What is it about?

We tested whether large language models (LLMs), the same technology used in advanced chatbots, could accurately identify relevant articles for systematic reviews of medical literature. We evaluated 18 different LLMs on their ability to extract relevant research articles, and their potential to streamline the analysis of scientific literature. Moreover, we investigated how the models make decisions when selecting relevant literature.

Featured Image

Why is it important?

Our findings show that AI has significant potential to transform evidence-based medicine. By speeding up the creation of systematic reviews, AI facilitates quicker access to the latest medical knowledge. AI can also help to more rapidly identify gaps in current research. This can accelerate advancements in treatments, suggest directions for future research, and ultimately improve patient care.

Perspectives

This research has been exciting, highlighting AI's potential in medical research. I'm particularly intrigued that AI models agreed more with each other than human reviewers often do, suggesting both a potential for consistency and a need to investigate possible biases. This reinforces my belief that AI will play a crucial role in evidence-based medicine, but as a collaborator with, not a replacement for, human expertise. I hope this work sparks discussion about the responsible integration of AI into research to achieve more robust and impactful healthcare outcomes.

Fernando Miguel Delgado-Chaves
Universitat Hamburg

Read the Original

This page is a summary of: Transforming literature screening: The emerging role of large language models in systematic reviews, Proceedings of the National Academy of Sciences, January 2025, Proceedings of the National Academy of Sciences,
DOI: 10.1073/pnas.2411962122.
You can read the full text:

Read

Contributors

The following have contributed to this page