American Journal of Therapeutics 10, 93–103 (2003) Provision of Drug Information to Patients by Pharmacists: The Impact of the Omnibus Budget Reconciliation Act of 1990 a Decade Later Robin Schatz, Robert J. Belloto, Jr., Donald B. White, and Kenneth Bachmann* Drug-related illness in the United States factors substantially in health care costs, although often these illnesses and their attendant costs are preventable. One strategy for minimizing adverse drug reactions is to provide drug information to consumers in the form of prescription counseling at pharmacies. The Omnibus Budget Reconciliation Act of 1990 (OBRA 1990) contained provisions for mandating such counseling to Medicaid patients. OBRA 1990 was implemented in 1993, but most states acted quickly to extend counseling services to all patients receiving prescription drugs. We looked at the extent and quality of prescription counseling available in community pharmacies 1 decade after OBRA 1990 was written. We evaluated the counseling services afforded at large chain pharmacies, independent community pharmacies, and on-line pharmacies for a hydrochlorothiazide prescription. We found that most (69%) pharmacies offered to provide prescription counseling service, and that average counseling index scores, a measure of the quality or extent of information provided as determined by a Rasch analysis, were generally satisfactory. Our observations based on a single prescription for hydrochlorothiazide, along with other studies, suggest that there is a positive upward trend in the number of pharmacies providing prescription drug information, and that the extent of information provided suggests that the objectives of OBRA 1990 and related legislation to reduce ADRs are being fundamentally satisfied. Keywords: OBRA 1990, drug information, patient counseling, pharmacists. INTRODUCTION Drug-related illness in the United States is considered to be a substantial factor in health care costs. In 1995, drug-related illness was responsible for an estimated 17 million emergency room visits, 8.7 million hospital admissions, and $76.6 billion in hospital costs.1 The US Food and Drug Administration (FDA) reported 12,000 deaths caused by adverse drug reactions (ADRs) in 1987.2 Department of Pharmacology, College of Pharmacy, The University of Toledo, Toledo, Ohio, USA. *Address for correspondence: Department of Pharmacology, College of Pharmacy, The University of Toledo, 2801 W. Bancroft St., Toledo, OH 43606, USA. E-mail: kbachma@utnet.utoledo.edu 1075–2765 © 2003 Lippincott Williams & Wilkins, Inc. Over one-fourth of the hospitalizations of the elderly have been ascribed to either ADRs or medication noncompliance.3 Many of these ADRs and their sequelae of emergency room visits, hospitalizations, and attendant costs are preventable, considering that in approximately one-third of emergency room visits precipitated by ADRs, patients had little or no understanding of how to properly self-administer their medications.4 Lakshmanan et al reported that 5% of admissions to an acute care facility were iatrogenic, and 48% of those were preventable.5 The Adverse Drug Event (ADE) Prevention Study Group demonstrated that nearly 30% of ADEs in a hospital setting are preventable, especially if intercepted early in the process.6 With regard to ambulatory elderly patients, it has been estimated that potentially inappropriate medications are prescribed for nearly one-fourth of this cohort,7 and that at least one potentially inappropriate medication is prescribed in 4.5% of outpatient 94 visits by the elderly.8 In an analysis of 414 communitydwelling subjects aged 75 years and older in Santa Monica, California who were not terminally ill, and had no severe cognitive or functional impairments, 14% were found to be using at least one inappropriate drug.9 In an effort to reverse the escalating costs of drugrelated illness, Congress included language in the Omnibus Budget Reconciliation Act of 1990 (OBRA 1990)10 that required pharmacists to offer to provide drug information about their prescriptions (ie, to counsel) to Medicaid patients. These requirements in OBRA 1990 became effective on January 1, 1993. most states quickly followed suit, but also required pharmacists to offer to counsel all patients, not just Medicaid patients. Ultimately, the obligation of pharmacists to counsel patients was included in the American Pharmaceutical Association (APhA) Standard of Practice for the Profession of Pharmacy.11 As recently as 1997, an effort was made to assess the impact of OBRA 1990 on the counseling practices of community pharmacists in Nebraska.12 Almost half (45%) of the pharmacy respondents reported that time devoted to patient counseling had increased; however, statutory obligations and practice standards to offer counseling services have been in place for over 7 years now. In view of the potential for effective pharmaceutical counseling to preclude or mitigate ADRs, this study was undertaken with two objectives: • To evaluate the extent to which pharmacists are meeting their mandate to offer prescription counseling to patients; • To determine whether practice setting (ie, independent community pharmacy, chain pharmacy, or on-line pharmacy) is a factor in the quality or quantity of