Wuhan University researchers develop AI to predict emergency room surges

Researchers from Wuhan University and the Renmin Hospital of Wuhan University have developed an artificial intelligence model designed to predict hourly patient surges in emergency departments. By analyzing over 213,000 emergency contacts, the study found that incorporating Social Determinants of Health (SDoH)—such as local demographic and social characteristics—improves prediction accuracy more effectively than relying on meteorological data.

The study, published in npj Digital Medicine, utilized the XGBoost machine learning algorithm to evaluate 22,883 hourly observations. The findings suggest that understanding the specific age structures and disease profiles of different residential areas allows hospitals to better manage limited resources, including staffing and treatment capacity.