A study was recently done to assess the accuracy of self-reported conflict-of-interest (COI) disclosures in the New England Journal of Medicine (NEJM) and the Journal of the American Medical Association (JAMA) within the disclosure period, prior to article submission. Authors of the study reviewed 206 original clinical-trial research articles published in NEJM and 188 published in JAMA during the 2017 calendar year and compared them to self-reported COI disclosure forms submitted to the journals with the published articles and the Open Payments website.
According to the study, only 31 articles from each of the publications (62 total) actually met inclusion criteria. Additionally, the physician authors (of which there were 118) received a combined total of more than $7 million. Of the 106 authors who received payments (89.8%), 86 received undisclosed payments. The top 23 highest compensated individuals received $6.32 million, of which almost half ($3 million) was undisclosed. Disclosure rates were comparable between the top 23 and the entire sample.
In conducting their research, the authors of the study coded all payments into one of four categories: disclosed, undisclosed, unrelated, or indeterminate. Payments that were self-reported by the medical researchers were considered “disclosed”; payments that were omitted from the self-reports that came from companies with products relevant to the given study, were labeled “undisclosed”; other payments omitted from the self-reports without products relevant to the given study, were classified as “unrelated.” If it was unclear whether the company offered related products at the time of the study, the payments were labeled “indeterminate.”
In the overall distribution of payments, 51.3% were found to be disclosed (59.8% in NEJM and 47.6% in JAMA), 45.6% were undisclosed (38.5% in NEJM and 48.3% in JAMA), 2.8% were unrelated (1.1% in NEJM and 4.0% in JAMA) and 0.4% (0.6% in NEJM and 0.1% in JAMA) were found to be indeterminate.
Considering slightly less than half of the medical researchers’ payments were not reported to the journals, the study concluded that high payment amounts and high proportions of undisclosed financial compensation resulted in possible conflicts of interest for two “influential” United States medical journals.
Interestingly, non-disclosure trends were similar with both the NEJM and JAMA, despite different procedures in how self-disclosures are handled with each journal. The authors of the study go on to note, therefore, that they “take no position on the physician-authors’ intentions in non-disclosure of their COIs; we characterize the issue as a ‘process problem’ rather than a ‘people problem,’ especially in light of the patterns observed in COI disclosure rates regardless of the journal’s disclosure process and regardless of payment amount.”
The authors of the study called for further research into why a large proportion of general payments went undisclosed, and whether the medical journals that rely on self-reported COI disclosure need to reconsider and/or revise their policies.