prescription counseling services provided by pharmacists. METHODS Investigational Review Board We submitted an application for review to the Human Subjects Research Review Committee (HSRC) of the University of Toledo. The HSRC reviewed the application and determined that this project was exempted from HSRC review in accordance with Code 45CFR 46.101 of the Federal Regulations. Pharmacies We selected a total of 90 pharmacies for study as follows: 30 were independent community pharmacies, 30 American Journal of Therapeutics (2003) 10(2) SCHATZ ET AL were large chain pharmacies, and 30 were on-line pharmacies. We selected on-line pharmacies from a search engine’s listing. All other pharmacies were selected from the yellow pages of a Toledo, Ohio-area phone book, and were located in northwest Ohio and southeast Michigan. Each pharmacy was randomly assigned a coded number, and all data was collected with reference only to those coded numbers. Data Form Using OBRA guidelines, the patient counseling information section of Clinical Pharmacology Online,13 and the patient information section from Drug Facts and Comparisons,14 we created a checklist-formatted data form (Fig. 1). We used this form to capture responses in a dichotomous fashion. That is, for each item on the checklist, we scored pharmacies as providing the information or service, or not providing it. We weighted each item equally. Prescriptions Five different area physicians wrote a total of 90 prescriptions for hydrochlorothiazide. Each prescription was for 30 tablets, 25 mg, to be taken once daily with no refills. Nine separate, written prescriptions were given to each of 10 monitors. We selected hydrochlorothiazide as the sole medication because (1) it is used to treat a prevalent condition (eg, hypertension), (2) it can be used as monotherapy in first-step therapy for hypertension, (3) it is devoid of abuse potential, (4) it is used to treat a symptomless condition, and (5) it is inexpensive. Filled prescriptions were returned to the University of Toledo College of Pharmacy for identification, counting, temporary storage, and subsequent destruction. Monitors We used a total of 10 monitors to present prescriptions at the 90 pharmacies, and to complete the data forms immediately thereafter. Monitors were drawn from a homogeneous demographic pool of college students. This step was necessary to assure that nothing about the presentation of any monitor at any pharmacy would signal special needs vis-á-vis drug information. Thus, all monitors were between the ages of 18 to 45 years, and appeared healthy. Moreover, each monitor was instructed to present himself or herself as taking no other medications, having no known allergies, and having no other medical conditions apart from hypertension. Each monitor was assigned to nine different pharmacies, three in each category (ie, chain, community independent, and on-line). Each monitor was OBRA 1990 95 Fig. 1. Data form for determining counseling index. provided with a total of nine different prescriptions, and charged with taking each of those to a different assigned pharmacy. Assignments were not random because it was intended that monitors were not to present a prescription at any pharmacy where they had been a customer or were otherwise known. Experimental Design In order to assure inter-rater reliability among the 10 monitors, all monitors received in-service training about the data form. After this instruction, all monitors were presented with three different vignettes, and asked to complete the data form for each vignette. One scenario depicted a pharmacist providing thorough prescription counseling information, one depicted a pharmacist providing virtually no counseling information, and one depicted a pharmacist providing an intermediate amount of counseling information. Individual monitors’ assessments of the quantity and quality of prescription information in each of these vignettes was linked to the assessments of the quality and quantity of prescription information in the field, thus providing a statistical assessment of the pharmacies corrected for any possible differences in monitors that may have existed.15 We used a many-facet Rasch model to objectively assess the ability of each pharmacist to provide American Journal of Therapeutics (2003) 10(2) 96 SCHATZ ET AL prescription information based upon the checklist. This model allows simultaneous quantification of the ability of each pharmacist (representing each pharmacy), the ease or difficulty of each item within the checklist, and the rating tendency of each monitor. The Rasch model used is shown below (Equation 1) log [PijkⲐ(1 − Pijk)] = Bi − Dj − Mk (1) where Pijk is the probability of the ith pharmacist satisfying the jth checklist item as recorded by the kth monitor, Bi is the ability of the ith pharmacist, Dj is the ease or difficulty of the jth item on the checklist, and Mk is the ability of the kth monitor. To estimate these parameters, it is necessary for all elements of all parameters to be linked.15 To accomplish this task, and to simultaneously train the monitors on the data form, all 10 monitors were presented with three different vignettes (ideal, adequate, and unsatisfactory prescrip- tion counseling), and asked to complete the