A Decade of Value-Based Payment: Lessons Learned And Implications For The Center For Medicare And Medicaid Innovation, Part 1

Over the past decade, adoption of value-based payment (VBP) models has increased substantially. The Center for Medicare and Medicaid Innovation’s has launched multiple new models since its authorization, commercial payers have structured almost one-third of their payments as alternative payment models (APMs), and state Medicaid programs have increasingly included payment reforms in their waivers and Medicaid managed care contracts.

Nevertheless, questions remain on how payment and care reforms can best reduce overall costs and improve quality, outcomes, and people’s experience of care. Overall evidence on cost and quality outcomes of the diverse reforms to date is mixed, with some payment models performing better than others. A recent assessment from the Centers for Medicare and Medicaid Services (CMS) has also raised questions about the impact of many of the pilot reforms implemented by the Center for Medicare and Medicaid Innovation (CMMI) during its first decade.

Despite these challenges, interest in payment reform and pressure to engage in reform continue to increase. The experience of COVID-19 has reinforced strong interest in VBP, given the limited ability of fee-for-service payment to support providers during the pandemic. COVID-19 will likely drive down health care spending for 2020 because patients decided to delay or forgo usual care, especially prevention-oriented care. This in turn may contribute to additional worsening of health outcomes, and the need to build on the shift to home-based and virtual care. Continuation of emergency telehealth and site-of-service waivers can continue to provide incremental support for these shifts, but bigger reforms are needed to support home-based services like remote monitoring, digital apps, and coordination with community care teams that are hard to sustain under fee-for-service payment.

In addition, COVID-19 accelerated steps to understand and address social factors that contribute to poor health and health inequities. While such steps may not reduce costs, especially in the short term, health care organizations increasingly view them as critical components of value and their mission.

Further, the underlying health spending growth trends before the pandemic were approximately 4.6-4.7 percent. This was lower than prior years, but still higher than inflation and economic growth, indicating a continued need to deal with high and rising health care spending. Health care spending did decrease during the pandemic but is expected to rebound in 2021, potentially surpassing pre-pandemic spending growth.

The years of reform experience, recent trends, plus a new Administration that has expressed a commitment to effective payment reform have led to a number of thoughtful assessments of how to proceed from here. In this post, we reflect on “lessons learned” from the evidence gained from the past decade and beyond to promote the effectiveness of APMs. We describe the implications for future VBP efforts and the implications for CMS and CMMI.

As Exhibit 1 summarizes, these lessons span the development, implementation, and maturity and refinement of VBP models. Part 1 of this post discusses lessons concerning development and implementation; part 2 will discuss lessons regarding the maturity and refinement of VBP models.

Exhibit 1: Twelve lessons learned from VBP in the last decade.

Source: Authors’ analysis

Designing Models To Meet System-Level Goals

Value-Based Payment Models With Small Shifts Away From Fee-For-Service (FFS) Generally Have Small Impacts

Evaluations have repeatedly highlighted how small shifts away from fee for service payments create only modest impacts. This makes intuitive sense—if an APM only affects a small percentage of an organization’s payments, there is little business case for investing in the infrastructure and personnel needed to transform care. The majority of the VBP model options available to date haven’t required large shifts away from FFS and only a handful of Medicare models (e.g. one part of the Bundled Payments for Care Improvement [BPCI] program) have been mandatory. The lack of advanced and mandatory models has likely limited the impact of payment reforms on costs and outcomes.

Creating a bigger shift from fee-for-service can happen in two ways. First, each individual model should have more meaningful accountability, which is seen in models like the Comprehensive Care for Joint Replacement bundled payment program, the recent iteration of the Medicare Shared Savings Pathways to Success Program, and commercial models with large shifts from FFS, as well as global budget models. Second, there is a need for more action across payers, so that more of a provider’s book of business is in advanced APMs. Anecdotal feedback from care delivery organizations we’ve spoken with suggests that at least 50 percent of a book of business needs to be in a significant value-based arrangement to support the substantial shifts in care delivery needed to succeed in new care models.

VBP Enables Provider Resiliency In The Face Of Emergencies

VBP can be a tool for building a better health care system during and after the pandemic. The pandemic highlighted the fragility of fee-for-service payment, with FFS-dependent clinics and facilities facing disastrous financial situations when their utilization and revenue dropped during shelter-in-place orders. Subsequently, fear of COVID infection and other stressors have led patients to defer services, particularly non-urgent preventive and primary care services.

In contrast, providers under VBP were better positioned to respond, as they could leverage other revenue sources (e.g., shared savings, monthly capitated payments) for financial stability. Further, organizations participating in VBP had already developed new capabilities as a part of their transition to VBP (e.g., care management staff, more robust data infrastructures) and were able to modify or redeploy those capabilities for the care reforms needed during COVID.

