AI-powered system comes to CHI Franciscan

Setting schedules at CHI Franciscan Health is about to look a lot more like rocket science.

"Mission Control" is now 18 screens filled with constantly-updating data about patients, medical staff, and procedures.  The system is intended to help hospital staff figure out where doctors and nurses are most needed at any given time, when patients are scheduled to go home, and what specific care each patient is likely to need.

"Our goal is to be able to look forward into the future," explains Dr. Jessica Schlicher.  She's the medical director for Mission Control, and says the system is designed to get better as CHI Franciscan uses it.

"When you talk about AI, we're simply talking about algorithms that can learn." Schlicher says. "The algorithm that we use today to care for patients will be better in one week; will be better in one year; because it will tell us: 'well we predicted there would be 45 ICU patients at St. Joe's on this date, but there were 46, so we're going to predict next year a slightly different number based on the factors in the algorithm.'"

If mentions of artificial intelligence (AI) get you thinking 'Skynet' - don't worry.  Schlicher says Mission Control is just a tool that helps human staffers streamline CHI Franciscan's inner workings.

"A great example would be if we know we're transferring an ICU patient... we don't have an ICU bed. We might have a patient sitting in the ICU who does not need ICU care. So if we can put that patient within the place where they would get the best care, making space for the patient who's coming in, and then expediting discharges to get patients home sooner - that's the best care for every patient involved."

The system is also intended to keep everyone on the same page about which tests are needed, and hopefully minimize administrative delays.

Of course, Mission Control does bring a lot of very sensitive medical information into one central location.

CHI Franciscan says it's keeping all of that data as secure as possible, and limiting what human staffers can see to a need-to-know basis.