1. Introduction
Traditional Medicare is a fee-for-service (FFS) federal health insurance programme for individuals aged 65 and older in the United States, consisting of Part A (inpatient hospital benefits) and Part B (outpatient service benefits), which are financially covered by payroll taxes and monthly premiums. Challenges faced by this Medicare programme feature physicians not wanting to take more Medicare patients due to low reimbursement rates, making care difficult to find for elderly patients, and practices becoming concierge services with high annual fees for direct access (Findeiss, Reference Findeiss2023; Doherty and Physicians, Reference Doherty2015; Cottrill et al., Reference Cottrill, Cubanski and Neuman2024). In an attempt to address these concerns, alternative payment models (APMs) such as bundled payments and shared savings programmes have been developed since 2010 by the Center for Medicare and Medicaid Innovation (CMMI) (Fowler et al., Reference Fowler, Rudolph, Davidson, Finke, Flood, Bernheim and Rawal2023). Payments are tied to outcomes in these APMs – instead of volume of service provided as in traditional insurance programmes – with the ambition that the newer reimbursement methodologies will motivate providers to give higher-quality, higher-value care than before.
Value-based primary care, characterised by a shift from traditional volume-incentivizing FFS models to value-based care (VBC) approaches focusing on quality and outcomes, has led to innovations such as the new Making Care Primary (MCP) model. The MCP model was introduced by the CMS Innovation Center (CMMI) as a pilot programme in eight states on 1 July 2024, for a 10.5-year period. This initiative was designed to test feasibility, financial viability, and patient outcomes before considering national implementation. These states had been selected by CMS on the basis of factors such as their ability to align with state Medicaid agencies, diversity of geographies, population demographics, existing CMS Innovation Center presence, and applicability to the broader Medicare population. While Medicare patients are eligible for services nationwide, this initial statewide rollout allows CMS to assess how the model performs across different healthcare environments.
The updated model also acknowledges the financial pressures facing privately owned primary care practices, particularly those struggling with Medicare’s low reimbursement rates. Many independent practices have either stopped accepting new Medicare patients or transitioned to concierge medicine, requiring patients to pay significant out-of-pocket fees – often exceeding $2000 annually – to cover unreimbursed costs. The MCP model aims to provide a more stable financial structure for primary care practices through prospective capitation payments and risk-adjusted reimbursements, reducing reliance on traditional FFS payments. However, its long-term effectiveness in addressing these financial pressures remains to be evaluated.
The MCP model’s eligible participants feature primary care practices serving FFS Medicare beneficiaries who choose to opt in. These independent practices are included alongside hospital-based ambulatory clinics, as long as they each meet eligibility criteria for serving FFS Medicare beneficiaries. Indeed, the practice must be Medicare-enrolled and bill health services for at least 125 Medicare beneficiaries. Also, at minimum 51 per cent of the practice’s primary care sites must be in one of the eight states that are formally participating: Colorado, Massachusetts, Minnesota, New Jersey, New Mexico, New York, North Carolina, or Washington. Specific organisations are not eligible to participate in the MCP model, including rural health clinics, concierge practices, grandfathered Tribal Federally Qualified Health Centers, and other entities that, as of 31 May 2023, were already participating in other models such as Primary Care First (PCF) or Accountable Care Organization (ACO) Realizing Equity, Access, and Community Health (REACH). Urgent care clinics, unless they offer longitudinal primary care services, are also not eligible for participation. Medicare Advantage (MA) plans, which serve approximately 30 million beneficiaries, are not directly included in MCP, as the model applies only to FFS beneficiaries. However, MCP is designed as a multi-payer model that actively encourages MA plans to align with its methodology. As such, individual MA carriers may choose to adopt aspects of MCP within their existing managed care structures. Of note, beneficiaries dually enrolled in both Traditional Medicare and Medicaid are eligible for MCP, provided they are not served by Medicare-Medicaid Plans under the Financial Alignment Initiative, which operates in some MCP states. While participation in MCP is voluntary for practices, those serving dual-eligible patients must consider the higher costs associated with this population. To address financial risks, MCP incorporates risk adjustment methodologies to ensure more equitable reimbursement for providers caring for complex patient needs (Centers for Medicare & Medicaid Services, 2025). This approach allows for a bespoke implementation, ensuring that only those practices willing to commit to the MCP model’s requirements participate. Importantly, the MCP model does not affect the number of Medicare beneficiaries at a particular practice; it works with the existing pool of Medicare beneficiaries and places them into a new care model. Additionally, the MCP does not necessarily broaden access to needed medical care in emergency settings – it provides rewards only for primary care services. Finally, MCP offers participants three tracks with increasing levels of accountability for cost and quality outcomes based on their prior experience with VBC. As providers move through these time-limited tracks throughout their participation in the programme, the MCP model seeks to gradually incentivize providers to optimise their care coordination workflows and address the clinical and social needs of their patients (Harris, Reference Harris2023).