Many VBP-related reforms can support health care organizations during emergencies. For example, organizations used their population health information systems to identify individuals at high risk of COVID-19 complications or who may have had unmet health needs due to social distancing; deployed their care coordinators for contract tracing or reaching out to high-risk individuals; and ramped up telehealth as part of more comprehensive virtual-care models. Beyond enabling greater access to care, support for enhanced virtual care is also a more efficient way to enable hospitals and health systems to flex in an emergency than simply trying to pay for more capacity under FFS.

Some commercial-payer and state-led payment reforms in the pandemic have gone further, linking financial relief to further progress on payment reform. For example, BlueCross BlueShield of North Carolina’s (BCBS NC) Accelerate to Value program provided upfront financial assistance to independent primary care practices for a commitment to later participate in a VBP model. Practices have the option to join an existing accountable care organization in BCBS NC’s value-based care program, Blue Premier, or can partner with Aledade, an organization that helps primary care practices move to value-based care.

VBP Helps Providers Take A Population View, But More Focused Accountability And Intentional Model Design Is Needed To Improve Health Equity

Shifting to VBP could allow the health care delivery system to improve long-standing health disparities by broadening the focus beyond health care for the patient who comes into the clinic to include the factors that are most important for improving outcomes. VBP models based around prospective payments to providers have more flexibility to support the delivery of services (e.g., nutritional support, transportation services) that can address many of the social factors that impact health outcomes and contribute to disparities in outcomes.

Intentional design considerations can help assure that payment models support health equity. First, models need to include more equity-focused quality measures (such as “disparities-sensitive” measures). Programs can start by improving available data on race, ethnicity, and other sociodemographic characteristics, to allow for stratified reporting by race and ethnicity. These measures can be linked to payment in ways that support providers caring for higher-risk populations.

For example, Rhode Island’s Medicaid program supports social needs screenings and health disparities, and Michigan Medicaid supports evidence-based interventions to improve disparities in birthweight. These payment models could be reinforced by Federal support for regional infrastructure investments to coordinate between VBP organizations and community-based social service providers—for example, through Section 1115 waivers modeled after North Carolina’s Healthy Opportunities pilots, or CMMI initiatives based on findings from the Accountable Health Communities pilot.

Additionally, adjustments of benchmarks for the social needs of a patient population would enhance implementation without compromising incentives to achieve equitable outcomes. For example, Massachusetts Medicaid provides higher global budgets to plans with higher-risk patients, determined by a combination of medical and behavioral diagnoses, social factors, and neighborhood risk scores. In Minnesota, providers in the Integrated Health Partnerships demonstration receive quarterly population-based payments that are risk adjusted on social factors such as homelessness, mental illness, and past incarceration.

Finally, safety-net organizations and lower-resourced organizations need opportunities to participate and succeed under new payment models. For example, most Federally-Qualified Health Centers do not have a path to shift to accountable population-based payments.

Implementing Models to Impact Patient Care

VBP Models Need To Be Framed Around Improvements For Patients, With Benefit Design Aimed At Improving Access And Outcomes While Lowering Costs

To date, value-based payment initiatives have focused much more on providers than engaging the public they serve, even though such engagement would improve the effectiveness of new payment models. Particularly in Medicare, changes in payment models generally are not linked to complementary “value-based” changes in benefit designs or premiums, such as shared savings or new benefits if patients select providers participating in APMs. Developers of future payment models should examine where the models can link to and build on other patient-focused incentives , implement value-based insurance design, and learn from the lessons of behavioral economics on incentives. For example, CMMI could expand the incentives for patients to participate in models (such as lower copays for certain services or providers), and all payers could increase patient and consumer involvement in model design and implementation.

Further, public information about value-based payment is often confusing, which leads to consumer skepticism that such efforts are really about improving care rather than denying needed services or limiting choices. Evidence from the Medicare Shared Savings Program has highlighted that many people are confused when informed that they are attributed to an Accountable Care Organization (ACO), which may suggest that few people will enroll through voluntary attribution efforts. For Medicare beneficiaries, CMMI can help providers design clear communications on the benefits of participating in a value-based model and the positive effects on outcomes; employers should do the same in commercial models. For example, Walmart has developed a communication and engagement strategy to address employee questions related to tiered benefits tied to Centers of Excellence.

Finally, in advancing APMs, it’s important for policymakers and health care leaders to highlight that cost savings is not the only goal of care reforms. As payment reforms allow more flexibility in the way providers deliver care, it’s likely that some interventions will add to costs, and not just in the near term. It is still valuable, for example, if a targeted expenditure on social services leads to a significant improvement in quality of life or reductions in hospitalization risk for a patient, even if the cost isn’t fully recovered. Patients don’t expect their provider to forgo a highly cost-effective drug therapy just because it adds to costs. Implementing such interventions will build trust for implementing the care reforms needed to reduce costs, such as reducing use of low-value procedures and services.

