What V28 is

V28 is the 2024 CMS-HCC risk adjustment model. CMS-HCC stands for the Centers for Medicare and Medicaid Services Hierarchical Condition Categories, the framework that translates a member's documented diagnoses into a risk score used to set Medicare Advantage (MA) payment. V28 is the successor to V24, the prior model version.

Two structural numbers define V28. It expands the HCC category set to 115, and it reduces the ICD-10-to-HCC crosswalk to about 7,770 codes, down from roughly 9,700 codes under V24. More categories, but fewer diagnosis codes map into them. A diagnosis that triggered an HCC under V24 may no longer map to any HCC under V28.

V28 also introduces or tightens constrained groups, where multiple related conditions share a single coefficient, and it lowers many individual coefficients relative to V24.

Why V28 matters

V28 changes how much revenue a given panel of patients generates, without any change in the patients themselves. CMS's Office of the Actuary projected an aggregate average MA risk score reduction of about 3.12% under V28. At the plan level, the realized impact runs between 2 and 6% depending on member mix and prior coding intensity. For a risk-bearing organization, a few points of risk score compression flows straight through to capitated revenue.

The operational stakes are just as concrete. Capture playbooks tuned to V24 mappings will quietly lose ground as the V28 blend takes over, because diagnoses that no longer map produce no score regardless of how well they are documented. Teams that do not re-run their populations under V28 mappings tend to discover the gap only after the submission window has closed.

How V28 is phased in and calculated

CMS did not switch from V24 to V28 in a single year. It blended the two models across three payment years: about one-third V28 in 2024, about two-thirds V28 in 2025, and full V28 in 2026. A plan's risk score in the transition years is a weighted combination of its V24 score and its V28 score.

The mechanics that drive the score difference are:

  • Category expansion. The set grows to 115 HCCs, with several conditions split into more granular tiers.
  • Crosswalk reduction. The ICD-10-to-HCC map shrinks to about 7,770 codes. Codes that previously mapped may now map to nothing.
  • Coefficient changes. Many coefficients are lowered, which is the main driver of the projected ~3.12% average reduction.
  • Constrained groups. Related conditions share one coefficient, so documenting additional members of the group does not raise the score beyond the shared value.

Common mistakes teams make with V28

  • Running V24 capture logic into 2026. Worklists built on the V24 crosswalk chase diagnoses that no longer map under full V28.
  • Ignoring constrained groups. Coding three related conditions in a constrained group when only the shared coefficient counts wastes encounter time and adds audit surface with no revenue.
  • Treating the phase-in as a single cliff. The blend shifted each year, so the impact in 2024 understated what arrives in 2026.
  • Not re-baselining documentation. Notes that supported a diagnosis under V24's coefficient logic may not survive a V28-era review.
  • Estimating with the 3.12% average instead of plan-level analysis. Plan-level impact spans 2 to 6%, and higher-acuity or dual-eligible panels tend to sit at the high end.

How Pelica handles V28

Pelica's Risk Adjustment Copilot maps every member's diagnoses under V28, surfaces at-risk HCCs at the point of care, applies V28 trumping and constrained-group logic in real time, and tracks each submitted diagnosis through the full lifecycle. Across Pelica deployments, customers have lifted RAF by roughly +0.4 in two quarters with no new headcount.

Related terms

Sources