Predicting the Length of Stay of Patients in the Emergency Department of the Hospital using Machine Learning Models

The length of patient stay in the emergency department is crucial for optimizing resource allocation, reducing costs, and enhancing operations.

This study aimed to predict patients’ length of stay in the emergency department using machine learning models.

This retrospective cohort study collected data from patients referred to the emergency room of a selected hospital in Tehran, including vital signs, diagnoses, and demographic information such as age and gender, during December 2022. After data preparation, ensemble models—Random Forest, Light GBM, Cat Boost, and Ada Boost—were employed to predict patients’ length of stay.

The study found that vital signs, age, and Emergency Severity Index level 1 significantly influence patient length of stay. The Cat Boost model, with an accuracy of 0.87, precision of 0.91, recall of 0.83, and F1-score of 0.87, outperformed other models in predictive performance.

This study demonstrated that ensemble models effectively predict emergency room patient length of stay, with Boosting methods outperforming Bagging methods.