Stroke rehabilitation: a model predicting return home.

We undertook this study to describe the changes in functional status for patients in a rehabilitation program for acute stroke and to identify the variables that best predict discharge home. Of 282 patients, 75% were discharged home. Increases in functional status were found for all 18 activities of the Functional Independence Measure from admission to discharge. Significant predictors of discharge disposition in a logistic regression model were the admission and discharge functional status scores, length of stay, and living arrangement before the stroke. The functional status at discharge was the most important predictor. Knowledge of these predictors can contribute to more appropriate treatment and discharge planning.