The Revenue Shock No One Should Be Surprised By
Health plans that built their risk adjustment programs around volume-based code capture are now facing a revenue correction. CMS-HCC V28 went to 100% implementation on January 1, 2026, ending the blended transition period that had softened the blow since 2024. The full impact is here, and it’s hitting hardest at organizations that treated risk adjustment as a code-counting exercise.
V28 dropped over 2,000 ICD-10 codes that previously mapped to HCCs. It eliminated entire HCC categories. It recalibrated coefficients so that certain high-value diagnoses, the ones aggressive coding programs prioritized, now generate less revenue. Plans that chased volume without investing in documentation quality are seeing their RAF scores decline even as their coding activity stays the same.
This wasn’t a surprise. CMS published the V28 final rule with clear intent: reduce overpayments driven by coding intensity that doesn’t reflect genuine patient complexity. The plans that prepared for this shift are performing better now than they were under V24. The plans that didn’t are scrambling.
What the Coefficient Changes Actually Mean
V28 didn’t just remove codes. It reshaped the value distribution across the entire model. Conditions like vascular disease and specified heart arrhythmias saw reduced coefficients. Certain diabetes complication categories lost significant weight. These were the categories that many chart review programs had prioritized specifically because of their high RAF impact.
At the same time, V28 increased the relative value of conditions tied to genuine, well-documented clinical complexity. Chronic kidney disease, heart failure, and major depressive disorder now carry more weight when properly documented with encounter-linked evidence. The model is rewarding plans that can prove their members are genuinely complex, not plans that can find the most codes.
For finance teams, this means revenue models built on V24 assumptions are unreliable. RAF projections need recalibration. Coding priority lists need updating. The conditions worth investing chart review effort into have shifted, and plans using outdated prioritization logic are misallocating resources.
The Compliance Multiplier
V28 doesn’t just change revenue. It amplifies compliance risk. With fewer codes mapping to HCCs, each submitted code carries proportionally more weight in a plan’s risk profile. A single unsupported diagnosis has a larger impact on audit outcomes than it did under V24. The margin for error has shrunk.
CMS’s population-level coding pattern analysis makes this worse. When a plan submits fewer total HCCs but concentrates them in high-value categories, that pattern stands out. It signals targeted coding activity rather than organic clinical documentation. Plans that haven’t adjusted their coding distribution to reflect V28’s new value map are creating patterns that invite scrutiny.
Compliance teams need to be running V28-specific risk assessments on their submitted data. Which categories are you concentrated in? How does your coding distribution compare to your clinical outcomes data? Are you still submitting codes in categories V28 devalued? These questions matter now in ways they didn’t two years ago.
Adapting to the New Reality
V28 rewards what should have been standard practice all along: accurate, encounter-linked, clinically evidenced coding. Plans that invested in documentation quality, two-way retrospective review, and explainable AI-assisted coding are finding that V28 works in their favor. Their codes are supported. Their risk scores reflect real patient complexity. Their audit exposure is manageable.
The plans still asking how to recover lost revenue under CMS HCC V28 are framing the problem wrong. The model didn’t take revenue away from plans with genuinely complex populations. It took revenue away from plans whose coding practices inflated complexity beyond what the documentation could support. That’s a distinction worth understanding, because the right response depends entirely on which side of it you’re on.
