Survey on the applicability of SF-36 version-2 (SF-36v2) in assessment quality of life among urban residents in Chengdu city

Abstract

To explore the psychometric performances and applicability of SF-36v2 in assessment quality of life among urban residents in Chengdu. During Oct. to Dec., 2012, 2 186 adult urban residents with clear mind and well self-express were recruited in the study by multistage stratified cluster sampling method in Chengdu urban area. The survey questionnaires included general health condition and quality of life, which was adopted the SF-36v2. Internal consistency reliability, test-retest reliability and construct validity were all analyzed as indicators of the psychometric performance. The survey released 2 186 questionnaires, with 2 182 ones returned and 2 178(99.8%) met the data standard. The scores of 8 scales in SF-36v2, including physical function (PF), role-physical (RP), bodily pain (BP), general health (GH), vitality (VT), social function (SF), role-emotion (RE) and mental health (MH), were 89.15 ± 17.56, 85.18 ± 22.52, 76.64 ± 17.80, 64.13 ± 19.56, 70.39 ± 17.31, 86.43 ± 17.35, 87.79 ± 19.24 and 80.61 ± 13.49, respectively; the floor effects were 0.28%, 0.41%, 0.23%, 0.28%, 0.09%, 0.05%, 0.14% and 0.23%, respectively; and the ceiling effects were 51.38%, 60.60%, 58.08%, 0.83%, 2.94%, 50.32%, 64.00% and 3.95%, respectively. The item-convergent validities were all achieved the standard (r = 0.40) except the item MH5 (Have you been happy?), and the total scaling success rate of item-convergent validity was 97.14%. The scales' success rates of item-discriminant validities for the SF, VT and MH scales were 93.75%, 56.25% and 97.50% respectively, while the rates of others were 100.00% and the total success rate was 96.43%. The internal reliability ranged from 0.724 to 0.974 across all the scales, except for SF (r = 0.603) and VT (r = 0.697). The two-week test-retest reliability ranged from 0.610 to 0.845. Within factor analysis, two common factors were confirmed, separately representing physical health and mental health, altogether contributing 64.4% of the total variance. As a revised version of SF-36v1, the SF-36v2 seemed to be more preferable in layout for questions and answers and could reduce the ceiling and floor effect. Additionally, it also showed comparatively well reliability and validity. And thereby we believed the SF-36v2 could be applied to assess the life quality among urban residents in Chengdu.