Retrieval-to-abstraction LLM-assisted literature synthesis: Case demonstrations in digital ...

This study presents a novel retrieval-to-abstraction framework using large language models (LLMs) to accelerate literature synthesis and systematic review of digital health technology-enabled randomized controlled trials (RCTs). The approach shows high accuracy comparable to human annotators in screening and extracting data, supporting scalable and rigorous evidence synthesis to improve reporting guideline adherence.

This Wire brief sits within Fusion42's coverage of Generative AI, Digital Health and Data & Analytics.