The Successful Alternative Payment Models from CMMI, To Date


The Center for Medicare & Medicaid Innovation (CMMI) has launched more than 50 alternative payment and care delivery model tests, with 33 models now or still operational, according to CMMI’s sixth report to Congress on its progress.

These model tests have impacted the lives of millions of patients. In the two-year period covering the report, CMMI reports that operational model tests have included over 314,000 healthcare providers and/or plans alone that have impacted the medical care of more than 41.5 million Medicare and Medicaid beneficiaries, as well as commercially-insured individuals. CMMI aims for alternative payment and care delivery model tests to improve the quality of healthcare for these participants while making the delivery of medical care more affordable for all stakeholders.

Qualities of care and cost savings have been the two metrics by which CMMI has evaluated its current and previous model demonstrations. However, few alternative payment and care delivery models have been able to achieve both and even fewer have qualified for expansion based on their cost reductions.

To date, just six model tests have delivered statistically significant savings, net of any incentive or operational payments, CMMI reports. Further, only two of those six models have shown significant improvements in quality and four have met the criteria to be eligible for expansion.

CMMI plans to broaden the definition of success for its models, adding metrics around health equity, person-centered care, and health system transformation. But under the traditional measures, here are the most successful alternative payment models to date, according to the most recent report CMMI sent to Congress.


The Pioneer ACO Model is one of the first accountable care organization (ACO) programs. CMS launched the Model in January 2012, shortly before the Medicare Shared Savings Program (MSSP). It is also one of six CMMI model tests to have delivered statistically significant savings.

The Pioneer ACO Model was designed for healthcare organizations and providers who had care coordination experience. CMS selected 32 organizations to become Pioneer ACOs through a total of five performance years. In its first two performance years, the Model netted shared savings of $134 million in 2012 and $99 million in 2013, the Government Accountability Office (GAO) reported.

The Pioneer ACO Model has also been credited with lowering emergency department and inpatient utilization, improving quality of care, increasing patient and clinician satisfaction, and expanding access to care.

In 2015, the CMS Office of the Actuary certified the expansion of the Model as a permanent part of the Medicare program, saying it would reduce net program spending without negatively affecting quality of care.


The ACO Investment Model (AIM) was designed to encourage new ACOs, especially those in rural and underserved areas, to form and for current MSSP ACOs to transition to greater financial risk. Forty-five ACOs participated in the program throughout its tenure.

The final evaluation report for AIM showed that the Model produced gross and net savings and improved utilization across inpatient admissions, emergency department visits, post-acute care, and readmissions. A 2022 study published in Health Affairs estimated that AIM had net savings to Medicare of $381.5 million over three years, driven by utilization reductions in inpatient and other institutional care and by the absence of shared risk for potential increases in Medicare spending incurred by participants.

Stakeholders have commended CMS’ upfront investment in ACOs. However, the Health Affairs study also showed that the majority of AIM ACOs exited the MSSP when faced with the requirement to assume downside financial risk.


The final evaluation report for the Medicare Prior Authorization Model: Repetitive Scheduled Non-Emergent Ambulance Transport (RSNAT) Model revealed $1 billion in total Medicare savings among Medicare beneficiaries with end-stage renal disease (ESRD) and/or pressure ulcers who benefited from the demonstration during its first 20 quarters.

The RSNAT Model is a prior authorization program that requires free-standing suppliers with ambulances in certain states to get prior authorization for repetitive, scheduled, non-emergent ambulance transport services. Medicare Administrative Contractors can sign off on the services, or suppliers can bypass the prior authorization and have the claim undergo prepayment review.

The prior authorization is meant to reduce improper use of these services—a major contributor to wasteful healthcare spending—while maintaining quality of care for beneficiaries, according to CMS.

Since its launch in 2014, the RSNAT Model has reduced use of the services and related expenditures by 72 and 76 percent, respectively. In light of its reduction of costs and quality improvements, HHS greenlit the model for nationwide expansion in 2019.


