ipopba - stock.adobe.com
Why one clinic lets physicians choose when & how to use ambient AI
Austin Regional Clinic has achieved a 97% AI documentation tool engagement rate among adopting clinicians by using a flexible, peer-led approach, leaders shared.
As health systems scale ambient AI beyond pilot programs, getting clinicians to incorporate the technology into their everyday workflows remains a key implementation challenge.
Despite growing enthusiasm for AI, physicians express concerns. A recent survey found that 77% of physicians believe that clinical deskilling is a significant risk to overusing AI, and 52% have concerns about its quality and reliability in healthcare.
This is why Austin Regional Clinic has taken a clinician-led approach to ambient AI adoption, giving physicians and advanced practice clinicians the flexibility to determine where the technology fits within their documentation workflows. ARC is a physician-owned, multispecialty medical group serving approximately 700,000 patients across 40 locations in Central Texas.
"Everybody, kind of, has a different approach. So because of that, we didn't say, 'Here's Suki, everybody use Suki now,'" said Jacob Childers, M.D., a practicing urgent care physician and physician lead for the clinic's EMR.
Instead, clinicians were encouraged to experiment with Suki and determine when it made sense for their individual workflows.
At the time of the interview, about 126 of the clinic's roughly 450 clinicians were using Suki, and 97% of those users met the vendor's benchmark for high engagement.
Letting clinicians find their own workflow
Austin Regional Clinic recognized early that documentation workflows vary by clinician and specialty. Some clinicians dictate or type, while others rely heavily on templates. Childers, for example, said he is more likely to use ambient AI for complex urgent care encounters and may stick with other documentation methods for simpler visits.
"We're not big on mandating things to people," Childers said. "In general, we try to leave people's individual practice up to their preferences. And so Suki is just an extension of that."
Suki is an ambient AI platform that captures patient-clinician conversations and generates clinical documentation for the EHR. Clinicians can try the Suki technology for three months before committing to it. Most who do try it continue using it, Childers said.
But the adoption was not without its challenges. Some early users found that the generated notes did not match how they preferred to document. Clinicians wanted different levels of detail and different formatting. Several surgeons found the tool did not fit their workflows.
The clinic leaders brought that feedback to Suki, which added more flexibility around note output. Childers shared. Some clinicians who initially stopped using the tool later returned as the technology evolved.
Reducing workflow burdens related to using the technology also helped the clinic encourage adoption of the ambient AI tool. Its IT team configured Epic so Suki-generated information flows into the appropriate portions of the clinical note. Without that backend integration, Childers said clinicians would have to manually move information between systems.
As clinicians gain confidence that ambient AI will note the information they need for documentation, Childers said they can focus more of their attention on the patient.
"Over time, you build trust with this technology that it's going to capture that stuff for you and you don't have to worry about it," Childers said.
Peer experience drives adoption
The experience of physician leaders and other early adopters has had a major impact on the adoption of ambient AI at the organization.
"What's really selling Suki to our clinicians is they're looking at ... the doctor on their left who's getting their charts finished more efficiently, who's very happy with Suki and speaking highly of it," Childers said.
All Austin Regional Clinic clinicians have been told the technology is available, but expansion has largely been driven by clinicians opting in as word spreads. Childers said new clinicians contact IT each week asking to be added, with adoption sometimes spreading through a department after one clinician has a positive experience.
The clinic also monitors utilization to identify clinicians who may be struggling. Rather than assuming low use means resistance, leaders look for individual barriers, such as technical problems or a prior frustrating experience.
"You have to meet them where they are and make it available, help them understand this is a tool that you can use for your practice," Childers said. "But I don't think you mandate."
For organizations facing their own ambient AI adoption challenges, Austin Regional Clinic's experience shows how a physician-led approach can give clinicians room to build trust, adapt their workflows and ultimately drive adoption among their peers.
Elizabeth Stricker, BSN, RN, comes from a nursing and healthcare leadership background, and covers health technology and leadership trends for B2B audiences.
