• Sajid Anwar
• Dr. PAWAN MEENA
• YOGESH RAI

The rapid advancement of generative artificial intelligence (AI), particularly large language models (LLMs), has created significant opportunities for improving healthcare decision support. These models can interpret natural language, summarize clinical information, answer medical questions, and assist healthcare professionals in accessing complex knowledge. However, the direct application of generative models in healthcare remains challenging because their responses may contain hallucinated facts, outdated knowledge, insufficient evidence, and limited transparency. Such limitations can reduce trust and introduce substantial risks when AI-generated information is used in clinical environments. Retrieval-Augmented Generation (RAG) has emerged as a promising approach for addressing these challenges by combining the generative capabilities of LLMs with external, domain-specific knowledge retrieval. Instead of relying exclusively on parameters learned during training, a RAG system retrieves relevant evidence from trusted sources and incorporates this information into the generation process. This architecture can improve factual grounding, knowledge freshness, source traceability, and explainability. This survey presents a comprehensive examination of RAG-enabled generative AI for reliable and explainable healthcare decision support. The paper reviews the fundamental principles of generative AI, healthcare decision support systems, and retrieval-augmented architectures. It further analyzes the major components of healthcare RAG pipelines, including medical knowledge acquisition, document preprocessing, indexing, retrieval, reranking, context integration, evidence-grounded generation, and response verification. Particular emphasis is placed on reliability issues such as hallucination mitigation, factual consistency, retrieval quality, uncertainty estimation, robustness, and knowledge updating. The survey also investigates explainability mechanisms, including source attribution, retrieval traceability, evidence-based reasoning, and clinician-oriented explanations. In addition, the paper examines major healthcare applications of RAG, including clinical question answering, diagnostic assistance, treatment recommendation, electronic health record analysis, drug information retrieval, and personalized decision support. Existing evaluation strategies, privacy and security concerns, ethical challenges, and regulatory considerations are discussed. Finally, key research gaps are identified, and future directions are proposed for developing trustworthy, clinically useful, and human-centered RAG systems capable of supporting healthcare professionals without replacing clinical judgment.

Retrieval-Augmented Generation, Generative Artificial Intelligence, Large Language Models, Healthcare AI, Clinical Decision Support Systems, Explainable AI, Reliable AI, Medical Question Answering, Trustworthy AI

"Retrieval-Augmented Generative AI for Reliable and Explainable Healthcare Decision Support: A Comprehensive Survey", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.13, Issue 10, page no.a936-a946, October-2026, Available :http://www.jetir.org/papers/JETIR2610097.pdf

"Retrieval-Augmented Generative AI for Reliable and Explainable Healthcare Decision Support: A Comprehensive Survey", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.13, Issue 10, page no. ppa936-a946, October-2026, Available at : http://www.jetir.org/papers/JETIR2610097.pdf