Palo Alto, Calif.-based Stanford Health Care piloted an in-house AI agent that generates hospital discharge summaries, finding it reduced physician burnout.
Researchers at Stanford (Calif.) Medicine built the tool, calling it MedAgentBrief, and deployed it at Sequoia Hospital in Redwood City, Calif. Over 10 weeks, 11 hospitalist physicians received AI-generated discharge summaries for each patient each morning via secure email. Physicians could use, edit or ignore the drafts.
Adoption was higher than expected. “Historically, it’s very hard to deploy technology and have it adopted very quickly, especially in medicine,” said François Grolleau, MD, PhD, a postdoctoral scholar at Stanford’s Division of Computational Medicine and the paper’s corresponding author, in a June 8 news release. “For this tool, that wasn’t the case.”
Physician review of 100 AI-generated summaries identified omissions in 25% of cases and inaccuracies in 20%, but hallucinations were rare at 2%, per the May study in JAMA Network Open. One summary was rated as posing potential moderate harm; independent reviewers determined it posed no actual risk.
Actual time savings were modest — about 3 minutes per summary — though physicians estimated saving more than 10 minutes. Researchers attributed the gap to the cognitive difference between editing a draft and writing from scratch. Physicians participating in the pilot reported meaningful reductions in burnout scores.
At the Becker's 11th Annual IT + Revenue Cycle Conference: The Future of AI & Digital Health, taking place September 14–17 in Chicago, healthcare executives and digital leaders from across the country will come together to explore how AI, interoperability, cybersecurity, and revenue cycle innovation are transforming care delivery, strengthening financial performance, and driving the next era of digital health. Apply for complimentary registration now.