MedCoT-RAG: Causal Chain-of-Thought RAG for Medical Question Answering

2025-08-20

AI Summary

To address hallucination, shallow reasoning, and poor clinical interpretability in large language models (LLMs) for medical question answering, this paper proposes the first causality-aware retrieval-augmented generation (RAG) framework. Our method integrates causal reasoning into medical RAG: (1) causal-aware document retrieval, aligned with diagnostic logic, to enhance evidence relevance; and (2) domain-customized structured chain-of-thought prompting to enable stepwise, traceable clinical reasoning. Unlike conventional semantic-matching RAG, our framework explicitly models causal relationships among symptoms, diseases, and treatments, thereby strengthening reasoning depth and interpretability. Evaluated on three medical QA benchmarks, our approach achieves 10.3% and 6.4% absolute accuracy gains over baseline RAG and state-of-the-art domain-adapted methods, respectively, while significantly improving reasoning consistency.

Technology Category

• Reasoning under Uncertainty: Causality
• Knowledge Representation and Reasoning: Diagnosis and Abductive Reasoning
• Machine Learning: Causal Learning

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Large language models (LLMs) have shown promise in medical question answering but often struggle with hallucinations and shallow reasoning, particularly in tasks requiring nuanced clinical understanding. Retrieval-augmented generation (RAG) offers a practical and privacy-preserving way to enhance LLMs with external medical knowledge. However, most existing approaches rely on surface-level semantic retrieval and lack the structured reasoning needed for clinical decision support. We introduce MedCoT-RAG, a domain-specific framework that combines causal-aware document retrieval with structured chain-of-thought prompting tailored to medical workflows. This design enables models to retrieve evidence aligned with diagnostic logic and generate step-by-step causal reasoning reflective of real-world clinical practice. Experiments on three diverse medical QA benchmarks show that MedCoT-RAG outperforms strong baselines by up to 10.3% over vanilla RAG and 6.4% over advanced domain-adapted methods, improving accuracy, interpretability, and consistency in complex medical tasks.

Problem

• Addresses hallucinations and shallow reasoning in medical LLMs
• Enhances retrieval with causal-aware evidence for clinical logic
• Improves structured reasoning in medical question answering

Innovation

• Causal-aware document retrieval for medical workflows
• Structured chain-of-thought prompting for clinical reasoning
• Step-by-step causal reasoning aligned with diagnostic logic