On March 12, 2025, the Centers for Medicare & Medicaid Services’ (CMS’s) Center for Medicare and Medicaid Innovation (CMMI) announced that it intends to end early four payment models to achieve $750 million in savings: the Primary Care First (PCF) and Making Care Primary (MCP) Models; the mandatory ESRD Treatment Choices (ETC) Model; and the Maryland Total Cost of Care (TCOC) Model. CMMI also announced that it would not move forward with two previously-announced models, the Medicare $2 Drug List and the Accelerating Clinical Evidence Models. CMMI further indicated that it is “considering options” to reduce the size of the Integrated Care for Kids Model or to make other changes to the model.
While this is not the first time that CMMI has ended a model before its initially targeted expiration date — indeed, CMMI announced an early end to the Medicare Advantage Value-Based Insurance Design (MA VBID) Model this past December — a simultaneous early closing of four models is atypical. CMMI has indicated that, at present, it intends to continue moving forward with its other active models, including the Accountable Care Organization (ACO) Primary Care Flex (PC Flex) Model. Though, in a separate statement (also issued on March 12), CMMI indicated that it plans to announce a “new strategy” to streamline the focus of demonstration models based on guiding principles of making Americans healthier by preventing disease through evidence-based practices, empowering people with information to make better decisions, and driving choice and competition.
Below, we review the context of the four terminated models — including the implications of the closure of the two primary care models (PCF and MCP) models and the ETC and Maryland TCOC Models.
Primary Care Models (PCF and MCP Models)
Over the past decade and a half, CMMI has trialed numerous payment and delivery models increasingly focused on prospective payments for primary care services. The PCF and MCP Models were two of those primary care models. While the early termination of the PCF and MCP Models apparently reflects a lack of cost savings, it does not appear to reflect a deviation from CMMI’s primary care strategy — especially as CMMI chose to continue the ACO PC Flex Model, as well as other ACO models open to primary care physicians.
The PCF and MCP Models are two in a line of CMMI models that have focused on enhancing primary care but have failed to generate savings. In 2012, CMMI developed the Comprehensive Primary Care Initiative (CPCI) as an effort to invest in primary care practices to improve the quality of care they provided by paying out monthly care management fees. After CPCI failed to report savings, CMS developed the Comprehensive Primary Care Plus (CPC Plus) Model in 2020, which added some capitated payments for more advanced primary care practices, though it too failed to demonstrate savings. The PCF Model, which began in 2021, combined flat per-visit payments, prospective population-based payments, and a performance-based adjustment based on a set of fixed quality measures. The Trump Administration had initially introduced the PCF Model prior to the COVID-19 pandemic as a part of its Direct Contracting Model (later retitled and reformed into the ACO REACH Model), and the Biden Administration eventually launched the model as a standalone in 2021. The more recent MCP Model began in July 2024 and introduced a 100% capitation option for more advanced practices. Both the PCF and MCP Models maintained the general theory that care quality improvements would lead to cost savings that offset additional investments in primary care.
In terminating the PCF and MCP Models, CMMI expressed concern about their costs. An evaluation report for 2022 showed that the PCF Model increased Medicare costs without affecting patient outcomes. The authors of the evaluation report wrote that they “did not expect to detect improvements in the primary outcomes after only two performance years” and that it was “too early to draw conclusions” about the PCF Model’s effectiveness. No evaluation has yet been posted of the MCP Model, which is only mid-way through its initial performance year.
Even in canceling the PCF and MCP Models, CMMI expressly emphasized its ongoing commitment to expanding and strengthening primary care. CMMI stated, “[p]rimary care remains a foundational component of the Center’s strategy. The early termination of Primary Care First and Making Care Primary does not signal a retreat from the Center’s support of primary care providers, but rather a need to focus on different approaches that are consistent with the CMS Innovation Center’s statutory mandate and produce savings.”
