ASNC and AMA push back against AI prior authorization denials
The American Society of Nuclear Cardiology (ASNC), in collaboration with the American Medical Association (AMA), is taking a firm stance against health insurance companies’ growing use of artificial intelligence (AI) to automate prior authorization denials—an approach that physicians say is undermining patient care and contributing to professional burnout.
Suman Tandon, MD, FASNC, a cardiac imaging specialist at St. Francis Hospital and Heart Center in Roslyn, New York, represented ASNC at the AMA House of Delegates (HOD) meeting this week, where Resolution 226 was adopted. The measure calls for AMA advocacy to regulate algorithm-based utilization review processes employed by insurers. The resolution seeks federal oversight to transparently evaluate potential biases in how AI systems review and deny requests for diagnostic tests, procedures and medications.
"Prior authorization has been a major issue for many physicians. Any physician you ask will tell you that prior authorization has gotten worse, not better. So I think it is important that the House of Delegates continually assess, Tandon explained.
ASNC was among the supporters of AMA HOD Resolution 226 at the annual meeting in June, which calls for regulating AI algorithmic-based health insurance utilization review. It asks the AMA to advocate for oversight and legislation to improve patient safety and the transparency of the prior authorization processes of insurers.
The use of AI by insurance companies has accelerated—particularly in Medicare Advantage plans, where automation is used to approve or deny coverage. Tandon said this practice removes essential clinical judgment and often ignores patient-specific nuances that only a trained provider can assess in person.
ASNC is also backing bipartisan legislation recently introduced in Congress (H.R. 3514). The bill would require Medicare Advantage plans that utilize AI in prior authorization processes to report detailed statistics to the Centers for Medicare and Medicaid Services (CMS). These include the number of AI-reviewed cases, denial rates, reasons for denials, and impacts on vulnerable communities, such as low-income, rural and minority patients, Tandon explained.
The need for such legislation is supported by troubling statistics. In 2023, there were 15 million prior authorization requests filed under Medicare Advantage, of which 3.2 million were either partially or fully denied. Most physicians and patients do not appeal denials—only around 12% are appealed. But, for those denials that were appealed, 82% were ultimately overturned, suggesting widespread inappropriate or premature rejections, Tandon said.
"The growth of AI and its predictive abilities require better over sight, transparency and accountability on the part of the health plans that utilize this technologies to review requests for prior authorization. This issue is deeply concerning to the physicians and it has been presented a number of times to the AMA House of Delegates because AI is increasingly being used to review prior authorization of tests and medications," Tandon said.
She said there is a lot of nuance in clinical decision making, and reducing it to an algorithm by an insurer undermines patient care.
"Overall, I would say prior authorization increases practice costs, it increases administrative burden, it undermines patient care, and causes delays in patient care. I also think that it erodes patient trust. So if you have a patient, you've just sat with them 20 minutes, recommended a test to them after they listening to you, and now they get denied by the insurer, they're very conflicted. It can affect your patient physician relationship and we do not want that," Tandon explained.
She also warned that clinical decision-making is being replaced by rigid AI models lacking the ability to weigh individual patient needs. Tandon said medical decisions are based on experience, guidelines, expert consensus and patient interaction—not AI models working for payers.
ASNC and other medical societies are urging the AMA and Congress to press insurers for greater accountability, urging reforms that ensure AI support, rather than replacing clinical expertise.
For more, watch the full video interview above.