Test-retest performance of a mailed version of the medical outcomes study 36-item short-form health survey among older adults

OBJECTIVES. This study reports the reliability, internal consistency, and response patterns for a mailed version of the Medical Outcomes Study 36-item Short-Form Health Survey (SF-36) among older adults. METHODS. The SF-36 surveys were mailed to patients living in the community aged 65 years and older who were enrolled in two primary care practice clinics. Comorbidity scores also were assigned to patients based on their diagnoses from computerized clinic encounter forms using Deyo et al's modification of the Charlson Index. Subjects repeated the SF-36 by mail after 1 month. RESULTS. Four-hundred twenty-two subjects were mailed surveys and 253 returned them (60.0%). Missing items further reduced the number of subjects with scores on all SF-36 scales. A total of 186 subjects completed both baseline and retest SF-36 surveys. Intraclass correlation coefficients generally were high and ranged from 0.648 to 0.868. Internal consistency of scales also was high (0.802 to 0.924). Mean SF-36 scale scores decreased significantly with increasing comorbidity levels. CONCLUSIONS. The SF-36 demonstrated good retest reliability and internal consistency among these older adults and also showed a strong relationship to an external measure of comorbidity/health status. Mailed surveys pose a problem of response among older adults, a problem not unique to the SF-36, and methods for increasing response are needed when personal interviews are not feasible.