American Medical Association President Jesse Ehrenfeld has some advice for artificial intelligence developers if they want to see their tech used by doctors nationwide. He explained how docs see the tech — and what they’re pushing for — at the AI in health care summit at the Healthcare Information and Management Systems Society conference in San Diego last week. The AMA recently asked doctors how they felt about using AI in their practices, Ehrenfeld said. The survey revealed their excitement that AI could improve care and ease burdens — but also some trepidation. Cautionary tales: Prior experience with new tech tools has some doctors worried, he said. Clinicians were promised big improvements with the advent of electronic health records — now loathed by many providers. And some tools had problems baked into them that were never fixed, he said. Pulse oximeters, for example, are known to misread blood oxygen levels in people of color but continue to be used. “We’ve already seen too many examples of systems that have had biased training data or a failure to imagine other design flaws that have unintentionally and invisibly caused harm to patients,” he said. To convince doctors to adopt the new technology, these questions must have good answers, he said: — Does it work? — Will I get proper payment? — Will I be liable? — Will it work in my practice? New systems: Ehrenfeld argued that new regulatory and safety systems must be created — including within the Food and Drug Administration — to ensure AI products are reliable and trustworthy. The AMA isn’t sitting back amid the changes. The group is working to form, validate, fund and ultimately influence how AI makes its way into medical practice.
| 
