Medicare Advantage addendum fraud uses later medical-record changes to claim payment for diagnoses unsupported by the original encounter. A physician query can ask you to put a condition into a visit note months after you saw the patient. The government alleged that some added conditions had nothing to do with the visit. For the physician receiving the request or the coder processing it, the important comparison is between what happened during care and what the amended record later told Medicare.
This form of risk adjustment fraud puts the treating provider’s judgment at the center of a payment process. Who asked for the change, who approved it and what happened after an objection can explain how an unsupported diagnosis became revenue.
Medicare Advantage plans receive monthly payments that reflect their enrollees’ expected health care costs. Diagnosis codes feed risk scores, and a higher score can increase the plan’s payment. The path from diagnosis to revenue runs through the provider’s documentation, the plan’s submission and the payment calculation.
Chart reviewers look back through existing records for diagnoses. A health risk assessment, often conducted at home by plan or vendor staff, supplies another source. A retrospective amendment changes the documentation attached to an earlier encounter.
The financial pull is visible across chart review. In its review of the industry’s practices, the HHS Office of Inspector General found that over 99 percent of the chart reviews it examined added diagnoses. Those were reviews from 2016. The failure to remove unsupported codes completes the payment problem: additions raise revenue while errors already favoring the plan remain.
A retrospective physician query asks the provider to revisit an earlier encounter and document a diagnosis. In the alleged schemes, the request arrives with the condition already selected. The physician’s response can give that selection the appearance of a clinical finding from the original visit.
The Justice Department’s Kaiser Permanente settlement announcement describes the Centers for Medicare & Medicaid Services’ outpatient requirement plainly: a diagnosis must have support in the medical record of a face-to-face visit and must have required or affected care, treatment or management at that visit. That makes the original note central to the inquiry. An old problem-list entry, a later request and the treating provider’s assessment each tell a different part of the story.
The Inspector General’s February 2026 compliance guidance expressly identifies electronic medical record queries, including prompts generated by artificial intelligence, as a way to solicit diagnoses patients did not have or that did not affect their care. The agency’s concern is the diagnosis the prompt seeks and its connection to the visit. The same comparison follows the request whether a person selected the condition or an algorithm suggested it.
The Kaiser allegations concerned pressure on physicians to change their records. The Villages allegations reached changes whose initiation, timing or approval failed the requirements described in the settlement. Both put the original encounter and the later amendment side by side.
Kaiser agreed to a $556 million settlement announced in January 2026. The Justice Department said the whistleblowers would receive $95 million of the recovery.
The government alleged that Kaiser mined patients’ medical histories for potential diagnoses that had not reached Medicare, then sent queries urging providers to add them to earlier visit records. Some requests came months after the encounter, and some came more than a year later.
The government alleged that physicians had neither considered nor addressed the added conditions during the visits. In that account, taxpayers paid for diagnoses the doctors had not addressed during care.
The Justice Department alleged that Kaiser tied physician and facility incentives to diagnosis goals and singled out those who fell short. Those risk-score targets and financial incentives explain the pressure behind the query. The changed note shows what the pressure produced.
The department also alleged that Kaiser ignored physicians’ warnings and compliance audits identifying inappropriate amendments. The objection therefore matters alongside the signature: it can show that someone challenged the practice before the organization continued using it.
The Villages Health System, a Florida healthcare provider group, agreed to a $541.5 million settlement announced in August 2026. The government alleged that the group sent false diagnosis codes to Medicare Advantage organizations, which submitted them to Medicare and received inflated payments.
The government alleged that codes lacked support in the chart or came from defective amendments. Some changes were both late and not initiated by the treating provider; others lacked the provider’s approval. For the clinician, those details make the amendment’s origin and approval history as important as its final wording.
The incentive can reach the clinic: under risk-sharing arrangements, provider groups can receive a share of the plan’s payment, so a higher risk score can increase the group’s revenue too.