data form for each vignette. Thus, each monitor became linked to every pharmacy, even though each monitor completed checklists on nine different pharmacies at most. A partial matrix created for this linkage is shown in Figure 2. Data Analysis Data was initially fitted to a three-facet Rasch model.15 In this analysis, each item checked on the form represents a pass; each item not checked represents a fail. The three-facet rating scale model taking the form previously shown (Equation 1) elicits continuous scales that simultaneously measure item difficulty, overall pharmacy counseling, and monitor ability. Within the framework of this model, all items are ranked for difficulty based upon their aggregate pass scores for all pharmacies, and pharmacies are ranked based upon Fig. 2. Matrix design for Rasch analysis. An X indicates that the monitors recorded something for all items (1–21) for the test pharmacies (ie, three vignettes) and the nine assigned (field) pharmacies. The linkage of the monitors, items, and pharmacists can easily be discerned in the data matrix. American Journal of Therapeutics (2003) 10(2) OBRA 1990 the number of items performed and the post hoc determination of item difficulty built into the Rasch analysis. Although a Rasch analysis basically omits items that all pharmacies passed or all pharmacies failed, it does assign them extreme scores (ie, > 3.0 or < −3.0). The resultant score for each pharmacy was denoted as a counseling index score. Model fitting and related computation of counseling index scores was accomplished with Facets software version 3.22 (Winsteps, Chicago, IL, USA).16 The analysis provides a measure of instrument (data form) reliability and validity.16 The range of counseling index scores could be expected to fall between +3.0 and −3.0, depending upon sample size and event rarity for each item. Once counseling index scores had been determined, the pharmacy code assignments were broken to enable the data to be stratified by practice setting (ie, chain, independent, or on-line). Differences in the counseling indices between these groupings were analyzed by an analysis of variance using JMP software (Cary, NC, USA).17 A Tukey–Kramer HSD post hoc test was applied if significant differences were found. Monitor reliability was calculated as the correlation coefficient for monitors’ responses (ie, assignment of a 1 if a checklist item had been performed or assignment of a 0 if not performed by a test pharmacy) versus true score of each test pharmacy on each item for each vignette. Validity or accuracy of monitors’ ratings was defined by the inter-monitor agreement for each checklist item performed or not performed by each of the test pharmacies (ie, in each vignette). Specifically, the 95% prediction interval for the slope of this regression line should embrace unity.18 RESULTS Seventy pharmacies (77.8%) actually filled prescriptions, and thus could be considered as respondents. One independent community pharmacy, even though listed under “pharmacies” in the yellow pages of the local telephone directory, could not fill the prescription because it sold over-the-counter (OTC) items exclusively. The 19 other nonrespondents were on-line pharmacies. Thus, prescriptions were filled at all 30 chain pharmacies, 29 community independent pharmacies, and 11 on-line pharmacies. Each of these filled prescriptions opened the door for a pharmacist– “patient” counseling experience that could be scored via the data form. Counseling was provided by on-line pharmacies either on-line, by e-mail, or by telephone. On-line pharmacies were initially provided a 1-month time frame in which to acknowledge receipt of the 97 prescription. If after 1 month there was no response, monitors were instructed to resubmit the prescription. If after a second month there was no acknowledgment of the prescription from an on-line pharmacy, it was contacted directly by the prescriber’s office. A third month was allowed to elapse before failure to acknowledge the prescription was considered to be a nonresponse. Counseling index scores computed by Rasch analysis are listed in Table 2 by rank. The more positive scores indicate a greater number of items were addressed. The more negative scores indicate fewer items being addressed. Even without a specific request from monitors, 19 pharmacies (27%) spontaneously initiated counseling. Another 29 of the respondent pharmacies (41%) actually offered to counsel our monitors without being requested to do so. The minimum requirement of OBRA 1990 is, in fact, that prescription counseling be offered on all prescriptions. This requirement was considered to have been met if monitors were asked if they wished to speak with a pharmacist, or if they had any questions, or if, at on-line pharmacies, e-mail or written instructions were provided about contacting a pharmacist. Thus, prescription counseling was provided without request by 48 (69%) of 70 pharmacies. Written pamphlets about hydrochlorothiazide were provided by 55 pharmacies (79%). Pharmacists were not available for counseling, even upon request of monitors, at 9 pharmacies (13%). All prescriptions were filled correctly with 25 mg tablets of hydrochlorothiazide. No single item on the data form was performed by all