Smart Machines Help Hospitals Prepare for Patient Surges

Emergency rooms are busy places where people come for urgent care. But hospitals are getting more crowded all over the world. This overcrowding makes it hard to give good care to patients. Urban hospitals are especially strained because there aren't enough beds for inpatients. Getting ahead of who will need hospital care could help a lot.

The researchers looked at past patient visits from January 2015 to December 2022. They studied adults who went to emergency departments at two academic hospitals in the McGill University Health Centre. They only included visits where patients were admitted or discharged. Other visit types were left out of the analysis. This gave them a clear picture of what happened at triage.

The team built machine learning models to guess who would be admitted after triage. They used the data that hospitals already collect in a standard way. The study focused on picking the right features, making the model accurate, and explaining why it made its predictions. These are things that were missing from earlier studies.

By predicting admissions early, hospitals can plan better. They can send more staff and beds to the right places. This means shorter wait times for everyone. The goal is to make healthcare more efficient and fair for all patients.