Evaluation of the EQ-5D-3L and 5L versions in low back pain patients.

BACKGROUND: The EuroQol EQ-5D is one of the most widely researched and applied patient-reported outcome measures worldwide. The original EQ-5D-3L and more recent EQ-5D-5L include three and five response categories respectively. Evidence from healthy and sick populations shows that the additional two response categories improve measurement properties but there has not been a concurrent comparison of the two versions in patients with low back pain (LBP). METHODS: LBP patients taking part in a multicenter randomized controlled trial of lumbar total disc replacement and conservative treatment completed the EQ-5D-3L and 5L in an eight-year follow-up questionnaire. The 3L and 5L were assessed for aspects of data quality including missing data, floor and ceiling effects, response consistency, and based on a priori hypotheses, associations with the Oswestry Disability Index (ODI), Pain-Visual Analogue Scales and Hopkins Symptom Checklist (HSCL-25). RESULTS: At the eight-year follow-up, 151 (87%) patients were available and 146 completed both the 3L and 5L. Levels of missing data were the same for the two versions. Compared to the EQ-5D-5L, the 3L had significantly higher floor (pain discomfort) and ceiling effects (mobility, self-care, pain/discomfort, anxiety/depression). For these patients the EQ-5D-5L described 73 health states compared to 28 for the 3L. Shannon's indices showed the 5L outperformed the 3L in tests of classification efficiency. Correlations with the ODI, Pain-VAS and HSCL-25 were largely as hypothesized, the 5L having slightly higher correlations than the 3L. CONCLUSION: The EQ-5D assesses important aspect of health in LBP patients and the 5L improves upon the 3L in this respect. The EQ-5D-5L is recommended in preference to the 3L version, however, further testing in other back pain populations together with additional measurement properties, including responsiveness to change, is recommended. TRIAL REGISTRATION: retrospectively registered: https://clinicaltrials.gov/ct2/show/NCT01704677 .

Measurement properties for the revised patient-reported OsteoArthritis Quality Indicator questionnaire.

OBJECTIVES: To assess validity, reliability, responsiveness and interpretability of the revised OsteoArthritis Quality Indicator (OA-QI) questionnaire version 2 (v2) assessing patient-reported quality of osteoarthritis care. METHODS: The OA-QI v2 (16 items, score range 0-100 (100 = best score)) was included in a longitudinal cohort study. Attendees of a 4.5 h osteoarthritis patient education programme at Diakonhjemmet Hospital, Norway, completed the OA-QI at four time points: 2 weeks before, immediately before, immediately after, and 3 months after the programme. Test-retest reliability and measurement error over a 2-week time period were assessed in those that had not seen health professionals in the interim. Construct validity and responsiveness were assessed with predefined hypotheses. Floor and ceiling effects, smallest detectable change (SDC95%) and minimal important change (MIC) were assessed to evaluate interpretability. RESULTS: The intraclass correlation coefficient for all 16 items was 0.89. For single items the test-retest kappa estimates ranged 0.38-0.85 and percent agreement 69-92%. Construct validity was acceptable with all six predefined hypotheses confirmed. Responsiveness was acceptable with 33 of 48 and three of four predefined hypotheses confirmed for single items and all items, respectively. There were no floor or ceiling effects. The SDC95% was 29.1 and 3.0 at the individual and group levels, respectively. MIC was 20.4. CONCLUSIONS: The OA-QI v2 had higher reliability estimates compared to v1, showed acceptable validity, and is the recommended version for future use. The results of responsiveness testing further support the use of the OA-QI v2 as an outcome measure in studies aiming to improve osteoarthritis care.

Measurement properties and normative data for the Norwegian SF-36: results from a general population survey.

