Navigating the Intersection of AI and Medicare Advantage: A Call for Oversight
In a recent and pivotal development, Democratic lawmakers in the U.S. House of Representatives have raised profound concerns about the use of artificial intelligence (AI) in the realm of Medicare Advantage (MA) plans. Their primary focus is on urging the Centers for Medicare and Medicaid Services (CMS) to critically analyze the impact of AI on claims denial rates, particularly in light of the 2024 final Part C and D rule. This rule introduces new prior authorization requirements, prompting worries that it may not adequately address the escalating reliance on AI within MA plans for guiding coverage decisions.
The Complex Landscape
To appreciate the concerns raised, it’s essential to recognize the significant differences in processes between traditional Medicare and the MA program. In traditional Medicare, items and services are submitted to Medicare Administrative Contractors for payment without prior authorization. However, the MA program extensively employs prior authorization, especially for higher-cost services. A 2018 report by the Department of Health and Human Services’ Office of Inspector General highlighted widespread issues concerning denials of care and payment in Medicare Advantage. Subsequently, a 2022 report unveiled that 13% of denied prior authorization requests met Medicare coverage rules.
AI’s Amplification of Prior Authorization Challenges
Lawmakers contend that the integration of AI into MA plans, with specific reference to software firms like naviHealth, myNexus, and CareCentrix, exacerbates challenges related to prior authorization. According to their assertions, the use of AI in coverage determinations within certain care settings results in more restrictive decisions compared to the guidelines laid out by traditional Medicare. The consequence is a discernibly higher rate of denied care, a phenomenon that has raised significant alarm.
The Call for Rigorous Oversight
In response to these concerns, lawmakers are advocating for increased transparency and oversight to ensure the responsible and ethical use of AI tools within MA plans. Their recommendations include a mandate for CMS to require MA plans to report comprehensive prior authorization data. This data would encompass reasons for denial, types of services, and beneficiary characteristics such as health conditions. The overarching objective is to create a comprehensive dataset that facilitates a robust evaluation of AI tools by comparing the guidance generated with the actual coverage decisions made by MA plans.
Additionally, the legislators propose an assessment of the frequency of denials for the same individual within the same episode of care. This analysis, based on data from Quality Improvement Organizations and Independent Review Entities, aims to identify trends in MA appeals related to hospital discharges, skilled nursing facility discharges, and home health terminations.
Moreover, the lawmakers suggest a detailed evaluation of how AI determinations are adjusted in response to unanticipated changes in a patient’s condition. They are keen on requiring attestation from MA plans and their contractors, including care management firms like naviHealth and myNexus, ensuring that their coverage guidelines are not more restrictive than those outlined by traditional Medicare. Importantly, they advocate for the establishment of an enforcement mechanism in cases where discrepancies are identified.
Addressing Bias and Ensuring Fairness
Recognizing the broader concerns about potential bias and the homogeneity of patient testing populations when developing AI algorithms, lawmakers emphasize the necessity for CMS to scrutinize the data used by plans in making determinations. Specifically, they want CMS to investigate whether plans are inappropriately incorporating factors such as race into their algorithms, aiming to ensure fairness and equity in healthcare decision-making.
Commendation for Executive Action
The lawmakers express their appreciation for the recent executive order by the Biden administration, released on October 30. This order, which establishes the first set of standards for the use of AI in healthcare, aligns closely with their call for increased oversight and regulation of AI tools within the healthcare landscape. It signals a positive step toward ensuring responsible and ethical AI integration in the healthcare sector.
The Larger Trend
Against the backdrop of these concerns and recommendations, it’s crucial to contextualize the larger trend in Medicare Advantage. The program has experienced substantial growth, now covering over half of all Medicare beneficiaries. A recent Health Affairs analysis reveals that the growth has been driven, in part, by individuals switching from traditional Medicare to Medicare Advantage, especially among younger and healthier patients. However, the increasing spotlight on prior authorization has prompted industry groups, including the American Hospital Association and the American Medical Association, to request a reconsideration of regulatory proposals that may introduce complexities into the prior authorization process.