The need for APMs like MCP is urgent, given post-COVID-19 challenges with staffing issues and reduced FFS payments. States are now taking risks with new models to strengthen primary care infrastructure. Its structure offers foundational support to newcomers and advanced management for seasoned practices, reflecting an understanding of diverse primary care needs. While this initiative signifies a significant step forward in the evolution of value-based primary care, we write to explore challenges that CMMI should be attentive to that could undermine the success of MCP.
2. Past lessons
Previous value-based primary care models (Table 1) launched by CMMI aimed to deliver better care coordination, increased quality, and lower costs for beneficiaries with complex needs. Programmes like the Pioneer ACO and the Medicare Shared Savings Program (MSSP) demonstrated notable cost savings and quality improvements by having providers be incentivized to take on more risk in patient outcomes compared to traditional FFS care (McWilliams et al., Reference McWilliams, Chernew, Landon and Schwartz2015, Reference McWilliams, Hatfield, Landon and Chernew2020; Centers for Medicare & Medicaid Services, 2024e, 2024g). However, the Pioneer ACO faced an issue termed ‘leakage’, whereby patients went out of the ACO network for care, leading to inefficiencies. MCP has improved on these aspects by fostering community partnerships that help mitigate such leakage (McWilliams et al., Reference McWilliams, Chernew, Landon and Schwartz2015; Centers for Medicare & Medicaid Services, 2024c). Moreover, MCP corrects the broad eligibility and limited downside risk of the MSSP by introducing three accountability tracks, which phase in accountability, prevent premature dropout, and curb financial inefficiencies among advanced-value practices (McWilliams et al., Reference McWilliams, Chernew, Landon and Schwartz2015, Reference McWilliams, Hatfield, Landon and Chernew2020; Centers for Medicare & Medicaid Services, 2024g).
Table 1. Previous CMMI value-based primary care models
In contrast, programmes like Independence at Home and PCF, targeting smaller practices, suffered from high attrition and financial losses due to insufficient risk adjustment and misaligned incentives. These issues are addressed by the MCP model by integrating improved risk adjustments and incentives that align more closely with the needs of smaller practices (Mathematica, 2024; Centers for Medicare & Medicaid Services, 2024f). Moreover, unlike the PCF model that had been facing funding issues due to inadequate adjustments for patients with complex illnesses, MCP adopts a broader risk adjustment approach, improving equitable compensation by incorporating social determinants of health (SDOH) (Haft and Berenson, Reference Haft and Berenson2023; Centers for Medicare & Medicaid Services, 2024f).
The Next-Generation ACO model scaled up ACO programmes of the past by increasing financial risk and enhancing primary care delivery, leading to participants having substantial savings while underscoring the importance of tailoring care delivery to population needs. More recently, the ACO REACH model has extended these principles and is currently under implementation and evaluation (Foster et al., Reference Foster, Durand, Johnson, Glover and Narayan2024; Centers for Medicare & Medicaid Services, 2024a). While previous ACO models focused on savings-based incentives, MCP diversifies rewards across various performance measures, facilitating participation for less experienced and independent practices in VBC delivery.
Overall, existing VBC programmes have demonstrated mixed success in balancing cost savings and quality improvements, highlighting essential areas for improvement, such as risk management and incentive alignment.
3. Challenges and steps forward
3.1 Coordinating HRSNs
The ACO REACH model underscored the need to address SDOH, such as housing and food insecurity, particularly in underserved communities, but lacked structured guidance for primary care providers to engage with community organisations. MCP advances this by requiring providers to screen patients for SDOH and establish partnerships with community-based organisations (CBOs), supported by funding incentives (Bibbins-Domingo, Reference Bibbins-Domingo2019). This approach integrates ‘social prescribing’, enabling providers to formally refer patients to social services for essential needs like housing and food support, thereby creating a cohesive method to manage SDOH within primary care (Moore et al., Reference Moore, Unwin, Evans and Howie2022; NHS England, 2024).
While MCP mandates screening and referrals for social health needs, primary care clinics often lack the resources to address these issues directly. Though clinics can identify concerns like food insecurity and social isolation, CBOs and public health organisations are better equipped for hands-on assistance. MCP could improve collaboration by allowing providers to contract directly with CBOs, leveraging each stakeholder’s strengths in meeting patients’ health-related social needs (HRSNs). Additionally, social prescribing, as practised in the UK and other parts of Europe, could be embedded in MCP by integrating social care link workers into primary care, forming robust partnerships to address SDOH.
