AI May Finally Shorten the ER Wait!

Mount Sinai's AI predicts ER admissions hours early, offering hope to ease delays and hospital strain.

27 Aug 2025

Emergency departments across the US are under growing strain, but new research suggests artificial intelligence could help relieve pressure. A study by Mount Sinai has shown that an AI system can forecast whether patients arriving at emergency rooms will require hospital admission, hours before doctors make that decision.

The model, trained on data from almost 50,000 emergency visits at multiple hospitals, processes routine information gathered at check-in. By identifying likely admissions early, hospitals could prepare beds, adjust staffing levels and shorten delays that often leave patients waiting.

“This is a major step forward,” said a senior Mount Sinai researcher. “Hospitals finally have a tool that can ease bottlenecks and help deliver faster, safer care to patients.”

The system has so far been tested retrospectively and is not yet embedded in clinical workflows. Researchers said the results indicate it is ready for live evaluation, though challenges remain in monitoring accuracy and ensuring equitable treatment across patient groups.

The findings come as emergency rooms contend with record patient volumes, persistent staff shortages and rising financial strain. Forecasting tools such as Mount Sinai’s offer the possibility of optimising resources without additional spending, potentially reducing diversions and improving compliance with insurance requirements.

Healthcare leaders have responded positively, viewing such models as a natural extension of hospital command centre technologies. Unlike previous studies, Mount Sinai’s work tested the AI under real hospital conditions, strengthening the case for future adoption.

While full-scale deployment is still pending, the research underscores the growing role of artificial intelligence in hospital operations, with the prospect of easing congestion and improving care delivery.