Are English- and Chinese-language versions of the SF-6D equivalent? A comparison from a population-based study

doi:10.1016/S0149-2918(04)90186-5

Clinical Therapeutics

Volume 26, Issue 7, July 2004, Pages 1137-1148

https://doi.org/10.1016/S0149-2918(04)90186-5 Get rights and content

Abstract

Objective: The goal of this study was to assess the equivalence of English- and Chinese-language versions of the SF-6D (a 6-dimensional health classification system based on the 36-Item Short Form Health Survey) using a model of equivalence proposed previously.

Methods: We analyzed data from a previously published, cross-sectional, population-based survey of ethnic Chinese in Singapore, using linear regression models to adjust for the influence of potential confounding variables. Based on equivalence clinical trial methods, measurement (ie, scale) and item equivalence were assessed by comparing 90% CIs of differences in scores due to language with predefined equivalence margins, that corresponded to the minimum clinically important difference for SF-6D utility and item scores.

Results: Data from 2558 respondents (aged 21–65 years; 48.8% completed the English-language version) were analyzed. The utility scores of respondents using the English- or Chinese-language versions of the SF-6D had similar distribution patterns, with a mean (SD) utility score of 0.8 (0.12). Adjusted 90% CIs for differences in utility and item scores due to language fell within predefined equivalence margins, suggesting measurement and item level equivalence. The 90% CI for scale scores was −0.0089 to 0.0065 (range, 0.0154; equivalence margin, 0.033); the 90% CI for item scores varied from −0.0046 to −0.0020 (range, 0.0026; equivalence margin, 0.0036) for vitality to −0.0024 to 0.0037 (range, 0.0061; equivalence margin, 0.0088) for social functioning. Functional equivalence was suggested because the various aspects of equivalence proposed previously were demonstrated in this study.

Conclusions: English- and Chinese-language versions of the SF-6D demonstrated item, measurement, and functional equivalence in this population-based study comparing 2 widely used languages with very different linguistic structures. This suggests that English and Chinese SF-6D scores can be pooled, thus increasing the representativeness and power of studies using the SF-6D, and providing a basis for studies to value health by obtaining SF-6D utility scores in Asian populations.

References (32)

J. Brazier et al.
Deriving a preference-based single index from the UK SF-36 Health Survey
J Clin Epidemiol
(1998)
J. Brazier et al.
The estimation of a preference-based measure of health from the SF-36
J Health Econ
(2002)
M.A. Choe et al.
Quality of life for patients with diabetes in Korea—I: The meaning of health-related quality of life
Int J Nurs Stud
(2001)
C.L. Lam et al.
Tests of scaling assumptions and construct validity of the Chinese (HK) version of the SF-36 Health Survey
J Clin Epidemiol
(1998)
N. Luo et al.
Do English and Chinese EQ-5D versions demonstrate measurement equivalence? An exploratory study
Health Qual Life Outcomes
(2003)
J.E. Ware et al.
C.A. McHorney et al.
The MOS 36-item Short-Form Health Survey (SF-36): III. Tests of data quality, scaling assumptions, and reliability across diverse patient groups
Med Care
(1994)
J.E. Ware et al.
The MOS 36-item shortform health survey (SF-36). I. Conceptual framework and item selection
Med Care
(1992)
C.A. McHorney et al.
The MOS 36-Item Short-Form Health Survey (SF-36): II. Psychometric and clinical tests of validity in measuring physical and mental health constructs
Med Care
(1993)
J.E. Ware et al.

J. Bernhard et al.

Applying quality of life principles in international cancer clinical trials

M. Herdman et al.

A model of equivalence in the cultural adaptation of HRQoL instruments: The universalist approach

Qual Life Res

(1998)

J. Yu et al.

Equivalence of Chinese and US-English versions of the SF-36 health survey

Qual Life Res

(2003)

J. Thumboo et al.

Does being bilingual in English and Chinese influence responses to quality-of-life scales?

