Abstract
Large Language Models (LLMs) like ChatGPT, Gemini, and Claude gain traction in healthcare simulation; this paper offers simulationists a practical guide to effective prompt design. Grounded in a structured literature review and iterative prompt testing, this paper proposes best practices for developing calibrated prompts, explores various prompt types and techniques with use cases, and addresses the challenges, including ethical considerations for using LLMs in healthcare simulation. This guide helps bridge the knowledge gap for simulationists on LLM use in simulation-based education, offering tailored guidance on prompt design. Examples were created through iterative testing to ensure alignment with simulation objectives, covering use cases such as clinical scenario development, OSCE station creation, simulated person scripting, and debriefing facilitation. These use cases provide easy-to-apply methods to enhance realism, engagement, and educational alignment in simulations. Key challenges associated with LLM integration, including bias, privacy concerns, hallucinations, lack of transparency, and the need for robust oversight and evaluation, are discussed alongside ethical considerations unique to healthcare education. Recommendations are provided to help simulationists craft prompts that align with educational objectives while mitigating these challenges. By offering these insights, this paper contributes valuable, timely knowledge for simulationists seeking to leverage generative AI's capabilities in healthcare education responsibly.