Impact of the Restoring Invisible & Abandoned Trials (RIAT) initiative on Cochrane evidence products
The Restoring Invisible & Abandoned Trials (RIAT) initiative is an international effort to tackle bias in the way research is reported. Cochrane should pay attention to and benefit from this project.
Objectives: To identify the Cochrane evidence products directly concerned by two trials, one on evolocumab and the other on plasma rich in growth factors (PRGF), recently restored through RIAT projects, and also explore potential changes to be implemented.
Methods: Two searches in the ‘All Text’ box of the Cochrane Library were performed on 15/02/2023 using ‘evolocumab’ and ‘PRGF’ as search terms. All identified results were screened looking for Cochrane products including the original trials [FOURIER (doi:10.1056/NEJMoa1615664); BTI-01-EC/07/ART (doi:10.1016/j.arthro.2012.05.011)] subsequently restored by RIAT researchers (doi:10.1136/bmjopen-2021-060172; doi:10.1186/s13063-022-07049-3), highlighting areas in need of updating.
Results: One Cochrane review (CD011748), one protocol (CD012917) and one Clinical Answer (CA) (doi:10.1002/cca.3409) were identified as linked to the FOURIER trial. One Cochrane protocol (CD013341) was retrieved as linked to the BTI-01-EC/07/ART trial. Regarding the Cochrane review, the FOURIER trial accounted for more than 95% of the total sample size in the evolocumab versus placebo comparison. After assessing only published data, FOURIER was labeled as low risk of bias in all domains, supporting the global ‘high certainty of evidence’ according to GRADE. This is contradicted by the subsequent published restored trial based on its Clinical Study Report, which will likely lead to not being categorized as low risk in ‘Reporting’ and ‘Other bias’ domains. The CA, specifically designed to widen the Cochrane review audience, reiterated the challenged message of evidence with high certainty. Both ongoing protocols identified are expected to consider the restored trials for inclusion.
Conclusions: In the examples above, the certainty of evidence would likely be downgraded when restored versions of trials are considered. Thus, Cochrane reviews including restored trials are in need of a rapid update to reflect the best available evidence, particularly when restored trials had a major weight on their reviews. In parallel, Cochrane Central could link restored and original versions of trials to assist authors with the search and identification of restored trials.
Patient, public and/or healthcare consumer involvement: None.