The Impact of Financial Conflicts of Interest in Plastic Surgery: Are They All Created Equal?

ABSTRACT

BACKGROUND: Recently, several studies have demonstrated that articles that disclose conflicts of interests (COI) are associated with publication of positive results. The purpose of this study was to learn more about the different types of COI as they relate to the general topic of COI in plastic surgery. Specifically, we aimed to examine whether different types of COI are more likely than others to be associated with the presentation of positive findings. METHODS: We reviewed all original articles in Annals of Plastic Surgery, Journal of Plastic, Reconstructive, and Aesthetic Surgery, and Plastic & Reconstructive Surgery from January 1, 2012, to December 31, 2013. All scientific articles were analyzed, and several article characteristics were extracted. Disclosed COI were categorized into the following categories consultant/employee, royalties/stock options, and research support. The findings reported in each article abstract were blindly graded as reporting a positive, negative, neutral, or not applicable result. A multivariable analysis was performed to determine whether an association existed between certain types of COI and publication of positive conclusions. RESULTS: A total of 3124 articles were identified of which 1185 fulfilled the inclusion criteria. Financial COI were reported in 153 studies (12.9%). The most common type of COI was "research support" (7.3%), whereas the least common was "royalties/stock options" (1.2%). Rates of different types of COI varied significantly by plastic surgery subspecialty field (P < 0.001). In the multivariable analysis, authors who disclosed COI related to research support, consultant/employee, and royalties/stock options were 1.31, 6.62, and 8.72 times more likely, respectively, to publish positive findings when compared with authors that disclosed no COI after correcting for potential confounding factors. However, consultancy/employee status was the only COI category statistically associated with publication of positive results (P < 0.001). CONCLUSIONS: Self-reported COI are uncommon in plastic surgery research. Our results provide evidence that certain types of financial COI are more likely than others to be associated with the presentation of positive findings. This analysis suggests that certain investigators may be more biased, consciously or unconsciously, by the type of financial benefit offered by industry.