Is health preference derived from the SF-36v2 equivalent to that measured by the SF-6D Health Survey?

Abstract

AIMS: The SF-6D is a preference based measure of health (PBMH) derived from the SF-36 Health Survey. The aim of this study was to find out whether the health preference value extracted from SF-36v2 data was equivalent to that measured by the SF-6D PBMH. METHODS: A cross-sectional survey was carried out on 589(86% of 684 eligible) Chinese chronic hepatitis B (CHB) patients in Hong Kong. Each subject completed the SF-6D PBMH, the SF-36v2 Health Survey and a structured questionnaire on sociodemographics, and was classified by the clinician into one of four disease severity groups: uncomplicated with normal liver function (AHB), uncomplicated impaired liver function (ILF), complicated with cirrhosis (CC) or complicated with hepatocellular carcinoma (HCC). Health preference values were measured by applying the Hong Kong population specific SF-6D scoring algorithm to data obtained by the Chinese (HK) versions of the SF-6D PBMH and SF-36v2 Health Survey, respectively. The difference in preference values by the two measures were tested by paired t tests. The relative efficiency in differentiating between disease groups of the two measures were tested by the t statistics ratio. RESULTS: There was significant difference between the SF-6D health preference measured by the SF-6D PBMH and that extracted from SF-36v2 data, overall and by groups. The SF-36v2 derived health preference values explained around 50% of the total variance measured by the SF-6D PBMH. The relative efficiency of the SF-36v2 was much lower (0.2 - 0.3) compared with the SF-6D PBMH in detecting a difference in health preference between CHB disease groups. CONCLUSIONS: The SF-6D Health Survey should be used to collect data for the calculation of health preference because it is more accurate and sensitive.