Nudging Toward Precision Medical Education: Linking Diagnostic Exposures to Tailored Learning

The article presents DxMentor, a platform integrated into electronic health records that automatically captures diagnostic exposures of medical students and residents based on ICD-10 codes. Artificial intelligence matches each diagnosis with curator-selected educational materials, practice questions, and simplified summaries of scientific articles from PubMed. Personalized email notifications are sent before morning rounds. From July 2024 to April 2026, the platform processed 32,846 patient encounters from 335 students and 346 residents, delivering 17,340 notifications with 63,754 educational resources. Physician-educators rated 91.5% of diagnosis-resource pairs as relevant. Overall, 78.5% of notifications were opened and 11.3% received at least one click. Residents showed higher engagement than students (open: 80.7% vs. 65.8%; clicks: 12.7% vs. 3.1%). Next steps include linking notifications to learning outcomes and testing alternative strategies and timing.