Highlights
- Transitioning to risk-based contracts in Medicare Advantage (MA) yields inconsistent reductions in healthcare utilization, with primary effects limited to emergency department visits and cardiovascular stress testing in upside-only models.
- The adoption of two-sided risk contracts (including financial penalties) did not demonstrate significant changes in utilization after accounting for pre-existing temporal trends.
- Crucially, neither upside-only nor two-sided risk models were associated with a reduction in the use of 26 measured low-value services across six clinical domains.
- The findings suggest that financial risk transfer to healthcare organizations is insufficient, on its own, to eliminate wasteful clinical practices or systematically optimize care delivery within the MA framework.
Background
The transition from fee-for-service (FFS) reimbursement to value-based care (VBC) represents the most significant paradigm shift in U.S. healthcare financing over the last two decades. At the heart of this transition is the use of risk-based contracts, which shift financial responsibility for patient outcomes and costs from insurers to healthcare organizations. Within the Medicare landscape, Medicare Advantage (MA) has become the primary laboratory for these models. Unlike traditional Medicare, where Accountable Care Organizations (ACOs) often struggle with voluntary participation and patient churn, MA plans utilize Health Maintenance Organization (HMO) and Preferred Provider Organization (PPO) structures that are theoretically better suited for managed care.
Risk-based contracts generally fall into two categories: upside-only risk (where providers share in savings but lose nothing if costs exceed benchmarks) and two-sided risk (where providers are liable for losses). While the economic rationale suggests that these incentives should lead to more efficient utilization and a reduction in “low-value” services—defined as care that provides little to no clinical benefit and may cause harm—empirical evidence has been sparse. Understanding the impact of these contracts is critical as the Centers for Medicare & Medicaid Services (CMS) aims to have 100% of Original Medicare beneficiaries in an accountable care relationship by 2030.
Key Content
Methodological Advances in Assessing Risk-Based Impacts
In a landmark retrospective cohort study published in JAMA Internal Medicine (Schwartz et al., 2026), researchers addressed the evidence gap by analyzing claims data from 2015 through 2021 for beneficiaries enrolled in Humana’s MA HMO plans. The study utilized a difference-in-differences (DiD) framework, a robust econometric method that compares the changes in outcomes over time between a group that transitioned to a new contract and a control group that did not. This approach is essential for isolating the effect of the contract transition from broader secular trends in healthcare, such as the overall decline in certain services during the COVID-19 pandemic.
The study analyzed 658 organizations moving from FFS to upside-only risk and 114 moving to two-sided risk, covering millions of beneficiary-years. This scale provides high statistical power to detect even modest changes in clinical behavior.
Impact on Healthcare Utilization
The analysis of nine utilization measures across three domains (inpatient, outpatient, and testing) revealed nuanced results:
- Upside-Only Risk: Initially, four outcomes showed reductions (ED visits, primary care visits, advanced imaging, and cardiac stress testing). However, after adjusting for differential temporal trends prior to the contract change, only two remained statistically significant: emergency department visits (-8.4% of baseline) and cardiovascular stress testing (-12.1%).
- Two-Sided Risk: Despite the higher financial stakes associated with potential penalties, organizations transitioning to two-sided risk did not show consistent, statistically significant reductions in utilization after trend adjustment. Initial observations of reduced specialty visits and advanced imaging were found to be continuations of pre-existing trends rather than direct results of the contract change.
These findings suggest that while risk contracts may prompt some organizations to tighten protocols around discretionary high-cost items like stress tests or steer patients away from the ED, the broad “utilization dampening” effect expected by policymakers is not universally realized.
The Persistence of Low-Value Care
One of the most sobering findings of recent research is the negligible impact of risk-based contracts on low-value services. Schwartz et al. tracked 26 specific measures, including unnecessary cancer screenings in older adults, preoperative testing for low-risk surgeries, and redundant imaging for low back pain.
Neither the transition to upside-only risk nor the move to two-sided risk resulted in a meaningful decrease in the aggregate use of these services. This lack of change was consistent across all six domains (cancer screening, diagnostic/preventive testing, preoperative testing, imaging, cardiovascular procedures, and other surgeries). This suggests that financial incentives at the organizational level do not easily translate into clinical decision-making changes at the point of care.
Comparison with Traditional Medicare ACOs
The findings in Medicare Advantage contrast slightly with some data from the Medicare Shared Savings Program (MSSP) in traditional Medicare, where some modest reductions in low-value care have been documented. However, those reductions were often small and took several years to manifest. The MA context is unique because many organizations are already operating under some form of managed care oversight, which may create a “ceiling effect” where the most egregious low-value care has already been mitigated, leaving a harder-to-address core of entrenched clinical habits.
Expert Commentary
The failure of risk-based contracts to significantly move the needle on low-value care points to several systemic barriers. First, there is the incentive dilution effect. While a healthcare organization (a large group or hospital system) may be at risk, the individual physician seeing the patient often remains compensated on a productivity basis (e.g., Work Relative Value Units or wRVUs). If the financial risk does not reach the person ordering the test, the behavior is unlikely to change.
Second, there is the information gap. Many clinicians are not aware of which services qualify as “low-value” according to specific claims-based metrics. Without real-time clinical decision support (CDS) embedded in the electronic health record (EHR) that flags an unnecessary preoperative EKG or a redundant vitamin D screening, physicians will continue to practice defensive medicine or follow outdated protocols.
Third, the voluntary nature of transitions may lead to selection bias. Organizations that choose to enter risk-based contracts may already be “leaner” in their utilization, leaving less room for further improvement. Conversely, those with the highest rates of low-value care may avoid risk-based models to protect their revenue streams.
From a biological and mechanistic perspective, the reduction in cardiovascular stress testing is notable. It reflects a growing clinical consensus that routine stress testing in asymptomatic or low-risk patients often leads to a cascade of unnecessary invasive procedures (e.g., cardiac catheterization) without improving survival or preventing myocardial infarction. The fact that risk contracts *did* impact this specific measure suggests that highly publicized, evidence-based guidelines can be synergistic with financial incentives.
Conclusion
The transition to risk-based payment models in Medicare Advantage has not yet proven to be the panacea for healthcare inefficiency. While there is evidence of modest reductions in specific areas like ED visits and cardiovascular testing under certain contract designs, the broader goal of eliminating low-value services remains elusive.
For clinicians and policymakers, these findings emphasize that changing how we pay for care is only the first step. To truly improve value, healthcare organizations must pair financial risk with robust clinical redesign, including physician-level incentives, advanced data analytics to identify waste, and point-of-care decision support. Future research should focus on whether longer-term exposure to risk or higher-magnitude financial penalties eventually forces the clinical changes that initial voluntary transitions have failed to produce.
References
- Schwartz AL, Kim S, Chhatre S, et al. Changes in Health Care Utilization and Low-Value Service Use After Risk-Based Contract Adoption in Medicare Advantage. JAMA Intern Med. 2026;186(1):98-107. doi:10.1001/jamainternmed.2025.5917. PMID: 41212579.
- McWilliams JM, Hatfield LA, Chernew ME, Landon BE, Schwartz AL. Probing the Limits of Value-Based Payment—Underperformance of the Shared Savings Program. N Engl J Med. 2020;383(11):1083-1090.
- Colla CH, Mainor AJ, Hargreaves C, et al. Changes in Use of Low-Value Services and Hospital Outcomes in Accountable Care Organizations. JAMA Intern Med. 2016;176(10):1503-1510.