Med Care

(2002)

Cited by (28)

A Comparison of UK and Brazilian SF-6D Preference Weights When Applied to a Brazilian Urban Population
2019, Value in Health Regional Issues
The 6-dimensional health state short form (SF-6D) is a health preference measure used in economic evaluations of many treatments.
To compare the results provided by the SF-6D index, when applied to a representative sample of the Brazilian population, using Brazilian and UK preference weights.
Five thousand individuals were assessed in the 5 regions of Brazil. Preference measures in healthcare were assessed using the SF-6D Brazil, version 2002. To calculate the single utility score, 2 preference weights were used: one established for the Brazilian population (SF-6D Brazil) and the other for the UK population (SF-6D UK). Agreement between the SF-6D Brazil and the SF-6D UK was assessed using the intraclass correlation coefficient, the Wilcoxon signed rank test, confidence intervals (CIs), and the Bland-Altman method.
The mean values of the SF-6D Brazil and the SF-6D UK were 0.83 ± 0.15 and 0.84 ± 0.15, respectively. The intraclass correlation coefficient was 0.952 (CI 0.942-0.960; P<.010). The Wilcoxon signed rank test and CI showed a statistically significant difference between the 2 measures; this difference was, however, very small and considered clinically irrelevant (CI 0.011-0.013; P<.010). Using the Bland-Altman method resulted in a mean difference of 0.012 and the limits of agreement were between −0.077 and 0.101.
The present study identified very small quantitative differences between UK- and Brazilian-derived SF-6D scores. Tests of agreement, however, showed that the impact of using different sets of preference weights in the construction of quality-adjusted life-year might be considered irrelevant.
A comparison of the reliability and validity of SF-6D, EQ-5D and HUI3 utility measures in patients with schizophrenia and patients with depression in Singapore
2019, Psychiatry Research
Citation Excerpt :
Utility scores generated by the SF-6D range from 0.29 to 1, with 1 representing full health and 0.29 representing the worst possible health state defined by the SF-6D (i.e., all domains being at the worst level). The utility scores derived from English and Chinese versions of the SF-6D have been demonstrated to be equivalent in Singapore (Wee et al., 2004). The HUI3 (Feeny et al., 1995) is a generic comprehensive health status classification and HRQoL utility scoring system, using a utility scoring function derived from a representative sample of the Canadian general population based on the SG and visual analog scale methods (Horsman et al., 2003).
There is limited evidence of a direct comparison of the psychometric performance of generic preference-based measures in patients with mental illness in an Asian patient population. The current study aimed to compare the test–retest reliability, convergent and known-group validity and magnitude of change in scores of the EuroQol Five-Dimension, Health Utility Index Mark 3 (HUI3) and Short-Form Six-Dimension (SF-6D) measures in patients with depression and patients with schizophrenia spectrum disorder. 500 patients were recruited from a tertiary psychiatric institution in Singapore. The Schizophrenia Quality of Life Scale (SQLS), 8-item Patient Health Questionnaire, Quality of Life Enjoyment and Satisfaction Questionnaire (Q-LES-Q), and Positive and Negative Syndrome Scale were also included. In the schizophrenia sample, the SF-6D was found to have higher test–retest validity, convergent validity with SQLS domain scores, known-group validity and magnitude of change in scores over 6-month follow up than other measures. In the depression sample, the HUI3 was found to have higher test–retest reliability, convergent validity with Q-LES-Q, known group validity and magnitude of change in scores than other measures. Results suggest that the SF-6D and HUI3 to be more suitable as a utility measure for patients with schizophrenia and depression in an Asian patient population.
The Short Form 36 English and Chinese versions were equivalent in a multiethnic Asian population
2013, Journal of Clinical Epidemiology
Citation Excerpt :
Domain equivalence of the SF-36v2 was taken as being present if the 90% CI of the differences in domain scores due to language fell within the five-point difference on the norm-based score [3]. As per our previous work [21], we defined that the equivalence of SF-6D scale would be achieved if the 90% CI of the differences in scores due to language version for SF-6D utility scores fell within the predefined equivalence margin. Item equivalence was assessed similarly using linear regression models with the SF-6D items as dependent variable and language as the independent variable, with and without adjustment for other known determinants of HRQoL.
The primary aim of this article was to evaluate measurement equivalence of the English and Chinese versions of the Short Form 36 version 2 (SF-36v2) and Short Form 6D (SF-6D).
