In late June, over fifty members of Congress sent a letter to the Centers for Medicare and Medicaid Services (CMS) regarding the use of artificial intelligence (AI) in Medicare Advantage prior authorization determinations. The letter followed another letter sent in November regarding the same topic.
Both letters addressed Medicare Advantage plans’ reliance on AI tools to make coverage decisions and offered additional measures members of Congress would like CMS to implement. Here’s an overview of some of the Congressional requests:
Add More Information to Medicare Advantage Prior Authorization Denial Notices
In the 2024 Interoperability and Prior Authorization Rule, CMS indicated that payers must report metrics about their prior authorization processes. But members of Congress want health plans to include additional information related to denials, including:
- The denial reason
- Type of service
- Beneficiary characteristics, like health conditions
- Timelines of prior authorization decisions
The congressional members also want more information included on denial notices, including:
- Beneficiary-specific information about why a service is denied or terminated
- What information is lacking
- The criteria used to make the decisions
- The unmet regulatory requirement
- The health professional who reviewed the request
- Specific denial codes
Review and Approve AI Tools
The members of Congress prefer that CMS proactively provide oversight over AI tools to determine their impact on care. Specifically, these members of Congress would prefer that CMS review case files across MA plans, comparing AI tools used by plans to actual coverage determinations and the beneficiary’s profile to the AI tool’s general population.
Disallow the Use of AI Tools From Use in Denials Until Completing a Review of Their Use
Members of Congress would also like CMS to limit or disallow the use of AI tools in coverage denials until CMS can review them. Instead of relying on health plans to police their AI tools to ensure they don’t introduce discrimination and bias, the members of Congress would prefer that CMS take ownership and review solutions before allowing them to make coverage decisions. They take issue with the broad generalizations that AI can introduce instead of focusing on person-centered approaches to care, which they see as a core mission of Medicare.
Clarify how CMS Distinguishes Between Uses of Algorithms that Account for Individual Circumstances and Those That Do Not
To protect from erroneous care decisions, these members of Congress also believe CMS should delineate which AI solutions account for individual circumstances and those that don’t. Because plans often make initial care determinations and fail to reassess as a beneficiary’s condition changes, they also want CMS to clarify how plans ensure reassessments occur and how plans communicate with beneficiaries and providers about those reassessments.
Clarify When MA Organizations Can Use Internal Coverage Criteria When Making Medical Necessity Determinations
Because AI tools generally use large populations to determine treatments, the Congressional letter writers believe plans should base decisions on a beneficiary’s medical condition. As a result, they believe CMS should more fully explain in what limited circumstances plans should use external coverage criteria to make medical necessity determinations.
In this case, they would like CMS to consider requiring plans to use sub-regulatory guidance when making coverage decisions.
Create a Minimum Period During Which MA Plans Can’t Issue a Termination Notice After a Reversal of a Prior Termination Decision
Finally, the members of Congress would prefer that CMS implement a grace period after a reversed denial. The goal is to ensure a proper reassessment of the beneficiary’s condition in a reasonable amount of time.
Using AI can streamline the prior authorization process for health plans. Unfortunately, as the members of Congress contend, it can also introduce bias and overlook a beneficiary’s unique condition in favor of external conditions and treatment results. The role of CMS in reviewing and approving AI prior authorization solutions is still to be determined, but some members of Congress are pushing for such oversight.
Certifi’s health insurance billing and payment solutions help payers improve member satisfaction while reducing administrative costs.
