Moving Toward High-Value Health Care: Integrating Delivery System Reform into 2020 Policy Proposals

Abstract

• Issue: Delivery system reform has been a focus of regulatory and legislative policy to date, but it is unclear how policymakers will integrate reforms into their plans for 2020 and beyond.
• Goal: To present and evaluate options for integrating delivery system reform into upcoming legislative proposals.
• Methods: Literature review.
• Findings and Conclusions: Policymakers should integrate delivery system reform into their 2020 plans to continue driving value in the health care system. Several options exist for promoting delivery system reform either through a state-based block grant approach or federal public plan approach. We identify three main principles that are critical for success of reform efforts: information sharing and infrastructure, flexibility to innovate, and alignment and stability of efforts.

Watch archived video of live Health Affairs event held Nov. 16, 2018. (Registration required.)

Introduction

Increasing quality. Rewarding value instead of volume. Coordinating and integrating care. All are common refrains of health care delivery system reform from both Democrats and Republicans. In a 2016 statement to the House Ways and Means Committee, Sylvia Burwell, the U.S. Secretary of Health and Human Services (HHS) under President Obama, noted, “We are focused on moving towards a health care system that delivers better quality of care, spends dollars in a smarter way, and keeps people healthy.”¹ Two years later, in remarks to the Federation of American Hospitals, current HHS Secretary Alex Azar declared, “There is no turning back to an unsustainable system that pays for procedures rather than value. In fact, the only option is to charge forward — for HHS to take bolder action, and for providers and payers to join with us.”²

Despite considerable efforts across the private and public sectors to change how the U.S. health care system provides and pays for care, more progress is needed. Health care spending for 2018 will rise a projected 5.3 percent, up from a 4.7 percent rise in 2017.³ And even though alternative payment models (APMs) such as accountable care organizations (ACOs) are designed to encourage greater risk-sharing and better coordination of care, a recent survey found that providers in only half of ACOs share in downside financial risk.⁴

As policymakers from both parties look to improve population health and contain spending, delivery system reform can and should be incorporated into any future proposal. In this issue brief, we first describe the fundamentals of delivery system reform. We then outline the likely health reform proposals for the 2020 presidential election — including a Republican bill that focuses on state block grants and two Democratic proposals that incorporate broader public options — and discuss how these plans could promote delivery system reform. Finally, we offer a framework for assessing these proposals.

Background

The Fundamentals of Delivery System Reform

Reform involves altering care delivery, payment incentives, or both to stimulate and sustain delivery system changes. Following existing literature, we organize delivery system reform loosely into three general categories: accountable care organizations (population-based care management), medical homes (individual-level care coordination), and bundled payments (episode-specific coordination) (Exhibit 1).⁵

The evidence base for these payment and care delivery models demonstrates promising trends but is still growing as models continue to be developed and evaluated. Actual care delivery improvements have lagged payment changes, particularly in managing population health beyond the walls of the physician’s office. Indeed, the authors of a recent ACO survey noted that, “ACOs are slowly becoming willing to accept increased financial risk, but they are largely still learning how to actually manage populations.”⁶

Given the lack of any “gold standard” health care delivery model as well as the variation in populations, markets, and geographies across the country, delivery system reform continues to be an evolving process of innovation and evaluation. In designing policies that incorporate new payment and care delivery models, we have identified three main principles that are critical for success: information sharing and infrastructure, flexibility to innovate, and alignment and stability of efforts.

Information Sharing and Infrastructure. Electronic health record (EHR) interoperability and the development of patient-owned medical records are crucial for providers to better manage their patient populations across different sites of care, including primary care and specialty clinics. Coupled with improved interoperability, the development of health information exchanges can provide more macro-level data for population management, such as tracking readmissions to hospitals in different health systems. Data can catalyze improvement, including provider-specific and patient-level information on the processes, cost, and outcomes of care. Additionally, bringing such information and data to the point of care can better engage patients in clinical decision-making, addressing a challenge in current delivery models.

Flexibility to Innovate. Various provider types, patient populations, and local markets respond to different incentives. Moreover, providers and patients across the country have different expectations of how they interact with each other and navigate the health care system — interactions that are affected by the history of the region, market fundamentals such as provider and plan concentration, geographic characteristics, and patient socioeconomic characteristics. Enabling flexibility to adjust models to the needs of particular environments can contribute to success. To date, the Centers for Medicare and Medicaid Services (CMS) has created a variety of Medicare models for different types of providers and patients, from disease-specific models to approaches tailored to the needs of rural areas. State Innovation Models Initiative grants have provided states with the opportunity to implement multipayer health care delivery reforms across Medicaid, Medicare, and the Children’s Health Insurance Program (CHIP). Maintaining a balance of tailored and nationwide approaches will further enable policymakers to meet the diverse needs across the country.