In this cross-sectional study, health-related quality of life (HRQoL) was measured from 4,973 ethnic Chinese subjects using the SF-36v2 questionnaire. Measurement equivalence of domain and utility scores for the English- and Chinese-language SF-36v2 and SF-6D were assessed by examining the score differences between the two languages using linear regression models, with and without adjustment for known determinants of HRQoL. Equivalence was achieved if the 90% confidence interval (CI) of the differences in scores, due to language, fell within a predefined equivalence margin.
Compared with English-speaking Chinese, Chinese-speaking Chinese were significantly older (47.6 vs. 55.5 years). All SF-36v2 domains were equivalent after adjusting for known HRQoL. SF-6D utility/items had the 90% CI either fully or partially overlap their predefined equivalence margin.
The English- and Chinese-language versions of the SF-36v2 and SF-6D demonstrated equivalence.
Are English and Chinese Versions of the Audit of Diabetes-Dependent Quality of Life Equivalent? An Exploratory Study Based on the Universalist Approach
2012, Value in Health Regional Issues
Citation Excerpt :
Respondents who completed the Chinese ADDQOL were more likely to be unemployed (P < 0.01), have received fewer years of education (P < 0.05), and had a lower level of housing (P < 0.01) and poorer family functioning (P < 0.01). The perception of QOL among Chinese-speaking Singaporeans has been well demonstrated by several local studies [4,16,19]. The relevance of 19 domains to the Chinese-speaking Singaporean diabetes patients was confirmed by two local diabetes experts and respondents during the CD.
To translate and culturally adapt the UK English Audit of Diabetes-Dependent Quality of Life (ADDQOL) into Chinese for Singapore.
Translation was integrated into investigation of conceptual, item, semantic, and operation equivalence. Conceptual equivalence, item equivalence, and operation equivalence were assessed by literature review, expert judgment, and cognitive debriefing. Semantic equivalence was studied by using an optimized procedure including forward and backward translation, clinician review, and cognitive debriefings. Cognitive debriefings were done with five Chinese-speaking diabetic patients at polyclinics. Reliability, responsiveness, and construct validity tests were used to evaluate measurement equivalence. English- and Chinese-speaking diabetic patients by convenient sampling at a Diabetes Society of Singapore's public event were recruited for the measurement equivalence study. Mann-Whitney U tests, chi-square tests, and descriptive analyses were used for group comparisons and Spearman's correlation coefficients for construct validity tests.
Forty-two English-speaking and 26 Chinese-speaking diabetic patients (45.5% females) with a mean age of 54.2 ± 10.07 years were recruited. Chinese-speaking respondents were more likely than the English-speaking group to be unemployed, less educated, and with poorer family functioning (P < 0.05). Conceptual equivalence, item equivalence, operation equivalence, and semantic equivalence were all demonstrated. Cronbach's alpha for internal consistency and intraclass correlation coefficient for test-retest reliability were 0.94 and 0.65, respectively. Distributions of responses were found to be similar except for some difference that can be justified by different demographic background. Convergent validity was suggested by weak to moderate correlations between “Present QOL” on the ADDQOL and EuroQol five-dimensional questionnaire (r = 0.268; P = 0.185) and six-dimensional health state short form (derived from short form 36 health survey) scores (r = 0.351; P = 0.078); divergent validity was shown by a weak correlation between ADDQOL average weighted impact (AWI) and ADDQOL “Present QOL” scores (r = 0.027, P = 0.896), a moderate correlation between ADDQOL AWI and six-dimensional health state short form (derived from short form 36 health survey) mental scores (r = 0.247; P = 0.224), and a positive correlation between ADDQOL AWI and family functioning scores (r = 0.288; P = 0.182).
The ADDQOL has been translated and culturally adapted successfully into a Chinese version for Singapore. Our study provides justification for further research with large sample sizes among the Chinese-speaking population in Singapore.
An exploratory study of response shift in health-related quality of life and utility assessment among patients with osteoarthritis undergoing total knee replacement surgery in a tertiary hospital in Singapore
2012, Value in Health
Citation Excerpt :
The SF-6D score or the SF-6D utility index is scaled from 0.26 to 1.00 continuously, with 0.26 representing the worst health state (all dimensions being at the worst level) and 1.00 representing full health (all dimensions being at full functional level). The validity and equivalence of the SF-6D in English and Chinese versions have been previously demonstrated in a population-based study in Singapore [25]. The EQ-5D is a preference-based HRQOL measure with five domains (mobility, self-care, usual activities, pain/discomfort, and anxiety/depression) for respondents to self-classify and rate their health status [26].