pharmacies or provided by all pharmacies. Only two items of information on the data form were not provided by a single pharmacy. No pharmacy provided drug storage information (item 19), and no pharmacy provided advice on steps to take in the event of a missed dose (item 21). The degree of difficulty of the 21 items on the data form ranged from 7.98 (i.e. item 21) to −4.08 (item 3), with larger positive values representing a higher degree of difficulty, and larger negative values representing items having been asked or provided by a large number of pharmacies. The Rasch estimates of item difficulty ± standard error (SE), corrected for monitor differences are listed in Table 1. Counseling indices ranged from +1.84 to −3.64 with a reliability of 0.78 (maximum possible 1.0). The mean indices ± standard error (SE) for each practice setting were: chain pharmacies, −1.34 (0.18); community independent pharmacies, −1.05 (0.29); and on-line pharmacies, −1.51 (0.36). These means were not significantly different by analysis of variance (ANOVA). The American Journal of Therapeutics (2003) 10(2) 98 SCHATZ ET AL Table 1. Rasch item analysis. Objective Rasch estimate of difficulty SE Spontaneously provided informaton Offered prescription counseling Pharmacist was available for counseling Written information provided Checked about other meds Checked on medical conditions Checked on allergies to meds Identified prescription as hydrochlorothiazide Identified strength as 25 mg Instructed to take once a day Instructed to take in AM Checked work shift Identified the indications as blood pressure or edema Warned of increased urination Warned of photosensitivity Warned of muscle cramps Advised potassium supplementation Advised to monitor blood pressure Advised on storage conditions Stipulated no refills Instructed what to do in case of missed dose 0.37 −0.94 −4.08 −1.88 0.29 0.81 −2.39 −0.88 1.97 −1.76 −1.76 2.58 −0.94 −1.23 .00 1.80 0.54 2.35 7.98 1.80 3.36 0.28 0.24 0.38 0.25 0.28 0.31 0.26 0.24 0.42 0.24 0.24 0.48 0.24 0.24 0.26 0.40 0.29 0.46 1.85 0.40 0.54 Data form item #* 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 *Item numbers correspond with those shown in Figure 1. † The more positive the score, the more difficult the item (ie, the less likely that this item would be performed by pharmacies); the more negative the score, less difficult the item (refer to Equation 1). SE indicates standard error. individual variates and their medians are depicted graphically in Figure 2. The ␹2 test for monitor equivalency embedded in the many-facet Rasch analysis yielded a P value less than 0.01, suggesting that all monitors were not equivalent.18 These differences are taken into account, however, by the many-facet Rasch model. Furthermore, the intermonitor estimate of reliability was 0.88. The slope value for the estimate of validity (accuracy) was 0.94 with 95% prediction limits of 0.60 to 1.29. This suggests a slight bias on the part of monitors to fail to record an item that was performed, but an adjustment for this bias is automatically included in the scoring of pharmacies. A ␹2 test for the variability of the pharmacies yielded a P value of 0.52, suggesting that each pharmacy sampled was drawn from a normally distributed population of pharmacies. Figure 3 shows the effect of the practice setting on counseling index scores. DISCUSSION The OBRA 1990 mandate to offer prescription drug counseling by pharmacists to their Medicaid clients is American Journal of Therapeutics (2003) 10(2) Fig. 3. Effect of practice setting on counseling index scores. Within each Box Plot for chain, independent, and on-line pharmacies, the individual plotting symbols denote individual pharmacy scores. Lines within box plots denote medians, and upper and lower box boundaries depict 75th and 25th percentiles, respectively. For test or vignette pharmacies, the plotting symbols represent individual pharmacy scores, and the upper and lower box boundaries depict the maximum and minimum values, respectively. The solid line across all four different groups depicts the grand mean score. 99 OBRA 1990 Table 2. Counseling index scores. Rank 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Setting Index SE Rank Setting Index SE Tank Setting Index SE Independent Independent Independent Chain Independent Independent Independent Independent Independent Chain On-line Independent Independent Chain Independent On-line On-line Chain Independent On-line Independent Chain Chain Chain 1.84 1.66 1.25 0.89 0.54 0.54 0.54 1.19 0.19 0.19 0.02 −0.07 −0.31 −0.31 −0.32 −0.32 −0.32 −0.43 −0.47 −0.48 −0.48 −0.48 −0.48 −0.48 0.64 0.64 0.60 0.59 0.59 0.59 0.59 0.59 0.59 0.59 0.59 0.59 0.58 0.58 0.58 0.58 0.58 0.61 0.58 0.58 0.58 0.58 0.58 0.58 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 Chain Chain On-line Independent Chain Chain Chain Independent Chain Independent On-line Independent Chain Chain Chain Chain Chain Chain Independent Chain Independent Chain Chain Chain −0.65 −0.66 −0.82 −0.82 −0.82 −0.88 −1.17 −1.17 −1.18 −1.24 −1.34 −1.34 −1.34 −1.34 −1.34 −1.34 −1.34 −1.34 −1.53 −1.53 −1.56 −1.70 −1.74 −1.91 0.58 0.58 0.58 0.58 0.58 0.59 0.59 0.59 0.61 0.67 0.59 0.59 0.59 0.59 0.59 0.59 0.59 0.59 0.61 0.61 0.63 0.61 0.63 0.63 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 