diabetes research and clinical practice 80 (2008) 171–184 available at www.sciencedirect.com journal homepage: www.elsevier.com/locate/diabres Review Disease-specific health-related quality of life instruments among adults diabetic: A systematic review Youness El Achhab a,b,c,*, Chakib Nejjari a, Mohamed Chikri b, Badiaa Lyoussi c a Laboratory of Epidemiology, Clinical Research and Community Health, Faculty of Medicine and Pharmacy, Fez, Morocco Laboratory of Biochemistry and Molecular Biology, Faculty of Medicine and Pharmacy, Fez, Morocco c UFR of Physiopathology and Pharmacology, Faculty of Science-Dhar El Mahraz, Fez, Morocco b article info abstract Article history: This paper provides a systematic review on health-related quality of life (HRQoL) measures Received 27 February 2007 in diabetic patients. For each included study, a description of the measure and its psycho- Accepted 31 December 2007 metric findings is provided. To evaluate these measures, a databases search (Medline, Published on line 14 February 2008 Scopus and Proqolid) was undertaken to identify relevant publications. Instruments were assessed according to predefined inclusion and exclusion criteria. Keywords: Sixteen instruments met the inclusion criteria among 1049 references produced: apprai- Health-related quality of life sal of diabetes scale (ADS), audit of diabetes-dependent quality of life (ADDQoL), diabetes-39 Diabetes mellitus (D-39), diabetes care profile (DCP), diabetes distress scale (DDS), diabetes health profile (DHP- Review 1, DHP-18), diabetes impact measurement scales (DIMS), diabetes quality of life measure (DQOL), diabetes quality of life clinical trial questionnaire-revised (DQLCTQ-R), diabetesspecific quality of life scale (DSQOLS), elderly diabetes burden scale (EDBS), insulin delivery system rating questionnaire (IDSRQ), quality of life with diabetes questionnaire (LQD), problem areas in diabetes scale (PAID), questionnaire on stress in diabetic patients-revised (QSD-R) and well-being enquiry for diabetics (WED). All those instruments have been developed in northern countries. The shortest instrument (ADS) has seven items and the longest (IDSRQ) has 67 items. ADDQoL was widely translated followed by DHP and PAID. Only authors of ADS and DIMS have not involved patients in the construction of instruments. The authors of instruments: ADS, ADDQoL, DHP, D-39, and PAID reported the itemtotal correlation which is ranged from 0.28 to 0.84. The ADS, DQOL, EDBS, IDSRQ, LQD, PAID, QSD-R, and WED have been assessed for test–retest reliability which varies between 0.27 and 0.99. The DQLCTQ-R, DQOL and IDSRQ were not subjected to factor analysis. Responsiveness was assessed in PAID with effect sizes and ranged from 0.32 to 0.65 for interventions. Four domains were responsive to clinical change in metabolic control in DQLCTQ-R. The other instruments were not been formally assessed for responsiveness. This review found evidence that the instruments: ADDQoL, D-39, DDS, DHP1/18, DSQOLS, EDBS and QSD-R had adequate psychometric properties. For future research, responsiveness should be a priority and further study is also required to examine the effect of ethnicity and to determine the validity of these scales in developing countries. # 2008 Elsevier Ireland Ltd. All rights reserved. * Corresponding author at: Laboratory of Epidemiology, Clinical Research and Community Health, Faculty of Medicine and Pharmacy of Fez, B.P 1893, Km 2.2 Route Sidi Harazem, Fez 30000, Morocco. Tel.: +212 69015590; fax: +212 35619321. E-mail address: youness_elachhab@yahoo.fr (Y. El Achhab). 0168-8227/$ – see front matter # 2008 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.diabres.2007.12.020 172 diabetes research and clinical practice 80 (2008) 171–184 Contents 1. 2. 3. 4. 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . Materials and methods . . . . . . . . . . . . . . 2.1. Search strategy . . . . . . . . . . . . . . . 2.2. Inclusion criteria . . . . . . . . . . . . . . 2.3. Data extraction . . . . . . . . . . . . . . . Results . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. Patients and study characteristics 3.2. Description of instruments . . . . . . 3.3. Psychometric finding . . . . . . . . . . . Discussion . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Introduction The research field in quality of life (QoL) has increased enormously since 1990. As QoL represents the effect of an illness on a patient, as perceived by the patient, and yields complementary information to medical or epidemiological data, it is often used as an outcomes measurement. QoL is commonly recognised as a multidimensional concept including domains of physical health and functioning, mental health, social functioning, satisfaction with treatment, concerns about the future and general well-being. However, QoL is a central issue for patients, providers, and policy makers, and interest in health-related quality of life (HRQoL) has increased markedly in recent years [1]. The term ‘Healthrelated Quality of Life’ (HRQoL) is used because aspects of life exist that are not generally considered as ‘health’ [2]. HRQoL is the value assigned to duration of life as modified by the impairments, functional states, perceptions, and social opportunities that are influenced by disease, injury, treatment, or policy [3]. Health outcomes research for chronic illness is becoming increasingly concerned with patient’s evaluations of the clinical effectiveness of care and treatment. From the point of view of the patient, relevant health outcomes include not only physiological measures, but also subjective factors such as disease self-management burden, social and role functioning, emotional health and physical functioning [4]. These subjective factors are especially important for people with diabetes mellitus because the disease is primarily selfmanaged and self-management regimens affect virtually all aspects of daily life. The HRQoL may be studied by generic or disease-specific questionnaires, depending on the research question [1]. Generic instruments are used in general population to assess a wide range of domains applicable to a variety of health states, conditions and diseases [5]. Disease-specific instruments can include aspects of health considered by patients or clinicians to be of greatest importance. The targeted focus of disease-specific instruments has the potential to make them more responsive to changes in health and together with more detailed and accurate assessment of patients concerns, this makes them important primary endpoints in clinical trials designed to measure changes in HRQoL. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172 172 172 172 173 173 173 173 176 181 182 182 There are a considerable number of measures of HRQoL specific to diabetes. This can be confusing for clinicians and researchers who are interested in measuring the HRQoL of patients with diabetes but are faced with several instruments offering different approaches to measurement. This paper provides a systematic review of diabetes-specific measure of HRQoL, with a focus on content and measurement properties. 2. Materials and methods 2.1. Search strategy Search strategies for this review were designed to retrieve references relating to the development of measures of HRQoL for people with diabetes, including reviews of such instruments. MEDLINE was searched through PubMed (30 January 2007) using the following MeSH terms: ‘quality of life’, and ‘diabetes mellitus’. Other electronic databases including Scopus and Proqolid were searched. A hand-search was conducted of relevant journals (e.g. Quality of Life Research, from 1996 to January 2007; Health and Quality of life Outcomes, from 2003 to January 2007; Diabetes Care, from 1978 to January 2007; Diabetes Research and Clinical Practice, from 1985 to January 2007; Diabetologia, from 1965 to January 2007; and the Diabetes Educator, from 1980 to January 2007). The names of identified instruments were used as terms for further search of the electronic databases. Additional searches were also performed by PubMed for ‘related articles’ to those already found, and by reading the reference lists of the articles found. The reference lists of existing reviews of HRQoL measures using in diabetes [6–8] were also reviewed. 