ENHANCING VALUE IN MEDICARE: DEMONSTRATIONS AND OTHER INITIATIVES TO IMPROVE THE PROGRAM Stuart Guterman and Michelle P. Serber The Commonwealth Fund January 2007 Prepared for The Commonwealth Fund/Alliance for Health Reform 2007 Bipartisan Congressional Health Policy Conference ABSTRACT: Medicare was designed to deal primarily with the effects of acute illness, which was seen at the time of the program’s inception as the major threat to the health and financial security of the aged. While it has fulfilled this purpose reasonably well, demographic and other changes pose new challenges to Medicare and the health care system as a whole. Moreover, that system must deal with sub-par performance, both on cost and quality. This report, which examines Medicare’s efforts to play a more proactive role in the purchase of appropriate, high- quality, and efficient health care for its beneficiaries, provides an overview of Medicare demonstrations, pilots, and other initiatives in two categories: chronic care and provider performance. The process of identifying, testing, evaluating, and implementing Medicare policy improvements is also discussed, and recommendations for improving that process are offered. Support for this research was provided by The Commonwealth Fund. The views presented here are those of the authors and not necessarily those of The Commonwealth Fund or its directors, officers, or staff, or of The Commonwealth Fund Commission on a High Performance Health System or its members. This and other Fund publications are available online at www.cmwf.org. To learn more about new publications when they become available, visit the Fund’s Web site and register to receive e-mail alerts. Commonwealth Fund pub. no. 990. CONTENTS List of Figures and Tables................................................................................................ iv About the Authors........................................................................................................... v Executive Summary........................................................................................................ vi Introduction .................................................................................................................... 1 The Need for Enhanced Value......................................................................................... 2 Initiatives to Improve Effectiveness and Efficiency ........................................................... 6 Chronic Care Initiatives ............................................................................................. 6 Initiatives to Improve Provider Performance ............................................................ 13 Moving from Demonstrations and Pilots to Program Improvements .............................. 22 Conclusions ................................................................................................................... 25 Notes............................................................................................................................. 27 iii LIST OF FIGURES AND TABLES Figure 1 Two-Thirds of Medicare Spending Is for People with Five or More Chronic Conditions .......................................................... 2 Figure 2 Growth in Medicare Expenditures, 1970–2015............................................... 4 Figure 3 Medicare Spending Per Enrollee and Mortality Rate by State, 2003................ 5 Figure 4 Medicare Health Support Organizations and Locations ................................. 10 Figure 5 Medicare Premier Hospital Demonstration: Higher Quality Hospitals Have Fewer Readmissions ............................................................. 18 Figure 6 Medicare Physician Group Practice Demonstration....................................... 19 Table 1 Chronic Care Initiatives ................................................................................. 8 Table 2 Aligning Payment and Quality: Initiatives Focused on the Reporting of Performance Measures............................................................. 14 Table 3 Aligning Payment and Quality: Financial Incentives to Improve Provider Performance ................................................................ 16 iv ABOUT THE AUTHORS Stuart Guterman has been senior program director for The Commonwealth Fund’s Program on Medicare’s Future since May 2005. He is responsible for the Fund’s research agenda on Medicare issues, the development, management, and review of grants to be funded under the program, analyses related to the current performance and future improvements in the Medicare program, and disseminating the results of the Fund’s activities and other research related to these topics. Prior to joining the Fund, Mr. Guterman was director of the Office of Research, Development, and Information at the Centers for Medicare and Medicaid Services from 2002 to 2005. Before that, he was a senior analyst at the Congressional Budget Office, a principal research associate in the Health Policy Center at the Urban Institute, and deputy director of the Medicare Payment Advisory Commission (and its predecessor, the Prospective Payment Assessment Commission) from 1988 through 1999. Previously, Mr. Guterman was chief of institutional studies in the Health Care Financing Administration’s Office of Research, where he directed the evaluation of the Medicare Prospective Payment System for inpatient hospital services and other intramural and extramural research on hospital payment. Michelle P. Serber is program assistant for the Fund’s Program on Medicare’s Future, based at AcademyHealth in Washington, D.C. Ms. Serber graduated magna cum laude in 2002 from Duke University with a degree in psychology. She interned at Burson- Marsteller’s Health Care Practice, conducting research on prospective new business opportunities. In February 2003, she joined the Advisory Board Company as a research associate, working primarily with hospital executives in the cardiovascular service line. Ms. Serber was responsible for preparing briefs for health care executives on a wide range of topics and conducted