Improving hospital bed management with AI

By | 2 minute read | August 16, 2018

doctors, nurses, and patients in hospital

My company, Glintt, based in Portugal, is an IT solutions integrator with deep healthcare expertise. We regularly develop solutions that can reduce healthcare costs, increase efficiencies and improve the patient experience.

Working with IBM, we recently developed WiseWard, a solution that helps achieve these goals using predictive analytics. The solution helps streamline hospital bed assignment decisions by predicting when patients will be discharged to free up bed space.

We plan to make WiseWard even more useful by integrating IBM Watson technology and extending the solution to related departments to further tune the predictions.

Flying blind in a critical process

Quality care depends on having adequate bed space and the ability to move patients through hour by hour and day by day. No patient wants to wait hours for a bed, or have a surgery canceled for lack of bed space. And hospital revenues depend on efficient utilization of hospital resources.

Bed managers at our client hospital had tools that told them which patient was in which bed, but they couldn’t predict the number and time of expected discharges. Relying on tedious methods such as phone calls and paper lists, the managers could guess just a day or two ahead, but in this critical aspect of workflow, admissions and patient care, they were mostly flying blind. You can imagine the stress they felt.

Predicting bed availability

By combining administrative data about bed usage with clinical data about the patients in those beds, WiseWard can automatically predict discharges and suggest actions five to seven days ahead. This is no easy task, given complex variables that include ward gender, surgery schedules, patient age and condition, type of bed needed, number of beds and nursing schedules.

The solution increases bed manager productivity by 30 to 50 percent and reducing stress. More accurate scheduling means better utilization of the hospital’s resources and higher revenues – no need to cancel surgeries for lack of beds. And a smoother workflow improves the patient experience.

Adding AI intelligence

The current version relies on analytical and business rules technology for insights, but we are in the process of integrating Watson artificial intelligence. With Watson, WiseWard should be able to understand natural language notes written by doctors and nurses, adding to its clinical intelligence, and Watson’s self-learning capabilities can make the solution smarter over time. We’re also exploring extending WiseWard to the operating theater, where Watson’s capacity to absorb surgical information can help predict the need for post-surgery beds.

We’re excited that our solution can help hospital professionals make better decisions. It’s a terrific feeling when people thank us for making their jobs easier and helping them in their care for patients.

Watch Glintt executives discuss key challenges in healthcare today: