logistic regression was used to assess the risk to undergo surgery. Results were reported as Odds Ratios (OR) with 95% confidence intervals. RESULTS: We identified 11,991 cases of IBD (4,279 and 7,712 for CD and UC, respectively). Intestinal surgery was more frequent in patients aged 18–64 years, in males and in subjects with CD. In the multivariate logistic regression the risk of any intestinal surgical procedure was higher for CD then UC patients (83%), in males vs. females (41%) and in 40–64 age class vs. older people (30%). Only in the case of UC, the youngest age class (0–17 years) had a major risk to undergo surgery of rectum (p-values 0.001). Furthermore, in the youngest age group (0–17 years) the difference between CD and UC risk to undergo any intestinal surgery was much less than adult age classes. No important differences were detected in the risk of at least one intestinal operation across the Apulia. CONCLUSIONS: The routine availability of a comprehensive, detailed, reliable system of record-linked databases allows to monitor timely risk of severe surgical procedures thus favoring the adoption of well targeted and updated health policy measures. A472 PSY62 HEALTH TECHNOLOGY ASSESSMENT FOR RARE DISEASES: A MARKOVIAN MODEL FOR FABRY DISEASE (FD) CONSIDERING ENZYME REPLACEMENT THERAPY(ERT) Vinhas de Souza M1, Schwartz I2 1 Universidade Federal do Rio Grande do Sul/UFRGS, Porto Alegre, RS, Brazil; 2Universidade Federal do Rio Grande do Sul/UFRGS and HCPA, Porto Alegre, RS, Brazil OBJECTIVES: Evaluate, using a markovian probabilistic model, the development of renal disease in Fabry Disease(FD), comparing the probability of developing the different stages of renal disease in those with FD using Enzyme replacement therapy (ERT) with those not using ERT. METHODS: A model with pre-defined 3 possible stages was used for patients with FD and renal dysfunction. Progression to a different stage or having a stable disease were possible. Less severe ‘stages’ could only migrate to more severe ‘stages’. Three year time cycles were used. The probabilities ERT x without ERT were compared. RESULTS: Probabilities between the 2 arms: 1)The probability of having proteinuria only and being stable in 3 y was 56% in those treated versus 48% in those untreated; 2)Also for those with proteinuria only, with ERT the probability of progressing to renal impairment in 3 y was reduced, to 24% compared to 32% in those untreated; and 3)For those receiving ERT already with renal impairment, but not in dialysis, the probability of postponing the dialysis in 3 y time was not modified. CONCLUSIONS: This model did show a relevant improvement with ERT for those that were treated at initial stages of renal disease (the subgroup with proteinuria only). For this group the probability of remaining in a less severe stage of the disease in a 3 y time, postponing renal impairment was significant for those using ERT. This was only a preliminar model, considering that the maintenance of less severe states of renal disease postpone the need for dialysis in this sub-group, aspects such as quality of life will be added in order to do a more complete evaluation. This model was build considering clinical outcomes, something new on rare diseases. The discussion on how to evaluate and on the value of ‘classical modeling’ in rare diseases are still an on going issue with a lot to be elucidated. PSY63 WORK AND DAILY ACTIVITY IMPAIRMENT IN PATIENTS WITH DIGITAL ULCERS (DUS)—RESULTS FROM THE DUO REGISTRY Hunsche E1, Denton C2, Krieg T3, Guillevin L4, Schwierin B1, Matucci-Cerinic M5 1 Actelion Pharmaceuticals Ltd., Allschwil, Switzerland; 2Royal Free Hospital, London, UK; 3 University of Koln, Koln, Germany; 4University Paris, Paris, France; 5University of Florence, Florence, Italy BACKGROUND: Digital ulcers (DU) are difficult to heal, and can lead to severe, disabling complications. The DUO Registry is an international, multi-centre, observational study to assess outcomes in patients with DU disease associated with Scleroderma. OBJECTIVES: To assess the impact of DU on patients’ work, daily activities, and dependency. METHODS: A work and daily activities assessment questionnaire was completed by the patients at enrollment visit into the DUO registry from April 2008 to April 2010. An overall score measuring the impairment in daily activities and in work was calculated. In addition, the number of hours of help needed was determined. Overall mean (SD) as well as stratified analyses by number of DU were performed. RESULTS: At the cut-off date for this analysis, 1770 questionnaires were received from patients enrolled in the registry. The questionnaire was completed by 71% (n = 1250), of which 35% patients were employed. The majority of patients completing the questionnaire were female (82%) with a mean age of 54 years. Problems associated with DU reduced the ability of patients to perform regular daily activities by 52% (SD 30). More than 50% of the patients needed help due to their DU during the previous month, with an average of 7 hours (SD 29) of paid help and 36 hours (SD 101) of unpaid help. The productivity of patients working was reduced by 40% (SD 31). Patients with one or more DU present at enrolment and completing the questionnaire (72% of all completers) were more impaired in their daily activities (38%, 52%, 55%, 63% for patients with 0, 1, 2, ≥3 DUs, respectively) and work (32%, 40%, 41%, 47% for patients with 0, 1, 2, ≥3 DU, respectively). CONCLUSIONS: DU impose a significant burden on patients’ by impacting their daily life and ability to work as shown in patients enrolled in the DUO Registry. PSY64 SEQUENCE OF TREATMENT IN IMMUNE THROMBOCYTOPENIA (ITP) PATIENTS: RESULTS OF A MEDICAL RECORD REVIEW FROM EIGHT HOSPITALS IN SPAIN Sanz MÁ1, Jarque I1, Yuste VJ2, Julià A3, Gómez RV4, García CG5, Mateos MV6, Montero EO7, Páramo JA8, Mathias SD9 1 University Hospital La Fe, Valencia, Spain; 2University Hospital La Paz, Madrid, Spain; 3 University Hospital Vall d’Hebron, Barcelona, Spain; 4C. H. U. a Coruña, a Coruña, Spain; 5 H. U. 12 de Octubre, Madrid, Spain; 6H. Clínico Salamanca, Salamanca, Spain; 7H. U. Germans Trias i Pujol, Badalona, Spain; 8Clínica Universidad de Navarra, Pamplona, Spain; 9 Health Outcomes Solutions, Winterpark, FL, USA OBJECTIVES: To better understand current practice patterns and determine how current practice fits recently published guidelines, a treatment sequencing analysis of ITP patients in Spain was conducted. METHODS: A retrospective chart review of patients diagnosed with ITP was undertaken at 8 hospitals in Spain, collecting demographic and clinical data, drug treatment (up to fifth line) and its duration. Thrombopoetin-receptor agonists were not available at the time of the review. RESULTS: Data on 60 patients [90% (n = 54) with baseline platelet counts <30 × 109/L and 10% (n = 6) <50 × 109/L with bleeding episodes] were available. Patients were 54.4 ± 20.7 13th Euro Abstracts (SD) years of age, 63% female, and 6.0 ± 7.2 (SD) years since diagnosis. The most common first-line treatment was corticosteroids (n = 39, 65%), followed by intravenous immunoglobulins (IVIg; n = 19, 32%). The mean duration of first-line treatment with corticosteroids was 28.2 ± 39.0 (SD) days. The most common sequence of treatment for patients initially treated with corticosteroids was administration of four additional courses with corticosteroids (18/39, 46%), followed by alternation between IVIg and corticosteroids until the fifth-line of treatment (7/39, 18%). Of the 19 patients with initial IVIg, 5 (26%) received corticosteroids in the 4 remaining treatments, 5 (26%) continued with IVIg, and 4 (21%) alternated corticosteroids and IVIg. In total, 41 patients (68%) received ≥2 treatments with corticosteroids and 10 (17%) received treatments other than corticosteroids and IVIg (3 patients received azathioprine, 3 rituximab, 2 platelet transfusions, 1 anti-D immunoglobulin, 1 danazol and 1 mycophenolate mofetil). CONCLUSIONS: Patterns of treatment of ITP in Spain usually followed recently introduced recommendations from international consensus guidelines. However, in most patients, corticosteroid treatment was given repeatedly, which exceeds current recommendations, i.e., rapidly tapering corticosteroid dose and stopping after 4 weeks. Future research is needed with a larger sample size, to explore the place of splenectomy in treatment sequencing, and better understand the role of combination therapy. SYSTEMIC DISORDERS/CONDITIONS – Conceptual Papers & Research on Methods PSY66 WEB VERSUS FACE-TO-FACE (FTF) ADMINISTRATION OF A HEALTH UTILITY SURVEY IN THE GENERAL PUBLIC: RESULTS FROM A TIME TRADE-OFF (TTO) SURVEY ON IDIOPATHIC THROMBOCYTOPENIC PURPURA (ITP) Szende A1, Schaefer C2, Brazier J3 1 Covance, Leeds, West Yorkshire, UK; 2Covance, Gaithersburg, MD, USA; 3University of Sheffield, Sheffield, UK OBJECTIVES: Web-based administration of health utility interviews offers the potential to recruit larger, more representative samples with reduced time and costs compared to FTF administration. This analysis compared health utilities elicited through web vs. FTF administration. METHODS: Six distinct ITP health states were included in a TTO and visual analogue scale (VAS)-based health utility valuation, which was administered FTF (n = 63) and via web (n = 359) to members of the UK general public. The Wilcoxon rank-sum test was used to compare utilities between methods. The interaction between administration method and respondent characteristics was assessed by regression analyses on each pair of health utilities. An additional analysis of exclusion criteria was conducted from least strict to most strict. RESULTS: Demographic characteristics in the FTF and web survey were generally comparable to the UK general population 2001 census data. The mean time to complete the TTO survey was 10.2 minutes in the FTF and 9.9 minutes in the web survey. Valid TTO response rates were higher in the FTF sample (85% to 96%) compared to the web sample (58% to 80%) across health states. Higher proportions of web respondents reported that the TTO exercise was ‘very’ or ‘somewhat unclear’ (17 % vs. none) or that all or most decisions were difficult to make (41% vs. 30%) compared to the FTF sample. Utilities were statistically significantly lower in the web vs. the FTF survey (P < 0.05). TTO scores were sensitive to the selection of exclusion criteria. The variation in TTO scores was smaller among FTF vs. web respondents. VAS ratings were similar across the two administration methods and less sensitive to exclusion criteria selection. CONCLUSIONS: Our study highlighted trade-offs between the advantages and challenges of web administration. More research is warranted to further improve data quality in web-based utility surveys. PSY67 PREDICTING EQ-5D UTILITIES FROM NEUROPATHIC PAIN SCORES: COMPARING INDIRECT MAPPING OF PREDICTED ITEM RESPONSES WITH DIRECT MAPPING OF SCORES Gu NY1, Bell C2, Ji X1, Carter J1, Botteman M1, Van Hout BA3 1 Pharmerit North America, LLC, Bethesda, MD, USA; 2GlaxoSmithBell, Research Triangle Park, NC, USA; 3University of Sheffield, Sheffield, UK OBJECTIVES: To predict EQ-5D utilities from neuropathic pain scores using indirect mapping of pain scores to EQ-5D item responses subsequently converted into utilities or direct mapping of pain scores into utilities. METHODS: Mappings were based on baseline data from three longitudinal surveys of adults (n = 2,719) who had ≥3 months of painful diabetic peripheral neuropathy (pDPN) or post-herpetic neuralgia (PHN), were receiving pain medications, and completed EQ-5D and pain questionnaires. In indirect mapping, ordered logit regression was used to predict and simulate EQ-5D responses using the following predictors: age, gender, and pain scores ranging from 0 (“no pain”) to 10 (“pain as bad as you can imagine”). Utilities were computed based on predicted responses using a U.S. algorithm. In direct mapping, OLS regression was used to directly predict utilities using the same predictors. Cross-validations were conducted separately in pDPN and PHN respondents. Comparisons were made between actual and estimated values on mean utilities, mean square/absolute errors (MSE/MAE). RESULTS: Both estimated utilities were consistent with actual values along the increment/decrement of the pain scores. Direct mapping explained 29% of the variance and had an estimated mean utility close to the observed data [0.594 (MSE = 0.31; MAE = 0.148)]. Indirect mapping resulted in lower mean utility [0.588 (MSE = 0.054; MAE = 0.184)] but its distribution was more consistent with the actual values. 