EHR-RAG: Bridging Long-Horizon Structured Electronic Health Records and Large Language Models via Enhanced Retrieval-Augmented Generation

TL;DR

EHR-RAG is a novel retrieval-augmented framework that enhances large language models' ability to interpret long-horizon structured electronic health records by preserving clinical structure, refining evidence retrieval, and jointly reasoning over factual and counterfactual data.

Contribution

The paper introduces EHR-RAG, a new framework with specialized retrieval and reasoning components tailored for long-term clinical EHR prediction tasks, outperforming existing LLM baselines.

Findings

EHR-RAG achieves an average Macro-F1 improvement of 10.76% over baselines.

The framework effectively preserves temporal and clinical structure in EHR data.

EHR-RAG outperforms existing methods across four long-horizon EHR prediction tasks.

Abstract

Electronic Health Records (EHRs) provide rich longitudinal clinical evidence that is central to medical decision-making, motivating the use of retrieval-augmented generation (RAG) to ground large language model (LLM) predictions. However, long-horizon EHRs often exceed LLM context limits, and existing approaches commonly rely on truncation or vanilla retrieval strategies that discard clinically relevant events and temporal dependencies. To address these challenges, we propose EHR-RAG, a retrieval-augmented framework designed for accurate interpretation of long-horizon structured EHR data. EHR-RAG introduces three components tailored to longitudinal clinical prediction tasks: Event- and Time-Aware Hybrid EHR Retrieval to preserve clinical structure and temporal dynamics, Adaptive Iterative Retrieval to progressively refine queries in order to expand broad evidence coverage, and Dual-Path Evidence Retrieval and Reasoning to jointly retrieves and reasons over both factual and counterfactual evidence. Experiments across four long-horizon EHR prediction tasks show that EHR-RAG consistently outperforms the strongest LLM-based baselines, achieving an average Macro-F1 improvement of 10.76%. Overall, our work highlights the potential of retrieval-augmented LLMs to advance clinical prediction on structured EHR data in practice.