There is a lot of talk about generational AI, but how is it actually being used?
Hollywood may be ecstatic about Barry Keoghan's meteoric rise to wealth, but generative AI is seeing a similar rise in healthcare. But can either of them live up to their potential while avoiding bombshells and tabloid headlines?
Today, AI is on everyone's radar as ViVE 2024 kicks off in earnest at the Los Angeles Convention Center. And while we've heard endless talk over the past year about the promise that large-scale language models and artificial AI will eliminate the healthcare industry's biggest pain points, the talk in the room is: What will it actually do? It's time to show everyone.
“All health systems are making significant investments in AI solutions,” said Harjinder Sandhu, chief technology officer for health and life sciences platforms and solutions at Microsoft, which partners with many health systems to test AI. states. “You have to. You really don't have a choice. But whether you can actually handle it depends on the complexity and risk of your use case.”
“What I'm seeing is quite a bit of alarm,” he added. “[Health systems] We are very tentative about these use cases. For now, we're seeing a lot more confusion than it seems like we're jumping too fast. ”
[Also read: How to Create a Game Plan for Assessing AI Readiness.]
Sandhu, who will be part of a ViVE panel this week discussing AI readiness, said healthcare industry leaders are pushing ahead with projects that carry a lot of risk, especially as the industry is still figuring out governance. He said he was naturally cautious about proceeding. But that doesn't mean they didn't come up with lots of ideas.
“I once sat in on a literally four-hour session where a group put post-it notes on a board with 300 use cases,” he says. “and [they’ll] “Here are all the different areas we want to impact,” he says.
Some of the biggest and best Post-Its focus on using Gen AI to capture conversations and turn them into valuable clinical information within the EHR. It is these practices that are causing the most conflict today, as healthcare organizations across the board grapple with clinician shortages and skyrocketing stress and burnout. Those in the know call this “pajama time.” This is the time clinicians spend each night at home reviewing notes from the day's patient encounters and converting them into information that can be used for care management.
Microsoft and Nuance recently released DAX Copilot, which integrates directly into Epic EHR. This is one of several tools offered by many AI vendors for specific pain points.
“This is probably one of the fastest growing products we've ever seen in terms of how quickly doctors are taking notice and adopting it,” Sandhu said of the overall market. “We are starting to see a major shift in the way physicians view their work and work-life balance.”
This particular tool captures conversations immediately after they occur and provides notes to clinicians immediately after meeting with a patient while the conversations are still fresh. You can review these notes for accuracy before submitting them to your EHR.
[Also read: Study: Ambient AI Scribes Are Good, But Not Yet Ready For Pime Time.]
This review process is very important. Since this technology is still relatively new, there is still the potential for mistakes to be made. As such, the use of AI in healthcare currently requires a “human stakeholder” to review and approve the final product.
“You have to approach [AI tools] There is always doubt,” Sandhu points out. “Be a little skeptical about what they make and double-check and triple-check.”
However, the benefits are great. Tools that can be easily integrated into a clinician's workflow and reduce translational tasks, especially during nights and weekends, can take time away from the computer and put it back where it belongs: with patients and families. Sandhu said these tools should see the highest adoption rate over the next year or two, especially as clinicians test them and vendors work to fine-tune their processes. It's worth noting that gen AI is designed to learn, so that the tool learns the habits and language of the clinician and becomes better at transcribing.
Beyond that, Sandu says, healthcare decision makers are eager to apply generational AI to another important problem: nursing workflow. Nurses are struggling just as much, if not more, than other health care providers, and she needs AI to relieve that overload and get it back in front of patients. But they also need technology designed for them.
[Also read: How to Create AI Programs With Clinical ROI.]
Beyond that, Sandhu sees a growing market for generational AI technologies that can capture and, more importantly, analyze conversations. Consider tools to study the signs of mental health distress that patients encounter. This allows specialists and even primary care providers to identify patients who need help just by the way they talk to someone.
“We’re kind of in a new era of AI literacy,” he says.
So for now, as ViVE attendees lounge in the crisp California sun, much of the conversation is about what's currently happening on stage and how it's performing in the marketplace. , and we don't really talk about the next big thing or the future. A big hit. It's good to see where this goes, but there is a view among executives that new ideas need to show their ROI now rather than later. Healthcare needs help now.
Eric Wicklund is an associate content manager and senior editor for Innovation, Technology, and Pharma for HealthLeaders.