Alignment and Stability of Efforts. Lack of alignment on the expectations, incentives, and measures of accountability across private and public payers, purchasers, and providers could dilute the focus of reform efforts and severely hamper systemwide change. If providers are held accountable to completely different quality metrics and payment structures depending on what type of insurance a patient has, they are less likely to consistently change their behaviors and how they provide care. Alignment of program characteristics — such as how a patient gets “assigned” to a particular provider, what quality metrics are used for performance evaluation, and how financial rewards or penalties are calculated and allocated — will play an important role in ensuring delivery system reform efforts are as effective as possible. Alignment could occur at different levels, such as state or federal, or across payers or providers. Although alignment is important to optimize investment in these models and reduce burden on providers, it should be balanced with the need for flexibility as discussed above.

Sharing lessons from successful delivery system reform efforts with those designing and participating in such initiatives will allow for stability and improvement over time. To this end, a public–private partnership, the Health Care Payment Learning and Action Network, has developed an APM framework and white papers to provide coordinated and consistent guidance on the various aspects of APMs.⁷

Incorporating Delivery System Reform Fundamentals into Health Reform Proposals

Recent debates provide insights on the central elements of each political party’s approach to health reform. The Republican plan likely would focus on replacing the Affordable Care Act’s (ACA’s) Medicaid expansion and health insurance marketplaces with state block grants for health care services and a federal per-enrollee spending cap on the traditional Medicaid program, much like the bill introduced by Sens. Lindsey Graham (R–S.C.), Bill Cassidy (R–La.), Dean Heller (R–Nev.), and Ron Johnson (R–Wis.) in 2017. This approach also would repeal the employer mandate and promote the use of health savings accounts through tax breaks.

Meanwhile, most Democratic plans follow the broad approach of building on the ACA and developing some form of public plan option. The Medicare for All Act (S. 1804), introduced by Sen. Bernie Sanders (I–Vt.) in 2017, would largely replace private insurance and Medicaid with a taxpayer-funded, Medicare-like program. Several “Medicare for more” proposals also have been made, including “Medicare Part E” (S. 2708) introduced by Sen. Jeff Merkley (D–Ore.). This bill would make Medicare an option for “everyone,” including individuals and small and large businesses. Other approaches include making Medicare available in areas with little insurance competition or provider shortages (“Medicare X”) and introducing a Medicare buy-in for individuals ages 50 to 64.⁸

Below we discuss the potential to incorporate delivery system reform into Republican and Democratic reform proposals.

Delivery System Reform in the Graham–Cassidy–Heller–Johnson (GCHJ) Bill. The defining feature of the GCHJ bill is that states are funded through block grants to design their own health care reform initiatives. Given the state-centric nature of this and other Republican proposals, incorporating delivery system reform into these plans will require incentives for states to engage in care delivery and payment models. These models could be existing Medicare models, such as ACOs, bundled payments, and medical homes (Exhibit 1), or new models recommended by the Health Care Payment Learning and Action Network or the Physician-Focused Payment Model Technical Advisory Committee.⁹

Adjustments to block grants could incentivize states to embrace innovative care delivery and payment models (Exhibit 2).

For example, states with 20 percent of their Medicaid and marketplace provider payments in APMs could receive a 2 percent bonus on their block grant, and states with 50 percent of their provider payments in APMs could receive a 5 percent bonus on their block grant. States also could receive block grants with an annual growth rate that is lower than current health care cost growth, providing further motivation for states to embrace delivery system reform. Over time, as APMs become more prevalent, the eligibility thresholds for bonuses could increase.

Another potential path for ensuring delivery system reform is to require states to incorporate performance-based payments into their contracts with Medicaid managed care and marketplace private plans (Exhibit 3).

To further incentivize states, the federal government could promote a “race to the top” approach in which states that make substantial progress with delivery system reform (e.g., improvements in quality metrics) receive a financial reward or enhanced shared savings opportunities (Exhibit 4). For example, states that achieve a threshold level for their populations on well-established metrics (e.g., those used by current ACO models) could be eligible for a 5 percent bonus on their block grant.

To encourage continued development of programs promoting population and community health, policymakers could provide incentives for the adoption of these models (Exhibit 5).

Finally, state promotion of information exchanges, all-payer databases, and provider resource networks could be supported with additional funding through block grants. Alternatively, an investment in data sharing could be required for receiving a block grant, along with mandating payer and provider participation in information networks.

