Back

The Epistemonikos Database of Trials: A Comprehensive and Freely Accessible Bibliographic Source of Randomized Trials in Health

Rada, G.; Avila-Oliver, C.; Ramos-Rojas, J. T.; Vasquez-Laval, J.; Vergara, C.; Cantor-Cruz, F.; Rojas-Gomez, A. M.; Veloso, V.; Bosio, C.; Bignon, M.; Pinto, S.; Silva, I.; Zambrano-Achig, P.; Fernandez-Sandoval, S.; Jara, I.; Nava-Mosler, D.; Verdugo-Paiva, F.; Pena, J.; Novillo, F.; Silva-Zapata, A.; Biscay, D.

2025-11-17 health informatics
10.1101/2025.11.16.25340330 medRxiv
Show abstract

Randomized controlled trials (RCTs) are the cornerstone of evidence-based healthcare, providing the most reliable means of assessing the efficacy and safety of health interventions. However, their rapidly growing volume and dispersion across diverse sources, such as bibliographic databases, preprint servers, and trial registries, pose major challenges for evidence synthesis, clinical guideline development, and health technology assessment. Despite longstanding recognition of their importance and multiple initiatives to improve accessibility, there is still no comprehensive, freely available, and continuously updated resource of trial records. To address this gap, the Epistemonikos Foundation developed the Epistemonikos Database of Trials (ED-Trials), a continuously updated repository of RCTs identified from more than 30 sources, including bibliographic databases, preprint servers, international trial registries, and other search sources. This paper describes the methods used to identify and classify trial records, outlines the key features of the search interface, and reports the database status as of October 2025. ED-Trials now includes more than 1.5 million unique trial records, many of which are not indexed in MEDLINE or other conventional bibliographic databases. It represents a significant scholarly and technical achievement, offering a freely accessible and user-friendly resource that enables more efficient, comprehensive, and up-to-date identification of randomized controlled trials, thereby strengthening the foundation of evidence-based healthcare.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.