A Multipayer Approach to Health Care Reform

A Multipayer Approach to Health Care Reform A N I S S U E B R I E F

OFFICERS J. Barclay Collins II Chairman James R. Tallon, Jr. President William M. Evarts, Jr. Patricia S. Levinson Frederick W. Telling, PhD Vice Chairmen Sheila M. Abrams Treasurer David A. Gould Sally J. Rogers Senior Vice Presidents Sheila M. Abrams Deborah E. Halper Vice Presidents Stephanie L. Davis Corporate Secretary DIRECTORS Richard A. Berman Jo Ivey Boufford, MD Rev. John E. Carrington Philip Chapman J. Barclay Collins II Richard Cotton Richard K. DeScherer William M. Evarts, Jr. Paul Francis Michael R. Golding, MD Josh N. Kuriloff Patricia S. Levinson Howard P. Milstein Susana R. Morales, MD Robert C. Osborne Peter J. Powers Katherine Osborn Roberts Mary H. Schachne John C. Simons Howard Smith Michael A. Stocker, MD, MPH Most Rev. Joseph M. Sullivan James R. Tallon, Jr. Frederick W. Telling, PhD Mary Beth C. Tully Howard Smith Chairman Emeritus United Hospital Fund The United Hospital Fund is a health services research and philanthropic organization whose mission is to shape positive change in health care for the people of New York. We advance policies and support programs that promote high-quality, patient-centered health care services that are accessible to all. We undertake research and policy analysis to improve the financing and delivery of care in hospitals, clinics, nursing homes, and other care settings. We raise funds and give grants to examine emerging issues and stimulate innovative programs. And we work collaboratively with civic, professional, and volunteer leaders to identify and realize opportunities for change. HONORARY DIRECTORS Donald M. Elliman Douglas T. Yates Honorary Chairmen Herbert C. Bernard John K. Castle Timothy C. Forbes Barbara P. Gimbel Rosalie B. Greenberg Allan Weissglass

A Multipayer Approach to Health Care Reform Sean Cavanaugh D IRECTOR OF H EALTH C ARE F INANCE U NITED H OSPITAL F UND Gregory Burke C ONSULTANT UNITED HOSPITAL FUND

Copyright 2010 by United Hospital Fund ISBN 1-933881-05-4 Free electronic copies of this report are available at the United Hospital Fund s website, www.uhfnyc.org.

Contents FOREWORD iv INTRODUCTION 1 THE RATIONALE FOR MULTIPAYER COLLABORATION 3 IMPEDIMENTS TO MULTIPAYER ACTION 6 IMPLEMENTATION ISSUES 8 OPPORTUNITIES FOR MULTIPAYER ACTION 10 MOVING FORWARD: MODELS OF MULTIPAYER ACTION 10 THE IMPORTANCE OF STATE LEADERSHIP AND PARTICIPATION 12 CONCLUSION 14 REFERENCES 15 iii

Foreword Health reform has many moving parts. Although much attention has focused on insurance expansions, bigger changes may occur in the redesign of the service delivery system. In this paper, Sean Cavanaugh, the Fund s director of health care finance, and Gregory Burke, an independent consultant, investigate the need for health care payers to work together in new ways to realize the promise of health care reform. The Affordable Care Act provides a wide array of new payment and delivery system experiments all targeted at reducing waste, improving quality, and controlling costs. To achieve these results, health care providers will need to fundamentally reengineer how they deliver care. But we cannot expect providers to undertake such fundamental changes unless they receive broad-based support from the numerous payers with whom they do business. This paper identifies several models for payer collaboration. We do not know how these new relationships will evolve and which ones will succeed, but we can gain some insights from cost containment efforts of the past and a few experiments in payer cooperation already occurring in New York. Whether the new model for service delivery turns out to be medical homes, accountable care organizations, or something altogether different, the provider community needs clear and consistent payment to support positive changes. With this paper, the United Hospital Fund begins a conversation about how to make this happen. JAMES R. TALLON, JR. President United Hospital Fund iv

