Outcome measurement in functional somatic syndromes: SF-36 summary scores and some scales were not valid

OBJECTIVE: This study aimed to test the validity of the 36-item Short-Form Health Survey (SF-36) scales and summaries in patients with severe functional somatic syndromes (FSS), such as fibromyalgia and irritable bowel syndrome. STUDY DESIGN AND SETTING: One hundred twenty patients with severe FSS enrolled in a randomized controlled trial filled in the SF-36 questionnaire. We tested for data quality, central scaling assumptions, and agreement with the conceptual model. RESULTS: Most SF-36 scales were found to be valid; however, three scales (role physical, role emotional, and general health) did not satisfy predefined criteria for construct validity, internal consistency, or targeting to the sample. The correlations between SF-36 scales differed considerably from those reported in the general population. As a consequence, the SF-36 summaries, physical component summary (PCS) and mental component summary (MCS), did not accurately reflect their underlying scales and were negatively correlated (r=-0.46, 95% CI [-0.60 to -0.31]). CONCLUSION: Although the SF-36 is a valuable instrument to assess perceived health in patients with severe FSS, there are problems with some of the scales and with the scoring procedure of the summaries. The SF-36 PCS may, therefore, not accurately measure the physical health status of these patients. Alternative summary measures are needed.