EdTech & Innovation
Redesigning Clinical Reasoning Education in the Generative AI Era: Five Core Curricular Strategies
Schaye et alAcademic Medicine

Clinical reasoning develops across training and comprises diagnostic reasoning, which classifies a patient's condition to assign a diagnostic label, and management reasoning, which weighs multiple defensible care plans while accounting for patient preferences and contextual constraints. Historically, clinical reasoning curricula have prioritized "in-the-head" cognitive processes—including information gathering, hypothesis generation, problem representation, and differential prioritization—over "out-in-the-world" contextualized processes such as system navigation, tool utilization, and interprofessional care.
Generative artificial intelligence (GAI) is reshaping how, where, and with whom trainees learn to reason, shifting the primary educational unit from the individual learner to the learner-AI dyad. This transition introduces opportunities for clinical upskilling, but also skilling risks across training stages, including never-skilling, mis-skilling, and deskilling.
Drawing on information processing theories and situativity frameworks, the authors recommend five cross-cutting strategies to redesign clinical reasoning curricula: 1. Introduce GAI tools early in the curriculum. 2. Equip clinical faculty to supervise GAI-assisted reasoning. 5. Rethink clinical reasoning assessment methods.
A deliberate, theory-informed integration of these approaches is intended to prepare learners to incorporate GAI into clinical reasoning skillfully and responsibly.
Key Takeaways
- To address risks of trainee deskilling, mis-skilling, and never-skilling within the learner-AI dyad, the authors propose five cross-cutting strategies for clinical reasoning education: early generative AI introduction, reasoning-first workflows, graduated clinical autonomy, faculty supervision training, and updated reasoning assessments.
Read the original sourceAcademic Medicine
