Using Value-Based Care to Inform Physician Payment Reform
New research from The Parity Center explores how healthcare utilization patterns under full financial risk contracts could guide changes to Medicare’s physician fee schedule.
Medicare’s physician fee schedule does more than determine reimbursement rates: it shapes financial incentives for the services clinicians provide. Payments largely reflect the resources required to deliver each service, but policymakers have long questioned whether those rates adequately encourage high-value care.
A study published in JAMA Health Forum by Aaron Schwartz, Amol Navathe, and colleagues from The Parity Center and collaborating organizations uses a novel approach to payment reform: examining the care delivered when physician organizations take full financial responsibility for patients’ medical spending. In particular, this is a novel way to utilize empirical data on utilization to identify under- and over-valued services. The research was supported by Arnold Ventures.
Comparing practice patterns across payment arrangements
Using Humana Medicare Advantage data from 2015–2019, the researchers compared utilization of 1,271 individual services among beneficiaries receiving care under full-risk contracts to those under fee-for-service arrangements. The analysis adjusted for detailed beneficiary characteristics, geography, and time.
Under full-risk contracts, physician organizations are responsible for total medical spending, giving them incentives to consider how individual services affect patients’ broader care needs and costs. Examining their practice patterns offers a way to identify services that these organizations use more or less frequently than organizations paid through fee-for-service.
What the study found
Adjusted service use varied substantially across payment arrangements:
- 74% of services had lower utilization under full-risk contracts.
- 26% of services had higher utilization under full-risk contracts.
- Lower-level office visits and laboratory services were generally more common, while higher-level office visits, hospital and emergency services, and rehabilitative therapies were less common.
These differences varied across individual services, demonstrating why only examining broad categories of care alone can overlook patterns relevant to payment reform.
Translating those patterns into payment benchmarks
The researchers used the observed utilization differences to simulate a new possible Medicare fee schedule. The simulation estimated how payment rates could change to encourage service-use patterns that resemble those observed under full-risk contracts, based on an assumption about how clinicians would respond to reimbursement changes.
A budget-neutral version designed to keep aggregate modeled spending unchanged yielded a possible 6% increase in reimbursement to general practice physicians. This illustrates how the approach could rebalance payments toward services associated with primary care.
Because fee-for-service remains a foundation of health care payment, including many value-based payment models, these benchmarks could help policymakers align incentives for individual services with broader goals for spending and outcomes.
What these benchmarks can and cannot tell us
The study offers evidence to inform fee-schedule reform rather than a definitive set of payment rates for adoption.
The observed differences reflect both payment arrangements and the capabilities and care approaches of organizations that voluntarily participate in full-risk contracts. They do not isolate the causal effects of those contracts. Higher or lower utilization also does not, by itself, establish whether a service is clinically appropriate or overvalued or undervalued.
Nevertheless, the findings provide a framework for incorporating real-world practice patterns into decisions about physician payment—and for testing whether new rates can better support the goals of value-based care.
Read the study
Schwartz AL, Bulat T, Ben-Michael E, et al. JAMA Health Forum. 2026;7(10). doi:10.1001/jamahealthforum.2026.3533.