Redundant trials: do researchers register clinical trials for outcomes that have a 'high' GRADE evidence rating?

Session: 

Oral session: Understanding and using evidence (1)

Date: 

Tuesday 22 October 2019 - 16:00 to 17:30

Location: 

All authors in correct order:

Vergara Merino L1, Verdejo C1, Pérez P1, Sánchez L1, Klabunde R2, Franco J3, Escobar Liquitay C4, Madrid E5
1 Cochrane Chile, Universidad de Valparaíso, Chile
2 Cochrane Chile, Chile
3 Centro Cochrane Asociado, Instituto Universitario Hospital Italiano, Argentina
4 Centro Cochrane Asociado, Biblioteca Central, Instituto Universitario Hospital Italiano, Argentina
5 Cochrane Chile, Interdisciplinary Centre for Health Studies Universidad de Valparaíso, Chile
Presenting author and contact person

Presenting author:

Laura Vergara Merino

Contact person:

Abstract text
Background: the GRADE rating system presented in the 'Summary of findings' (SoF) table in Cochrane Systematic Reviews (SR) serves to measure the quality of evidence in an SR. When a primary outcome in an SoF table is rated 'high', this indicates that there is confidence that the true effect of the intervention on this primary outcome is close to that of the presented estimate, and it is very unlikely to change based on future research. Despite this, clinical research may continue to study these outcomes.

Objectives: identify registered clinical trials studying an intervention and primary outcome for which a 2015 Cochrane SR reported a 'high' GRADE rating of evidence quality.

Methods: we searched the Cochrane Library for all publications from 2015 (full-year). Editorials, overviews, and qualitative, diagnostic or methodological reviews were excluded from this analysis. We analyzed the remaining intervention SRs, and excluded withdrawn reviews or those without SoF tables. We extracted GRADE ratings of SR primary outcomes SoF tables, and identified those with 'high' GRADE ratings. We analyzed updates of these SRs to identify changes in GRADE ratings between publications. Subsequently, we completed a search in the World Health Organization International Clinical Trials Registry Platform (WHO ICTRP) to identify relevant clinical studies registered after the date of publication of the 2015 SR. We analyzed titles and relevant registries of resulting studies to identify studies having a primary outcome matching one of the 'high'-rated primary outcomes of the 2015 SR. We noted these in the study database.

Results: we identified a total of 1006 publications from 2015, of which there were 11 editorials, six overviews, and one qualitative, 26 diagnostic, and one methodological reviews. We excluded these from analysis. There remained 961 intervention SR, 52 being withdrawn reviews and 307 not containing SoF tables. We extracted primary outcome data and GRADE ratings from 602 intervention SR. Of these, 81 SR contained a 'high' rating on at least one primary outcome. Of 290 primary outcomes from these 81 SR, 147 had 'high' GRADE ratings. We searched outcomes from 46 SR published in 2015 in the ICTRP database, and have identified 45 registered trials thus far that have the same intervention and 'high'-rated primary outcome as a 2015 SR. The search in the ICTRP database is ongoing.

Conclusions: this study demonstrates the existence of registered clinical trials investigating primary outcomes that have already been rated as 'high'-quality evidence in a 2015 Cochrane SR. There was a significant proportion of SR of intervention studies with no SoF table presented, which is notable as this is currently a recommended Cochrane practice.

Patient or healthcare consumer involvement: the redundant conduct of clinical trials places unnecessary risk or discomfort on patients without the generation of novel knowledge, and it represents a waste of financing and resources for scientific inquiry.