BACKGROUND: The interpretation of the SF-36 in Norwegian populations largely uses normative data from 1996. This study presents data for the general population from 2002-2003 which has been used for comparative purposes but has not been assessed for measurement properties. METHODS: As part of the Norwegian Level of Living Survey 2002-2003, a postal survey was conducted comprising 9,164 members of the general population aged 16 years and over representative for Norway who received the Norwegian SF-36 version 1.2. The SF-36 was assessed against widely applied criteria including data completeness and assumptions relating to the construction and scoring of multi-item scales. Normative data are given for the eight SF-36 scales and the two summary scales (PCS, MCS) for eight age groups and gender. RESULTS: There were 5,396 (58.9%) respondents. Item levels of missing data ranged from 0.6 to 3.0% with scale scores computable for 97.5 to 99.8% of respondents. All item-total correlations were above 0.4 and were of a similar level with the exceptions of the easiest and most difficult physical function items and two general health items. Cronbach's alpha exceeded 0.8 for all scales. Under 5% of respondents scored at the floor for five scales. Role-physical had the highest floor effect (14.6%) and together with role-emotional had the highest ceiling effects (66.3-76.8%). With three exceptions for the eight age groups, females had lower scores than males across the eight health scales. The two youngest age groups (<30 years) had the highest scores for physical aspects of health; physical function, role-physical, bodily pain and general health. The age groups 40-49 and 60-69 years had the highest scores for role-emotional and mental health respectively. CONCLUSIONS: This SF-36 data meet necessary criteria for applications of normative data. The data is more recent, has more respondents including older people than the original Norwegian normative data from 1996, and can help the interpretation of SF-36 scores in applications that include clinical and health services research.

Rasch analysis of the Norwegian version of the Arthritis Self-Efficacy Scale (ASES).

OBJECTIVES: To apply the Rasch model to the Norwegian version of the Arthritis Self-Efficacy Scale (ASES). METHOD: The ASES was included in a self-administered questionnaire sent to 224 patients attending one of three rehabilitation centres for rheumatic diseases in Norway in 2009. The fit of the ASES to the Rasch model was assessed together with hypothesized associations with the Modified Health Assessment Questionnaire (MHAQ), the 36-item Short Form Health Survey (SF-36), the numerical rating scale (NRS) for pain, and NRS fatigue. RESULTS: A total of 145 (64.7%) patients responded to the questionnaire. The two scales of other symptoms and pain showed good fit to the Rasch model with no evidence for differential item functioning (DIF) according to eight sociodemographic and disease-related variables. The Person Separation Index (PSI), which is equivalent to Cronbach's alpha, ranged from 0.74 to 0.78. Correlations with scores for other instruments were as hypothesized: ASES pain had the highest correlations with SF-36 pain and NRS pain and ASES other symptoms had the highest correlations with other aspects of the SF-36 and NRS fatigue. CONCLUSIONS: The ASES had good fit to the Rasch model and correlations with other instrument scores that followed hypotheses, lending further support to the application of the instrument in patients with rheumatic diseases.

Patients with well-established ankylosing spondylitis show limited deterioration in a ten-year prospective cohort study.

This study aims to describe the deterioration of well-established ankylosing spondylitis (AS) over a 10-year period. Patients with well-established AS under secondary care review were assessed at baseline and re-assessed after a 10-year period. Data on patient characteristics, spinal mobility and self-reported health status (AS Quality of Life, Bath AS Disease Activity Index, EuroQoL, Revised Leeds Disability Questionnaire and Short Form-12) were compared between the two time points. One hundred fifty-nine patients took part in the initial assessment. After 10 years, 69/117 patients still under secondary care follow-up attended a second assessment. At the second assessment, this cohort was predominately male (92.8 %), mean age of 48.7 years (SD 9.7) and mean disease and symptom duration of 14.9 (SD 9.0) and 21.9 years (SD 13.3), respectively. Amongst the physical and patient-reported measures, only tragus to wall (p < 0.001), cervical rotation (p = 0.001) and disability (p = 0.02) had significantly deteriorated over time. The percentage of patients who had withdrawn from the workforce before normal retirement age increased from 37 to 53 % over the 10-year period. Many of the measures used in the assessment of AS do not generally deteriorate over time in those with well-established disease. This suggests that deterioration may plateau in established disease. However, AS has a long-term social impact with levels of employment in this cohort 30 % lower than the general population. Interventions directed at preventing deterioration may be more effective earlier in their disease course, before this plateau is reached.

