Court Blocks CMS Ability to Recover Overpayments from Medicare Advantage

Overpayments in the Medicare Advantage (MA) program continue to be a significant problem for the federal government, and the problem may have just gotten harder to tackle. Last month, a federal district court vacated the Centers for Medicare & Medicaid Services’ (CMS) 2023 Risk Adjustment Data Validation (RADV) rule – a rule that the CMS uses to recoup MA overpayments – ruling that it was a technical violation of the Administrative Procedures Act (APA).

Congress should step in to empower CMS’ oversight of MA plans and help encourage overpayment recovery by codifying their RADV authority. Alternatively, or additionally, the Trump Administration could restart the rulemaking process to reinstate the authority in an APA-compliant manner.

The RADV program is a tool CMS uses to identify and recover overpayments to MA. Only a few months ago, CMS touted its commitment to beefing up this program, leveraging the 2023 final rule. Through RADV, CMS audits medical records to ensure that the diagnoses listed are appropriately supported with documentation. Without these audits, MA could add more diagnoses and receive higher risk-adjusted payments. This is important because plans are paid more when enrollees have multiple or serious diagnoses. Past audits have shown that MA plans often list diagnosis codes for enrollees that are not reflected by their medical records, garnering higher federal payments than the plans deserve.

On September 25, 2025, the U.S. District Court for the Northern District of Texas vacated the 2023 rule that lays out RADV requirements. The court found that CMS violated the Administrative Procedure Act’s requirements to follow a thorough notice and comment process. Specifically, the court ruled that the final 2023 rule did not sufficiently stem from the preliminary rule put forward in 2018. As a result, the opinion halted two audit tools that were key to CMS recovering significant overpayments.

The first tool involves CMS’s authority to extrapolate audit findings. CMS divides plan members into different tiers based on risk, identifies the percentage of errors within each, and extrapolates that percentage to the entire tier. In this way, CMS can recover greater overpayments with fewer audits using statistically valid sampling methods.

The second tool is called the “Fee-for-Service Adjuster (FFS Adjuster),” and sets the FFS error rate as a benchmark to compare MA to FFS. When CMS applies the FFS Adjuster, MA plans are only penalized if errors exceed the FFS Adjuster rate. CMS’ final rule removed the FFS Adjuster, holding MA to a higher standard and increasing recoveries. However, the court ruling forces CMS to reimplement the FFS Adjuster, so CMS will recover less in overpayments.

CMS has not announced whether they will appeal the court’s decision. Without an appeal, CMS would likely have to go through the full rulemaking process again and delay implementation of this commonsense program integrity measure.

Alternatively, Congress could avoid a long rulemaking process by codifying key audit authorities in statute. Specifically, Congress could mandate that CMS implement the extrapolation methods and eliminate the FFS Adjuster. Congress could go even further to reduce waste in the MA program, for example, by lowering the benchmark payment to MA plans, raising the minimum “upcoding” adjustment from 5.9 percent, and eliminating the techniques plans use to make beneficiaries look sicker, such as health risk assessments.

Given our estimate of $1.2 trillion of MA overpayments over ten years, Congress should take action now to exercise oversight over the MA program.