Enhancing hematology-oncology education through a structured AI-based board review series: An ASCO Leadership Development Education Scholars initiative.

Abstract

9028 Background: As an ASCO Leadership Development Education Scholars Initiative, structured board review is essential for hematology–oncology fellows to consolidate knowledge, support examination readiness, and strengthen clinical decision-making in an increasingly complex therapeutic landscape. Existing approaches are often fragmented and insufficiently tailored to trainee-specific knowledge gaps. To address this need, we developed a pilot, structured board review series within a single academic fellowship program integrating peer-led teaching, faculty mentorship, and anonymous question-based learning aligned with board examination content and real-world oncology practice. Methods: A longitudinal, fellow-led board review curriculum was implemented within a hematology–oncology fellowship program with subspecialty faculty mentorship to ensure clinical accuracy and guideline-based context. Fellows engaged through weekly anonymous board-style multiple-choice questions distributed via Microsoft Teams polls. Questions were curated from ASCO Self-Evaluation Program materials based on high-yield topics identified from ASCO and ASH in-training examinations, with AI-assisted topic prioritization. Objective knowledge acquisition was assessed using identical pre- and post-session questions. Each topic spanned one month with eight pre-session questions and the same eight questions administered post-session. Across five sessions, five fellows answered eight questions per session, yielding 200 learner responses for both pre- and post-intervention assessments. This initiative was conducted as an educational quality improvement project and did not meet criteria for human subjects research. Results: Educational gains were observed across multiple high-yield oncology domains, including molecular biomarker interpretation, guideline-directed therapy selection, toxicity recognition and management, and familiarity with emerging treatment modalities. In the pre-test assessment, 63% of questions were answered correctly (126 of 200 responses). Following the educational intervention, 98% of questions were answered correctly (196 of 200 responses), representing an absolute improvement of 70 correct responses (35%). Conclusions: This structured, fellow-led, faculty-supported board review initiative represents a feasible and effective model for delivering high-yield oncology education within hematology–oncology fellowship training. The curriculum demonstrated measurable knowledge gains, high trainee engagement, and successful integration of AI-informed topic selection with in-training examination data, highlighting the potential for learner-centered curricula to modernize subspecialty medical education and expand to more fellowship programs.