ARTESIA — Dr. Peter Jewell, a family medicine physician at Artesia General Hospital, isn’t a computer whiz.
“I’m too old,” he said. “I’m 58 years old, and most of the younger kids these days, they can run circles around me with these computers.”
Margaret O’Hara/The New Mexican
And yet, Jewell now uses Microsoft Dragon Copilot, a clinical assistant — powered by artificial intelligence — that records and transcribes his conversations with patients and automatically updates their electronic health records.
That’s right, AI has made its way into New Mexico’s small, rural hospitals. Artesia General Hospital, a 25-bed facility with clinics throughout southeastern New Mexico, is an early adopter of the technology, which Jewell framed as a useful tool that has helped return his focus to patients rather than taking notes from behind a computer.
“There’s a lot of fear in AI these days,” he said. “But honestly, this isn’t really doing my job. It’s helping me take notes.”
Time for patients
As a family medicine practitioner at a small community hospital, Jewell said he sees a little bit of everything.
“Diabetes, hypertension, back pain, headaches, seizures — you name it, I see them,” he said.
As is common for rural hospitals, Jewell said his patients come from all over oil-rich southeastern New Mexico, with some making the trip from Lovington, Ruidoso, Alamogordo and Cloudcroft. On average, he sees between 20 and 25 patients per day.
That patient load translates to a lot of notetaking and recordkeeping for providers. Typically, Jewell said he would spend a couple extra hours finishing his charts after the end of his shift.
Artesia General Hospital adopted the Microsoft Dragon Copilot tool about two years ago, Jewell recalled. Providers use the ambient listening assistant — which is embedded in TruBridge, the technology the hospital uses to manage patients’ electronic health records — to update clinical documentation.
“Hospitals like Artesia prove that advanced technology doesn’t belong only in large urban systems,” David Harse, general manager of patient care at TruBridge, said in a news release.
He added, “We’re proud to support Artesia with a seamless integration of ambient AI directly into the [electronic health record] clinical workflow, helping lift the administrative weight from physicians’ shoulders and giving them back time, energy, and the ability to be fully present with their patients.”
So far, Jewell said the new tech has helped assuage one of patients’ biggest complaints: “When you were staring at the computer like this and not looking them in the eye, they feel like you’re not listening to them. They feel like you’re not paying them attention,” he said.
The ambient AI listener and automatic note-taking function allow Jewell to keep his focus on patients, assess their body language and level of pain, and judge whether they’re understanding his medical recommendations, he said.
“You can’t judge that by not looking at a patient, by not touching a patient, by not listening to their heart. … [It] gives you time to do those kinds of things as well, too,” Jewell said. “If you’re spending all your time buried in your computer, you’ve got to throw that out the window. There’s just not enough time.”
The technology also helps him finish his work day on time. While he used to stick around until 7 or 7:30 p.m., Jewell said most days now end on time, around 5 or 5:30 p.m.
“It’s very important to have that,” he said. “You get burned out really quick as a physician if you don’t have that work/personal life balance.”
‘It’s just listening’
Of course, there is no shortage of critiques about AI, ranging from its lack of regulation to the projected greenhouse gas emissions and the amount of water required to run AI data centers. The HBO medical drama The Pitt recently featured a story line involving ambient listening technology similar to the Microsoft Dragon Copilot that introduces errors into patient records.
Jewell estimated the AI tool is likely more accurate than his notetaking, though he noted he can catch and fix any mistakes — which are typically limited to doctors’ names and missing medical terminology required for billing codes.
Patients are repeatedly warned about the hospital’s adoption of the AI technology. Signs at check-in counters alert patients to the use of AI, Jewell said, while medical assistants confirm patient consent for the system.
In the thousands of patients Jewell sees each year, he could only recall one who asked him not to use the AI companion.
“It’s not making decisions; I’m making all the decisions still. It’s just listening,” Jewell said.