Skip to content
Back to Blog
March 8, 2026·11 min read

How RAF Scores Work: Risk Adjustment Factor Scores Explained

Learn how Risk Adjustment Factor scores are calculated, which member inputs matter, and what the CMS-HCC V28 PY2026 RAF tool requires.

RAF ScoreRisk AdjustmentCalculatorHCC Coding

By the HCC Buddy Coding Team
Updated: July 26, 2026

How RAF Scores Work: Risk Adjustment Factor Scores Explained

What Is a RAF Score?

A Risk Adjustment Factor score, commonly called a RAF score, represents expected relative cost under a specified model and member context. The score is an input to CMS's risk-adjusted payment calculation. It is not the plan payment by itself.

CMS calibrates and normalizes risk models for specific populations and payment years. Do not use 1.0 as a universal Medicare Advantage member average or translate a raw score into a simple payment percentage without the applicable model, segment, normalization, and payment context.

RAF scores vary with the demographics, enrollment status, documented conditions, hierarchy results, and interaction terms in the applicable model. Understanding how those pieces are calculated is essential for any coder working in risk adjustment. For foundational concepts, see our introduction to HCC coding.

Why RAF Scores Matter

RAF scores feed into Medicare Advantage payment, but there is no universal dollar value for a 0.01 change. The payment effect depends on factors outside the raw score, including the county benchmark, plan bid, normalization, and coding intensity adjustment.

This financial significance creates several important dynamics:

  • Accurate RAF scores ensure plans receive appropriate funding, A plan that cares for a disproportionately sick population needs adequate funding to provide quality care. Undercoding leads to underfunding.
  • Inaccurate RAF scores trigger CMS action, RAF scores that are systematically too high attract Risk Adjustment Data Validation audits. CMS can extrapolate audit findings across the entire membership, resulting in payment clawbacks that dwarf the value of any individual coding error.
  • RAF scores affect quality ratings, CMS Star Ratings and quality metrics are partially risk-adjusted. Inaccurate RAF scores distort quality comparisons between plans.
  • Coders are the primary influence on the diagnosis component, While demographic factors are fixed, the diagnosis-based portion of the RAF score depends entirely on what conditions are coded from clinical encounters.
  • Components of a RAF Score

    A RAF score has two major components: demographic factors and diagnosis factors.

    Demographic Factors

    Demographic factors establish a baseline Risk Adjustment Factor score before any diagnoses are considered. These are assigned automatically based on CMS enrollment data, coders do not influence them:

  • Age and sex, CMS assigns a baseline coefficient based on the member's age band and sex. Older patients and males generally have higher baselines because they statistically use more healthcare resources.
  • Dual-eligible status, Members who are eligible for both Medicare and Medicaid (dual-eligible) receive a higher baseline coefficient because dual-eligible populations have higher average healthcare costs.
  • Institutional status, Members residing in long-term care facilities have substantially higher baselines than community-dwelling members, reflecting the intensive care needs of institutionalized populations.
  • Originally disabled status, Members whose original reason for Medicare entitlement was disability (rather than age) receive a modified baseline, as the disability population has different cost patterns than the aged population.
  • A community-dwelling, non-dual, non-disabled male aged 70 has a V28 community non-dual demographic coefficient of 0.396. This is the starting point before diagnosis coefficients are added.

    Diagnosis Factors (Where Coders Come In)

    The diagnosis component is where medical coders have direct impact. Each HCC category has a coefficient (weight) that adds to the demographic baseline:

  • Each HCC's coefficient is added to the baseline, If a patient has HCC 37 (Diabetes with Chronic Complications) with a V28 CNA coefficient of 0.166, that 0.166 is added to the demographic baseline.
  • Multiple HCCs sum together, A patient with HCC 37 (Diabetes with Chronic Complications), HCC 226 (heart failure), and HCC 280 (Chronic Obstructive Pulmonary Disease) would have all three coefficients added to the demographic baseline.
  • Hierarchies remove lower-severity duplicates, If the same patient qualifies for both HCC 37 and HCC 38, the hierarchy removes HCC 38 because HCC 37 is the higher-severity diabetes category. You cannot count both. Note that in V28, CMS constrained HCC 37 and HCC 38 to the same 0.166 CNA coefficient, so the hierarchy has no effect on the total RAF value here. The reason to capture the complication-specific code is clinical accuracy and RADV defensibility, not a weight gain.
  • Worked member-score example

    > RAF Calculator: The RAF Calculator supports CMS-HCC V28 for Payment Year 2026 and requires complete member context. HCC Buddy shows a score only after the server confirms the required source and calculation checks. If a check is unavailable or does not pass, no score is shown.

