Developers of AI/machine learning (ML)-enabled medical devices paid $39.7 million for radiology devices between 2017 and 2023, according to a study published September 22 in JAMA.
The payouts were part of an overall $179 million bestowed to healthcare professionals and hospitals during the timeframe, and were second only to $59.4 million paid for cardiovascular devices, noted study lead author David-Dan Nguyen, MD, of the University of Toronto in Ontario, Canada, and colleagues.
“Financial relationships with manufacturers may exert significant influence on decisions to adopt, implement, or promote these technologies,” the group wrote.
The rapid authorization AI/ML medical devices by the U.S. Food and Drug Administration (FDA) over the last decade raises “familiar concerns” about conflicts of interest, the researchers wrote. As such, in this analysis, they examined the extent and nature of payments made to health care professionals and teaching hospitals by manufacturers of FDA-authorized AI/ML–enabled devices.
The group first identified 846 devices on the FDA’s AI/ML–Enabled Medical Device List. Of these, 79 (9.3%) were linked to payments to healthcare professionals and hospitals between 2017 and 2023, based on an analysis of the U.S. Centers for Medicare and Medicaid Services’ Open Payments Database.
In total over the period, $120.2 million was dished out in general payments and $59.1 million in research payments. On an annual basis, total payments increased from $17.3 million in 2017 to $24.6 million in 2023, driven primarily by increasing general payments, which doubled from $6.6 million to $13.3 million, the researchers reported.
In addition, the number of devices for which manufacturers reported payments increased from 32 in 2017 to 53 in 2023, and the median seven-year total payment amount per product was $142,538 for general and $238,362 for research.
Lastly, the researchers noted that the number of recipients increased from 7,911 physicians and 116 teaching hospitals in 2017 to 14,066 physicians, 4,091 nonphysician health care professionals and 153 teaching hospitals in 2023.
Ultimately, these reported payments may be just the tip of the iceberg, the group noted, as manufacturers of more than 90% of FDA-listed AI/ML–enabled devices publicly disclosed no such payments.
“For example, manufacturers of radiology devices may be less likely to disclose payments because many fall outside current reporting requirements due to differing clearance pathways, reimbursement models, and uncertain regulatory classification,” they wrote.
In conclusion, the researchers suggested that reporting requirements should be modernized to comprehensively capture AI/ML applications and ensure adequate transparency and oversight.
The full study is available here.
