
Beth Israel Deaconess Medical Center has a problem familiar to many hospitals: Too many patients aren’t being discharged when they’re healthy enough to go home.
It’s an issue with wide-ranging ripple effects: Because of operational holdups, much-needed beds remain filled when they don’t need to be, leading emergency rooms to become congested. Patients must often be sent wherever there’s a free bed, even when it’s not the best ward for them to be treated, which might further delay their discharge.
In a bid to tackle that problem, the Boston hospital is developing an artificial intelligence model designed to suggest a bed assignment for every patient from the get-go. If there isn’t a space for them in the most appropriate area, the model recommends other options.