Assessment of the SF-36 version 2 in the United Kingdom.

OBJECTIVES To introduce the UK SF36 Version II (SF36-II), and to (a) gain population norms for the UK SF36-II in a large community sample as well as to explore the instrument's internal consistency reliability and construct validity, and (b) to derive the Physical Component Summary (PCS) and Mental Component Summary (MCS) algorithms for the UK SF36-II. DESIGN Postal survey using a questionnaire booklet, containing the SF-36-II and questions on demographics and long term illness. SETTING The sample was drawn from General Practitioner Records held by the Health Authorities for Berkshire, Buckinghamshire, Northamptonshire, and Oxfordshire. SAMPLE The questionnaire was sent to 13,800 randomly selected subjects between the ages of 18-64 inclusive. OUTCOME MEASURES Scores for the eight dimensions of the UK SF36-II and the PCS and MCS summary scores. RESULTS The survey achieved a response rate of 64.4% (n = 8889). Internal consistency of the different dimensions of the questionnaire were found to be high. Normative data for the SF-36 are reported, broken down by age and sex, and social class. Factor analysis of the eight domains produced a two factor solution and provided weights for the UK SF36-II. CONCLUSION The SF36-II domains were shown to have improved reliability over the previous version of the UK SF36. Furthermore, enhancements to wording and response categories reduces the extent of floor and ceiling effects in the role performance dimensions. These advances are likely to lead to better precision as well as greater responsiveness in longitudinal studies.