• Posted March 14, 2025

AI-Written Doctor's Notes? Patients Don't Mind, Study Says

Patients generally don’t mind getting AI-written notes from their doctor’s office, unless they know the note came from a computer program, a new study says.

Patients shown messages written by either AI, otherwise known as artificial intelligence, or a human doc tended to prefer the responses drafted by AI, although overall satisfaction was high for both, researchers reported March 11 in JAMA Network Open.

And that preference only diminished, but did not disappear, when patients were informed that the message had been crafted by AI, researchers said.

The results suggest that letting patients know AI is involved in clinical communications does not greatly reduce confidence in the message, researchers concluded.

“Every health system is grappling with this issue of whether we disclose the use of AI and how,” senior researcher Dr. Ananda Chowdhury, an assistant professor of medicine at Duke University, said in a news release.

“There is a desire to be transparent, and a desire to have satisfied patients. If we disclose AI, what do we lose? That is what our study intended to measure,” Chowdhury added.

For the study, nearly 1,500 patients with the Duke University Medical System were handed one to three surveys comparing AI-generated doctor’s notes to those written by a human physician.

The three topics for the notes were a routine medication refill request, a question about a medication’s side effects, and potential cancer found in an imaging scan.

The human responses came from a team of physicians asked to write realistic responses to each survey scenario, based on how they typically communicate with patients.

ChatGPT generated the AI responses, which were reviewed for accuracy by doctors participating in the study. Minimal changes were required for the AI notes, researchers said.

Patients preferred AI-drafted messages by an average difference of 0.3 points on a 5-point scale for satisfaction, results show. AI notes also scored better on usefulness by 0.3 points, and on empathy toward patients by 0.4 points.

The AI notes tended to be longer, included more details, and seemed more empathetic to the patients than human-drafted messages, researchers said.

However, satisfaction fell for ChatGPT compared to human doctors when patients were told which note came from AI, with AI scoring 0.1 points lower than docs.

“Our study shows us that patients have a slight preference for messages written by AI, even though they are slightly less satisfied when the disclosure informs them that AI was involved,” lead researcher Dr. Joanna Cavalier, an assistant professor of medicine at Duke University School of Medicine.

Regardless of the author, more than 75% of patients were satisfied with the response they received, results show -- even if they knew it came from an AI.

In addition, the seriousness of the message being conveyed didn't have any impact on patients' preferences when they were told AI drafted the note.

“While we hypothesized that patients would be more accepting of messages they are told come from AI via disclosure in less serious topics, such as a medication refill, there was no interaction,” researchers wrote.

In other words, people were as fine with AI drafting a medication refill message as they were with a note discussing potential cancer found on imaging.

“These findings are particularly important in the context of research showing that patients have higher satisfaction when they can connect electronically with their clinicians,” Chowdhury said.

“At the same time, clinicians express burnout when their in-basket is full, making the use of automated tools highly attractive to ease that burden,” he said. “Ultimately these findings give us confidence to use technologies like this to potentially help our clinicians reduce burnout, while still doing the right thing and telling our patients when we use AI.”

More information

The Mayo Clinic has more on AI in health care.

SOURCE: Duke University, news release, March 11, 2025