Potential Bias in Image-Guided Procedure Research: A Retrospective Analysis of Disclosed Conflicts of Interest and Open Payment Records.

PURPOSE To assess the prevalence of positive conflict of interest (COI) disclosures in U.S.-based interventional radiology (IR) research as well as the level of agreement between disclosed financial relationships and open payment data for top-cited image-guided procedure research. MATERIALS AND METHODS All publications in volume 30 (2019) of the Journal of Vascular and Interventional Radiology (JVIR) were reviewed to estimate the prevalence of COI disclosures in IR research. Publications were categorized as primary research, systematic review, or other. Prevalence was then compared across JVIR publication subtype, categories, and whether they were device-focused with chi-squared tests. Additionally, the Web of Science database was searched for the top 10 cited studies of 10 common image-guided procedures with available U.S. physician payment data. Payments were categorized as historical (>1 year prior to publication) or active (<1 year prior to publication) and compared to disclosed financial COIs with one-way ANOVA. RESULTS Positive COI disclosures were present in 29% (114 of 397) of publications in JVIR volume 30. Positive COI disclosures were most prevalent in Standards of Practice (50%, p = 0.01) and more prevalent in device-focused publications (54% vs 23%, p < 0.01). Among the 396 authors of 100 U.S.-based top-cited image-guided procedure publications, 383 (97%) failed to disclose at least one active financial relationship, with an average of $57,937 in undisclosed payments per publication. CONCLUSION Conflicts of interest are prevalent in IR, like other areas of healthcare research, and conflicts of interest in top-cited image-guided procedure research are often underreported.