To further strengthen these partnerships, CMMI could consider providing seed funding and aligning financial incentives by allocating a portion of MCP capitation payments to CBOs. Contract structures could reward CBOs based on measurable outcomes, like reduced hospital readmission rates, to promote comprehensive patient care and foster accountability across stakeholders. Providers are required to utilise part of these payments to strengthen referral networks with specialists, conduct behavioural health screening, and connect patients to CBOs for HRSNs, such as housing and transportation. MCP does not expect providers to directly deliver these services but instead to act as coordinators, leveraging the resources of CBOs.
3.2 Track-based approach and adverse incentives
MCP’s three-track approach offers practices time and resources for improved primary care investment. The MSSP introduced track options with varying risk levels to give providers flexibility; however, its lack of structured progression meant that new practices often struggled with high-risk tracks and chose to exit early, which limited the programme’s effectiveness. MCP addresses this by establishing a clear three-track system where practices start in lower-risk tracks and gradually increase their risk and reward levels over time. Unlike prior models such as MSSP, MCP also includes an Upfront Infrastructure Payment (UIP) in Track 1 to support technology and staffing investments, which tapers off as practices advance. This structure provides smoother transitions and more tailored support, reducing early exits and promoting sustained engagement. Practices entering the MCP model in Track 1 or 2 are required to progress to the subsequent track after 2–2.5 years, ultimately spending 6–10.5 years in the final track depending on their starting track. Tying track progression to fixed time periods and risk capabilities ensures greater accountability commensurate with the organisation’s experience delivering VBC. Despite the fact that this new regulatory system is less burdensome in time to report and bill than existing frameworks, since less specific standards for billing are required, the currently restrictive track design may adversely incentivize inexperienced practices to exit the model early or experienced practices to opt out entirely.
To qualify for Track 1, providers must have no prior experience in VBC delivery. According to the MCP documentation, experience in VBC is defined as having participated in performance-based Medicare APMs, such as PCF, Comprehensive Primary Care Plus, Next-Generation ACO, Direct Contracting, or ACO REACH. Although Tracks 2 and 3 provide greater financial incentives for performance and successful care coordination, only Track 1 offers eligible participants a UIP. Upon progression, the MCP model cuts UIPs for staffing, partnerships, and IT investments like e-consults and patient monitoring – all of which are central to successful primary care delivery independent of practice maturity. To mitigate concerns about reduced compensation, MCP incorporates robust risk adjustment methodologies that account for patient complexity and social determinants of health, ensuring equitable reimbursement. Additionally, practices receive upfront infrastructure payments to support technology and staffing investments, with periodic reviews to adjust payments based on evolving practice needs and performance outcomes. Moreover, periodic peer review where larger organisations share best practices could help under-resourced providers learn without incurring additional costs.
3.3 Multi-payer alignment
A part of MCP, CMMI is providing an opportunity for commercial insurers and Medicaid agencies to align with the model; however, achieving such alignment needs granular guidance beyond the broad parameters of the existing model. The Comprehensive Primary Care and Multi-Payer Advanced Primary Care Practice demonstrations underscore the pitfalls of broad directives. These efforts were hampered by permissive frameworks lacking specific payer engagement and risk adjustment standards, leading to limited or inconsistent results on practice behaviour and patient outcomes (Table 1) (Leung et al., Reference Leung, Beadles, Romaire and Gulledge2019; Jacobs et al., Reference Jacobs, Schreiber, Seshamani, Tsai, Fowler and Fleisher2023).
One example of prior successful multi-payer alignment is the Washington Multi-Payer Collaborative, with a Primary Care Transformation Initiative (PCTI) that aligned payment methodologies across Medicaid, Medicare, and private insurers (Washington State Health Care Authority, 2024; Conrad et al., Reference Conrad, Grembowski, Hernandez, Lau and Marcus-Smith2014). The PCTI includes the Primary Care Practice Recognition (PCPR) programme, which assesses the operational capability of providers and assigns a level (1–3) based on performance. Another example is a voluntary, four-year All-Payer ACO Model initiated in 2017 by the Vermont Legislature. Similar to the PCTI in Washington, Vermont formed the Green Mountain Care Board, an entity responsible for overseeing the development and implementation of the All-Payer ACO Model in the state. The Board’s regulatory authority includes provider rate-setting, health information technology plan approval, hospital and ACO budget approval, insurer rate approval, and oversight of the state’s all-payer claims database. By 2022, the Vermont model had set a target for the state to attribute 70 per cent of the state’s population and 90 per cent of Medicare beneficiaries to an ACO. Following the mandate of the model, Vermont also capped per capita growth in health expenditures at 3.5 per cent and successfully kept Medicare spending growth at 0.1–0.2 per cent below national spending (Grembowski and Marcus-Smith, Reference Grembowski and Marcus-Smith2018; Centers for Medicare & Medicaid Services, 2024h).
