Risk adjustment pays health plans according to how sick their members are, using diagnoses coded into Hierarchical Condition Categories (HCCs). Arkangel AI reviews charts against the configured model, links each possible HCC to the chart text that supports it, and leaves every coding decision to your coders and clinicians.
Last updated:
HCC coding is the documentation and coding of diagnoses that map to Hierarchical Condition Categories (HCCs). In Medicare Advantage, CMS calculates each enrollee’s risk score with the CMS-HCC risk adjustment model, from demographics and documented diagnoses, so every reported condition must be supported in the medical record.
Summary of public CMS documents, checked November 1, 2026. It is not coding or legal advice.
Sources
AI-powered chart analysis across all patient records for unreported conditions and coding gaps
Experience the power of AI-driven risk adjustment with our comprehensive trial program:
Streamline your Risk Adjustment Data Validation process with AI-powered workflows
RADV audits validate that submitted diagnosis codes are supported by medical records. To prepare, AI reviews claims data against documentation, flags unsupported or missing codes, and routes exceptions to coders with prioritized action items. It then generates audit-ready summaries—so when CMS requests records, your HCC coding is already defensible and fully traceable.
AI validates submitted diagnosis codes against medical records, identifying discrepancies and documentation gaps before audits.
Streamlined workflow for exceptions and missed opportunities with prioritized action items for your coding team.
Detailed summaries flagging risks and revenue opportunities with audit-ready documentation.
Industry-leading technology backed by clinical expertise
Review received records for documented conditions and supporting evidence under the configured risk-adjustment model. Clinicians and coders validate candidates, exclusions and applicable reporting rules. Confirm the EHR connector and permissions separately; coverage depends on available records. Measure supported corrections and actual model outcomes without assuming recall, RAF improvement or unchanged staffing.
Findings are linked to source records for validation by coders and clinicians.
Confirm the specific EHR, connector version, permissions and data formats in a synthetic pilot before integration.
Data-driven improvements with clear metrics and trending analysis.
Try our platform with no commitment. See results before you decide.
How AI-powered risk adjustment compares with manual coding on capture, speed, and audit defensibility.
Compare manual and assisted HCC review on the same received cohort, applicable model and evidence criteria. Measure supported candidates, missed records, adjudicated errors and reviewer effort separately. Assistance can organize source evidence while clinicians resolve exceptions. Neither review method establishes complete capture or a RAF increase without validated coding and model-specific outcome data.
| Capability | Arkangel AI | Manual HCC review |
|---|---|---|
| Chart coverage | Received records with risk-adjustment candidates and exceptions | Reviewer examines the agreed sample or cohort |
| Turnaround | Measure processing and review time on the pilot | Measure reviewer time on the same records |
| Coding accuracy | Link candidates to documentation and the configured model | Reviewer records model criteria and supported decisions |
| Cost to scale | Measure assisted throughput and clinical approval workload | Measure reviewer capacity, staffing and review time |
| RADV readiness | Retain sources and exceptions for reviewer approval | Reviewer prepares the supporting documentation and decision |
Join leading healthcare organizations using AI to optimize risk adjustment and improve patient outcomes.
Request a DemoEverything you need to know about AI-powered chart intelligence for risk adjustment
See how chart intelligence captures the HCCs buried in your records