Chain Chain Independent On-line Independent Independent Independent On-line Chain Chain Independent On-line On-line Independent Chain On-line Independent Independent Chain Chain Independent Independent −1.91 −1.98 −2.08 −2.08 −2.08 −2.08 −2.08 −2.33 −2.48 −2.51 −2.51 −2.83 −3.00 −3.12 −3.12 −3.12 −3.30 −3.35 −3.35 −3.35 −3.64 −3.64 0.63 0.67 0.63 0.63 0.63 0.63 0.63 0.67 0.74 0.67 0.67 0.74 0.74 0.87 0.87 0.87 0.87 1.15 1.15 1.15 0.87 0.87 SE indicates standard error. a recognition by government that the provision of sufficient prescription information is likely to lower ADRs and their related health care costs. Some states, beginning even before the implementation of OBRA 1990, have extended this mandate to all patients, and this mandate has also been incorporated into Pharmacy’s professional standards. The mandate of OBRA 1990 was implemented in 1993. The study that we conducted was intended to provide a snapshot of the extent and quality of counseling on prescription drugs currently available to consumers. Although the mandate to provide drug information can be technically met by providing written information, our evaluation pertained to face-to-face interactions. A report to the FDA investigators concluded that of the eight informational elements required for written material to provide complete drug information for patients, only three were included 60% of the time or more, and that there was considerable variability in the quantity and quality of drug information contained therein.19 We were also interested to know whether the practice setting (ie, community independent pharmacy, large chain pharmacy, or on-line pharmacy) might influence any counseling services that are provided. We selected a total of 90 pharmacies, 30 from each category, as sites for evaluating counseling services. Our analysis was limited to 70 respondent pharmacies. Every chain store pharmacy initially selected was evaluable. Of the 30 independent community pharmacies, 29 were evaluable. However, only 11 of the 30 on-line pharmacies responded to our prescription submission, even allowing three attempts over 3 months; therefore, these were the only evaluable on-line pharmacies. The data form that we prepared asked whether any of 21 different informational procedures about hydrochlorothiazide were undertaken at any given pharmacy. A three-facet rating scale model that constitutes the Rasch analysis yielded a counseling index for each pharmacy, and simultaneously ranked each item on the data form for its degree of difficulty. Any of the 21 items on the data form that were either omitted by every pharmacy or that were implemented by every pharmacy do not influence the counseling index obtained. The monitors who presented prescriptions at each pharmacy were between the ages of 18 and 45, appeared to be in good health, and presented themselves as receiving this prescription for the first time, having no known allergies, taking no other medications, and having no other medical conditions. The monitors received training on the completion of the American Journal of Therapeutics (2003) 10(2) 100 data forms, and when tested on their scoring of three different vignettes, their rankings were highly correlated. Differences between monitors were modeled in the many-facet Rasch model, which, in fact, may be an advantage to this approach. Monitors were instructed to complete data forms immediately after receiving their prescriptions (or their prescription information from on-line pharmacies) to reduce recall bias. Practice setting did not influence the average counseling index. These values averaged −1.33, −1.51 ,and −1.05 for large chain, on-line, and community independent pharmacies, respectively. These differences were not statistically significant, and were consistent in that all practice settings exhibited, on average, comparably adequate counseling indices. Provision of relatively thorough drug information is not out of the reach of community pharmacies. Eleven pharmacies exhibited high scores (>0), and eight of those (73%) were community independent pharmacies, two (18%) were large chain pharmacies, and one (9%) was an on-line pharmacy. Only 27% of the pharmacies spontaneously counseled our monitors, and this is in line with the frequency of spontaneous counseling reported by others.20 This study has several limitations. We studied only a limited number of pharmacies in northwest Ohio and southeast Michigan. It is impossible to know whether these findings would be reproduced elsewhere in the US. Additionally, we evaluated the prescription counseling for a single drug, hydrochlorothiazide, under the least complicated conditions of use. It cannot be known whether the likelihood of pharmacist engagement in more extensive prescription counseling activity would have been greater had the patients’ condition or the drug regimen been more complicated. There is no consensus about those elements of drug information that might be viewed as essential or “ideal.” Just how much information might be necessary to reduce ADEs is an open question. On the other hand, the inclusion on our data form of drug information about hydrochlorothiazide that may be regarded by some as either unnecessary or as too much detail for customary