Worked examples
Each row is scored with the same engine as the calculator when this page was built. Open a row to load it above and see every factor; the raw score is what the CMS software outputs, and the last column is after the CY2027 normalization factor and MA coding adjustment.
| Example | Segment and age | ICD-10-CM | Payment HCCs | Raw score | Adjusted |
|---|---|---|---|---|---|
| Diabetes with kidney disease and heart failure | Community, non-dual, aged, female 72 | E11.22, N18.4, I50.22 | HCC 37, HCC 226, HCC 327 | 1.723 | 1.503 |
| Two diabetes codes: the hierarchy keeps one | Community, non-dual, aged, male 68 | E11.9, E11.22, I48.0 | HCC 37, HCC 238 | 0.797 | 0.695 |
| Heart failure, atrial fibrillation, COPD and respiratory failure | Community, full-benefit dual, aged, female 80 | I50.22, I48.0, J44.9, J96.11 | HCC 213, HCC 226, HCC 238, HCC 280 | 3.030 | 2.642 |
| Disabled long-term institutional resident with Medicaid | Long-term institutional, male 58 | G35.A, I50.9, J44.9, L89.154 | HCC 198, HCC 226, HCC 280, HCC 379 | 4.896 | 4.270 |
| Breast cancer code under age 50 | Community, non-dual, disabled, female 45 | C50.911 | HCC 22 | 0.706 | 0.615 |
| The same code at age 70 | Community, non-dual, aged, female 70 | C50.911 | HCC 23 | 0.581 | 0.506 |
| New enrollee with Medicaid | New enrollee, female 66 | E11.22, I50.22 | not used | 0.990 | 0.864 |
| New enrollee turning 65 during 2027 | New enrollee, female 64, entitled by age | E11.22, I50.22 | not used | 0.532 | 0.464 |
Take the first row. A 72-year-old woman living in the community without Medicaid is scored in the community, non-dual, aged segment, whose factor for her age band is 0.395. E11.22 maps to HCC 37 (0.166), N18.4 maps to HCC 327 (0.514) and I50.22 maps to HCC 226 (0.360). The diabetes and heart failure interaction adds 0.112 and the heart failure and kidney disease interaction adds 0.176. Three payment HCCs carry no count factor in this segment, so the raw score is 1.723; divided by 1.079 it is 1.597, and after the 5.90% coding adjustment 1.503. The second row shows the hierarchy at work: E11.9 maps to HCC 38, but HCC 37 from E11.22 overrides it, so only one diabetes factor is paid. The breast cancer rows show an age edit: the same code maps to HCC 22 before age 50 and to HCC 23 from 50. The new enrollee rows show that diagnoses do not change a new enrollee's score, and that a woman of 64 entitled by age is scored in the age-65 cell (0.532), not the 60-64 cell CMS uses for a new enrollee of that age entitled by disability (1.212).
How the RAF score is calculated
Payment year 2027 risk score = (raw score / 1.079) x (1 - 0.059)
The calculator follows the steps of the CMS SAS macro V2826T1M. Age is taken on February 1 of the payment year. Each diagnosis assigns its condition categories, after the edits that make some categories depend on age or sex. A heart failure rule then drops HCC 223 unless another heart failure category (HCC 221, HCC 222, HCC 224, HCC 225, HCC 226) is also present. The hierarchy runs next, one rule at a time in CMS's order, so only the most severe category of a related set is paid. Diagnostic groups such as diabetes, heart failure, chronic lung disorder and kidney disease are then flagged, the interaction terms are switched on when two groups are both present, and the number of payment HCCs left after the hierarchy selects the count factor. The score is the sum of every factor that applies in the beneficiary's segment.
Interaction terms pay an extra factor when both condition groups are documented for the same beneficiary. The community, non-dual, aged values are shown; the calculator uses the values of whichever segment applies, and the substance use and psychiatric interaction exists only in the disabled segments.
| Interaction | CMS variable | Factor (CNA) |
|---|---|---|
| Diabetes and heart failure | DIABETES_HF_V28 | 0.112 |
| Heart failure and chronic lung disorder | HF_CHR_LUNG_V28 | 0.078 |
| Heart failure and kidney disease | HF_KIDNEY_V28 | 0.176 |
| Chronic lung disorder and cardio-respiratory failure | CHR_LUNG_CARD_RESP_FAIL_V28 | 0.254 |
| Heart failure and specified heart arrhythmias | HF_HCC238_V28 | 0.077 |
The payment HCC count adds nothing in the community segments until a beneficiary has five categories after the hierarchy; from there the community, non-dual, aged factor is 0.050 for 5, 0.102 for 6, 0.188 for 7, 0.316 for 8, 0.444 for 9, 0.728 for ten or more. In the long-term institutional segment the count factor applies only at ten or more (0.373), and that segment carries its own interaction terms for beneficiaries under 65 (disabled) with cancer, a neurological condition, heart failure, a chronic lung disorder or a skin ulcer.