Evaluation of a modified arthritis self-efficacy scale for an ankylosing spondylitis UK population.

OBJECTIVES: To evaluate an Ankylosing Spondylitis-specific Arthritis Self-Efficacy Scale (ASES-AS) United Kingdom (UK) secondary care population. METHODS: The ASES-AS is based on the 8-item ASES with minor alterations in phraseology. Patients from ten secondary care rheumatology centres across England were asked to complete a postal questionnaire concerning sociodemographic and clinical characteristics: Bath AS Functional Index (BASFI), Bath AS Disease Activity Index (BASDAI), numerical pain rating scale (NRS), Hospital Anxiety and Depression Scale (HADS), Short Form 36 (SF-36), Evaluation of AS Quality of Life questionnaire (EASi-QoL) and ASES-AS. Respondents received repeat questionnaires at 2 weeks and 6 months including health transition questions assessing change in AS-specific and general health. The ASES-AS was assessed for data quality, reliability, validity, and responsiveness. RESULTS: Response rate was 64% (n=612), 72% (n=438) were male, mean age 50.8yrs (SD 12.2 yrs), mean disease duration 17.3 yrs (SD 11.7 yrs) and mean symptom duration 22.4 yrs (SD 12.4 yrs). Missing data for each item/total score range was 0.7%-3.1%. Item-total correlations range was 0.66 to 0.83. Cronbach's alpha was 0.93 and test-retest reliability (intraclass correlation coefficient) 0.77. A priori hypothesised associations between ASAS-AS and disease status measures were supported. Social variables potentially related to self-efficacy demonstrated evidence of convergent validity (employment p<0.001, educational level p<0.005). A Modified Standard Response Mean (MSRM) of 0.44 and 0.26 in AS-specific and general health respectively at 6 months indicates moderate responsiveness. CONCLUSIONS: ASES-AS has good evidence supporting its application as an AS-specific self-efficacy measure in research including clinical trials at a group level.

Patient experiences of rehabilitation in rheumatology: association with process and outcome.

OBJECTIVE: To evaluate health-care experiences of patients following inpatient rheumatology rehabilitation and to assess the association between these experiences and aspects of health-care delivery and patient characteristics. METHODS: Data were collected from 435 patients with a rehabilitation stay of >or= 1 week at 12 institutions in Norway in 2006. At discharge, patients completed the Rehabilitation Patient Experiences Questionnaire (Re-PEQ), which includes four important aspects of patient experiences. Multiple regression analysis was used to assess associations between Re-PEQ scores, health-care process, health and sociodemographic variables. RESULTS: A total of 412 (94.7%) patients completed the Re-PEQ; scores ranged from 69 (social environment) to 83 (care/organization) on the 0-100 scale, where 100 represents the best possible experience. The social environment scale had the largest component of variation explained by the independent variables, which included number of doctor visits, amount of group education, and individual exercise (p < 0.01). The type of institution, number of doctor visits, mental health, and gender also explained significant components of variation in the other Re-PEQ scale scores. CONCLUSION: Patients reported good experiences with rheumatology rehabilitation. Areas where poorer experiences emerged can help target areas for future initiatives aimed at improving the quality of care. Health and sociodemographic variables should be controlled for in studies of patient experiences.

The OutPatient Experiences Questionnaire (OPEQ): data quality, reliability, and validity in patients attending 52 Norwegian hospitals.

