Prediction of Independent Walking Ability for Severely Hemiplegic Stroke Patients at Discharge from a Rehabilitation Hospital

Background It is important to predict walking ability for stroke patients, because rehabilitation programs are planned on such predictions. We therefore examined predictive factors that are available before discharge from a rehabilitation hospital. Methods Seventy-two consecutive patients with a first attack of stroke with severe hemiplegia were included in this study. We retrospectively evaluated background factors (age, gender, time from stroke onset, paresis side, and stroke type). Other neurological and physical parameters were collected by means of the modified National Institutes of Health Stroke Scale, the Mini–Mental State Examination, the Trunk Control Test (TCT), and the knee extension strength/body weight ratio on the unaffected side (KES/BW-US) at the time of admission. We divided the patients into 2 groups, the independent group (n = 49) and the dependent group (n = 23), on the basis of the Barthel Index of mobility at the time of discharge. We then compared the 2 groups with respect to the aforementioned parameters. We also performed stepwise discriminant analyses to ascertain which parameters are the best predictors of walking ability at the time of discharge. Results Age, TCT score, and the KES/BW-US ratio were significantly different between the groups. Discriminant analysis revealed that younger age and a higher KES/BW-US ratio were significantly associated with walking ability at discharge, which could be precisely predicted using the following formula: Y = .093 × (age) − 4.316 × (KES/BW-US) − 4.984. Conclusions At the time of admission, age and the KES/BW-US ratio permit the prediction of independent walking ability at the time of discharge. Our formula predicts walking ability with an accuracy of more than 91%.