Choosing the segment and the demographic inputs
CMS scores every beneficiary in one of nine segments, each a separate regression with its own factors. Continuing enrollees living in the community are split by Medicaid dual status (non-dual, partial-benefit dual, full-benefit dual) and by age: the aged segments apply from 65 and the disabled segments below 65. Beneficiaries who have been in a long-term care facility for 90 days or more are scored in the institutional segment, where Medicaid is a separate factor. New enrollees, with fewer than 12 months of Medicare Part B in the data collection year, are scored on age, sex, Medicaid and originally-disabled status only, with a separate set of factors for new enrollees of chronic condition special needs plans. The originally-disabled flag matters only from 65: it adds a factor in the community aged segments and changes the new enrollee cell. Age 64 has one more rule for new enrollees: someone who is 64 on February 1 and entitled to Medicare by age, because they turn 65 later in the payment year, is placed in the age-65 new enrollee cell rather than the 60-64 cell, which the CMS software (macro AGESEXV2) uses at that age only when the original reason for entitlement (OREC) is not age. The calculator asks for it only when a new enrollee segment and age 64 are selected. CMS determines all of these from enrollment data, not from the plan.
| Segment | Diagnoses | Medicaid | Originally disabled (65+) |
|---|---|---|---|
| Community (non-dual, partial or full dual) | Yes: HCCs, interactions, count | Through the segment choice | Yes, aged segments |
| Long-term institutional | Yes, plus disabled interactions under 65 | Yes (LTIMCAID) | Yes (ORIGDS) |
| New enrollee and SNP new enrollee | No | Yes (NEMCAID) | Yes; at 64, entitlement by age selects the age-65 cell |
What the code lookup shows
The lookup reads the ICD-10-CM to condition category crosswalk in the CMS V28 software and the PY2027 initial ICD-10-CM mapping file, which agree code for code. 280 codes map to two categories, for example B37.7 (candidal sepsis) to HCC 2 and HCC 6, and both are listed. 408 codes carry an age or sex condition from the CMS edit macro or the Medicare Code Editor age ranges, shown in brackets next to the category: C50 breast cancer codes map to HCC 22 before age 50 and HCC 23 after, COPD codes count only from age 18, and two hemophilia codes (D66, D67) map to HCC 112 for women and HCC 111 for men. 4,056 of the mapped codes also carry a Part D category, linked to the RxHCC V08 category hub; codes such as I10 (essential hypertension) that have an RxHCC but no V28 payment category are reported as such. Every category links to its HCC reference page with the full code list and factors in all seven continuing enrollee segments.
Final payment year 2027 risk scores use diagnoses from 2026 dates of service (CY2027 Rate Announcement); the initial scores CMS calculates earlier draw on an earlier window, so the PY2027 initial mapping lists codes valid in FY2025 and FY2026. 12 mapped codes were deleted on October 1, 2026; they still count for earlier dates of service and the lookup marks them. Three more (D71, E72.53 and G35) were split into more specific codes on October 1, 2025, according to the FY2026 conversion table; they are valid only for dates of service before then, so they cannot support a 2026 diagnosis, and the lookup marks them with the codes that replaced them. 45 codes created on October 1, 2026 replace a code that maps to an HCC; the lookup names the code each one replaces and its category, but CMS has not yet assigned the new codes, so they add nothing to the estimate until a later mapping does. The V24 to V28 crosswalk report shows which codes gained or lost a category when V28 replaced the 2020 model.
What the estimate leaves out
The score covers one beneficiary and the codes typed in. It does not decide whether a diagnosis is eligible for risk adjustment: CMS accepts diagnoses from face-to-face or audio-video encounters with acceptable provider types, and for 2027 the Rate Announcement excludes diagnoses from audio-only services (modifiers 93 and FQ) and from chart review records that are not linked to an encounter, except for beneficiaries who switch MA organizations. It does not compute the ESRD dialysis or functioning graft models, the PACE blend with the 2017 model, frailty factors, Medicare secondary payer adjustments or partial-year enrollment, and it does not turn the score into a payment, which also depends on the county rate and the plan's bid. The RxHCC Part D score uses a different model whose factors are on the RxHCC hub. The glossary explains the RAF score and the V28 model in more detail.
Where QuickIntell fits beyond the calculator
The calculator scores the codes you type. In QuickIntell's risk adjustment workflow, proposed HCC findings are reviewed against the current documentation, the source records and reviewer decisions stay together for audit, and open clinical questions go to the responsible clinician.
Frequently asked questions
How is a RAF score calculated?