Introduction Since the passage of the Patient Protection and Affordable Care Act (ACA) in March 2010, debates have started about whether the new law can successfully control health care costs and improve health system performance. Some critics have claimed that the new law is long on coverage provisions but short on cost controls, but supporters point out that the bill includes essentially every major cost control proposal that has been circulating in policy circles in recent years (Orszag and Emanuel 2010). Indeed, the law provides for primary care medical homes, accountable care organizations, an independent payment advisory board for Medicare, and numerous other cost control mechanisms. But an essential element of this debate has been overlooked. The ACA reaffirmed our nation s dependence on multiple health plans, both public and private, to provide health coverage to different segments of the population. While a multipayer system allows for more consumer choice and innovation, analysts have argued convincingly that this division of responsibility among multiple payers contributes to high costs in American health care (Anderson, Reinhardt, Hussey, and Petrosyan 2003; Vladeck and Rice 2009). Efforts by individual payers to control costs have repeatedly run aground against a delivery system resistant to change or have been circumvented through cost shifting. Consequently, no strategy is likely to be successful in controlling health care costs unless payers work together to give providers consistent incentives to improve performance. Motivating a diverse, multipayer health care delivery system to make the needed changes will require a broader effort than anything any one payer could undertake by itself. Since the demise of health planning in the 1980s, federal efforts to control health care costs and improve quality have focused largely on payer-specific initiatives developed and implemented by Medicare or the Veterans Administration to benefit that program s beneficiaries. It now appears a shift in this approach may be imminent. Increasingly, policymakers acknowledge that improving health system performance will require fundamental changes in the structure and performance of the health care delivery system (Guterman and Drake 2010). The Obama administration signaled its shift in strategy from government programonly to broader-based health system reform efforts with the announcement of the Medicare Advanced Primary Care Practice Demonstration in late 2009. 1 This is the first Medicare demonstration of a new payment system that explicitly requires the participation of Medicaid, private health plans, and self-insured plans since the allpayer hospital reimbursement demonstrations of the 1970s and 1980s. This strategy is carried further in the ACA. The federal reform legislation supports multipayer approaches to service delivery reform in several important instances: Preference is given to accountable care organizations that have multipayer support (Section 3022); The new Center for Medicare and Medicaid Innovation is instructed to allow states to test and evaluate systems 1 The Centers for Medicare and Medicaid Services (CMS) provides details of the demonstration at http://www.cms.gov/demoprojectsevalrpts/md/itemdetail.asp?itemid=cms1230016 (accessed October 25, 2010). A Multipayer Approach to Health Care Reform 1

of all-payer payment reform (Section 3021); and The Secretary of Health and Human Services is directed to align the efforts of public and private payers with regard to quality and patient safety initiatives (Section 3011). This new vision for change appears to be taking hold among private health plans, too. During the negotiations surrounding the ACA, commercial health plans collectively agreed to simplify the administrative requirements they impose on providers in order to reduce health care costs. Many payers are now collaborating on physician credentialing through the new Council on Affordable Quality Healthcare. And it appears this collaboration is now slowly spreading to collaborative efforts at payment reform. Learning From the Adirondack Medical Home Demonstration In this paper, we used the Adirondack Medical Home Demonstration (AMHD) as a starting point for thinking about multipayer reform efforts. The AMHD is a multiprovider, multipayer collaboration involving essentially all primary care practices serving about 200,000 residents in a five-county region of northeast New York State. 2 The aim of this five-year demonstration project is to strengthen the Adirondack region s ability to attract and retain primary care physicians by improving their quality of work-life and increasing their income. At same time, the demonstration seeks to transform the delivery of primary care in the region by increasing quality and continuity of care, reducing avoidable utilization and costs, and improving the health of the community. To achieve these ambitious goals, a critical mass of physicians in the Adirondacks have committed to implementing the National Committee for Quality Assurance s Physician Practice Connections Patient-Centered Medical Home standards and guidelines in their practices, and to achieving a variety of access, quality, and utilization targets over the next five years. Medical homes provide a range of enhanced services to patients, such as more timely access to appointments, more assistance with self-management of chronic diseases, and a team-based approach to patient care. In return, the participating payers will reimburse those providers $7.00 per member, per month. This payment over and above the normal payments for office visits is intended to cover the costs of the expanded services of a medical home. The impetus for this initiative was a crisis in this rural region s ability to attract and retain an adequate supply of primary care physicians. One rural hospital participating in this demonstration noted that it had lost 18 out of 65 primary care physicians in the 12 months prior to the demonstration. The physician community identified the patientcentered medical home model as its best chance of improving primary care for physicians and patients alike. Over a fouryear planning period, physicians, payers, state officials and community leaders worked to achieve broad consensus on the structure and funding of this demonstration. The demonstration officially began on January 1, 2010. Over the next five years, the physician practices will be held to a series of performance benchmarks to ensure the new payments are generating savings by reducing avoidable emergency department visits and inpatient stays that will finance these improvements in patient care. This paper examines the reasons why multipayer collaborations such as the AMHD are growing and identifies the conceptual and practical challenges to continued growth in 2 The Adirondack Medical Home Demonstration spans four counties within the Adirondack region in their entirety (Clinton, Essex, Franklin, and Hamilton), and parts of a fifth (Warren). 2 United Hospital Fund

payer collaboration. The underlying analysis relies on interviews conducted by the authors with leaders from participants in the AMHD, including health plans, provider organizations, and state officials. We also spoke with participants in other payment reform demonstrations in the Hudson Valley and Rochester, and with one large national payer involved in several projects across the country. Many of the ideas in this paper come from those individuals and organizations interviewed, and we are greatly indebted to their insights. Nevertheless, the opinions expressed in this paper are the authors alone, unless otherwise indicated. The Rationale for Multipayer Collaboration Market Fragmentation The structure and performance of the health care delivery system is influenced by many factors, including the financial incentives embedded in the payment systems used by public and private insurers. Payers have tried numerous strategies for improving health system performance, yet most are frustrated with the level of progress made in controlling costs and making fundamental improvements in quality. This limited success is due, in part, to the fact that individual payers typically constitute a small share of any provider s revenue and thus have limited leverage to effect meaningful change. The data in Figure 1 show the share of patients involved in the AMHD from each major payer. Figure 1: Payer Mix of the Practices Participating in the Adirondack Demonstration Other, Including Uninsured, 21% A, 16% Medicare, 18% D, 6% B, 10% C, 8% A D: Private Plans with >6,000 Enrollees E: Three Private Plans with <6,000 Enrollees Medicaid, 13% E, 8% Participating payers represented 62% of all active patients in participating practices. A Multipayer Approach to Health Care Reform 3

