Evaluating the efficacy of chatGPT in near-peer simulation for resident doctors in the emergency department

Abstract

Introduction: Near-peer teaching is gaining popularity as a newer teaching tool, as it improves the learner’s comprehension, targets the right audience and promotes familiarity with the clinical situation and enhances critical thinking.(1) This study was initiated to evaluate the efficacy of chatGPT in near-peer simulation using AI-enabled scenarios in the residency training programme of an emergency department. (2) Methods: ChatGPT an LLM was asked to generate clinical scenarios as per prompts given to evaluate its efficacy in generating real-time and realistic scenarios with information on stepwise approach and critical treatment decisions for the patient. Results: In our study, ChatGPT was able to successfully generate real-time and realistic scenarios based on the prompts given with detailed treatment approaches and critical decisions for all patients/scenarios given, serving as a successful tool to consider in near-peer education using simulation enabled by AI. CONCLUSION: Near-peer simulation training was found to be a valuable method of teaching residents for increasing hands-on experience, skill assessment, confidence in diagnosis and practical thinking. Integration of AI into near-peer simulations aids in creating a wider range of scenarios with prompt treatment decisions.

Ähnliche Arbeiten