COVID-19 has sparked a rapid change in the medical research landscape, with preprints and open access rapidly becoming the norm. However, a recent article has highlighted the suboptimal and non-transparent scientific practices that have coincided with this revolution, contributing to a “dysfunctional scientific process” for COVID-19 research.
Besançon and colleagues analysed reviewing times, conflicts of interest and misuse of non-peer reviewed material in COVID-19 publishing data over recent months. They grouped the flaws found into three main groups, and suggested solutions using open science principles:
- Data collection and interpretation: Methodological and statistical issues (such as in the famous preprint on the use if hydroxychloroquine for the early treatment of COVID-19), duplication of research, and ethical concerns were identified as the main issues in the early stages of COVID-19 studies. The authors suggested that study preregistration may help reduce duplication of research and reduce bias, and that the adoption of registered reports may improve the quality and robustness of the scientific evidence and reduce wasted resources.
- Publication process: There is concern that shortcuts have been taken when publishing the massive volume of COVID-19 research publications. These include the quality of review under expedited timelines, handling of conflicts of interest, and distrust of the results – with retractions occurring even in ‘top-tier’ journals. Suggested solutions include open review (sharing reviewers’ reports and authors’ responses) to better balance reviewing quality and reviewing time, and data sharing to improve peer review quality as well as facilitating reproducibility.
- Science communication: Although preprints have been widely, and often beneficially adopted around COVID-19 (7-fold more than the previous 6 months), their misuse, particularly by the media, has led to confusion among the public and the scientific community. The authors call for more reasonable communication from academic institutions, and for journalists and news editors to be cautious in their reporting of findings.
Open science cannot solve broader societal issues such as scientific illiteracy and the flawed system for the evaluation of researchers that contribute to these imperfect research practices
The authors note that open science cannot solve broader societal issues such as scientific illiteracy and the flawed system for the evaluation of researchers that contribute to these imperfect research practices, but they hope that the adoption of open science principles will lead to more a rigorous research process, and more reliable and transparent science during this pandemic.
With thanks to our sponsor, Aspire Scientific Ltd