13th Euro Abstracts Higher MSEs found in indirect mapping were mostly pronounced in lower utilities (<0.2). Predicted utilities were slightly higher than actual values when population average was used for input (0.3–3%). Similar findings were noted in sub-samples. CONCLUSIONS: Both methods produced robust results. Compared to direct mapping, indirect mapping better represents the EQ-5D’s descriptive information, although with higher MSE/MAE. This research provides algorithms for estimating EQ-5D item responses and utilities on the basis of pain scores in absence of direct utility evidence. Further investigation using out-of-sample predictions is encouraged. PSY68 HEPATITIS C-SPECIFIC QUALITY OF LIFE IS NOT PROPERLY MEASURED BY EXISTING INSTRUMENTS Duracinsky M, Armstrong AR, Lalanne C, Chassany O Assistance Publique-Hopitaux de Paris, Paris, France OBJECTIVES: Measures commonly used in Hepatitis C (HCV) Health-Related Quality of Life (HRQL) research were investigated to determine whether they effectively encompass the disease-specific experience of HCV. METHODS: Important HCV HRQL dimensions were identified via a review of qualitative and quantitative literature. Medline and Embase were interrogated for appropriate studies. Quality and relevance of HCV HRQL studies were assessed against FDA and EMEA guidelines. The review facilitated construction of a conceptual framework of HCV HRQL. The framework was compared against HRQL measures used in HCV research, including the SF-36, Hepatitis Quality of Life Questionnaire (HQLQ), Chronic Liver Disease Questionnaire (CLDQ), Liver Disease Symptom Index (LDSI 2.0), and Hepatitis B Quality of Life instrument (HBQOL). RESULTS: Numerous dimensions encompassing the HCV HRQL experience were not adequately represented by common measures. Absent from the measures were consideration of treatment adherence and management, and management of side effects; HIV/HCV co-infection issues; drug addiction; resilience and coping; contagiousness and transmission-related issues; illness uncertainty and unpredictability; and changes in body image. HCV dimensions needing further attention were liver disease severity; psychiatric complications including emotional volatility; cognitive impairment during daily activity; sexual dysfunction; the multidimensional nature of stigma; and fatigue variability. CONCLUSIONS: The disease-specific experience of HCV is not fully addressed by any single existing measure. An HCV HRQL instrument that is sensitive to the identified dimensions and issues would be of considerable benefit. Such a measure would help health care providers plan individual interventions for problematic HRQL domains, as well as improve patient monitoring during treatment and clinical research trials, and contribute to determining the value and efficacy of treatment programs. URINARY/KIDNEY DISORDERS – Clinical Outcomes Studies PUK1 CONTRAST-INDUCED NEPHROPATHY IN PATIENTS WITH CHRONIC KIDNEY DISEASE UNDERGOING COMPUTED TOMOGRAPHY: A COMPARATIVE SAFETY META-ANALYSIS OF RANDOMIZED CONTROLLED TRIALS Catalá-López F1, Arana E2 1 Spanish Medicines and Healthcare Products Agency (AEMPS), Madrid, Spain; 2Fundación Instituto de Investigación en Servicios de Salud, Valencia, Spain OBJECTIVES: We conducted a systematic review and meta-analysis comparing the incidence of contrast-induced nephropathy (CIN) in patients at risk with chronic kidney disease (CKD) undergoing computed tomography (CT). METHODS: Randomized controlled clinical trials designated to evaluate the nephrotoxicity related to iso-osmolar contrast media (IOCM) compared to low-osmolar contrast media (LOCM) were searched in the following electronic databases: PubMed/MEDLINE, EMBASE, ISI Web of Knowledge and Virtual Health Library (BVS-BIREME), as well as abstracts presented at related scientific societies meetings. Prior to data extraction, definitions of nephrotoxicity were established. We applied a random effects model of DerSimonian and Laird, with heterogeneity (Q statistic), publication bias (Egger and Begg test) and sensitivity analyses. RESULTS: Five studies were included with 716 randomized patients. When CIN was defined as increased serum creatinine (SCr) ≥ 25%, the relative risk (RR) was 0.71 (CI95% 0.40 to 1.26)—in favor of IOCM—and when it was defined as SCr ≥ 0.5 mg/dL it showed a RR 1.48 (CI95% 0.37 to 5.87)— favoring LOCM—in the four studies used this criterion. CONCLUSIONS: In patients with CKD undergoing CT there is a similar risk of CIN with the administration of any contrast media studied. CIN incidence depends on the chosen criteria and is lower with the definition of SCr ≥ 0.5 mg/dL at 24–72 h. PUK2 COMORBIDITY EFFECT ON HOSPITAL READMISSION RATES IN PATIENTS WITH RENAL FAILURE Rhee Y1, Walker DR2, Inglese G2 1 Northwestern University, Chicago, IL, USA; 2Baxter Healthcare Corporation, McGaw Park, IL, USA OBJECTIVES: To examine the effects of comorbidities on the hospital readmission rate in renal failure (RF) patients on dialysis. METHODS: We used 2005–2007 U.S. MarketScan claims database to identify RF patients under 64 years old. RF dialysis patients were identified using ICD-9 and CPT codes. RF patients had to be continuously eligible for at least 6 months after the initial dialysis diagnosis date (index dates). A473 Comorbidity scores were measured by the Charlson Comorbidity Index (CCI). We identified readmission rates to the hospital within 15 days after the index date. ANOVA tests and logistic regression were performed to compare outcomes by CCI. RESULTS: A total of 6117 patients were diagnosed with RF. Among those who initiated dialysis treatment, 13.9% and 86.1% were treated with peritoneal dialysis and hemodialysis, respectively. Over half of all patients were male (55.2%) and the average age was 52.4 years. The most frequent comorbid conditions were hypertension (69.1%), diabetes (34.5%), congestive heart failure (CHF) (26.3%), and anemia (26.4 %). Most patients had low (78%, score < = 3) or moderate (12.0%, score 4–5) CCI scores. 7.2% of patients had a score of 6–7 and 125 (2.0%) had a score > 8. Older patients had significantly more severe CCI scores (P < 0.001). The average age in the moderate and severe co-morbidity groups was 51.7 and 57.4 years old, respectively. a total of 1.3% of RF dialysis patients were readmitted to the hospital within 15 days. Their readmission rates were significantly associated with the CCI score (P < 0.001). Patients over age 50 were significantly more likely to readmit to the hospital (P < 0.001). Those with a higher CCI score were more likely to readmit to the hospital within 15 days (moderate group OR. 3.0, P < 0.001, very high group OR 4.04, P < 0.001). CONCLUSIONS: For patients with RF, the presence of severe comorbid physical conditions was significantly associated with hospital readmission within 15 days. PUK3 ASSESSMENT OF COMORBIDITIES IN PATIENTS WITH OVERACTIVE BLADDER (OAB) DISORDER: AN ELECTRONIC MEDICAL RECORD (EMR) DATA ANALYSIS Asche C1, Kim J2, Chakravarti P3, Andersson KE4 1 University of Utah College of Pharmacy, Salt Lake City, UT, USA; 2University of Utah, Salt Lake City, UT, USA; 3Novartis Pharmaceuticals, East Hanover, NJ, USA; 4Wake Forest University Baptist Medical Center, Winston-Salem, NC, USA OBJECTIVES: To compare OAB patients to non-OAB patients by assessing their pre-existing comorbidities prior to their OAB diagnosis or first OAB treatment. METHODS: This retrospective cohort study used the General Electric (GE) Centricity EMR database. The study subjects were from between January 1, 1996 to March 30, 2007. The index date for OAB patients was defined as the date of their first prescription for an antimuscarinic agent or a diagnosis for OAB identified by ICD-9 codes. The index date of non-OAB subjects without diagnosis or pharmacy claim was defined as a year after the first activity date in the EMR. Subjects ≥18 years old were included and had 395 days of continuous enrollment before and after the index date. Non-OAB subjects were matched to OAB subjects on 1:1 propensity score matching based on age, body mass index (BMI) and gender at baseline. Two linear regressions were constructed using the outcome variables of the Charlson Comorbidity Index (CCI), using ICD-9 codes, and the Chronic Disease Score (CDS), using prescribed drugs, respectfully. RESULTS: There were 38,739 OAB subjects [mean age 61.18 (SD:13.26) years; 85.67% women] and 38,739 matched non-OAB subjects [mean age 61.17 (SD: 13.24) years; 85.70% women]. Patients with OAB had higher mean CCI and CDS than subjects without OAB [(CCI: 1.17 vs. 1.11 (p-value < 0.001); CDS: 2.95 vs. 1.74 (p-value < 0.001)]. After controlling for other covariates, the linear regressions (n = 22,544) showed that OAB patients had higher CCI and CDS than subjects without OAB by 0.037 (p-value < 0.001) and by 0.881 (p-value: < 0.001), respectively. CONCLUSIONS: This study determined that pre-existing comorbidities were more prevalent in OAB patients than in non-OAB patients. These comorbidities should to be taken into account when making the decision on the most appropriate treatment option for each individual patient. PUK4 A SYSTEMATIC REVIEW OF IMMUNOSUPPRESSIVE REGIMENS IN LOWER IMMUNOLOGICAL RISK RENAL TRANSPLANT RECIPIENTS Kang MH, Kim HJ, Ko RK, Ko SK Pfizer Pharmaceuticals Korea Ltd., Seoul, South Korea OBJECTIVES: In this study, we conducted a systematic review of three immunosuppressive regimens in lower risk renal transplantation and compared their efficacy. METHODS: MEDLINE were searched and two independent reviewers assessed studies. We limited the search to English, Randomized controlled trial, Human and publication between January 1, 1999~May 31, 2010. Studies which were conducted with adult renal transplant patients (>18 years) with lower immunological risk were included. Sirolimus + steroid regimen with 3 months cyclosporine, and CNI (Calcineurin inhibitor: cyclosporine or tacrolimus) regimen with MMF and steroid were eligible for inclusion. a total of 434 studies were retrieved from MEDLINE firstly. By reviewing title, abstracts and full text, 10 studies were eligible for the inclusion finally. As outcome data, we extracted patient survival and graft survival. RESULTS: There was no head-to-head clinical trial which compared the three regimens concurrently. But results from 2 studies shows trend of higher patient and graft survival with low dose CNI regimen than standard dose cyclosporine. ELITE-Symphony Study illustrated patient and graft survival of 96.5% and 89.3% for standard dose cyclosporine, 98.2% and 93.1% for low dose cyclosporine, 97.2% and 94.2% for low dose tacrolimus at 1 year. And CAESAR Study showed 97.1% and 92.4% for standard dose cyclosporine, 97.8% and 94.5% for low dose cyclosporine at 1 year. 5 reports from ‘Rapamune Maintenance Regimen Study’ showed outcomes for sirolimus based regimen during 5 years. At 1 year, patient and graft survival were 98.1% and 97.2% and decreased thereafter. One