These data demonstrate the fundamental problem of scale. Because health coverage is fragmented among multiple public and private health plans, any individual health plan typically represents a small share of any practice s patients and, therefore, has insufficient leverage to spur improvements in quality and to control costs. Often when one payer has put pressure on reimbursement, providers have responded by extracting offsetting rate increases from other, less powerful payers, with the net result that costs continue to escalate. By the same token, health care providers are often frustrated in trying to deal with the myriad conflicting payment structures, quality measures, and incentive programs offered by individual public and private health plans. Providers cannot effectively respond to incentives unless they are clear and consistent and cover most of the practice. This is seldom the case when they are dealing with multiple payers. Because of this market fragmentation, pay-for-performance initiatives and incentives offered by any individual payer are unlikely to be of sufficient magnitude or scale to make it worth the provider s time and energy to fundamentally change the design or operations of his or her practice. From the providers perspective, making any substantive operational changes will be attractive only if the new behavior is something that they agree has value for their patients, if a critical mass of the payers are supporting it, and if the incentive payment is sufficiently large to get their attention. In a world with many competing payers all going it alone, these conditions are seldom met. Thus, while pay-for-performance initiatives sound sensible in theory, they tend to underachieve in practice, mainly because of lack of scale and leverage in the physician practice. For example, many health plans disseminate reports on the quality and efficiency of individual providers in the hopes of encouraging improvement. Providers may find these reports of little use because they often reflect performance on a small subset of their patients or they do not sufficiently risk-adjust performance. (In New York, payers that share reporting on physician performance with consumers generally follow minimum guidelines established by the Attorney General.) 3 In addition, when reforming the delivery system requires providers to invest in new technology or services (e.g., care managers or dietitians), providers may not be able to afford these investments unless enough payers support and reward this new form of care (Network for Regional Healthcare Improvement, Pittsburgh Regional Health Initiative, and the Robert Wood Johnson Foundation 2008). Payer Interests In our interviews, health plans advanced the view that individual payer efforts were often unsuccessful, as were many insurance company-based care management programs. Most were interested in collaborating on payment reforms that involved transformative changes in the structure and function of physician practices or larger care systems, such as primary care medical homes, accountable care organizations, and other risk-sharing models. In these cases, payers realize that a physician practice cannot make fundamental changes for one segment of its patients. These models entail the wholesale reorientation of the practice, which requires support from a critical mass of the practice s payers. Payers involved in the AMHD cited two areas where they were involved in collaborations with other health plans or 3 For more information on the Attorney General s guidelines, see http://www.ag.ny.gov/bureaus/health_care/hit2/doctor_ranking.html (accessed October 25, 2010). 4 United Hospital Fund

PCMH and ACO The Primary Care Medical Home (PCMH). The PCMH is a model for delivering primary care that facilitates partnerships between patients, physicians, and family caregivers. In the PCMH, each patient has an ongoing relationship with a personal physician who leads a team of appropriate providers to give or coordinate all of the care an individual needs. PCMHs have processes to ensure quality and patient safety, utilize health information technology, and provide ready access to services. The basic services and attributes of the PCMH show promise in addressing a number of key problems facing the nation s health care system: Reducing the costs of care, particularly for the chronically ill, by better prevention and management of those conditions, in line with established best practices; Reducing the avoidable use of hospital emergency departments for conditions that could be cared for in a physician office; Reducing avoidable hospital admissions and readmissions; and Reducing the costs of drugs, and the frequency and costs of adverse drug events. Five physician specialty organizations articulated the joint principles of the PCMH in February 2007; they can be found online at www.pcpcc.net. The National Committee on Quality Assurance has the best known process for accrediting PCMHs, though other organizations are starting accrediting processes, too. Accountable Care Organization (ACO). ACO is a new term that policymakers have used to describe a variety of care models; however, in the Patient Protection and Affordable Care Act, the definition narrows: an ACO is an organization of health care providers that agrees to be accountable for the cost, quality, and overall health care of a designated group of Medicare beneficiaries. ACOs that achieve specified cost savings targets will be allowed to share in those savings. ACOs could be physician group practices, networks of physician practices, physician/hospital organizations, hospitals and their employed physicians, or other forms as the Secretary of Health and Human Services may designate. * * Section 3022 of the Patient Protection and Affordable Care Act. were willing to consider a collaboration: the primary care medical home (PCMH) and accountable care organizations (ACOs). (For definitions of these terms, see the box above.) The reasons payers cited for their willingness to collaborate on these initiatives included the following: Impact. Both the PCMH and ACO models have the potential to transform the way the health care delivery system performs, improving quality and continuity of care for populations of patients (particularly for the chronically ill, who make up the highest-cost segment), improving provider and patient satisfaction, and reducing avoidable utilization and costs by the rigorous application of best practices in primary care settings and across the delivery system. A Multipayer Approach to Health Care Reform 5

