Variation in Large Language Model Recommendations in Challenging Inpatient Management Scenarios
For nuanced inpatient management questions, widely used LLMs produced inter- and intra-model variation in their recommendations. Clinicians should view LLM output as one perspective among many, consider sampling multiple models or re-prompting, and retain final responsibility for bedside decisions. Prospective studies are needed to test designs that surface model uncertainty and support safe integration of generative AI into complex decision-making.
Authors
- Eric Bressman (ORCID: https://orcid.org/0000-0003-4688-0747)
- S. Ryan Greysen (ORCID: https://orcid.org/0000-0002-0123-0831)
- Susan N. Landon (ORCID: https://orcid.org/0000-0003-4026-3668)
- Thomas Savage (ORCID: https://orcid.org/0000-0003-4828-5802)
Institutions
- Philadelphia VA Medical Center (US)
- University of Pennsylvania Health System (US)
- University of Pennsylvania (US)
Publication Details
- Journal
- Journal of General Internal Medicine
- Published
- 2025-10-07
- DOI
- https://doi.org/10.1007/s11606-025-09888-7
- Citations
- 8
- Primary Topic
- Artificial Intelligence in Healthcare and Education
- Type
- article
- Field-Weighted Citation Impact
- 5.23