Written and maintained by CASRAI Editorial Board
Last updated
Hierarchical Condition Category (HCC) coding is the diagnosis-based prospective risk adjustment methodology used by the Centers for Medicare & Medicaid Services (CMS) under Social Security Act § 1853(a)(1)(C) (42 U.S.C. § 1395w-23) to estimate future healthcare expenditures for Medicare Advantage (Part C) enrollees and Accountable Care Organization (ACO) populations. Rather than reimbursing providers strictly per procedure, the CMS-HCC model maps qualifying chronic and severe acute ICD-10-CM diagnosis codes from eligible clinical encounters into hierarchical disease groups that determine each beneficiary’s Risk Adjustment Factor (RAF) score.
Beyond capitated reimbursement, CMS-HCC classifications serve as a standard comorbidity adjustment index across health services research, Medicare outcomes studies, and comparative effectiveness evaluations using administrative claims.
What Is HCC Coding and How Does the CMS-HCC Model Work?
The CMS-HCC model is a prospective statistical model: ICD-10-CM diagnosis codes collected during a baseline 12-month data collection period (January 1 through December 31) are combined with demographic variables to predict expected medical expenditures in the subsequent payment year. Each beneficiary’s composite RAF score is calculated as an additive linear sum of three components:
- Demographic Coefficients: Baseline relative factors assigned for age, sex, Medicaid dual-eligibility status (full benefit vs. partial benefit vs. non-dual), original reason for Medicare entitlement (age vs. disability or end-stage renal disease), and residence setting (community vs. long-term institutional facility).
- Hierarchical Disease Category (HCC) Coefficients: Relative weights assigned to each distinct payment HCC documented during the collection year. Unrelated chronic conditions are additive—for example, a patient with both heart failure and major depressive disorder accumulates the coefficient weights of both disease categories.
- Disease Interaction Terms: Additional risk weights triggered when specific high-cost physiological combinations co-occur (such as congestive heart failure combined with chronic obstructive pulmonary disease, or congestive heart failure combined with renal failure) to account for super-additive clinical complexity.
Clinical Hierarchy (“Trumping”) Logic
Within a single physiological disease family, HCC categories are arranged in a strict clinical hierarchy from most severe to least severe. If a patient’s medical record contains ICD-10-CM codes mapping to multiple HCCs within the same hierarchy during a calendar year, the higher-severity HCC supersedes (trumps) the lower-severity HCC so that only the highest-weighted category contributes to the RAF score. Conversely, if two diagnoses belong to separate disease groups, both HCC weights are summed.
Annual Calendar-Year Reset and Clinical Encounter Eligibility
Unlike problem lists inside an electronic health record that persist indefinitely, CMS-HCC risk profiles reset every January 1. Even permanent conditions—such as a below-knee amputation status, chronic systolic heart failure, or type 2 diabetes mellitus—must be recaptured at least once each calendar year during an eligible face-to-face clinical encounter (or qualifying real-time audio-video telehealth encounter) with an acceptable rendering provider (MD, DO, NP, PA, or CNS). Every submitted diagnosis must be substantiated in the progress note under standard medical record documentation and coding guidelines (ICD-10-CM Official Guidelines for Coding and Reporting, Section IV), typically audited via the MEAT framework (Monitor, Evaluate, Assess, Treat).