    Static code-level coefficients are references, not member totals. Historical PY2025 V24 stays reference-only.

    How the RAF Score Calculation Works (Step by Step)

    Here is the complete calculation flow from encounter to payment:

  • Step 1: Start with the demographic baseline, CMS assigns the age/sex/status coefficient automatically from enrollment data.
  • Step 2: Collect all diagnosis codes, Every ICD-10-CM code submitted from face-to-face encounters during the data collection period (typically the calendar year prior to the payment year) is gathered.
  • Step 3: Map codes to HCCs, Each ICD-10-CM code is crosswalked through the CMS-HCC model mapping table. Codes that map to an HCC are retained; codes that do not map are discarded for risk adjustment purposes.
  • Step 4: Apply hierarchies, Within each disease hierarchy, only the highest-severity HCC is kept. Lower-severity HCCs in the same group are removed.
  • Step 5: Sum HCC coefficients, The remaining unique HCC coefficients are added together.
  • Step 6: Apply disease interaction factors, Certain combinations of HCCs have interaction terms that add bonus coefficients. For example, having both heart failure and Chronic Obstructive Pulmonary Disease together may produce an interaction coefficient that recognizes the compounding effect of these conditions.
  • Step 7: Total all components, Demographic baseline + HCC coefficients + interaction terms = final RAF score.
  • The entire calculation happens at the CMS level, coders do not manually compute RAF scores. But understanding the process helps coders recognize which coding decisions have the greatest impact.

    V24 vs V28 RAF Weights

    For non-PACE organizations, CMS blended V24 and V28 risk scores in CY2024 and CY2025. CY2026 uses V28 at full weight. PACE follows a separate blend:

  • Payment Year 2024: 67% V24 + 33% V28
  • Payment Year 2025: 33% V24 + 67% V28
  • CY2026, non-PACE: V28 at full weight (transition complete)
  • The models differ in their coefficient values, even for HCCs that exist in both. Some notable patterns:

  • V28 generally has lower individual HCC weights but compensates with more categories and interaction terms
  • Severe conditions have higher weights in V28, Advanced cancers, organ failure, and acute conditions saw weight increases
  • Common chronic conditions may have lower weights, Conditions that V24 weighted heavily may carry less individual weight in V28
  • Restructured V28 categories changed value, Historical PY2025 V24 category numbers do not carry into CY2026 V28 by number alone
  • For a comprehensive comparison of what changed, see our V24 vs V28 guide.

    HCC Buddy RAF Calculator

    > RAF Calculator: The RAF Calculator supports CMS-HCC V28 for Payment Year 2026 and requires complete member context. HCC Buddy shows a score only after the server confirms the required source and calculation checks. If a check is unavailable or does not pass, no score is shown.

    Common RAF Score Misconceptions

    Several misconceptions about RAF scores persist in the coding community:

    "Higher RAF is always better." This is incorrect and potentially dangerous. Deliberately inflating RAF scores through inaccurate coding is healthcare fraud. The goal is an accurate RAF score that reflects the patient's true clinical burden, no more, no less. Plans with artificially inflated RAF scores face Risk Adjustment Data Validation audit exposure, payment clawbacks, and potential fraud investigations.

    "RAF scores only matter for Medicare Advantage." While the CMS-HCC model applies to Medicare Advantage, the concept of risk adjustment extends beyond it. The Affordable Care Act marketplace uses a different risk adjustment model (HHS-HCC) that operates on similar principles. Medicaid managed care programs also use risk adjustment models in many states. The skills transfer across programs.

