Skip to main content

Free billing screening

Get screening →
Resource

The V28 Risk Adjustment Model (2024 CMS-HCC), Explained

By · Published

V28 is the common name for the 2024 CMS-HCC risk adjustment model, the formula CMS uses to adjust Medicare Advantage payments for each member's expected cost. It has 115 payment HCCs, up from 86, and 2,236 diagnosis codes that counted toward payment under the 2020 model no longer do. It has set 100% of risk scores for non-PACE plans since 2026, and CMS kept it for 2027 while it stopped counting diagnoses from unlinked chart reviews and audio-only visits. This guide sets out what changed, year by year, and what each change means for the documentation a practice produces.

Quick Answer

What is the V28 risk adjustment model?

V28 is the 2024 CMS-HCC model, which CMS finalized in the CY 2024 Rate Announcement and phased in for Medicare Advantage organizations over three years: 33% of the risk score in 2024, 67% in 2025 and 100% in 2026. It is calibrated on 2018 diagnoses and 2019 expenditures, has 115 payment HCCs out of 266, and maps 7,770 ICD-10 codes to payment HCCs against 9,797 in the 2020 model. For 2027, CMS kept it rather than adopt the recalibrated version it had proposed, and stopped counting diagnoses from unlinked chart review records and audio-only encounters.

  • 115 payment HCCs, up from 86 in the 2020 model
  • Phase-in for non-PACE plans: 33% in 2024, 67% in 2025, 100% in 2026
  • 2027: same model; unlinked chart-review and audio-only diagnoses excluded
  • RADV: CMS announced audits of about 550 MA plans a year (May 2025)

V28 at a Glance: the 2020 and 2024 CMS-HCC Models Compared

"V24" and "V28" are the version labels coders and vendors use. CMS's own documents call them the 2020 CMS-HCC model and the 2024 CMS-HCC model.

2020 CMS-HCC model ("V24")2024 CMS-HCC model ("V28")
Calibration data2014 diagnoses, 2015 expenditures2018 diagnoses, 2019 expenditures
Payment HCCs86115 (of 266 HCCs; the other 151 are non-payment)
ICD-10 codes that map to a payment HCC9,797 (13% of ICD-10 codes)7,770 (10.5% of ICD-10 codes)
Codes that no longer map to a payment HCC—2,236: 2,161 from ICD-10 clinical updates and 75 from changes aimed at discretionary coding

Source: CMS fact sheet on the 2024 Medicare Advantage and Part D Rate Announcement, 31 March 2023.

CMS gave two reasons for the rebuild. It moved the model onto newer data and reclassified the conditions, and its fact sheet says "The goal of updating the underlying data and the clinical reclassification was to improve predictive ability." It also points out that "ICD-10 has a much larger code set, which provides more specificity." The practical result for coders is that HCC numbers and mappings from the 2020 model do not carry over. Protein-calorie malnutrition, for example, was HCC 21 in the 2020 model and is HCC 47 in the 2024 model, and angina pectoris moved from HCC 88 to HCC 230, so a crosswalk built on V24 numbering points at the wrong categories.

How the V28 Phase-In Ran, Year by Year

CMS blended the two models' risk scores for Medicare Advantage organizations other than PACE over three years, then dropped the old model.

Payment yearRisk score blend for non-PACE MA organizationsSource
202433% 2024 model, 67% 2020 modelCY 2024 fact sheet
202567% 2024 model, 33% 2020 modelCY 2025 fact sheet
2026100% 2024 modelCY 2026 fact sheet
2027100% 2024 model, still calibrated on 2018 diagnoses and 2019 expendituresCY 2027 fact sheet

PACE organizations are on a slower blend with the older 2017 CMS-HCC model: 10% 2024 model and 90% 2017 model in 2026, and 50% of each in 2027.

For 2027, CMS had proposed an updated model calibrated on 2023 diagnoses and 2024 expenditures. In the final CY 2027 Rate Announcement of 6 April 2026 it kept the 2024 model instead, to give the MA market more time to adjust to the phase-in it had just completed. So the model that set 2026 risk scores also sets 2027 risk scores.

What Changed Inside the Model: Removed, Constrained and Narrowed HCCs

Beyond the renumbering, CMS made targeted changes to conditions it found were coded inconsistently. Its fact sheet says CMS "removed certain codes where there is wider variation in diagnosing and coding," and that "These diagnoses don’t predict costs as well as a result of the variation in their coding." The full text of the CY 2024 Rate Announcement names the main ones.

