Abstract
OBJECTIVE: This study aimed to examine agreement between the Activity Measure for Postacute Care (AM-PAC) Computer Adaptive Test (CAT) and the "6-Clicks" short forms across the basic mobility (BM), daily activity (DA), and applied cognition (AC) domains in individuals with ischemic stroke during acute hospitalization. DESIGN: Secondary analysis of a larger prospective, longitudinal cohort study. SETTING: Acute care hospital. PARTICIPANTS: A total of 137 Individuals (N=137) aged ≥18 years who had a National Institute of Health Stroke Scale score and a "6-Clicks" and CAT within 24 hours of each other. INTERVENTIONS: Not applicable. MAIN OUTCOME MEASURES: Agreement was assessed in 4 ways; (1) intraclass correlations coefficients; (2) weighted κ statistics, quartiles; (3) equivalence using the minimal detectable difference of the domains; (4) Bland-Altman plots; and (5) we calibrated the "6-Clicks" t scores using identity linking with the CAT t scores. RESULTS: The study included 59% men, and the median age was 66 years. The "6-Clicks" and CAT scores had poor agreement according to the intraclass correlations coefficient (BM=0.58; DA=0.64; AC=0.48). Weighted κ values were 0.74 and 0.63 for BM and DA, respectively, indicating good agreement, and 0.53 for AC, indicating moderate agreement. Assessment of equivalence found they are equivalent with the mean differences. Bland-Altman plot for BM and DA had a negative trend, indicating that, on average, the "6-Clicks" had a systematic bias and overestimated compared with the CAT score and AC had a positive trend. CONCLUSIONS: We found poor to moderate agreement between the CAT and the "6-Clicks," for all 3 domains, and systematic bias at higher and lower scores in this cohort of individual's poststroke.