Interprofessional Clinical Reasoning on Common Health Problems Using AI-Assisted Virtual Patient Cases
AI-supported virtual patient cases foster interprofessional clinical reasoning among medical and APN students by providing personalized feedback.
Duration: December 2025 – November 2026
Status: Ongoing
Educational Level: Tertiary Level
Topic: Artificial Intelligence AI
Keywords: Clinical Reasoning, Virtual Patients, Interprofessional Education, Artificial Intelligence, Medical Education
Initial Situation
Clinical reasoning is a core competence for both medical students and Advanced Practice Nurses (APN), yet existing virtual patient (VP) cases offer only model answers instead of personalised feedback. This limits learning and does not support diagnostic decision-making effectively. Moreover, there is currently no interprofessional approach, although joint clinical reasoning is essential in practice. At the same time, growing student cohorts and new, case-based curricula require scalable digital formats. Prior BeLEARN work showed that LLMs can provide highly individualised feedback on student inputs, addressing a key deficit of VP cases. However, no curriculum of common health problems using LLM-supported mono- and interprofessional VP cases exists. This project fills that gap by developing a 20-topic curriculum and three exemplar LLM-enhanced VP cases.
Objectives
The project aims to create an interprofessional clinical reasoning curriculum covering 20 common health problems and to implement three LLM-enhanced VP cases. These cases will support mono- and interprofessional learning and be evaluated for their effectiveness compared with conventional VP cases.
Method
A 20-topic curriculum is developed, and three existing VP cases are extended with LLM-based feedback that analyses and responds to student free-text reasoning. Separate learning paths for medical students, APNs, and mixed pairs are designed. A mixed-methods study compares learning gains across cohorts using a Key-Feature test and explores learning experiences through focus groups.
Planned Translation
Because the project builds on an established VP format, results can be directly integrated into medical and APN curricula. Data from the study will guide decisions on broader implementation, including interprofessional use. Collaboration between the University of Bern and BFH ensures long-term curricular adoption and supports the expansion of VP-based learning in both programmes. The project improves learning by providing personalised AI feedback and strengthens interprofessional diagnostic skills. It also supports scalable, case-based curricula. Impact will be measured through learning gain comparisons, qualitative analysis of student experiences, and feedback on feasibility and acceptance within both study programmes.
