Predicting models of outcome stratified by age after first stroke rehabilitation in Japan

A multivariate model predicting the function at discharge following inpatient rehabilitation has been previously produced. The aim of this study is to determine predictors of function at discharge for stroke outcome and examine their accuracy of prediction. Four hundred sixty-four stroke patients were enrolled. Sex, the nature of the stroke, age, onset to rehabilitation admission interval and length of rehabilitation hospital stay were obtained from their medical records. Patients were divided into the following five groups according to age: < or = 49, 50-59, 60-69, 70-79, and > or = 80 yr. Disability was assessed on admission and at discharge by the FIM. Stepwise multiple regression analysis was performed in each group. The model for patients aged 60-69 yr was best for accuracy of prediction and explained 76% of variation for discharge FIM total score. The equation: (expected discharge FIM total score) = 111.88 + 0.08 x (the type of stroke) - 0.11 x (age) + 0.81 x (admission FIM total score) - 0.12 x (onset to rehabilitation admission interval), R = 0.87, R2 = 0.76, P < 0.0001. The type of stroke = 1 for cerebral infarction and 0 otherwise. Length of rehabilitation stay is not selected as a predictor. The stratification of patients by age is useful to determine predictors of function at discharge for stroke outcome and to improve their accuracy of prediction.