Investigation of Financial Conflict of Interest among Published Ventral Hernia Research

Abstract

Background: Discordance exists between author self-disclosure and the Open Payments Database in various surgical fields, but the effects of this discordance on study design and presentation are unknown. We hypothesized that, among ventral hernia publications, discordance exists between industry and physician self-reported conflicts of interest (COIs); authors disclose relevant COIs; and disclosure and relevant COIs affect study favorability.

Study design: We conducted a double-blinded, prospective, observational study of published articles. PubMed was searched in reverse chronological order for clinical articles pertaining to ventral hernias. Authors' self-disclosed conflicts were compared with those on the Open Payments Database. Two reviewers blinded to article disclosure status determined jointly whether the COIs were relevant to the article. Three blinded referees independently voted whether each article was favorable to discussed subject matter. The primary end point was study favorability. Secondary outcomes included disclosure status and relevance.

Results: One hundred articles were included. Compared with authors with no COIs, authors with a COI, self-disclosed or not, were twice as likely to write results favorable to industry. Of those with a COI, most of the articles had a relevant COI (37 of 45 [82.2%]), and 25% of relevant COIs were not disclosed by authors. Among authors with a relevant COI, study favorability remained unchanged at 68.5% (control: no COI 33.3%; p < 0.001).

Conclusions: Within the ventral hernia literature, 70% of articles have a COI. Self-reporting of COI is discordant in 63% of articles. Twenty-five percent of relevant COI are not disclosed. Having a COI increases the chances that an article will cast a favorable impression on the company paying the authors by 200%.