OBJECTIVE: To describe the development and evaluation of the OutPatient Experiences Questionnaire (OPEQ) for somatic outpatients. DESIGN: Literature review, patient interviews, pretesting of questionnaire items, and a cross sectional survey. SETTING: Postal survey of adult outpatient clinics at 52 hospitals in all five regions of Norway during 2003 and 2004. SUBJECTS: 35,719 patients who had attended an outpatient clinic within the previous 3 weeks. RESULTS: 19,266 patients (53.9%) responded to the questionnaire. Low levels of missing data suggest that the questionnaire is acceptable to patients. Factor analysis of items applicable to all patients produced three factors: clinic access (two items), communication (six items), and organisation (four items). The remaining items contributed to the hypothesised scales of hospital standards (three items), information (six items), and pre-visit communication (three items). With the exception of the pre-visit communication scale, the levels of Cronbach's alpha were >0.7. With the exception of the hospital standards scale, all produced test-retest correlations that exceeded 0.7. Most of the results of validity testing were as hypothesised. Correlations between the OPEQ scores ranged from 0.30 (clinic access and hospital standards) to 0.73 (communication and information). As hypothesised, scores were significantly related to patient responses to questions about overall satisfaction, general health and age. CONCLUSIONS: The OPEQ is a self-administered questionnaire that includes the most important aspects of patient experience from an outpatient perspective. It has good evidence for internal consistency, test-retest reliability, and validity.

Quality of life in older people: a structured review of generic self-assessed health instruments.

OBJECTIVE: To review evidence relating to the measurement properties of multi-item generic patient or self-assessed measures of health in older people. METHODS: Systematic literature searches to identify instruments. Pre-defined criteria relating to reliability, validity and responsiveness. RESULTS: 122 articles relating to 15 instruments met the inclusion criteria. The most extensive evidence was found for the SF-36, COOP Charts, EQ-5D, Nottingham Health Profile (NHP) and SIP. Four instruments have evidence of both internal consistency and test-retest reliability--NHP, SF-12, SF-20, SF-36. Four instruments lack evidence of reliability--HSQ-12, IHQL, QWB, SQL. Most instruments were assessed for validity through comparisons with other instruments, global judgements of health, or clinical and socio-demographic variables. Five instruments lack evidence of responsiveness--GQL, HSQ-12, IHQL, QLI and QWB. CONCLUSION: There is good evidence for reliability, validity and responsiveness for the SF-36, EQ-5D and NHP. There is more limited evidence for the COOP, SF-12 and SIP. The SF-36 is recommended where a detailed and broad ranging assessment of health is required, particularly in community dwelling older people with limited morbidity. The EQ-5D is recommended where a more succinct assessment is required, particularly where a substantial change in health is expected. Instrument content should be assessed for relevance before application. The concurrent evaluation of generic instruments in older people is recommended.

Patient-assessed health in ankylosing spondylitis: a structured review.

OBJECTIVE: To review evidence relating to the measurement properties for all disease-specific, multi-item, patient-assessed health instruments in patients with ankylosing spondylitis (AS). METHODS: Systematic literature searches were made to identify instruments, using predefined criteria relating to reliability, validity, responsiveness and precision. RESULTS: Twelve AS-specific and three arthritis-specific instruments met the inclusion criteria. Three AS-specific instruments that measure health-related quality of life (HRQL) were reviewed. The Bath Ankylosing Spondylitis Disease Activity Index (BASDAI), the Bath Ankylosing Spondylitis Functional Index (BASFI) and the Dougados Functional Index (DFI) had the greatest amount of evidence for reliability, validity and responsiveness across a range of settings. Four instruments lacked evidence for test-retest or internal consistency reliability. Most were assessed for validity through comparisons with other instruments, global judgements of health, mobility or clinical and sociodemographic variables. Most were assessed for responsiveness through mean score changes. Three instruments lacked evidence of responsiveness. CONCLUSION: This review provides a contribution to AS assessment. AS-specific multi-item measures specific to the assessment of pain, stiffness, fatigue and global health were not identified; where assessed, these domains were largely measured with single-item visual analogue scales. Single items may provide a limited reflection of these important domains. The BASFI and DFI remain the instruments of choice for functional assessment. HRQL is recommended as a core assessment domain. Further concurrent evaluation is recommended.

