EOSC-ENTRUST workshop: Support of IPD meta-analyses through trusted research environments (TREs)
As part of the EOSC-ENTRUST project, ECRIN in collaboration with the University of Oslo, Health Data Research UK and the University of Dundee, hosted the workshop “Support of individual patient data (IPD) meta-analyses through trusted research environments (TREs)” which focused on enabling secure, legally and ethically compliant cross-border re-use of clinical research data.
It brought together international experts from different backgrounds including statistics, data sharing, meta-analysis, TREs, data repositories, e-infrastructures, federated analysis, legal and ethical frameworks, Health Technology Assessment, data standards, and representatives from the DG-SANTE.
IPD meta-analysis is a specific type of systematic review. Rather than extracting summary (aggregate) data from study publications or from investigators, the original research data are sought directly from the researchers responsible for each study. These data can then be re-analysed centrally and combined, if appropriate, in meta-analyses. IPD meta-analyses of clinical trial data enhance analytical potential, yielding more robust and reliable evidence.
There are many hurdles for IPD meta-analyses, including the findability, the accessibility and the re-usability of datasets. This is reflected in the low data-sharing rates for IPD meta-analyses, which have been demonstrated in several studies and are still observed despite the various data-sharing initiatives and platforms. This situation is going to be improved by the new and forthcoming regulation (EHDS), better provision of Trusted Research Environments (TRE) for data analysis, and further developments of federated approaches and analyses.
Over the course of the workshop held on January 15, 42 participants explored opportunities and addressed challenges for improving meta-analysis of clinical trial data through faster, safer and more comprehensive data sharing within federated Trusted Research Environments (TREs).
