Generative AI Support for Faculty During Live Simulation

Read full text Summarise with VIRAT
Authors: David Power
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Delivering high-fidelity clinical simulation places substantial cognitive demands on faculty, particularly during emergency scenarios requiring real-time clinical oversight, scenario management, and preparation for structured debriefing. While generative artificial intelligence (AI) has been increasingly explored in medical education, most applications are learner-facing or used post hoc. This short report describes the development and pilot deployment of a faculty-facing generative AI assistant designed to support live simulation delivery and debriefing. The system integrates a large language model accessed via API and operates in two modes: (1) real-time clinical support during simulation and (2) structured debrief scaffolding aligned with the PEARLS framework. Safety features include grounding to allowlisted clinical sources and explicit human-in-the-loop control. The tool was implemented during emergency simulation courses for newly qualified doctors and evaluated as a feasibility and usability pilot. Faculty reported that the system was feasible to use without disrupting scenario flow and was particularly beneficial for less experienced facilitators, supporting clinical confirmation, escalation prompts, and structured debriefing. These findings suggest that generative AI may have a practical role in supporting faculty cognition during live simulation, warranting further evaluation in larger and more formal studies.

Fields of Study & Subjects

Simulation-Based Education in Healthcare Artificial Intelligence in Healthcare and Education Innovations in Medical Education

Publication details

Journal / source
EdArXiv (OSF Preprints)
Publication date
27 March 2026
Language
EN
Document type
Preprint