Patient-assessed health instruments for the knee: a structured review.

OBJECTIVES: To identify patient-assessed health instruments specific to the knee and review evidence for reliability, validity and responsiveness. METHODS: Instruments were identified through systematic searches of the literature. Information relating to instrument content, patient population, reliability, validity and responsiveness was extracted from published papers. RESULTS: The 16 instruments that met the inclusion criteria varied in length from 4 to 42 items. The majority form a single index; six produce a profile of scores. Eight have been evaluated in patients with a variety of knee problems. All instruments have satisfactory internal or test-retest reliability. However, there is limited empirical support for the health domains of six instruments. Patients informed the development of items within just five instruments. Few authors gave explicit consideration to the size of expected relationships in tests of construct validity. Eleven instruments have evidence for responsiveness to changes in health. The minimally important difference was not determined for any of the instruments. CONCLUSIONS: In the absence of comparative evidence, the large number of patient-assessed instruments for knee problems makes instrument selection difficult. The Knee Injury and Osteoarthritis Outcome Score (KOOS), Knee Pain Scale and Oxford Knee Score have good evidence for reliability, content validity and construct validity. The KOOS and Oxford instruments also have evidence for responsiveness. The instruments have not been evaluated for all knee problems, and instrument appropriateness, including content relevance, must be assessed before application. The comparative evaluation of instruments is recommended.

Spinal mobility in ankylosing spondylitis: reliability, validity and responsiveness.

OBJECTIVE: To evaluate the measurement properties of an evidence-based selection of measures of spinal mobility in patients with ankylosing spondylitis (AS). METHODS: Measurements of spinal mobility were taken by trained observers within a UK rheumatology centre. Inter-observer reliability was assessed. Intra-observer reliability was assessed in patients reporting no change in AS-specific health at 2 weeks. Validity was assessed and scores were correlated with responses to health transition questions. Responsiveness was evaluated for patients reporting change in health at 6 months. RESULTS: Reliability estimates support the use of all measures in individual evaluation (intraclass correlation>0.90). Correlations between measures of spinal mobility were in the hypothesized direction; the largest was between the modified Schober index (15 cm) (MSI) and the other measures. As hypothesized, small to moderate levels of correlation were found between mobility measures and patient-assessed health status. There was no significant linear relationship between mobility measures and self-reported health transition. Fingertip-to-floor distance following trunk forward flexion (FFD) was the most responsive mobility measure but was not as responsive as two AS-specific patient-assessed instruments, the Ankylosing Spondylitis Quality of Life Questionnaire (ASQoL) and the Bath Ankylosing Spondylitis Disease Activity Index (BASDAI). The MSI and cervical rotation (Crot) also had evidence of responsiveness. Low levels of responsiveness were found for the remaining measures. CONCLUSION: All mobility measures had adequate levels of reliability and validity. The MSI had a strong relationship with all mobility measures, and the FFD and Crot were the most responsive to self-perceived changes in health at 6 months. The MSI, FFD and Crot are recommended for clinical practice and research.

Generic measures of health-related quality of life in ankylosing spondylitis: reliability, validity and responsiveness.

