Auditors say Humana and UnitedHealthcare plans overbilled Medicare about $180 million and both refuse to repay.

Image Credit: José Luis Filpo Cabana - CC BY 3.0/Wiki Commons/

Two federal audits released within days of each other found that a Humana Medicare Advantage plan and a UnitedHealthcare of Wisconsin plan together may have collected an estimated $177.8 million more than they were owed from the Medicare program for 2020 and 2021. The HHS Office of Inspector General traced the money to diagnosis codes that plan-submitted medical records did not actually support, and both insurers have told the inspector general’s office they see no obligation to send any of it back, despite formal recommendations that they do exactly that.

A Nationwide Sweep of High-Risk Diagnosis Codes

Medicare pays Medicare Advantage insurers more each month for enrollees whose paperwork makes them look sicker, using diagnosis codes the plans collect from providers and submit to CMS for its risk-adjustment formula. A code for a severe complication, a major organ failure or an active cancer carries a materially higher monthly payment than a routine chronic-condition code, which is precisely why OIG treats certain high-risk diagnoses as worth auditing on their own. CMS itself estimates that roughly 9.5 percent of everything it pays Medicare Advantage organizations is improper, mainly because of diagnoses the medical record does not actually back up. That figure is why HHS-OIG runs a standing, five-project audit series targeting the codes its own data flag as most prone to being upcoded, among them acute stroke, heart attack, vascular embolism and several cancers. The HumanaChoice and UnitedHealthcare of Wisconsin reports published this month closed out two of those five projects; three more, covering other Medicare Advantage contracts, remain open and unreported, so the total recovered or disputed across the series is still growing, as the OIG’s own series tracker shows.

Neither company is a stranger to this kind of review. OIG’s public report archive shows the office has audited other HumanaChoice contracts under the same category of diagnosis-code compliance work in past years, separate from the current five-project series, meaning this month’s finding extends a pattern of scrutiny rather than opening a new one.


Free download: IRMAA, the Medicare Savings Programs and Extra Help explained side by side, with the official form and contact for each. Get the free Medicare premium route finder.

From a Chart Review to a $177.8 Million Estimate

Neither report claims to have found nine-figure billing errors sitting in a single ledger. Investigators pulled a random, statistically valid sample of enrollee-years from each contract, had physician reviewers check whether the underlying medical record actually supported the diagnosis code the plan had billed to CMS, then used the error rate found in that sample to project a total across every enrollee CMS paid the plan for in 2020 and 2021. That projection produced the two headline figures: an estimated $130.9 million for HumanaChoice’s contract H5216 and at least $46.9 million for UnitedHealthcare of Wisconsin’s contract H5253, a combined $177.8 million. Extrapolation is the standard method OIG applies across this entire audit series rather than a shortcut built for these two contracts, but it also means the number represents a statistical estimate of the underlying error rate, not a claim-by-claim tally of confirmed overbilling that CMS has separately verified dollar for dollar. OIG’s reports note that as more of the sampled diagnoses fail the medical-record check, the projected total climbs accordingly, and both audits found that most of the codes each company selected for review did not hold up, which is what pushed each estimate into nine figures rather than a smaller, contained sum.

Two Refusals, Two Different Objections

Humana, Inc., HumanaChoice’s parent company, disagreed with several of OIG’s findings and rejected all four of the audit’s recommendations, which included refunding the $130.9 million, examining 212 additional enrollee-years the auditors had not personally reviewed, and checking for the same coding problems in years after the audit period. UnitedHealthcare of Wisconsin went further: rather than negotiate over the $46.9 million, the company disagreed with the findings and asked OIG to withdraw every one of its recommendations outright, arguing the sampling approach overstated the error rate. Both insurers’ objections centered on the same target, the statistical projection itself, rather than a dispute over any single enrollee’s chart, and neither company’s response disputed that CMS’s risk-adjustment payments to it were based on the diagnosis codes OIG reviewed.

What an OIG “No” Actually Means for the Money

An OIG recommendation carries no independent power to pull money out of an insurer’s account. The reports refer the matter back to CMS, which decides how hard to press for recovery, whether through negotiated settlement, a formal risk-adjustment data validation action, or simply logging the disagreement and moving to the next contract in the series. That gap matters to Medicare Advantage enrollees even though neither of them personally received an overpayment: the risk-adjustment dollars at the center of the dispute are the same pool that funds the extra benefits, such as reduced drug costs or supplemental coverage, that MA plans use to compete for members, and every dollar an insurer keeps after being told to return it is a dollar that never flows back into the Medicare Trust Fund those same enrollees rely on. HumanaChoice’s report also leaves 212 enrollee-years still unexamined, and OIG’s own work-plan tracker shows three more contracts still moving through the identical review process, meaning the two refusals logged this month are unlikely to be the last word from this series. For now, the official record on both contracts reads the same way: an audit finding a shortfall, and an insurer declining to close it.


The Coding Fight Enrollees Never See

The dispute over these two contracts plays out entirely between a federal auditor and an insurer’s compliance office; the enrollee whose diagnosis history triggered the extra risk-adjustment payment has no seat at that table and no way to check whether a similar coding question is sitting inside a plan’s own current filings. The rules that decide what a Medicare Advantage enrollee actually pays out of pocket next year, from drug costs to prior-authorization hurdles, sit on the other side of that same paperwork gap.

The Medicare Cost & Coverage Protection Kit is a 10-page kit covering the new Part D out-of-pocket cap, the prior-authorization appeal steps, and 51 state Medicare cost-help packs, plus a medication and cost tracker.

See the The Medicare Cost & Coverage Protection Kit.

This article was researched and drafted with the assistance of AI and reviewed by an editor.