RAG-Coding: Enhancing LLM Medical Coding with Structured External Knowledge
For medical coding professionals and healthcare systems, this work improves automated ICD coding accuracy and compliance by integrating structured external knowledge, though gains over the SOTA PLM-ICD are mixed.
RAG-Coding uses four LLM agents with external knowledge sources to improve ICD-10-CM coding, outperforming LLM baselines by 8-13% micro-F1 and 2-8% macro-F1 on the MDACE dataset, and achieving comparable F1 to the SOTA PLM-ICD method.
We present RAG-Coding, an agentic method for automated ICD-10-CM coding. RAG-Coding orchestrates four large language model (LLM) agents and grounds their coding decisions in external knowledge sources (e.g. the official coding tabular list and guidelines). By retrieving and cross-referencing relevant knowledge in these sources, the agents enhance coding accuracy and ensure clinical compliance. On the MDACE dataset, RAG-Coding outperforms the best LLM-based baseline by 8-13\% in micro-F1 and 2-8\% in macro-F1 across multiple LLM backbones. Compared to the state-of-the-art pretrained language model method, PLM-ICD, RAG-Coding exhibits higher micro recall (+11\%), while PLM-ICD exhibits higher micro precision (+6\%), yielding comparable micro- and macro-F1. Ablations show stepwise gains, highlighting the importance of incorporating external knowledge. We also release MDACE-2025, updating the original dataset with expert re-annotations with the latest 2025 ICD-10-CM guidelines. This update features more fine-grained code labels and enables evaluation against current clinical standards.