Contents
Contents
PACE risk adjustment models change Medicare payment through diagnosis-based scores and a separate frailty adjustment. If you work in coding, clinical operations or finance, the diagnosis you question may sit in a chart, a submission file or a payment calculation. Those records show different parts of the story. This page follows the calculation from the clinical record and survey response to the model output, paid month and later adjustment.
PACE organizations receive monthly payments in advance and assume full financial risk for an integrated care benefit. Medicare and Medicaid provide separate payment streams, and both can support the same participant. The Medicaid monthly amount and enrollment reconciliation follow the organization’s agreement with the state. That arrangement puts a great deal of weight on the information used to calculate the monthly payment: the organization receives the money while remaining responsible for the care.
PACE uses its own blend of Medicare risk models for calendar year 2026. Other Medicare Advantage organizations use the 2024 CMS-HCC model for the full non-ESRD risk score, while PACE combines an older model with that newer one. CMS-HCC refers to the hierarchical condition categories used in the diagnosis-based calculation. Coding teams will recognize the 2017 model as V22 and the 2024 model as V28.
The CMS implementation memorandum distinguishes the non-ESRD calculation from the ESRD calculation. The annual payment announcement applies a corresponding blend to the frailty adjustment. The weights apply to calculated scores, rather than to a count of diagnoses.
| PACE Payment Component for CY2026 | Older Component | Newer Component |
|---|---|---|
| Non-ESRD Parts A and B risk score | 90% of the 2017 CMS-HCC score | 10% of the 2024 CMS-HCC score |
| End-stage renal disease risk score | 90% of the 2019 ESRD score | 10% of the 2023 ESRD score |
| Frailty adjustment | 90% of the score using the 2017 model’s frailty factors | 10% of the score using the 2024 model’s frailty factors |
For non-ESRD Parts A and B, CMS’s historical PACE payment manual describes a county rate multiplied by the sum of an individual risk score and an organization frailty score. That architecture explains why a participant’s diagnoses and the organization’s survey results enter different parts of the payment. The annual announcements supply the models and factors for the particular payment year.
The ESRD route has its own models and factor types for dialysis, transplant and post-graft status. CMS’s PACE factor-type list distinguishes community from institutional dialysis and full dual eligibility from partial or non-dual eligibility. It also identifies new-enrollee dialysis separately. For an insider comparing reports, those participant categories identify which calculation the payment used. They belong beside the diagnoses when a score changes between model runs.
Encounter submissions feed the PACE risk adjustment model through the diagnosis sources CMS allows for each component. For the older CY2026 components, CMS pools eligible diagnoses from the Risk Adjustment Processing System, known as RAPS, encounter data and fee-for-service claims. The newer components use encounter data and fee-for-service claims only.
This distinction matters inside a PACE organization moving its data into the Encounter Data System, or EDS. A diagnosis submitted through RAPS reaches a different source pool from one included in eligible encounter data. We follow the actual outbound record and the model that used it when assessing a disputed payment.
The transition also changes how the submission is screened. EDS uses procedure-code filtering, while RAPS uses specialty filtering. The source record, submission format and filtering result together explain how a diagnosis reached the calculation. A file’s acceptance message answers a narrower question than the clinical support for the diagnosis or its effect on payment.
For services provided in a PACE center that generate no provider claim, CMS permits either an encounter data record or an unlinked chart-review record. The PACE encounter-submission instruction in CMS’s January 2024 archive limits that option to services at the center. Where there is a provider claim, the organization still needs an encounter record.
Diagnoses added through that process must come from a face-to-face visit, have medical-record support and meet the risk-adjustment criteria. For a coder reviewing a disputed addition, the setting and the underlying visit therefore matter alongside the code itself.
The annual announcement also explains why CMS tied the model transition to encounter submissions: moving fully to the updated model requires all diagnoses to reach EDS. CMS says an organization can stop submitting to RAPS once it submits complete diagnostic data to EDS. For an insider, the useful comparison is between what the clinical record supports and what the organization actually sent through each system.
PACE frailty scores use activities-of-daily-living information from the Health Outcomes Survey-Modified, or HOS-M. The survey responses feed a contract-level frailty score. The care team’s clinical assessment and the participant’s chart can describe the same limitations while serving different purposes from the survey used for payment.
The historical PACE payment manual limits the frailty add-on to community and short-term institutional participants and assigns zero frailty adjustment to long-term institutional participants. We read that distinction alongside the applicable annual model instructions when tracing a participant’s payment. The contract’s survey score and the participant’s status both matter.
CMS explains that frailty factors account for additional costs the diagnosis-based model does not explain. That makes the survey consequential even where the chart already describes extensive health care needs. A payment review connects the survey input to the contract’s factor and then to the applicable payment year.
The 2026 HOS-M administration memorandum places survey administration from late July through November. Vendors may not give organizations patient-level reports or data during collection or between administrations. Someone who works with the survey may know which responses, access permissions or vendor communications deserve closer examination.
CMS’s annual announcement also addresses proxy assistance. A proxy response is at the PACE participant’s discretion, and staff may inform a family member or caregiver of the right to request a proxy when a participant with dementia needs help completing the survey.
