What's Your Problem (List)? Using AI to Augment Clinical Problem-Solving Education in Morning Report

Artificial intelligence (AI) and large language models (LLMs) like ChatGPT have garnered significant interest in medical education because of their ability to pass clinical licensure examinations and solve complex clinicopathological challenges. Chatbots have also demonstrated proficiency in crucial clinical problem-solving skills fundamental to a trainee's development: generating prioritized problem lists and differential diagnoses. AI use by trainees has outpaced educator efforts to integrate it into clinical curricula. As a case-based problem-solving exercise, morning report presents a key venue to teach trainees practical applications of AI. By controlling the specificity of their input, trainees can use AI to improve potentially biased problem-solving approaches while improving their clinical knowledge. Contextualizing this discussion in familiar tasks like building a problem list or differential diagnosis would give trainees a starting place for building new AI-based skills. This workshop will equip rising chiefs with a systematic approach to integrate AI into morning report sessions. Participants will learn the benefits of utilizing AI for clinical problem-solving education and will gain practical experience creating medical prompts within AI to elicit desired responses. The workshop will be framed around common scenarios in morning report (e.g. building a problem list) so that chief residents can easily adapt AI-teaching to fit their needs. Participants will leave with valuable strategies for effectively using AI in their morning report sessions, supported by a handout featuring example scenarios and best practices for handling each situation.