Risk adjustment is the financial mechanism that compensates Medicare Advantage and ACA marketplace payers for the clinical risk in their member populations. Accurate HCC (Hierarchical Condition Category) coding from clinical encounters drives the payments. Traditional risk adjustment uses claims-derived HCC inference combined with chart abstraction; FHIR-based risk adjustment uses clinical data resources directly. Five patterns have emerged for FHIR-driven risk adjustment in 2026. For FHIR for analytics and Stars guides coverage on this site, these are the practical patterns.
1. Condition Resource as Primary HCC Source
The foundational pattern. FHIR Condition resources captured from EHR data carry diagnosis codes (ICD-10-CM) that map to HCC categories. Risk adjustment computation reads Condition resources, applies the CMS HCC model to derive risk scores, and produces member-level risk adjustment factors.
The pattern requires the Condition resources to be present and accurate. EHR-sourced Condition data is more complete than claims-derived diagnoses (claims only carry diagnoses when they affected the billable visit; EHR Condition resources capture ongoing diagnoses regardless of billing).
2. Observation-Driven HCC Validation
A pattern that uses Observation resources to validate HCC coding. A claim or EHR-Condition resource indicates diabetes; the Observation resources should show A1c values, blood glucose monitoring, related lab work. Inconsistencies (HCC claim with no supporting Observation evidence) flag for review.
The pattern improves risk adjustment accuracy by catching coding errors. CMS audit pressure on risk adjustment makes this validation increasingly important.
3. Encounter-Based Documentation Sufficiency Check
A pattern that uses Encounter resources to verify that HCC coding has adequate documentation. CMS requires that HCC coding be supported by face-to-face encounters with qualified providers. Risk adjustment computation can check that each HCC has at least one qualifying Encounter in the relevant period.
The pattern protects against audit findings. Plans with weak documentation sufficiency checks risk HCC reclassification during CMS audits.
4. AI-Augmented Chart Review
A pattern that uses ML on FHIR data to prioritize chart review for risk adjustment. The model identifies members whose claims-based HCC coding may be incomplete, based on patterns in their Observation, Procedure, and Medication resources. Chart abstractors focus their effort on the highest-leverage cases.
The pattern increases the impact of limited chart abstraction capacity. Production deployments in 2026 are growing as both FHIR data availability and ML tooling mature.
5. Provider Engagement Through FHIR Data Sharing
A pattern that uses CMS-0057-F Provider Access to share clinical data with in-network providers, supporting provider-side documentation that drives HCC coding accuracy. The provider sees the payer's view of the member's HCC profile and can address documentation gaps during clinical care.
The pattern requires Provider Access to work well operationally. Plans deploying this pattern alongside CMS-0057-F compliance get value from the same infrastructure investment.
How Risk Adjustment Differs From Stars and HEDIS
Risk adjustment is financial: better HCC accuracy means better risk-adjusted payments. Stars and HEDIS are performance measurement: better measure performance affects ratings and bonuses. The technical patterns overlap (FHIR clinical resources drive both); the operational frameworks differ.
Plans that build FHIR-based risk adjustment alongside Stars and HEDIS computation get cross-domain value from the clinical data investment. The same Observation resources that support BP control measurement for Stars also support HCC coding validation for risk adjustment.
The CMS Audit Pressure That Shapes the Patterns
CMS Risk Adjustment Data Validation (RADV) audits have tightened in 2025 and 2026. Plans with weak documentation defense lose HCC payments during audit. The patterns above are partly motivated by audit defensibility: every HCC coded should have supporting evidence in the FHIR data, traceable from the risk adjustment computation back to source.
Plans building FHIR-based risk adjustment with strong audit trails defend RADV audits more effectively than plans relying on claims-only HCC inference.
What Production Implementations Combine
A production-grade FHIR risk adjustment implementation typically combines several patterns. Condition resources drive primary HCC sourcing (Pattern 1). Observation-based validation catches errors (Pattern 2). Encounter sufficiency checks protect against audit findings (Pattern 3). AI augmentation prioritizes chart review (Pattern 4). Provider engagement closes the documentation loop (Pattern 5).
For the broader Stars and HEDIS work that uses the same FHIR data, the Top 5 FHIR-based Stars rating measurement patterns for 2026 covers the related domain. For care management work that builds on the same FHIR clinical data, the Best care management platforms built on FHIR data stores covers the related use case.