GenAI needs guardrails to flourish in healthcare

Left to right: Brian Spisak, independent consultant and research associate at National Preparedness Leadership Initiative at Harvard University; Humberto Quintanar, vice president and chief technology officer at Memorial Healthcare System; Srinivasan Suresh, vice president, chief information officer and chief medical information medical officer at UPMC Children's Hospital of Pittsburgh; and Deb Muro, chief information officer at El Camino Health, speak at the HIMSS24 conference in Orlando Tuesday.

Photo: Jeff Lagasse/Healthcare Finance News

ORLANDO – Generative AI, or GenAI, holds the potential to drastically transform healthcare, and some trends and use cases are beginning to generate excitement among health leaders, who see the promise of improved diagnostics, personalized treatment plans and operational efficiencies.

In a panel discussion at HIMSS24 here Tuesday, "Generative Al in Healthcare: Hype, Integration, and Future Outlook," the excitement was palpable – but was tempered with caution.

"Because we're in a fragile state where this could go one way or another, it requires leaders to be at the top of their game now … because of the threats and opportunities we have," said Brian Spisak, independent consultant and research associate at National Preparedness Leadership Initiative at Harvard University. "We need to understand – who are our people and what are our processes?"

For Humberto Quintanar, vice president and chief technology officer at Memorial Healthcare System, the concern is not so much about the technology itself, but about the data at the core of it.

"What worries me and keeps me up at night is the security of that data," said Quintanar. "Today we are looking at that very carefully, because it's so easy to get that information and just share it internally, but it's also becoming easier to share it externally. 

"Generative AI is going to create a hell of a lot more information than we already have," he said, "because it takes all the data and generates more data. We're going to create so much data … that we're going to be overwhelmed. We won't have the ability to absorb that data."

And yet, despite the reservations, the technology is already creeping into healthcare – and already yielding positive results.

Deb Muro, chief information officer at El Camino Health, sees hope in a use case at her health system that seeks to help doctors avoid working on their notes at night – what Muro dubbed "pajama time." 

"We did a pilot with one of our physicians, and we just turned on 'ambient listening,'" said Muro. "It's a recording of the visit, and it generated a note. And the physician does have to review it, but the physician was really amazed. So I'm hopeful this is really going to make a difference for them."

There's another active use case in radiology, with GenAI tech that can generate notes from X-rays and even pull up images of X-rays from similar patients. Muro estimated this saves radiologists about an hour per day.

"They're thrilled to have this virtual partner, if you will, in their daily work," she said.

In order for these promising results to continue, expand and grow in ethically responsible ways, there need to be guardrails, said Spisak.

"This tech might save us a bunch of time – say 15% of time – but what's going to happen with that gap? Is leadership just going to fill it with something else? They would just be finding another way to burn you out," said Spisak. "You need to build guardrails for not just the technology but the leadership."

According to Muro, everyone in the organization, and especially leadership, needs to be educated on AI's promises as well as its risks. The biggest near-term potential, she said, is in reducing administrative burden and burnout.

"Let's get rid of those things and save time for what human beings need – the compassion and the care," she said. "That's why we got into the profession in the first place."
 

Jeff Lagasse is editor of Healthcare Finance News.
Email: jlagasse@himss.org
Healthcare Finance News is a HIMSS Media publication.