Predicting mobility outcome one year after stroke: a prospective cohort study.

OBJECTIVE To develop a prognostic model to predict mobility outcome one year post-stroke. DESIGN Prospective cohort study in patients with a first-ever stroke admitted for inpatient rehabilitation. PATIENTS A total of 217 patients with stroke (mean age 58 years) following inpatient rehabilitation in 4 rehabilitation centres across the Netherlands. METHODS Mobility was measured using the Rivermead Mobility Index at one year poststroke. Included independent variables were: patient and stroke characteristics, functional status, urinary incontinence, sitting balance, motor and cognitive function. Univariate and multivariate linear regression analyses were performed in a model-developing set (n=174) and the model was validated in cross-validation set (n=43). RESULTS Total Rivermead Mobility Index score at one year post-stroke was predicted by functional status, sitting balance, time between stroke onset and measurement, and age. The derived model predicted 48% of the variance, while validation in the cross-validation set resulted in an adjusted R(2) of 0.47. CONCLUSION The present prospective study shows that outcome of mobility one year after stroke can be predicted validly by including functional status, sitting balance, moment of admission to the rehabilitation centre after stroke onset and age.