Consistent incentives. Payers realize that providers cannot make fundamental changes for just one segment of their patients. The PCMH and ACO models entail the wholesale reorientation of the practice, which will require consistent incentives from a critical mass of any practice s payers, aligned with and supporting the desired behavior change. Better performance measurement. Aggregating data from multiple payers can allow more meaningful and accurate performance measurement of PCMHs and ACOs. This was an essential component of the demonstrations we examined in New York. However, creating new means of aggregating claims and clinical data for every new payment system entails new resources. Some interviewees held out hope that regional health information organizations and other health information exchange mechanisms could reduce these burdens in the future. Reducing the number of free riders. Free riders are payers that do not participate in a new payment system but whose patients may benefit from the changes implemented by providers. The PCMH model often involves investment up front by payers. Since the practice transformation will benefit all patients within a given practice, a single payer may not want to subsidize the new infrastructure unless other payers are participating as well. Thus, payers may look to collaborate to reduce or eliminate free riders. Impediments to Multipayer Action There are reasons why payers have not collaborated historically. As one of our interviewees phrased it, multipayer collaboration is not a natural act. It is our society s cultural and legal expectation that payers will compete, not cooperate. Payers compete with each other to sell insurance policies, and their success is based on their ability to argue convincingly that their product is superior to others available in their marketplace, based on such things as quality, network, service, and price. Corporate Structure and Philosophy Whether a given payer will be interested in a regional, multipayer initiative is likely to be influenced by its corporate structure and philosophy. Payers differ in their outlook on different models of service delivery reform. Payers that have embraced experimentation may be more likely to participate in a regional, multipayer initiative than more traditional payers. Also, differences may occur between national and regional health plans. Regional health plans can find it difficult to collaborate with national plans that do not have a strong local presence and, therefore, may be more comfortable collaborating with other regional plans. A local presence is essential to facilitate negotiations around delivery system models, payment models, performance standards for payers and providers, and other defining issues. Generally, competing firms seek to differentiate their products from their competitors rather than make them more similar. Payers consider their utilization management and disease management programs an essential part of the services they offer to employers. In this context, it is understandable that any payer would not seek out opportunities to collaborate with other payers. Legal Issues Beyond the business reasons impeding collaboration among payers, another and perhaps more compelling disincentive is 6 United Hospital Fund

that in most circumstances it is illegal. Many forms of cooperation are specifically prohibited by antitrust laws, especially any concerted action around payment. While federal antitrust regulators have attempted to define acceptable forms of collaboration among health care providers through the creation of safety zones and the issuance of advisory opinions, very little has been written about acceptable forms of cooperation among payers (U.S. Department of Justice and Federal Trade Commission 1996). This lack of attention by regulators probably reflects the limited interest among payers in the ability to cooperate. Market Position Payers will also vary in how much they have at stake in any multipayer collaborative because of differences in market share. Payers that have a substantial number of enrollees in a given region have a bigger stake in improving the performance of the local delivery system than those with a smaller share, and thus may be more likely to participate in a regional, multipayer initiative. In the AMHD, the largest payers were early and strong supporters. On the other hand, a number of the payers cited the small share of their total enrollment that resided in the Adirondacks as a factor in their participation in this demonstration. They were interested in the demonstration, and felt that it could provide them with important insights, but their risk, in the event the experiment was not successful, was not great. The Empire Plan (the state employees health plan) had additional reasons to support the demonstration. According to state officials, participation in the state employees health plan is unusually stable over long periods of time, with a large number of retirees continuing to participate in the plan. Unlike some other health plans where enrollment churns from year to year, the Empire Plan could expect to reap any longterm benefits that accrued from current investments. Differences in the Product Most health plans today are actually in two different lines of business: first, insurance products for employers and individuals, in which the plan accepts financial risk; and second, administrative services only (ASO) products provided to self-insured employers, who retain financial risk themselves and contract with a plan to administer its provider network, fee schedules, and other related services. When a health plan decides to participate in a payment reform collaborative, the question arises as to which lines of business will participate. Generally, the insurer can decide on its own whether to change the payment methodology for an insured product; but when health plans are providing administrative services, they are paying claims with someone else s money. In the AMHD, some health plans felt that they could not commit self-insured employer groups to a new payment system particularly PCMH models that involve upfront investments without getting their agreement to participate. This is not an inconsequential issue. One national payer told us that the five PCMH demonstrations in which they participate involve over 1,500 self-insured companies. While this payer said that over 60 percent of its employer groups ultimately agreed to participate in these demonstrations, contacting each of them involved a significant administrative burden. Similar issues may arise for experience-rated insurance products in which today s health claims are built into tomorrow s premiums. In the AMHD, health plans offering administrative services to self-insured companies and other entities reported that such arrangements represented roughly half of their covered lives in the region. A Multipayer Approach to Health Care Reform 7