The CMS Chief Actuary certified in October 2020 that the expansion of the Home Health Value-Based Purchasing (HHVBP) Model would cut net program spending, according to CMMI’s report to Congress.

The HHVBP Model had been implemented in nine states since 2016. The Model tied Medicare reimbursements to participating home health agencies to quality performance metrics, including utilization, outcome, process, and patient experience metrics. Based on their performance, home health agencies would have their payments adjusted upward or downward by 7 percent in the Model’s last year of limited operation.

The fifth annual evaluation report of the HHVBP Model, published in April 2022, showed cumulative net savings of $949 million in the first five years. The report also found that the model has led to higher quality care in home health agencies within model states compared to home health agencies in non-model states, as well as a reduction in unplanned hospitalizations and use of skilled nursing facilities in model states compared to non-model states.

The HHVBP Model is one of the two models that have demonstrated significant quality improvements in addition to cost reductions.


The Medicare Care Choices Model (MCCM) started in 2016 and, through its five-year run, reduced total Medicare expenditures by 14 percent, even returning $33.2 million to Medicare after accounting for provider payments.

MCCM offered eligible beneficiaries the option to receive hospice service while continuing treatment for their terminal condition. Current Medicare payment rules state that Medicare and dually eligible beneficiaries must forgo Medicare payment for care related to their terminal condition if they want to receive the Medicare or Medicaid hospice benefit. Through MCCM, CMS tested whether a new option would improve quality of life and care, boost patient satisfaction, and reduce spending.

CMMI reports that MCCM significantly reduced inpatient care through increased use of the Medicare hospice benefit and improved quality of end-of-life care for beneficiaries. For example, beneficiaries in MCCM were 26 percent less likely to receive aggressive life-prolonging treatment in the last 30 days of life and spend six more days at home. However, CMMI notes that 60 percent of participants exited the model during its tenure, limiting the demonstration to a small number of hospices.

MCCM is the other alternative payment and care delivery model that has shown significant improvements in quality.


The Maryland All-Payer Model (MDAPM) Model is an ongoing CMMI test that aims to determine whether a rate-setting system for hospital services will improve patient outcomes and reduce costs. So far, global budgets for hospitals have achieved significant savings for Medicare while improving quality, CMMI reports.

The final evaluation report for the MDAPM Model states that hospitals reduced utilization and lowered spending on beneficiaries with multiple chronic conditions and dually eligible beneficiaries more than their counterparts. Commercial insurance also had 6.1 percent slower growth in total hospital expenditures.

Additionally, total Medicare expenditures fell by 2.8 percent and hospital expenditures declined by 4.1 percent without shifting costs to other parts of the healthcare system. A 17.2 percent reduction in outpatient department service expenditures drove Medicare hospital savings.

The MDAPM Model was limited to hospital services, making general applicability difficult to determine. However, CMMI is testing the Maryland Total Cost of Care Model (TCOC), which sets a per-capita limit on Medicare’s total cost of care for beneficiaries in the state.


While just a handful of alternative and care delivery model tests are producing net savings for Medicare, others that have not are still making meaningful improvements in healthcare. CMMI reports that some models have “shown improvements in quality or reduced [low-value] care that generated gross, but not net, savings.” Those models include:

  • The Comprehensive Care for Joint Replacement (CJR) Model
  • The Comprehensive End-Stage Renal Disease (ESRD) Care (CEC) Model
  • The Oncology Care Model (OCM)
  • The Million Hearts®: Cardiovascular Disease Risk Reduction Model

These models—and some others—did not qualify for “expansion of successful payment models” as defined by federal regulations. However, CMMI believes that when a model ends, it can continue to impact the Center’s work, which now explicitly includes advancing health equity, along with quality, affordable care.

For More Information:  The most successful alternative payment models from cmmi to date