Objective：We analyzed whether activities of daily living (ADL) had an additive effect on the discharge destination of hospitalized older patients with internal medicine.Methods：Of the 691 patients hospitalized for medical illness aged 65 years or older who received physical therapy during this study period, 186 patients were included in the analysis. The main outcome was the discharge destination. Participants were categorized in the home discharge group and other institution group. The Barthel Index (BI) was used to assess the ADL at the first physical therapy session. Multivariate logistic regression analysis was used to estimate the influence of the BI on determining the discharge destination. The cut-off point of the BI score was evaluated using a receiver operating characteristic curve.Results：During the follow-up period, 17 participants (9.1%) could not be discharged to their homes. Logistic regression analysis showed that the BI influenced discharge destinations (odds ratio：1.54；95% confidence intervals：1.23-1.89). The cut-off point of the BI score for determining home discharge was 72.5 (sensitivity, 0.80；specificity, 0.94；area under the curve, 0.94；positive predictive value：0.99, negative predictive value：0.32).Conclusions：These findings suggest that the BI is a useful predictor for determining the potential destination of hospitalized older patients following discharge. However, the results of this study have limitations such as a low negative predictive value and a limited number of subjects.

Objective：We analyzed whether activities of daily living (ADL) had an additive effect on the discharge destination of hospitalized older patients with internal medicine.Methods：Of the 691 patients hospitalized for medical illness aged 65 years or older who received physical therapy during this study period, 186 patients were included in the analysis. The main outcome was the discharge destination. Participants were categorized in the home discharge group and other institution group. The Barthel Index (BI) was used to assess the ADL at the first physical therapy session. Multivariate logistic regression analysis was used to estimate the influence of the BI on determining the discharge destination. The cut-off point of the BI score was evaluated using a receiver operating characteristic curve.Results：During the follow-up period, 17 participants (9.1%) could not be discharged to their homes. Logistic regression analysis showed that the BI influenced discharge destinations (odds ratio：1.54；95% confidence intervals：1.23-1.89). The cut-off point of the BI score for determining home discharge was 72.5 (sensitivity, 0.80；specificity, 0.94；area under the curve, 0.94；positive predictive value：0.99, negative predictive value：0.32).Conclusions：These findings suggest that the BI is a useful predictor for determining the potential destination of hospitalized older patients following discharge. However, the results of this study have limitations such as a low negative predictive value and a limited number of subjects.