Unsupported PACE Diagnoses and Risk Adjustment Fraud: When the Chart Changes but the Payment Does Not

An unsupported diagnosis can drive PACE risk adjustment fraud when an organization knowingly uses it to obtain or keep Medicare money. That takes taxpayer money intended for the health care of frail elderly participants. The evidence follows the diagnosis from the visit and medical record into the submitted data, the risk calculation and the payment. A corrected chart, an accepted deletion and a corrected payment are separate events. If you work in coding, clinical documentation, compliance or finance, you may see exactly where that correction stopped.

This page follows unsupported diagnosis submissions and the money affected by them. PACE’s separate frailty adjustment uses survey information about limitations in activities of daily living. That survey process has its own evidence trail. The diagnosis question begins with the encounter and ends with what Medicare paid.

Which Records Can Support a PACE Diagnosis Submission?

A PACE diagnosis submission needs an eligible source visit and medical-record support, carried through the permitted submission route. Under the Centers for Medicare & Medicaid Services’ PACE encounter-data instruction, diagnoses added through chart review must come from a face-to-face visit, have support in the medical record and meet risk-adjustment criteria. An unsupported diagnosis can enter a submission when a chart review adds a code that the source visit and medical record do not support.

The route depends on where the service occurred and whether it generated a provider claim. Provider claims require encounter data records. Services furnished in a PACE center that generate no provider claim may use an encounter data record or an unlinked chart review record. The option to use an unlinked record without an associated encounter applies to services provided in the PACE center.

That distinction matters to someone reviewing a coding queue. A center service without a provider claim has a recognized submission route, and the medical record still has to support the diagnosis. We connect the visit, the source documentation, the selected code and the record type used to send it to CMS. Each answers a different question about how the diagnosis entered the payment system.

The Encounter Data System uses procedure-code filtering, while the legacy Risk Adjustment Processing System uses specialty filtering. A submission can pass technical checks while the clinical support still needs examination. Technical acceptance tells us the system accepted the record. The source visit and medical record tell us whether the diagnosis belongs there. The model and payment reports answer what happened to the money.

Does PACE Use the Same Risk Model as Medicare Advantage?

PACE uses its own annual blend of risk models and diagnosis sources, so its payment calculation differs from the general Medicare Advantage calculation. The PACE risk adjustment model and encounter submissions must be read together for the payment year at issue. A diagnosis available to one component may reach that component through a different set of source records.

Medicare pays PACE prospective monthly capitation for entitled participants. Diagnosis risk and frailty affect the medical payment calculation, making the participant-month the relevant payment unit. For a coder, the disputed item may be a diagnosis attached to a particular encounter. For finance, the question is how that diagnosis affected the amount paid for the participant.

For ordinary medical risk scores in calendar year 2026, CMS’s implementation memorandum specifies 90% of the score calculated with the 2017 CMS-HCC model and 10% with the 2024 CMS-HCC model. The older component uses diagnoses from RAPS, encounter data and fee-for-service claims. The newer component uses encounter data and fee-for-service claims only.

The distinction is useful when an insider knows that a diagnosis went through RAPS but cannot find it in the encounter submission. The relevant question is which component used that diagnosis and what effect it had on the blended result. A model name alone leaves out the source-data question. A submission receipt alone leaves out the calculation.

CMS’s final policy for calendar year 2027 changes the PACE medical-risk blend to 50% of the older model and 50% of the newer model. The announcement also keeps diagnoses from unlinked chart review records available for PACE risk-score calculation, while excluding diagnoses identified as coming from audio-only services. Those source rules belong beside the model weights when we examine a submission.

For payments in 2024 and 2025, PACE used the 2017 CMS-HCC model and its associated frailty factors, with eligible encounter, RAPS and fee-for-service diagnoses pooled. A coding warning about those paid months belongs with that historical calculation. We match the submission and correction to the model that governed the payment, including the permitted diagnosis sources for that year.

How Does a Submitted Diagnosis Affect the Payment?

A submitted diagnosis affects payment through the eligible data, filtering and model calculation for the participant and payment period. The model’s hierarchy and output matter because the question is the additional payment attributable to the unsupported diagnosis.