Moved out of the payment model

  • Protein-calorie malnutrition (HCC 47). Removed from payment. CMS: "Empirical data show that coding for malnutrition is variable," and removing it "allows expenditures to flow to the well-specified underlying conditions," such as cancer and HIV.
  • Angina pectoris (HCC 230). Angina pectoris now maps to a non-payment HCC; in CMS's words, "the more variable manifestation is now mapped to a non-payment HCC." Unstable angina stays in a payment HCC.
  • Atherosclerosis of the arteries of the extremities with intermittent claudication (HCC 265). A new non-payment HCC, because CMS found "it is not a reliable predictor of prospective medical expenditures."

Constrained to equal weights

  • Diabetes: all three diabetes payment HCCs (36, 37 and 38) are constrained "such that each HCC carries the same weight." The 2020 model constrained only its top two diabetes HCCs. Moving a member from one of these diabetes HCCs to another no longer changes the score by itself.
  • Congestive heart failure: the 2024 model splits heart failure into five HCCs, and HCCs 224, 225 and 226 are constrained the same way. CMS: "The most severe manifestation is left unconstrained, while the less severe split outs of the HCC are constrained."

Depression narrowed

Major depression counts for payment only at moderate or severe levels: the 2024 payment model is "limited to ICD-10 codes specifying moderate or severe major depression." A note that supports only an unspecified or mild depression code does not reach a payment HCC.

What Changed for 2027: Which Diagnoses Count

The bigger 2027 change is not the model. It is which diagnoses CMS will accept from MA organizations' risk adjustment data. The final CY 2027 Rate Announcement and its fact sheet:

  • Exclude diagnoses from unlinked chart review records. A chart review record is unlinked when the MA organization does not tie it to a previously submitted encounter data record, so it is not clear which service produced the diagnosis. The one exception: unlinked chart review diagnoses still count for members who switch from one MA organization to another. CMS puts the impact at -1.53%, or -1.78% without the exception.
  • Exclude diagnoses from audio-only encounters, which CMS identifies by modifier 93 or FQ. CMS estimates the average impact at 0%.
  • Show a -1.12% impact for the risk model and normalization, which CMS says is entirely due to the normalization factor update, because no new model was adopted.
  • Project average MA payments up 2.48% before risk score trend, or 4.98% once CMS's expected average risk score increase of 2.50% is counted.

Because the model is prospective (next section), 2027 risk scores are built from 2026 dates of service. So the exclusions reach diagnoses recorded during 2026. A condition captured only on an audio-only call does not count toward the member's 2027 risk score, and neither does one a plan added only through an unlinked chart review, unless the member switched MA organizations.

Why Chronic Conditions Need Documenting Every Year

The CY 2027 Rate Announcement states that "CMS-HCC risk adjustment models are prospective, meaning that they use diagnostic information from a base year" to predict costs for the following year. A condition documented and coded in 2026 informs the member's 2027 payment. A condition that is managed but never documented and coded during a year contributes nothing to the next one.

That is why chronic conditions have to be documented every calendar year, not once, and why the source of the diagnosis now matters as much as the code. For practices in Medicare Advantage value-based or shared-risk contracts, this is the link between the visit note and the plan's revenue; our Medicare Advantage billing page covers how MA claims and HCC recapture run alongside each other.

RADV Audits: Why the Record Has to Support Every HCC

Risk adjustment data validation (RADV) audits check that the diagnoses an MA organization submitted for payment are supported by the medical record. On 21 May 2025 CMS announced that, in newly initiated audits, it would:

  • audit all eligible MA contracts for each payment year, moving from about 60 plans a year to about 550;
  • review between 35 and 200 records per plan per year, depending on plan size, instead of 35;
  • grow its medical coder team from 40 to about 2,000 by 1 September 2025;
  • complete all remaining RADV audits for payment years 2018 to 2024 by early 2026.

How far an audit's findings reach is less settled. CMS's 2023 RADV final rule (CMS-4185-F2) would extrapolate sample findings across a contract from payment year 2018 without a fee-for-service adjuster, but a federal district court vacated that rule on 25 September 2025 (Humana Inc. v. Becerra, N.D. Tex.), and CMS appealed to the Fifth Circuit. Either way, a diagnosis the record does not support is removed on review. The plan carries the payment risk, but the evidence it needs is the practice's note.

Our risk adjustment coding services code HCCs to what the record supports, remove unsupported HCCs as readily as they recapture supported ones, and organize the records behind each sampled HCC when a RADV audit pulls them. For the audit side of coding generally, see the medical coding audit guide.

A V28 Documentation Checklist for Practices

Each item follows from the CMS rules above. None is a coding guarantee, and none replaces the ICD-10-CM Official Guidelines.

