BACKGROUND: Generic health-related quality of life measures are often applied to disease groups without assessment of their psychometric properties. The current work assesses the properties of the Short Form 36-item (SF-36) questionnaire in a British sample of adolescents and adults with cystic fibrosis (CF). METHODS: Two hundred and twenty-three adolescents and adults with CF completed the SF-36 with a further 185 approached and not responding by non-completion of the questionnaire. The structure and internal reliability of the instrument was assessed by principal components analysis, Cronbach alpha coefficients and item to domain correlations. Differences between disease severity groups were assessed by analysis of variance. RESULTS: Factor analysis of the SF-36 scores broadly confirmed domain structures for the SF-36. Cronbach alpha coefficients were high (range 0.82-0.91) and item-to-same domain correlations were stronger than item-to-unrelated domain correlations. Examination of differences between mild, moderate and severe disease states revealed four significant main effects for: physical functioning, role limitation due to physical functioning, general health perceptions and energy and vitality. The analysis also revealed the presence of numerous ceiling effects across domains. CONCLUSIONS: The domain structure of the SF-36 was demonstrated to be robust. However, the discriminatory ability of the measure was disappointing. The presence of ceiling effects and the low frequency of differences between intermediate disease severity groups indicated that the SF-36 was not discriminatory with respect to mild disease states or progression of illness.
BACKGROUND: Generic health-related quality of life measures are often applied to disease groups without assessment of their psychometric properties. The current work assesses the properties of the Short Form 36-item (SF-36) questionnaire in a British sample of adolescents and adults with cystic fibrosis (CF). METHODS: Two hundred and twenty-three adolescents and adults with CF completed the SF-36 with a further 185 approached and not responding by non-completion of the questionnaire. The structure and internal reliability of the instrument was assessed by principal components analysis, Cronbach alpha coefficients and item to domain correlations. Differences between disease severity groups were assessed by analysis of variance. RESULTS: Factor analysis of the SF-36 scores broadly confirmed domain structures for the SF-36. Cronbach alpha coefficients were high (range 0.82-0.91) and item-to-same domain correlations were stronger than item-to-unrelated domain correlations. Examination of differences between mild, moderate and severe disease states revealed four significant main effects for: physical functioning, role limitation due to physical functioning, general health perceptions and energy and vitality. The analysis also revealed the presence of numerous ceiling effects across domains. CONCLUSIONS: The domain structure of the SF-36 was demonstrated to be robust. However, the discriminatory ability of the measure was disappointing. The presence of ceiling effects and the low frequency of differences between intermediate disease severity groups indicated that the SF-36 was not discriminatory with respect to mild disease states or progression of illness.
Authors: Donna K McClish; Lynne T Penberthy; Viktor E Bovbjerg; John D Roberts; Imoigele P Aisiku; James L Levenson; Susan D Roseff; Wally R Smith Journal: Health Qual Life Outcomes Date: 2005-08-29 Impact factor: 3.186