Medicare Advantage Works Until the Math Changes
UnitedHealth Group announced it will exit Medicare Advantage plans covering about 600,000 beneficiaries after rising utilization and drug spending disrupted its financial projections. The company pointed to higher emergency department intensity, expanded diagnostic testing, and rapidly growing costs from oncology drugs, obesity treatments, and gene therapies. Leadership also described expanded auditing and tighter payment integrity controls to manage expenses.
The decision matters because Medicare Advantage exists as a substitute administrative layer for a public entitlement. Congress did not design it as a niche option. The program now enrolls more than half of Medicare beneficiaries. When the largest carrier retreats from segments of that population, the event signals a mismatch between payment design and medical reality rather than a routine product adjustment.
The mechanics sit in statute. Medicare Advantage pays insurers a benchmark rate tied to county level fee for service spending. Plans then receive additional revenue based on risk adjustment scores calculated from diagnosis coding. The system assumes private plans will manage utilization more efficiently than traditional Medicare and therefore profit while maintaining benefits. That assumption holds only when utilization trends remain predictable and drug pricing grows at a pace aligned with annual benchmark updates.
The current moment breaks that alignment. Emergency departments now function as the diagnostic front door of the healthcare system. Hospitals increased imaging and laboratory testing during acute encounters partly because malpractice exposure and quality reporting encourage comprehensive workups. Each additional test increases both claim cost and risk score documentation. The plan pays immediately and may recover some revenue in the following year through risk adjustment, but the recovery never matches the speed of expense growth.
Prescription drugs introduce a second instability. Oncology biologics, GLP 1 metabolic agents, and gene therapies enter the market with launch prices that exceed the assumptions built into payment formulas written years earlier. A therapy costing $300,000 in a single episode erases the margin generated by dozens of routine primary care encounters. Capitation works only when variance stays within actuarial tolerance. Catastrophic pharmacy spending destroys that tolerance.
The insurer response follows the contract. Exit products with broader provider choice because they limit utilization control. Increase audit intensity because retrospective claim review shifts financial exposure to providers. Adjust pricing because premiums and bids represent the only levers available under federal rules. Each action protects the balance sheet without changing underlying medical pricing.
This behavior did not begin this quarter. The same pattern appeared during the early managed care era in the 1990s when plans entered markets aggressively and withdrew once hospital utilization exceeded projections. It repeated after the introduction of risk adjustment in the mid 2000s when coding accuracy improved faster than payment recalibration. The program oscillates between expansion and contraction because its architecture assumes medical inflation behaves gradually. Modern therapeutic innovation behaves discontinuously.
Patients experience the oscillation as administrative instability. A beneficiary selects a plan believing enrollment reflects a durable coverage relationship. The withdrawal notice reframes coverage as conditional participation in a financial instrument. Formularies change. Authorizations restart. Continuity of care depends on the next carrier’s contracting decisions rather than clinical necessity.
Hospitals and physicians will argue that insurers should absorb higher costs because they accepted capitation risk. Insurers will argue providers expanded diagnostic intensity beyond reasonable expectations. Pharmaceutical manufacturers will argue research investment justifies pricing. Each position remains internally rational. None address the structural mismatch. Payment benchmarks update annually while medical technology advances unpredictably.
The program claims to transfer efficiency from private management to public spending. In practice it transfers volatility from national budgets to individual beneficiaries. Traditional Medicare spreads extreme costs across the entire population. Medicare Advantage concentrates them inside discrete plan contracts. When spending spikes, the contract breaks rather than the budget.
The announced exit represents an early signal of a broader adjustment cycle. More plans will narrow networks and reduce geographic coverage as high cost therapies proliferate. Congress will face pressure to modify risk corridors or carve out specific drug categories from capitation calculations. Without such changes insurers will continue treating enrollment as an option rather than an obligation.
The system currently performs exactly as designed. It invites private capital to administer a public benefit while retaining public responsibility for affordability. Private actors enter when margins exist and depart when they vanish. Beneficiaries remain the constant variable moving between administrative regimes.
The real choice lies in deciding which entity absorbs medical innovation risk. If society wants stable coverage regardless of therapeutic price volatility, the government must directly finance catastrophic treatments. If society prefers private administration without additional public spending, coverage instability will continue. The present arrangement attempts both goals simultaneously and therefore achieves neither.
The withdrawal affecting 600,000 people does not represent a failure of management. It represents a predictable outcome of placing a fixed payment model on top of an accelerating cost curve. Unless payment design changes, future announcements will follow the same script and beneficiaries will continue learning that enrollment in a Medicare Advantage plan guarantees access only until the math changes.