  1. Document and code each chronic condition you manage at least once every calendar year, at a visit that is not audio-only. Diagnoses from audio-only encounters stop counting for 2027 payment, which is built from 2026 dates of service.
  2. Document the severity of major depression when the clinical picture supports it. Only moderate and severe major depression codes are in the payment model.
  3. Do not rely on HCC 47 (protein-calorie malnutrition), HCC 230 (angina pectoris) or HCC 265 (claudication) to carry weight. For malnutrition, CMS's stated aim is for payment to flow to the well-specified underlying condition, such as cancer or HIV, so document that condition fully.
  4. Code diabetes and heart failure to the specificity the note supports. Within each constrained group the weight is now the same, so the gain from a more specific code is accuracy, not a higher score.
  5. Make sure every diagnosis the plan submits traces to an encounter. For members who stay with the same plan, diagnoses that live only in an unlinked chart review stop counting for 2027 payment.
  6. Keep the record behind every HCC retrievable. A RADV sample is checked against the medical record, not against the claim.

Where note templates don't prompt for severity, specificity or the status of each chronic condition, that gap is a documentation problem before it is a coding one; our clinical documentation improvement services review the record and query the provider when a note is ambiguous or incomplete.

Free Billing Screening · No obligation

Get HCC Capture Coded on the Current V28 Model

Tell us your Medicare Advantage panel and chart volume. A free billing screening reviews a permitted 12-month practice-management export; a chart-based risk adjustment coding review is scoped and quoted in writing before any charts are pulled.

Prefer to talk? Book a 15-minute call

A free screening reviews a permitted 12-month practice-management export and provides a short written memo within five business days after receipt. It includes no chart review, coding opinion or extrapolated recovery figure. Chart-based audits are separately scoped and quoted.

Solo provider or group practice?

Practice details only · No obligation to continue · We reply within 1 business day

Common Questions

Common questions about the v28 risk adjustment model (2024 cms-hcc).

Get a Free Billing Screening

Our billing specialists can walk you through this and more.

Request a Free Screening

What is V28 in risk adjustment?

V28 is the common name for the 2024 CMS-HCC model, which CMS uses to risk-adjust Medicare Advantage payments. It has 115 payment HCCs and is calibrated on 2018 diagnoses and 2019 Original Medicare expenditures. CMS's own documents call it the 2024 CMS-HCC model rather than V28.

What is the difference between V24 and V28?

V24 is the 2020 CMS-HCC model and V28 is the 2024 model. V28 is calibrated on newer data (2018 diagnoses and 2019 expenditures, against 2014 and 2015), has 115 payment HCCs against 86, and maps 7,770 ICD-10 codes to payment HCCs against 9,797. It also removes or constrains conditions CMS found were coded inconsistently, including protein-calorie malnutrition, angina pectoris and the diabetes and heart failure groups.

When did V28 take full effect?

For non-PACE Medicare Advantage organizations, the 2024 model was 33% of the risk score in 2024, 67% in 2025 and 100% in 2026. PACE organizations are on a separate, slower blend with the 2017 model: 10% 2024 model in 2026 and 50% in 2027.

Is V28 still used in 2027?

Yes. In the CY 2027 Rate Announcement CMS kept the 2024 model rather than the recalibrated version it had proposed, which would have used 2023 diagnoses and 2024 expenditures. It also excluded diagnoses from unlinked chart review records, with an exception for members who switch MA organizations, and from audio-only encounters.

How many HCCs are in V28?

266 in total. 115 are in the payment model and 151 are not. The 2020 model had 86 payment HCCs.

Which conditions were removed from payment in V28?

Among the changes CMS named in the CY 2024 Rate Announcement: protein-calorie malnutrition (HCC 47) was removed from the payment model, angina pectoris now maps to non-payment HCC 230 while unstable angina stays in payment, and atherosclerosis of the extremities with intermittent claudication is non-payment HCC 265. Diabetes HCCs 36 to 38 and heart failure HCCs 224 to 226 were constrained to equal weights, and depression in the payment model is limited to moderate or severe major depression.

Does V28 change what a practice is paid?

V28 sets Medicare Advantage plan payments, not fee-for-service claim payments. It reaches a practice through value-based and shared-risk contracts with MA plans, whose terms vary, and through the documentation plans and CMS auditors review. Our Medicare Advantage billing page covers the claim side of MA plans.

Free billing screening

Get HCC Capture Coded on the Current V28 Model

Tell us your Medicare Advantage panel and chart volume. A free billing screening reviews a permitted 12-month practice-management export; a chart-based risk adjustment coding review is scoped and quoted in writing before any charts are pulled.

A free screening reviews a permitted 12-month practice-management export and provides a short written memo within five business days after receipt. It includes no chart review, coding opinion or extrapolated recovery figure. Chart-based audits are separately scoped and quoted.

  • No obligation to continue
  • Practice details only
  • Reply within 1 business day
Call us Free screening