Assessment of Physician Preferences for Large Language Model–Generated Responses Across Geographic Regions and Clinical Experience Levels: Preliminary Survey Study
This exploratory study demonstrates that physicians across diverse regions and experience levels generally prefer LLM-generated responses to human-authored ones. The consistency of this finding across continents and practice durations underscores growing professional acceptance of AI as a viable tool for patient communication. These results suggest that modern LLMs, particularly GPT-4.0, may provide clinically acceptable, contextually relevant, and user-trusted health information, with potential to augment physician workflows and patient education.
Authors
- Paa‐Kwesi Blankson (ORCID: https://orcid.org/0000-0003-3726-3501)
- Anny Rodriguez (ORCID: https://orcid.org/0000-0001-5863-1129)
- Tijani Idris Ahmad Oseni (ORCID: https://orcid.org/0000-0001-5301-1983)
- J. S. Brooks (ORCID: https://orcid.org/0009-0003-9642-8203)
- Tye Young Yang
- R. Adams Cowley (ORCID: https://orcid.org/0000-0003-0996-5223)
- Muhammed Y. Idris (ORCID: https://orcid.org/0000-0002-3899-3728)
- Peter Campbell (ORCID: https://orcid.org/0009-0002-7802-7628)
Institutions
- Georgetown University (US)
- MedStar Georgetown University Hospital (US)
- Cathay General Hospital (TW)
- Edo State University Uzairue (NG)
- Morehouse School of Medicine (US)
- Icahn School of Medicine at Mount Sinai (US)
Publication Details
- Journal
- JMIR Formative Research
- Published
- 2025-11-24
- DOI
- https://doi.org/10.2196/82487
- Primary Topic
- Artificial Intelligence in Healthcare and Education
- Type
- article
- Field-Weighted Citation Impact
- 0.00