How health systems are easing patient fears surrounding AI

As artificial intelligence (AI) becomes more embedded into the healthcare system, hospital executives are strategizing ways to help patients feel comfortable with the technology when it is used as part of their care.

Gaining patient trust may be a difficult task, though, as three out of four respondents of a recent survey from clinical data management company Carta Healthcare said they don’t trust AI in a healthcare setting. However, the survey also found that a lack of education on how AI is used plays a big role in patient skepticism.

NewYork–Presbyterian, one of the largest health systems in New York, has prioritized patient education as a way to alleviate the fears surrounding AI.

“We want to be transparent about where we are and aren’t using technology,” Peter Fleischut, group SVP and chief transformation and information officer at the health system, previously told Healthcare Brew. “We need to be able to communicate with our patients, and we need to also make it a part of the training that our doctors receive.”

NewYork–Presbyterian—which has around 120 AI initiatives across clinical and nonclinical use cases—is developing strategies to integrate patient education on AI into the health system’s care model, Fleischut said, adding that that includes explaining what clinicians are doing with AI as well as why they’re doing it.

The health system is also engaging patients in testing the technology and getting their feedback before it’s deployed widely, Fleischut said.

Nearby, at New York City’s Mount Sinai, health system leaders promote AI-enabled tools as “augmented intelligence”—an approach that David Reich, president of the Mount Sinai Hospital and Mount Sinai Queens, told Healthcare Brew looks to demystify the technology.

“Augmented intelligence sounds affirming, it sounds supportive, it sounds great,” he said. “Artificial intelligence is, by definition, a little scary.”

Reich said Mount Sinai uses AI to better connect patients with the proper physician or specialist, as well as predict which patients may be at risk for conditions like malnutrition, among other things.

Intermountain Health, based in Salt Lake City, has also prioritized AI in the last couple years, Mona Baset, VP of digital services, told Healthcare Brew. Since 2021, the health system has implemented more than 4,000 automations in clinical and nonclinical areas, she said.

The health system has prioritized creating what it calls an “AI playbook,” which is made up of protocols to vet any new AI technology introduced to the system, according to Baset. The playbook guides the team on how to deliver AI responsibly, and how to make sure there’s no bias in the algorithms, she said.

Intermountain also makes sure that “physicians are always involved” in the decision-making process when using AI for clinical purposes, which gives patients “a lot of confidence,” Baset said.

The health system is in the “very early” stages of strategizing how to have conversations with patients about AI, and it hasn’t yet developed a policy that requires clinicians to disclose to patients when an automation tool is used a part of clinical decision-making. This is because the AI doesn’t “do the work or take the lead,” Baset said.

However, she added that as the technology becomes more advanced, having a disclosure policy may be necessary, saying that talking to patients about AI is “certainly something that [the system has] thought about.”

Shannon Young contributed to this report.