Reproducibility Policies in Cardiology Journals: The REPLICA Cross-Sectional Study

Abstract

AbstractImportanceTransparency and data sharing are valuable practices in research, contributing to improved precision and flexibility in cumulative evidence; and ultimately expanding the research ecosystem by addressing one of the philosophical research norms that implies that knowledge belongs to society.ObjectivesThe objective of the Reproducibility Policies In Cardiology Journals (REPLICA) study was to estimate the proportions of policies and guidance for reproducibility and transparency practices among Cardiology journals, as well as to determine details of completeness of reporting and data sharing conditions whenever disclosed.DesignCross-sectional analysis.SettingCross-sectional study through analyses of journals deposited in the National Library of Medicine (NLM) Catalog tagged with the “Cardiology” and “Vascular Diseases” entry terms.Eligibility CriteriaCardiology journals from the NLM Catalog database that published at least one randomized clinical trial in 2018. Journals that published articles in English, Spanish, French, or Portuguese and were available in MEDLINE/PubMed were eligible for inclusion.ExposuresThe exposures were mainly related to journal’s characteristics such as publisher operations characteristics (e.g., journal access only by subscription), indexation in the DOAJPlus, requirement for registration for RCTs, and others.Main outcomes and measuresWe prespecified a primary composite outcome composed of data-sharing policy or guidance. Secondary outcomes were proportions of reporting guidelines within the journal’s instructions for the author’s section (e.g., CONSORT), and also other components of sharing practices.ResultsWe assessed 148 journals. Of them, 74 (50.0%, 95%CI 41.9% to 58.1%) presented policy or guidance for data sharing. We found guidance for data sharing in 68 journals (47.5% 95%CI 39.4% to 55.8%). Notably, among them, only two mentioned sharing individual participant data (IPD). Regarding guidelines for article reporting, we identified that 132 journals displayed guidance for authors, in which 27 (20.45%, 95%CI 14.34% to 28.29%) had CONSORT and EQUATOR Network guidance content.Conclusion and relevanceIn summary, we found a mild proportion of policies and guidance for data-sharing. Moreover, transparency practices inclined to RCTs are suboptimal, as mirrored by the very low prevalence of IPD data-sharing policy and guidance as well as specific reporting guidelines instructions for RCTs.Key PointsQuestionWhat is the proportion of journals displaying policies and guidance about data sharing in cardiology journals?FindingsWe found a low prevalence of policy and guidance for data sharing in Cardiology journals, as well as transparency and reproducibility practices; details, individual participant data sharing, registration, and completeness of reporting, for example.MeaningJournals play a role in driving reproducibility and transparency among scientific areas. Stakeholders involved in the editorial processes should be open to understand the valuable impact of data-sharing practices and learn how to implement such mechanisms, that being the case.