Comparison of Prediction Accuracy Between Models Using Motor and Cognitive FIM Scores and Those Using Individual FIM Items in Multiple Regression Analysis

Key Points

• The study aims to compare the predictive accuracy of models using motor and cognitive FIM scores against those using individual FIM items.
• Performed multiple regression analysis with motor FIM and cognitive FIM as independent variables.
• Compared predictive accuracy of two approaches in stroke patients during rehabilitation.
• Used discharge motor FIM as the dependent variable.
• The model using total motor and cognitive FIM scores showed higher predictive accuracy than using individual FIM items.
• Statistical significance for the improvement in prediction accuracy was achieved, indicating a better fit for the combined scores.