SF-36 scales, and simple sums of scales, were reliable quality-of-life summaries for patients with schizophrenia

OBJECTIVE: To examine the feasibility and reliability of the Medical Outcomes Study 36 Item Short Form Health Survey (SF-36) for patients with schizophrenia, focusing on the eight scales and various aggregate summary measures. STUDY DESIGN AND SETTING: Secondary analysis of data from the European multicenter QUATRO medication adherence trial and the Regional Psychosis Project from The Netherlands. Methods included exploratory and confirmatory factor analyses and estimation of aggregate score reliability using Cronbach's alpha and Tarkkonen's generalized reliability index. The aggregate scores that were compared included two sets based on factor analyses, the standard "physical and mental health component summary scores" (PCS and MCS) and scores based on the original conceptual model of the SF-36 (simple sum of first four scales for physical health, last four for mental health). RESULTS: The eight SF-36 scales were feasible to administer and reliable. Factor analyses of the QUATRO baseline scale data suggested two or three factors, the latter solution including a general "role limitation" factor. Aggregate scores based on the conceptual model had the highest generalized reliability of those compared. CONCLUSION: SF-36 scales are suitable for patients with schizophrenia. Aggregate scores based on the conceptual model may be preferable to the MCS and PCS for such patients. Further investigation of factor structure is advisable.