2.2. Inclusion criteria The selection of studies included in this review was restricted to those with a primary focus on the development, reliability and validity of disease-specific HRQoL measures. Studies that focused on children and adolescents were not considered, as descriptive reports that did not address methodology and measurement issues. The review was restricted to instruments that have been evaluated in adult patients. diabetes research and clinical practice 80 (2008) 171–184 2.3. Data extraction Data extraction followed predefined criteria which are considered important in the assessing quality of life instruments [9] including reliability and validity. Study and patients characteristics have been also extracted. Reliability is concerned with the temporal stability of instrument scores and, for multi-item instruments, internal consistency. Test–retest reliability assesses score consistency over two points in time, assuming no change in the underlying health state. Internal consistency reliability is assessed following a single application and evaluates the relationship between all items and their ability to measure a single underlying domain [9,10]. Reliability estimates falling between 0.70 and 0.90 are recommended for instruments intended for groups and individuals respectively [9,10]. Validity assesses whether an instrument measures what it purports to measure. Validity can be evaluated qualitatively through examination of instrument content and quantitatively through factor analysis and comparisons with related variables. These two forms of validity are qualitative matters of judging whether an instrument is suitable for its proposed application [10]. Factor analysis and principal component analysis provide empirical support for the dimensionality or internal construct validity of an instrument. External construct validation includes comparisons with other instruments and relating instrument scores with clinical and sociodemographic variables. The results of these comparisons have been extracted. Responsiveness refers to an instrument’s ability to detect change [9,10]. Responsiveness is considered as the longitudinal construct validation process. Assessment of responsiveness involves statistical estimation of an effect size or change score. For assessing the relative size of change; an effect size of 0.2 being considered small, 0.5 as medium and 0.8 or greater as large [10]. 3. Results The initial search of the MEDLINE database used the MeSH term ‘‘quality of life’’ yielded 58,712 articles. The addition of ‘‘diabetes mellitus’’ diminished the number to 1433 articles, of which 378 were potentially suitable for inclusion. On Scopus, 119,724 articles were yielded using the term ‘‘quality of life’’ which is diminished to 3994 when the term ‘‘diabetes mellitus’’ was added. Among them 536 articles were potentially suitable for inclusion. Thirty-one instruments were founded in the database Proqolid in the field of nutritional and metabolic diseases using the term ‘‘diabetes mellitus’’. Hand searching in relevant journals produced 104 articles. The search using instrument names in Pubmed and Scopus produced 1396 articles and 2752 articles, respectively. Among the 1049 references, 47 were potentially suitable for inclusion. Thirty instruments were eligible for review. Sixteen instruments met the inclusion criteria [11–26]. Fourteen instruments were excluded from the review for different reasons [27–40]. The well-being questionnaire, although widely used in diabetes, is not diabetes-specific and focuses on psychological well being [41]. 3.1. 173 Patients and study characteristics The populations in which the instruments have been developed or evaluated are shown in Table 1. The following topics are addressed: country, sample size, mean age, gender, diabetes type and/or treatment, duration of diabetes. 3.2. Description of instruments The instrument contents, number of items, time taken, original language and number of translations are shown in Table 2. The appraisal of diabetes scale (ADS) is a seven items scale based on theory and previous research [11]. It was developed to assess an individual’s appraisal of his or her diabetes. The seven items use a five-point scale and measure control, uncertainty, coping, affect of diabetes on life goals, predictive view of diabetes and the degree of distress caused by diabetes. The ADS has been used to assess the effects of family environment and work environment on glycaemic control and psychosocial adaptation of adults with diabetes [52,53]. Less than 5 min is the time taken to cover this questionnaire. The audit of diabetes-dependent quality of life (ADDQoL) is a 19 items scale based on a review of existing instruments, discussions with health professionals and interviews with diabetics patients [12,54–56]. The instrument is designed to measure individual’s perceptions of the impact of diabetes on their quality of life. Nineteen diabetes specific QoL domains address projected social, physical and emotional functioning. Each item is scored on a seven-point scale and then the respondent indicates whether the item is very important, important, quite important, or not at all important. The instrument is translated into more than 20 languages. The diabetes-39 (D-39) is a 39 item scale designed to assess the quality of life of patients with diabetes, and covers five dimensions of health: energy and mobility, diabetes control, anxiety and worry, social burden, and sexual functioning [13]. Item selection is based on literature review and interviews with health professionals. This tool uses a visual analogue scale. The D-39 has been translated into more than two languages. The diabetes care profile (DCP) is developed as an instrument to assess social and psychological factors related to diabetes and its treatment [14]. The questionnaire is a comprehensive one and deals with 234 items. It is derived from the diabetes educational profile and the health belief model, and so to a great extent concerns issues related to diabetes knowledge, beliefs, and treatment. However, six subscales of the DCP measure diabetes-specific QoL domains including perceptions of control, personal, social and emotional functioning. This questionnaire takes 30–40 min to be completed. The diabetes distress scale (DDS) is a 17 items scale which builds on the strengths of previously developed instruments (ATT39 [57], QSD-R [25], and PAID [24]) and address at least some of their limitations [15]. The instrument is developed to assess the diabetes-related emotional distress for use in research and clinical practice. In consultation with patients and professionals from multiples disciplines, a preliminary scale of 28 items is developed, based a priori on four distress- 174 diabetes research and clinical practice 80 (2008) 171–184 Table 1 – Populations in which the instruments were evaluated Instrument ADS: appraisal of diabetes scale ADDQoL: audit of diabetesdependent quality of life Country n Mean age Gender (year) (%male) Diabetes type/ treatment (n) Duration (year) USA [11] 200 58.4 100 Insulin (132) 15 UK, Bromley [12] 102 61.6 54 7.3 52 52.4 54 12.7 Portugal [42] 100 61.3 46 Insulin/diet (38), tablet/diet (33), diet (30) Insulin/diet (32), tablet/diet (14), diet (6) Type 2 (73), Type 1 (27) D-39: diabetes-39 USA [13] 516 165 262 52.4 61.7 55.3 46.5 44.8 35.5 Type 1 (159), Type 2 (330) Type 1 (31), Type 2 (128) Type 1 (25), Type 2 (218) 14.2 11.5 10.1 DCP: diabetes care profile USA-community study [14] USA-medical centre study [14] USA [43] USA-Hispanic [44] USA-non-Hispanic white [44] 440 61 45 