However, the MCP needs clear definitions for commercial insurer data sharing with CMMI. Indeed, the Vermont All-Payer ACO Model attempted to align Medicare, Medicaid, and private insurers to ensure uniformity in care quality and reduce costs across payer types, but it struggled due to inconsistent payer participation and vague data-sharing standards. MCP builds on this by mandating detailed data-sharing requirements among commercial insurers, Medicaid, and CMMI, along with clear guidelines on how payers should share quality data and engage in performance-based payments. MCP also requires that all payers adopt standardised quality metrics from a CMMI-approved compendium, ensuring consistency across payers and improving accountability while facilitating a more integrated multi-payer approach. Current MCP documentation lacks detailed payer data-sharing requirements for aligning with CMMI, instead focusing on payer-provider data exchange for interoperability and care coordination. Detailed blueprints for how to collaborate with public health organisations are also vital for the early identification and support of at-risk populations. Moreover, payer contracts should require the selection of quality metrics from a CMMI-approved compendium to eliminate the ambiguity that hinders comparability and accountability. Enforcing these requirements can bolster multi-payer alignment, ensuring that the flexibility intended to promote payer participation does not dilute the model’s efficacy.
3.4 Evaluating and refining quality metrics
The Comprehensive Primary Care (CPC) and CPC+ models included quality metrics focused on core primary care functions, but they primarily emphasised managing hypertension and diabetes without expanding to a broad range of prevalent health conditions (Anglin et al., Reference Anglin, Tu, Liao, Sessums and Taylor2017; Centers for Medicare & Medicaid Services, 2024b). This limited the ability to comprehensively assess the impact of primary care on patient health.
MCP addresses this limitation by including traditional metrics like hypertension and diabetes control alongside a wider array of clinical and social health metrics such as provider performance based on quality metrics like clinical and patient-reported outcomes, screening for SDOH, and financial and utilisation indicators (Table 2). However, while these metrics are valuable, the clinical measures still only focus on managing hypertension and diabetes, ensuring timely colorectal cancer screenings, and monitoring mental health through depression screening and follow-up. Absent are other prevalent conditions like arthritis, obesity, heart disease, substance use disorder, osteoporosis, and thyroid disorders, all of which ought to be included in a forward-looking primary care model (Jacobs et al., Reference Jacobs, Schreiber, Seshamani, Tsai, Fowler and Fleisher2023).
Table 2. Quality metrics in making care primary (Kumar et al., Reference Kumar, Adashi and Kocher2023)
To truly launch into a new era of primary care, the MCP could include 45 patient-centred outcome measures from the International Consortium for Health Outcomes Measurement (ICHOM), covering a wide spectrum of global health conditions (Benning et al., Reference Benning, Das-Gupta, Sousa Fialho, Wissig, Tapela and Gaunt2022). A glide path approach could progressively introduce these adult and paediatric health standard sets, focusing on ICHOM’s metrics across cardiometabolic health, infectious disease, maternal and child health, geriatrics, and oncology.
3.5 Looking ahead: ongoing assessment and iterative improvement
In the PCF model, providers were evaluated periodically to gauge performance, but there was not a structured mechanism for frequent adaptation of programme requirements based on early results, which limited the model’s ability to course correct (Centers for Medicare & Medicaid Services, 2024f). The tiered system of the MCP provides a beneficial update through its inherent iterative approach across the tiers of participation. CMMI should regularly evaluate each of the tracks over 10.5 years to ensure quality improvement and cost reduction. This involves actively assessing each track’s performance, particularly in the first 2–3 years, focusing on spending, outcomes, and disparities. CMMI should promptly modify underperforming tracks to align with objectives before broadening participation. CMMI should also track participating primary care practices’ performance longitudinally in 2–3-year increments, evaluating their ability to meet quality and cost benchmarks. Practices that significantly underperform in the first measurement period should be given a targeted improvement plan, as providers still underperforming by the end of years 5–6 may be unlikely to subsequently perform well in years 7–10.
Ensuring the model’s success requires monitoring its impact on underserved populations while maintaining adaptability in response to the evolving landscape of clinical medicine over the next 10.5 years. This evolution, marked by artificial intelligence-supported clinical workflows and expanded roles for nurse practitioners and physician assistants in primary care, underscores the need for continuous adaptation (Lin et al., Reference Lin, Sattler and Smith2020). Despite existing barriers, effective oversight in the MCP could revolutionise primary care payments and pave the way for broader Medicare reforms.
Acknowledgements
None.
Financial support
None.
Competing interests
None.
Open access
