What Is an HCC, and What Does the CMS-HCC V28 Mapping Tell a Coder About an ICD-10-CM Code?

CMS risk-adjusts Medicare Advantage payment with the CMS-HCC model, which groups ICD-10-CM diagnosis codes into Hierarchical Condition Categories. Each category has a number and a label — HCC 226 is Heart Failure, Except End-Stage and Acute; HCC 326 is Chronic Kidney Disease, Stage 5 — and belongs to a family, and some categories suppress others. CMS publishes the code-to-category mapping, the category labels and the hierarchy in its mappings and model-software releases on its risk-adjustment page. The release MedCoder holds is CMS-HCC V28, PY2026 mid-year final release, taken from the 2026 Mid-Year Final ICD-10 Mappings and Model Software (CMS-HCC V28, V22; RxHCC V08); the exact release files are listed on Data sources.

The mapping reads in one direction. A code either is or is not assigned to a category under the model; nothing in the mapping says which code to assign. That is decided by the record and by the Guidelines, which do not mention risk adjustment at all. Section I.A.19:

Official sourceThe assignment of a diagnosis code is based on the provider’s diagnostic statement that the condition exists. The provider’s statement that the patient has a particular condition is sufficient. Code assignment is not based on clinical criteria used by the provider to establish the diagnosis. If there is conflicting medical record documentation, query the provider.

That fixes the order of operations. The provider documents; the coder assigns the code the documentation supports; the model, applied by CMS after submission, assigns the category. The active-versus-history guide puts it the same way: coding history as active overstates risk, coding active disease as history understates it, and neither changes the rule — the test decides the code and the risk-adjustment consequence follows it.

Three consequences follow for a working coder. A code that maps to no category is not an error and not a prompt to find one that does. A category that another category suppresses is not a reason to leave a code off the claim. And a category is not a dollar figure: MedCoder code pages show the mapping CMS published and do not show a coefficient or a score.

Each V28 category can suppress other categories: when the higher category is present, CMS zeroes out the lower one. The example this site carries is HCC 211, Respirator Dependence/Tracheostomy Status/Complications, which Z93.0 reaches; it suppresses HCC 212 (Respiratory Arrest) and HCC 213 (Cardio-Respiratory Failure and Shock), which J96.01 reaches. The code page states it in those terms: when the higher category is present, CMS zeroes out the categories it suppresses, and the hierarchy applies between categories, not between individual codes.

The hierarchy is often described as codes superseding codes. It is not. Z93.0 and J96.01 are both reported when both are documented and both meet Section III; the suppression is arithmetic CMS performs on the categories after the claim is submitted. It is not an Excludes note, not a sequencing instruction, and not a reason to drop the second code. A coder who omits J96.01 because “the trach already covers it” has removed a reportable diagnosis from the record for a reason the Guidelines do not recognize.

1. Assign the code from the provider’s statement. Section I.A.19: the provider’s statement that the condition exists is sufficient, and code assignment is not based on clinical criteria. Conflicting documentation is a query, not a judgment call.
2. Decide reportability by the reporting rules, not by the mapping. For an inpatient stay, Section III defines other diagnoses for reporting purposes as additional conditions that affect patient care in terms of requiring clinical evaluation, therapeutic treatment, diagnostic procedures, extended length of hospital stay, or increased nursing care and/or monitoring. For an outpatient encounter, Section IV.J:

   Official sourceCode all documented conditions that coexist at the time of the encounter/visit and that require or affect patient care, treatment or management. Do not code conditions that were previously treated and no longer exist. However, history codes (categories Z80-Z87) may be used as secondary codes if the historical condition or family history has an impact on current care or influences treatment.

   The passage answers both halves of the question a mapping cannot: a condition that affects care is reported whether or not it maps, and a resolved condition is not reported however it maps.
3. Code to the specificity the record supports. The category follows the code, so specificity moves it: the documented CKD stage selects N18.31 (HCC 329), N18.4 (HCC 327) or N18.6 (HCC 326); documented dementia severity selects F02.80 (HCC 127, Dementia, Mild or Unspecified) or F02.B0 (HCC 126, Dementia, Moderate); documented diabetic CKD selects E11.22 (HCC 37) over E11.9 (HCC 38); documented remission selects E11.A (no category). In each pair the record chooses; the category is what CMS says about the result.
4. Read the mapping afterwards, as a check on a fact. Mapped, not mapped, which category, which hierarchy. None of it reopens the code selection.
5. Read the model name. CMS-HCC V28 is the model for payment year 2026; CMS-HCC V22 and RxHCC V08 are other published models and are shown separately.
6. Do not read a value into a category. CMS publishes a separate coefficient for each population segment, so no single number describes a category, which is why none is shown.

