Private beta โ€” Lynceus is in active testing; approved members get full access. Request access

HCC V28 Changes by Condition: A Coder's Quick-Reference

The V28 CMS-HCC model is fully phased in for 2026 payment, and the changes are easiest to understand condition by condition rather than as a list of numbers.

What changed overall

  • The model was recalibrated on newer data and restructured, with many HCCs renumbered.
  • CMS removed a large number of diagnosis codes from payment, including many with high discretion or variable coding.
  • Some related HCCs were constrained to equal coefficients.
  • The transition ran in thirds: blended V24/V28 in 2024 and 2025, then 100% V28 for 2026.

Always confirm specifics in the CMS Rate Announcement and model software; the table below is a coderโ€™s orientation, not a payment schedule.

Condition quick-reference

Condition areaWhat changed in V28Coder focus
DiabetesThree diabetes HCCs share one coefficientDocument type and complications anyway; see diabetes V24 vs V28
Heart failureSingle V24 category split into graded heart failure HCCsCapture type and acuity; see heart failure mapping
AnginaRemoved from paymentDo not expect HCC value; still code accurately
DepressionMilder major depressive disorder severities no longer risk-adjustDocument severity and remission status
MalnutritionNarrowed to more severe presentationsSupport with clinical criteria
Vascular diseaseSeveral unspecified atherosclerosis codes droppedSpecify site and complication

What did not change

Documentation rules. A diagnosis must still be supported by a face-to-face encounter, evaluated or treated, and documented by an acceptable provider. Removal of a code from payment never makes it optional to code.

Using Lynceus to check

  1. Search the code in Lynceus and open its page, for example E11.22.
  2. Open the HCC Crosswalk to compare V24 and V28.
  3. Test a patient profile in the RAF Calculator.

For deeper reading, see CMS-HCC V24 vs V28: Key Differences.

Takeaway

V28 shifts value away from vague or discretionary diagnoses and toward clinically specific ones. Code what is documented, query for specificity, and verify mapping before you assume a diagnosis carries weight.