Use of the SF-36 health status survey with a chronically disabled back pain population: Strengths and limitations

The SF-36 is a widely used self-report health status survey. If has repeatedly demonstrated sound psychometric properties with regard to certain aspects of reliability and validity, and has proven effective as a means of tracking group outcomes. There were three major objectives of the present investigation: (I) to document group changes in a sample of chronically work disabled low back pain patients who underwent tertiary rehabilitation; (2) to evaluate the clinical utility of the SF-36 in measuring changes in patients as a result of treatment; and (3) to compare the SF-36 scores found for this sample to the United States norms for back pain/sciatica patients published elsewhere. All patients (n = 286) received an initial evaluation consisting of several assessment procedures, including the SF-36. First, the sample was broken down into a "completer" group (n = 188), a "noncompleter" group (n = 31), and a "nonstarter" group (n 67). Next the preprogram scores for the physical and mental component summary scales were compared to the norms for back pain/sciatica patients. Changes in the "completer" group's scores from preprogram to postprogram were then examined. To assess clinical utility three additional self-report questionnaires were correlated to the eight scales and two summary scales of the SF-36. Results revealed no statistically significant differences with regard to demographics among the three groups. The preprogram scores for the physical and mental component summary scales were significantly lower when compared to the norms for back pain/sciatica patients. Further patients who completed the program showed improvement on the SF-36 from preprogram to postprogram administration. When the correlations between SF-36 scales and other self-report questionnaires were examined, many of the highest correlations were found between the self-reported pain and disability measurements and the physical components of health on the SF-36. In terms of clinical usefulness, the SF-36 demonstrated limited clinical utility when attempts were made to use it on an individual patient basis. This is consistent with the psychometric limitations of the SF-36. Because of the brevity of each of the eight scales and the limited number of score levels, the reliability coefficients have been shown to be low in other studies (20), thereby increasing confidence intervals around an individual's scores on each of the scales.