CMS-HCC V24 vs V28: Complete Model Comparison for Risk Adjustment Teams

Why the V24 to V28 Transition Matters

CMS-HCC V28 fully replaced V24 as the sole risk adjustment model for Medicare Advantage capitation payments in 2026. This was not a minor update — it was a structural overhaul of how Medicare Advantage plans are compensated for the clinical complexity of their enrolled populations.

The transition followed a three-year phase-in: 67% V24 / 33% V28 in payment year 2024, 33% V24 / 67% V28 in 2025, and 100% V28 starting in 2026. Each step of the blend shifted more revenue weight toward V28's mechanics, giving plans a window to adapt their documentation, coding, and analytics workflows.

For risk adjustment teams, the distinction between V24 and V28 is not academic. Every difference in HCC structure, coefficient methodology, and ICD-10 mapping directly affects RAF scores, revenue forecasts, and operational priorities.

HCC Count and Structure

V24 included 86 Hierarchical Condition Categories. V28 expanded this to 115 HCCs organized into 26 disease families. The increase reflects several structural decisions by CMS:

• Category splits: Several V24 HCCs were divided into more granular categories to better differentiate clinical severity and resource utilization.
• Category consolidation: Some V24 categories with overlapping clinical profiles were merged into unified HCCs under V28.
• Behavioral health expansion: V28 significantly expanded coverage of behavioral health conditions, adding new HCCs for substance use disorders and mental health diagnoses that were underrepresented in V24.
• Disease interaction factors: V28 introduced new interaction terms that account for the compounding effect of co-occurring conditions, reflecting the higher cost burden of members with multiple complex diagnoses.

Disease Family Organization

The 26 disease families in V28 group clinically related HCCs together. This organizational structure is foundational to V28's constraining methodology — within each family, HCCs share coefficient values regardless of severity differentiation. Understanding which HCCs belong to which disease family is essential for accurate revenue forecasting under V28.

Coefficient and Constraining Changes

This is the most consequential structural change between V24 and V28. Under V24, coefficients were severity-differentiated: higher-severity HCCs within a disease hierarchy earned proportionally higher weights. A more severe diabetes diagnosis carried a larger coefficient than a less severe one.

V28 introduced constraining — related HCCs within the same disease family now carry identical coefficients regardless of severity level. The practical impact is significant:

• V24 diabetes coefficients: ranged from 0.105 (least severe) to 0.368 (most severe)
• V28 diabetes coefficients: all share approximately 0.166 (except pancreas transplant status)

For plans that historically relied on capturing the most severe diagnosis within a hierarchy to maximize RAF scores, constraining eliminates that differential. The revenue uplift from documenting a higher-severity condition within a constrained family is zero — all severity levels produce the same RAF contribution.

Strategic Implications

Constraining shifts the strategic focus from severity capture to breadth of capture. Under V28, documenting conditions across multiple disease families matters more than documenting the most severe condition within a single family. Risk adjustment teams should prioritize identifying and recapturing conditions across all 26 disease families rather than concentrating on severity escalation within familiar hierarchies.

ICD-10 Code Mapping Reductions

V24 mapped 9,797 ICD-10-CM codes to HCCs. V28 maps 7,770 — a net reduction of 2,027 codes. This 21% reduction in mappable diagnoses has direct implications for coding and documentation workflows.

What Was Removed

The majority of eliminated codes were unspecified diagnosis codes — ICD-10 codes that lacked the clinical specificity CMS requires for risk adjustment. V28 enforces a higher documentation standard by only accepting codes that clearly identify the clinical condition and its characteristics.

Documentation Workflow Impact

Codes that mapped to HCCs under V24 may no longer map under V28. If documentation and coding practices haven't been updated, conditions that previously contributed to RAF scores will silently drop from risk profiles. This is particularly common in:

• Chronic condition recapture: Historical ICD-10 codes used for annual recapture may no longer be valid V28 mappings
• Provider documentation templates: Pre-built templates referencing V24-era codes may produce unmapped diagnoses under V28
• Automated coding workflows: NLP and CAC systems trained on V24 mappings need revalidation against V28's reduced code set

Revenue and RAF Score Impact

CMS projected a 3.12% average reduction in Medicare Advantage risk scores under V28. However, actual industry experience has varied widely:

• Best-prepared plans: 5-8% RAF score decline, offset by improved documentation and coding specificity
• Average plans: 10-15% RAF score decline, reflecting partial adaptation to V28 mechanics
• Least-prepared plans: 20-30% RAF score decline, driven by reliance on V24 coding patterns and unspecified diagnoses

Quantifying the Revenue Impact

For a 50,000-member Medicare Advantage plan, a systematic 0.1 RAF score understatement translates to approximately $5.2 million in unrealized annual revenue. The magnitude of this impact makes V28 adaptation a financial imperative, not an optional operational improvement.

Plans that invested in V28 readiness during the phase-in period — running parallel scoring, retraining coders, updating documentation templates — consistently outperformed those that deferred action until V28 became the sole model in 2026.

What This Means for Your Organization

Whether your organization has already adapted to V28 or is still closing gaps, these actions remain relevant:

• Audit ICD-10 code specificity: Review your most frequently submitted diagnosis codes against V28 mappings. Identify any codes that mapped under V24 but no longer map under V28, and establish updated coding guidance.
• Retrain coders on V28 mappings: Ensure coding teams understand which ICD-10 codes map to V28 HCCs and the specificity requirements for each. Legacy V24 coding habits are a primary source of preventable revenue loss.
• Run parallel scoring: Score your population under both V24 and V28 logic to quantify the exact RAF score differential. This identifies the specific members and conditions driving the gap.
• Update revenue forecasts: Replace V24-based RAF projections with V28-calibrated models. Revenue forecasts that haven't been updated for constraining and reduced mappings will systematically overstate expected capitation payments.

V24 vs V28 at a Glance

Tools for V28 Analysis