Background: Since 2005, the International Committee of Medical Journal Editors (ICMJE) requires researchers to prospectively register their clinical trials in a publicly accessible trial registry. The Consolidated Standards of Reporting Trials (CONSORT) statement has supported this policy since 2010. We aimed to evaluate to what extent biomedical journals have incorporated ICMJE's clinical trial registration policy into their editorial and peer review process.
Methods: We searched journals' instructions to authors and performed an internet survey among all journals publishing reports of randomised controlled trials that follow ICMJE's Recommendations for the Conduct, Reporting, Editing and Publication of Scholarly Work in Medical Journals (n = 695), and/or that endorse the CONSORT statement (n = 404) accessed in January 2011. Survey invitations were sent to the email addresses of the editorial offices and/or editors-in-chief of included journals in June 2011.
Results: For 757 ICMJE and/or CONSORT journals, we identified that they published RCT reports. We could assess the instructions to authors of 747 of these; 384 (51%) included a statement of requiring trial registration, and 33 (4%) recommended this. We invited 692 editorial offices for our survey; 253 (37%) responded, of which 50% indicated that trial registration was required; 18% cross-checked submitted papers against registered records to identify discrepancies; 67% would consider retrospectively registered studies for publication. Survey responses and specifications in instructions to authors were often discordant.
Conclusion: At least half of the responding journals did not adhere to ICMJE's trial registration policy. Registration should be further promoted among authors, editors and peer reviewers.