The Centers for Medicare & Medicaid Services distributes Model Output Reports identifying the condition categories used to calculate each enrollee’s risk score. The Monthly Membership Report identifies the risk-factor type used for a payment or adjustment and contains the actual risk factors. Together, those reports connect the submitted diagnoses to the calculation. The payment and reconciliation records complete the connection to dollars received or returned.

What Is the Difference Between a Chart Correction, an Accepted Deletion and a Payment Correction?

A chart correction changes the clinical record, an accepted deletion changes the submitted data, and a payment correction changes the financial result. The evidence has to connect all three. A coding reviewer may know the diagnosis was removed from the chart while a finance employee sees the risk factor still used for a paid month.

The clinical correction establishes what the organization changed in its account of the patient’s condition. The coding decision and outbound correction file show whether that change reached the data sent to CMS. The accepted deletion shows what the submission system processed. Those records let us distinguish a correction that stayed in the chart from one that reached the payer.

The next question is what CMS did with the corrected data. Its risk-data accuracy memorandum describes recalculating risk scores and adjusting payments through final reconciliation when diagnoses are corrected before that deadline. It also describes later reruns for closed periods. A deletion acknowledgment and a reconciled payment therefore document different parts of the correction.

CMS uses diagnosis data from defined service periods to calculate PACE risk scores. The visit date, correction date and payment-adjustment date each have a job to do.

We match the records to the participant, payment year and model run before assessing the financial effect. The initial score, midyear update and final reconciliation can contain different information about the same participant. The question for finance is which score actually drove the payment under examination and what the later adjustment changed.

If management closes a correction after changing only the chart, we see the same diagnosis survive in the payment system. The chart got the correction. The payment data kept the diagnosis.

What Happens to an Unsupported Diagnosis After the Deadline?

Corrections that reduce overpayments remain available after the final risk-data deadline, while new diagnoses cannot be added to obtain additional payment. CMS’s accuracy memorandum describes deleting incorrect diagnosis data through RAPS or the Encounter Data Processing System before or after that deadline. The financial follow-through depends on the payment period and the correction CMS processes.

For a period whose final reconciliation deadline has closed, CMS describes periodically rerunning risk scores and adjusting payments to account for deleted diagnoses. The organization’s correction record, the later model output and the payment adjustment belong together. A closed period can therefore have a later correction trail, including evidence of what the organization knew before the money changed.

The memorandum also describes an auditable-estimate route when an organization identifies incorrect data after the final submission deadline but lacks the data needed for a RAPS or encounter-system submission. For an insider reviewing such a correction, the estimate and its supporting explanation show the payment problem the organization recognized.

The estimate belongs beside CMS’s eventual adjustment. The memo describes reruns and payment adjustments based on both closed-period deletions and auditable estimates. We follow the reported amount through that adjustment to establish what remained with the organization. An estimate, a processed deletion and money returned have distinct evidence behind them.

The Centers for Medicare & Medicaid Services’ rule incorporating PACE risk-data requirements described the change as codifying longstanding submission practice. For the insider who raised a coding warning, we examine what the organization understood, which correction it accepted as necessary and what happened to the affected payment.

What Connects an Unsupported PACE Diagnosis to a Whistleblower Case?

A healthcare fraud whistleblower case involving PACE diagnoses connects the false submission or retained overpayment to the organization’s knowledge and government money. The central records show what was represented, who understood the problem, which payment depended on the diagnosis and what correction followed. We work with the insider who can explain the decision behind those records.

A clinician or documentation reviewer may know why the source record cannot support the diagnosis. A coder may know how the disputed code reached the outbound file, and a finance employee may know which risk factor remained in use. Compliance warnings and management instructions can connect those observations to what the organization knew at the time. Together, they let us examine the clinical support, the submission decision and the financial consequence.

The amount at issue follows the diagnosis through the model and reconciliation. We establish the additional payment or the known amount left unreturned, tied to the responsible organization and payment period. That is the connection between a coding concern and a case about taxpayer money.

A whistleblower case under the False Claims Act can pay 15% to 30% of what the government recovers. We assess the submission history, payment effect and decisions inside the organization with you, then bring in whistleblower lawyers who understand PACE risk-adjustment cases.

When You Know Where the Correction Stopped

If you have seen an unsupported PACE diagnosis survive a clinical correction and remain in the payment data, speak with our whistleblower experts about the submission and the money. We build the case with you across the clinical, coding and finance records, bring in PACE fraud whistleblower lawyers and stay beside you through the process.