Medicare Advantage RADV Audits: Finding the Error Is Only the Beginning

A federal court threw out the Medicare Advantage audit rule in September 2025 because the final version departed too far from the proposal the Centers for Medicare & Medicaid Services had put before the public. The separate duty to return identified overpayments remained. For someone working in risk adjustment, compliance or audit response, the question reaches beyond the court fight. Your team may have found the unsupported diagnoses already. What happened to those findings, and to the money they brought in?

The agency announced a major audit expansion, but the Government Accountability Office later found that it lacked a detailed plan for accelerating the work. The distance between finding an error and recovering the payment is the story behind Medicare Advantage RADV audits.

Who Audits Medicare Advantage Plans?

The Centers for Medicare & Medicaid Services conducts Risk Adjustment Data Validation audits to check the medical-record support for diagnoses used to pay Medicare Advantage plans. The Department of Health and Human Services Office of Inspector General also audits plans. Plans and providers commission internal and outside coding reviews, and the Justice Department pursues False Claims Act cases involving alleged Medicare Advantage fraud. These reviews answer different questions about the same money.

The payment arrangement explains why the record matters. Medicare pays a plan a monthly amount adjusted for its members’ expected healthcare costs. Diagnoses feed into disease categories and risk scores. A higher score means a higher payment. The path from diagnosis to payment starts with information the plan submits about its own members.

The agency acknowledged in its final-rule explanation that it uses those diagnoses before validating them against the medical records. That gap before payment puts considerable weight on the later audit. By the time a reviewer asks whether the chart supports the condition, Medicare has already paid on the diagnosis.

RADV’s purpose is to identify improper risk-adjustment payments. A risk adjustment fraud inquiry also follows what the organization knew about an unsupported diagnosis and what it submitted or retained afterward. An audit report supplies one part of that account. The decisions inside the business supply another.

From Sample Errors to Contract Recoveries

Extrapolation uses the errors found in a sample to estimate overpayments across the larger group the sample represents. That changes the financial consequence of an audit: recovery can reach beyond the particular members whose charts a reviewer read.

The agency laid out that approach in its 2012 methodology, intended to begin with payment year 2011. Reviewers would examine a sample of members’ records, calculate payment errors and use the results to estimate the contract’s payment error.

That methodology also included a fee-for-service adjuster, an offset to the preliminary recovery amount. The dispute behind it concerned errors in traditional Medicare diagnosis data used to calibrate Medicare Advantage payments. Insurers argued that CMS should account for those errors when applying medical-record standards to their own diagnoses.

In its later rule, the agency said that it had not yet issued final RADV results for that starting payment year or any subsequent year, and had therefore never applied the proposed adjuster. The calculation had been published. The final audit results still had not followed.

What Did the 2023 RADV Rule Change?

The February 2023 RADV rule provided for extrapolated recoveries beginning with payment year 2018 and rejected the fee-for-service adjuster. For payment years 2011 through 2017, the agency limited recovery to overpayments identified in the audited samples. The boundary determined whether an error found in a small set of records could support a much larger recovery.

The rule published on February 1, 2023 projected approximately $4.7 billion in transfers from plans to the government over ten years.

The adjuster decision mattered for the same reason. An offset would reduce the preliminary audit recovery. The agency rejected it, arguing that errors in traditional Medicare claims did not systematically distort the risk scores produced by its model. Plans disputed the relationship between the data used to set their payments and the standard used to audit them.

Why Did the Court Throw Out the Medicare Advantage Audit Rule?

The court threw out the Medicare Advantage audit rule because the final version did not follow sufficiently from the proposal on which CMS had sought public comment. In the challenge brought by Humana, the federal district court in the Northern District of Texas set aside the rule on September 25, 2025.

The decision concerned how the agency made the rule. It did not prohibit extrapolation itself.

The agency’s payment-year 2020 instructions, dated March 20, 2026, show what that meant in practice. It designed the audits to support extrapolation but said it had not decided whether it would recover extrapolated amounts or only the overpayments associated with sampled enrollees. It described collecting sample-level payment errors while reserving broader recovery for later, if legally permitted.

The medical-record review could proceed while the larger recovery question remained open. A plan’s response to an audit and the government’s authority to extrapolate its findings were separate pieces of the work.

Will CMS Audit All Medicare Advantage Contracts?

The Centers for Medicare & Medicaid Services announced that it would audit all eligible Medicare Advantage contracts for each payment year in newly initiated audits. Its May 21, 2025 announcement set an early 2026 target for finishing the remaining payment-year 2018 through 2024 audits.

The agency said it would move from roughly 60 plans audited annually to approximately 550. It also promised more medical coders and larger record samples.