literature searches and interviews with industry experts, leading clinicians, and hospital executives. v EXECUTIVE SUMMARY Medicare was designed to deal primarily with the effects of acute illness, which was seen at the time of the program’s inception as the major threat to the health and financial security of the aged. While it has fulfilled this purpose, demographic and other changes pose new challenges to Medicare and the health care system as a whole. As with many other countries, the population of the United States is aging, and the prevalence of chronic conditions is increasing. Yet our nation’s health care delivery and financing system is not set up to care for a population with complex, long-term medical needs. The performance of the U.S. health system, according to many cost and quality indicators, is subpar. Moreover, the Medicare program is faced with insolvency by the end of the next decade. To meet these challenges, Medicare must play a more proactive role in the purchase of appropriate, high-quality, and efficient health care for the elderly and disabled. In response to these imperatives, the Centers for Medicare and Medicaid Services (CMS) is developing an array of initiatives to address the evolving needs of the Medicare program and its beneficiaries. Many of these initiatives have been developed under CMS’s demonstration authority, which allows the agency to waive certain Medicare payment rules that determine what services are covered and how they are paid in order to test potential improvements; others have been specifically mandated by Congress. Initiatives to Improve Medicare’s Effectiveness and Efficiency Medicare has a number of initiatives, both under way and in development, aimed at improving the quality and coordination of services provided to its beneficiaries. These fall primarily into two categories: improving the availability and coordination of care for beneficiaries with chronic conditions; and improving the alignment between payment (as well as other incentives) and the quality and effectiveness of care. The majority of Medicare’s chronic care initiatives are focused on better ways to coordinate care for beneficiaries in the traditional Medicare fee-for-service program, but several such initiatives address the structural impediments that Medicare managed care plans have faced when it comes to serving beneficiaries with chronic conditions. CMS has launched several initiatives in recent years to encourage improved quality of care, placing emphasis first on public reporting of quality indicators in a variety of health care settings. The agency also is providing technical assistance to a wide range of vi providers through its Quality Improvement Organizations (QIOs). In addition, CMS has developed demonstration projects to test ways of using financial incentives to encourage better performance by hospitals, physicians, and other providers. Moving from Demonstrations and Pilots to Program Improvements With pressure mounting to find ways to improve quality while also controlling the growth of Medicare spending, it is important to know what these initiatives have to tell us about whether policy should be changed and, if so, how. But resources currently available for that purpose are scarce. The availability of more funding—to help identify potential improvements and assess their likelihood of success, to develop appropriate design and implementation strategies, and to evaluate results in a timely but rigorous manner—would enhance our ability not only to identify and develop more (and more appropriate) initiatives, but also to translate those initiatives into better policy. In identifying, developing, testing, evaluating, and implementing Medicare improvements, policymakers face a number of hurdles. Suggestions for improving the process include the following: • Increase clarity and flexibility in the waiver approval process. The purpose of Medicare demonstrations should be to enhance the program’s value, and the process for identifying, developing, approving, and implementing them should be more transparent and explicitly based on criteria consistent with that purpose. One important requirement for approval of Medicare demonstrations is budget neutrality, which rightly was intended to protect the fiscal integrity of the program. However, budget neutrality as currently applied has become a strict and narrowly defined rule that can hinder the development of important new initiatives. The application of this requirement should take into account the fact that some policy changes may involve short-term costs but long-term benefits. Cross-program savings and costs must be considered as well, so that policy changes that reduce overall spending can be tested, even if they require spending increases in some individual programs; the same reasoning should be applied across levels of government. In addition, there should be allowances for considering policy changes that increase value, even if they do not reduce spending per se. • Expand the array of methods available for evaluating demonstration results. Medicare demonstrations are not conducted in laboratories but in a world in which the policy environment is constantly changing. Consequently, evaluations must deal with imperfect controls and incomplete data with which to account for mitigating factors. Moreover, in many cases some of the major objectives of the policy vii change being tested are themselves difficult to measure, either because they are qualitative in nature or because no baseline data exist to determine whether the policy in question has had the hoped-for effect. Putting mechanisms in place that allow for continuous monitoring of demonstrations would help indicate directions not only for the development of new policies when the trials are completed but also for changes in the trials themselves as they proceed. The importance and time-sensitivity of information on potential policy changes calls attention to the shortcomings of the methodology currently available for evaluating demonstration results. The imperfect controls and incomplete data available in the real