Type 1 (48), insulin (198) 10 352 54 40 Type 1 (116), insulin (236) 14 672 83 238 63 64.4 66 42 96 96 Type 2 (349) Type 2 Type 2 12.5 15.2 14.2 DDS: diabetes distress scale USA [15] 683 56.3 52.3 12.8 DHP: diabetes health profile UK [16] UK [16] UK [16] UK [45] Denmark [45] 239 2239 233 426 460 40.9 39.8 51.5 61.6 63.6 NR 51 52 57 53.9 Insulin (344), tablet (290), diet (49) Type 1/insulin (239) Type 1/insulin (2239) Type 1 Type 2 Type 2 DIMS: diabetes impact measurement scales USA [17] China [46] 130 219 45 63.5 42 35 Type 1 (51), Type 2 (77) Type 2 11 8 DQLCTQ-R: diabetes quality of life clinical trial questionnaire-revised Multinational study (Canada, France, Germany, and the US) [18] 942 46 56.6 Type 1 (468), Type 2 (474) 12.6 DQOL: diabetes quality of life measure USA [19] USA [47] China [48] 192 240 70 NR 52.6 67.6 59.4 49.2 44.3 Type 1 Type 1 (111), Type 2 (129) Type 2 8 15.2 11 DSQOLS: diabetes specific quality of life scale Germany [20] Germany [49] 657 424 36 36.9 57.9 54.5 Type 1 Type 1 18 12.5 EDBS: elderly diabetes burden scale Japan [21] 455 75.2 36 14.2 IDSRQ: insulin delivery system rating questionnaire LQD: quality of life with diabetes questionnaire USA [22] 197 46.4 47.2 Diet (135), tablet (258), insulin (62) Type 1 (142), Type 2 (45) German [23] 144 57.2 48 Insulin (72), without insulin (72) 12.8 451 1472 36.2 51 0 51 15.9 16 256 52 48 Type 1 (370), Type 2 (81) Insulin (1241), tablet/diet (199) Type 1 (136) 1930 NR 54 Type 1 (915), Type 2 (1015) 12.3 267 52.3 49 Type 1 (70), Type 2 (197) 7.1 UK, Cambridge [12] PAID: problem area in diabetes scale USA [24] Netherlands [50] USA [51] QSD-R: questionnaire on stress in patients with diabetes-revised WED: well-being enquiry for diabetics Germany [25] Italy [26] 12 13.7 13.1 NR NR NR 22.6 15 NR: not reported. The populations in which the instruments have been developed or evaluated are shown in the table. The following topics are addressed: country, sample size, mean age, gender, diabetes type and/or treatment, duration of diabetes. related domains: emotional burden subscale, physicianrelated distress subscale, regimen-related subscale, and diabetes-related interpersonal distress. The diabetes health profile (DHP) is developed to examine psychological well being associated with having diabetes, with specific emphases on psychological distress, barriers to activity and dietary perceptions and behaviour [16]. The instrument content is derived following a literature review, a review of available instruments, interviews with diabetic patients and discussions with diabetes health care profes- 175 diabetes research and clinical practice 80 (2008) 171–184 Table 2 – Instrument contents, number of items, time needed, original language and number of translations Instrument Appraisal of diabetes scale [11] Audit of diabetesdependent quality of life [12] Diabetes-39 [13] Diabetes care profile [14] Diabetes distress scale [15] Diabetes health profile [16] Diabetes impact measurement scales [17] Diabetes quality of life clinical trial questionnaire-revised [18] Diabetes quality of life measure [19] Diabetes specific quality of life scale [20] Elderly diabetes burden scale [21] Insulin delivery system rating questionnaire [22] Quality of life with diabetes questionnaire [23] Problem area in diabetes scale [24] Questionnaire on stress in patients with diabetes-revised [25] Well-being enquiry for diabetics [26] Contents (items) Single index (7) Impact of diabetes (19); global (2) (the original measure was 13 items and revised to 18 items) Energy and mobility (15); diabetes control (12); anxiety and worry (4); social/peer burden (5); sexual functioning (3) General (234); QoL domains: control problems (18), social and personal factors (13), positive attitude (5), negative attitude (6), self-care ability (4) Emotional burden (5); physicianrelated distress (4); regimenrelated distress (5); diabetesrelated interpersonal distress (3) Psychological distress (14 in DHP-1, 6 in DHP-18); barriers to activity (13 in DHP-1, 7 in DHP-18); disinhibited eating (5 in DHP-1, 6 in DHP-18) Diabetes-specific symptoms (5); nonspecific symptoms (11); wellbeing (10); diabetes-related morale (9); social role fulfilment (5) Generic QoL from SF-20 and 36 (28); satisfaction from DQOL (9); treatment satisfaction (3); treatment flexibility (10); frequency of symptoms (7) Treatment satisfaction (15); impact of treatment (20); worry: diabetesrelated (4); worry: social/vocational (7); overall well-being (1) Perceived burden of diabetes (44); treatment goals (10); treatment satisfaction (10) Social burden (5); dietary restrictions (4); worry about diabetes (4); burden by tablets/insulin (3); treatment dissatisfaction (3); symptom burden (4) Treatment satisfaction (15); daily activity interference (11); clinical efficacy (9); diabetes worries (6); psychological well being (15); social burden (7); overall preference (4) Diabetes satisfaction (7); diabetes stress (7); blood glucose stress (3) Diabetes-related emotional problems (12); treatment-related problems (3); food-related problems (3); social support-related problems (2) Leisure time (4); depression/fear of future (6); hypoglycaemia (4); treatment regimen/diet (9); physical complaints (6); work (6); partner (6); doctor–patient relationship (4) Symptoms (20); impact (20); discomfort (10); serenity (10) Time needed Original language Number of translations <5 min – English for the USA English for the UK – More than 20a – English for the USA More than 2 30–40 min English for the USA – – English for the USA – – English for the UK 14a 15–20 min English for the USA 3a 10 min English for the USA 2a – English for the USA 4a 10–20 min German 1a 5–10 min Japan 1 – English for the USA – – German – 3–5 min English for the USA 9a 5–15 min German – – Italian – Table shows the description of each instrument, included in the review, in terms of: instrument contents, number of items, time taken, original language and number of translations. a PROQOLID: www.qolid.org (last accessed on November 2007). 176 diabetes research and clinical practice 80 (2008) 171–184 sionals. Original validation studies resulted in a 32-item, three factor questionnaire, developed for use among adult insulindependent and insulin-requiring patients in an ambulatory care setting (DHP-1) [16]; this is later revised in a cross-cultural study to an 18-item version with the same three factors, modified for use within type 2 diabetics patients (DHP-18) [45]. This questionnaire has been translated into 14 languages. The diabetes impact measurement scales (DIMS) is a 40item scale designed to measure longitudinal changes in health status in diabetes patients for application in clinical trials [17]. The instrument is developed following a review of the literature and existing instruments and discussions with diabetes health-care professionals. The five factors of the DIMS reflect general well being, physical symptoms, diabetesrelated morale, and social functioning. Items use between four and six-point scales. The time taken to complete this questionnaire ranged from 15 to 20 min. It is translated into Chinese, French and Italian. The diabetes quality of life clinical trial questionnairerevised (DQLCTQ-R) is developed based on DQLCTQ. The revised version contains only 57 questions and 8 generic as well as disease-specific domains: physical function, energy/ fatigue, health distress, mental health, satisfaction, treatment satisfaction, treatment flexibility, and frequency of symptoms [18]. The developers of the draft questionnaire included previously validated measures (SF-36 and DQOL) and developed new items as needed. The DQLCTQ questionnaire is originally composed of 142 items and is designed for use in a clinical trial to measure the QoL changes in patients receiving insulin lispro [58]. These items are Likert scaled and takes 10 min to be completed. The instrument is translated into French and German. The diabetes quality of life