Public Plan Participation To be successful, new payment systems may need participation by public health plans such as Medicare and Medicaid to reach a critical mass. This is true in the AMHD, where Medicaid accounts for 13 percent of patients in participating practices and Medicare accounts for another 18 percent. Securing public plan participation may require statutory changes or regulatory waivers. In the case of the AMHD, New York Medicaid had already made a policy decision to support the primary care medical home, but the demonstration still required legislation and budgetary support for the program to participate in the demonstration. Medicare remains the only substantial payer in the region not currently participating in the project. New York State has applied under the new Medicare Advanced Primary Care Practice demonstration to have Medicare participate in the Adirondack demonstration. If approved, this would bring Medicare into the demonstration as a full partner, paying the same rate as the other payers for its enrollees, for the same set of services designed and agreed upon for the project. However, as the state prepared the Medicare application, it became clear that Medicare s participation will bring requirements consistent with the multipayer demonstration as well as some new priorities and requirements to which the AMHD will need to respond. These requirements include a formal evaluation of Medicare s participation (as opposed to a demonstration-wide evaluation) and restrictions on participating in other Medicare demonstrations. Restructuring the Offering Over the past decade, payers have invested in the creation of disease management programs, disease registries, and care management programs. These programs have often become featured elements of payers product design and marketing to employers. (Employer groups also buy these services directly from other vendors.) Although the results of these payer-based services are mixed, they remain part of the payers administrative overhead and elements in their offering. When these payers participate in a multipayer medical home initiative under which such capacities are considered core competencies of a PCMH they no longer need to offer those services in the region where the demonstration is taking place, but the scale of these demonstrations is not sufficient to allow them to downsize those programs within their companies. The net result is that the payers must spread those costs over a somewhat smaller enrollee base. As these demonstrations grow and perhaps become the norm payers may be able to generate savings through a reduction in the scope of services they provide directly. Implementation Issues If the impediments can be overcome, a number of technical issues can arise that complicate the design and implementation of multipayer collaborations: Forum for Collaboration. To collaborate, providers and payers need a forum in which payers and providers can meet to define the new payment system, establish the obligations of providers and payers, and resolve disputes. This forum could consist of a temporary or permanent governance structure for the demonstration, which may entail costs and the commitment of additional resources. This decision-making process also needs to be able to deal with imbalances in size or market position, as payers with a large market share may believe they deserve a larger voice in the design of the multipayer collaborative. There is also the related question of who can and should convene the multiple stakeholders to discuss new models of payment and service delivery. In the Adirondack 8 United Hospital Fund

demonstration, the state became the convening entity to ensure compliance with antitrust laws and because it was seen as the closest thing to a disinterested third party. One of the features of collaboration is that it often requires compromise. Payers are independent corporations used to making their own decisions. Collaboration reduces their degrees of freedom. A health plan with an idea for an innovative payment system can pursue its own vision most easily by working alone; working with other plans, the plan might need to accommodate its partners differing visions. Collaborations can also be slow. Several observers of the Adirondack demonstration identified the time- and resource-consuming meetings as a significant drawback of the project. A payer hoping to implement or test a new idea quickly may prefer to move ahead by itself. Benefit design. Differences in the structure of benefit packages are a potential obstacle to payer collaboration. In the Adirondacks, for example, the payers agreed in principle to create a monthly care management fee to cover the costs of new services. Some insurance products offered by the participating plans, however, did not include care management as a covered benefit. These plans are working to revise contracts with employers as they come up for renewal. These plans will need to renegotiate their contracts to reflect the services and payments involved in the demonstration. Patient attribution. Whenever a new payment system is linked to patients rather than to specific services as in the PCMHs and ACOs the question of properly attributing members, and thus payments, to the correct provider will arise. While many benefit packages require patients to affirmatively choose a primary care provider (PCP), a significant number of open access plans do not require participants to do so. This inconsistency creates problems in accounting for members in a demonstration like this. In some cases, a member may have a primary care provider, but that provider is not participating in the demonstration. In other cases, plans may allow enrollees to use a specialist (e.g., a cardiologist who also delivers primary care services) as their primary care provider. In both such cases, that member would not have access to the augmented services of the PCMH, and his or her physician would not receive the PCMH payment. In most demonstrations, and in the AMHD, the attribution methodology is based on historical utilization patterns. The member is assigned to the physician whom the member has seen for one or more primary care visits over the past 18 months for primary care services. This approach is problematic for nonusers, people enrolled in a health plan and paying for health insurance who may not have generated a billable visit to any primary care physician in two or more years. Nonusers who are presumably healthier, and not using any services typically represent roughly 20 percent of the enrolled population in a given health plan. In most demonstrations, and in the AMHD, they are not being assigned to any provider, nor would they generate a PCMH payment for any primary care practice. While attribution may not be a major problem in some new forms of payment that are event-driven (e.g., bundled rates) it will certainly arise whenever there is a population-based payment or budget, as in medical homes and various forms of risk-sharing and capitation. Claims system. Payers rely on different claims payment systems that may or may not support the new payment methodology developed by a multipayer collaborative. In the Adirondacks, several payers had claims payment systems that were unable to generate a monthly care management fee that was not generated by a bill for a physician A Multipayer Approach to Health Care Reform 9

