SafeRx-Agent: A Knowledge-Grounded Multi-Agent Framework for Safe and Explainable Medication Recommendation

2026-05-27

AI Summary

This work addresses the limitations of existing drug recommendation methods, which often rely on coarse-grained drug categories and lack safety validation and traceability mechanisms, leading to potential overestimation of risk. To overcome these issues, the authors propose a multi-agent framework that integrates patient context, external clinical knowledge, and rigorous safety checks, achieving fine-grained drug recommendations at the fourth-level ATC code granularity for the first time. The approach leverages large language models, drug–drug interaction and contraindication detection, structured ATC code generation, and a knowledge-guided multi-agent collaboration mechanism to ensure both interpretability and safety of recommendations. Experiments on MIMIC-III and MIMIC-IV demonstrate that the method significantly improves fine-grained recommendation accuracy while effectively controlling risks related to drug interactions, contraindications, and recommendation set size.

Abstract

Medication recommendation predicts medications for patient visits, but existing methods still face two key challenges. At the model level, traditional drug recommendation methods only predict structured drug codes with limited evidence grounding, while LLM agents can use richer clinical context but may lack safety verification and traceability. At the task level, existing benchmarks often use broad medication categories, which ignore subgroup-level safety differences and can lead to risk overestimation. We introduce the first fine-grained medication recommendation setting based on fourth-level ATC code generation. We propose Safe Prescription Agent (SafeRx-Agent), a knowledge-grounded multi-agent framework that uses patient context, external clinical knowledge, and safety verification to recommend traceable medication sets. Experimental results on MIMIC-III and MIMIC-IV datasets show that SafeRx-Agent improves fine-grained medication prediction accuracy while controlling drug interactions, contraindications, and medication set size.

Problem

• medication recommendation
• safety verification
• fine-grained ATC codes
• explainability
• clinical knowledge grounding

Innovation

• fine-grained medication recommendation
• knowledge-grounded multi-agent
• ATC code generation
• safety verification
• explainable AI