Delivery System Reform in Medicare for All/Medicare Part E. Sen. Sanders’ Medicare for All proposal replaces private insurance with a combination of original Medicare and Medicare Advantage, while Sen. Merkley’s Medicare Part E bill expands original Medicare as a private insurance alternative. Both original Medicare and Medicare Advantage already contain some approaches to delivery system reform, including Medicare APMs and Medicare Advantage capitation and quality bonuses. Medicare for All and Medicare Part E could build upon this by expanding payment models in the new populations covered by original Medicare.

For instance, these proposals could require that 50 percent of payments in the new populations are through APMs by 2024. As discussed above, these payment models could be existing Medicare models or emerging models. Because Medicare Part E does not replace private insurance, it could extend Center for Medicare and Medicaid Innovation authority for delivery system reform demonstrations to marketplace plans, so demonstration projects could involve both the public plan option and qualified private marketplace plans.

Under the expansion of Medicare Advantage in Sen. Sanders’ bill, delivery system reform could be further embedded in contracts with private insurers, similar to the approach described for GCHJ and the example of Covered California (Exhibit 3). The public plan also could fund population health and community programs (Exhibit 5). Participation in health information exchanges and all-payer databases could be incentivized or required for contracted plans and providers as well. Given that Medicare Part E does not replace private insurance, the legislation could adopt a similar approach for all marketplace plans.

Discussion

Exhibit 6 presents an assessment of how the different health reform proposals discussed here reflect the three main principles of information sharing and infrastructure, flexibility to innovate, and alignment and stability of efforts — all of which are required to generate more value from the health care system.

The GCHJ approach maximizes flexibility because of its state block grant nature. States have substantial freedom to tailor their reforms to their specific health insurance and delivery system environments. However, while the state-based approach maximizes flexibility, the small number of people in some states could limit providers’ ability to share financial risk and change their care delivery model. Further, many states have not been the traditional source of design and implementation support for delivery system reform efforts, often delegating these initiatives to managed care plans, including those in Medicaid. Depending on the engagement of managed care plans going forward, it may be challenging for some states to introduce delivery system reform projects.

The state-based block grant approach also facilitates improvements in information and data sharing. To date, states have been the primary organizers of health information exchanges and all-payer data systems, and block grants provide a path for funding such an infrastructure. Through the block grant approach, states also have the potential to engage multiple payers to improve participation in information sharing.

The state-based approach of GCHJ may be less likely to promote alignment and stability. Provider systems that cross state lines would be subject to several different programs, increasing provider burden and making it harder to drive change. With 50 different state-based “laboratories,” it also will be difficult to track progress and ensure stability over time without a concerted effort by the federal government to encourage sharing of best practices. Consistency across payers also may be an issue. If provider rates under public insurance drop substantially because of lowered funding, then the public and private insurance markets may become more segmented, making it less feasible to align efforts and incentives.

In contrast to GCHJ, the Sanders and Merkley approaches are nationwide in scope. This national focus would ensure alignment of efforts across state lines and populations, as well as the ability to test multiple models and scale those that are successful. The Medicare program has developed strategies to tailor efforts to local provider and patient characteristics, including through local technical assistance and initiatives such as the Quality Improvement Organization program. Nevertheless, a Medicare-centric approach likely would be less state-specific and have less flexibility than the GCHJ approach.

In terms of information sharing, the nationwide development of an informational infrastructure could promote data exchange, particularly across state lines. However, the focus on expanding original Medicare in the Sanders and Merkley proposals creates some limitations in engaging multiple payers to create an information exchange. The public plan proposals could improve information sharing and build an informational infrastructure if private payers are involved in the expansion of Medicare Advantage (as part of Sen. Sanders’ Medicare for All) or marketplace plans (as part of Sen. Merkley’s Medicare Part E).

Conclusion

Ongoing and future delivery system reform efforts should continue to build the evidence base of what works and what does not in the move toward improved health and smarter spending in our health care system. Policymakers and health care leaders also should recognize and strive to overcome limitations of APMs and other value-based payment models to date. First, reforms should address the continued lag in meaningful care improvements. Second, methodologies for calculating incentives and tracking spending and financial savings over time should continue to evolve to address regional variations in costs.¹⁰ Third, increasing patient engagement in these models should be an ongoing focus.

The health reform proposals in the 2020 presidential election present a prime opportunity to continue the push for value in our care delivery system. With health care a key issue for many Americans and concerns about health costs growing, we have a window of opportunity to not only address existing limitations but to break new ground in health system performance improvement.

Acknowledgments

The authors thank Jeanne Lambrew of the Century Foundation and David Blumenthal of the Commonwealth Fund for their thoughtful comments and suggestions.