office visit or procedure. To work around this problem, these payers are increasing their FFS payments for certain evaluation and management codes in a way that should approximate the budgeted monthly payment. However, this will be administratively complex, requiring a periodic reconciliation of actual payments and possibly a redistribution of funds among providers in order to make the actual payments equate with the contractually required amounts. Opportunities for Multipayer Action In New York, multipayer action is not a new phenomenon. The New York Prospective Hospital Reimbursement Methodology (NYPHRM), which regulated hospital rates from 1983 to 1996, was a significant effort to control hospital costs and ensure universal access to hospital services by mandating consistent payment from all payers. Over the years, a variety of initiatives in Rochester, including the Hospital Experimental Payment Program of the 1980s and, more recently, the Community Technology Assessment Advisory Board (CTAAB), have involved collaboration among multiple payers (Hall and Griner 1993). 4 In recent years, state-funded regional health information organizations throughout the state have brought together multiple payers working with providers to further the dissemination and use of interoperable health information technology. There are also a growing number of efforts focused on improving the performance of the delivery system, including several medical home demonstrations underway in communities across the state, some of which involve multiple private health plans working together. On the whole, however, there are comparatively few multipayer initiatives. A recent national study of medical home initiatives across the United States found that most of those surveyed (69 percent, 17 out of 26 demonstrations) involved only one payer (Bitton, Martin, and Landon 2010). Among medical home demonstrations under way in New York, most are single-payer, but certain examples in the Adirondacks and Rochester involve multiple private health plans working together, as does a pay-for-performance initiative in the Hudson Valley and the activities of the P 2 Collaborative of Western New York. Moving Forward: Models of Multipayer Action Payers can work together in a variety of methods. Each type of collaboration has its own strengths and weaknesses and each raises a different set of legal and regulatory issues. We identified three models in our interviews: regulatory, follow-the-leader, and voluntary. A Regulatory Model All-payer rate regulation is a form of multipayer action that maximizes leverage with the delivery system by presenting providers with a uniform set of payment rules and incentives. In the 1970s and 1980s, many states, including New York, experimented with all-payer hospital rate regulation. Generally, states that regulated hospital charges had lower rates of hospital cost growth than their unregulated counterparts (Davis, Anderson, Rowland, and Steinberg 1990; Atkinson 2009). Most of these regulatory frameworks were dismantled in the 1990s as managed care grew and political trends favored private-sector efforts to control health care costs (McDonough 1997). 4 For more information about CTAAB, see www.ctaab.org (accessed October 25, 2010). 10 United Hospital Fund

All-payer rate regulation has the advantage of eliminating the free-rider problem and providing universally consistent incentives. In some cases, all-payer rate regulation can also address other social needs and market failures, such as providing a broad-based mechanism for financing uncompensated hospital care and by ensuring equity in reimbursements among payers and purchasers. On the other hand, regulation can be a heavyhanded and slow tool at a time when payment and delivery system models may need to adapt quickly. Also, in the United States, there are no models of all-payer rate setting for noninstitutional providers. Last year in Massachusetts, a state commission called for the elimination of fee-for-service payment in favor of global payments. One model for implementing this recommendation could be all-payer rate regulation, but the state has so far been unable to implement this policy into law and practice. The ACA directs the Center for Medicare and Medicaid Innovation to allow states to test and evaluate systems of all-payer payment reform. While there appears to be little appetite in most states for all-payer rate regulation, interest could revive as concerns about cost-shifting continue to grow. In New York, there has been some discussion of using limited rate regulation as a means to ensure adequate commercial payment levels for certain financially distressed hospitals and health centers. Follow-the-Leader Collaboration Another model of multipayer collaboration is one in which a single, large payer implements a new payment system and other payers follow suit. This model does not require any formal agreements and it raises no antitrust issues, yet it can achieve many of the advantages of other forms of collaboration. The best example of this model is Medicare s implementation of its DRG-based inpatient prospective payment system (IPPS) in 1983. Before the introduction of the IPPS, Medicare reimbursed hospitals based on their reported costs, which provided no incentive for hospitals to operate efficiently. In 1983, Medicare instituted a new system of prospectively set prices based on diagnosisrelated groups (DRGs). In the years following the IPPS implementation, Medicare costs were reduced, hospital length of stay decreased, and hospital margins improved all without diminishing the quality of hospital care (Mayes and Berenson 2006). Because of Medicare s success with inpatient prospective payment, many commercial payers soon adopted similar DRG-based payment systems. The result is that hospitals now face a fairly consistent set of clear incentives for the efficient use of resources. Medicare was a sufficiently dominant payer for hospital services that it could test this new payment methodology without explicit collaboration with other payers. While Medicare is also a large payer of hospital outpatient services, its new outpatient prospective payment system, implemented in 2000, has not been imitated by other payers. Many hospital analysts believe it does not reward efficiency because the outpatient payment system does not make a single payment for all the services involved in each patient visit the way the inpatient system does. Thus, a large payer may have sufficient leverage to test a new payment system, but others will adopt it only if the new payment system is successful in achieving cost containment and other goals (Averill, Goldfield, Vertrees, McCullough, Fuller, and Eisenhandler 2010). No payer appears to be big enough to advance a flawed payment system. The ACO model may revive the practice of commercial payers and Medicaid following Medicare s lead in payment reforms. In our interviews, several payers said they were eagerly awaiting Medicare s new shared savings program, which will require the A Multipayer Approach to Health Care Reform 11