OBJECTIVE: To assess the acceptability and measurement properties of two generic measures of health-related quality of life (HRQL): the EuroQol and the Short Form 12-item Health Survey Questionnaire (SF-12) in ankylosing spondylitis (AS). METHODS: Instruments were administered by means of a self-completed questionnaire to AS patients recruited from across the United Kingdom. Instruments were assessed for data quality and scaling assumptions. Test-retest reliability was assessed in those patients reporting no change in general health at 2 weeks. The convergent validity of both instruments was assessed and scores were correlated with responses to health transition questions. Responsiveness was assessed for patients reporting change in health at 6 months. RESULTS: The instruments had high completion rates. Although slightly skewed towards better levels of health, scores covered the available range for both sections of the EuroQol [EQ-5D and EQ-visual analogue scale (EQ-VAS)]. Score distributions approximated normality for the SF-12. Test-retest reliability estimates support the use of both instruments in group evaluation and the SF-12 Physical Component Summary score (PCS) in individual evaluation (>0.90). Correlations between instruments were in the hypothesized direction and were of a moderate level. The EQ-VAS had the strongest linear relationship, with responses to both specific and general health transition questions (P<0.01). The EQ-VAS and SF-12 PCS were the most responsive instruments. The EQ-5D was the least responsive instrument. CONCLUSION: The instruments have undergone a comprehensive comparative evaluation to assess the measurement properties required for patient-assessed measures of health outcome in AS. Adequate levels of acceptability, reliability and validity were found for both instruments. Although evidence supporting instrument responsiveness was strong for the EQ-VAS and SF-12 PCS, it was very weak for the EQ-5D and SF-12 Mental Component Summary Scale (MCS). The EQ-VAS and SF-12 PCS can both be recommended for use in group evaluation, and the SF-12 PCS is recommended in routine practice or research. However, the lower reliability of the SF-12 MCS and the limited ability of both the EQ-5D and SF-12 MCS to detect change in health may restrict these roles.

Child/parent-assessed population health outcome measures: a structured review.

AIMS: To identify generic measures of health-related quality of life (HRQL) for children and adolescents developed for use within general populations. Instruments are evaluated on the basis of evidence relating to their reliability and validity. METHODS: Systematic literature searches were used to identify instruments, which were then assessed against predefined criteria. Information relating to instrument content, population, reliability and validity was extracted from published papers. RESULTS: Sixteen instruments were identified that had been evaluated among a general population of children or adolescents. Four instruments had reported data on both internal consistency and test-retest reliability. All except two instruments had undergone some degree of construct validation. CONCLUSIONS: The evidence suggests that the Child Health Questionnaire (CHQ) has been the most extensively evaluated for younger populations but is available as a parent-completed measure only. The new version of the Child Health and Illness Profile (CHIP-CE) is particularly promising and has parallel child- and parent-completed versions for young ages. The weight of evidence suggests that versions of these two instruments are suitable for older children. The Warwick Child Health and Morbidity Profile could be used where information on morbidity and health service contacts is required. Once basic psychometric criteria are fulfilled, instruments should be chosen by assessing their content and design in the light of the prospective application.

Patient-assessed health outcome measures for diabetes: a structured review.

AIMS: To identify available disease-specific measures of health-related quality of life (HRQL) for diabetes and to review evidence for the reliability, validity and responsiveness of instruments. METHODS: Systematic searches were used to identify instruments. Instruments were assessed against predefined inclusion and exclusion criteria. Letters were sent to authors requesting details of further instrument evaluation. Information relating to instrument content, patients, reliability, validity and responsiveness to change was extracted from published papers. RESULTS: The search produced 252 references. Nine instruments met the inclusion criteria: Appraisal of Diabetes Scale (ADS), Audit of Diabetes-Dependent Quality of Life (ADDQoL), Diabetes Health Profile (DHP-1, DHP-18), Diabetes Impact Measurement Scales (DIMS), Diabetes Quality of Life Measure (DQOL), Diabetes-Specific Quality of Life Scale (DSQOLS), Questionnaire on Stress in Diabetic Patients-Revised (QSD-R), Diabetes-39 (D-39) and Well-being Enquiry for Diabetics (WED). The shortest instrument (ADS) has seven items and the longest (WED) has 50 items. The ADS and ADDQoL are single-index measures. The seven multidimensional instruments have dimensions covering psychological well-being and social functioning but vary in the remainder of their content. The DHP-1 and DSQOLS are specific to Type 1 diabetes patients. The DHP-18 is specific to Type 2 diabetes patients. The DIMS and DQOL have weaker evidence for reliability and internal construct validity. Patients contributed to the content of the ADDQoL, DHP-1/18, DQOL, DSQOLS, D-39, QSD-R and WED. The authors of the ADDQoL, DHP-1/18, DQOL, DSQOLS gave explicit consideration to content validity. The construct validity of instruments was assessed through comparisons with instruments measuring related constructs and clinical and sociodemographic variables. None of the instruments has been formally assessed for responsiveness to changes in health. CONCLUSIONS: Five of the diabetes-specific instruments have good evidence for reliability and internal and external construct validity: the ADDQoL, DHP-1/18, DSQOLS, D-39 and QSD-R. Instrument content should be assessed for relevance before application. The instruments should be evaluated concurrently for validity and responsiveness to important changes in health.

