RAG-Coding: Enhancing LLM Medical Coding with Structured External Knowledge
Quick Answer
RAG-Coding enhances ICD-10-CM coding accuracy by 8-13% in micro-F1 and 2-8% in macro-F1 using four LLM agents grounded in external knowledge.
Quick Take
It outperforms the PLM-ICD model in micro recall by 11%, while releasing the updated MDACE-2025 dataset with expert re-annotations for current clinical standards.
Key Points
- -Coding orchestrates four agents for automated ICD-10-CM coding.
- Achieves 8-13% improvement in micro-F1 on the MDACE dataset.
- Outperforms PLM-ICD in micro recall by 11%, while PLM-ICD has higher precision.
- Releases MDACE-2025 dataset with expert re-annotations for 2025 ICD-10-CM guidelines.
- Demonstrates the importance of integrating external knowledge for coding accuracy.
Paper Resources
Article Excerpt
From source RSS / original summaryarXiv:2605. 27377v1 Announce Type: new Abstract: We present -Coding, an agentic method for automated ICD-10-CM coding. RAG-Coding orchestrates four (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.
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from arXiv cs.CL
See more →TriAgent: Divergence-Aware Committees for Cost-Efficient Financial Sentiment Analysis
TriAgent introduces a cost-efficient multi-agent system for financial sentiment analysis, combining VADER, FinBERT, and Qwen2.5. It achieves an F1 score of ~0.87 with significant savings of $9.3M/year at a 10M-user scale compared to GPT-4o-mini, while also detecting hallucinations with an AUC of 0.90.