Workflow Validation of the (semi-) automated Review Features of the Epistemonikos Sustainable Knowledge Platform: A Prospective Validation Study

Evidence synthesis is crucial for evaluating healthcare interventions, but the growing volume of scientific literature creates challenges in maintaining both speed and rigor. This study evaluates the performance of the Epistemonikos Sustainable Knowledge Platform (SKP), which integrates artificial intelligence to streamline evidence synthesis. Specifically, we assess four semi-automated SKP features: literature searches, abstract screening, full-text screening, and data extraction. The investigation compares SKP's performance against human reviewers, examining each feature individually and in combination, with particular focus on their impact on results and evidence certainty.