measure (DQOL) is a 46 item scale. It is developed for use in the diabetes control and complications trial (DCCT) to compare two treatment regimens for chronic complications in patients with Type 1 diabetes [19]. However, its structure allows application in Type 2 patients [59]. Instrument content is derived from literature review and consultation with patients and clinicians. Item scoring is a five-point Likert scale. The instrument is translated into Chinese, French, Spanish and Turkish. The diabetes-specific quality of life scale (DSQOLS) is a 64item scale based on a review of existing diabetes-specific QoL questionnaires, group discussions with type 1 diabetes patients and review by diabetes healthcare professionals [20]. The instrument is developed to measure the quality of life of type 1 diabetes patients. The 39 quality of life items of this instrument form six dimensions: social relations, leisure tile flexibility, physical complaints, worries about the future, diet restrictions and daily hassles. Items use six-point scales. Filling in the questionnaire takes less than 20 min. In a later study, five subscales were added to the questionnaire [49]. It is translated from German to English for UK. The elderly diabetes burden scale (EDBS) is a short version of the elderly diabetes impact scale (EDIS) which consisted of 37 items [60]. The authors are selected 23 items that rated on a four-point multiple-choice scale and are developed the EDBS [21]. The EDBS is consisted of six subscales: symptom burden, social burden, dietary restrictions, worry about diabetes, treatment (dis)-satisfaction, and burden by tablets or insulin. Less than 5 min is the time taken to cover this questionnaire. It is translated from Japan to English. The insulin delivery system rating questionnaire (IDSRQ) is a 67-item developed through a three-step procedure basing upon literature review, interviews with diabetic patients and experience of authors [22]. The IDSRQ is comprised seven multi-item subscales, one for each section of the questionnaire. Three subscales are asked questions specific to the respondent’s insulin delivery system and other three subscales are more general. One subscale is assessed overall preference for the insulin delivery system. The score for each item is a metric ranging from 0 for the lowest response option to 100 for the highest response option, with equal distance between response categories. Scale scores are computed as well as the mean of the completed items. The quality of life with diabetes (LQD) questionnaire is a 17item scale developed from DQOL as well as a result from the feedback of the patients [23]. The LQD is addresses the satisfaction with life and the burdens of diabetes and its treatment. The 17 items, which are answered on a five-point scale, refer to the month before the test. The problem areas in diabetes scale (PAID) is a 20-item scale, single-factor measure of diabetes related distress and developed by researchers associated with the Joslin Diabetes Center and Harvard Medical School [24]. The 20-item cover a range of emotional problems. Items were developed from patient interviews, input from diabetes health care professionals and pilot testing. Filling in the questionnaire takes less than 5 min. It is now translated into 9 languages. The questionnaire on stress in patients with diabetesrevised (QSD-R) is a 45-item scale designed to assess psychological stress associated with problems in daily living with diabetes [25]. The items define eight stress scales for patients with diabetes: leisure time, depression/fear of future, hypoglycaemia, treatment regimen/diet, physical complaints, work, partner, and doctor–patient relationship. The original QSD comprises ninety items selected following literature reviews and interviews with diabetologists and patients [61]. Respondents are asked to indicate if the situation causes them stress, and if so, to rate the severity of stress on a five-point scale. The time taken to complete this questionnaire is ranged from 15 to 20 min. The well-being enquiry for diabetics (WED) is a 60-item scale measure disease-related quality of life [26]. The items provide an evaluation of four areas of quality of life: symptoms, discomfort, serenity, and impact. Instrument content was developed from a review of existing diabetesspecific QoL measures and input from patients and diabetes health care professionals. The items are Likert scaled. 3.3. Psychometric finding Psychometric information about HRQoL measures is presented in Table 3. All studies of the 16 measurement report the internal consistency reliability (Cronbach’s a coefficient). With the exception of DCP, DIMS, DQOL, IDSRQ, EDBS and QSD-R the level of Cronbach’s a exceed 0.7, the criterion recommended for studies involving groups of patients [10,63]. Another approach to establish internal consistency of items is simply to examine the correlation of individual items to the Table 3 – Psychometric evaluation of diabetes-specific health-related QoL measures Instrument Reliability Cronbach’s a Item total correlation Validity Test– retest Scale analyses Convergent/ discriminatory validity 0.73 0.28–0.59 0.85–0.89 Single factor explaining 39% of variance ADDQoL [12,42] 0.84 [12], 0.89–0.90 [42] 0.37–0.67 [12] NR D-39 [13] 0.81–0.93 0.45–0.84 NR All items loading >0.4 on one factor [12,42]; mean weighted score with QoL (r = 0.31) and QoL without diabetes (r = 0.47) [12] Five factors accounted for approximately 90% of the total variance DCP [14,43,44] 0.60–0.95 [14], 0.54–0.97 [44] NR NR GFI = 0.92 HSS (r = 0.27 to 0.32); CESD (r = 0.53 to 0.48); SPS (r = 0.34 to 0.32); BDI (r = 0.53 to 0.45); DFBC (r = 0.33 to 0.36) DDS [15] 0.93 NR NR CESD (r = 0.56) DHP [16,45] 0.77–0.88 [16], 0.70–0.88 [45] 0.47–0.75 [16], 0.40 [45] NR Correlation between 28-item/17-item (r = 0.99). Mean correlation between subscales/ 17-item (r = 0.82) Three factors explained 33–35% [16] and 40–46%[45] of the total variance; SI: 0.30–0.70 DDHS (r = 0.59); DRAQ-R (r = 0.17); DHBQ (r = 0.31–0.42); PSS (r = 0.49); PSI (r = 0.39–0.55) NR SF-36: related variables (r = 0.15 to 0.71); unrelated variables (r = 0.20 to 0.68) HADS (r = 0.28–0.62); SF-36 (r = 0.17 to 0.62) Differences between groups Related variables NR HbA1c (r = 0.18) NR Greater impact of diabetes on QoL in insulin treated; microvascular complications [12,42]; perceptions of hypoglycaemia [12] NR NR Diabetes severity (r = 0.15–0.56); group insulin/oral therapies had the least desirable scale score; elderly scored lower for diabetes control, anxiety/worry, social burden; males scored higher for sexual functioning T2D not using insulin associated with less impact on social/ personal life; T1D reported more control problems. No differences between Hispanic and non-Hispanic NR NR NR GHb values (r = 0.33 to 0.21) NR Age (r = 0.29); MP (r = 0.30); exercise (r = 0.13); TC (r = 0.20) NR NR NR Psychological distress and disturbed eating affect younger women then men diabetes research and clinical practice 80 (2008) 171–184 ADS [11] Responsiveness 177 Cronbach’s a Reliability Item total correlation Validity Test– retest 0.60–0.85 [17], 0.61–0.86 [46] NR NR [17], 0.55–0.92 [46] DQLCTQ-R [18] 0.77–0.90 ICCa > 0.70 DQOL [19,47] 0.66–0.92 DSQOLS [20] Convergent/ discriminatory validity NR One factor accounting for 32% variance [17]; SI: 0.49–0.97 [17], 0.32–0.81 [46] NR PRDC (r = 0.22–0.55); CRDC (r = 0.24–0.35); PRGW (r = 0.27–0.47); CRGW (r = 0.29–0.45) NR NR 0.78–0.92 SI: 0.26–0.68 SC-90 (r = 0.40–0.77); ABS (r = 0.25 to 0.67); PAIS (r = 0.06–0.81); SF-36 (r = 0.003 to 0.59) [47] 0.70–0.88 NR NR GFI = 0.81; six-factor solution explaining 50.1% of the total variance; SI: r = 0.28–0.66 Positive well-being scale (r = 0.35–0.53) EDBS [21] 0.55–0.89 NR 0.94–0.99 Six-factor solution explaining 69.4% of variance PGC morale scale (r = 0.51); GDS (r = 0.27–0.57); MMSE and AFD (not significant) IDSRQ [22] 0.67–0.92 NR 0.67–0.94 NR NR LQD [23] 