Discriminative ability of the short-form 36 health survey: a tale of two versions.

BACKGROUND AND OBJECTIVES: Previous studies suggested that the Chinese version of the Short-Form 36 Health Survey (SF-36) had smaller variability in the physical functioning (PF) and physical component summary (PCS) scores than the English SF-36. This may translate into differences in discriminative ability and sample size requirement. MATERIALS AND METHODS: Data were drawn from a community-based survey and a randomised crossover study of Singaporeans bilingual in the Chinese and English languages. The abilities of the two PF and PCS versions in discriminating subjects who reported chronic illness and acute disease symptoms versus those who did not were compared. RESULTS: In all four comparisons (i.e. two health criteria in two studies) the Chinese version of PF showed a larger effect size than the English version. In three out of four comparisons the Chinese version of PCS showed a larger effect size than the English version. CONCLUSIONS: The Chinese version appeared more efficient in detecting a statistically significant difference between groups. Other factors being the same, the Chinese SF-36 may require a smaller sample size than the English SF-36 for the studies of physical aspects of health-related quality of life.

Modelling variability of quality of life scores: a study of questionnaire version and bilingualism.

BACKGROUND AND OBJECTIVES: Modelling variability of quality of life scores can not only improve our understanding of the characteristics of the measurement tools, but also shed light on sample size requirements. Although the English and Chinese versions of the Short Form 36 Health Survey (SF-36) are commonly considered equivalent, they have not been compared in terms of variability. Furthermore, bilingual and monolingual persons may differ in cognition and responses to questionnaires. METHODS: In a community-based survey of quality of life in Singapore, a society where both English and Chinese are widely used and bilingualism is prevalent, 2590 respondents answered either version of the SF-36. We studied the impact of questionnaire version and bilingualism on the variability of SF-36 scores by regression modelling, with adjustment for covariates. RESULTS: The Chinese version had smaller variances in the physical functioning (PF) and the physical component summary scores than the English version. The variance ratios (VRs) were respectively 0.32 and 0.60 (each p < 0.01), controlling for covariates. Bilingualism was not associated with variability in SF-36 scores except PF (VR = 0.78; p < 0.05). CONCLUSIONS: As a result of a smaller variance, using the Chinese version of SF-36 among bilingual Chinese people may require a smaller sample size than using the English version.

Measuring health-related quality of life in Singapore: normal values for the English and Chinese SF-36 Health Survey.

OBJECTIVE: To determine norms for assessing Health-related Quality of Life (HRQOL) in Singapore using the Short Form 36 Health Survey (SF-36). MATERIALS AND METHODS: Mean SF-36 scores were calculated for 24 population subgroups (categorised by age, gender, ethnicity and questionnaire language) and for subjects with self-reported co-morbid conditions using data from a community-based survey in Singapore. RESULTS: The English and Chinese SF-36 was completed by 4122 and 1381 subjects, respectively, 58% (n = 3188) of whom had self-reported co-morbid conditions. SF-36 scores varied in subgroups differing in age, gender and ethnicity. In general, subjects with self-reported co-morbid conditions had lower SF-36 scores than those without these conditions, the magnitude of which exceeded 20 points in several instances. A method for calculation of SF-36 scores adjusted for age, gender, ethnicity and questionnaire language is described. CONCLUSION: We present norms for English and Chinese SF-36 versions in Singapore and describe potential uses for these data in assessing HRQOL in Singapore.

A Chinese version of the Rheumatology Attitudes Index is a valid and reliable measure of learned helplessness in patients with SLE.