CMMI continues to promote and study primary care through its ACO models, which have primary care providers at the center of patient assignment and cost control methodologies. Thus, it is notable that CMMI continued the ACO PC Flex Model, CMMI’s most recent primary care model, which is aimed at addressing health care spending through both primary care management and ACO formation—thereby combining two of CMMI’s long-time pursuits into a single model design. CMMI initially anticipated selecting approximately 130 ACOs to participate in the model, though just 24 ACOs currently participate and CMMI “does not intend to offer another round of applications for the ACO PC Flex Model.” The model is anticipated to run from 2025 to 2030.
ETC Model
CMMI also announced its intent to end early the ETC Model. Since that model was initiated via notice-and-comment rulemaking, CMMI indicated that it will propose ending the model early through the rulemaking process.
The ETC Model, which began January 1, 2021, made several changes to reimbursement rates for certain clinicians and approximately thirty percent (30%) of end stage renal disease (ESRD) facilities, assigned based on their location. The changes included an increase for three years in home dialysis reimbursement; a performance-based adjustment to incentivize increasing home dialysis and kidney transplant rates; and an “Equity Incentive” for “significant improvement in the home dialysis rate or transplant rate of beneficiaries who are dual eligible for Medicare and Medicaid (DE) or Low Income Subsidy (LIS) recipients[.]” See also 42 C.F.R. § 512.300 et seq. After the first two years, the model showed an increase in deceased donor transplants, but no statistically significant improvement in home dialysis rates, Medicare spending and utilization, or most objective measures of care quality.
The decision to end the ETC Model does not necessarily indicate a broader desire to limit kidney care reform. During President Trump’s first term, he issued an executive order on Advancing American Kidney Health and CMMI launched the optional Kidney Care Choices (KCC) Model, which allows Kidney Care Entities to take financial risk for the cost and quality of care for individuals with ESRD. CMMI’s March 12 announcement did not affect the KCC Model, and the model is set to continue until 2026.
Maryland TCOC
Since the 1970s, CMS has used a series of demonstration models to exempt Maryland from normal hospital payment rules and allow the State to establish hospital rates across different payors (i.e., to allow Maryland to set the amount that payors in the state would pay hospitals). The Maryland TCOC Model is the latest iteration in this long-standing rate-setting experiment; it began in January 2019 and was originally set to conclude in December 2026.
The TCOC Model involves a per capita limit on Medicare total cost of care in Maryland, and builds upon and expands its immediate predecessor, the Maryland All-Payer Model, which launched in 2014. The TCOC Model includes a hospital payment program, a care redesign program, and a primary care program modeled after the PCF Model to incentivize the provision of advanced primary care services.
When the TCOC Model was established, Maryland committed to generating $2 billion in Medicare savings over the planned eight-year performance period, but performance has not reached that level. According to a progress report, from 2019 to 2021, the model generated $689 million in net savings to Medicare, primarily from hospital global budgets that lagged behind national spending growth from 2014 to 2019. For the first two years of the model, this encouraged hospitals to shift care to non-hospital settings, but, more recently, reductions in hospital spending have failed to offset the corresponding increases in non-hospital spending. The primary care component of the TCOC Model generated estimated net losses of $90.3 million annually from 2019 to 2022.
CMMI had already announced that Maryland would be the first state to join the Advancing All-Payer Health Equity Approaches and Development (AHEAD) Model, beginning in January 2026. As such, pending discussions with the state, CMMI anticipates that the termination of the TCOC Model (on December 31, 2025) would lead to a transition into the AHEAD Model starting in 2026. However, parts of the AHEAD Model, such as the statewide, all-payer TCOC target, are not set to be implemented until 2027. We will await CMMI guidance on this transition.
Conclusion
CMMI’s announcement of its intent to terminate four demonstration models reflects the new Administration’s emphasis on regularly assessing whether demonstration models are achieving anticipated cost-savings, while preserving or enhancing quality of care. Providers participating in other demonstration models should also account for the possibility of other changes related to their models, especially if CMMI moves forward with its plans to announce a “new strategy” on Innovation Center demonstration models.
[View source.]
