Predicting motor gains with home-based telerehabilitation after stroke.

INTRODUCTION: Telerehabilitation (TR) may be useful for rehabilitation therapy after stroke. However, stroke is a heterogeneous condition, and not all patients can be expected to derive the same benefit from TR, underscoring the need to identify predictors of response to TR. METHODS: A prior trial provided patients with 6 weeks of intensive rehabilitation therapy targeting arm movement, randomly assigned to be provided in the home via TR (current focus) or in clinic. Eligible patients had moderate arm motor deficits and were in the subacute-chronic stage post stroke. Behavioral gains were measured as change in the arm motor Fugl-Meyer score from baseline to 30 days post therapy. To delineate predictors of TR response, multivariable linear regression was performed, advancing the most significant predictor from each of eight categories: patient demographics, stroke characteristics, medical history, rehabilitation therapy outside of study procedures, motivation, sensorimotor impairment, cognitive/affective deficits, and functional status. RESULTS: The primary focus was on patients starting TR >90 days post stroke onset (n = 44), among whom female sex, less spasticity, and less visual field defects predicted greater motor gains. This model explained 39.3% of the variance in treatment-related gains. In secondary analysis that also included TR patients enrolled ≤90 days post stroke (total n = 59), only female sex was a predictor of treatment gains. A separate secondary analysis examined patients >90 days post stroke (n = 34) randomized to in-clinic therapy, among whom starting therapy earlier post stroke and less ataxia predicted greater motor gains. DISCUSSION: Response to TR varies across patients, emphasizing the need to identify characteristics that predict treatment-related behavioral gain. The current study highlights factors that might be important to patient selection for home-based TR after stroke.