study of tacrolimus based regimen reported 2-year A474 patient and graft survival of 92.9% and 92.4% for biopsy arm, 97.9% and 94.1% for control arm. Both of the treatment arms were standard dose tacrolimus regimen. CONCLUSIONS: Three of immunosuppressive regimens are considered to have similar efficacy in short-term patient and graft survival in lower immunological risk renal transplant patients. PUK5 CLINICAL AND DEMOGRAPHIC CHARACTERISTICS OF PATIENTS IN RENAL REPLACEMENT THERAPY IN SÃO PAULO, BRAZIL Takemoto ML, Fernandes RA, Passos RB, Tolentino ACM, Takemoto MM, Moretti AIP, Cukier FN ANOVA—Knowledge Translation, Rio de Janeiro, RJ, Brazil OBJECTIVES: This study aims to describe clinical and demographic characteristics of end-stage renal disease (ESRD) patients on renal replacement therapy in the São Paulo state, Brazil. METHODS: Cross-sectional analysis of São Paulo renal replacement therapy claims as reported in Brazilian Ambulatory Information System (SIA/ DATASUS) database in January 2009. Repeated records were excluded using identification code, age, sex, and first treatment date as compatibility criteria. The following variables were investigated: diseases associated to ESRD, anemia, glucose levels, HCV, HbSAg, HIV, urea reduction ratio (URR), vascular access, and type of renal replacement therapy. RESULTS: A total of 18,360 patients were identified among 18,886 available claims, with a mean age of 54.23 years (SD = 15.53) and 57.7% male. Among 8,305 patients for whom secondary ICD-10 codes in addition to those related to renal failure itself were available, the more frequent conditions associated to ESRD were diabetes mellitus (17.5%), hypertension (26.4%) and glomerulonephritis (8.81%). Continuous Ambulatory Peritoneal Dialysis (CAPD) was the therapeutic strategy for 9.9% of patients compared to hemodialysis in 90.1%. 58.4% of all patients had hemodialysis vascular access and 50.9% had URR > 65%. Prevalence of positive HCV, HbSAg and HIV serology tests was 4.3%, 0.9% and 0.6%, respectively. Anemia and glucose levels > 126 were present for 45.4% and 19.6% of patients. The total amount paid for renal replacement therapy procedures in January 2009 in São Paulo state was 36,073,377 BRL. CONCLUSIONS: Although diabetes and hypertension renal complications can be prevented, they accounted for 43.9% of all conditions related to renal disease among the studied population. The findings also indicate that CAPD is still underused in Brazilian health care public system compared to other countries, as reported by previous local studies. PUK6 POST RENAL TRANSPLANT PATIENT SURVIVAL AND THE COMBINED ADVERSE EFFECT OF POOR RENAL FUNCTION AND NEW ONSET DIABETES MELLITUS McEwan P1, Baboolal K2 1 CRC, Cardiff, UK; 2Cardiff and Vale NHS Trust, Cardiff, UK OBJECTIVES: A major challenge in the field of renal transplantation is to prolong patient survival. Previous studies have demonstrated that renal function at 1 year is a major determinant of long term patient survival. Furthermore, the development of new onset diabetes after transplantation identifies patients at high risk of cardiovascular events and mortality. This study aims to quantify any additive effect of impaired renal function and elevated glucose levels at 1 year post transplant on patient mortality. METHODS: Consecutive renal transplants from a single UK transplant centre over a 10 year period were analyzed using multivariate logistic regression in SPlus 8; the risk of patient mortality was assessed after stratification by renal function (measured by glomerular filtration rate (GFR)) and impaired fasting glucose levels. RESULTS: Data were available on 307 patients with fasting glucose and GFR measurements at 1 year post transplant. Overall 37% (n = 114) had a GFR of less than 40 mL/minute and of these, 24% (n = 27) had fasting glucose levels greater than 7 mmol/L. After adjusting for age, sex and donor factors patients with fasting glucose greater than 7 mmol/L and a GFR less than 40 mL/minute had a mortality odds ratio of 5.47 (P < 0.01) compared to those with glucose levels less than 5.6 mmol/l and a GFR greater than 40 mL/minute. CONCLUSIONS: This study demonstrates that the development of impaired fasting glucose post transplants is associated with a 35% increase risk of mortality and the development of new onset diabetes associated with a 2-fold increased risk. Our study further suggests that there is a negative synergistic effect of deteriorating renal function and progressive impaired glucose regulations on patient survival. Consequently, therapeutic strategies that could both improve GFR at one year and the incidence of diabetes might be expected to improve long term patient survival. URINARY/KIDNEY DISORDERS – Cost Studies PUK7 THE PHARMACY BUDGET IMPACT OF EXTENDING REIMBURSEMENT OF LANTHANUM CARBONATE TO TREATMENT OF HYPERPHOSPHATEMIA (>1.78 MMOL/L) IN PATIENTS WITH CHRONIC KIDNEY DISEASE PRE-DIALYSIS IN FRANCE AND THE UNITED KINGDOM Campbell J1, Tao CY1, Keith MS2, Leahy KJ1, Russo L2 1 i3 Innovus, Medford, MA, USA; 2Shire Pharmaceuticals, Wayne, PA, USA OBJECTIVES: To examine the pharmacy budget impact (PBI) of extending reimbursement of non-calcium lanthanum carbonate (LC) to treatment of hyperphosphatemia 13th Euro Abstracts (serum phosphorus > 1.78 mmol/L) in patients with chronic kidney disease pre-dialysis (CKD-ND) in France and the UK over five years. METHODS: The treated prevalence of CKD-ND and hyperphosphatemia, and the use of pharmacologic therapies, were estimated using published literature and EU nephrologist surveys. Drug costs were estimated from published sources. Market share changes were estimated from market research. Base-case analyses assumed complete medication compliance. Annual PBI was calculated as the difference in total drug costs between scenarios with and without the reimbursement extension for LC. Alternate scenarios and deterministic sensitivity analyses were also estimated. RESULTS: A small percentage of CKD-ND patients, 14,500 (7%) and 64,600 (14.6%), are estimated as hyperphosphatemic in France and the UK, respectively. Of these patients, 81% in France versus 30% in the UK are estimated to receive phosphate binder therapy. CKD-ND market share for LC is estimated at 12% in France and 1–2% in the UK. The label extension, adding hyperphosphatemic CKD-ND patients, is estimated to increase LC use primarily at the expense of calcium-based phosphate binders. The annual PBI from the label extension is estimated to grow over Years 1–5 from 0 to c1.1 M in France; and to be <£25,000 in Years 1–5 in the UK. Results are most sensitive to LC market share changes postlabel extension. CONCLUSIONS: The number of CKD-ND patients with hyperphosphatemia (>1.78 mmol/L) eligible for LC treatment is minimal in France and the UK. Assuming complete compliance, the annual PBI after 5 years of a label extension for LC to CKD-ND patients is estimated to be c1.1 M in France and <£25,000 in the UK. Calcium is the predominant therapy in CKD-ND; however, adding LC may result in a low pharmacy budget impact. PUK8 BUDGET IMPACT ANALYSIS OF ALISKIRENO IN TYPE 2 DIABETES PATIENTS WITH HYPERTENSION AND NEPHROPATHY IN THE MEXICAN INSTITUTE OF SOCIAL SECURITY García-Contreras F1, Nevarez A1, Olvera K2 1 Mexican Institute of Social Security, Mexico, D.F., Mexico; 2Novartis, Mexico, D.F., Mexico OBJECTIVES: To determine the budget impact of Losartan + Aliskiren versus Losartan in type 2 diabetes patients with hypertension and nephropathy in the Mexican Institute of Social Security (IMSS). METHODS: An annual budget impact analysis of the use of aliskiren in the treatment of diabetic nephropathy in the IMSS was made considering four scenarios: a) Losartan + Aliskireno; b) Losartan; c) real prevalence of diabetic nephropaty and d) estimated prevalence of diabetic nephropaty. The source of epidemiologic data was the Health National Survey (ENSA 2006). Information source to estimate health benefits was AVOID study. Costing was performed through the technique of microcosting for replacement therapy with hemodialysis and peritoneal dialysis. The use of resources was the standard IMSS supportive care, costs are in 2009 USD. There was no discount rate considered due to annual temporal horizon. a probabilistic and non-probabilistic sensitivity analysis was performed. RESULTS: Aliskiren alternative prevents the use of dialysis in the real scenario of 5,946 patients, with expected savings of $ 95,461,538.46 USD, while in the estimated alternative prevents dialysis in 11.765 with expected savings of $188,846,153.85 USD. CONCLUSIONS: Aliskiren use investment prevents up to 11,765 patients requiring dialysis with savings for the health sector of more than 188 million USD. PUK9 REAL-LIFE DOSING OF ERYTHROPOIESIS-STIMULATING AGENTS (ESAS) IN CHRONIC KIDNEY DISEASE (CKD)-ASSOCIATED ANEMIA: BUDGET IMPACT Walsh CM1, Bilton R1, Robinson JE2 1 Complete Medical Group, Macclesfield, UK; 2BioTrends Research Group, Exton, PA, USA OBJECTIVES: Effective management of chronic kidney disease (CKD)-associated anaemia has been established for all erythropoiesis-stimulating agents (ESAs). The aim of this study was to establish the real-life doses and comparative costs of both shortacting ESAs and the newer longer-acting ESAs, darbepoetin alfa and the continuous erythropoietin receptor activator (C.E.R.A), and to forecast the impact of a change in current prescribing practice on health care budgets in Europe. METHODS: a retrospective chart review of CKD-associated anaemia management in 1014 pre-dialysis and 1016 dialysis patients was carried out by 261 nephrologists across the UK, Italy, Germany, Spain and France, between August 2008 and September 2009. Data collected on ESA dosing were used to populate a UK budget impact model comparing current prescribing practice with specific focus on C.E.R.A and darbepoetin alfa, over a five-year time horizon. Data on the number of patients, ESA market shares, and unit drug costs were derived from UK sources. RESULTS: The average monthly doses of ESAs in the combined five European markets (5EU) were as follows: epoetin alfa (pre-dialysis, 19,350 International Units (IU); dialysis, 35,404 IU); epoetin beta (predialysis, 18,230 IU; dialysis, 36,789 IU); darbepoetin alfa (pre-dialysis, 107 µg; dialysis, 169 µg); C.E.R.A (pre-dialysis, 98 µg; dialysis, 150 µg). The model based on UK doses showed, that after five years, an increase in the use of C.E.R.A to 40% would reduce the overall ESA budget for CKD-associated anaemia by 15% compared to current prescribing practice. CONCLUSIONS: Based on real-life data, an increase in C.E.R.A prescribing may result in a cost-saving for the management of CKD-associated anaemia in the UK. This model can be applied to other European markets to forecast the local impact of a change in current prescribing practice. 