Risk adjustment is not a coding problem. It is a clinical documentation problem, and most organizations treat it as the former when it demands the full attention of both.
What HCC Risk Scores Actually Measure
The Centers for Medicare & Medicaid Services (CMS) uses Hierarchical Condition Category (HCC) models to predict the expected cost of care for a Medicare Advantage enrollee based on their documented diagnoses. Each qualifying ICD-10 code maps to an HCC category, which carries a Risk Adjustment Factor (RAF) score. The sum of those scores, combined with demographic factors, produces the member’s RAF, which directly determines plan payment.
Two patients with identical clinical complexity can carry very different RAF scores depending solely on how thoroughly their conditions were documented in the prior measurement year.
The Documentation Gap
Most provider groups with unexplained performance gaps in value-based contracts have the same root cause: conditions present, treated, and managed, but never captured in the claim record with the specificity CMS requires.
Common failures:
- Chronic conditions documented in the chart narrative but not coded on the claim
- Unspecified codes used when specificity is available (and required)
- Conditions resolved in prior years recaptured annually without updated status
- HCC-relevant diagnoses buried in specialist notes, never reconciled to the PCP claim
What High-Performing Groups Do Differently
The programs that consistently achieve RAF accuracy above 0.95 share several structural traits:
1. Annual wellness visit discipline. AWVs are the single highest-yield opportunity for prospective capture. High performers schedule and complete them at rates above 80% for their Medicare population.
2. Retrospective review as a feedback loop. Chart-to-claim audits identify documentation patterns, by provider, by specialty, by site, that can be corrected through targeted education rather than broad mandates.
3. Prospective gap identification. Predictive models flag members whose documented RAF is below their expected clinical complexity before the measurement year closes. Outreach is targeted, not blanket.
4. Coder-clinician collaboration. The most durable programs create structured pathways for certified coders to query clinicians, not flag errors after the fact, but align on documentation standards prospectively.
The Regulatory Dimension
RADV audits are not hypothetical. CMS continues to expand its Risk Adjustment Data Validation program, and the extrapolation methodology finalized in recent rulemaking means that identified error rates can produce significant retroactive payment adjustments. Documentation quality is both a revenue opportunity and a compliance obligation.
Where Prizm Fits In
Our Risk Adjustment practice works with provider groups and health plans to design programs that close the gap between clinical reality and documented risk, prospectively. We combine retrospective audit findings with targeted clinician education, coding workflow redesign, and ongoing performance tracking to build programs that hold up under RADV scrutiny.
If your RAF scores don’t reflect the clinical complexity of your population, that gap has a cause, and a fix.