world in which policy is implemented and evaluated must be balanced with the need for rigorous testing of potential policy improvements. More resources are needed to develop new ways of providing timely results that meet the needs of policymakers while maintaining research standards that ensure scientific validity. • Establish a more explicit and transparent mechanism for moving from pilot to policy. Finally, once the demonstrations are done and evaluated, the process would benefit from a more explicit mechanism for translating what we learn from them into new policy. In some cases—the Medicare Health Support pilot is the most recent example—Congress gives Medicare (through the Secretary of Health and Human Services) the authority to continue or expand a trial, but most often additional action is required to effect the policy changes suggested by demonstration results. Even when independent evaluations of a mandated demonstration’s results are required by law, it can take years for the evaluation report to be cleared by the Executive Branch and transmitted to the Congress. Again, making the process transparent would help considerably, allowing more open discussion of policy changes of interest and their potential impacts. The designation of a regular vehicle for reporting the findings from demonstrations and other initiatives would provide a visible source of information and a platform for the open discussion of policy implications. Overall, a strategy for dissemination is needed, so that potential policy improvements can be implemented and their benefits realized in a timely and effective manner. New initiatives to improve the program should build on the experiences and lessons learned from demonstrations, with the ability to reshape interventions as they are implemented to maximize their effectiveness. These changes will help Medicare improve the quality and effectiveness of care while controlling the cost growth that threatens the program’s fiscal viability. viii ENHANCING VALUE IN MEDICARE: DEMONSTRATIONS AND OTHER INITIATIVES TO IMPROVE THE PROGRAM INTRODUCTION The Medicare Program, created in 1965, was designed to ensure access to needed health care for the elderly—half of whom lacked insurance to protect them against the potentially catastrophic costs of major illness.1 It has served that purpose well for more than 40 years. Over that time, Medicare has become one of the most popular government programs, generating consistently high satisfaction levels among its now 43 million elderly and disabled beneficiaries. Medicare was designed to deal primarily with the effects of acute illness, which was seen at the time of its implementation as the major threat to the health and financial security of the aged. While the health care delivery and financing system in the United States remains largely oriented toward acute care, demographic and other trends are putting pressure on that system—and on Medicare particularly—to change. Health care spending overall is growing more rapidly than our economy can sustain, and Medicare faces the additional pressure of a wave of post-World War II baby boomers set to begin retiring within the next few years.2 At the same time, for all we spend on health care, there are significant issues with the safety, quality, and efficiency of care, and that care is poorly coordinated across providers.3 This problem is especially important for Medicare, whose aged and disabled beneficiaries need and use more health care and are more likely to have chronic conditions than the rest of the population.4 Consequently, Medicare must play a more proactive role in making sure that appropriate, high-quality, and efficient health care is available for the elderly and disabled. In response to these imperatives, the Centers for Medicare and Medicaid Services (CMS) is developing an array of initiatives to address the evolving needs of the Medicare program and its beneficiaries. Many of these initiatives have been developed under CMS’s demonstration authority, which allows the agency to waive certain Medicare payment rules that determine what services are covered and how they are paid in order to test potential improvements; others have been specifically mandated by Congress. 1 This paper will describe two sets of initiatives: those intended to improve care for chronic conditions; and those aimed at providing incentives for better provider performance. In addition, the paper will discuss the process for identifying, developing, implementing, and evaluating these kinds of initiatives and using them to enhance value in the Medicare program, and it will also offer some suggestions for improving that process. THE NEED FOR ENHANCED VALUE Like many other countries, the United States population is aging. In 2000, the proportion of individuals age 65 and older in the U.S. was 12.5 percent; this share is projected to grow to 16.6 percent by 2020, an increase of one-third.5 Older individuals are more likely to have one or more chronic conditions. A 2004 Commonwealth Fund survey of older adults asked respondents if a physician had told them that they had any of six conditions— hypertension or high blood pressure, heart disease or heart attack, cancer, diabetes, arthritis, or high cholesterol—and the rate of reported conditions increased significantly with age: 67 percent of respondents aged 50 to 64 cited at least one chronic condition, versus 84 percent of those aged 65 to 70.6 Other studies have shown that the 20 percent of Medicare beneficiaries with five or more chronic conditions account for 66 percent of Medicare spending (Figure 1)—and they receive services from an average of almost 14 physicians in a given year.7 Figure 1. Two-Thirds of Medicare Spending Is for People with Five or More Chronic Conditions No chronic conditions 1–2 chronic 1% conditions 10% 3 chronic conditions 10% 5 or more chronic conditions 4 chronic 66% conditions 13% THE COMMONWEALTH FUND Source: G. Anderson and J. Horvath, Chronic Conditions: Making the Case for Ongoing Care (Baltimore, Md.: Partnership for Solutions, Dec. 2002) 2 The health care delivery and financing