Despite the prognostic importance of learned helplessness (LH) in rheumatic diseases, there are no validated measures of LH in Chinese or other Asian languages. We therefore assessed the validity of a Chinese translation of the Rheumatology Attitudes Index (CRAI; a widely used measure of LH) and its Helplessness (CHS) and Internality (CIS) subscales in patients with SLE. Chinese-speaking SLE patients (n = 69) completed identical, self-administered questionnaires containing the CRAI and assessing demographic/socio-economic variables twice within 2 weeks. SLE activity, damage and quality of life were assessed using the BILAG, SLICC/ACR Damage Index and SF-36 respectively. Scale psychometric properties were assessed through Cronbach's alpha, intra-class correlations, quantifying test-retest differences, factor analysis and known-groups construct validity. Internal consistency and reliability were acceptable, with Cronbach's alpha for the CHS, CIS and CRAI being 0.70, 0.69 and 0.74, respectively. Mean differences in test-retest scores spanned 1.6-2.4% of possible scale ranges and intra class correlations ranged from 0.72 to 0.88. Factor analysis identified two major factors corresponding to the CHS and CIS subscales of the CRAI. Eight of 10 a priori hypotheses relating the CRAI and CHS to demographic, disease and quality of life variables were confirmed, supporting the construct validity of these scales. The CRAI and its helplessness subscale are valid and reliable measures of learned helplessness in Chinese-speaking SLE patients.

A community-based study of scaling assumptions and construct validity of the English (UK) and Chinese (HK) SF-36 in Singapore.

Scaling assumptions and validity of the English (UK) and Chinese (HK) short form 36 health survey (SF-36) were assessed in a community-based survey of 5,503 Chinese, Malays and Indians in Singapore using the international quality of life assessment project approach of item and scale level validation. Missing data for SF-36 items and scales occurred in less than 1.0% of subjects. Item level validation of both versions generally supported assumptions underlying Likert scoring and hypothesised item-order clustering. Item level factor analysis supported the eight-scale structure of the SF-36. In scale level validation, SF-36 scale scores showed wide variability and acceptable internal-consistency reliability (Cronbach's alpha > 0.70 for six English and seven Chinese scales), conformed to hypothesised patterns and generally varied according to hypotheses in subjects known to differ in quality of life. Scale level factor analysis of both versions yielded very similar patterns of factor correlation, comparable to that found in Japan, but differing from that seen in Western populations. Taken together, these results support the validity of the English (UK) and Chinese (HK) SF-36 versions in the multi-ethnic Asian socio-cultural context of Singapore.

A comparative study of the clinical manifestations of systemic lupus erythematosus in Caucasians in Rochester, Minnesota, and Chinese in Singapore, from 1980 to 1992.

OBJECTIVE: To examine the relationship between ethnicity and major organ involvement at and after diagnosis in community-based cohorts of Caucasian and Chinese systemic lupus erythematosus (SLE) patients resident in Rochester, Minnesota, and Singapore, respectively. METHODS: Clinical manifestations at and after diagnosis were compared in Caucasian and Chinese SLE patients. The association between ethnicity and disease manifestations at and after diagnosis was determined using logistic regression and Cox proportional hazards models, respectively, adjusting for the influence of demographic, socioeconomic, disease-related, and therapy-related factors. RESULTS: At diagnosis, Caucasian SLE patients were 3 times more likely than Chinese SLE patients to have serositis (odds ratio [OR] 3.11, 95% confidence interval [CI] 1.01-9.71), nearly 7 times more likely to have a hematologic disorder (OR 6.95, 95% CI 2.20-21.97), and far less likely to have a malar rash (OR 0.19, 95% CI 0.07-0.54) or positive antinuclear antibodies (OR 0.11, 95% CI 0.03-0.52). Ethnicity was not associated with the prevalence of proteinuria or central nervous system (CSN) and other major organ involvement at diagnosis. After diagnosis, there was a trend toward less development of proteinuria and other major organ involvement in Caucasians (relative risk [RR] 0.47, 95% CI 0.19-1.15, and RR 0.22, 95% CI 0.05-1.04, respectively). CONCLUSION: Chinese SLE patients are far less likely to have serositis or a hematologic disorder at diagnosis and may be more likely to develop proteinuria or CNS or other major organ involvement over the course of the disease, compared with Caucasian SLE patients. This may contribute to the increased mortality seen in Chinese SLE patients.

