Development and validation of a prediction model for home discharge in patients with moderate stroke: The Korean stroke cohort for functioning and rehabilitation study

Background: Previous studies have investigated the predictors for home discharge without considering stroke severity. Objectives: To develop a practical assessment tool that predicts home discharge for moderate stroke patients after subacute rehabilitation therapy in the tertiary hospitals. Methods: Stroke patients with National Institutes of Health Stroke Scale scores of 6 to 13 were included in this prospective cohort study. Various demographic, clinical, and functional factors were analyzed as potential predictive factors. A weighted scoring model was developed through the following three-step process: 1) selection of the factors by logistic regression analyses, 2) development of a weighted scoring model, and 3) validation of the generalizability of the model. Results: The home discharge rate was 51% (n = 372), and the overall mean length of stay of hospitalization was 32.5 days. 1) The Cognitive Functional Independence Measure, 2) the Functional Ambulation Categories, 3) the modified Charlson Comorbidity Index, and 4) marital status were independent predictors of home discharge. The coefficient value for marital status was adjusted to 1 in the scoring system, and the values of the other parameters were proportionally converted to the nearest integer. Possible total scores ranged from 0 to 13 in the model, with a higher score indicating a higher probability of home discharge. With a cutoff point of 7, this model showed 87.0% sensitivity and 86.2% specificity (area under the curve = 0.90). Conclusions: This novel assessment tool can be useful in predicting home discharge after subacute rehabilitation of moderate stroke patients.