Improving interoperability and addressing data-sharing concerns in evidence synthesis: lessons from Cochrane reviews

Article type
Authors
Flemyng E1, van Valkenhoef G2
1Cochrane, UK
2Cochrane, Denmark
Abstract
Background:
Recent improvements in how Cochrane manages and shares data in its authoring platform (RevMan) and in publications in the Cochrane Library have helped continue its mission to produce trusted, accessible, and synthesized evidence. These developments not only foster collaboration but also combat research waste, streamline systematic review workflows, and fortify transparency and quality.

Objective:
This presentation will provide a detailed look at how changes to Cochrane reviews have (1) improved data interoperability in the evidence ecosystem and increased the potential impact of Cochrane evidence and (2) addressed researchers’ apprehensions about data sharing.

Main Content:
Recent research (Page et al. 2022) found that only 29% of randomly sampled systematic reviews incorporate a data availability statement, with just 4% offering downloadable data. The new data packages for Cochrane reviews include more data associated with the review, including data on included studies, analyses, and references, in user-friendly, reusable formats. These are downloadable both from the published review and from RevMan. This facilitates data reuse in other Cochrane systematic reviews, guideline development, and meta-research. How the data are shared is also novel and addresses data-sharing concerns raised by researchers, for example, through automatic data curation and mandatory terms and conditions for those downloading data to safeguard authors' rights.

In addition, modifications to data structures within RevMan bolster interoperability across the evidence ecosystem. This means data can more easily flow between data extraction tools, like Covidence and Excel, to RevMan and from RevMan to tools for summarizing findings, such as GRADEpro GDT. With the ability to import extracted data into RevMan, this also supports the automatic creation of included studies' tables and inclusion of data in analyses, better facilitating the production of timely and living systematic reviews.

Conclusion:
These transformations are poised to revolutionize the efficiency of producing and reusing underlying data from Cochrane reviews. They mark another milestone in Cochrane's journey towards embracing the FAIR—Findable, Accessible, Interoperable, Reusable—data principles, unlocking prospects for collaborations that will improve data flow across the evidence ecosystem and invigorate innovations based on Cochrane evidence.