Requirements for trial registration and adherence to reporting guidelines in critical care journals: A meta-epidemiological study of journals' instructions for authors

Background: The purpose of this study was to investigate the policies of critical care journals with regard to guideline adoption and clinical trial registration to understand the extent to which journals use these mechanisms to improve reporting practices. Methods: The current study's sample comprised 37 critical care journals cataloged in the Expanded Science Citation Index of the 2015 Journal Citation Reports and Google Scholar Metrics h5-index critical care subcategory. A webbased data abstraction was performed to identify which journals required, recommended, or made no mention of 17 different reporting guidelines. We also extracted whether journals required or recommended trial registration. Authors were blinded to one another's ratings until completion of the data validation. Cross tabulations and descriptive statistics were calculated by using STATA 13. Results: Of the 37 critical care journals, 15 (15/37, 40.5%) did not mention a single guideline within their instructions for authors, whereas the remaining 22 (22/37, 59.5%) mentioned one or more guidelines. The Quality of Reporting of Meta-analyses statement and Standards for Reporting Qualitative Research were not mentioned by any journals, whereas the International Committee of Medical Journal Editors Uniform Requirements for Manuscripts (26/37, 70.3%) and Consolidated Standards of Reporting Trials statement (17/37, 45.9%) were mentioned most often. Of the 37 critical care journals, 21 (21/37, 56.8%) did not mention trial or review registration, but the remaining 16 (16/37, 43.2%) mentioned at least one of the two. Trial registration through ClinicalTrials.gov was mentioned by six (6/37, 16.2%) journals, whereas the WHO registry was mentioned by five (5/37, 13.5%). Sixteen (16/37, 43.2%) journals mentioned trial registration through a registry platform. Conclusion: Nearly half of the journals in our sample did not mention a reporting guideline, and only a small percentage of journals required the registration of clinical trials as a condition for publication. Implementing these two mechanisms may limit bias, and their adoption should be considered by journal editors in critical care.