The Prevalence and Disclosure of Financial Conflicts of Interest among Endovascular Specialists on Medical Social Media

Purpose: To evaluate the prevalence of conflicts of interest (COI) among interventional radiologists and physicians from other specialties who mention specific devices or companies on medical social media.

Materials and Methods: Twitter hashtags commonly used by interventional radiologists and other specialties who perform endovascular procedures were used to identify posts by US physicians mentioning a specific device or company as well as random paired sample posts not mentioning a device or company. A total of 12,028 tweets posted between 10/28/2021 and 12/31/2021 were evaluated. The Centers for Medicare & Medicaid Services Open Payment database was used to identify potential financial COIs among post authors. An active, positive COI was pre-defined as receiving a payment from the device manufacturer or company within 12 months prior to posting on social media. The prevalence and amount of positive COIs were assessed and compared between posts mentioning a device or company and the paired control group using descriptive statistics and chi-squared tests.

Results: A total of 90 posts were identified mentioning a specific device or company with positive COIs present in 52% (47/90) of these posts. A total of 180 posts (90 including device/company mention and 90 time-matched controls) were analyzed. By specialty, 37% of posts by interventional radiologists mentioning a device or company had a positive COI relative to 65% among neurosurgeons, 80% among vascular surgeons, and 100% among interventional cardiologists. Physicians who mentioned a specific device or company were 5 times more likely to have a positive COI relative to the paired control group (47/90 vs 9/90, p< 0.001). On average, physicians with a positive COI received $11,461 in payments. None of the positive COIs were disclosed in the posts.

Conclusion: COI’s are prevalent among physicians posting about specific devices or companies on medical social media. None of these COIs were disclosed, prohibiting followers from weighing the potential bias in the content being posted.