Will the Epistemonikos Trials Database avoid the need of searching in multiple databases?

Article type
Authors
Rada G1, Ávila-Oliver C1, Vergara C1, Vásquez J1, Verdugo-Paiva F1, Peña J1, Novillo F1, Ramos J1, Cantor-Cruz F1, Rojas A1, Veloso V1, Bosio C1, Biscay D1, Bignon M1, Pinto S1, Silva I1, Zambrano P1, Jara I1, Nava D1, Garrido D1, Silva A1
1Epistemonikos Foundation, Santiago, Chile
Abstract
Background:
Randomized controlled trials (RCTs) are the gold standard for evaluating healthcare interventions. However, identifying all relevant RCTs for systematic reviews is a time-consuming and resource-intensive process. Searching across multiple, fragmented databases leads to inefficiencies and potentially incomplete evidence syntheses.
To address these challenges, the Epistemonikos Foundation, building upon its experience in evidence synthesis and leveraging cutting-edge technologies (including artificial intelligence), will launch the Epistemonikos Trials Database in July 2024.

Objective:
The Epistemonikos Trials Database aspires to be a one-stop shop for identifying RCTs.

Methods:
To construct and maintain the database, our plan involves regularly harvesting from over 50 diverse sources, including electronic databases, trial registries, and preprint servers. Additionally, we manually check the list of included studies in reviews identified in the Epistemonikos Database of Systematic Reviews and add the trials identified through this process. The records are deduplicated automatically and then classified using different technologies, such as artificial intelligence, alongside a network of human collaborators.
We will assess the database's comprehensiveness using the relative recall method, employing a reference standard derived from a substantial sample of trials included in systematic reviews.

Results:
As of submitting this abstract, we have identified 1,119,912 trials. Among these, 103,862 records come from the Epistemonikos database of Systematic Revies (9.3%). The uploading process from various sources is ongoing, and the manual identification of trials will be fully implemented by July 2024. Complete figures and a formal evaluation of the database's comprehensiveness will be presented at the Global Evidence Summit.

Conclusions:
The Epistemonikos Trials Database, launching in July 2024, aims to streamline RCT identification for systematic reviews using AI and other innovative approaches. This one-stop shop has the potential to improve evidence synthesis efficiency and comprehensiveness.

Importance to patients:
While this project is not directly oriented towards patients as users, its potential impact on enhancing the efficiency and quality of systematic reviews could greatly benefit patients by facilitating better health-related decision-making.