• <- Back to List

Financial conflicts of interest in orthopaedic research

Kanu Okike, M.D.

This thesis seeks to examine three aspects of financial conflict of interest in orthopaedic research: 1) the association between financial conflict of interest and manuscript publication, 2) the association between financial conflict of interest and the frequency of article citation after publication, and 3) the accuracy of conflict of interest disclosure as self-reported by orthopaedic surgeons.

To investigate the association between financial conflict of interest and manuscript publication, we examined manuscripts that were submitted to a prestigious orthopaedic journal but not accepted. Data were extracted on conflict of interest disclosure as well as other factors plausibly related to publication, and PubMed searches were conducted to determine which manuscripts went on to achieve full publication within five years. To investigate the association between financial conflict of interest and article citation, we examined articles published in three prestigious general orthopaedics journals. Data were extracted on conflict of interest disclosure as well as other factors plausibly related to citation, and the number of citations at five years was determined. To determine the accuracy of conflict of interest disclosure, we transcribed the reports of payments made to orthopaedic surgeons by five hip and knee arthroplasty manufacturers and compared them to the disclosure statements of individuals participating in an AAOS Annual Meeting. To ascertain the reasons for nondisclosure, a survey was administered to individuals in receipt of payments that were not disclosed.

Disclosure of a for-profit or non-profit conflict of interest was associated with a significantly higher likelihood of publication following rejection by a prestigious orthopaedic journal (p=0.028 and 0.027, respectively). Disclosure of a for-profit or non-profit conflict of interest was also associated with a significantly higher rate of citation after publication (p=0.011 and 0.027, respectively). The overall rate of payment disclosure was 71.2%, including 79.3% for directly-related payments, 50.0% for indirectly-related payments, and 49.2% for unrelated payments. The reasons most commonly given for non-disclosure were that the payment was unrelated to the topic of presentation (38.9% of respondents), that the disclosure requirements had been misunderstood (13.9%), or that the payment had been disclosed but mistakenly omitted from the program (11.1%).

Disclosure of a financial conflict of interest was associated with a higher likelihood of publication following rejection by a prestigious orthopaedic journal, as well as a higher rate of citation following publication. When conflicts of interest were self-reported by orthopaedic surgeons, the rate of disclosure was 79.3% for directly-related payments and 50.0% for indirectly-related payments.