Disclosures Undisclosed

Category: Ethics Introduction/Purpose: The objective of our study was to analyze COI nondisclosure for US based research articles that were published to three prevalent orthopaedic journals from the beginning of 2014 to the end of 2016. Methods: All US-based research articles published to FAI, JBJS, and JOA from the beginning of 2014 to the end of 2016 were reviewed. COI disclosure sections were analyzed to determine if a disclosure was made for first and/or last authors. Authors disclosing any financial relationship involving employment, royalties/licensing, speaking, and consulting fees were recorded as having disclosed a potential COI. First and last authors were then searched for using CMS Open Payments search tool to determine if they had received any of the aforementioned payment types. To determine if a COI nondisclosure (disclosure discrepancy) was present, an author’s disclosure statement for a published article was compared to CMS Open Payments data from the year prior to publication of the article. We used CMS data from the year prior to a publication to account for the time it takes to construct a paper and have it published. Results: Across all journals and years, we obtained disclosure accuracy data for 3,465 total first and last authors publishing 1,770 research articles. Within this sample, 7.1% (245/3,465) of the authors had a recorded undisclosed conflict-of-interest and 13.2% (233/1,770) of articles had a first and/or last author with an undisclosed potential conflict-of-interest. When looking at each journal individually over the three-year period, FAI contained the highest percentage of authors with undisclosed COI’s (42.3%), JBJS contained the lowest percent of authors with an undisclosed COI (4.6%), and JOA had an intermediate amount of authors with an undisclosed COI (7.0%). Conclusion: Discrepancies between payment disclosures made by authors and those published in the CMS database were present in all three journals reviewed in this study. The percentage of articles containing an author with a disclosure discrepancy varies widely between these journals. However, when analyzing the percentage of disclosure discrepancies by year, no trend was found.