counseling neither prejudges nor affects the ability of any pharmacy to achieve a high counseling index because Rasch analysis simply omits from its computation of individual counseling indices any of those counseling items that are simply not undertaken by any pharmacy. Among the pharmacies in this study, 69% actually provided prescription counseling. This may signal an improvement over the extent of counseling provided just 3 to 4 years ago. In the greater Kansas City Missouri area, only 30% of pharmacies provided prescripAmerican Journal of Therapeutics (2003) 10(2) SCHATZ ET AL tion counseling.21 The experience with community pharmacies in Kansas City and also in Nebraska12 (described previously) may have been fairly representative of all pharmacies, including those in acute care settings. A national clinical pharmacy services study conducted in 1998 reported that between 40% and 50% of hospital pharmacies provided prescription counseling to inpatients either during hospitalization or at discharge,22 whereas another analysis undertaken at hospitals in Columbus, Ohio indicated that fewer than one-third of hospital pharmacists counseled patients.23 Though at first blanche these rates of prescription counseling may appear low, they do signal a clear improvement over those observed before the implementation of OBRA 1990. For example, a study of 80 community pharmacists in New Mexico that was undertaken just before the implementation of OBRA 1990 found that prescription counseling was provided with about 20% of the prescriptions. For patients whose prescriptions were paid for by capitation, the counseling rate dropped to about 8%.24 Thus, the extent of counseling offered seems to be trending upward as follows: 10% to 20% pre-OBRA, 30% to 50% in the late 1990s, and ∼70% currently. If our assessment of prescription counseling for hydrochlorothiazide can be extrapolated, it would suggest that the objectives of OBRA 1990 on the quality of counseling services provided by pharmacies are adequately being met. Ultimately, however, the issue of prescription counseling effectiveness must be addressed by outcome studies. This conclusion is supported by the findings of others. For example, an analysis by the FDA of 12year trends in prescription counseling both in pharmacies and in physician offices has likewise led to the conclusion that there has been moderate progress on verbal counseling stemming from OBRA 1990.25 On the other hand, a survey of community pharmacy clients with asthma indicated that community pharmacists rarely provided information regarding either the role of their asthma medications, techniques for appropriate inhaler use, or general advice regarding asthma control.26 Taken together, these findings suggest that prescription counseling can generally be characterized as adequate or moderate, though not good or excellent. This should be cause for concern, particularly when many studies that have evaluated outcomes in patient safety and well-being with direct pharmacist involvement in the therapeutic regimen, have demonstrated benefits.27–31 In-house counseling of hospitalized patients receiving combination lipidlowering therapy after coronary artery bypass graft 101 OBRA 1990 surgery coupled with weekly pharmacist-patient telephone contact for the next 12 weeks dramatically increased patient compliance measured both 1 and 2 years later.32 The quality of counseling services rendered by pharmacists in Great Britain has also been criticized, though in these studies it was advice pertaining to over-the-counter (OTC) drugs that was evaluated.33 Canadian pharmacists do not seem particularly eager to become more engaged in providing information about OTC drugs, either.34 Whether the barriers that limit pharmacist advicegiving on OTC drugs in Great Britain and Canada are similar to those that might be limiting prescription counseling in US pharmacies remains to be determined; however, a substantial number of potential barriers to widespread prescription counseling in the U.S. have begun to be identified. For example, a survey of asthma patients revealed that only 36% even desired any information on potential druginteractions or side effects, whereas 33% believed that drug information should be provided by their physicians.25 When these same patients were asked whether they would be willing to pay for prescription counseling or to have their insurance pay for it, 76% indicated that they would not be willing to pay and 46% indicated that they would not be willing to have their insurance pay for this service. In another survey of pharmacy provision of asthma patient care, pharmacists were viewed by patients as not substantially contributing to improved control of their condition.35 If willingness to pay is used as a gauge of consumer desire for pharmacy services—including prescription counseling—then a few recent surveys suggest a lack of demand for these services from pharmacists. Metge et al reported that most patients would accept pharmacy services only if they were free.36 Others reported similar findings.37 In contrast, 96.7% of surveyed consumers (n = 197) indicated a willingness to pay for counseling services, but valued this service at $2 to $3.38 More recently, an overall (out-of-pocket) willingness-to-pay figure for pharmacy services was established to range from about $4.00 to $5.50 per prescription.39 Though drug information from pharmacists may not be a highly valued commodity, most adult outpatients do want to learn from their physicians about adverse drug reactions.40 Existing workload exigencies may be responsible for an inertia that is difficult to overcome in common practice settings.41 Prescription counseling accounted for 6.6% of pharmacists’ work-related functions in a grocery chain, or about 30 minutes per shift, with nonjudgmental functions dominating the workday. 