Prospective trial registration in the Research Integrity Assessment (RIA) of randomized controlled trials (RCTs)
2Evidence-based Medicine, Department I of Internal Medicine, Faculty of Medicine and University Hospital Cologne, University of Cologne, Germany
3Cochrane Injuries Group, London School of Hygiene & Tropical Medicine, Germany
Background:
The WHO and the Declaration of Helsinki require that every clinical study must be prospectively registered in a publicly accessible study registry (i.e., registration before enrolling the first patient). However, most evidence syntheses ignore violations of this fundamental moral and ethical principle. The Research Integrity Assessment (RIA) (doi:10.1002/jrsm.1599), a tool to assess adherence of randomized-controlled trials (RCTs) to the principles of Good Clinical Practice, and key elements of research integrity in clinical trials, assesses prospective registration and may help evidence synthesis authors to deal with studies which do not comply with the requirements. RIA was piloted in the updated Cochrane review “Ivermectin for preventing and treating COVID-19” revealing that lack of prospective registration was the most frequent reason for the exclusion of studies according to RIA.
Objectives:
To evaluate trial registration in RCTs included in the study pool of COVID-19 systematic reviews (SRs) on interventional drugs and to analyse the impact with the application of RIA.
Methods:
We searched for SRs investigating at least 1 of 13 different interventions for treatment or prevention of COVID-19 in Medline/PubMed until 09 June 2022. We selected Cochrane and non-Cochrane SRs with the largest RCT pool, extracted all included RCTs, evaluated trial registration data, and applied the RIA domain 2 (prospective trial registration) to all RCTs. RIA was independently applied by two reviewers; conflicts were resolved by a third reviewer.
Results:
We identified 188 RCTs for RIA domain 2 assessment. Twelve RCTs were not registered and 176 RCTs were registered in at least 1 of 10 clinical trial registries. Using RIA, we assessed 109 prospectively registered RCTs as ‘no concernʼ, 51 non- or retrospectively registered RCTs as ‘excludeʼ, and 28 RCTs with inconsistent information between publication and registry record, missing information, inaccessible registry record, or no publication as ‘awaiting classificationʼ (Table 1).
Conclusion: Almost one in three studies on interventional treatments for COVID-19 has not complied with the international standard of prospective trial registration. Authors of evidence synthesis can use the RIA tool as a transparent mechanism to manage studies that do not adhere to the international standard.
Patient and/or public involvement: none.