13th Euro Abstracts PUK10 GENERIC AZATHIOPRIN FOR KYDNEY TRANSPLANT PATIENTS— ANALYSIS OF COST SAVINGS Georgieva SS1, Stoimenova A2, Petrova GI2 1 Aleksandrovska Hospital, Sofia, Bulgaria; 2Medical University, Faculty of Pharmacy, Sofia, Bulgaria OBJECTIVES: To analyze the cost savings after an introduction of the generic version of azathioprin for patients with transplanted kidney. METHODS: Prospective observational study of the cost of pharmacotherapy of patients was performed during 2009–2010. It was collected information for 121 patients previously treated with the originator and transferred to the generic medicines. The patient sample was systematized according to patient age, gender, additional medicines used for main therapy, and monthly cost of pharmacotherapy. RESULTS: The average monthly cost of the combined immunosuppressing therapy when the original product was included was c320.99 and after the introduction of the generic version it became c311.29. On the other hand the monthly cost of the therapy only with azathioprin changed from c6.71 to c5.78 per patient per month. The patients’ number varies among 121 to 96 during different months due to the drop out. After the introduction of the generic version 7 patients were switched to another immunosuppressive agent, while for the originator the corresponding figure is 2 patients. For the switched patients the cost of pharmacotherapy did not increase. CONCLUSIONS: In spite of the contradictory introduction of generic versions of immunosuppressors, due to their narrow therapeutic index the drop out of the patients is not higher and the savings for the health care system are possible. PUK11 A COST EVALUATION OF PERITONEAL DIALYSIS AND HEMODIALYSIS IN THE TREATMENT OF ESRD IN SÃO PAULO, BRAZIL Ferraz MB1, Mendes de Abreu M2, Walker DR3, de Castro Cintra Sesso R1 1 Universidade Federal de São Paulo, Sao Paulo, SP, Brazil; 2Universidade Federal de Sao Carlos, Sao Carlos, SP, Brazil; 3Baxter Healthcare Corporation, McGaw Park, IL, USA OBJECTIVES: ESRD patient survival is similar for hemodialysis (HD) and peritoneal dialysis (PD). In Brazil, access to dialysis is universal, although the resources consumed and their costs are poorly understood. We compare the resources used for the treatment of patients undergoing HD or PD who are covered by public insurance. METHODS: A one-year prospective study comparing resource use and total costs of prevalent patients treated with HD (n = 210) and PD (n = 194) was conducted in 5 dialysis units in Sao Paulo, Brazil. Inclusion criteria: ≥18 years of age and clinically stable on chronic dialysis. The study period was April 2007 to February 2009. Data were obtained at baseline, 6 and 12 months using surveys and medical records. Cost categories included hospitalizations, diagnostic and therapeutic procedures, medications, professional fees, transportation, and lost productivity (current homemakers who stopped working due to dialysis). Government reimbursement rate was used as a proxy for the direct costs related to the act of dialysis (maintenance). The study took the societal perspective. RESULTS: Approximately 50% of HD and 48% of PD patients were female (p = 0.75); 54% and 58% were white (p = 0.48); mean age was 55.2 and 60.6 (P < 0.001); 62% and 71% had diabetes (p = 0.0528); and 59% and 55% had coronary heart disease (p = 0.37), respectively for HD and PD. Overall average costs per patient-year of follow up was US$23,283 for HD and US$23,285 for PD patients. The average annual cost per patient-year, per category, for HD and PD were respectively, US$11,774 and US$14,058 for maintenance dialysis costs; US$9,208 and US$7,559 for medications; US$94 and US$43 for hospitalization, US$796 and US$487 for travel costs and $US323 and $264 for current homemakers lost income due to dialysis. CONCLUSIONS: Mean annual total cost of PD and HD are nearly identical, even though the former were significantly older and more diabetic. PUK12 THE ECONOMIC IMPACT OF RENAL GRAFT FAILURE: A COST ANALYSIS IN A UK SETTING Sun Y1, Cerri K2 1 London School of Economics and Political Science, London, UK; 2Bristol-Myers Squibb, Braine-l’Alleud, Belgium OBJECTIVES: Compared to dialysis, kidney transplantation is a highly cost-effective choice for most patients with ESRD. Post transplantation, a key objective is to maintain a functioning graft. When graft failure occurs, the majority of patients return to dialysis. This study is performed to assess the cost of renal graft failure in a UK setting. METHODS: A model was built using data from the UK renal registry (2007–2008) to estimate the number of graft failures occurring in the first year after transplantation. Costs for procurement, transplantation, and for the treatment of graft failure, were derived from the result of a systematic review. This study adopted an investment perspective—all the medical resource used from organ procurement to the treatment of graft failure were taken into consideration. RESULTS: In the UK, the cost of renal graft failure was approximately £58,847 when taking account the medical resource used from an investment point of view (including transplantation cost, immunosuppressive medication cost and resource to treat post transplantation adverse events for graft failure patients). The post graft failure cost was £28,179. The most important cost contributors are dialysis cost, transplantation cost and post transplantation immunosuppressive medication cost. CONCLUSIONS: Estimating the economic impact of graft loss should take into account the cost of management of patients post graft failure, as well as previous medical investment that is lost with the graft (includ- A475 ing costs associated with procurement of the organ and transplantation). Improvements in the management of renal transplant patients are needed to reduce the risk of graft loss and the economic burden of graft failure to the health care system. PUK13 INCIDENCE AND COST OF HOSPITALIZATIONS FOR ACUTE URINARY RETENTION ATTRIBUTABLE TO BENIGN PROSTATE HYPERPLASIA IN FRANCE Cotté FE1, Torreton E2, Lafuma A3 1 GlaxoSmithKline, Marly le Roi, France; 2CEMKA-EVAL, Paris, France; 3CEMKA-EVAL, Bourg la Reine, France OBJECTIVES: Acute urinary retention (AUR) is one of the most significant complications of long-term benign prostatic hyperplasia (BPH) and often leads to prostate surgery. AUR also represents an important and feared event that needs to be quantified from an economic perspective as well as from the viewpoint of BPH patients. METHODS: French hospital information program (PMSI) databases from 2005 to 2008 were used to assess the number of hospitalizations for AUR and their management among males ≥50 years old presenting with a diagnosis of BPH. Number of patients concerned and rates of re-hospitalizations and deaths due to AUR were estimated using the linking system of the PMSI. Cost estimation was performed for the year 2008 adopting the Sickness Funds perspective. RESULTS: During the 4-year period, AUR frequency increased to 38,914 hospitalizations (+20.5%, +8.2%, +1.2% compared with 2005, 2006 and 2007, respectively). In 2008, mean length of stay was 5.5 ± 6.9 days and decreased of 6.6% within the study period. Proportion of stays referred by emergency department was 28.8% (14.1%, 20.6%, 25.3% in 2005, 2006 and 2007, respectively). 26,581 males were concerned by at least one AUR (+15.4%, +5.9%, +0.0% compared with 2005, 2006 and 2007, respectively) of whom 32.6% were concerned by a recurrence (28.9%, 30.7%, 31.4% in 2005, 2006 and 2007, respectively). Mean age was 74.1 years (SD:10.4) (stable over the period) and 232 patients (0.9%) deceased during their hospitalization in 2008. Mean cost per patient in 2008 was c2400 (c2663 vs. c1997 in public and private hospitals, respectively). Global cost of hospitalizations for RAU was estimated at c93.4 million (67.2% in public hospitals). CONCLUSIONS: AUR attributable to complicated BPH globally increased but tend to become stable in 2008. However, proportion of emergency utilization and recurrent stays amplified. Despite appropriate available therapies, prevention of BPH complications remains suboptimal. PUK14 HEALTH CARE COST OF RENAL REPLACEMENT THERAPY IN HUNGARY Kalo Z1, Kiss Z2, Gerendy P2, Nagyjanosi L3, Vokó Z1 1 Eötvös Loránd University, Budapest, Hungary; 2National Health Insurance Fund Administration, Budapest, Hungary; 3Syreon Research Institute, Budapest, Hungary OBJECTIVES: Annual cost of renal replacement therapy is an important benchmark for public reimbursement of all health care services. The last report was presented more than 10 years ago in Hungary, so our objective was to calculate the current annual cost of waiting listed dialysis and renal transplantation. METHODS: We selected all waiting listed or renal transplanted patients between July 2004 and March 2008. Resource utilization of all health care services with public reimbursement per patient in Q1 2008 were aggregated by linking claims records with anonimised patient IDs. We calculated health care costs of waiting listed and renal transplanted patients. Results were adjusted to gender, age and onset of ESRD by multivariate regression analysis. a total of 135 HUF/USD GDP specific PPP exchange rate was employed to convert results into USD. 2008 cost calculations were compared to results of the 1997 analysis. RESULTS: A total of 2209 patients were selected to the analysis. 3 year cost of waiting listed dialysis and renal transplantation was US$110,742 and US$87,420 respectively. Renal transplantation is cost-saving within 2 years compared to dialysis. CONCLUSIONS: Between 1997 and 2008 the 3-year cost of waiting listed dialysis increased by 60.3%, 3-year cost of renal transplantation increased by 96,8% without correction for inflation. In real values the health care costs waiting listed dialysis and renal transplantation is reduced by 26.7% and 10,0%. During this period the costcontainment measures of the National Health Insurance Fund were successful. PUK15 RESOURCE USE AND COSTS OF PATIENTS UNDERGOING DIALYSIS IN BELGIUM Caekelbergh K1, Lamotte M1, Dratwa M2, Bogaert AM3, Bouman K4, Laplante S5 1 IMS Health Consulting, Brussels, Belgium; 2CHU Brugmann, Brussels, Belgium; 3AZ St Elisabeth, Zottegem, Belgium; 4ZNA Middelheim, Antwerpen, Belgium; 5Baxter Healthcare Corporation, Braine l’Alleud, Belgium OBJECTIVES: This study was conducted to identify and compare resources and costs used by dialysis patients in Belgium. METHODS: The records of 130 patients undergoing dialysis were retrospectively reviewed to identify direct medical and non-medical resources used over year 2006. Data collected: baseline medical characteristics, dialysis-related information and resources used (hospitalizations, ambulatory care, medication, transport). Official tariffs were used for costing. RESULTS: Patients were hospitalized 1.5 ± 1.7 times for 18.1 ± 29.1 days. Laboratory tests were performed more frequently in hemodialysis (HD) patients than in peritoneal dialysis (PD) patients (295.6 ± 137.7 vs. 120.1 ± 75.5; P < 0.0001). Patients on HD took more medications (12.4 ± 3.7 vs. 10.7 ± 4.3; P = 0.0254). 