Add the relative factors that apply to the beneficiary: one demographic factor for age and sex in the right segment, plus originally-disabled or Medicaid factors where the segment has them, one factor for each payment HCC left after the hierarchy, the interaction factors when two condition groups are both present, and the payment HCC count factor. That sum is the raw risk score the CMS software outputs. For payment year 2027 CMS divides it by the normalization factor and then reduces it by the MA coding pattern difference adjustment.
What normalization factor applies in 2027?
The CY2027 Rate Announcement (April 6, 2026) sets 1.079 for the 2024 CMS-HCC model, which is V28, and a 5.90 percent MA coding pattern difference adjustment, the statutory minimum. The calculator shows the raw score first and then each adjustment, so the CMS software output stays visible.
Why does one ICD-10-CM code show two HCCs?
CMS maps some codes to two condition categories, for example a diabetes code with a retinal complication to both the diabetes category and the eye category, or candidal sepsis to both sepsis and opportunistic infection. Both count unless the hierarchy drops one. The lookup lists every category a code maps to.
Why do some codes map to an HCC only at certain ages or for one sex?
The CMS software applies edits before it assigns categories: breast cancer codes map to a higher category under age 50, COPD and some chronic lung codes count only from age 18, certain newborn codes only under age 2, two hemophilia codes map to a different category for women, and Medicare Code Editor age ranges (adult, pediatric, maternity, newborn) apply to 303 codes. Counting both kinds, 408 codes carry a condition, and the lookup shows each one next to the category.
Does the calculator include RxHCC, ESRD or frailty scores?
No. It scores the CMS-HCC V28 model that Medicare Advantage Part C payment uses for most beneficiaries. The lookup still lists each code's RxHCC category, and the RxHCC hub has the Part D factors; ESRD dialysis and graft models, PACE blending and frailty factors are separate CMS calculations.
Which codes count for payment year 2027?
Diagnoses from acceptable encounters in the data collection period. The CY2027 Rate Announcement bases final 2027 payments on diagnoses from 2026 dates of service; the initial scores CMS calculates earlier use an earlier window, which is why the PY2027 initial mapping covers codes valid in FY2025 and FY2026. A code counts only for dates of service when it was valid: codes deleted on October 1, 2025 can appear only in that earlier window, codes deleted on October 1, 2026 still count for 2026 dates of service, and the new FY2027 codes wait for a later CMS mapping. For 2027 CMS also excludes diagnoses from audio-only services and from chart review records not linked to an encounter.
Browse the categories on the CMS-HCC V28 hub, compare inpatient groups for the same diagnoses on the MS-DRG reference, or see the other free revenue-cycle tools.
Sources
Every figure on this page is taken from the CMS publications below, as released by the Centers for Medicare & Medicaid Services. Projection built 2026-10-10. Verify against the primary file before billing or contracting decisions. The reference methodology explains how each figure is computed from these files.
- ICD-10-CM to CMS-HCC V28 mappings, payment year 2027 initialVersion PY2027 initial ICD-10-CM mappings (dates of service 2026) · effective 2027-01-01 · file 2027 Initial ICD-10-CM Mappings.csvSHA-256 9c260b06d0dd2806…
- CMS-HCC V28 model software: labels, hierarchies and relative factorsVersion CMS-HCC V28 software V2826.115.T2 (PY2027 initial) · effective 2027-01-01 · file C2824T2N.csvSHA-256 20f38d537493b41c…
- CMS-HCC V28 Python model software, PY2027 initial: ICD-10-CM to CC mapping with the age and sex edit conditionsVersion CMS-HCC V28 Python package 2027 T1 initial, package v1 · effective 2027-01-01 · file ICD10_CC_mappings_CMS_HCC_2027_v28_initial.csvSHA-256 93307f974301b2a5…
- Announcement of CY2027 Medicare Advantage capitation rates and Part C and Part D payment policiesVersion CY2027 Rate Announcement (2026-04-06) · effective 2027-01-01 · file 2027-announcement.pdfSHA-256 8bb37a0969062c57…
- ICD-10-CM FY2026 conversion table (new codes and the codes they replace)Version FY2026 (April 1, 2026 update) · effective 2026-04-01 · file 508-Version-ICD-10-CM-CONVERSION-TABLE-FY2026-April 1 2026.csvSHA-256 fabda06a0993f92c…
Disclaimer
Operational estimate built from CMS public files. It reproduces the relative factors, edits, hierarchies and interactions of the CMS software for the inputs given; it is not a CMS risk score, does not establish that any diagnosis is supported by the medical record, and is not legal, clinical, coding or billing advice.
Found a figure that does not match the CMS file? Report a data correction with the CMS file and the value you expected; the request reaches the editorial team through the contact form.