system, however, is not set up to serve individuals with multiple chronic conditions. Studies have shown that Medicare beneficiaries with these conditions are more likely to have preventable hospitalizations, experience adverse drug interactions, undergo duplicate tests, and receive contradictory information from doctors.8 Moreover, the high Medicare costs they incur appear to be consistent over time: a 2005 Congressional Budget Office report found that nearly half of the beneficiaries in the top 25 percent of the Medicare population with respect to cost in 1997 (a group that accounted for approximately 85 percent of total Medicare spending) were again in the top 25 percent the following year.9 That report also determined that of the high-cost beneficiaries in 2001, more than 75 percent had been diagnosed with one or more of seven major chronic conditions. Neither traditional fee-for-service Medicare nor Medicare Advantage (MA) is currently configured to provide adequate care for these beneficiaries. The fee-for-service payment model still dominates in the United States—particularly in Medicare. Although the proportion of Medicare beneficiaries enrolled in managed care arrangements has grown recently, about 85 percent of them remain in the traditional fee-for-service program, which provides no incentive for the coordinated care needed by the chronically ill.10 Additionally, fee-for-service payment encourages specific, condition-oriented care, by which an individual with multiple conditions is treated by multiple providers.11 Moreover, the fee-for-service model allots more generous payments for procedures and specialists’ services, thereby discouraging physicians from entering the primary care fields that are more compatible with the role of care coordination. Although managed care would appear to be better suited to providing the kind of coordinated care needed by chronically ill Medicare beneficiaries, the MA program and its predecessors historically also have been flawed in this respect. Capitation can provide a strong incentive to avoid chronically ill enrollees if the payment system fails to adjust properly for the costliness of the individual enrollee and, although MA plan payment rates will be fully risk-adjusted in 2007, recent analyses indicate that the incentive to avoid sicker enrollees may persist.12 Meanwhile, Medicare is likely to face increased fiscal pressure over the next few years: as baby boomers approach retirement, the country’s ratio of workers to beneficiaries is declining. As a result of the aging population and the new drug benefit, the Medicare Trustees estimate that program expenditures will grow from $336 billion in 2005 to $817 billion in 2015 (Figure 2).13 Medicare spending as a share of gross domestic product (GDP)—at 2.7 percent in 2005—is expected to rise to 4.7 percent by 2020. In addition, the Medicare 3 Hospital Insurance Trust Fund is projected to be insolvent by 2018. These projections will soon be pushed to the forefront of the political debate: the next (2007) Medicare Trustees’ Report is expected to trigger a “Medicare funding warning,” which by law requires that the president submit a proposal to Congress to address Medicare spending growth.14 Figure 2. Growth in Medicare Expenditures, 1970–2015 Dollars in billions 900 817.2 800 700 572.9 600 500 400 336.4 300 221.8 184.2 200 111 72.3 100 16.3 36.8 7.5 0 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 THE COMMONWEALTH Note: Figures for 2010 and 2015 are projected. FUND Source: 2006 Medicare Trustees’ Report. In addition to an aging population, the increased prevalence of chronic conditions, and rapid spending growth, the Medicare program and the health care system as a whole must also deal with sub-par performance on many cost and quality indicators. The National Scorecard on U.S. Health System Performance compiled by the Commonwealth Fund’s Commission on a High Performance Health System indicates that there is much room for improvement.15 The 16 percent of the United States’ GDP attributable to health spending is double the proportion of most industrialized countries; after a pause in the late 1990s, this percentage has been growing more rapidly in recent years.16 Yet these greater expenditures do not appear to translate into better care, with the United States lagging behind other countries on indicators such as mortality and healthy life expectancy. Moreover, both the quality of care and the efficiency with which it is provided are highly variable.17 Multiple quality indicators demonstrate large variation between top and bottom groups of hospitals, states, and health plans. For example, although top- performing hospitals reach 100 percent adherence to basic clinical guidelines for treating patients with heart attacks, congestive heart failure, and pneumonia, the national average 4 is only 84 percent.18 Variations also exist in mortality rates: an analysis of Medicare beneficiaries’ mortality rates over the years 2000–2002 indicates a spread of 33 percentage points between the risk-adjusted mortality ratios in the 10 percent of hospitals with the lowest rates and the 10 percent of hospitals with the highest rates.19 This highly variable quality of care is delivered by a system that is too often poorly coordinated, which puts patients at risk and raises costs. Care coordination is necessary at the time of hospital discharge and during transitions following discharge. Yet, according to a 2005 Commonwealth Fund survey, only 67 percent of hospitalized patients in the United States reported having their medications reviewed at time of discharge, compared to as much as 86 percent in Germany.20 Additionally, a lack of discharge planning occurs all too frequently. On average, U.S. patients with congestive heart failure receive hospital discharge instructions only 50 percent of the time.21 There is also wide variation across the country in spending per Medicare beneficiary and the quality of care (Figure 3). For beneficiaries with acute myocardial infarction, hip fracture, and colorectal cancer with resection, there are substantial differences in one-year