79% of patients received EPO (average dose: 10,587 ± 14,114 IU). Patients on PD had more ambulatory consultations (9.7 ± 8.8 A476 vs. 4.6 ± 8.9; P = 0.0029) and interventions (5.8 ± 11.1 vs. 0.7 ± 1.1; P = 0.0042). Only 11.6% of PD patients used an ambulance or transport organized by the dialysis center/sickness fund, compared with 67.8% for HD. The estimated annual cost for the public payer (PP) was c72,350 per HD and c55,343 per PD patient (i.e., 31% more). As in 2006 there were approximately 6400 patients on dialysis (90% on HD, 10% on PD), the PP total cost is estimated to be around c452 million (2.45% of 2006 health care budget). The dialysis procedure was the main cost driver (66% of costs) being 27% more expensive for HD. Hospital and ambulatory services were respectively 28% and 45% more expensive for HD. CONCLUSIONS: The economic burden of dialysis is important in Belgium. Considering that survival of PD patients is at least as good as that of HD patients and that home-treatment reduces exposure to hospital pathogens, PD represents good value for money and should be considered in more patients. PUK16 COST OF ILLNESS ASSOCIATED WITH RENAL TRANSPLANTATION AND DIALYSIS IN END STAGE RENAL DISEASE IN THE UNITED STATES Srivastava K, Sharma S, Ahuja A, Taneja A Heron Health Private Ltd, Chandigarh, India OBJECTIVES: A patient suffering from end stage renal disease (ESRD) has two treatment options, lifelong dialysis or renal transplantation. The aim of this review is to determine economic cost of illness associated with renal transplantation and dialysis in ESRD in the United States (US). METHODS: The information was retrieved from databases including Medline, EMBASE, United States Renal Data System (USRDS), WHO and relevant grey literature. Studies reporting data for cost associated with transplantation and dialysis in ESRD in the US were included. RESULTS: In 2007, hemodialysis (HD) was initiated in 99,886 patients, peritoneal dialysis (PD) in 6376 patients and transplantation in 2500 patients in the US. Total Medicare costs associated with these were $17.6 billion for HD, $949 million for PD and $1.9 billion for transplantation (USRDS 2009). Unadjusted average annual Medicare expenditure (2004 US$) for PD and HD as first modalities was $53,277 and $72,189 respectively (Shih 2005). Patients with HD were twice as likely to be hospitalised over a 12-month period compared to matched PD patients. The median health care costs associated with hospitalization were $173,507 for HD patients vs. $129,997 for PD patients (Berger 2009). The mean length of stay was significantly less for PD with 6.57 days (P < 0.0001) vs. 7.25 days for HD (Walker 2009). The mean cost of treating S. aureus bacteraemia in HD patients, including readmissions and outpatient costs, was $24,034 per episode (Engemann 2005). Over a 25 year time horizon, renal transplantation resulted in significant cost savings with a cost of $376,577/patient and life expectancy of 7.4 years compared to $568,670/patient and life expectancy of 6.7 years with long term dialysis (Quinn 2007). CONCLUSIONS: Renal transplantation results in significant cost savings compared to long term dialysis. The total health care costs associated with hemodialysis are higher compared to peritoneal dialysis. PUK17 PATIENT CO-MORBIDITIES AFFECT THE COST OF DIALYSIS PATIENTS IN BELGIUM Laplante S1, Walker DR2, Caekelbergh K3, Lamotte M3, Dratwa M4, Bogaert AM5, Bouman K6 1 Baxter Healthcare Corporation, Braine l’Alleud, Belgium; 2Baxter Healthcare Corporation, McGaw Park, IL, USA; 3IMS Health Consulting, Brussels, Belgium; 4CHU Brugmann, Brussels, Belgium; 5AZ St Elisabeth, Zottegem, Belgium; 6ZNA Middelheim, Antwerpen, Belgium OBJECTIVES: This analysis was done to assess if co-morbidities influence the public payer (PP) cost of dialysis patients in Belgium. METHODS: The data from a cost study (retrospective chart review of 130 Belgian patients undergoing dialysis in 2006) was analyzed a posteriori. Baseline medical characteristics were used to compute the Charlson co-morbidity score (CCMS). Costs included: dialysis procedure and medical management (i.e., hospitalizations, outpatient visits and procedures, laboratory and imaging tests, and transport). Multivariate analyses were performed with the logarithmic transformation of costs as the dependent variable and CCMS, dialysis modality (hemodialysis: HD or peritoneal dialysis: PD) and gender as the independent variables. The regression model was weighted by number of patient months in the study. CCMS was categorized as low (<4), moderate (4–5), high (6–7) and very high (> = 8). RESULTS: All 3 variables had a significant impact on costs. Total costs to the PP were 16% higher for HD than for PD patients (p = 0.0039) and were 13% higher in women than in men (p = 0.0207). The costs in patients with a very high CCMS were 21% higher than those with a low or moderate score (p = 0.0072 and p = 0.0094 respectively) and 10.7% higher than those with a high score, but this latter difference did not reach statistical significance (p = 0.1160). The differences were larger when excluding the cost of dialysis procedure and considering medical management only, but only reached statistical significance or patients having a very high CCMS score vs. low or moderate CCMS (p = 0.0036 and 0.0056 respectively). CONCLUSIONS: This analysis showed that patient co-morbidities have a significant impact on medical management and total costs of dialysis patients. It is therefore important to take this into consideration when studying the costs of dialysis patients, especially if a total cost approach (i.e., procedure plus medical management) is taken. 13th Euro Abstracts PUK18 COST-EFFECTIVENESS ANALYSIS OF TIMELY VERSUS LATE DIALYSIS REFERRAL AFTER RENAL TRANSPLANT FAILURE IN SPAIN Villa G1, Fernández-Ortiz L1, Cuervo J1, Rebollo P1, Sánchez-Álvarez E2, Ortega F2 1 BAP Health Outcomes Research, Oviedo, Spain; 2Hospital Universitario Central de Asturias, Oviedo, Spain OBJECTIVES: Complications due to late dialysis referral after graft loss involve higher medical costs, together with a worsened health status and higher mortality rates. The efficiency of timely (TDR) versus late dialysis referral (LDR) after renal transplant failure is evaluated for the Spanish case. METHODS: A Markov model was developed and 6 health states were defined: hemodialysis (HD), peritoneal dialysis (PD), transplant (TX), late referral hemodialysis (LRHD), late referral peritoneal dialysis (LRPD) and death (D). a hypothetical cohort of patients aged 45 was observed during 40 years, considering age-dependent mortality rates. Transition probabilities were estimated using data from the Spanish Nephrology Society registry. Costs (in 2009 EUR) were obtained from a comprehensive literature review and included both direct (DC) (medical and nonmedical) and indirect costs (IC) (lost labor productivity due to mortality and morbidity). Effectiveness was measured in terms of Quality Adjusted Life-years (QALYs). Health utilities were estimated from a proprietary database. a discount rate of 3.5% was considered for both cost and effectiveness figures. All the model parameters were supported by an expert panel. Incremental Cost-Effectiveness Ratios (ICERs) and Net Health Benefits (NHBs) were computed. a willingness-to-pay threshold of c35,000/ QALY was taken into account. Both univariate and Monte Carlo multivariate sensitivity analyses were performed. RESULTS: The ICER was c27,385/QALY (IC not included) and c34,051/QALY [IC included], providing NHBs of (0.08) [0.01]. TDR yielded 0.37 additional QALY/patient. The multivariate sensitivity analysis showed that TDR was efficient in (54%) [53%] and dominant in (28%) [27%] of the simulations. The probability of accepting TDR was (55%) [50%]. CONCLUSIONS: TDR is an efficient scenario when compared to LDR, providing a greater number of QALYs with yet an affordable increase in costs. Our results, however, raise the debate on the suitability of the willingness-to-pay threshold as a rigid decision tool. PUK19 ECONOMIC EVALUATION OF ALISKIREN IN TYPE 2 DIABETES AND HYPERTENSION PATIENTS WITH NEPHROPATHY IN MEXICO Nevarez A1, García-Contreras F1, Olvera K2 1 Mexican Institute of Social Security, Mexico City, D.F., Mexico; 2Novartis, Mexico City, D.F., Mexico OBJECTIVES: To determine the most cost-effective alternative between a) Losartan, and b) Losartan + Aliskiren in type 2 diabetes and hypertension patients with microalbuminuria in the Mexican Institute of Social Security. METHODS: A complete economic evaluation was performed from institutional perspective, using a Markov model as analytical tool with semi-annual cycles and follow up until death, with transversal analyses at 10, 15 and 20 years. Simulating a cohort with a 53 years old patient with type 2 diabetes, hypertension, and microalbuminuria and using a discounting rate of 5% in costs and effectiveness. One assumption is that all patients that require dialysis receive it. Proportion of patients who have not received dialysis, as well as survival and quality of life were considered as effectiveness end points. Transition probabilities were obtained from AVOID study and IMSS information. Resource use was obtained from IMSS data and costs are considered in 2009 USD. Probabilistic and non-probabilistic sensitivity analysis was performed. RESULTS: Keeping a patient in stages prior to dialysis at 20 years of follow-up requires an investment of $19,647 with Losartan and $18,774 with Losartan + Aliskiren. After 14 years of follow up, Aliskiren + Losartan is dominant versus the use of Losartan. CONCLUSIONS: Aliskiren + Losartan is a cost-saving alternative if administered for prolonged periods, being the most effective regardless the period of monitoring and effectiveness measurement used. PUK20 THE COST-EFFECTIVENESS OF LANTHANUM CARBONATE VS. SEVELAMER HYDROCHLORIDE IN PATIENTS WITH END-STAGE RENAL DISEASE Park H1, Rascati KL1, Keith MS2, Hodgkins PS2, Smyth MD3, Goldsmith D4, Akehurst RL5 1 The University of Texas at Austin, Austin, TX, USA; 2Shire Pharmaceuticals, Wayne, PA, USA; 3Shire Pharmaceuticals, Basingstoke, UK; 4Guy’s Hospital London, London, UK; 5The University of Sheffield, Sheffield, UK OBJECTIVES: To assess the cost-effectiveness between two non-calcium binders, lanthanum carbonate (LC) and sevelamer hydrochloride (SH), in end-stage renal disease (ESRD) patients previously treated with calcium-based binders. METHODS: A Markov model was developed to estimate incremental costs for three health outcomes: 1) quality-adjusted life-years (QALYs), 2) Life-years saved (LYS) and 3) percent who successfully met serum phosphorus (SP) level goals (3.5–5.5 mg/dl) between the two non-Ca binders. The model incorporated patient-level data from a randomized head-to-head crossover study which compared the reduction of SP using fixed doses of LC for 4 weeks. For this analysis the model included patients previously treated with calcium-based binders. The ‘intent-to-treat’ (ITT) population and the ‘completer’ population were assessed. Baseline risks of cardiovascular disease (CVD), overall mortality, and CVD mortality were derived from a large US epidemiological study. Utilities, costs and relative risks of CVD were derived from published sources. Patient outcomes were modeled for 10 years, and incremental cost-effective ratios (ICERs) were calculated for LC relative to SH. Clinical and economic outcomes were dis- 13th Euro Abstracts counted at 5% per year. Deterministic and probabilistic sensitivity analyses (PSA) were performed to test the model robustness. RESULTS: Using clinical results of patients previously treated with a calcium-based binder, for the ITT population, the ICERs were $24,724/QALY, $15,053/LYS and $6,700/additional success (i.e., SP goal achievement), respectively. For patients who completed the randomized controlled trial, the ICERs of LC compared with SH were $15,285/QALY, $9,337/LYS and $8,265/additional success respectively. The PSAs indicated a 61.9% and 85.8% probability of LC being cost-effective at the $50,000/QALY threshold, for the ITT population and completer populations, respectively. Sensitivity analyses support the robustness of the model results. CONCLUSIONS: LC is a cost-effective strategy compared to SH in ESRD patients previously treated with calcium-based binders. PUK21 LONG-TERM COST-EFFECTIVENESS OF SIROLIMUS BASED REGIMEN COMPARED WITH CALCINEURIN INHIBITOR BASED REGIMENS IN LOWER IMMUNOLOGICAL RISK RENAL TRANSPLANT RECIPIENTS IN KOREA Kang MH1, Song HJ2, Park SY2, Kang SH2, Ko SK1, Lee EK2 1 Pfizer Pharmaceuticals Korea Ltd, Seoul, South Korea; 2SookMyung Women’s University, Seoul, South Korea OBJECTIVES: The calcineurin inhibitor (CNI) based regimens have improved shortterm patient and graft survival. However, the nephrotoxicity of CNI has limited the long-term improvement of outcomes. Recently, sirolimus, a novel class of immunosuppressant, got an approval in Korea. In this report, the cost-effectiveness of sirolimus based regimen compared with CNI based regimens was evaluated especially in the lower immunological risk renal transplant patients under Korea health care system. METHODS: A Markov model was developed to simulate costs and outcomes of a renal transplant recipient for 20 years with 5% discount rate in the societal perspective. Health states were defined as patient with graft survival, patient