Accuracy of medical record scoring of the SLICC/ACR damage index for systemic lupus erythematosus.

OBJECTIVE: To study the accuracy of medical record scoring of the SLICC/ACR Damage Index (DI) for SLE. METHODS: Medical record DI (MDI) scoring of 60 SLE patients by one physician was compared with prospective direct DI (DDI) scoring a median of 2 y earlier. Agreement between DDI and MDI total and organ system scores was compared using Cohen's kappa, the degree of misclassification by MDI and the significance of differences between DDI and MDI scores. RESULTS: The mean duration of SLE at DDI scoring was 5.2 y. The median/mean (SD) DDI and MDI total scores were 0/0.78 (1.38) and 0/0. 85 (1.44). Damage was present on DDI and MDI scoring in 36.7% (22/60) and 38.3% (23/60) of patients. Qualitative MDI and DDI total scores (damage present/absent) showed good agreement (kappa=0.61, 95% CI=0.40-0.82, 18.3% misclassification). Quantitative MDI and DDI total scores (degree of damage) showed moderate agreement (kappa=0. 47, 95% CI=0.28-0.66, 36.7% misclassification). 9 of 12 MDI organ system scores misclassified <10% of subjects. There was no statistically significant difference between DDI and MDI total or organ system scores. CONCLUSION: MDI total scoring is a good qualitative and moderate quantitative reflection of SLE related damage. MDI organ system scores are more accurate than MDI total scores.

A prospective study of factors affecting quality of life in systemic lupus erythematosus.

OBJECTIVE: To prospectively identify factors influencing quality of life (QOL) over 6 months in patients with systemic lupus erythematosus (SLE). METHODS: Ninety ethnically diverse patients with SLE completed questionnaires administered 6 months apart assessing QOL (using the Medical Outcomes Study Short Form-36) and demographic, socioeconomic, psychosocial, and behavioral factors. Disease activity, damage, and treatment were recorded at both evaluations. Multiple linear regression (adjusting for baseline health status) was used to identify factors influencing mental and physical health. RESULTS: Improved physical health after 6 months was associated with reductions in learned helplessness (p = 0.034), improved mental health (p<0.001), longer disease duration (p = 0.009), and better physical health at baseline (p = 0.027). Improved mental health after 6 months was associated with better family support (p = 0.002), improvements in physical health (p<0.001), disease activity, and prednisolone dose (interaction term p = 0.019), less disease related damage (p<0.001), non-use of cytotoxic drugs (p = 0.02), and older age at diagnosis (p = 0.007). CONCLUSION: Potentially modifiable psychosocial, disease, and therapy related factors influence QOL in patients with SLE.

Validation of the medical outcomes study family and marital functioning measures in SLE patients in Singapore.

OBJECTIVE: To validate the Medical Outcomes Study Family and Marital Functioning Measures (FMM and MFM) in a multi-ethnic, urban Asian population in Singapore. METHODS: English speaking Chinese, Malay or Indian SLE patients (n=120) completed a self-administered questionnaire containing the FFM and MFM at baseline, after 2 weeks and after 6 months. Lupus activity, disease-related damage and quality of life were assessed using the British Isles Lupus Assessment Group (BILAG), Systemic Lupus International Collaborating Clinics/American College of Rheumatology (SLICC/ACR) Damage Index and SF-36 Health Survey respectively. Scale psychometric properties were assessed through factor analysis, Cronbach's alpha, quantifying test-retest differences and known-groups construct validity. RESULTS: Factor analysis of scores obtained at baseline and after 6 months identified 3 factors corresponding to the FFM (1 factor) and the MFM (2 factors). Both scales showed acceptable internal consistency, with Cronbach's alpha of 0.95 for the FFM and 0.70 for the MFM. Mean (s.d.) test-retest differences were -0.31 (3.82) points for the FFM and -0.70 (4.26) points for the MFM. Eleven out of 13 a priori hypotheses relating both the FFM and MFM to demographic, disease and quality of life variables were confirmed, supporting the construct validity of these scales. CONCLUSION: The FFM and MFM are valid and reliable measures of family and marital functioning in a multi-ethnic cohort of Asian SLE patients in Singapore.