Problems in using health survey questionnaires in older patients with physical disabilities. The reliability and validity of the SF-36 and the effect of cognitive impairment.

Reliability and validity of the SF-36 Health Survey Questionnaire was assessed in older rehabilitation patients, comparing cognitively impaired with cognitively normal subjects. The SF-36 was administered by face-to-face interview to 314 patients (58-93 years) in the day hospital and rehabilitation wards of a department of medicine for the elderly. Reliability was measured using Cronbach's alpha (for internal consistency) on the main sample and intraclass correlation coefficients on a test-retest sample; correlations with functional independence measure (FIM) were examined to assess validity. In 203 cognitively normal patients (Mini-Mental State Examination > or =24), Cronbach's alpha scores on the eight dimensions of the SF-36 ranged from 0.545 (social function) to 0.933 (bodily pain). The range for the 111 cognitively impaired patients was 0.413-0.861. Cronbach's alpha values were significantly higher (i.e. reliability was better) in the cognitively normal group for bodily pain (P = 0.003), mental health (P = 0.03) and role emotional (P = 0.04). In test-retest studies on a further 67 patients, an intraclass correlation coefficient of 0.7 was attained for five out of eight dimensions in cognitively normal patients, and four out of eight dimensions in the cognitively impaired. Only the physical function dimension in the cognitively normal group attained the criterion level (r > 0.4) for construct validity when correlated with the FIM. In this group of older physically disabled patients, levels of reliability and validity previously reported for the SF-36 in younger subjects were not attained, even on face-to-face testing. Patients with coexistent cognitive impairment performed worse than those who were cognitively normal.

Problems in using health survey questionnaires in older patients with physical disabilities. Can proxies be used to complete the SF-36?

BACKGROUND: The SF-36 Health Survey questionnaire has been proposed as a generic measure of health outcome. However, poor rates of return and high levels of missing data have been found in elderly subjects and, even with face-to-face interview, reliability and validity may still be disappointing, particularly in cognitively impaired patients. These patients may be the very patients whose quality of life is most affected by their illness and their exclusion will lead to biased evaluation of health status. A possible alternative to total exclusion is the use of a proxy to answer on the patient's behalf, but few studies of older people have systematically studied patient-proxy agreement. OBJECTIVE: To compare the agreement between patients, lay and professional proxies when assessing the health status of patients with the SF-36. METHODS: The SF-36 was administered by interview to 164 cognitively normal, elderly patients (Mini-mental State Examination 24 or more) referred for physical rehabilitation. The SF-36 was also completed by a patient-designated lay proxy (by post) and a professional proxy. Agreement between proxies and patients was measured by intraclass correlation coefficients (ICCs), and a bias index. RESULTS: Professional proxies were better able to predict the patients' responses than were the lay proxies. Criterion levels of agreement (ICC 0.4 or over) were attained for four of the eight dimensions of the SF-36 by professional proxies, but for only one dimension by lay proxies. In professional proxies, the magnitude of the bias was absent or slight (<0.2) for six of the eight dimensions of the SF-36 with a small (0.2-0.49) negative bias for the other two. Lay proxies showed a negative bias (i.e. they reported poorer function than did the patients themselves) for seven of the eight dimensions of the SF-36 (small in two and moderate (0.5-0.79) in five). CONCLUSIONS: For group comparisons using the SF-36, professional proxies might be considered when patients cannot answer reliably for themselves. However, in the present study, lay proxy performance on a postal questionnaire showed a strong tendency to negative bias. Further research is required to define the limitations and potentials of proxy completion of health status questionnaires.