The announcement also said the receiving plans were returning overpayments through code deletions or agreements to return funds. The allegedly false codes had traveled into another organization’s payment data.
Independent Health’s vendor DxID allegedly brought the proposed diagnoses to the physician. The government’s complaint concerning its retrospective review program alleged that DxID used suggestive forms to obtain approval for conditions that providers supposedly assessed but had not adequately documented.
The complaint alleged that, when providers did not complete the forms themselves, DxID staff visited their offices, reviewed records, checked off diagnoses and presented prefilled forms for signature. The filing reproduces what one form said about the visit:
This is an addendum to the original encounter note dated 9/13/2012. It serves to further clarify and detail the diagnosis and conditions that existed at the time of this visit. These diagnosis [sic] coexisted and required or affected the care, treatment, or management of [Patient] at the time of this encounter.
According to the complaint, the form had already written the conclusion that made the diagnosis useful for payment. The physician was being asked to sign it. The complaint challenged the connection to care that the form asserted.
The same complaint alleged that reviewers captured conditions from old problem lists and laboratory results without evidence that the treating provider considered them during care. It also quoted staff reporting that physicians often signed from memory without reviewing records.
For retrospective amendments, the evidence needed to report Medicare fraud connects the original visit, the later query, the provider’s response and the diagnosis submitted for payment. We examine that sequence with the physician, coder or risk-adjustment employee who understands how the organization handled it.
The original signed note supplies the account of the visit before the amendment. The query shows the requested diagnosis and who requested it. The amendment history supplies the later wording, its author and date, and the provider’s approval. Reading those records together lets us distinguish the person who proposed a condition from the clinician whose judgment supposedly supported it.
The DxID complaint alleged that Independent Health withdrew kidney-disease codes for an earlier service year after receiving notice of the problem, but left later codes in place and continued the same practice. It also alleged that the company temporarily stopped the vendor’s work over amendment concerns, then restarted it. The later submission and the response to an earlier warning belong in the same account.
In retrospective amendment programs, the red flags for Medicare fraud include queries detached from the original visit, preselected diagnoses, missing provider approval and objections followed by continued submissions. The strongest connection is a dated contradiction between what the organization learned and what it did afterward.
A physician may know why a requested condition did not belong to the visit. A coder may know which code was added. A compliance employee may know who received the warning. Together, those accounts can explain why a documented problem survived internal review.
The visit date identifies the care supporting the code. The payment year identifies when Medicare paid on it. Keeping both visible prevents a later submission from obscuring the encounter it supposedly describes.
An Inspector General audit of Anthem Community Insurance Company, published in May 2021, examined seven groups of high-risk diagnosis codes. The service years were 2014 and 2015, and the corresponding payment years were 2015 and 2016. The Inspector General estimated at least $3.47 million in net overpayments for the audited high-risk codes across the audit period.
The reviewers connected the medical record to the diagnosis category, recalculated the risk score and determined the payment difference. That connection is also what makes an amendment’s history financially meaningful.
A 2019 study in Health Management Policy and Innovation concluded that upcoding may have overstated how much riskier Medicare Advantage enrollees looked than fee-for-service patients, raising plan payments, the pattern behind coding intensity in Medicare Advantage.
For a particular query, the clinical question remains specific: what supported the condition at that visit? The original note, the later form and the provider’s response supply the detail behind the broader payment pattern.
The Kaiser settlement over unsupported diagnoses and The Villages settlement over a provider group’s false codes show how the False Claims Act applies to Medicare Advantage.
A whistleblower case under the False Claims Act can pay 15% to 30% of what the government recovers. For physicians and coders, the value of their knowledge lies in explaining how the requested diagnosis departed from the encounter and how the organization handled that difference.
If you know how a later diagnosis request changed an earlier encounter, speak with our whistleblower experts about the request and what followed. We work with you to build the case around the clinical contradiction, the approval trail and the payment it produced. We bring in Medicare Advantage fraud whistleblower lawyers who understand retrospective coding and remain your partners as the case develops.