risk-adjusted mortality rates following the initial hospital admission and in resource use over that period. Medicare annual payments for patients with three chronic conditions—diabetes, chronic obstructive pulmonary disease, and congestive heart failure—also vary greatly across regions.22 Figure 3. Medicare Spending per Enrollee and Mortality Rate by State, 2003 Medicare Spending per Enrollee $4,500 $5,500 $6,500 $7,500 $8,500 Mortality Rate of Medicare Enrollees 2.5% HI 3.5% MN 4.5% OR UT CO AK ID NH AZ DC WY NM WI VT DE NV FLMD CA IA SD MTWA VA IN SCKS US NY ND NE ME MI TX CT MA NJ 5.5% NC WV KY OH IL AR GA MS MO AL PA TN OK LA RI 6.5% THE COMMONWEALTH FUND Source: Data from The Dartmouth Atlas of Health Care, www.dartmouthatlas.org. 5 Another contributor to inefficiency is the lack of timely access to physicians for after-hours care and advice, which can lead to unnecessary use of the emergency room (ER). A cross-national survey of six nations found that 26 percent of U.S. adults use the ER for conditions that could have been treated by a primary care doctor, compared with 6 percent to 9 percent in the lowest-rate countries.23 Medicare’s role in addressing these issues is unique: comprising one-fifth of all personal health care spending, it is both highly vulnerable to the forces affecting the broader health system and potentially an important driver of change.24 The fact that Medicare is financed by a near-universal payroll tax and also by general tax revenues, together with the fact that almost everyone who turns 65 will become a Medicare beneficiary, make it particularly visible, important, and accountable to the American people. It is readily apparent that changes are needed, and Medicare can and must serve as a springboard for policies that improve health care, not only for its beneficiaries but also for the entire population. CMS has already begun to respond by developing a variety of initiatives aimed at improving the quality and coordination of services provided to Medicare beneficiaries. These initiatives are discussed below. INITIATIVES TO IMPROVE EFFECTIVENESS AND EFFICIENCY25 This paper considers demonstrations, pilots, and other initiatives that fall primarily into two categories: improving the availability and coordination of care for beneficiaries with chronic conditions; and improving the alignment between payment (as well as other incentives) and the quality and effectiveness of care. Chronic Care Initiatives One dimension on which CMS has focused is improving its ability to provide the chronic care beneficiaries increasingly need.26 Beneficiaries with chronic conditions make up an increasing proportion of the Medicare population. Moreover, as previously discussed, patients with chronic conditions typically receive fragmented health care from multiple providers and multiple sites of care; this problem is amplified for beneficiaries with multiple chronic conditions. Not only is such disjointed care confusing and ultimately ineffective, it can present difficulties for patients, including an increased risk of medical errors. Additionally, the repeated hospitalizations that frequently accompany such care are extremely costly to both patients and Medicare. As the nation’s population ages, the number of chronically ill Medicare beneficiaries is expected to grow dramatically, with serious implications for access, quality, and Medicare spending. 6 In the private sector, managed care entities such as health maintenance organizations, as well as private insurers, disease management organizations, and academic medical centers, have developed a wide array of programs that combine adherence to evidence-based medical practices with better coordination of care across providers. These initiatives are based on the belief that disease management programs can improve medical treatment plans, reduce avoidable hospital admissions, and promote other desirable outcomes without increasing costs.27 Two features make the case for disease management even stronger in the Medicare context than in the private sector. First, the greater prevalence of chronic illnesses among the Medicare population provides more opportunity for improving the appropriateness, effectiveness, and efficiency of care. Second, unlike private insurers, the Medicare program keeps its enrollees for life. This means that efforts to improve the coordination of care for chronic conditions can be consistently and continuously applied over a long period; it also means that the benefits of such efforts will accrue to the program rather than to some other payer. The demonstration projects conducted by CMS in this area are intended to test the value of alternative approaches to improving care for beneficiaries with chronic conditions, while also making Medicare a more aggressive and effective purchaser of this care.28 The majority of Medicare’s chronic care initiatives have focused on the coordination of care for chronically ill beneficiaries in the traditional Medicare fee-for- service program, but several of them have addressed the structural impediments that managed care plans have faced in attempting to provide appropriate care to this population. These initiatives are summarized in Table 1, and in the following discussion. 