0.71–0.83 NR 0.53–0.77 (3 days) 0.27–0.74 (30 days) Three factors having high loadings on the three factors (>0.63) NR Differences between groups Women reporting more symptoms Better HRQoL associated with: tight metabolic control; T1D; male patients, and high selfperceived control of diabetes Adult females reporting more worries and impact of diabetes; taking insulin associated with less satisfaction, greater impact, and fewer worries Age (r = 0.23–0.01); social status (r = 0.04–0.24); better QoL associated with greater flexibility of insulin treatment, fewer complications and use of rapid-acting insulin Women had higher scores of dietary restrictions, worry, and less satisfaction of treatment but more adaptive feeling to diabetes than men Differences between injection users and CSII users ranged from 0.3 to 1.4 Diabetic with complications had a lower QoL on diabetes satisfaction and diabetes stress Related variables HbA1c negatively correlated; age positively correlated NR NR NR NR HbA1c (r = 0.24– 0.00); DD (r = 0.22–0.02); age at onset (r = 0.14–0.11); FH (r = 0.23 to 0.03) Age (r = 0.14) HbA1c and FH (P < 0.05) NR Four domains were responsive to clinical change in metabolic control NR NR NR NR NR diabetes research and clinical practice 80 (2008) 171–184 DIMS [17,46] Scale analyses Responsiveness 178 Table 3 (Continued ) Instrument 0.93–0.95 0.32–0.84 0.83 Large single factor explaining 50–52% of variance QSD-R [25] 0.69–0.81 NR 0.45–0.73 Eight factors explaining 51% of variance WED [26] 0.81–0.84 NR 0.68–0.89 Four factors explaining 50% of variance GSIBSI (r = 0.63); ATT39 (r = 0.22 to 0.81); DCM-a (r = 0.05–0.59); DCM-pr (r = 0.01 to 0.70); DCM-ts (r = 0.13 to 0.82); WBQ (r = 0.50 to 0.53); HFS (r = 0.53 –0.57); STAI (r = 0.61) STAI (r = 0.62);s BDI (r = 0.61) DQOL (r = 0.05 to 0.68); SATA (r = 0.13 to 0.63); HDRS (r = 0.29 to 0.49); BITE (r = 0.26 to 0.35) T1D reported more diabetes-related distress than T2D HbA1c (r = 0.30); self-monitoring blood glucose frequency (r = 0.13) Age (r = 0.11); DD (r = 0.12) Patients with poor metabolic control had higher stress; T1D reported high stress score, having more complications, and inpatient status Better QoL associated with younger age and being male; worse QoL associated with using insulin in T2D and mental disorders NR NR HbA1c (r = 0.06–0.35) NR Effect sizes range from 0.32 to 0.65 for interventions [62] r: correlation coefficient; ABS: affect balance scale; AFD: adaptive feeling to diabetes; BDI: beck depression inventory; BITE: bulimic investigation test Edinburgh; CESD: the centre for epidemiological studies depression scale; CRDC: clinician-rated diabetes control; CSII: continuous subcutaneous; DCM-a: diabetes coping measure-avoidance; DCM-pr: diabetes coping measure-passive resignation; DCM-ts: diabetes coping measure-tackling spirit DD: diabetes duration; DDHS: diabetic daily hassles scale; DFBC: diabetes family behaviour checklist; DHBQ: diabetes health belief questionnaire; DRAQ-R: diabetes regimen adherence questionnaire-R; FH: frequency of hypoglycaemia; GDS: geriatric depression scale; GFI: goodness of fit index; GHb: glycated haemoglobin; GSIBSI: global severity index of brief symptom inventory; HbA1c: glycosylated haemoglobin; HDRS: Hamilton depression rating scale; HFS: hypoglycaemia fear survey insulin infusion; HSS: happiness and satisfaction scale; ICC: intraclass correlation coefficient; MMSE: mini-mental state examination; MP: adherence to meal planning; PAIS: psychosocial adjustment to illness scale; PGC: Philadelphia geriatric center; PRDC: patient-rated diabetes control; PRGW: patient-rated general wellness; PSI: psychiatric symptom index; PSS: perceived stress scale; SATA: state anxiety and trait anxiety; SC-90: symptom checklist-90; SI: scale intercorrelations; SPS: social provisions scale; STAI: state trait anxiety inventory; T1D: Type 1 diabetic; T2D: Type 1 diabetic; TC: total cholesterol; WBQ: well-being questionnaire. A description of psychometric information about HRQoL measures included in the review: reliability, validity and responsiveness. a Except for diabetes worry and social stigma. diabetes research and clinical practice 80 (2008) 171–184 PAID [24,50,51] 179 180 diabetes research and clinical practice 80 (2008) 171–184 scale as a whole. This approach assesses the scale homogeneity. Items should correlate at least 0.2 with the remainder of the scale [10,64]. The authors of instruments: ADS, ADDQoL, D-39, DHP, and PAID reported the item-total correlation, which is ranged from 0.28 to 0.84 [11–13,17,24]. Test–retest reliability or reproducibility is commonly examined by means of a correlation coefficient. As with internal consistency reliability, minimal standards for reproducibility coefficients are also typically considered to be 0.70 for group comparisons [9,10]. The ADS, DIMS, DQOL, EDBS, IDSRQ, LQD, PAID, QSD-R, and WED are assessed for test– retest reliability that varies between 0.27 and 0.99. The authors of DIMS, IDSRQ, LQD, QSD-R and WED are reported a poorly level (lower than 0.7) of test–retest reliability [22,23,25,26]. Factor analysis of the ADS suggests that is a onedimensional measure; users may wish to examine specific items focusing on diabetes-related distress, perceived control and interference with daily life. Thirty-nine percent of variance was explained by the single factor [11]. On ADDQoL, factor analysis with oblimin rotation is produced three factors and all items load greater then 0.4 [12]. Factor analysis with varimax rotation was performed on EDBS, LQD and DSQOLS. This statistical test is indicated a six-factor solution explaining 69.4% of the total variance on EDBS [21], and is represented three factors having their high loadings (more than 0.63) on the three factors on LQD [23]. On DSQOLS the scree-test indicated a six-factor solution explaining 50.1% of the total variance [20]. A forced three factor Principal Axis Factoring analysis (PAF) with varimax rotation was carried out on the 32 items of DHP-1 and accounted for no more than 35% of the total explained variance, and for no more than 46% on DHP-18, leaving over half of the variance unexplained [16]. Factor analysis of the D-39 revealed that five unrotated factors accounted for approximately 90% of the total variance [13]. An exploratory factor analysis was performed on DDS using principal factor analysis with Promax rotation, revealed four factors most consistent and interpretable [15]. Goodness of fit index (GFI) was searched by authors of DCP and DSQOLS through confirmatory factor analysis. This index was 0.92 for DCP and 0.82 for DSQOLS [14,20]. Principal component analysis of QSD-R yielded eight factors that all together explained 50.7% of the variance [25], but this analysis was performed on DIMS and authors recommend use of an average of subscale scores for use as a general index of HRQoL [17]. Principal components and confirmatory factor analyses of the PAID confirmed the existence of a large single factor explaining most of the item variance [51]. Factor analysis of WED revealed multiple factor loadings, but authors retained the original categorisation of items for conceptual reasons [26]. The DQLCTQ-R, DQOL and IDSRQ was not subjected to factor analysis or principal component analysis [18,19,22,43,44]. Only authors of ADS and DIMS have not involved patients in the construction of instruments [11,17]. Content or face validity is used in the development of the other instruments. Professional of health, diabetes experts and patients were contributed in the development of the instruments. Construct validity of the instruments included correlation with other instruments and global judgement of health and comparisons with related variables and between groups. Correlation between instrument’s score and HbA1c was searched by the authors of ADS, DCP, DDS, DIMS, DSQOLS, EDBS, PAID, QSD-R, and WED. Pearson correlation coefficients were computed by the authors of ADS and were ranged