42 Thus, a perception by pharmacists that prescription counseling is not highly valued by consumers,43 coupled with lack of compensation for those services, the difficulty of overcoming the inertia of extant work routines including a lack of time, staffing, and space to accommodate this function, might all serve as barriers to a more active role for pharmacists in prescription counseling.44,45 Even though much of the information about drugs placed on consumer-targeted websites is accurate, consumers have difficulty in interpreting the information, and reliable sources of information are still required.41 Safety-minded statutory changes such as those in OBRA 1990 might be more likely to effectively meet their objectives if they include provisions for reimbursement for fundamentally new services. CONCLUSIONS In conclusion, we found that when presented with a nonrefillable prescription for oral hydrochlorothiazide (25 mg once a day), pharmacies generally provided adequate prescription counseling services. This was true whether the pharmacies were community independent, large chain, or on-line pharmacies. Sixty-nine percent of these pharmacies met minimum state standards for counseling by offering to provide prescription counseling. Nine pharmacies (13%) declined to make a pharmacist available on request, and this probably was a function of manpower shortages (too few pharmacists per client) rather than an unwillingness to provide prescription counseling. The Rasch analysis that we used yielded counseling index scores that simultaneously provided both a quantitative and qualitative assessment of the prescription counseling function. This analysis did not require any subjective ranking of the service whatsoever, and the outcome is therefore free from rater or investigator biases. The Nominal Group Technique is one of several more subjective methods developed for assessing other types of health care delivery that is currently being used in the evaluation of pharmacy counseling services, too.46 This study was not designed to assess outcomes associated with prescription counseling. Because of the limitations of our study, our findings can only be considered as highly preliminary. But if they are validated by similar protocols or by other strategies, we will have to address the structural problems in the delivery of prescription counseling and reimbursement thereof in a much more concerted way if we are likely to have any appreciable success in reducing the adverse consequences tied to prescription drug use, as was envisioned when OBRA 1990 was fashioned. American Journal of Therapeutics (2003) 10(2) 102 Although the objectives of OBRA 1990 vis-á-vis prescription counseling services appear to be adequately met, only proper outcomes research can establish whether the objectives of reduced ADEs and related health care costs are also being met. ACKNOWLEDGMENTS The authors gratefully acknowledge the assistance of Drs. Monica Holiday-Goodman, James Tita, Luis Jauregui, Vijay Mahajan, Albert Rafanan, and Srinivas Katragadda. REFERENCES 1. Johnson JA, Bootman JL: Drug-related morbidity and mortality. A cost of illness model. Arch Intern Med 1995; 155:1949–1956. 2. Hepler CD, Strand LM: Opportunities and responsibilities in pharmaceutical care. Am J Hosp Pharm 1990;47: 533–543. 3. Col N, Fanale JE, Kronholm P: The role of medication noncompliance and adverse drug reactions in hospitalizations of the elderly. Arch Intern Med 1990;150:841– 845. 4. Dennehy CE, Kishi DT, Louie C: Drug-related illness in emergency department patients. Am J Health-System Pharm 1996;53:1422–1426. 5. Lakshmanan MC, Hershey CO, Breslau D: Hospital admissions caused by iatrogenic disease. Arch Intern Med 1986;146:1931–1934. 6. Bates DW, Cullen DJ, Laird N, et al: Incidence of adverse drug events and potential adverse events. Implications for prevention. ADE Prevention Study Group. JAMA 1995;274:29–34. 7. Willcox SM, Himmelstein DU, Woolhandler S: Inappropriate drug prescribing for the community-dwelling elderly. JAMA 1994;272:292–296. 8. Aparasu RR, Sitzman SJ: Inappropriate prescribing for elderly outpatients. Am J Health Syst Pharm 1999;56: 433–439. 9. Stuck AE, Beers MH, Steiner A, et al: Inappropriate medication use in community-residing older persons. Arch Intern Med 1994;154:2195–2200. 10. Pub L No. 101–508, §4401, 1927(g) (November 5, 1990); OBRA 1990 Regulations, Federal Register, November 2, 1992;57FR(212):4937–4940. 11. Molzon JA: What kinds of patient counseling are required? Am Pharm 1992;NS32:50–57. 12. Scott DM, Wessels MJ: Impact of OBRA ’90 on pharmacists’ patient counseling practices. J Am Pharm Assoc (Wash) 1997;NS37:401–406. 