7 Table 1. Chronic Care Initiatives Initiative Description Medicare Case Management The first of the Medicare chronic care initiatives, designed to test case man- Demonstration agement for beneficiaries with catastrophic illnesses and high medical costs. Medicare Coordinated Care To examine whether coordinated care programs can improve medical Demonstration treatment plans, decrease avoidable hospital admissions, and further benefit chronically ill beneficiaries without increasing program costs. Medicare Disease Management To evaluate the effect of disease management services, coupled with a Demonstration prescription drug benefit, on the health outcomes of Medicare beneficiaries diagnosed with advanced-stage congestive heart failure, diabetes, or coronary disease. Medicare Health Support Pilot program to test population-based chronic care programs that provide self-care support, education, and coordination of care to beneficiaries. Care Management for High-Cost To study a variety of provider-centered care management models— Beneficiaries Demonstration including intensive-care management, increased provider availability, structured chronic care programs, restructured physician practices, and greater flexibility in care settings—for high-cost beneficiaries. Special Needs Plans (SNPs) Authorized by the Medicare Modernization Act to focus on individuals with special needs, including beneficiaries who are institutionalized, dually eligible for Medicare and Medicaid, or suffering from severe or disabling chronic conditions. End-Stage Renal Disease (ESRD) To test the feasibility of year-round open enrollment in managed care for Managed Care Demonstration beneficiaries with ESRD. Each site provides service integration, case management, and extra benefits, and is paid a higher rate to reflect the additional costliness of enrollees with ESRD. ESRD Disease Management To test the effectiveness of disease management models for increasing Demonstration quality of care for ESRD patients while ensuring that this care is provided more effectively and efficiently. Source: CMS Web site, http://www.cms.hhs.gov. Improving Chronic Care in Fee-for-Service Medicare The first of the Medicare chronic care initiatives was the Medicare Case Management Demonstration, which studied the appropriateness of providing case management services to beneficiaries with catastrophic illnesses and high medical costs. This demonstration was implemented at three sites beginning in October 1993 and continued through November 1995.29 The target conditions and case management protocols differed across the sites, but all three generally focused on increased education regarding proper patient monitoring and management of the target condition. 30 The project evaluation found that, while the projects successfully identified and enrolled populations of Medicare beneficiaries likely to have much higher than average Medicare costs, there was an unexpectedly low level of enthusiasm for the project from beneficiaries. This was attributed to the lack of physician involvement or sufficiently focused interventions, and to the lack of a financial incentive to reduce Medicare spending. 8 The Medicare Coordinated Care Demonstration was mandated by Congress in the Balanced Budget Act of 1997. This project was designed to test whether providing coordinated care services to Medicare fee-for-service beneficiaries with complex chronic conditions could yield better patient outcomes without increasing program costs. The demonstration (and a similar parallel project) originally involved a total of 15 sites, both in urban and rural areas, that focused on complex chronic conditions, including: congestive heart failure; heart, liver, and lung diseases; diabetes; psychiatric disorders; Alzheimer’s disease or other dementias; and cancer.31 Enrollment in these programs began in April 2002, and at its maximum reached about 21,000 patients in the intervention and control groups combined. However, the five largest programs accounted for almost 60 percent of the total enrollment, while three of them enrolled fewer than 100 beneficiaries in their intervention groups. Among the initial findings from the demonstration was that beneficiary recruitment in the fee-for-service market can be a challenge. The most successful of the programs had close ties to physicians and other providers, which helped reach the appropriate beneficiaries and overcome skepticism about enrolling. Although the evaluation results are not yet available, the programs overall appear to have been very well received both by enrolled beneficiaries and participating physicians. The initial term of the demonstration has ended, but most of the programs have been continued until more complete evidence on their performance can be collected and analyzed. The Medicare Disease Management Demonstration, mandated in the Benefits Improvement and Protection Act of 2000, was intended to provide disease management, as well as a comprehensive drug benefit, for up to 30,000 eligible beneficiaries. This project, which began in Spring 2004, was of particular interest because it was designed to provide the first indication of how well prescription drugs can be used to help chronically ill beneficiaries in the context of the Medicare program. The three sites selected were fully at risk for any increase in Medicare spending among their enrollees.32 The sites encountered greater-than-anticipated difficulties in identifying and enrolling beneficiaries, however; and, given the magnitude of the risk they faced, the demonstration was unable to continue to conclusion. A major initiative mandated in the Medicare Modernization Act of 2003 (MMA) was the Medicare Voluntary Chronic Care Improvement Program, now known as Medicare Health Support. This pilot program, which was implemented in August 2005 and will run for three years, was expected to involve about 160,000 beneficiaries at eight sites (Figure 4) around the United States with high prevalence of diabetes and congestive 9 heart failure.33 The participating organizations are responsible for increasing adherence to evidence-based care and reducing unnecessary hospital stays and emergency room visits in an entire geographic area.34 They each receive a per-beneficiary-per-month fee for their care coordination services, and in return are responsible for meeting quality, outcome, and patient satisfaction objectives while reducing total spending for their populations by at least 5 percent. If they fail to meet these requirements, they are responsible for reimbursing Medicare up to the total amount of their fees. One unique aspect of this project is that, unlike the other initiatives described here—in which beneficiaries were recruited to participate by explicitly indicating a willingness to “opt in”—eligible beneficiaries in Medicare Health Support areas were assumed to be participating in the demonstration unless they explicitly indicated that they wanted to “opt out” of it. The evaluation