Apr 15, 2024 - Health

Gaps emerge in how doctors and patients see AI

Illustration of binary numbers inside of speech bubbles.

Illustration: Shoshana Gordon/Axios

Four in 10 U.S. doctors say they're ready to start using generative AI in their interactions with patients, according to a Wolters Kluwer Health survey shared first with Axios.

Why it matters: Most of those doctors said they'd changed their minds about the technology in the past year, showing a growing acceptance even as patients are still wary of it.

  • A previous patient survey from Wolters Kluwer showed most would be concerned if their doctor was using GenAI to help diagnose them or set treatment.

What they're saying: "What's a little bit interesting is, doctors have traditionally been somewhat reluctant to adopt new technologies," said Peter Bonis, the group's chief medical officer.

What they found: 8 in 10 physicians think GenAI will improve interactions with patients.

  • Over half believe it will save them substantial time by quickly searching medical literature or summarizing patient data from electronic health records.
  • The survey included 100 physicians who work in large hospitals, see patients and use at least one clinical support tool.

However, physicians said they're still being careful about how they proceed.

  • 9 in 10 said they would need to know what data GenAI tools were trained on and that vendors must be transparent about their sourcing.
  • That's a key point of agreement between doctors and patients, Bonis said.
  • They also agree that health care professionals should oversee how these tools are used.

My thought bubble: While it's a relatively small survey, one thing really stood out: Most doctors indicated their organizations haven't fully grappled with GenAI yet.

  • More than a third said their organizations don't have guidelines about how to use it, while almost half (46%) said they didn't know if there were guidelines.
Go deeper