Dr. Harrison Jackson shows his patient how DAX uses artificial intelligence to transcribe clinical notes during an appointment. Credit: Courtesy of The Ohio State University Wexner Medical Center
There are certain desirables for every healthcare appointment, from efficiency to detailed clinical notes to direct attention from physicians. Artificial intelligence may be able to grant all these wishes and more.
The Ohio State Wexner Medical Center piloted the Microsoft Dragon Ambient eXperience — or DAX — Copilot application back in January; the tool, which uses AI to transcribe physician-client conversations in real-time, can save healthcare providers valuable time by eliminating the need for hand-typed notes, according to Microsoft’s website.
“I think there was a need to bring this tool into healthcare, the tool to decrease the documentation time that it takes a clinician to see a patient and to meet some of those requirements,” said Ravi Tripathi, chief health information officer and sponsor lead for the DAX project.
DAX has been fully implemented at the medical center since July 1, meaning it is available for use by any physician or provider in ambulatory care. Tripathi said there are 294 physicians and advanced practice providers using this application, saving them two to six minutes per patient on average.
DAX has been used on 17,874 patients since January — when the medical center started piloting — and 13,532 additional patients since July — when full implementation was achieved, Tripathi said.
“If you figure last week, we did 2,132 encounters, times, let’s say, three minutes per encounter — we saved 106 hours of work for physicians,” Tripathi said.
When DAX streamlines a visit’s documentation process, it opens up more opportunities for patients to bring up additional concerns without the worry of wasting appointment time, said Dr. Harrison Jackson, clinical assistant professor of general internal medicine and one of the original pilot users for DAX.
“As a primary care provider, I might be dealing with four or five or even 10 medical issues during an appointment, so an extra two minutes is enough time to address another concern,” Jackson said. “We’re able to do more comprehensive visits.”
The medical center has been trying to implement an AI note-taking component for years, but didn’t experience success until 2024, Tripathi said.
Two years ago, Jackson said a different AI-recording product was trialed at the medical center. However, he said the product required 100 patient recordings — which would have taken roughly three weeks to gather — in order to build a model usable for physicians, so the hospital pivoted to using DAX, which was available for immediate use.
“It took about two minutes for me to get access, look at the information and then use it for a clinic visit,” Jackson said.
Notably, Tripathi said DAX creates drafts for physicians to review instead of completely taking over clinical documentation without human interference.
“It’s not meant to replace the documentation that Dr. Jackson does, or the physicians,” Tripathi said. “It’s meant to draft a response, and then it is expected that he reviews that response.”
As more people expose themselves to AI technology — and the idea of its integration into healthcare — an application like DAX will soon be normalized in patient-care settings, Jackson said.
“It will become more routine as seeing a physician type on a computer,” Jackson said.
From physicians having the time to eat lunch for the first time in 10 years to clients having extra time to speak with their physicians, this application has proven to be beneficial on both the physician’s and client’s end, Jackson said.
“My patients love that I’m actually making eye contact with them and spending more time actually interacting with them and less time documenting during the visit,” Jackson said.