of the success of each site in meeting goals related to clinical quality outcomes, beneficiary satisfaction, and impact on program spending will be based on comparisons of beneficiaries who participated in the pilot programs with similar groups of beneficiaries who had indicated they were willing to participate but were instead randomly assigned to a control group. Based on these results, the Secretary of Health and Human Services has the authority to expand the breadth and scope of this program. Figure 4. Medicare Health Support Organizations and Locations • LifeMasters Supported SelfCare, Inc. (Okla.) • Health Dialogue Services Corp. (Western Pa.) • American Healthways, Inc. (Washington, D.C., and Md.) • McKesson Health Solutions, LLC (Miss.) • CIGNA Health Support, LLC (Northwest Ga.) • Aetna Health Management, LLC (Chicago, Ill.) • Green Ribbon Health (Central Fla.) • XLHealth Corp. (Select counties, Tenn.) THE COMMONWEALTH FUND Source: “Medicare Health Support.” www.cms.hhs.gov (Accessed Nov. 20, 2006). 10 Another project developed by CMS is the Care Management for High-Cost Beneficiaries Demonstration. This project, which began enrollment in Fall 2005 and is operating in six sites, aims to study various care management models for high-cost/high- risk beneficiaries.35 It is explicitly designed to use provider-directed, rather than third- party, models of chronic care management; and to test the ability of these sites to coordinate care for participating beneficiaries by providing them with clinical support beyond traditional settings to manage their conditions. As in Medicare Health Support, each of the sites in this demonstration receives a monthly fee for each beneficiary participating in the program and must achieve program savings while meeting established performance standards; otherwise, they must return all or part of their fee. The sites are employing a variety of features, including support programs for health care coordination, physician and nurse home visits, use of in-home monitoring devices, provider office medical records, self-care and caregiver support, education and outreach, tracking and reminders of individuals’ preventive care needs, 24-hour nurse telephone lines, behavioral health care management, and transportation services. Improving Chronic Care in Medicare Managed Care As mentioned earlier, several aspects of the financing mechanism that became an integral part of the managed care model—particularly in Medicare—are incompatible with the original vision of coordinated care as it applies to chronically ill enrollees. Although capitation should provide a strong incentive to help chronically ill enrollees manage their conditions and avoid expensive hospital stays, it also provides an even stronger incentive for plans to avoid chronically ill enrollees in the first place: they are much more costly than the average enrollee, and—although Medicare adjusts the payment rates that managed care plans receive for the higher anticipated costliness of some types of individual enrollees—this risk adjustment has been gradually phased in over 10 years (it will not take full effect until 2007). Thus, plans still face potentially severe financial penalties for making themselves attractive to chronically ill populations. Medicare managed care plans, moreover, were prohibited (until 2006) from specializing in subsets of the population. Consequently, a plan that was designed to be particularly well suited to treating beneficiaries with a particular condition or cluster of conditions (such as congestive heart failure, asthma, or other chronic respiratory diseases) also had to be prepared to offer the full range of services to the entire beneficiary population, which it might not have been prepared to do. One initiative intended to address this shortcoming is the inclusion in the MMA of a provision (Section 231) authorizing Special Needs Plans (SNPs). This provision allows for the creation of MA plans that focus on individuals with special needs, including 11 beneficiaries who are: institutionalized, dually eligible for Medicare and Medicaid, or suffering from severe or disabling chronic conditions. SNPs are not paid differently from other MA plans (so their payment will not be fully risk-adjusted until 2007), but—unlike other MA plans—they are permitted to target individuals in the specified groups, and CMS has been flexible in certain other MA administrative requirements as well. As of January 2006, there were 276 SNPs, with more than 600,000 enrollees: 226 SNPs, with nearly 500,000 enrollees, were approved for dual eligibles (a population that itself includes a high proportion of beneficiaries with chronic conditions); 37 SNPs, with about 40,000 enrollees, were focused on institutionalized beneficiaries (many of whom are both dually eligible and suffering from chronic conditions); and 13 SNPs, with about 70,000 enrollees, were focused specifically on beneficiaries with chronic conditions.36 A population that is particularly in need of better coordinated care is Medicare beneficiaries with End-Stage Renal Disease (ESRD); people with ESRD not only require dialysis but also have other chronic conditions. In 2003, there were 351,000 Medicare beneficiaries with ESRD, with Medicare spending an average of $46,330 per person for their health care.37 Despite their need for coordinated care, beneficiaries with ESRD are not permitted to enroll in MA plans unless they were enrolled prior to the onset of the condition, because of the extreme risk that this population presents. In an attempt to develop an approach that would permit these beneficiaries to participate in Medicare Advantage (then called the Medicare Risk Program), an ESRD Managed Care Demonstration was launched in 1996, with enrollment beginning in 1998. This demonstration was conducted at sites in California and Florida (with a third site in Tennessee discontinuing operations after enrolling just 50 beneficiaries).38 Each site provided service integration, case management, and extra benefits in exchange for being paid a higher payment rate (with adjustments to reflect the additional