RePrompT: Recurrent Prompt Tuning for Integrating Structured EHR Encoders with Large Language Models

TL;DR

RePrompT is a novel framework that enhances large language models with structured EHR data by recurrent prompt tuning, improving clinical prediction accuracy without altering model architectures.

Contribution

It introduces a time-aware prompt tuning method that integrates structured EHR encoders with LLMs, capturing longitudinal and population-level information effectively.

Findings

RePrompT outperforms baseline models on MIMIC datasets.

RePrompting improves longitudinal patient trajectory modeling.

Prompt tuning enhances LLMs for structured clinical data.

Abstract

Large Language Models (LLMs) have shown strong promise for mining Electronic Health Records (EHRs) by reasoning over longitudinal clinical information to capture context-rich patient trajectories. However, leveraging LLMs for structured EHRs (e.g., standardized diagnosis and medication codes) presents two key challenges. First, translating time-stamped EHR sequences into plain text can obscure both temporal structure and code identities, weakening the ability to capture code co-occurrence and longitudinal regularities. Second, unlike cohort-trained predictive models that learn a shared, task-aligned representation space across patients, LLMs are often applied in a case-isolated inference setting where each patient is processed independently without leveraging population-level patterns. To address these challenges, we introduce RePrompT, a time-aware LLM framework that integrates structured EHR encoders through prompt tuning, without modifying underlying architectures. Specifically, RePrompT recurrently incorporates latent states from prior visits to preserve longitudinal information, and injects population-level information through trainable prompt tokens derived from a cohort-trained, task-aligned EHR encoder. Experiments on MIMIC-III and MIMIC-IV demonstrate that RePrompT consistently outperforms both EHR-based and LLM-based baselines across multiple clinical prediction tasks.