Benchmarking Large Language Models on Answering and Explaining Challenging Medical Questions

LLMs have demonstrated impressive performance in answering medical questions, such as passing medical licensing examinations. However, most existing benchmarks rely on board exam questions or general medical questions, falling short in capturing the complexity of realistic clinical cases. Moreover, the lack of reference explanations for answers hampers the evaluation of model explanations, which are crucial to supporting doctors in making complex medical decisions. To address these challenges, we construct two new datasets: JAMA Clinical Challenge and Medbullets. JAMA Clinical Challenge consists of questions based on challenging clinical cases, while Medbullets comprises USMLE Step 2 3 style clinical questions. Both datasets are structured as multiple-choice question-answering tasks, where each question is accompanied by an expert-written explanation. We evaluate four LLMs on the two datasets using various prompts. Experiments demonstrate that our datasets are harder than previous benchmarks. The inconsistency between automatic and human evaluations of model-generated explanations highlights the need to develop new metrics to support future research on explainable medical QA.