1. Documentation: hypertension; CKD stage 4; nothing states the two are unrelated. Result:I12.9 → N18.4. Mapping: I12.9 maps to no CMS-HCC V28 category; N18.4 maps to HCC 327. Why it does not matter to the coding: I10, I11.9, I12.9 and I13.10 map to no HCC, and the companion I50.- and N18.- codes carry their own categories regardless. The hypertension-with-CKD rule chose I12.9; the stage chose N18.4; a documented stage 3a would have given N18.31 (HCC 329), and documented ESRD would have changed the pair to I12.0 → N18.6, both mapping to HCC 326. Only the documentation moves the category.
2. Documentation: hypertensive heart and chronic kidney disease with heart failure and ESRD. Result: I13.2 → the I50.- code for the heart-failure type → N18.6. Mapping: I13.2 maps to both HCC 226 and HCC 326, the one code in the hypertension family carrying both; I50.9 and I11.0 map to HCC 226; N18.6 maps to HCC 326. Point: the I50.- and N18.6 codes are required by I13’s Use Additional Code notes, not by the mapping, and the fact that N18.6 shares a category with I13.2 — a “same CMS-HCC risk category” pairing on the code page — removes nothing from the claim.
3. Documentation: type 2 diabetes; CKD stage 3a; no statement that the CKD is unrelated. Result: E11.22 → N18.31. Mapping: E11.22 maps to HCC 37, Diabetes with Chronic Complications; E11.9 would have mapped to HCC 38. Point: E11.22 is assigned because the classification presumes the diabetic link in documented diabetes with CKD, not because HCC 37 exists. In a different record — type 2 diabetes documented as in remission, no complication documented — E11.A would be the code, and it maps to no category; that is still the correct code.
4. Documentation: right breast cancer excised two years ago; no current treatment directed at the site; no evidence of disease. Result:Z85.3, not C50.911. Mapping: C50.911 maps to HCC 23, Prostate, Breast, and Other Cancers and Tumors; Z85.3 maps to no category. Point: the active-versus-history test decides the code, and the risk-adjustment consequence follows the code. Coding the history as active would overstate risk; coding active disease as history would understate it; the rule is the same in both directions.
5. Documentation (inpatient): tracheostomy status; acute respiratory failure with hypoxia, diagnosed and treated this admission. Result: J96.01 and Z93.0 both reported, sequenced by the circumstances of admission. Mapping: Z93.0 maps to HCC 211; J96.01 maps to HCC 213; HCC 211 suppresses HCC 213. Point: both conditions meet Section III, so both are reported. The suppression is CMS’s arithmetic between two categories after submission; it is not an instruction to the coder.

• Selecting a code for its category. The order runs provider statement → code → category. A code chosen because it maps, when the documentation supports a different code, is the error the Guidelines are written to prevent; Section I.A.19 bases assignment on the provider’s diagnostic statement, not on clinical criteria — and not on a mapping.
• Omitting a code because its category is suppressed. The hierarchy acts between categories after submission. A reportable diagnosis under Section III or Section IV.J stays on the claim whatever CMS later zeroes out.
• Reading “not mapped” as a problem. It means CMS assigns the code to no category under CMS-HCC V28. It is the answer for most ICD-10-CM codes, and it says nothing about whether the code is correct or reportable.
• Attaching a value to a category. No single value describes a category: CMS publishes a separate coefficient for each population segment, which is why no coefficient, score or demographic factor is shown anywhere on this site.
• Mixing models. CMS-HCC V28 is the model for payment year 2026; V22 and RxHCC V08 are different models shown separately. A category number from one model means nothing in another.
• Reading a family as if it mapped as a unit. Within M02, M02.3 (Reiter’s disease) maps to HCC 94 while M02.0 and M02.2 map to nothing, and all 24 M01 codes map to HCC 92. The subcategory, not the family, decides, and ranges and chapters are not looked up at all.
• Confusing an HCC with an MS-DRG complication or comorbidity. The CC/MCC designations belong to the inpatient grouper and the HCC categories to the risk-adjustment model; a code’s standing in one says nothing about the other.

• The provider’s diagnostic statement. Section I.A.19 and Section I.B.14: the code rests on the documentation of the physician or other qualified practitioner legally accountable for the diagnosis, and conflicting documentation is queried.
• That the condition affected this encounter. Section III for an inpatient additional diagnosis; Section IV.J for an outpatient visit. A condition previously treated and no longer existing is not coded, whatever it would map to.
• Specificity. CKD stage; dementia severity; a diabetic complication and its link to the diabetes; remission; active disease versus personal history. Each moves the code, and the category moves with it.
• Status conditions that affect care. A tracheostomy, gastrostomy or other artificial opening is reported with a Z93.- status code when it affects care. Section III and Section IV.J supply that reason; the mapping does not. Z93.0 maps to HCC 211, and that is a fact about the code after it is reported, not a reason to report it.
• Nothing the mapping asks for. The CMS file lists codes and categories. It contains no documentation requirement, and a record that supports the code supports the category.

A code page’s risk-adjustment block shows, for CMS-HCC V28 and payment year 2026, the category or categories CMS assigns the code, each category’s label, the categories it suppresses, and — under “Other published models” — the CMS-HCC V22 and RxHCC V08 categories, kept separate and not merged. “Not mapped” is stated explicitly. The number of codes the model maps is counted live from the loaded mapping rather than typed into the page, because a typed figure drifted within one CMS mid-year update; this guide states no count for the same reason.

It does not show risk-adjustment factor (RAF) scores, coefficients or demographic factors, because CMS assigns a different coefficient to each population segment and no single value describes a category. It does not show the ESRD model. And it is not billing advice: it shows what CMS published about a code, and the code itself is decided by the record and the Guidelines.