Risk-Based Contracts Reveal Wide Service-Level Shifts in Utilization
A large claims analysis in Medicare Advantage found that service-level utilization diverges sharply between full-risk organizations and traditional fee-for-service arrangements, with the gap playing out across individual physician services rather than as a uniform overall pattern. The study examined 1,271 services and showed that 940 of them, or 74%, were used less often in the full-risk cohort after adjustment for beneficiary characteristics. In 331 services, utilization was higher under risk-based contracting. Those differences were not trivial noise around a single payment model; they were distributed across office visits, laboratory testing, emergency care, hospital services, and rehabilitation in ways that suggest payment structure is shaping clinical behavior at the level of day-to-day service delivery.
The practical significance is easy to miss if attention stays fixed on aggregate spending alone. Fee schedule debates often focus on whether payments should rise or fall in broad categories, yet physicians and health systems make decisions service by service: how often to schedule follow-up, whether to order repeat labs, when to refer to higher-intensity settings, and how much rehabilitative care to use after acute illness. In a full-risk model with 100% downside risk for total medical spending, those choices carry direct financial consequences. The utilization pattern in this cohort suggests that organizations under full-risk contracts appear to lean toward lower-intensity outpatient management and away from more resource-intensive services when compared with fee-for-service counterparts.
A striking asymmetry emerged in the mix of services. Low-level office visits and laboratory services were more common in risk-based settings, while higher-level office visits, hospital and emergency services, and rehabilitative therapies were less common. In clinical terms, that pattern is consistent with care that is managed more proactively in the ambulatory setting, with fewer escalations to expensive sites of service. It also fits the logic of capitation-like accountability: when organizations bear total spending risk, there is a strong incentive to prevent avoidable acute care use and to substitute relatively low-cost monitoring for more intensive encounters where clinically appropriate.
Why Service Mix, Not Just Spending, May Matter for Fee Schedule Reform
The study used Humana Medicare Advantage claims, encounter, and administrative data from 2015 to 2019 and included 585,487 beneficiaries attributed to full-risk contracts and 1,153,455 beneficiaries in fee-for-service contract arrangements. The analysis was retrospective and cross-sectional, but the scale is large enough to make the service-level contrasts difficult to dismiss as idiosyncratic. Data were analyzed from September 2025 to June 2026, and the comparison was adjusted for detailed patient-level differences, an important feature because service use in different payer models can otherwise reflect case mix rather than payment incentives.
Service-level benchmarking matters because the physician fee schedule contains more than 1,000 prices, and policy proposals often struggle to identify which rates should change and by how much. A blunt across-the-board adjustment can miss the real leverage points. By looking at the utilization profile associated with full-risk contracts, the investigators created a different kind of reference point: not what services cost in theory, but what service mix tends to occur when organizations are paid to manage total risk. In that sense, the benchmark is behavioral as much as financial. It links payment policy to the clinical decisions that payment policy is supposed to influence.
The secondary simulation pushed that idea further. Investigators modeled a modified Medicare fee schedule using the observed utilization differences and a supply elasticity drawn from prior literature. Under that budget-neutral reform, reimbursement to general practice physicians could rise by about 6%. The estimate is not a prediction of what would happen in the real world tomorrow, but it does show how service-level utilization patterns can be translated into a plausible redistribution of fee schedule dollars. If lower-intensity services are more common in risk-based care and those patterns are used as a benchmark, primary care-oriented work may gain relative value in the fee schedule while more intensively priced services could lose ground.
The Policy Signal Is Strong, but the Clinical Question Is Narrower
The distribution of utilization differences across 1,271 services argues against a one-size-fits-all approach to fee schedule reform. Some services appear more common in risk-based contracts, some less common, and many may sit near the middle. The broad implication is that payment reform should not be built on a single average effect across all physician work. It should account for the fact that different services respond differently to financial incentives, especially when total-cost accountability is present.
For clinicians, the result also invites a more operational question: which service changes reflect better coordination and which reflect underuse? Lower emergency and hospital utilization may indicate more effective longitudinal management, better triage, and fewer avoidable acute episodes. Lower rehabilitative therapy use could reflect tighter care management, but it could also raise concerns about access if reductions exceed what clinical need would justify. The study design cannot separate those possibilities, and it does not measure patient outcomes such as symptom control, functional recovery, or mortality. Utilization is a useful policy signal, but it is not a direct measure of quality.
Still, the direction of the findings is clinically intelligible. Full-risk contracts reward substitution away from expensive downstream care and toward earlier, lower-cost contact when problems first surface. That kind of incentive structure can favor shorter office visits, basic monitoring, and restrained use of high-intensity services. It can also expose where the fee schedule may be misaligned with value, especially if payment levels continue to reward face-to-face intensity more than prevention, coordination, and longitudinal management. The simulation suggests that a benchmark rooted in full-risk utilization could support a redistribution of physician reimbursement without increasing total Medicare spending.
The study leaves an important policy limitation in plain view: service use patterns in one insurer’s Medicare Advantage population may not map cleanly onto the broader Medicare program, and observed differences do not prove that every lower-utilization service reflects efficiency. Even so, the analysis offers a concrete framework for fee schedule reform at a time when policymakers are again revisiting how more than 1,000 physician services should be priced. The more actionable question is not whether full-risk contracts differ from fee-for-service, but which of those differences should be imported into payment policy and which should remain a feature of a different delivery model altogether.