Risk Adjustment
Paying a health plan more for sicker enrollees. In Medicare Advantage the CMS-HCC risk score is built from submitted diagnoses, so coding accuracy is money.
Risk adjustment is the mechanism that pays a health plan according to how sick its members are expected to be, rather than a flat amount per head. CMS explains the reason in its 2024 report to Congress: without an adjustment for health status, a plan has “a strong incentive” to attract members who are healthier than average, and a plan that happened to enroll the highest-cost members “would have difficulty remaining viable.” The adjustment is what makes enrolling a frail member financially survivable.
In Medicare Advantage the arithmetic runs through the CMS-HCC model. Diagnoses reported for a member during the data-collection year map into hierarchical condition categories, those categories carry relative factors, and the resulting risk score scales the plan’s monthly payment. CMS estimates the factors from the fee-for-service population rather than from plan data. The design point to hold onto is that the money follows documented diagnoses, so what gets coded, and whether the medical record supports it, is the whole exposure.
That is why AI shows up here in a form unlike the rest of health insurance AI. The tools do not deny anyone care. They read charts to surface conditions that were treated but never coded, suggest codes to a reviewer, and rank charts for human review. An accuracy gain and an audit exposure are the same capability pointed in two directions, and the record that decides which one a regulator sees is the link from suggested code back to the encounter that supports it. Our health operations guide sets out that evidence chain next to the ones for prior authorization and claims.