The Rheumatology Attitudes Index and its helplessness subscale are valid and reliable measures of learned helplessness in Asian patients with systemic lupus erythematosus.

OBJECTIVE: To assess the internal consistency, reliability, and construct validity of the Rheumatology Attitudes Index (RAI) and its subscales in a cohort of Asian patients with systemic lupus erythematosus (SLE). METHODS: English speaking ethnic Chinese, Malay, or Indian patients with SLE (n = 120) seen at a rheumatology unit completed a questionnaire containing the RAI twice within a 2 week period. Lupus activity was assessed using the British Isles Lupus Activity Group (BILAG) score, disease related damage using the Systemic Lupus International Collaborating Clinics/American College of Rheumatology (SLICC/ACR) damage index, and quality of life using the Medical Outcome Survey Short Form 36 Health Survey (SF-36). Factor analysis and Cronbach's alpha were used to study the psychometric properties of the RAI. The magnitude of test-retest differences was assessed using the method of Bland and Altman. Relationships between the RAI, its helplessness (HS) and internality (IS) subscales, and BILAG, SLICC/ACR damage index and SF-36 scores were studied using Spearman's rank correlation. RESULTS: Factor analysis (n = 105) identified 2 factors corresponding to the HS and IS subscales of the RAI. All scales showed acceptable internal consistency, with Cronbach's alpha of 0.64 for the HS, 0.77 for the IS, and 0.74 for the RAI. Mean (SD) test-retest differences were 0.85 (3.96) points for the HS (n = 86), 0.81 (4.44) points for the IS (n = 85), and 1.46 (7.88) points for the RAI (n = 74). Six of 10 hypotheses relating the RAI and HS to demographic, disease, and quality of life variables were confirmed, supporting the construct validity of these scales. CONCLUSION: The RAI and its helplessness subscale are valid and reliable measures of learned helplessness in a multiethnic cohort of Asian patients with SLE in Singapore.

Validation of the MOS SF-36 for quality of life assessment of patients with systemic lupus erythematosus in Singapore.

OBJECTIVE: To assess the reliability and construct validity of the Medical Outcomes Study Short Form 36 Health Survey (SF-36) in a multiracial cohort of Asian patients with lupus in Singapore. METHODS: A cross sectional study was performed on 118 English speaking patients with lupus attending a specialist rheumatology unit between March and August 1996. Patients completed a questionnaire containing the UK standard version of the SF-36 twice within a 14 day period. All patients were assessed for disease activity using the British Isles Lupus Assessment Group score (BILAG), and for disease related damage using the Systemic Lupus International Collaborating Clinics/American College of Rheumatology (SLICC/ACR) damage index (DI). Relationships between SF-36, BILAG, and DI scores were studied using Spearman's rank correlation. Internal consistency of the SF-36 was assessed using Cronbach's alpha, and stability using the repeatability coefficient of Bland and Altman. RESULTS: SF-36 subscales showed high internal consistency, with Cronbach's alpha coefficient ranging from 0.84 to 0.94. Test-retest reliability was acceptable, with Spearman's rank correlation >0.70 for all subscales except role-physical, and mean differences in test scores of <2 points for 5 of 8 subscales. SF-36 subscale scores were weakly correlated with BILAG scores (Spearman's p -0.37 to 0.15) and SLICC/ACR DI scores (Spearman's p -0.25 to 0.23), suggesting divergent construct validity of the SF-36. CONCLUSION: These data suggest the SF-36 is a reliable and valid measure of the quality of life of patients with lupus in Singapore.