This thesis aims to map the use of Large Language Models (LLMs) in Clinical Decision Support, through a systematic literature review and a survey of physicians. The systematic review analyzed 39 articles from the Web of Science database published between 2023 and 2025. The findings suggest that LLMs hold significant potential to enhance clinical diagnosis and support treatment recommendations, functioning as assistive tools that complement rather than replace the physician’s role. The review categorizes the literature by technological approach, including General Purpose LLMs, Data Wrangling, Prompt Engineering, RAG, Imaging Analysis, and Multimodal Applications . It also addresses key barriers to LLM adoption in clinical decision support, such as their ’black box’ nature, hallucinations, data privacy concerns, regulatory challenges, ethics issues and algorithmic biases.