from 0.17 to 0.59 with related measures [11]. The correlation with HbA1c levels was moderate (r = 0.18) [11]. Weighted ADDQoL scores indicated greater negative impact of diabetes on QoL for insulin-treated respondents and reported complications [12]. Construct validity was evaluated in the development of D-39 by using the SF-36. Strong negative correlations of the five scales in the revised 39-item instrument with the nine scales in the SF-36 were determined and ranged from 0.15 to 0.71 [13]. Significant correlations were found between many of the DCP profile scales and independent psychological and social measures. Three scales were significantly correlated with glycated haemoglobin level. Using the DCP, worse QOL is associated with higher glycaemic levels, use of insulin versus tablets and having a greater number of diabetes complications [14,43,44]. Ethnicity had no impact on the DCP scale scores [44]. On the DDS instrument, subscale scores were mostly unrelated to HbA1c but were consistently and positively linked to total cholesterol [15]. This instrument has certain potential advantages over previous instruments, like the ATT39 [57], PAID [24] and QSD-R [25]. It is shorter, and the new subscales allow direct comparison of four different types of distress [15]. The DHP-1 scores were not associated with chronic complications [16]. Psychological distress and barriers to activity subscales were correlated with the hospital depression and anxiety scale and subscales of the SF-36 [16]. The DHP-18 scales of psychological distress and barriers to activity were significantly higher for insulin-treated patients compared with patients with less treatments demands [45]. The three subscale scores for each language group were correlated with age [45]. Total DIMS scores were significantly correlated with HbA1c levels, sex and diabetes complications index, although no significant correlations were found between DIMS scores and complications index [17]. Validity of DQOL was measured by comparison with three established instruments. The total DQOL score and the satisfaction and impact scales consistently showed significant correlations with these measures [19]. In a study on DQOL, the authors showed that the DQOL compares favourably to the SF-36, although they report that the SF-36 is less sensitive to lifestyle issues such as diet or treatment [47]. Women reported greater impact of diabetes and greater diabetes-related worries [19]. Among DQLCTQ-R, better HRQoL was associated with male patients and patients with type 1 diabetes, tight metabolic control, and high self-perceived control of diabetes [18]. The DSQOLS dimension score had small to moderate levels of correlations with the positive well-being scale (r = 0.35–0.53) [20]. The DSQOLS scores for physical complaints and treatment satisfaction were correlated with HbA1c levels. The majority of dimension scores were significantly related to the presence of late complications. Dimension scores for physical complaints and diet restriction had significant small levels of correlation with social status. Dimension scores for leisure time flexibility, worries about the future and diet restrictions were significantly related to the type of insulin treatment. diabetes research and clinical practice 80 (2008) 171–184 The high scores of some subscales and total EDBS were significantly associated with high HbA1c level, frequency of hypoglycaemia, and insulin therapy, showing construct validity [21]. Convergent validity of these measures was assessed with the correlation with the Philadelphia Geriatric Center morale scale (r = 0.51). The subscales of EDBS and total EDBS did not have any significant correlations with the adaptive feeling to diabetes and the Mini-Mental State Examination, showing a discriminant validity of EDBS [21]. Validity was assessed by comparing the difference between groups in IDSRQ and LQD. On IDSRQ, differences in means between injection users and continuous subcutaneous insulin infusion users groups were statistically significant (ranged from 0.3 to 1.4) [22]. People with more than one late complication had a lower QOL on the diabetes satisfaction and diabetes stress scales from the LQD [23]. The PAID was positively correlated with HbA1c levels and it scales were moderately to strongly associated with related measures of general and diabetes-specific stress [24]. PAID scores were not significantly different between subjects with type 1 versus type 2 diabetes and were only weakly correlated with age (r = 0.11) and disease duration (r = 0.12). QSD-R scores were significantly related to HbA1c levels and longterm complications [25]. Correlational analysis indicate significant relationships between the QSD-R total score and scores of the beck depression inventory (r = 0.61) and the statetrait anxiety inventory (r = 0.62). WED scores were significantly related to HbA1c levels for type 1 diabetes patients [26]. Among patients type 2, WED scores for symptoms and serenity were significantly related to sex. WED subscales were related to measures of similar constructs with a correlation coefficient ranging from 0.05 to 0.68 [26]. Responsiveness across the sixteen instruments for the diabetes population studied was evaluated only by authors of DQLCTQ-R and PAID. On DQLCTQ-R, four domains (Treatment Satisfaction, Health/Distress, Mental Health, and Satisfaction from the DQOL) were responsive to clinical change in metabolic control. The other four domains in this instrument were not responsive. This property was evaluated by comparing the baseline and endpoint (6 months) HRQoL scores [18]. Responsiveness across the seven studies on PAID, assessed by effect size, revealed a pattern of small to moderate effects ranged from 0.32 to 0.65 [62]. The other instruments were not been formally assessed for responsiveness. 4. Discussion The past few decades have witnessed considerable research about HRQoL, leading to the development and refinement of a number of diabetes-specific HRQoL measures. This systematic review has focused on instruments developed for assessing HRQoL in patients with diabetes. On the basis of the review, the following measures: DCP, DIMS, DQOL, DSQOLS and the D-39 are recommended in research in which a broad conceptualisation of diabetesspecific QoL is appropriate. DQLCTQ-R had relevant domains on HRQoL and is developed for use in multinational clinical trials. The remainder of instruments is appropriate for 181 investigation of one or more specific aspects of diabetesQoL. DDS, PAID and the QSD-R had a primary focus on diabetes-related distress; the DHP is focused also on diabetesrelated distress and other aspects like the activity and the eating behaviour; ADS and ADDQoL are a single scale questionnaire focused on stressful impact and life without diabetes respectively and the WED had focused on the perceptions of diabetes in relation with mental health. EDBS is a measure of diabetic-specific QOL in elderly people with diabetes mellitus. IDSRQ was focused on HRQoL and treatment preferences for insulin delivery systems. The LQD addresses the satisfaction with life and the burdens of diabetes and its treatment. On the 16 instruments reviewed, 12 were developed in English. The time taken to cover the items of the ADS, DQLCTQ-R, EDBS and PAID was shortened to be less than 10 min for every patient. The authors of ADDQoL, DHP-18, DQOL and LQD were not mentioned the needed time to cover the questionnaire, but these instruments with a reduced number of items do not require much time. Around the world, the ADDQoL is the more translated and validated questionnaire followed by DHP and PAID. Validation was lacking in one or more respects of the reviewed measures. Patients were not involved in the derivation of items for the ADS, DCP, DIMS and D-39. The authors of the ADDQoL, DHP, DQOL and DSQOLS included the concept of content or face validity in the development of their questionnaires. The DSQOLS is specific