Technology would flag unsupported diagnoses, and human coders would verify them. The agency also said it would work with the Inspector General to recover uncollected overpayments from earlier audits.

Its own account showed how much catching up it faced. The agency said the last significant recovery of Medicare Advantage overpayments had followed the payment-year 2007 audit. It also reported that completed audits for payment years 2011 through 2013 found overpayments between 5% and 8%. The government had evidence of payment errors while recovery remained far behind. Taxpayers were still waiting for the money back.

After that target had passed, the Government Accountability Office reported on July 21, 2026 that CMS lacked a detailed plan for expediting RADV audits and that the backlog contributed to significant delays in recovery.

More auditors are a promise. Money returned to Medicare is the result that matters.

What Does a RADV Audit Finding Show?

A RADV audit finding shows whether the submitted medical records support the diagnosis categories used in payment, and how the validated information changes the payment calculation. The payment-year 2020 instructions separate record intake, diagnosis review and calculation of the payment error. A record accepted for review has not yet passed the diagnosis check.

The Enrollee Data List identifies the members and diagnosis categories selected for audit. The plan obtains records from the hospitals and practitioners that treated the member. At intake, the agency checks the provider’s signature, credentials and encounter dates. The intake feedback report does not give the results of the diagnosis review.

The final findings can confirm the audited condition category, support a higher or lower category in the same hierarchy, or leave the audited category unsupported. The agency then uses the relevant payment model to calculate the financial effect. An error count alone does not tell that whole story: which category survived review affects the risk score and the money.

For a whistleblower, the audit result becomes especially useful when it connects to an internal decision. Who received the finding? Did the organization correct the data that fed payment? What did management decide about money already received? Those are questions about the company’s response to information it had.

The Refund Duty Survives the Rule Dispute

The Medicare 60-day overpayment rule requires a Medicare Advantage organization to report and return an identified overpayment within that period. The agency’s RADV rule described the refund duty as a longstanding obligation that applied when an insurer learned through any means that a diagnosis lacked medical-record support. That duty existed separately from the disputed extrapolation policy.

The payment-year 2020 instructions paused specified overpayment-reporting and data-correction submissions for the audit’s full sampling group, including members not selected for review. They kept the reporting requirement for identified overpayments outside that group.

That pause had a defined population and payment year. We examine the direction alongside the diagnoses the company identified, the people it covered and the decisions made afterward. An insider may know whether management applied it to the same population or used it to explain a much broader halt in corrections.

When an Audit Finding Reaches Management

The company’s response to a finding can explain more than the error rate. In the government’s complaint involving Independent Health and its chart-review vendor, DxID, an internal response treated the prospect of defending a diagnosis in an audit as a reason to submit it:

The government alleged that the coding program generated unsupported diagnoses. Its complaint also described an internal Group Health review in which at least 17 of 20 audited vendor codes failed the applicable documentation requirements. The review findings and the internal approval illustrate different parts of the alleged conduct: what the charts supported and what the business was willing to submit.

Sutter Health’s 2021 settlement for $90 million supplies another consequence of that gap. The case, started by a former employee, alleged unsupported diagnoses and a failure to take sufficient corrective action to find and delete additional unsupported codes after Sutter became aware of the problem. A whistleblower case under the False Claims Act can pay 15% to 30% of what the government recovers.

That is why reviews that process additions while leaving unsupported codes in place matter alongside government audits. An internal review may tell the company something about its payment data before CMS samples it. The finding, the people who received it and the action taken afterward belong in the same account.

Why Control of the Medical Chart Matters

Control over documentation can affect what an audit sees. The American Institutes for Research’s 2026 report, Medicare Advantage Fraud, Waste, and Abuse Policy Options, examines how organizations that own or control both health plans and providers can influence coding through record-system prompts, retrospective reviews and documentation incentives.

The report cites a study in which patients’ risk scores rose by 25% two years after their physicians joined corporate primary-care practices, without additional evidence of increased health care use. Its concern extends to the document the auditor eventually receives:

The report identifies a policy risk: the organization benefiting from the diagnosis may also influence the chart against which it is checked. For an insider, the history of that chart can matter alongside its final contents. Coding instructions, prompts and review decisions can explain how the diagnosis reached the record and then the payment system.

When You Know What Happened After the Audit

If your team found unsupported diagnoses and you know why the payment data stayed unchanged, speak with our whistleblower experts about the findings and the response. We build your strongest Medicare Advantage whistleblower case around the audit, the submission history and the decisions about repayment. We bring in Medicare Advantage fraud whistleblower lawyers who understand risk adjustment, and we stay beside you throughout the case.