Invited review: Completeness of reporting of experiments: REFLECTing on a year of animal trials in the Journal of Dairy Science

Reproducibility is an essential element of the scientific process, and it requires clear and complete reporting of study design, conduct, and analysis. In the human and animal health literature, incomplete reporting is associated with biased effect estimates. Moreover, incomplete reporting precludes knowledge synthesis and undervalues the resources allocated to the primary research. The Reporting Guidelines for Randomized Controlled Trials for Livestock and Food Safety (REFLECT) statement, published in 2010, is a checklist developed by expert consensus to provide guidance on what study elements should be reported in any intervention trial (designed experiment) involving livestock. The Journal of Dairy Science (JDS) has recently endorsed the use of reporting guidelines. To assess the status of reporting of controlled experiments in JDS and to provide a baseline for future comparison, we evaluated the reporting of 18 items from the REFLECT statement checklist in a sample of 137 controlled trials published in IDS in 2017. Two reviewers independently screened titles and abstracts for relevance and then evaluated a sample of 120 papers reporting controlled trials (experimental studies involving at least one intervention and one comparison or control group), using yes or no questions. Although some items, such as treatment details and statistical analysis, were well reported, other areas, including sample size justification, allocation concealment, blinding, study flow, baseline data, and ancillary analyses, were often not reported or were incompletely described. This work highlights the need for authors and reviewers to take advantage of guidelines and checklists for reporting. Adherence to reporting guidelines can help improve the completeness of reporting of research, expedite and better inform the peer-review process, increase clarity for the reader, and allow for knowledge synthesis, such as meta-analysis, all of which serve to increase the value of the work conducted.