program to define ACOs. In particular, payers were pleased the ACA requires ACOs to create a legal entity to represent the participating providers; payers believe this would help them negotiate directly with ACOs, eliminating the need for third-party conveners, which have been essential to some payment reform demonstrations. Voluntary Collaboration Lying somewhere between the follow-theleader model and the regulatory approach is an area of growing interest in which payers actively collaborate on a voluntary basis. For example, in New York we have seen voluntary multipayer collaboratives in the Adirondack medical home demonstration, the Taconic Health Information and Network Community (THINC) pay-for-performance demonstration, and the Rochester medical home demonstration. These collaborative models can arise from a variety of circumstances. In the Adirondacks, the demonstration was clearly a provider-initiated response to a crisis in the ability of several rural communities to recruit and retain primary care physicians. The providers identified the PCMH as a means for improving the quality of life for physicians and for increasing their income. The providers were adept at garnering public and political support for the demonstration, which provided much of the impetus for multiple payers to participate. The Importance of State Leadership and Participation The Adirondack demonstration also relies heavily on state leadership and state participation. (See the box on page 13 for an explanation of all of the roles the state played in the Adirondack Demonstration.) Collaboratives of this sort often need a third party that can serve as the impartial convener of payers and providers. Many interviewees believed that the state was the only party that could serve this role in the Adirondacks. (The state was not a perfect option because it was, in fact, not an impartial body; it was also participating in the demonstration as a large payer.) In other areas, such as the pay-for-performance demonstration in the Hudson Valley, a nonprofit organization such as THINC plays this role. In the Adirondacks, the state participation also provided antitrust immunity that allowed payers and providers to share data on costs and to openly discuss levels of financial support for the PCMH services. These actions are allowed under something called state action immunity. Essentially, if a state articulates a policy goal that requires taking actions that would otherwise violate antitrust laws, these actions will be allowable if the state actively supervises the implementation of the policy. In the Adirondacks, the articulation of state policy was accomplished by the passage of legislation specifying the terms of the demonstration, and the supervision requirement was satisfied through the state s active (and ongoing) role as the convener of the meetings between the payers and providers. A recent national survey of PCMH demonstrations identified only four that involved antitrust protection through state action immunity (Bitton, Martin, and Landon 2010). This antitrust protection allowed open discussions about the financing of the PCMH, but it also required a much greater level of state involvement than most multipayer collaboratives can expect (or may even want). Collaboratives can accomplish many of the same goals without antitrust immunity. Payers and providers can convene to define the practice transformation they are seeking and set performance benchmarks. Discussions around payment and paymentrelated terms for these changes, however, must be conducted separately. The discussions can happen one-on-one between 12 United Hospital Fund

THE ROLE OF THE STATE As discussed in this issue brief, the state has played a crucial role in the Adirondack demonstration. Several observers noted that the state wore many hats throughout the process of developing and implementing this demonstration: Policymaker. The state endorsed the PCMH as a preferred model by creating new Medicaid payments for accredited PCMHs and by enacting legislation that specifically authorized the Adirondack demonstration. Providers in the Adirondacks believed the state supported their demonstration because it had already endorsed the PCMH model. Payer. Between the Medicaid program and the Empire Plan, the state is probably the largest payer participating in the Adirondack demonstration. The participation of these two programs was essential to achieving the critical mass necessary to garner provider buy-in. Regulator. The state was a regulator in that only it could confer antitrust protection for the discussions around cost and reimbursement for the medical home services. In addition, the State Insurance Department and the Department of Health regulate the operations of the health plans participating in the project. Convener. The state was required to serve as the convener in order to maintain antitrust immunity. However, several participants noted that even in the absence of antitrust immunity, a convener would still have been needed and only the state could have played that role in the Adirondacks. Grantmaker. State grants for health information technology and exchange and rural health planning were essential foundations to the Adirondack demonstration. State grants were the catalyst in the Hudson Valley pay-for-performance project as well. each payer and each provider, or they can use the messenger model in which a disinterested third party, such as THINC, serves as an intermediary between payers and providers to facilitate negotiations. In the messenger model, both providers and payers can share cost and payment information with the messenger, which must aggregate the data and maintain the confidentiality of provider and payer-specific data. Depending on the number of providers and payers involved in a given collaboration, these negotiations can be time-consuming. The messenger model relieves this burden somewhat, largely by placing it on the messenger. New York State officials noted that serving as the convener of the AMHD meetings requires significant ongoing resource support. Although the state has likely learned lessons from the AMHD process that could make future demonstrations more efficient, it is unlikely that the state could play a similar role in multiple demonstrations across the state at the same time without additional or redirected resources. More thought needs to be given to whether the state needs to play such an active role or whether there are other means of achieving the same ends. A Multipayer Approach to Health Care Reform 13

