Conflicts of Interest in Bronchoscopy Research: Is Self-Reporting Sufficient?

BACKGROUND

Robotic assisted bronchoscopy has been enthusiastically adopted in the United States and has transformed the treatment of patients with indeterminate pulmonary nodules. Unprecedented industry investments in research, development, and marketing have profoundly affected the bronchoscopy landscape, leading to concerns that conflicts of interest could influence the validity of bronchoscopy studies. Disclosures of conflicts of interest in research are predicated on open and transparent self-reporting.

RESEARCH QUESTION

Are self-reported relevant conflicts of interest in articles pertaining to robotic assisted bronchoscopy accurate when compared with publicly available payments on the Open Payments Database of the Centers for Medicare & Medicaid Services?

STUDY DESIGN AND METHODS

All articles pertaining to robotic assisted bronchoscopy indexed on PubMed between 2016 and 2022 were screened for relevance. Articles appearing in the 5 journals with the most relevant publications were selected. General, research, and associated research payments reported in the Open Payments Database were recorded for each US physician-author with available data. "Relevant payments" refers to transactions made to authors by bronchoscopy-related companies. Documentation of all payments involving these companies during the 3 years before the submission date of an article was obtained. These payments were compared with the self-reported conflicts of interest for each author, per article, and the number and value of payments were categorized and totaled.

RESULTS

Twenty-seven articles were included, accounting for 75 US physicians with data reported in the Open Payments Database. Of the $17 million in relevant payments reported, $9.9 million were not disclosed (57%). Sixty-eight of 75 authors (91%) had incomplete physician disclosures. Excluding food and beverage payments, 60 authors had incomplete disclosures (80%).

INTERPRETATION

Relevant conflicts of interest appear to be inconsistently disclosed in publications on robotic assisted bronchoscopy, suggesting that self-reporting may be an insufficient strategy. A centralized disclosure process that is automated or easier to use should be considered.