From Scores to Steps: Diagnosing and Improving LLM Performance in Evidence-Based Medical Calculations
This work addresses the need for clinically trustworthy evaluation of LLMs in medical calculations, which is crucial for real-world medical applications, though it is incremental in improving existing methods.
The paper tackled the problem of evaluating LLMs in medical calculations by proposing a step-by-step evaluation pipeline that revealed GPT-4o's accuracy dropped from 62.7% to 43.6%, and introduced MedRaC, a modular agentic pipeline that improved LLM accuracy up to 53.19% without fine-tuning.
Large language models (LLMs) have demonstrated promising performance on medical benchmarks; however, their ability to perform medical calculations, a crucial aspect of clinical decision-making, remains underexplored and poorly evaluated. Existing benchmarks often assess only the final answer with a wide numerical tolerance, overlooking systematic reasoning failures and potentially causing serious clinical misjudgments. In this work, we revisit medical calculation evaluation with a stronger focus on clinical trustworthiness. First, we clean and restructure the MedCalc-Bench dataset and propose a new step-by-step evaluation pipeline that independently assesses formula selection, entity extraction, and arithmetic computation. Under this granular framework, the accuracy of GPT-4o drops from 62.7% to 43.6%, revealing errors masked by prior evaluations. Second, we introduce an automatic error analysis framework that generates structured attribution for each failure mode. Human evaluation confirms its alignment with expert judgment, enabling scalable and explainable diagnostics. Finally, we propose a modular agentic pipeline, MedRaC, that combines retrieval-augmented generation and Python-based code execution. Without any fine-tuning, MedRaC improves the accuracy of different LLMs from 16.35% up to 53.19%. Our work highlights the limitations of current benchmark practices and proposes a more clinically faithful methodology. By enabling transparent and transferable reasoning evaluation, we move closer to making LLM-based systems trustworthy for real-world medical applications.