13. Clinical Pharmacology 2000. Patient Education. Available at: http://www.cp.gsm.com. Accessed American Journal of Therapeutics (2003) 10(2) SCHATZ ET AL 14. Drug Facts and Comparisons, Wolters Kluwer Co., St. Louis, 2000, pp 619. 15. Linacre JM: Many-Facet Rasch Measurement, 2nd edition. MESA Press, Chicago, 1994. 16. Linacre JM: A User’s Guide to Facets, Version 3.0. MESA Press, Chicago, 1996. 17. JMP Statistics and Graphics Guide, version 3.1 of JMP. SAS Institute, Inc. Cary, NC, 1995. 18. Montgomery DC, Peck EA: Introduction to linear regression analysis. John Wiley and Sons, Inc., New York, 1982. 19. Svarstad B, Mount JK: Evalutation of written prescription information provided in community pharmacies, 2001 [Center for Drug Evaluation and Research Web site]. December 21, 2001. Available at: http://www.fda. gov/cder/reports/prescriptionInfo/default.htm. Accessed June 19, 2002. 20. Willison DJ, Muzzin LJ: Workload, data gathering, and quality of community pharmacists’ advice. Med Care 1995;33:29–40. 21. Fritsch MA, Lamp KC: Low pharmacist counseling rates in the Kansas City, Missouri, metropolitan area. Ann Pharmacother 1997;31:984–991. 22. Raehl CL, Bond CA: 1998 national clinical pharmacy services study. Pharmacotherapy 2000;20:436–460. 23. Griffith NL, Schommer JC, Wirsching RG: Survey of inpatient counseling by hospital pharmacists. Am J Health Syst Pharm 1998;55:1127–1133. 24. Raisch DW: Patient counseling in community pharmacy and its relationship with prescription payment methods and practice settings. Ann Pharmacother 1993;27:1173– 1179. 25. Morris LA, Tabak ER, Gondek K: Counseling patients about prescribed medication: 12-year trends. Med Care 1997;35:996–1007. 26. Liu MY, Jennings JP, Samuelson WM, et al: Asthma patients’ satisfaction with the frequency and content of pharmacist counseling. J Am Pharm Assoc (Wash) 1999; 39:493–498. 27. Brouker ME, Gallagher K, Larrat EP, et al: Patient compliance and blood pressure control on a nuclear powered aircraft carrier: impact of a pharmacy officer. Mil Med 2000;165:106–110. 28. Berringer R, Shibley MC, Cary CC, et al: Outcomes of a community pharmacy-based diabetes monitoring program. J Am Pharm Assoc (Wash) 1999;39:791–797. 29. Currie JD, Chrischilles EA, Kuehl AK, et al: Effect of a training program on community pharmacists’ detection of and intervention in drug-related problems. J Am Pharm Assoc (Wash) 1997;NS37:182–191. 30. Faulkner MA, Wadibia EC, Lucas BD, et al: Impact of pharmacy counseling on compliance and effectiveness of combination lipid-lowering therapy in patients undergoing coronary artery revascularization: a randomized, controlled trial. Pharmacotherapy 2000;20:410–416. 31. Birtcher KK, Bowden C, Ballantyne CM, et al: Strategies for implementing lipid-lowering therapy: pharmacybased approach. Am J Cardiol 2000;85:30A-35A. 32. Faulkner MA, Wadibia EC, Lucas BD, et al: Impact of OBRA 1990 33. 34. 35. 36. 37. 38. pharmacy counseling on compliance and effectiveness of combination lipid-lowering therapy in patients undergoing coronary artery revascularization: a randomized, controlled trial. Pharmacother 2000;20:410–416. Bissell P, Ward PR, Noyce PR: Appropriateness measurement: application to advice-giving in community pharmacies. Soc Sci Med 2000;51:343–359. Taylor J, Berger B, Anderson-Harper H, et al: Pharmacists’ readiness to assess consumers’ over-the-counter product selections. J Am Pharm Assoc (Wash) 2000;40: 487–494. Kradjan WA, Schulz R, Christensen DB, et al: Patients’ perceived benefit from and satisfaction with asthmarelated pharmacy services. J Am Pharm Assoc (Wash) 1999;39:658–666. Metge CJ, Hendricksen C, Maine L: Consumer attitudes, behaviors, and perceptions about pharmacies, pharmacists, and pharmaceutical care. J Am Pharm Assoc (Wash) 1998;38:37–47. Smith DL: APhA national survey: willingness of consumers to pay for pharmacists’ clinical services. Am Pharm 1983;23:58–64. Carroll NV, Perri M 3rd, Eve EE, et al: Estimating demand for health information: pharmacy counseling services. J Health Care Mark 1987;7:33–40. 103 39. Suh DC: Consumers’ willingness to pay for pharmacy services that reduce risk of medication-related problems. J Am Pharm Assoc (Wash) 2000;40:818–827. 40. Ziegler DK, Mosier MC, Buenaver M, et al: How much information about adverse effects of medication do patients want from physicians? Arch Intern Med 2001;161: 706–713. 41. Curran CF, Oh KE: Sources of drug information available to consumers. Drug Information Journal 2001;35: 539–546. 42. Dupclay L Jr, Rupp MT, Bennett RW, et al: Analysis of grocery chain pharmacists’ work-related behaviors. J Am Pharm Assoc (Wash) 1999;39:74–81. 43. Schommer JC, Wiederholt JB: Pharmacists’ perceptions of patients’ needs for counseling. Am J Hosp Pharm 1994;51:478–485 44. O’Loughlin J, Masson P, Dery V, et al: The role of community pharmacists in health education and disease prevention: a survey of their interests and needs in relation to cardiovascular disease. Prev Med 1999;28:324–331. 45. Raisch DW: Barriers to providing cognitive services. Am Pharm 1993;NS33:54–58. 46. Ward PR, Bissell P, Noyce PR: Criteria for assessing the appropriateness of patient counseling in community pharmacies. Ann Pharmacother 2000;34:170–175. American Journal of Therapeutics (2003) 10(2)