    "One missed HCC does not matter much." Every HCC carries real financial weight. The correct V28 CNA coefficient for HCC 37 (Diabetes with Chronic Complications) is 0.166, and that amount is added to the risk score of every qualifying patient for the year. Multiplied across a panel of hundreds of diabetic patients, systematic undercoding of documented complications adds up to meaningful lost risk-adjusted revenue. Every HCC matters.

    "RAF scores are permanent." RAF scores are recalculated annually based on diagnoses submitted during the data collection period. Chronic conditions must be re-documented and re-coded every calendar year to continue contributing to the RAF score. A diagnosis coded in 2025 does not carry forward to 2026 automatically, it must appear in a 2026 encounter.

    How Coders Impact RAF Scores

    Medical coders connect clinical documentation to the diagnosis data used in risk adjustment. A code's effect depends on the applicable model, complete diagnosis set, hierarchy, member segment, and interactions:

  • Accurate, specific coding produces accurate RAF scores, When every documented condition is captured at its highest supported specificity, the RAF score reflects the patient's true health burden.
  • Missing a legitimate HCC means the plan is underfunded, If a patient truly has diabetes with chronic kidney disease but the coder only captures diabetes without complications, the plan receives less funding than the patient's care requires.
  • Overcoding an HCC creates audit risk, If the documentation does not support the coded condition, the Risk Adjustment Data Validation audit finding results in payment recovery, often extrapolated across the plan membership.
  • Query programs improve both accuracy and RAF scores, When documentation is ambiguous, sending a clinical documentation improvement query to the provider is the appropriate action. Queries are expected and encouraged, they improve documentation quality and coding accuracy simultaneously.
  • Documentation support is separate from RAF math. See the MEAT review guide for the mnemonic's limits and links to official sources.

    For coders reviewing model context, HCC Buddy's encoder shows source-labeled HCC mappings and segment-specific base coefficients when the source supports them. It does not turn a code-level factor into a universal member score or payment amount.

    Conclusion

    RAF scores are the financial backbone of risk adjustment, the mechanism through which clinical complexity translates to appropriate plan reimbursement. Understanding how they work makes you a more effective, more valuable medical coder.

    The key principles: demographic factors set the baseline, HCC coefficients add diagnosis-based value, hierarchies prevent double-counting, and qualifying combinations add interaction terms. Payment Year 2026 uses V28; V24 remains available for earlier-model review.

    Use the RAF tool for a CMS-HCC V28 PY2026 member result only when you have the complete diagnosis set and member details. Historical V24 remains a reference, not a second member calculation.

    RAF score calculator FAQs

    Which CMS-HCC model applies to Payment Year 2026?

    CY2026 non-PACE Medicare Advantage risk scores use CMS-HCC V28 at full weight. PACE follows a separate blend. HCC Buddy labels V24 as a historical PY2025 reference, so choose the program, payment year, and source that match the work you are checking.

    What information does a RAF score calculator need?

    A RAF score calculator needs the applicable model, demographic factors such as age, sex, and enrollment status, and the confirmed HCCs that remain after hierarchy rules are applied. The final calculation can also include interaction terms when the selected HCC combination qualifies for one under that model.

    Does the HCC Buddy RAF score calculator apply HCC hierarchies?

    HCC Buddy's RAF Calculator supports CMS-HCC V28 for Payment Year 2026 and requires complete member context. It shows no score unless the required source and calculation checks pass. Historical PY2025 V24 stays reference-only.

    Can I calculate both V24 and V28 RAF scores?

    HCC Buddy's RAF Calculator supports CMS-HCC V28 for Payment Year 2026 and requires complete member context. It shows no score unless the required source and calculation checks pass. Historical PY2025 V24 stays reference-only.

    HCC Buddy

    HCC Buddy Coding Team

    Editorial

    Every HCC Buddy article is checked against the current CMS-HCC model and the active FY ICD-10-CM tabular release before it publishes.

    Get HCC Coding Tips in Your Inbox

    Join our newsletter for coding tips, guideline updates, and tool announcements.

    Related Articles