Preference-based SF-6D scores derived from the SF-36 and SF-12 have different discriminative power in a population health survey

Abstract

Objective: To compare the discriminative power of the SF-6D index scores derived from the SF-36 (SF-6D 36) and SF-12 (SF-6D 12) in the general population. Methods: Data from the National Health Measurement Study were used. The F statistic was used to compare the relative efficiency of the SF-6D 36 and SF-6D 12, as well as the EQ-5D, HUI2, and HUI3 index scores, in discriminating between respondents with and without 1 of the 11 chronic medical conditions. The efficiency of the multiattribute health classification systems of the study instruments was measured using the Shannon index (H′). The relative efficiency of the SF-6D 36 and SF-6D 12 was also compared in respondents who were on the ceilings of the EQ-5D, HUI2, and HUI3 scales. Results: The SF-6D 36 score was systematically lower than the SF-6D12 score at the group level (range, 0.022-0.036). The SF-6D 36 exhibited higher discriminative power in 8 and 5 conditions than the SF-6D 12 and all other index scores, respectively. The SF-6D 36 had higher H′ values than the SF-6D 12 in the dimensions of physical functioning (1.73 vs. 0.78), mental health (1.70 vs. 1.39), and bodily pain (2.16 vs. 1.56) as well as than all other instruments in similar health dimensions. In respondents reporting full health on the EQ-5D, HUI2, or HUI3, the SF-6D 36 better discriminated between those with and without medical conditions than the SF-6D 12. Conclusions: The SF-6D derived from the SF-36 is more discriminative than that derived from the SF-12 and is therefore preferred for use in population health surveys where a preference-based health index is needed. © 2012 by Lippincott Williams & Wilkins.