We examine the survey response, any proxy involvement, the contract-level score and the payment year together. The survey’s collection date identifies one stage in that sequence. The factor used in payment supplies the connection to Medicare money.
CMS uses diagnosis data from defined service periods for the initial, midyear and final calculations. Community or long-term institutional status and low-income status can change the factors used in those calculations.
A finance employee examining a correction can compare the accepted diagnosis data, factor type and model run with the adjustment recorded by the payer. We follow those records to the money the organization actually received or returned.
The Monthly Membership Report and Model Output Reports connect the calculated factors with the condition categories used for payment. The Monthly Membership Report, often called the MMR, shows both the risk-adjustment factor type and its numerical value.
The distinction matters when someone points to a code on a payment report as evidence of a changed score. The factor type identifies the category used for the payment or adjustment. The factor values carry the numerical result. Both belong in the comparison with the underlying diagnosis data.
CMS distributes separate Part C and Part D model-output files. Their Model Output Reports, or MORs, identify the HCCs or RxHCCs used to calculate each enrolled participant’s risk score. Separate report types show which model and source pool each score used. They connect a questioned input to the particular calculation CMS ran.
We use that connection when reviewing unsupported diagnoses or corrections that stopped inside the organization. The clinical change, the outbound correction, its acceptance, the model output and the payment adjustment each answer a different question. A correction has to travel farther than the chart to reach the payment.
PACE’s CY2026 Part D risk score blends two calibrations that CMS calls the 2026 RxHCC model. The newer component is 10% of the score from the 2022/2023 calibration, using encounter and fee-for-service diagnoses. The older component is 90% of the score from the 2018/2019 calibration, also drawing on RAPS.
Those calibration pairs identify the diagnosis and expenditure years used to build the models. Calling both components the 2026 model leaves out the distinction the payment calculation needs.
The drug-payment trail extends beyond the risk score to bids, prescription drug event data and reconciliation. In its final CY2026 Part D instructions, CMS says PACE organizations must account for the selected drug subsidy in both their bid estimates and their prescription drug event records. CMS reconciles prospective subsidy payments against the actual reported costs and adjusts the payment accordingly.
PACE also has its own drug-data field instructions. Under the PDE reporting guidance, the Patient Pay Amount field must contain zero for service dates beginning January 1, 2025, because PACE participants have no cost sharing. For someone working with pharmacy data, a familiar field name can carry a PACE-specific value.
The certification reaches the people responsible for the data. CMS’s August 2026 attestation memorandum required the chief executive, chief financial or chief operating officer to sign electronically for the prior contract year’s data. CMS expressly tied the signed attestations to receiving risk-sharing and reinsurance payment adjustments.
The PDE/DIR attestation covers the accuracy, completeness and truthfulness of data used for federal reimbursement. CMS also provides for an attachment describing a known or potential PDE error where the sponsor has already notified CMS and is working to correct it. The attachment describes the error, its magnitude and the expected correction timetable. We examine the signed statement, any error attachment and the resulting reconciliation together when an insider questions what management certified.
The requirement includes PACE organizations even when their Summary DIR submission reports zero dollars. CMS permits one parent-level attestation to cover multiple contracts. An insider who knows which contract’s data carried an error can connect that knowledge to the statement signed above it and the notification already sent to CMS.
CMS’s final CY2027 announcement moves PACE’s non-ESRD risk score and corresponding frailty score to a 50% share from the 2017 model and a 50% share from the 2024 model. The ESRD calculation likewise divides evenly between the 2019 and 2023 models. These are rules for the upcoming payment year, separate from the calculation for CY2026 paid months.
The future Part D blend uses 50% of the 2018/2019 calibration and 50% of the 2023/2024 calibration, with Medicare Advantage prescription drug relative factors for the newer portion. Only encounter and fee-for-service diagnoses feed the newer portion. The older portion also includes RAPS, alongside encounters and fee-for-service claims.
For CY2027, CMS also excludes diagnoses identified as coming from audio-only services through modifiers 93 or FQ. That exclusion applies to PACE’s medical and drug risk calculations. The general exclusion of diagnoses from unlinked chart-review records does not apply to PACE for that year, preserving the importance of the program’s own submission rules.
The useful payment trail connects what the clinical record supports, what the organization submitted, what CMS used and what the organization received. For a whistleblower inside coding or finance, knowledge of the decisions between those records can be as important as the records themselves. Who approved the submission, who received the correction request and what happened afterward are questions we work through with you.
A clinical correction belongs beside the coding decision and outbound file. An accepted deletion belongs beside the model output and the later payment adjustment. We follow the sequence through reconciliation to identify the payment effect and management’s response to the underlying problem. A whistleblower case under the False Claims Act can pay 15% to 30% of what the government recovers.
If you know why a PACE diagnosis stayed in the payment data after someone challenged it, talk with our whistleblower experts about the code, the correction and the money. We work beside you to build the payment evidence into your strongest whistleblower case and bring in healthcare fraud whistleblower lawyers who understand risk-adjustment submissions. You bring the knowledge of what happened inside; we bring the expertise to pursue it with you.