CMS-HCC Model Transition: Version 24 (V24) vs. Version 28 (V28)
Beginning in Calendar Year (CY) 2024 and completing its three-year phase-in at 100% weight in CY 2026 (following a 67% V24 / 33% V28 blend in CY 2024 and a 33% V24 / 67% V28 blend in CY 2025), CMS replaced the legacy Version 24 (V24) model with CMS-HCC Version 28 (V28). V28 recalibrated the model using ICD-10-CM clinical structures and updated fee-for-service expenditure data:
| Model Dimension | CMS-HCC Model V24 | CMS-HCC Model V28 | Operational & Research Impact |
|---|---|---|---|
| Clinical Classification Basis | Crosswalked from legacy ICD-9-CM clinical diagnostic structures | Built natively on ICD-10-CM clinical granularity and updated diagnostic hierarchies | Renumbered HCC categories and restructured disease hierarchies to reflect modern ICD-10-CM specificity. |
| Total Payment HCC Categories | 86 payment HCCs | 115 payment HCCs | Expanded granularity for severe conditions (including split severity tiers for dementia, pressure ulcers, and anorexia/bulimia). |
| Mapped ICD-10-CM Diagnosis Codes | 9,797 ICD-10-CM codes mapped to a payment HCC | 7,770 ICD-10-CM codes mapped to a payment HCC (with net removals and targeted additions) | Removed over 2,200 codes that no longer independently predict future Medicare costs or exhibited high discretionary coding variation (e.g., protein-calorie malnutrition without severity, angina pectoris, atherosclerosis of extremities without ulceration or gangrene). |
| Coefficient Constraints (Diabetes) | Distinct ascending weights for diabetes with chronic complications (HCC 18), acute complications (HCC 17), and without complications (HCC 19) | Constrained identical relative coefficients across diabetes severity tiers (HCC 36, HCC 37, HCC 38 in V28) | Reduces financial sensitivity to differential coding of mild or unspecified diabetic manifestations while preserving heart failure and chronic kidney disease interaction weights. |
How Is HCC Coding Used in Clinical Research and Healthcare Informatics?
While health plans and health systems use HCC coding for Medicare Advantage revenue integrity and CMS Risk Adjustment Data Validation (RADV) audits, clinical informaticists and epidemiologists rely on the CMS-HCC software (distributed publicly by CMS as SAS and Python-compatible mapping packages) for three core research applications:
- Confounder Adjustment in Real-World Evidence (RWE): Alongside the Charlson and Elixhauser comorbidity indices, the CMS-HCC RAF score provides a validated, expenditure-weighted composite morbidity score for propensity-score matching and multivariable regression in observational Medicare claims cohorts.
- Hospital Performance & Value-Based Quality Measures: CMS hospital-wide 30-day readmission and mortality models, as well as Medicare Shared Savings Program (MSSP) expenditure benchmarks, incorporate condition-category risk adjustment to compare facilities treating populations with different baseline chronic disease burdens.
- Algorithmic Bias and Coding Intensity Analysis: Health policy researchers audit longitudinal V24 and V28 claims datasets to distinguish true clinical morbidity shifts from diagnostic coding intensity differences between Medicare Advantage plans and traditional fee-for-service Medicare.
Frequently Asked Questions
What is HCC coding?
Hierarchical Condition Category (HCC) coding is a prospective risk adjustment methodology maintained by the Centers for Medicare & Medicaid Services (CMS) under Social Security Act § 1853(a)(1)(C). It maps specific chronic and severe acute ICD-10-CM diagnosis codes to weighted condition categories that combine with demographic factors to calculate a beneficiary’s Risk Adjustment Factor (RAF) score and prospective capitated payment.
How is HCC coding used in clinical research and healthcare?
In healthcare finance and population health, HCC coding adjusts capitated payments for Medicare Advantage (Part C) plans and Accountable Care Organizations (ACOs) based on patient morbidity. In clinical and health services research, investigators use CMS-HCC categories and RAF scores to control for baseline comorbidity burden, stratify high-risk observational cohorts in claims databases, and adjust hospital readmission and mortality outcomes.
What changed between CMS-HCC Model V24 and Model V28?
CMS-HCC Model Version 28 (V28), phased in from CY 2024 through 100% implementation in CY 2026, restructured condition hierarchies directly on ICD-10-CM clinical logic rather than legacy ICD-9-CM categories. V28 expanded payment HCC categories from 86 to 115, reduced the number of ICD-10-CM diagnosis codes mapping to an HCC from 9,797 to 7,770, and constrained coefficients across diabetes severity levels to reduce coding sensitivity.







