The buzz around artificial intelligence in healthcare can be wearying at times — Providence CFO Greg Hoffman has the inbox to prove it — but he and other health system CFOs are excited about AI's possibilities nonetheless.
"The context is very important when using the buzzword," the CFO of the Renton, Wash.-based system told Becker's.
Mr. Hoffman said he thinks AI has the chance to reduce a lot of administrative tasks. He said they're hearing about studies that have shown physicians and nurses still have 20% to 30% of their time tied up in administrative work. They've been exploring "casual, practical AI" solutions to focus on.
One focus in particular is on in-basket management for physicians, answering questions from patients.
"A small number you can answer automatically because it's something that doesn't require medical opinion," he said. "Or you can take that inbound message, combine it with data that's in the electronic health record and begin to pre-draft responses and it just makes it a lot quicker. Of course, the physician is still going to have to own the message and so forth, but it's opportunities like that that really get us excited."
Joanna Weiss, CFO of Tampa-based Moffitt Cancer Center, said AI "has the potential to really revolutionize the backend work that we do in finance."
"I don't anticipate it being a job-taker, I think it's going to be a job enhancer," she said.
Ms. Weiss has an accounting background and remembers the era of accounting with green bar paper.
"Excel didn't take away the jobs of accountants, it just made us a lot more efficient," she said. "And I think of AI as something similar. It's a tool that will make us more efficient and allow our teams to do the things that they went to school and have been trained for, as opposed to [being] data enters or aggregators of information."
John Beaman, CFO of Roseville, Calif.-based Adventist Health, said they're approaching AI by focusing on the problems they're trying to solve rather than allocating a dedicated budget.
"On the revenue cycle side, we use AI to identify payment and denial trends from payers, especially with prior authorizations," he said. "With healthcare now digitized, AI helps us make better business decisions. On the clinical side, it has helped us identify conditions like sepsis more quickly, literally saving lives."
Robert Broermann, CFO of Norfolk, Va.-based Sentara Healthcare, said his system is taking a similar approach, asking if it's evolved enough to be useful in certain areas, but we haven't set a specific AI spending goal."
"Valid use cases are emerging, and we've already invested in some, though we're still in the early stages," he said. AI excels when processing large amounts of information quickly, such as in the revenue cycle, where clinical documentation and coding often require human effort. This is a perfect AI application, especially in pushing back against insurance denials. At some point, it may even be AI versus AI, with systems exchanging medical records to facilitate fair decisions. On the clinical side, AI can quickly review years of records to inform better diagnoses, something human clinicians don't have time to do. It's still early in the game, but AI shows promise."