Reforming Care To Improve Outcomes And Equity And To Lower Costs Is The Goal, Not Payment Reform For Its Own Sake

The point of payment reform is to enable needed care redesign that otherwise can’t be implemented or sustained. Success requires health care organizations to commit to changing care models—to be driven by culture change to encourage a whole-patient focus and reduce low-value care.

For example, some organizations shift partially into APMs, but continue to compensate physicians using activity-based approaches like relative value units (RVUs), or they may share savings with their primary care physicians while still compensating specialist clinicians based on volume and intensity. With financial supports that don’t match the intended goals, .

It Takes Resources And Investments To Build The Organizational Competencies Needed To Redesign Care Successfully Under VBP Models

Our analyses of ACO data have shown that organizations often needed at least 3 years to generate consistent savings. This matches informal feedback we have heard from those organizations, where many ACOs report that the first year is spent trying to design and test reforms based on program rules and data, the second year is when they refine and scale promising care initiatives, and then they can start to demonstrate savings in their third year. Individual organizational performance varies considerably, indicating the lack of a clear and consistent path to success.

Ensuring that more organizations successfully implement the needed infrastructure and organizational changes for care redesign requires some up-front capital, technical assistance, and/or partnerships. Such supports are especially important for small, independent, or rural providers. To provide needed upfront funding, payers can offer model options similar to the ACO Investment Model or the new direct contracting models (which envision flexibility for technically capable organizations to support participating practices). Alternatively, many providers are seeking to develop needed capabilities such as data analytics, care managers, telehealth and digital supports, social service coordination, and behavioral health integration through partnerships with companies that specialize in such services. Providers can partner with such companies to gain access to upfront capital and technical assistance. Medicare Advantage and Medicaid incentives for health plans can also support care reforms by augmenting provider capabilities.

Readiness assessments may also help providers and payers determine feasible steps towards success in VBP models. These assessments can help stakeholders understand what infrastructure they need to build and where they may be lacking resources. For example, the Accountable Care Learning Collaborative offers an analytical model called the “Value-Based Readiness Assessment,” which measures organizational readiness for the shift to value-based payment. Similarly, the Health Care Transformation Task Force offers a “Partnership Evaluation Tool” and other resources to help organizations assess their readiness to participate in value-based partnerships.

Transparency And Predictability In Payment Reforms Enable Confidence By Health Care Organizations To Make Needed Investments

Given the level of investment, commitment, and capital resources needed to succeed under value-based payment models, health care delivery organizations will want some certainty before committing. At the same time, implementing these reforms is a complex process that should get simpler as evidence and experience accumulate. One way to provide more confidence under such conditions is to provide more transparency, so that organizations can more accurately predict and verify how they are likely to perform, and can have more confidence in the evidence base for deciding when modifications are needed.

However, the details of payment reforms such as methods and data for calculating performance against benchmarks are often not publicly disclosed, with some payers viewing the information as proprietary. To set the tone for increased transparency, CMMI could publicly release more information regarding key features of demonstrations like attribution rules and benchmarks, and provide more timely data sharing using improved “Blue Button” application programming interfaces (APIs), which would help organizations know what to expect should they join an APM, and the reasons for and consequences of adjustments in the APM.

Moreover, there is limited information available on model evaluation, especially by commercial payers. CMMI could improve the available evidence by synthesizing the key themes from its diverse set of evaluations and make that regularly updated data source publicly available. CMMI could also provide templates for other participants in its multipayer models to share more findings.

Summing Up

The past decade of payment reform experience has provided valuable lessons regarding the design and implementation of value-based payments. Further progress in adopting VBP models should build on these lessons learned, to best reduce costs and improve quality, outcomes, and people’s experience of care. Tomorrow, in Part 2 of this post, we will describe lessons learned on maturing and expanding VBP models and summarize key themes from the past decade of payment reform.

Authors’ Note

This post is part of a collaborative initiative on value-based payment reform supported by and in partnership with West Health. In We want to thank the following people for their thoughtful discussions on these topics covered in this paper: David Brailer, Don Berwick, Michael Bailit, Don Crane, Rick Gilfillan, Mara McDermott, Elizabeth Mitchell, Mark Smith, Hemi Tewarson, and Judy Zerzan.

Robert Saunders has a consulting agreement with Yale-New Haven Health System for development of measures and development of quality measurement strategies for CMMI Alternative Payment Models under CMS Contract Number 75FCMC18D0042/Task Order Number 75FCMC19F0003, “Quality Measure Development and Analytic Support,” Base Period.

Aparna Higgins serves on the National Cardiovascular Disease Registry Management Oversight Committee and receives consulting fees from MITRE and IPRO.

Mark McClellan is an independent board member on the boards of Alignment Healthcare, Cigna, Johnson & Johnson, and PrognomIQ; cochairs the Guiding Committee for the Health Care Payment Learning and Action Network; and receives fees for serving as an adviser for Arsenal Capital Partners, Blackstone Life Sciences, and MITRE.