Performance comparison of Likert and binary formats of SF-36 version 1.6 across ECRHS II adults populations

Abstract

Objectives: To evaluate a binary response structure of SF-36 items assessing scaling assumptions, reliability, and validity of questionnaire. Methods: An optimal scaling accounting for the nonmetric properties of the data was used to reduce SF-36 Likert item responses to give a binary coding. The binary recoding was compared with the original format regarding item analysis, underlying latent components and know-groups clinical validity using ordered correlation/regression methods. Data from the European Community Respiratory Health Survey Follow-up (ECRHS II) of 8854 subjects from 25 centers were analyzed to cross-validate the binary coding proposal. Results: Overall, the testing comparison produces results indicating that the binary recoding of the SF-36 scales meets at least similar standards without jeopardizing the underling structure of the original format. Internal binary consistencyshows comparable values with the Likert ones and these are always higher than the minimum suggested. The Principal Component structure was well replicated and know-groups validity gives similar research findings for symptomatic, long-term illness and depression differences. Conclusions: Although there is lost of information due to the reduction of response's chance, our results indicate that the SF-36 binary recoding gives the possibility to suggest a new version of smarter and easier methodology of administration, compilation, score calculation, and data processing. Consequently, it may be an alternative to the existing shorter versions, suitable in administering in clinical setting and clinical trials, in subjects with serious diseases, and by telephone. © 2007, International Society for Pharmacoeconomics and Outcomes Research (ISPOR).