Skip to main navigation menu Skip to main content Skip to site footer

Abstracts & Posters

Vol. 1 No. S1 (2025): Special Issue: 2025 Geisel Research Poster Night

Clinical Disclosure Training With an AI Patient Actor

Submitted
21 December 2025
Published
27-12-2025

Abstract

Disclosure of medical errors is an essential yet under-taught clinical skill. This pilot project evaluated a compact, scalable curriculum combining self-directed learning modules with simulated encounters using Dartmouth’s AI Patient Actor 2.0. Eight second-year medical students completed a baseline AI encounter, video-based pre-work, a didactic lecture, and a second AI scenario. Objective rubric-based scoring improved by an average of 7%, with increases up to 20%, and no learners demonstrating decreased performance. Students rated the AI tool as more natural and effective than earlier models, and unanimously agreed the curriculum improved their confidence with harm-communication skills. Qualitative feedback supported feasibility for integration into On Doctoring or clerkship transitions. Future iterations will refine the rubric and expand testing to larger cohorts. This study demonstrates that AI-based simulation can enhance communication training while reducing the resource burden associated with standardized patients.

References

  1. References are available on the poster PDF.