How AI could ease burden of hospital discharge summaries

The hospital discharge summary — a document that outlines a patient’s hospital stay for their outpatient providers — can take up a lot of doctors’ time. It needs to comprehensively and succinctly summarize days, sometimes weeks, of medical details.

“One of the main problems in medicine is the amount of information in the system,” said François Grolleau, MD, PhD, a postdoctoral scholar at the Stanford Division of Computational Medicine. “It’s just too much for any human to process, yet that’s what we ask of physicians. They’re overwhelmed.”

But quickly processing and summarizing copious amounts of data is exactly what the artificial intelligence systems known as large language models have been trained to do.

To determine whether AI could safely and effectively ease the burden of writing hospital discharge summaries, Stanford researchers built an AI-enabled agent that was piloted last summer at Stanford Health Care. Their findings, which were published in JAMA Network Open on May 8, showed that the AI-generated summaries were safe, and using the tool was associated with lower providers’ burnout scores.

Hospital discharge summaries are extremely important for a patient’s ongoing care, said April Liang, MD, a clinical assistant professor of medicine and an author on the paper. “You want to get every detail right,” she said, “and really capture the clinical course accurately.”

Liang, who is a specialist in caring for hospitalized patients at Stanford Health Care as well as a clinical informaticist, estimates it takes her 15 to 30 minutes to write a single discharge summary, depending on the complexity of the patient’s case.

Hospital discharge summaries are extremely important for a patient’s ongoing care, said April Liang, MD, a clinical assistant professor of medicine and an author on the paper. “You want to get every detail right,” she said, “and really capture the clinical course accurately.”

Liang, who is a specialist in caring for hospitalized patients at Stanford Health Care as well as a clinical informaticist, estimates it takes her 15 to 30 minutes to write a single discharge summary, depending on the complexity of the patient’s case.

Because there were no comparable tools available when the project began, Stanford researchers built their own in-house hospital discharge summary agent, MedAgentBrief. After the system was tested, it was deployed at the Stanford Health Care patient care unit at Sequoia Hospital, a 24-bed unit staffed by Stanford Medicine hospital-based physicians, known as hospitalists, on Aug. 1, 2025.

During the 10-week pilot, the 11 physicians working on the unit received a secure email containing AI-generated discharge summaries for each of their patients every morning. The format of the summaries, borrowed from a best-practices discharge summary template, was designed by Stanford Medicine hospitalists. It included a one-liner about what brought the patient to the hospital, a high-level overview of their admission and a structured summary for each of the patient’s inpatient diagnoses.

The doctors could ignore the AI-generated summaries, Grolleau said. But, to the researchers’ surprise, they used the tool. “Historically, it’s very hard to deploy technology and have it adopted very quickly, especially in medicine,” said Grolleau, the paper’s corresponding author, a certified anesthesiologist and critical care medicine specialist. “For this tool, that wasn’t the case. There was so much demand.”

Physicians were asked to provide feedback on the AI-generated summaries — namely, whether there were inaccuracies, hallucinations or omissions in the summaries — and their perceived time savings, if applicable.

Feedback on 100 AI-enabled summaries found some omissions (25%) and inaccuracies (20%), but hallucinations were rare (2%). Physicians rated 88 unedited summaries as having no harm potential and 21 as having mild harm potential.

One summary was rated by a physician as likely to cause moderate harm. In that case, the summary noted that a patient was discontinued from intravenous antibiotics and transitioned to oral antibiotics. But it failed to mention that a nine-day course of treatment antibiotics had already been completed, and the oral antibiotics were prophylactic due to an unrelated condition. Independent reviewers determined the summary posed no risk.

No severe harm was reported.

Less exhaustion
While physicians presumed they had a time savings of more than 10 minutes per discharge summary with MedAgentBrief, the researchers determined, from checking time logs in the electronic medical record, that the time savings were modest (three minutes at best) and variable for every user. The discrepancy is likely because revising and checking the AI-generated summaries felt to physicians like less effort than creating the summary from scratch, the researchers said.

Physicians participating in the pilot reported significant reduction in their burnout scores.

Liang said she found the summaries especially helpful — and time saving — for complex patients with long hospital stays whom she met shortly before discharge. In one case, the AI-generated summary highlighted a patient’s anemia, which wasn’t central to their hospitalization but was important to convey to their primary care team.

“I didn’t save too much time,” she said, “but it definitely was very helpful and important that it highlighted these things that I wasn’t fully aware of.”

The pilot ended on Oct. 11, but the MedAgentBrief remained in routine clinical use for another six months. It was turned off in April in advance of the rollout of a similar tool by Stanford Medicine’s electronic medical record supplier.

Grolleau said the researchers plan to develop a system to evaluate AI tools from vendors. Stanford Medicine