MedPlan:A Two-Stage RAG-Based System for Personalized Medical Plan Generation
This work addresses the need for more effective and clinician-aligned treatment planning in healthcare AI, though it is incremental as it builds on existing RAG and LLM methods.
The paper tackled the problem of generating personalized medical treatment plans from electronic health records by addressing limitations in current LLM-based systems, such as lack of sequential reasoning and patient-specific context, and introduced a two-stage RAG-based framework that significantly outperformed baselines in assessment accuracy and treatment plan quality.
Despite recent success in applying large language models (LLMs) to electronic health records (EHR), most systems focus primarily on assessment rather than treatment planning. We identify three critical limitations in current approaches: they generate treatment plans in a single pass rather than following the sequential reasoning process used by clinicians; they rarely incorporate patient-specific historical context; and they fail to effectively distinguish between subjective and objective clinical information. Motivated by the SOAP methodology (Subjective, Objective, Assessment, Plan), we introduce \ours{}, a novel framework that structures LLM reasoning to align with real-life clinician workflows. Our approach employs a two-stage architecture that first generates a clinical assessment based on patient symptoms and objective data, then formulates a structured treatment plan informed by this assessment and enriched with patient-specific information through retrieval-augmented generation. Comprehensive evaluation demonstrates that our method significantly outperforms baseline approaches in both assessment accuracy and treatment plan quality.