One-Way Chart Review Fraud: Why Codes Are Added but Unsupported Diagnoses Stay

Medicare Advantage chart review fraud runs one way when an organization uses review findings to collect more money while ignoring findings that show it was overpaid. The same chart can support a missed diagnosis and expose an unsupported one already submitted. If you worked on those reviews, the telling part may be what happened after you reported the error. Someone wanted the additions. What happened to the deletions?

What Is a Medicare Risk Adjustment Chart Review?

A Medicare risk adjustment chart review checks existing medical records for diagnoses a provider omitted or submitted incorrectly. It can also uncover a submitted diagnosis the chart does not support.

The money follows the diagnosis. Medicare pays private plans a monthly amount for each enrollee, adjusted for expected health care costs. Eligible diagnoses feed the risk score through hierarchical condition categories, or HCCs. That score helps determine the plan’s payment. A later review can therefore change what the plan receives for the patient without a new visit taking place.

The payment incentive also reaches the healthcare provider. A medical group can supply the diagnoses to a plan and receive a share of the plan’s Medicare payments. Whoever reads the charts, whoever submits the codes and whoever shares the resulting revenue may work for different organizations.

The one-way form of risk adjustment fraud turns on how those organizations handle contrary information. The review has already done the work of opening the chart. An unsupported diagnosis is sitting in the same results used to seek more money.

The Same Selective Review Across Companies

HealthCare Partners, doing business as DaVita Medical Holdings, agreed to a $270 million settlement announced in October 2018 covering one-way reviews and other coding practices. The Justice Department’s announcement described whistleblower allegations that the group searched for missed diagnoses and sent them to Medicare Advantage plans, while ignoring inaccurate codes that should have been deleted. Those deletions would have reduced Medicare reimbursement or required the plans to repay Medicare. The provider group shared in the payments its diagnoses helped generate.

Cigna’s $172,294,350 settlement, announced in September 2023, covered chart reviews, home assessments and morbid-obesity coding. The government’s chart-review allegations concerned payment years 2014 through 2019. The government alleged that Cigna used coders’ reviews of services already provided to submit additional diagnoses. The Justice Department put the selective use of their work plainly:

Aetna’s March 2026 settlement announcement described the same choice for payment year 2015. The government alleged that Aetna submitted additional codes from its reviews but did not withdraw previously submitted codes those reviews failed to substantiate. The $117.7 million settlement also covered separate morbid-obesity allegations.

The important feature is the use of the same results. In each account, the review found information that could move payment in either direction. The government alleged that the organization acted on the direction that paid. Taxpayers paid for diagnoses the same reviews failed to support.

Where the Deletion Stops

A correction has to reach the data Medicare uses to pay the plan. In its April 2022 reminder to plans, the Centers for Medicare & Medicaid Services said organizations that identify incorrect submitted diagnosis data must delete it through the submission process. That obligation applies before or after the final risk-adjustment deadline. The agency also described recalculating risk scores and adjusting payments after corrections.

A reviewer, a vendor and a plan can each hold a different part of that history. The instructions say what work was ordered. The review output says what was found. The submission and payment records say whether the finding went anywhere. That is how an insider’s account can explain a missing correction that an additions total conceals.

Production Measures That Reward Only Additions

The government’s complaint against Anthem alleges a review program organized around new revenue. It quotes Anthem’s instruction to the chart-review vendor:

The complaint alleges that newly identified codes capable of generating additional payment received a second review. It also alleges Anthem had no process to compare previously submitted diagnoses with the vendor’s findings for the same patients and visits, even though its programmers could build that comparison. In the government’s account, extra scrutiny went to the new revenue. The comparison that could expose old errors was left undone.

The same complaint alleges that Anthem left an active-lung-cancer code in place even though its review did not substantiate the diagnosis and no other provider reported it that year. Meanwhile, Anthem allegedly added three other codes from the review. The complaint also describes transcription errors that turned a diabetes code into a leukemia code. That is what revenue-driven risk score targets can mean in a patient’s record.

Vendor contracts can reward that direction too. The government’s complaint concerning DxID alleges that Group Health Cooperative paid the vendor 20 percent of additional revenue generated by its reviews. It also alleges that Independent Health withdrew kidney-disease codes for an earlier service year after notice of the problem, but retained later years’ codes and continued the same practice. The vendor’s share of added revenue belongs beside its instructions, output and the plan’s response to errors. Together they show what the business paid to find and what it did with the rest.