to type 1 diabetes patients [20]. All the others instruments reviewed appear relevant measures of HRQoL for patients with type 1 and type 2 diabetes patients. Use of factor analysis to support construct validity was not reported in validation studies of the ADS, DQOL, DQLCTQ-R and IDSRQ. Differentiation between groups was not reported by authors of ADS. Linkages of ADDQoL, DQLCTQ-R, IDSRQ and LQD with other measures of QoL were not reported. With the exception of DQOL and DIMS, who has a weaker evidence for reliability, the remainder of instruments produces a good to acceptable reliability. Internal consistency measured by Cronbach’s a coefficient was reported by all the authors of instruments reviewed. Responsiveness to changes in health lacking in all instruments reviewed with exception of the DQLCTQ-R and PAID. This term has been understudied, mainly because of the crosssectional design of the majority of pilot studies. Future research should evaluate this key attribute of QoL instruments. Our methodology has focused only on the published data on Medline, Scopus and Proqolid database. We have not access on the following database: PsychInfo, Embase and Cinahl. As the intention was to study HRQoL and diabetes, the databases searched appear probably sufficient. We have analysed only published results, and sometimes some details were not available in papers. As it was reported by others reviews, we focused on psychometric findings. The criteria of ethnicity was taken into account in some studies, but was not in others. So, we have not considered these criteria in comparisons. Among studies reviewed, one HRQoL measure (EDBS) validated is devoted to elderly populations. Future research is needed focusing on older diabetic adults. 182 diabetes research and clinical practice 80 (2008) 171–184 Prior systematic reviews have been published on HRQoL among adults with diabetes. Our review represents an update. Garratt et al. [6] performed a systematic review of diabetesspecific HRQoL instruments, and identified nine instruments. They note that five of these instruments have demonstrated reliability as well as internal and external construct validity: ADDQoL, DHP-1/18, DSQOLS, D-39, and the QSD-R. In another systematic review, Watkins and Connel [8] selected a 12 measures addressed to the construct of HRQoL and diabetes. They concluded that the measures reviewed offered a broad array of conceptual and measurement approaches and their continued use in diabetes research will further the understanding of the construct of HRQoL. The most instruments we have discussed above were developed and validated in industrialised countries. Due to epidemiological transition, the prevalence of diabetes increases dramatically in low-income countries [65]. It is necessary to develop same specific instruments of HRQoL we can use in those countries. This is very pertinent for public health system, health economics and management of diabetes. Given the variation in the content of the sixteen diabetesspecific instruments, what should potential users do? So, researchers selecting a diabetes-specific QoL questionnaire should consider the content of instruments in relation to the research question. In conclusion, evidence is clear that diabetes has a major impact on the QoL of affected individuals. The instruments of disease-specific HRQoL in diabetes: ADDQoL, DDS, DHP1/18, DSQOLS, D-39, EDBS and QSD-R have good psychometric properties; patients are involved in the development of these instruments, then they offer good field for potential users. For future research, responsiveness should be a priority because it is lacking in the psychometric finding of the instruments included in this review. Further study is also required to examine the effect of ethnicity and to determine the validity of these scales in developing countries. Conflict of interest statement The authors state that they have no conflict of interest. Acknowledgements We thank Mr Ali BAIZ for correcting manuscript language and Mohamed BERRAHO for his assistance. Grant support: No sources of funding were used to assist in the preparation of this review. references [1] R.R. Rubin, M. Peyrot, Quality of life and diabetes, Diabet. Metab. Res. Rev. 5 (1999) 205–218. [2] G.H. Guyatt, D.H. Feeny, D.L. Patrick, Measuring healthrelated quality of life, Ann. Intern. Med. 118 (1993) 622–629. [3] Health outcomes methodology [no authors listed], Med. Care 38 (2000) SII7–II13. [4] J.T. Barr, The outcomes movement and health status measures, J. Allied Health 24 (1995) 13–28. [5] Mapi Research Institute, Instruments by Pathology/Disease, Available from www.qolid.org. Accessed on February 2006. [6] A.M. Garratt, L. Schmidt, R. Fitzpatrick, Patient-assessed health outcome measures for diabetes: a structured review, Diabet. Med. 19 (2002) 1–11. [7] S.L. Norris, Health-related quality of life among adults with diabetes, Curr. Diabet. Rep. 5 (2005) 124–130. [8] K. Watkins, C.M. Connell, Measurement of health-related QOL in diabetes mellitus, Pharmacoeconomics 22 (2004) 1109–1126. [9] Scientific Advisory Committee of the Medical Outcomes Trust, Assessing health status and quality of life instruments: attributes and review criteria, Qual. Life Res. 11 (2002) 193–205. [10] R. Fitzpatrick, C. Davey, M.J. Buxton, D.R. Jones, Evaluating patient based outcome measures for use in clinical trials, Health Technol. Assess 2 (1998) 1–74. [11] M.P. Carey, R.S. Jorgensen, R.S. Weinstock, R.P. Sprafkin, L.J. Lantinga, C.L.M. Carnrike, et al., Reliability and validity of the appraisal of diabetes scale, J. Behav. Med. 14 (1991) 43– 51. [12] C. Bradley, C. Todd, T. Gorton, E. Symonds, A. Martin, R. Plowright, The development of an individualized questionnaire measure of perceived impact of diabetes on quality of life: the ADDQoL, Qual. Life Res. 8 (1999) 79–91. [13] J.G. Boyer, J.A.L. Earp, The development of an instrument for assessing the quality of life of people with diabetes: diabetes-39, Med. Care 35 (1997) 440–453. [14] J.T. Fitzgerald, W.K. Davis, C.M. Connell, G.E. Hess, M.M. Funnell, R.G. Hiss, Development and validation of the Diabetes Care Profile, Eval. Health Prof. 19 (1996) 208–230. [15] W.H. Polonsky, L. Fisher, J. Earles, J. Dudl, J. Lees, J. Mullan, et al., Assessing psychosocial distress in diabetes: development of the Diabetes Distress Scale, Diabet. Care 28 (2005) 626–631. [16] K. Meadows, N. Steen, E. McColl, M. Eccles, C. Shiels, J. Hewison, et al., The diabetes health profile (DHP): a new instrument for assessing the psychosocial profile of insulin requiring patients: development and psychometric evaluation, Qual. Life Res. 5 (1996) 242–254. [17] G.S. Hammond, T.T. Aoki, Measurement of Health Status in diabetic patients: diabetes impact measurement scales, Diabet. Care 15 (1992) 469–477. [18] W. Shen, J.G. Kotsanos, W.J. Huster, S.D. Mathias, C.M. Andrejasich, D.L. Patrick, Development and validation of the diabetes quality of life clinical trial questionnaire, Med. Care 37 (1999) AS45–AS66 (Lilly Suppl). [19] A.M. Jacobson, I. Barofsky, P. Cleary, L.L. Rand, Reliability and validity of a diabetes quality-of-life measure for the diabetes control and complications trial (DCCT), Diabet. Care 11 (1988) 725–732. [20] U. Bott, I. Mühlhauser, H. Overmann, M. Berger, Validation of a diabetes-specific quality-of-life scale for patients with type 1 diabetes, Diabet. Care 21 (1998) 757–769. [21] A. Araki, H. Ito, Development of elderly diabetes burden scale for elderly patients with diabetes mellitus, Geriatr. Gerontol. Int. 3 (2003) 212–222. [22] M. Peyrot, R.R. Rubin, Validity and reliability of an instrument for assessing health-related quality of life and treatment preferences: the insulin delivery system rating questionnaire, Diabet. Care 28 (2005) 53–58. [23] A. Hirsch, C. Bartholomae, T. Volmer, Dimensions of quality of life in people with non-insulin-dependent diabetes, Qual. Life Res. 9 (2000) 207–218. [24] W.H. Polonsky, B.J. Anderson, P.A. Lohrer, G. Welch, A.M. Jacobson, J.E. Aponte, et al., Assessment of diabetes-related distress, Diabet. Care 18 (1995) 754–760. diabetes research and clinical practice 80 (2008) 171–184 [25] P. Herschbach, G. Duran, S. Waadt, A. Zettler, C. Amm, B. Marten-Mittag, Psychomentric properties of the questionnaire on stress in patients with diabetes: revised (QSD-R), Health Psychol. 