Testing the measurement properties of the Spanish version of the SF-36 Health Survey among male patients with chronic obstructive pulmonary disease

The aim of this article is to evaluate the measurement properties of the Spanish version of the SF-36 Health Survey (SF-36). In total, 321 male chronic obstructive pulmonary disease (COPD) patients attending hospital outpatient or primary health clinics were cross-sectionally administered the SF-36, the St. George's Respiratory Questionnaire (SGRQ), and a dyspnea scale. A clinical measure of respiratory function, the proportion of the predicted Forced Expiratory Volume in 1 second (%FEV1) was also obtained. Internal consistency, central tendency, and dispersion statistics of scores were calculated, as well as the percentage of respondents with the highest and lowest scores for each scale and correlations between health status and clinical measures. All patients completed the SF-36 questionnaire, and less than 1% of items were missing. The proportion of patients with the worst possible score (floor effect) ranged from 0.9-20.1% among the different scales. The proportion of patients achieving the best possible score (ceiling effect) ranged from 0-61%. Cronbach's alpha coefficients were above 0.75 except for one scale (Social Functioning, alpha = 0.55). SF-36 scores were moderate to highly correlated with SGRQ scores (coefficients ranged from -.41 to -.79). Correlations were moderate to high with dyspnea and lower but statistically significant with %FEV1. A clear gradient of SF-36 scores was found according to dyspnea levels and disease staging based on %FEV1 impairment, the gradient being more marked for the Physical Functioning, Role-Physical, and General Health scales. Data presented suggest that the Spanish version of the SF-36 is acceptable, valid, and reliable in COPD patients. Although more studies are needed, this version is adequate in international comparisons of health outcomes. (C) 1998 Elsevier Science Inc.