with graft failure, patient with regraft survival, patient with regraft failure and patient death. Quality adjusted life-years (QALYs) were taken as effectiveness parameter. Efficacy data were obtained from systematic review of immunosuppressive regimens in lower immunological risk renal transplant recipients. And the additional benefits of mTOR inhibitor (including sirolimus), reduced nephrotoxicity and lower relative risk to the incidence of malignancy, were reflected in the model. Utility weights were obtained from published literature. Costs were calculated by the Korean public institutional data and clinical expert opinions. Sensitivity analyses were performed on crucial parameters. RESULTS: Sirolimus based regimen costs KRW 130,067,696 for the 8.94QALYs for 20 years treatment duration, whereas KRW 140,185,240 for 8.67 QALYs for tacrolimus based regimen, and KRW 131,136,664 for 8.56 QALYs for cyclosporine based regimen. Sirolimus based regimen was shown to be dominant over the others, as it was more efficacious in QALYs and less costly. The sensitivity analysis showed that the results were quite robust across most parameters. CONCLUSIONS: Sirolimus based regimen is expected to be superior in the clinical effectiveness and cost-saving in lower risk renal transplant recipients compared with the most frequently prescribed CNI regimens in Korea. PUK22 HEALTH ECONOMIC ASPECTS OF USING SERUM CYSTATIN C FOR EARLY DETECTION OF CHRONIC KIDNEY DISEASE IN TYPE 2 DIABETICS IN GERMANY Walter E1, Lennartz L2, Hofmann W3, Herget-Rosenthal S4 1 Institute for Pharmaeconomic Research, Vienna, Austria; 2Abbott GmbH & Co. KG, Wiesbaden, Germany; 3Klinikum München, München, Germany; 4Rotes Kreuz Krankenhaus, Bremen, Germany OBJECTIVES: Diabetic nephropathy is a serious and common complication in diabetic patients; however the onset and the course can be favourably influenced by appropriate therapy when detected early. Current recommendations include testing of creatinine based estimation of glomerular filtration rate (eGFR) and for microalbuminuria. Due to substantial limitations of creatinine, cystatin C (cysC) has been proposed as alternative marker for early detection of (chronic kidney disease) CKD. METHODS: We developed a Cost-Utility Model to simulate the long-term consequences and disease progression for diabetic patients using serum cysC instead of Creatinine for monitoring of kidney function. Markov modelling techniques were used to estimate incidence of complications and disease progression, and Monte Carlo simulation accounted for uncertainty. The model includes 11 health states to describe the disease progression and occurrence of adverse events. Probabilities were derived from clinical and epidemiological studies. The cohort definition was adapted from the UKPDS study. Direct medical costs from published sources were used and expressed in 2009 Euro from the payer’s perspective. QALYs and total costs were projected over a life-time horizon and discounted at 5% p.a. RESULTS: Early detection of CKD with cysC leads to lifetime costs of c52,950 and 14.19 QALYs (c3,732 per QALY). Detection with Creatinine amounts to c64,912 and 12.82 QALYs (c5,063 per QALY). Cost saving amounts to c11,962 per patient and a QALY gain of 1.37; these arise due to slower disease progression and reduced complications. The cumulative incidence of complications was estimated to be lower for cysC than Creatinine (−15%). Early detection with cysC leads to a decreased number of patient-years (14%-points) spent in ESRD, due to a later onset. Probabilistic sensitivity analyses demonstrated the robustness of the model. CONCLUSIONS: The economic benefit of using cysC is substantial, due to reduction in complications and disease progression as well as lower long-term costs. A477 PUK23 ECONOMIC EVALUATION OF DIFFERENT TREATMENT MODALITIES IN ACUTE KIDNEY INJURY De Smedt D1, Elseviers M2, Lins R2, Annemans L3 1 University Ghent, Ghent, Belgium; 2University of Antwerp, Wilrijk, Belgium; 3Ghent University—Brussels University, Ghent, Belgium OBJECTIVES: Major controversy exists regarding the optimal treatment strategy in Acute Kidney Injury (AKI). The purpose of this study was to assess the incremental cost-effectiveness of continuous (CRRT) versus intermittent renal replacement therapy (IRRT) and conservative (CONS) AKI treatment in Belgium. METHODS: An area under the curve model was used, whereby patients are simulated using time-to-event data (survival analysis), combined with daily direct medical costs. Data were derived from the multi-centre SHARF4 study in which the three treatment modalities for AKI were compared. a time horizon of 2 years, starting from the admission to the hospital, was considered. Utility data were obtained from SF-36 assessment at 2 years and from literature data on in-hospital utility. RESULTS: Analyses indicated that mortality rates, cost and length of stay differed significantly between treatment mode during hospitalization. The mortality rate and the cost per day during the post-discharge follow-up period showed no significant difference between the treatment modes. Utility values, which improved gradually after admission to the hospital, revealed no significant differences between the three treatment strategies. Conservative treatment was associated with a 2-year cost of c34,090 and 0.49 QALYs. The CRRT was the most expensive therapy with a cost of c51,664 leading to 0.52 QALYs. The cost and QALYs associated with IRRT were c43,711 and 0.46. The ICER of CRRT versus IRRT was c131,604/QALY, while the ICER of CRRT versus CONS amounted to c651,318/QALY. Additionally the conservative treatment dominated IRRT. Sensitivity analysis—tornado diagram—confirmed the robustness of the results. CONCLUSIONS: This study has indicated that the most expensive treatment (CRRT) associated with an incremental cost of approximately c7952 generates only a minor increase in QALYs of 0.06 compared to IRRT. Additionally the results revealed that IRRT was dominated by CONS. From a health economic perspective conservative therapy seems to be the preferred treatment strategy. PUK24 A COST-UTILITY ANALYSIS OF CINACALCET IN SECONDARY HYPERPARATHYROIDISM (SHPT) IN FIVE EUROPEAN COUNTRIES Iannazzo S1, Pradelli L1, Chiroli S2 1 Adres srl, Torino, Italy; 2Amgen (Europe) GmbH, Zug, Switzerland OBJECTIVES: A probabilistic patient-level Markov model was developed to simulate lifetime clinical and economic outcomes of the cinacalcet treatment in secondary hyperparathyroidism (SHPT) patients. We present the adaptation of this model to five European countries: Italy, Spain, Portugal, Czech Republic and Switzerland. METHODS: In the simulation the treatment with cinacalcet influences the trend in time of individual parathyroid hormone (PTH), serum calcium (Ca) and phosphorous (P) levels, based on the OPTIMA study. Published correlations between these levels, mortality and morbidity (CV events, fractures, and parathyroidectomies) were incorporated as well as local epidemiologic and cost data for dialysis, drugs and events management. Simulation horizon was patient lifetime; standard treatment (vitamin D sterols and phosphate binders) and cinacalcet plus standard treatment were compared. The effectiveness was measured as life expectancy (LE) and quality-adjusted life expectancy (QALE). Health Utility Indexes derived from literature and took into account dialysis, CV events and fractures. RESULTS: The simulated mean LE extension was 1.20 life-years (LY) in Italy, 1.1 LY in Spain, 1.18 LY in Portugal, 1.10 LY in Czech Republic and 1.40 LY in Switzerland. QALE increase was 0.89, 0.82, 0.89, 0.80 and 1.01 quality-adjusted life-years (QALY) in the same countries, respectively. The lifetime average cost increase, calculated not taking into account the cost for dialysis, was c28,161 in Italy, c23,878 in Spain, c27,932 in Portugal, CZK 836,914 (corresponding to c32,751) in Czech Republic and CHF 48,908 (c34,630) in Switzerland. The incremental cost-effectiveness ratio (ICER) was c23,473/LY and c31,616/ QALY in Italy, c21,789/LY and c29,270/QALY in Spain, c23,680/LY and c31,249/ QALY in Portugal, CZK 759,600/LY (approx. c29,726/LY) and CZK 1,042,643/ QALY (approx. c40,802/QALY) in Czech Republic and CHF 34,858/LY (approx. c24,682/LY) and CHF 48,310/QALY (approx. c34,207/QALY) in Switzerland. CONCLUSIONS: Cinacalcet treatment could be considered a cost-effective treatment of SHPT in all the countries analyzed. Results seem more homogeneous in the three southern countries. PUK25 USING GENETIC POLYMORPHISM AS A STRATEGY TO ESTIMATE THE POTENTIAL COST-EFFECTIVENESS OF PHARMACOLOGICAL CCR5 BLOCKADE IN DIALYSIS PATIENTS Vegter S1, Muntinghe FL2, Verduijn M3, Boeschoten EW4, Dekker FW5, Navis G6, Postma MJ1 1 University of Groningen, Groningen, The Netherlands; 2University Medical Centre Groningen (UMCG), Groningen, The Netherlands; 3Leiden University Medical Center, Leiden, The Netherlands; 4Hans Mak Institute, Naarden, The Netherlands; 5Clinical Epidemiology, Leiden, The Netherlands, 6Division of Nephrology, Groningen, The Netherlands OBJECTIVES: Pharmacological interventions that are of benefit in non-dialysis populations have thus far been disappointing in dialysis patients. Since clinical trials are expensive and time-consuming, adjunct strategies are needed to support decision A478 making in prioritization of tracks for drug development. Genetic association studies may provide such a strategy when a genotype is associated with a well-defined molecular and functional phenotype. Previously an association with better survival was found in incident dialysis patients with systemic inflammation and a deletion variant of the CC-chemokine receptor 5 (CCR5Δ32). Thus, we hypothesized that pharmacological CCR5 blockade could protect against inflammation associated mortality and estimated if such a treatment would be cost-effective. METHODS: Ascreen-and-treat strategy was modelled in which patients were screened for the CCR5Δ32 polymorphism and patients with the wild-type genotype and high inflammation status were treated with CCR5 antagonists. a decision-analytic Markov model was used. Costs, utilities and clinical data on the association between CCR5 polymorphisms and mortality were all gathered from a single prospectively followed dialysis cohort (NECOSAD, n = 413). Univariate and probabilistic sensitivity analyses were performed. RESULTS: Pharmacological CCR5 blockade in patients with CCR5 wild-type and high inflammation status decreased mortality (RR = 0.61) and improved the probability of renal transplantation (RR = 2.41). The cost-effectiveness of the screenand-treat strategy was c18,557 per life-year gained and c21,896 per quality-adjusted life-year (QALY) gained. The main drivers of the cost-effectiveness were the costs of pharmacological CCR5 blockade and the reduction in relative risk of mortality. Threshold analyses were performed for these two parameters. CONCLUSIONS: Pharmacological blockade of the CCR5 receptor in inflamed dialysis patients can be incorporated in a potential cost-effective genetic screen-and-treat program. This study illustrates the potential of genetic association studies in drug-development programs, as a new source of Mendelian randomized evidence from an observational setting. PUK26 THE INFLUENCE OF FUTURE UNRELATED COSTS ON COSTEFFECTIVENESS ESTIMATES: TREATMENT OF HYPERPHOSPHATEMIA WITH LANTHANUM CARBONATE IN PRE-DIALYIS PATIENTS WITH CHRONIC KIDNEY DISEASE Vegter S1, Tolley K2, Keith M3, Postma MJ1 1 University of Groningen, Groningen, The Netherlands; 2Tolley Health Economics, Buxton, UK; 3Shire Pharmaceuticals, Wayne, PA, USA OBJECTIVES: A long-standing controversy in health-economics is whether future unrelated costs should be included in cost-effectiveness analyses. This discussion