Conclusion Because of the limited power of any one payer, even one as large as Medicare, to force meaningful changes in the shape and performance of the health care delivery system, multipayer collaboration will be essential to the success of health care reform. The Adirondacks experiment, which is focused on transforming primary care in a five-county region of New York State, has demonstrated that there is an appetite for ambitious initiatives on the part of both providers and payers under certain conditions. One of the important lessons of the AMHD is that while multipayer collaboratives are necessary to effective health system redesign and improvement, they are quite difficult to craft. There are several compelling reasons why payers might want to collaborate, but there are many difficult issues that must be addressed and overcome for multipayer efforts to succeed. Participation in a multipayer collaborative like the Adirondacks PCMH demonstration forces payers to deal with a basic conundrum to effectively influence provider behavior, payers must work together to align their incentives toward paying consistently for a specific set of behaviors and outcomes. But when payers cooperate with their competitors, they begin to lose their product differentiation, the ability to make the case that their offering is unique. Despite the numerous obstacles identified in this paper, the AMHD also demonstrates that with enough leadership and a continued focus on common goals, multipayer collaboratives can be developed and have a chance to succeed. The challenge for policymakers will be determining the right mixture of policy, leadership, and regulation to allow more of these projects to bloom. 14 United Hospital Fund

References Anderson GE, UE Reinhardt, PS Hussey, and V Petrosyan. 2003. It s the Prices, Stupid: Why the United States Is So Different From Other Countries. Health Affairs 22(3): 89-105. Atkinson G. 2009. State Hospital Rate-Setting Revisited. New York: Commonwealth Fund. Averill RF, NI Goldfield, JC Vertrees, EC McCullough, RL Fuller, and J Eisenhandler. 2010. Achieving Cost Control, Care Coordination, and Quality Improvement Through Incremental Payment System Reform. Journal of Ambulatory Care Management 33(1):2-23. Bitton A, C Martin, and BE Landon. 2010. A Nationwide Survey of Patient-Centered Medical Home Demonstration Projects. Journal of General Internal Medicine 25(6): 584-592. Davis K, GF Anderson, D Rowland, and E Steinberg. 1990. Health Care Cost Containment. Baltimore: Johns Hopkins University Press. Guterman S and H Drake. 2010. Developing Innovative Payment Approaches: Finding the Path to High Performance. New York: Commonwealth Fund. Hall WJ and PF Griner. 1993. Cost-Effective Health Care: The Rochester Experience. Health Affairs 12(1): 58-69. Mayes R and R Berenson. 2006. Medicare Prospective Payment and the Shaping of U.S. Health Care. Baltimore: Johns Hopkins University Press. McDonough JE. 1997. Tracking the Demise of State Hospital Rate Setting. Health Affairs 16(1): 142-149. Network for Regional Healthcare Improvement, Pittsburgh Regional Health Initiative, and the Robert Wood Johnson Foundation. 2008. From Volume to Value: Transforming Health Care Payment and Delivery Systems to Improve Quality and Reduce Costs. Recommendations of the 2008 NRHI Healthcare Payment Reform Summit, Pittsburgh, Pennsylvania. Princeton, NJ: Robert Wood Johnson Foundation. Available at http://www.chqpr.org/ (accessed October 25, 2010). Orszag P and EJ Emanuel. 2010. Health Care Reform and Cost Control. New England Journal of Medicine 363(7):601-603. U.S. Department of Justice and Federal Trade Commission. 1996. Statements of Antitrust Enforcement Policy in Health Care. Available at http://www.justice.gov/atr/public/guidelines/0000.htm (accessed October 25, 2010). Vladeck BC and T Rice. 2009. Market Failure and the Failure of Discourse: Facing Up to the Power of Sellers. Health Affairs 28(5): 1305-1315. A Multipayer Approach to Health Care Reform 15

Additional copies of A Multipayer Approach to Health Care Reform may be downloaded, at no charge, from the United Hospital Fund website, www.uhfnyc.org.

Shaping New York s Health Care: Information, Philanthropy, Policy. Empire State Building 350 Fifth Avenue, 23rd Floor New York, NY 10118 (212) 494-0700 http://www.uhfnyc.org ISBN 1-933881-05-4