16 (1997) 171–174. [26] E. Mannucci, V. Ricca, G. Bardini, C.M. Rotella, Well-being enquiry for diabetics: a new measure of diabetes-related quality of life, Diabet. Nutr. Metab. 9 (1996) 89–102. [27] J.L. Gilden, M.S. Hendryx, S. Clar, C. Casia, S.P. Singh, Diabetes support groups improve health care of older diabetic patients, J. Am. Geriatr. Soc. 40 (1992) 147–150. [28] M.J. Henderson, H. Tindall, Evaluation of consumer satisfaction and quality of life in patients changing to Novopen II, Pract. Diabet. 7 (1990) 206–208. [29] M.E. Wierenga, Life-style modification for weight control to improve diabetes health status, Patient Educ. Couns. 23 (1994) 33–40. [30] Michigan Diabetes Research and Training Centre (MDRTC), Diabetes History (DMH), Available from http:// www.med.umich.edu/mdrtc/textonly/educmast/ survey_intro.htm. Accessed on February 2006. [31] F. Talbot, A. Nouwen, J. Gingras, M. Gosselin, J. Audet, The assessment of diabetes-related cognitive and social factors: the multidimensional diabetes questionnaire, J. Behav. Med. 20 (1997) 291–312. [32] P.A. Grootenhuis, F.J. Snoek, R.J. Heine, L.M. Bouter, Development of a type 2 diabetes symptom checklist: a measure of symptom severity, Diabet. Med. 11 (1994) 253– 261. [33] E. McColl, I.N. Steen, K.A. Meadows, A. Hutchinson, M.P. Eccles, J. Hewison, et al., Developing outcome measures for ambulatory care: an application to asthma and diabetes, Soc. Sci. Med. 41 (1995) 1339–1348. [34] R.M. Anderson, M.M. Funnell, J.T. Fitzgerald, D.G. Marrero, The diabetes empowerment scale: a measure of psychosocial self-efficacy, Diabet. Care 23 (2000) 739–743. [35] N.C. Van Der Ven, K. Weinger, J. Yi, F. Pouwer, H. Ader, H.M. Van Der Ploeg, et al., The confidence in diabetes self-care scale: psychometric properties of a new measure of diabetes-specific self-efficacy in Dutch and US patients with type 1 diabetes, Diabet. Care 26 (2003) 713–718. [36] R.T. Anderson, S.E. Skovlund, D. Marrero, D.W. Levine, K. Meadows, M. Brod, et al., Development and validation of the insulin treatment satisfaction questionnaire, Clin. Therapeut. 26 (2004) 565–578. [37] E.P. Taylor, J.R. Crawford, A.E. Gold, Design and development of a scale measuring fear of complications in type 1 diabetes, Diabet. Metab. Res. Rev. 21 (2005) 264–270. [38] E.D. Mollema, F.J. Snoek, F. Pouwer, R.J. Heine, H.M. Van Der Ploeg, Diabetes fear of injecting and self-testing questionnaire: a psychometric evaluation, Diabet. Care 23 (2000) 765–769. [39] D.J. Cox, A. Irvine, L. Gonder-Frederick, G. Nowacek, J. Butterfield, Fear of hypoglycemia: quantification, validation, and utilization, Diabet. Care 10 (1987) 617–621. [40] R.M. Anderson, J.T. Fitzgerald, M.M. Funnell, L.D. Gruppen, The third version of the diabetes attitude scale, Diabet. Care 21 (1998) 1403–1407. [41] C. Bradley, K.S. Lewis, Measures of psychological well-being and treatment satisfaction developed from the responses of people with tablet-treated Type 2 diabetes, Diabet. Med. 7 (1990) 445–451. [42] F.A. Costa, J.P. Guerreiro, C. Duggan, An audit of diabetes dependent quality of life (ADDQoL) for Portugal: exploring validity and reliability, Pharm. Pract. 4 (2006) 123–128. [43] J.T. Fitzgerald, L.D. Gruppen, R.M. Anderson, M.M. Funnell, S.J. Jacober, G. Grunberger, et al., The influence of treatment modality and ethnicity on attitudes in Type 2 diabetes, Diabet. Care 23 (2000) 313–318. 183 [44] V. Cunningham, M.J. Mohler, C.S. Wendel, R.M. Hoffman, G.H. Murata, J.H. Shah, et al., Reliability and validity of the DCP among hispanic veterans, Eval. Health Prof. 28 (2005) 447–463. [45] K.A. Meadows, C. Abrams, A. Sandbaek, Adaptation of the diabetes health profile (DHP-1) for use with patients with Type 2 diabetes mellitus: psychometric evaluation and cross-cultural comparison, Diabet. Med. 17 (2000) 572– 580. [46] T.C. Li, C.C. Lin, C.S. Liu, C.I. Li, Y.D. Lee, Validation of the Chinese version of the diabetes impact measurement scales amongst people suffering from diabetes, Qual. Life Res. 15 (2006) 1613–1619. [47] A.M. Jacobson, M. De Groot, J.A. Samson, The evaluation of two measures of quality-of-life in patients with Type-I and Type-II diabetes, Diabet. Care 17 (1994) 267–274. [48] A.Y. Cheng, E.Y. Tsui, A.J.G. Hanley, B. Zinman, Developing a quality of life measure for Chinese patients with diabetes, Diabet. Res. Clin. Pract. 46 (1999) 259–267. [49] U. Bott, S. Ebrahim, S. Hirschberger, S.E. Scovlund, Effect of rapid-acting insulin analogue insulin aspart on quality of life and treatment satisfaction in patients with type 1 diabetes, Diabet. Med. 20 (2003) 626–634. [50] F.J. Snoek, F. Pouwer, G.W. Welch, W.H. Polonsky, Diabetesrelated emotional distress in Dutch and U.S. diabetic patients: cross-cultural validity of the problem areas in diabetes scale, Diabet. Care 23 (2000) 1305–1309. [51] G.W. Welch, A.M. Jacobson, W.H. Polonsky, The problem areas in diabetes scale: an evaluation of its clinical utility, Diabet. Care 20 (1997) 760–766. [52] P.M. Trief, W. Grant, K. Elbert, R.S. Weinstock, Family environment, glycemic control, and the psychosocial adaptation of adults with diabetes, Diabet. Care 21 (1998) 241–245. [53] P.M. Trief, C. Aquilino, K. Paradies, R.S. Weinstock, Impact of the work environment on glycemic control and adaptation to diabetes, Diabet. Care 22 (1999) 569–574. [54] C. Bradley, J. Speight, Patient perceptions of diabetes and diabetes therapy: assessing quality of life, Diabet. Metab. Res. Rev. 18 (2002) S64–S69. [55] DAFNE Study Group, Training in flexible, intensive insulin management to enable dietary freedom in people with type 1 diabetes: dose adjustment for normal eating (DAFNE) randomised controlled trial, BMJ 325 (2002) 746– 749. [56] H.L. Wee, C.E. Tan, S.Y. Goh, S.C. Li, Usefulness of the audit of diabetes-dependent quality-of-life (ADDQoL) questionnaire in patients with diabetes in a multiethnic Asian country, Pharmacoeconomics 24 (2006) 673– 682. [57] S.M. Dunn, H. Smartt, L. Beeney, J. Turtle, Measurement of emotional adjustment in diabetic patients: validity and reliability of ATT39, Diabet. Care 9 (1986) 480–489. [58] J.G. Kotsanos, L. Vignati, W. Huster, C. Andrejasich, M.B. Boggs, A.M. Jacobson, et al., Health-related quality-of-life results from multinational clinical trials of insulin lispro, Diabet. Care 20 (1997) 948–958. [59] A.M. Jacobson, Quality of life in patients with diabetes mellitus, Semin. Clin. Neuropsychol. 2 (1997) 82–93. [60] A. Araki, Y. Izumo, J. Inoue, R. Takahashi, K. Takanashi, T. Teshima, et al., Development of elderly diabetes impact scales (EDIS) in elderly patients with diabetes mellitus, Nippon Ronen Igakkai Zasshi (Jpn. J. Geriatr.) 32 (1995) 786– 796. [61] G. Duran, P. Herschbach, S. Waadt, F. Strain, A. Zettler, Assessing daily problems with diabetes: a subjectoriented approach to compliance, Psychol. Rep. 76 (1995) 515–521. 184 diabetes research and clinical practice 80 (2008) 171–184 [62] G.W. Welch, K. Weinger, B. Anderson, W.H. Polonsky, Responsiveness of the problem areas in diabetes (PAID) questionnaire, Diabet. Med. 20 (2003) 69–72. [63] D.A. Cook, T.J. Beckman, Current concepts in validity and reliability for psychometric instruments: theory and application, Am. J. Med. 119 (2006) 166.e7–166.e16. [64] D.L. Streiner, G.R. Norman, Health Measurement Scales: A Practical Guide to their Development and Use, 2nd ed., Oxford University Press, New York, NY, 1995. [65] S. Wild, G. Roglic, A. Green, R. Sicree, H. King, Global prevalence of diabetes: estimates for the year 2000 and projections for 2030, Diabet. Care 27 (2004) 1047–1053.