is relevant in Chronic Kidney Disease (CKD) for treatments that delay progression towards dialysis and prolong survival. In this study, we determined the influence of future unrelated costs on the cost-effectiveness of the non-calcium based phosphate binder lanthanum carbonate (LC) when used as second-line treatment for hyperphosphatemia in predialysis patients. METHODS: Time-dependent Markov models were constructed; cohorts of 1000 patients were followed lifelong. Patients not reaching target serum phosphate (SP) levels on first-line calcium based phosphate binders (CB) were treated with LC. This strategy was compared with continued CB treatment. Patient-level data were pooled from two clinical trials, one in predialysis and one in dialysis. Reductions in SP levels delayed progression towards dialysis and prolonged survival. RESULTS: For the predialysis cohort, 544 did not achieve target SP levels (<4.6 mg/dl) on CB treatment, and 230 were eligible for LC treatment. Of these, 43 (18.8%) now responded to an 8 week trial of LC. Compared with continued CB treatment, 150 life-years and 101 QALYs were gained with LC. Considerable future unrelated costs were accrued due to prolonged survival while on dialysis. Including these unrelated costs, the cost-effectiveness of second-line LC treatment was £17,150/ QALY. Excluding future unrelated costs, net health care cost-savings were estimated due to delayed progression towards dialysis. The net monetary benefit of LC treatment rose from £1302 to £4558 upon exclusion of future unrelated costs. CONCLUSIONS: Second-line treatment of hyperphosphatemia with LC in predialysis CKD patients is cost-effective irrespective of whether future unrelated costs are included. However, this analysis demonstrates the substantial impact of these costs on the cost-effectiveness ratio. We recommend cost-effectiveness guidelines used for reimbursement purposes should specify exclusion of future unrelated costs in the base case, with inclusion in sensitivity analysis. PUK27 TIME SAVINGS WITH Q4W (ONCE-MONTHLY) C.E.R.A.: A TIME AND MOTION STUDY CONDUCTED IN HEMODIALYSIS CENTRES IN FIVE EUROPEAN COUNTRIES Raluy M1, Irgl H2, De Cock E1 1 United BioSource Corporation, Barcelona, Spain; 2F. Hoffmann-La Roche Ltd., Basel, Switzerland BACKGROUND: Governments across Europe are increasingly focused on controlling rising health care costs. Within this context, a major challenge for hemodialysis centres is to maximise efficiencies while maintaining high standards of care. OBJECTIVES: The aim of this study was to document health care personnel time for anaemia management in maintenance therapy for both shorter-acting erythropoiesis-stimulating agents (ESA) and a continuous erythropoiesis receptor activator (C.E.R.A.) once monthly (Q4W), and model time savings following the introduction of Q4W C.E.R.A. in five European countries. METHODS: A multinational, multicentre time and motion study has been conducted in 16 centres across Germany, France, Italy, Poland and Spain. The time spent on frequent anaemia management-related tasks (preparation, distribution, injection, record-keeping) was recorded in each centre by trained observers. Time/pt/session was used to calculate time/pt/year, time/centre/year and for modelling potential time savings with a 100% uptake of C.E.R.A. RESULTS: Numbers of 13th Euro Abstracts ESA injections/pt/year, weighted by type of ESA, frequency and route of administration, ranged from 53–148. The mean uptake of C.E.R.A. Q4W across 16 centres was 29% (7–56%). The mean annual reduction in the number of ESA administrations following conversion to C.E.R.A. was 76 (41–136). Estimated observed time/pt/year ranged from 38–310 min for ESA and 6–68 min for C.E.R.A. Assuming a 100% uptake of Q4W C.E.R.A. maintenance therapy, annual time savings/centre for frequent anaemia management tasks would be 84% (79–91%) in Germany, 78% (74– 86%) in France, 88% (87–88%) in Italy, 85% (78–88%) in Poland and 77% (69–84%) in Spain. CONCLUSIONS: Substantial annual time savings on frequent anaemia management-related tasks were consistently found in hemodialysis centres across five European countries with a 100% uptake of Q4W C.E.R.A. maintenance therapy. This provides the opportunity to re-focus health care resources, at a critical time point in the dialysis procedure, on other important CKD care tasks in order to improve overall pt care. PUK28 THE ECONOMIC BURDEN OF POST-TRANSPLANT EVENTS IN RENAL TRANSPLANT PATIENTS (PORTRAIT STUDY) IN A SINGLE UK CENTRE McEwan P1, Baboolal K2, Cerri K3 1 CRC, Cardiff, UK; 2Cardiff and Vale NHS Trust, Cardiff, UK; 3Bristol-Myers Squibb, Braine-l’Alleud, Belgium OBJECTIVES: There is a paucity of information available regarding the prevalence of post-transplant events and resource utilization associated with such events in renal transplant patients in real-life treatment settings. The PORTRAIT study aims to describe the health care resources used and to estimate the cost of managing posttransplant patients using observational data from transplant databases and physician questionnaires from transplant centres across Europe. This abstract describes the preliminary results from a single UK centre. METHODS: A pilot retrospective observational study was undertaken in which resource usage over a three year period, calculated from Healthcare Resource Groups (HRGs), was employed to derive costs, with results stratified by glomerular filtration rate (GFR) status at one year post transplant. Comparison of costs was undertaken using the non-parametric MannWhitney test. RESULTS: There were 879 renal transplant patients who had a recorded GFR measurement at one year post transplant. Overall 130 (14.8%) had a GFR of <30 mL/min/1.73 m2; 535 (60.1%) had a GFR between >30 and ≤60 mL/min/1.73 m2 and 214 (24.3%) had a GFR > 60 mL/min/1.73 m2. Overall three-year HRG derived costs were significantly lower in the >60 mL/min/1.73 m2 group at £497 (SD = £111) compared to both the 30–60 and <30 mL/min/1.73 m2, at £1,323 (SD = £1245, p = 0.025) and £1448 (SD = 1726, p = 0.01) respectively. CONCLUSIONS: This study provides evidence that post-transplant resource usage in a real-life treatment setting (assessed using HRG tariffs) is approximately three times higher in those patients with lower post transplant GFR. Therefore management strategies that promote renal function post transplant are likely to provide important resource savings. An ongoing database and physician questionnaire study has been implemented to confirm these observations using a bottom-up costing approach. URINARY/KIDNEY DISORDERS – Patient-Reported Outcomes Studies PUK29 CHARACTERIZING THE RELATIONSHIP BETWEEN HEALTH UTILITY IN KIDNEY TRANSPLANT RECIPIENTS IN UK AND US WITH DIFFERENT LEVELS OF KIDNEY FUNCTION Neri L1, McEwan P2, Cerri K3, Baboolal K4 1 Saint Louis University, St. Louis, MO, USA; 2CRC, Cardiff, UK; 3Bristol-Myers Squibb, Braine-l’Alleud, Belgium; 4Cardiff and Vale University Health Board, Cardiff, UK OBJECTIVES: Little data is available describing the relationship between quality of life and levels of kidney function in renal transplant patients. We sought to assess the relationship between health utility and renal impairment in US and UK kidney transplant (KTX) recipients. METHODS: Data was obtained from KTX patients enrolled at the kidney transplant facilities of the Renal Unit at the Cardiff and Wales NHS Trust in Cardiff, UK (n = 209) and Saint Louis University Hospital, St. Louis, MO (n = 233). General linear models were used to estimate adjusted EQ-5D means across CKD stages (K/DOQI classification). Partial Spearman’s correlation was used to evaluate trend across CKD classes. We adjusted all models for age, gender, time since transplant and diagnosis of diabetes. We adjusted for center effect by including the variable indicating the center of enrollment and the appropriate interaction term in the model. RESULTS: Mean age of KTX recipients was 52.7 and 49.1 years and mean time since transplant was 5.6 and 3.3 years in the UK and US cohorts respectively. After adjustment, EQ-5D scores resulted as follows in UK and US samples respectively: CKD 1–2, 0.74 and 0.87; CKD 3, 0.69 and 0.88, CKD 4, 0.61 and 0.82; CKD 5, 0.39 and 0.79. The trend across CKD classes was statistically significant in both samples (UK: ρ = −0.24, P < 0.01; US: ρ = −0.19, p = 0.03). Center effect was statistically significant and mainly explained by inclusion of recipients with failed graft on dialysis in the UK sample. CONCLUSIONS: This study demonstrates a significant and independent relationship between levels of post transplant renal impairment and health utility. Management strategies that maximize post transplant renal function will have important implications for patients’ quality of life. 13th Euro Abstracts PUK30 MAPPING THE OVERACTIVE BLADDER QUESTIONNAIRE-SHORT FORM (OABQ-SF) TO EQ-5D PATIENT PREFERENCES IN PATIENTS WITH OVERACTIVE BLADDER IN SPAIN Villacampa-Aubá F1, Ruiz MA2, Arlandis-Guzmán S3, Errando-Smet C4, Lizarraga I5, Rejas J6 1 Department of Urology, Hospital 12 de Octubre, Madrid, Spain; 2Universidad Autonoma de Madrid, Madrid, Spain; 3Department of Urology, Hospital Universitario La Fe, Valencia, Spain; 4 Department of Urology, Fundación Puigvert, Barcelona, Spain; 5Medical Unit, Pfizer España, Alcobendas, Madrid, Spain; 6Pfizer España, Alcobendas/Madrid, Spain OBJECTIVES: To develop a mapping algorithm to transform the Overactive Bladder questionnaire-Short Form (OABq-SF) questionnaire responses into EQ-5D derived utilities in patients with OAB in Spain. METHODS: The culturally adapted Spanish versions of OABq-SF and EQ-5D questionnaires were self-administered on two occasions 3 months apart to a set of patients of both genders, above 18 years, diagnosed of OAB according with standard criteria and a score > 8 in the OAB-V8 scale and able to understand PRO instruments written in Spanish. Patients were recruited consecutively in clinics of Urology all over the country. Linear correlations between symptoms severity and health-related quality-of-life (HRQoL) domains of the OABq-SF and EQ-5D (VAS and TTO-tariffs) were computed. a combined score of both domains of OABq-SF was used to obtain a unique scoring per patient, although mapping was developed for both combined and separated punctuations of the OABq-SF using regression modelling. RESULTS: A total of 199 OAB patients (58.4 years, 74% women, 99% Caucasian, 37% workers and 36% primary schooling) from 18 urological clinics were evaluated. Utility values ranged from 1 to 0.01; mean value = 0.7623 (SD = 0.2292) and median = 0.79. Correlations with EQ-5D were −0.282 and −0.301 for the severity of symptoms domain at baseline an after 3 months of a medical intervention, and 0.329 and 0.369 for the HRQoL domain, respectively. Pearson coefficients for combined scoring were −0.331 and −0.351, respectively. All coefficients were significant at P < 0.001 level. Observed relation between utilities and combined score followed a linear although heterocedastic pattern. Utilities distribution was clearly positively skewed with benign utilities prevailing. CONCLUSIONS: Utilities values for Spanish patients with OAB were mostly mild to moderate in the 0–1 range. a utility map was obtained for different levels of scoring in the OABq-SF questionnaire, which is useful to derive Qualityadjusted Life-year gain in economic evaluations. PUK31 PSYCHOMETRIC VALIDATION OF THE SPANISH VERSION OF THE OVERACTIVE BLADDER QUESTIONNAIRE SHORT-FORM (OABQ-SF) TO ASSESS SEVERITY OF SYMPTOMS AND PATIENT HEALTH RELATED QUALITY OF LIFE IN SUBJECTS WITH OVERACTIVE BLADDER Arlandis-Guzmán S1, Ruiz MA2, Errando-Smet C3, Villacampa-Aubá F4, García-Vargas M5, Arumi D6 1 Department of Urology, Hospital Universitario La Fe, Valencia, Spain; 2Universidad