BACKGROUND AND PURPOSE: Individuals after stroke experience participation restrictions. The purpose of this study was to investigate whether measures collected early after stroke predict participation at 3 and 6 months and develop clinical prediction equations. METHODS: A prospective cohort study was conducted. Adults were initially assessed in person at the hospital (<7 days), followed by phone interviews at 28 days, 3, and 6 months. The outcome of interest was participation, measured using the Stroke Impact Scale Participation Subscale. Predictors were prestroke (age, activity/participation, environmental factors, and mobility) and poststroke factors (mobility, cognition, depression, functional independence, social support, and self-efficacy). RESULTS: A total of 251 participants (129 men) were included. The equation to predict participation at 3 months (P-3) explained 26% of the variance: P-3: Participation (y) = 25. 10 - (5. 81 × Modified Rankin Scale) + (1. 20 × stroke self-efficacy) (R2 = 0. 26). Its predictive performance showed a moderate positive predictive value (54%; 95% confidence interval CI, 0. 39-0. 68) and a high negative predictive value (83%; 95% CI, 0. 52-0. 98). The equation to predict participation at 6 months (P-6) explained 48% of the variance: P-6: Participation (y) = 64. 04 + (0. 29 × Stroke Impact Scale baseline) - (4. 33 × Modified Rankin Scale) - (1. 40 × Mini-Mental) + (0. 21 × life-space assessment) (R2 = 0. 48). The positive predictive value was moderate (55%; 95% CI, 0. 40-0. 68), and the negative predictive value was high (100%; 95% CI, 0. 59-1. 00). DISCUSSION AND CONCLUSIONS: This study provided simple equations that predict participation at 3 and 6 months after stroke, based on data collected early after stroke.
Correia et al. (Mon,) studied this question.