ClinicalAgent: Clinical Trial Multi-Agent System with Large Language Model-based Reasoning

TL;DR

ClinicalAgent is a multi-agent system utilizing GPT-4 and advanced reasoning techniques to improve clinical trial outcome prediction and enhance the utility of large language models in medical research.

Contribution

The paper introduces ClinicalAgent, a novel multi-agent system that integrates GPT-4, LEAST-TO-MOST, and ReAct reasoning to advance clinical trial applications.

Findings

Achieves 0.7908 PR-AUC in clinical outcome prediction

Outperforms standard prompt methods by 0.3326 PR-AUC

Demonstrates improved LLM performance in clinical contexts

Abstract

Large Language Models (LLMs) and multi-agent systems have shown impressive capabilities in natural language tasks but face challenges in clinical trial applications, primarily due to limited access to external knowledge. Recognizing the potential of advanced clinical trial tools that aggregate and predict based on the latest medical data, we propose an integrated solution to enhance their accessibility and utility. We introduce Clinical Agent System (ClinicalAgent), a clinical multi-agent system designed for clinical trial tasks, leveraging GPT-4, multi-agent architectures, LEAST-TO-MOST, and ReAct reasoning technology. This integration not only boosts LLM performance in clinical contexts but also introduces novel functionalities. The proposed method achieves competitive predictive performance in clinical trial outcome prediction (0.7908 PR-AUC), obtaining a 0.3326 improvement over the standard prompt Method. Publicly available code can be found at https://anonymous.4open.science/r/ClinicalAgent-6671.