Autonoma de Madrid, Madrid, Spain; 3Department of Urology, Fundación Puigvert, Barcelona, Spain; 4Department of Urology, Hospital 12 de Octubre, Madrid, Spain; 5Health Outcomes Research Department, Corporate Affairs and Market Access Unit, Pfizer Spain, Alcobendas, Madrid, Spain; 6Pfizer Inc., Madrid, Spain, Alcobendas, Madrid, Spain OBJECTIVES: To carry out the psychometric validation of the Spanish version of the self-administered OABq-SF questionnaire, a tool measuring severity of symptoms and health-related quality of life (HRQoL) of patients with symptomatic overactive bladder (OAB). METHODS: The culturally adapted Spanish version of OABq-SF was administered on two occasions separately 3 months to a set of patients of both genders, above 18 years old, diagnosed of OAB according with standard criteria and a score > 8 in the OAB-V8 scale and able to understand PRO instruments written in Spanish. Patients were recruited consecutively in clinics of Urology all over the country. Feasibility, internal consistency (Cronbach’s alpha), test-retest reliability, structure of instrument, criteria and construct validity and responsiveness were examined using classic test theory statistics. Minimally Important Difference (MID) was also calculated. RESULTS: Data from 199 OAB patients (58.4 years, 74% women, 99% Caucasian, 37% workers and 36% primary schooling) from 18 urological clinics were evaluated. Floor and ceiling effect ranged between 0.5% and 31.2%, and missing items was below 1%. Cronbach’s alphas attained 0.795 and 0.923 for severity and HRQoL domains. These two instrument domains showed to be one-dimensional with explained variances above 50% in both domains, which correlated moderately with EQ-5D [−0.335 and +0.405, respectively (P < 0.01 in both cases)] and with self-perceived general valuation symptoms of OAB. a significant change in OABq-SF domains scoring [−23.3 (−26.0; −20.6) and +17.1 (14.7; 19.5), P < 0.001 in both cases (effect sizes: 1.34 and 1.50)] were observed after a 3-month medical intervention. MID values were, respectively, 6.6 and 8.1 for severity and HRQoL domains. CONCLUSIONS: The Spanish version of OABq-SF demonstrated strong reliability, validity and responsiveness psychometric properties to be used in the measurement of severity and Healthrelated Quality of life symptoms of OAB. PUK32 CONTENT VALIDITY OF A PATIENT-REPORTED URINARY URGENCY RATING SCALE: RESULTS OF A COGNITIVE INTERVIEW STUDY Chen WH1, Notte S1, Marshall TS2, Lee M2, Hakimi Z3, Revicki DA1 1 United BioSource Corporation, Bethesda, MD, USA; 2Astellas Pharma Global Development—US, Deerfield, IL, USA; 3Astellas Pharma Global Development—EU, AC Leiderdorp, The Netherlands OBJECTIVES: The Patient Perception of Intensity of Urgency Scale (PPIUS) measures the urgency severity of urinary episodes, and was developed based on the definition of urgency A479 presented by the International Continence Society (ICS) (Abrams et al. 2002) as well as recommendations from the Committee for Proprietary Medical Products (CPMP) (EMA 2002). The PPIUS is a single item scale with five levels of urgency severity, from “No urgency” to “Urge incontinence,” each with its definition provided. While the PPIUS has been used in previous trials, its content validity has not been established in overactive bladder (OAB) patients. The objective of this study was to determine comprehensibility and understanding of the scale through cognitive interviews with OAB patients. METHODS: Forty-one OAB patients participated in a non-interventional study assessing reproducibility of the PPIUS by completing a three-day micturition diary each week for three consecutive weeks. Following successful completion of the testretest study, twelve participants were selected to participate in a 30-minute in-person interview discussing their experience with the PPIUS. Interviews were conducted by a trained interviewer and followed a semi-structured guide with think-aloud approach. RESULTS: N = 12; mean (SD) age = 61.6 (13.5) years; 66.7% Caucasian; 91.7% female; mean (SD) years with symptoms = 5.8 (4.6); 83.3% reported medication use for treatment. Nine participants found it simple to choose a PPIUS rating for each of their daily micturition episodes. Most agreed with the definitions provided for the ratings of “No urgency” (n = 9), “Mild urgency” (n = 7), “Moderate urgency” (n = 8), “Severe urgency” (n = 9) and “Urge incontinence” (n = 9). Three suggested distinction between daytime and nighttime episodes could be made clearer. CONCLUSIONS: Content validity of the PPIUS was supported by the results of the cognitive interviews. PPIUS was well understood and the urgency definitions were generally accepted. Participants found the diary easy to complete and did not have difficulty selecting their ratings. PUK33 TEST-RETEST RELIABILITY AND ACCLIMATION EFFECT OF PATIENT PERCEPTION OF INTENSITY OF URGENCY SCALE (PPIUS) FOR OVERACTIVE BLADDER Chen WH1, Notte S1, Marshall TS2, Lee M2, Hakimi Z3, Revicki DA1 1 United BioSource Corporation, Bethesda, MD, USA; 2Astellas Pharma Global Development—US, Deerfield, IL, USA; 3Astellas Pharma Global Development—EU, AC Leiderdorp, The Netherlands OBJECTIVES: A three-day micturition diary has been designed to capture information on the number of urinary and incontinence episodes in overactive bladder (OAB) patients. Within the diary is the Patient Perception of Intensity of Urgency Scale (PPIUS), which captures the urgency associated with each episode. The PPIUS is a single item with five levels of urgency severity: “No urgency,” “Mild urgency” “Moderate urgency,” “Severe urgency,” and “Urge incontinence.” The objective of this study was to assess the test-retest reliability and acclimation effect of the PPIUS. METHODS: OAB patients were recruited from five sites in the US for a non-interventional two-visit study. At the Screening Visit, participants were trained on how to complete the three-day diary at home for three consecutive weeks. In the diary, participants recorded the time of each micturition, a corresponding PPIUS rating, and whether the episode was accompanied by incontinence. Participants and clinicians completed overall treatment effect (OTE) questions at the Final Visit asking whether their bladder condition had changed since the Screening Visit. RESULTS: N = 39; mean age = 59.1 years; 69.2% Caucasian; 82.1% female; mean (SD) years with symptoms = 6.7 (6.78); 94.9% reported medication use for treatment. The mean number of micturitions per 24-hours = 10.5, mean number of incontinence episodes per 24-hours = 1.8; mean level of urgency severity = 1.9. There were 34 (87.2%) and 35 (89.7%) stable patients as defined by OTE-Patient and OTE-Clinician, respectively. Using Weeks 2 and 3 data from stable patients, the intra-class correlation (ICC) was 0.95 and the Spearman correlation was 0.89. The difference between the average ratings of any two weeks was non-significant using paired t-test. CONCLUSIONS: The PPIUS demonstrated excellent test-retest reliability as evidenced by ICC = 0.95. Acclimation effect was not observed in the PPIUS, as evidenced by non-significant differences in the average ratings across three weeks. PUK34 PSYCHOMETRIC VALIDATION OF THE SPANISH VERSION OF THE TREATMENT BENEFIT SCALE (TBS) FOR ASSESSING SUBJECTIVE OUTCOMES IN TREATING PATIENTS WITH OVERACTIVE BLADDER Errando-Smet C1, Ruiz MA2, Villacampa-Aubá F3, Arlandis-Guzmán S4, García-Vargas M5, Arumi D6 1 Department of Urology, Fundación Puigvert, Barcelona, Spain; 2Universidad Autonoma de Madrid, Madrid, Spain; 3Department of Urology, Hospital 12 de Octubre, Madrid, Spain; 4 Department of Urology, Hospital Universitario La Fe, Valencia, Spain; 5Health Outcomes Research Department, Corporate Affairs and Market Access Unit, Pfizer Spain, Alcobendas, Madrid, Spain; 6Pfizer Inc., Madrid, Spain, Alcobendas, Madrid, Spain OBJECTIVES: To carry out the psychometric validation of the Spanish version of the self-administered TBS scale, a tool assessing subjective outcomes in the treatment of subjects with overactive bladder (OAB). METHODS: The culturally adapted Spanish version of TBS was administered at the end-of-trial visit after 3 months of a medical intervention for OAB. Responses to the instruments were recorded in a set of patients of both genders, above 18 years old, diagnosed of OAB according with standard criteria and a score > 8 in the OAB-V8 scale, and able to understand PRO instruments written in Spanish. Patients were recruited consecutively in clinics of Urology all over the country. Scale variability, criteria, known-groups and construct validity and responsiveness (effect size) were examined using classic test theory statistics. RESULTS: Data from 199 OAB patients (58.4 years, 74% women, 99% Caucasian, 37% workers and 36% primary schooling) from 18 urological clinics were evaluated. The 2.5% of A480 patients did not fill in the scale, only 31% greatly improved, 57% improved, 10% did not changed and 2% worsened during medical intervention. TBS correlated moderately with clinical judgement of change; 0.749, P < 0.001, and also, linearity was observed, with changes at end-of-trial visit in both domains of the OABq-SF; Flineal = 23.02, P < 0.001 and Flineal = 31.02, P < 0.001. TBS was able to differentiate groups of patients with different levels of change in symptoms severity and health-related quality of life of the OABq-SF questionnaire; F = 8.09, P < 0.001 and F = 10.5, P < 0.001, respectively, with effect sizes between 1.69 and −0.44. CONCLUSIONS: The Spanish version of TBS scale demonstrated strong validity and responsiveness psychometric properties to be used in the subjective assessment of outcomes related with treatment of OAB. URINARY/KIDNEY DISORDERS – Health Care Use & Policy Studies PUK35 PATIENTS WITH BPH: RESULTS AT 6 MONTHS FOLLOWING TREATMENT WITH PHYTOTHERAPY VERSUS OTHER TREATMENTS Taieb C1, Auges M1, Perrin P2 1 PFSA, Boulogne, France; 2Lyon Sud, Pierre Benite, France OBJECTIVES: Assess the impact of the treatment of urinary disorders of the lower urinary tract (LUT) suggestive of benign prostatic hypertrophy (BPH) using medical 13th Euro Abstracts treatment under actual conditions of use. METHODS: A pragmatic cohort of 420 patients (France, Italy, and Portugal) treated medically, was followed up for 6 months, using 3 questionnaires: IPSS, MSF4 and SF12. RESULTS: 366 patients under medical treatment were assessed. 267 patients were treated with phytotherapy, versus 80 patients on “other treatments”. At inclusion, the patients treated with Serenoa repens versus “other treatments” were different on the following characteristics: Age (P = 0.0047), time since diagnosis (p = 0.006), country (P < 0.0001), IPSS score (P = 0.0087), physical dimension of the SF12 (P = 0.0071). The “change from baseline” for the scores of the IPSS and SF12 self-assessment questionnaires between the 2 treatment groups was comparable. a generalized linear model adjusted for age, time since diagnosis, country, IPSS score, and the physical dimension of the SF12 score at inclusion was used, making the 2 treatment groups comparable. We observed an improvement in the IPSS score from 6 weeks. We were not able to demonstrate a significant difference between the 2 treatment groups concerning the “change from baseline” of the IPSS score (P = 0.7288). The same applies to the analyses at 3 and 6 months where the p-values were 0.2047 and 0.0947 respectively. We did not observe any statistical difference in the scores for the 2 dimensions of the SF12 between the 2 treatment groups. CONCLUSIONS: We observed an improvement in the IPSS and SF12 scores from 6 weeks. This improvement was not significantly different between the 2 treatment groups. Under actual conditions of use, the various medical treatments gave similar improvements.