To investigate the influence of response shift (RS) on health-related quality of life (HRQOL) and utility assessment among patients undergoing total knee replacement.
Consenting patients undergoing total knee replacement were interviewed to determine their HRQOL by using the six-dimensional health state short form, derived from SF-36, and the EuroQol five-dimensional questionnaire at baseline (pretest 1) and the six-dimensional health state short form, derived from SF-36, at 6 (pretest 2) and 18 months after surgery (post-test). RS was studied by using a “then-test” approach by contacting participants 18 months after surgery and asking them to evaluate their HRQOL at baseline (then-test 1) and at 6 (then-test 2) and 18 months after surgery. RS was calculated as the score difference between pretest and then-test scores for a given time point. Relationships between RS and external variables were explored by using univariate and multiple liner regression analyses.
In 74 subjects (63% response rate, median age 68 years), median (interquantile range) six-dimensional health state short form, derived from SF-36, scores for then-tests at baseline (0.48 [0.42–0.49]) and at 6 months (0.72 [0.66–0.79]) after surgery were significantly different from respective pretest scores (0.61 [0.58–0.68] at baseline, P = 0.000; 0.69 [0.63–0.72] at 6 months, P = 0.000), showing RS at both time points. RS at baseline (0.14 [0.08–0.20]) was significantly larger than that at 6 months (−0.05 [0.14 to 0.00], P = 0.000). EuroQol five-dimensional questionnaire pretest and then-test scores at baseline also differed significantly (0.69 [0.17–0.73] vs. −0.18 [−0.23 to 0.00], P = 0.000). RS at baseline was not affected by assessed demographic or medical variables. RS at 6 months was greater in subjects with more years of education (16% of variance in multiple liner regression, P < 0.01).
RS was present and impacted HRQOL and utility assessment among patients undergoing total knee replacement before and 6 months after surgery.
Assessing differences in utility scores: A comparison of four widely used preference-based instruments
2007, Value in Health
To characterize the differences in utility scores (dUTY) among four commonly used preference-based Health-Related Quality of Life instruments, to evaluate the potential impact of these differences on cost-utility analyses (CUA), and to determine if sociodemographic/clinical factors influenced the magnitude of these differences.
Consenting adult Chinese, Malay and Indian subjects in Singapore were interviewed using Singapore English, Chinese, Malay or Tamil versions of the EQ-5D, Health Utilities Index Mark 2 (HUI2) and Mark 3 (HUI3), and SF-6D. Agreement between instruments was assessed using Bland–Altman (BA) plots. Changes in incremental cost-utility ratio (ICUR) from dUTY were investigated using eight hypothetical decision trees. The influence of sociodemographic/clinical factors on dUTY between instrument pairs was studied using multiple linear regression (MLR) models for English-speaking subjects (circumventing structural zero issues).
In 667 subjects (median age 48 years, 59% female), median utility scores ranged from 0.80 (95% confidenceinterval [CI] 0.80, 0.85) for the EQ-5D to 0.89 (95% CI 0.88, 0.89) for the SF-6D. BA plots: Mean differences (95% CI) exceeded the clinically important difference (CID) of 0.04 for four of six pairwise comparisons, with the exception of the HUI2/EQ-5D (0.03, CI: 0.02, 0.04) and SF-6D/HUI2 (0.02, CI: 0.006, 0.02). Decision trees: The ICER ranged from $94,661/QALY (quality-adjusted life-year; 6.3% difference from base case) to $100,693/QALY (0.3% difference from base case). MLR: Chronic medical conditions, marital status, and Family Functioning Measures scores significantly (P-value < 0.05) influenced dUTY for several instrument pairs.
Although CIDs in utility measurements were present for different preference-based instruments, the impact of these differences on CUA appeared relatively minor. Chronic medical conditions, marital status, and family functioning influenced the magnitude of these differences.

View all citing articles on Scopus

View full text

Are English- and Chinese-language versions of the SF-6D equivalent? A comparison from a population-based study

Abstract

J Clin Epidemiol

J Health Econ

Int J Nurs Stud

J Clin Epidemiol

Health Qual Life Outcomes

The MOS 36-item Short-Form Health Survey (SF-36): III. Tests of data quality, scaling assumptions, and reliability across diverse patient groups

Med Care

The MOS 36-item shortform health survey (SF-36). I. Conceptual framework and item selection

Med Care

The MOS 36-Item Short-Form Health Survey (SF-36): II. Psychometric and clinical tests of validity in measuring physical and mental health constructs

Med Care

Applying quality of life principles in international cancer clinical trials

A model of equivalence in the cultural adaptation of HRQoL instruments: The universalist approach

Qual Life Res

Equivalence of Chinese and US-English versions of the SF-36 health survey

Qual Life Res

Does being bilingual in English and Chinese influence responses to quality-of-life scales?

Med Care