The Study That Changed the Measuring Stick

A study published in Health Management Policy and Innovation in 2019 compared Medicare Advantage and traditional Medicare claims from 2010 through 2014. It assessed risk using both the diagnosis categories that drove Medicare Advantage payment and a different diagnosis-based measure, adjusted diagnostic groups, that did not determine payment.

With the payment measure, the Medicare Advantage population appeared substantially higher risk. With the alternative measure, its risk scores fell to approximately the traditional Medicare level. The point, as the study describes those adjusted diagnostic groups, is that they draw on patients’ diagnoses just as the payment categories do but play no part in Medicare Advantage payment.

The study found evidence suggesting that plans used greater coding intensity to drive higher payments. The company records supply the next level of detail: who directed the review, which findings produced revenue and which findings were left behind.

When an Audit Finds Too Much

The Sutter Health whistleblower’s 2015 complaint describes auditors at Palo Alto Medical Foundation reporting both added and deleted HCCs to management. The whistleblower alleged that reviewers found additional inaccuracies outside their original audit scope and recorded them in a spreadsheet. The complaint says a supervisor declined to see it.

The complaint then describes finance leadership questioning why auditors were reviewing data for which the organization had already been paid. According to the whistleblower, management directed a stop to corrections while risk scores were falling. Those allegations of compliance failures after internal warnings connect the work on the charts to the decisions above the coders’ desks.

The complaint also alleges a change that could look like correction from inside the wrong file. According to the whistleblower, auditors were told to change only the billing side of the electronic record, leaving physicians to correct the patient encounter. The whistleblower said that changing the billing file alone would leave inaccurate diagnoses in the clinical records used to generate Medicare Advantage payment data.

That is why the file changed matters as much as the task marked complete. An audit report, a management instruction and a billing edit can describe three different outcomes for the same diagnosis. The payment trail connects them.

OIG Measured How Far the Reviews Leaned

The HHS Office of Inspector General’s study shows how heavily chart reviews leaned toward additions:

Over 99 percent

of chart reviews in OIG’s analysis of 2016 Medicare Advantage encounter data added diagnoses.

HHS Office of Inspector General, Billions in Estimated Medicare Advantage Payments From Chart Reviews Raise Concerns, December 10, 2019

The report estimated $6.7 billion in 2017 payments from diagnoses reported only on chart reviews, with no corresponding diagnosis on a service record. Its concerns included whether diagnoses were valid and whether beneficiaries received care for the conditions reported.

The HHS Office of Inspector General’s 2021 follow-up found the payments were concentrated. One company accounted for 40 percent of payments from diagnoses reported only on chart reviews and health risk assessments while enrolling 22 percent of Medicare Advantage beneficiaries. Company size did not explain that difference.

The 2024 evaluation examined assessments and chart reviews linked to them. Diagnoses found only through those sources in 2022 generated an estimated $7.5 billion in 2023 payments. Missing follow-up service records left the inspector general with a serious question: were the diagnoses inaccurate, or were patients going without care for the conditions the plans reported?

Chart reviews that identify no encounter, known as unlinked chart review records, no longer count toward risk scores under the final payment announcement for 2027, apart from beneficiaries switching between Medicare Advantage organizations.

For one-way chart reviews, the evidence needed to report Medicare fraud has to show what the review found and what the organization did with it. Our whistleblower experts work with you to connect those findings to the decisions and payments behind them.

A coder may know which findings were dropped. An auditor may know who received the error report. A vendor employee may know which outputs the contract required. A plan’s risk-adjustment staff may know whether the correction reached CMS. Connecting those accounts is where we come in.

A whistleblower case under the False Claims Act can pay 15% to 30% of what the government recovers. In the HealthCare Partners matter, the Justice Department announced a $10.2 million whistleblower reward specifically for the one-way chart-review allegations. The fate of the deletion findings mattered enough to support a case of its own.

Tell Us Where the Corrections Stopped

Speak with our whistleblower experts if you know why review findings that would reduce payment disappeared. We connect the coding decisions, the payment trail and the people who knew, and bring in Medicare Advantage whistleblower lawyers who understand these cases. We stay beside you as that inside knowledge becomes the strongest case the evidence can support.