Abstract Keywords: Artificial Intelligence/Machine Learning, Communication Strategies, Disruptive and Innovative Technologies
Primary Track: AI and Care Outcomes
Programmatic Theme: Emerging Technology and Technical Infrastructure
Generative artificial intelligence (AI) is rapidly transforming patient care, with well over half of health systems now experimenting with generative AI applications. Prompt engineering is foundational to the success of these projects, enabling healthcare professionals to leverage large language models (LLMs) like OpenAI’s GPT for tasks such as summarizing clinical notes and enhancing patient communication. Despite the proliferation of generative AI resources, a significant gap remains in educational content tailored specifically to the healthcare context.
This session will address the need for clinician-focused education by exploring the principles of prompt engineering as they apply to healthcare. Attendees will gain a fundamental understanding of LLMs, including their training and fine-tuning processes, and the impact of prompt design on performance. The session will be updated to include the most relevant changes in 2025, including a deep dive into 3 techniques: few-shot learning, chain of thought, and clear communication.
The session builds on a workshop that debuted at AMIA CIC 2024 and has since been delivered at multiple national conferences and grand rounds. By emphasizing evidence-based strategies and providing practical, healthcare-oriented examples, this updated workshop aims to empower clinicians and informaticists to craft prompts that maximize the utility of generative AI in clinical workflows.
Participants will leave with actionable skills to improve LLM integration in healthcare, addressing one of the largest barriers to effective implementation: the perceived lack of clinician expertise in AI and prompt engineering.