Editor's note: This story has been corrected to reflect that the AI challenge is a prize competition.
The federal government is launching a challenge competition to explore the use of artificial intelligence to predict health outcomes and improve health care delivery.
Adam Boehler, director of the Center for Medicare & Medicaid Innovation (CMMI)—the innovation arm of the Centers for Medicare & Medicaid Services—said the federal agency is launching the Artificial Intelligence Health Outcomes Challenge.
The challenge, announced by Boehler during a media briefing at the Healthcare Information and Management Systems Society's (HIMSS) annual conference and exhibition, is being launched in partnership with the American Academy of Family Physicians.
The challenge, a prize competition, is expected to be rolled out in the next few weeks. Winners of the challenge will receive a significant monetary award, he said.
Boehler, also the CMS deputy administrator, said the goal of the initiative was to drive the private sector to develop AI tools that could better predict health outcomes to improve care quality. Better predictive power can help to identify the onset of dangerous conditions such as septic shock so physicians can intervene earlier, he said.
Last year, Google, in collaboration with the University of California San Francisco, released a study that found its AI-based software was more effective at predicting patient outcomes than other methods currently available.
“We want to get feedback from private industry on how we can redefine quality in a way that minimizes physician burden and predicts quality a lot better,” Boehler said.
CMMI is interested in exploring how AI can be applied to preventive care and incorporated into both current and new payment and service delivery models, he said. “Medicine is treated too much like an art, and it’s a science. I think you’ll see more on the prevention side to get us to a system where we can stop disease before it happens,” he said.
The aim of the challenge is to spur private sector innovation. “Our job is not to redefine AI for everybody; it’s to popularize it, provide data and create incentive. We can get people 60, 70, 80 percent there and then let the private market take it to 99 percent,” Boehler said.
The CMMI director also hinted that the agency might incorporate the interoperability requirements in newly proposed CMS rule into new CMMI voluntary payment and service delivery models. In that proposed rule, CMS is suggesting that Medicaid, the Children’s Health Insurance Program, Medicare Advantage plans, and Qualified Health Plans be required to make enrollee data immediately accessible by 2020.
“We want patients and their delegates to get free access to electronic, easy-to-use information within 24 hours. There’s a lot of ways to get there, and within CMMI, you can say look at our models and drive that sooner. These are advanced models and advanced participants, so that is something that we’re going to looking at closely,” he said.