[Reliability and validity of a Short-Form Health Survey Scale (SF-36), Chinese version used in an elderly population of Zhejiang province in China].

OBJECTIVE

To evaluate the validity and reliability of a Chinese version on the Short-Form Health Survey Scale (SF-36) among elderly population.

METHODS

Questionnaire including SF-36 Chinese version was administered in a cross-sectional study from October to December 2007. 4241 elderly people over 60 years old from rural and urban area in Zhejiang province were face-to-face interviewed by well-trained investigators, and then analyses on correlation, reliability, factors, t-test and one-way ANOVA were made to evaluate on reliability and validity of the scale.

RESULTS

The SF-36 Chinese version had good split-half reliability (r = 0.91, P < 0.001) and all the internal consistency Cronbach's alpha coefficients exceeded 0.8 except for VT (alpha = 0.65), SF (alpha = 0.65) and MH (alpha = 0.40). The correlative coefficients between each item and its domain were all greater than the 0.4 thresholds except item 9 - 2, and they were greater than the correlation between the item and other domains for all but item 9 - 8. These results demonstrated that the SF-36 Chinese version had good convergent validity and discriminant validity. The distribution of 35 items in 6 public factors (the cumulative variance = 67.04%) extracted from them was consistent with the basic conceived concept. All domains except MH had good validity to discriminate different categories.

CONCLUSION

The SF-36 Chinese version had good reliability and validity. It was acceptable for the evaluation on quality of life in elderly population, but the reliability and validity of MH were relative low and the items such as 9 - 2, 9 - 8 in MH and 3 - 1 in PF were not suitable for Chinese elderly population.