Document Type
Brief Report
Disciplines
Higher Education | Medicine and Health Sciences
Abstract
Background. Artificial Intelligence (AI) offers an opportunity to enhance formative assessment of clinical reasoning skills. Traditionally, these skills have been assessed through workplace-based assessment or simulated patient encounters, both of which are limited by variability in case exposure, faculty resources, scheduling logistics, and cost. AI-based simulated patient encounters provide a scalable solution for convenient, low-stakes practice. At the University of Nebraska Medical Center (UNMC), we piloted an AI-based simulated patient case and conducted a needs assessment to explore current practices and user experiences.
Methods. From June to August 2025, participants trialed an AI-based simulated patient case and received feedback on their clinical reasoning. They then completed a brief 20-question survey assessing acceptability ofthe technology, perceived educational value,potential future applications, and comparisonto existing assessment modalities. Invitationsto participate were sent to medical students(M1-M4), residents (PGY1-PGY3), andattending physicians in internal medicine,family medicine, and emergency medicine atUNMC.
Results. Thirty-four surveys were completed: 29% pre-clinical students (M1-M2), 32% clinical students (M3-M4), 12% residents, and 26% attending physicians. Most respondents found the experience acceptable. In particular, the AI-based clinical reasoning feedback was perceived positively, with most agreeing it identified gaps, provided actionable steps, and was useful for learning. While opinions on replacing traditional Observed Structured Clinical Examinations (OSCEs) with AI-based encounters were mixed, the majority expressed interest in using this technology to improve their clinical reasoning skills.
Conclusion. AI-based simulated patient encounters offer a feasible way to supplement and scale current methods in clinical reasoning assessment. While not intended to replace human encounters, they provide additional opportunities for on-demand, low-pressure practice – an especially valuable tool for busy trainees. Rather than viewing AI as a threat to clinical reasoning, personalized tools should be implemented and studied to strengthen, advance, and guide its development.
DOI
https://doi.org/10.32873/unmc.dc.gmerj.8.1.004
Keywords
Clinical reasoning, artificial intelligence, patient simulation, medical education, formative assessment, needs assessment
Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.
Recommended Citation
Sorrick, A., Liebovitz, D., Vick, S., Parsons, A. S., Locke, B. W., Herskovitz, J., , Rohlfsen, C. AI-Based Simulated Patient Encounters for Clinical Reasoning: A Needs Assessment. Graduate Medical Education Research Journal. 2026 May 30; 8(1).
https://digitalcommons.unmc.edu/gmerj/vol8/iss1/4
