When it comes to adoption among physicians, augmented intelligence (AI) has definitively turned the corner. According to findings from a survey issued by the AMA’s Center for Digital Health and AI, 81 percent of physicians are using the tools, making a significant uptick from recent years.

One of the key factors? Usability, according to Margaret Lozovatsky, MD. “Unlike in the past, when the perception was that the technology wasn’t sophisticated enough to be an enabler, now it feels like an enabler,” she said. “We’re finally starting to see clinicians say, ‘I can’t practice medicine without this.’”

Margaret Lozovatsky, MD

However, there is still reluctance around adopting AI, which leaders must be aware of. “Physicians need to trust these technologies,” she noted. During a recent interview, Lozovatsky, who joined the AMA as VP of Digital Health in early 2024, shared insights on how health systems leaders can work to build trust, which will become increasingly critical as AI initiatives continue to dominate strategic discussions. 

“We’re seeing such high rates of adoption,” whether it’s for summarization, predictive models, or other uses. “This is unprecedented for clinical technology in my experience,” she said. “Clinicians like it because it decreases their workload and their cognitive load. It’s easy to use things that are helpful.”

No ‘side projects’

What isn’t so easy is managing the pressure physicians face when it comes to patient safety. “There’s a lot of caution in the clinical environment to ensure that these tools are being rolled out safely,” she said. “We need to continue to build that trust.”

And while there’s no blueprint that will work across every organization, there are core elements that can help spur confidence and foster adoption.

The first is ensuring clinicians have a seat at the table. “If you have the right clinicians that are involved in the development and design of these technologies, it becomes integrated into their workflow, and you get buy-in,” she noted. Doing so can also help ensure AI initiatives are aligned with the organization’s strategic objectives, rather than being a ‘side project.’ “It’s about understanding what problem we’re trying to solve, and what’s the roadmap for solving that problem.”

Governance matters

That’s where governance comes in, noted Lozovatsky. At Novant Health, where she served as CHIO, the strategy was to create a “leadership triad” in which a physician, nurse, and informaticist worked together to identify and solve problems. One problem was the ever-increasing demand for tools, which she attributed to a lack of awareness around which functionalities are already available. 

“To me, that means that we’ve failed in the design perhaps, but also in the education and the training that was not intuitive.” Additionally, “there was concern that if something goes wrong, there’s not a way to report this issue,” she said. “And that if there were changes to clinical guidelines and workflows, there wasn’t a robust process to monitor the output.”

These issues, as she learned, were hardly unique in healthcare, which is part of what prompted her to join the AMA – an organization she had been involved with for nearly a decade as an advisor. When they reached out about a possible opening, Lozovatsky welcomed the opportunity to amplify the voice of physician leaders, who have traditionally been positioned “at the elbow helping people figure out which buttons to push,” she noted. 

That, however, is changing. In recent years, the CMIO/CHIO has evolved into an executive leader who can help drive clinical technology implementation. Having organizations like the AMA in their corner can “support the clinician voice,” and serve as a key advocate.

The next frontier

That support will become increasingly crucial as augmented intelligence continues to shape strategic planning. Once adoption has reached critical mass, the next frontier is to determine “how to support the next phase of technologies,” which can be challenging given the number of new products hitting the market.”

For leaders, it means cutting through the noise and figuring out how to integrate AI “in a thoughtful way so we can impact both our clinicians and our patients in a positive manner,” she noted. “That’s the endpoint we’re all after.”

And although many organizations have reported improvements after deployed AI tools – particularly in decreasing cognitive burden and boosting satisfaction – there will always be hesitancy, Lozovatsky said, urging leaders not to push too hard.

“Instead of creating an environment where we judge our clinicians for their cautious approach, we should listen to the reasons for the caution and partner with them to address their concerns so that we don’t inadvertently create patient safety issues,” she said. “We need to think about these issues as we continue to drive for AI utilization, adoption, and integration.

“Ultimately, we’re there to care for patients. And so, we need to make sure that whatever we do has a positive impact on patient outcomes. That has to be the North Star.”