Test-Retest Reliability and Responsiveness of the Computerized Adaptive Test of Social Functioning in Persons With Stroke.

OBJECTIVE:To examine the test-retest reliability, minimal detectable change (MDC), responsiveness, and efficiency of the Computerized Adaptive Test of Social Functioning (Social-CAT) in patients with stroke. DESIGN:Repeated-assessments design. SETTING:A department of rehabilitation of a medical center. PARTICIPANTS:In total, 31 patients with chronic stroke and 65 patients with subacute stroke were recruited. INTERVENTION:Not available. MAIN OUTCOME MEASURE:Social-CAT. RESULTS:The Social-CAT showed acceptable test-retest reliability (intraclass correlation coefficient, 0.80) and small random measurement error (MDC%: 18.0%). However, heteroscedasticity was found (r between the means and the absolute change scores: 0.32), so the MDC% adjusted cut-off score is recommended for determining real improvement. Regarding responsiveness, the Social-CAT showed large differences (Kazis' effect size and standardized mean response: 1.15 and 1.09, respectively) in subacute patients. Regarding efficiency, the Social-CAT required an average of 5 items and less than 2 minutes for completion. CONCLUSIONS:Our findings indicate that the Social-CAT is a reliable and efficient measure with good test-retest reliability, small random measurement error, and good responsiveness. Thus, the Social-CAT is a useful outcome measure for routine monitoring of the changes in social function of patients with stroke.