Healthcare AI Success Starts with a Bias Toward Action

The potential applications of AI are endless; there are so many solutions in the marketplace that health care organizations can get “paralyzed,” according to Kim Lynch, Founder and CEO at Metis Health Technologies. She advises organizations not to debate endless what to do, but to “try something” with “a bias toward action.” Jeffrey Nelson, Enterprise Chief Information Officer at Mosaic Health, similarly says to “try and fail,” but “move the ball forward.”

Of course, nobody advises just throwing money at fancy technologies. In our recent interview at the Navina Ascend user conference, Lynch and Nelson along with Dr. Ronen Gordon, Vice President of Medical Product at Navina, lay out the real benefits of AI.

Lynch gives two basic criteria that a technology solution should meet: it has to solve a real problem, and the organization has to see a way to develop a return on investment (ROI) from it. But Lynch and Nelson point out that the organization has to be creative in defining ROI.

The revenue cycle, according to Nelson, makes it hard within a year to determine whether a product has generated an ROI, and organizations must also look at “soft ROI” such as effects on patient satisfaction, doctor satisfaction, and burnout. Lynch says to “throw out” the 12-month ROI measurement, and look over a period of time for continuous improvement.

How is AI being used now? Gordon claims that 80% of physicians are using it, but mostly for search and summarization. He predicts that AI will evolve to start taking action, so the doctor can focus on working with the patient. The reality is that most healthcare organizations are not at the transformative stage where it alters the way the organization works, but we are heading there quickly.

Lynch mentions some areas of potential for AI: risk capture, organizing information that otherwise would take a long time to find, and capturing accurately “what’s going on” with a patient. Nelson also stresses having a longitudinal view of the patient.

Much of our interview with these experts is devoted to the process of choosing and deploying an AI solution. Nelson defines a three-part framework used at Mosaic Health: a physician informatics committee, a governance committee, and pilots.

Pilots are particularly important, because they show what a tool is capable of accomplishing, and because the organization can work through common problems such as bias and hallucinations. The clinicians must be involved early in the pilot, so that they can improve the product and then become champions if it’s effective. When pilots are unpopular, Nelson believes, it’s because people can’t see what the tool will accomplish.

Winning over clinicians is also a key task. Gordon says that important factors are explainability (seeing the data behind AI output) and seeing the product improve from physician feedback.

Lynch calls clinicians “the real edge of innovation” and says they are looking forward to adopting AI. This contrasts with EHRs, where she didn’t see clinician enthusiasm at their adoption. Overall, though, Lynch and Gordon are seeing clinicians take the lead in exploring tech opportunities, something that was rare a decade ago.

Check out our interview with Navina, Metis Health Technologies, and Mosaic Health to learn more about their experience implementing AI in healthcare.

Learn more about Mosaic Health: https://mosaichealth.com/

Learn more about Metis Health Technologies: https://www.metishealthtechnologies.com/

Learn more about Navina: https://www.navina.ai/

Listen and subscribe to the Healthcare IT Today Interviews Podcast to hear all the latest insights from experts in healthcare IT.

And for an exclusive look at our top stories, subscribe to our newsletter and YouTube.

Tell us what you think. Contact us here or on Twitter at @hcitoday. And if you’re interested in advertising with us, check out our various advertising packages and request our Media Kit.

Navina is a proud sponsor of Healthcare Scene.