The UK version of the Seattle Angina Questionnaire (SAQ-UK): reliability, validity and responsiveness.

The study assesses the reliability, validity and responsiveness of the UK version of the Seattle Angina Questionnaire (SAQ-UK). The instrument was anglicised and administered by self-completed postal questionnaire to 959 patients recruited from general practices in the North East of England. A total of 655 (68.3%) patients returned a completed questionnaire. Principal component analysis produced three important dimensions: physical limitations, anginal frequency and perception and treatment satisfaction. Four items that performed poorly were removed from the instrument. The removal of items greatly increased the number of computable scale scores. Estimates of internal reliability for the three dimensions ranged from 0.83-0.92. Estimates of test-retest reliability were above accepted standards. The correlation between the SAQ-UK scores and the SF-12, EuroQol and health transition was evidence for the validity of the instrument. The SAQ-UK produced responsiveness statistics that were comparable to the parent instrument and was more responsive to improvements in health than the generic instruments. The SAQ-UK is recommended as a measure of health outcome for the evaluation of angina management.

Responsiveness of generic and specific measures of health outcome in low back pain.

STUDY DESIGN: A longitudinal study using patient questionnaires was performed. OBJECTIVE: To compare the discriminatory power and responsiveness of the Aberdeen Back Pain Scale (ABPS), the Roland Disability Questionnaire (RDQ), and the EuroQol in patients with low back pain. SUMMARY OF BACKGROUND DATA: A number instruments specific to low back pain have not been compared for measurement properties. The EuroQol is a widely used generic instrument that has not been compared with specific instruments in patients with back pain. METHODS: A questionnaire incorporating the Aberdeen Back Pain Scale, the Roland Disability Questionnaire, and the EuroQol was completed by patients taking part in a clinical trial of exercise treatments for back pain. Patients completed follow-up questionnaires at 6 weeks, 6 months, and 1 year. The discriminatory power of these instruments was assessed against variables relating to activity limitations, medication, and comorbidity. Responsiveness was assessed using standardized response means. RESULTS: The questionnaire was completed by 187 patients taking part in the clinical trial. The Aberdeen instrument was found to be the most powerful at discriminating between different groups of patients on variables relating to activity limitations, medication, and comorbidity. The specific instruments demonstrated good levels of responsiveness, with the Aberdeen instrument producing the largest standardized response means. The Aberdeen instrument was more responsive to the smaller changes experienced by the control group, but was less powerful than the Roland at measuring differences in the levels of change between the two groups of patients at two of the three follow-up assessments in the trial. The EuroQol demonstrated a moderate level of responsiveness. CONCLUSIONS: The two specific instruments are capable of greater levels of discrimination between groups of patients, and are more responsive over time than the generic EuroQol. The Aberdeen instrument performed most satisfactorily in relation to these criteria, but the Roland instrument was more sensitive to differences between the two groups in the clinical trial. The measurement properties of these two instruments reflect their origin: The Aberdeen instrument is based on clinical questions, whereas the Roland instrument is based on the generic Sickness Impact Profile. Instrument content should be carefullyconsidered when selecting instruments for applications, including clinical trials.