costliness of enrollees with ESRD). The evaluation concluded that enrollees in the demonstration fared as well as, or in some cases better than, a representative sample of fee-for-service comparison beneficiaries. However, government expenditures were found to be higher than if the same enrollees had remained in fee-for-service Medicare; this was because the demonstration enrollees were, on average, younger and healthier than the general ESRD population. Moreover, despite the increased payment by the government, the demonstration sites experienced financial losses in one case and only small gains in the other. With an extensively reworked risk adjustment mechanism that was thought to reflect better the costliness of ESRD enrollees, CMS in 2005 announced an ESRD 12 Disease Management Demonstration to test the capability of disease management models to increase quality of care while ensuring that this care is provided more effectively and efficiently.39 Enrollment in this new demonstration began at three sites in the fall of 2005, with coverage beginning in January 2006.40 Under this demonstration, 5 percent of the plans’ fees are reserved for incentive payments related to quality improvement. Initiatives to Improve Provider Performance The need for explicit financial incentives to improve quality and effectiveness of care in Medicare is widely recognized. The Institute of Medicine (IOM) made such a recommendation in 2002 and the Medicare Payment Advisory Commission (MedPAC) called for incentive payments in 2004.41 In an open letter published in 2003, prominent health care leaders—including several former Administrators of CMS and its predecessor, the Health Care Financing Administration—called for Medicare to create financial incentives that promote the pursuit of improved quality.42 They noted that Medicare demonstration authority gives the agency the power to expand its programs for the testing of models, and they suggested that the Medicare Program should make quality improvement for hospital care in particular its immediate priority. Also, in the past few years, there have been several efforts on the part of private sector employers and payers to improve the quality of care through incentive programs. More recently, in its report on rewarding provider performance in Medicare, the IOM concluded that: “The overall quality of health care delivered to Americans is worse than it should be. While many quality improvement efforts have been undertaken, their success has been limited by current payment systems. The existing systems do not reflect the relative value of health care services in important aspects of quality . . . . Nor do current payment systems recognize or reward care coordination . . . . Fundamental changes in approaches to health care payment are necessary to remove impediments to and create incentives for significant quality improvement.”43 These and similar findings have spurred an acceleration in the pace of activities aimed at improving the quality of care provided to Medicare beneficiaries. Moreover, with Medicare spending exceeding $375 billion in 2006, there is increasing pressure to improve the effectiveness of the program in obtaining the best possible outcomes for the resources that are expended. CMS has undertaken several initiatives in recent years to encourage improved quality of care, placing emphasis first on public reporting of quality indicators in a variety of health care settings. The agency is providing technical assistance to a wide range of 13 providers through its QIOs. Additionally, CMS has developed demonstration projects to test ways of providing financial incentives to encourage better performance. Initiatives Focused on the Reporting of Performance Measures There is evidence that the reporting of performance measures can enhance quality. A 2006 study by the National Committee for Quality Assurance found that health plans that collected and publicly reported performance data demonstrated broad-based improvements.44 Public reporting can be an incentive for group practices as well; groups that participate in public reporting tend to have higher performance.45 CMS has developed and implemented several initiatives involving the reporting of quality measures for private plans, nursing homes, home health agencies, hospitals, and physicians.46 These initiatives are summarized in Table 2 and in the discussion that follows. Table 2. Aligning Payment and Quality: Initiatives Focused on the Reporting of Performance Measures Initiative Description Nursing Home Quality Initiative Launched nationally in 2002, this initiative involves the reporting of post-acute and chronic care quality measures on the Nursing Home Compare Web site. Home Health Quality Initiative Launched in 2003, this initiative involves the reporting of quality measures relating to improvements in patient functionality on the Home Health Compare Web site. Hospital Quality Initiative This initiative began in 2003 with the voluntary reporting of a starter set of 10 hospital quality measures on the Hospital Compare Web site. The MMA provided a financial incentive for hospitals to report these measures, and the Deficit Reduction Act of 2005 expanded the set of measures and increased the amount of the financial incentive. ESRD Quality Initiative Measures of quality for dialysis facilities are available on the Dialysis Facility Compare Web site. Physician-Focused Quality Initiative The goals of this initiative, currently under development, are to develop measures of ambulatory care quality; develop and test means for measuring the quality of care for chronic diseases and preventive services provided in doctors’ offices; and support physicians’ adoption and effective use of information technology. Source: CMS Web site, http://www.cms.hhs.gov. The Nursing Home Quality Initiative was launched nationally in 2002 with the availability of post-acute and chronic care quality measures on the Nursing Home Compare Web site. This initiative focuses on both regulation and enforcement, as well as collaboration with Medicare’s QIOs, to improve the quality of care in nursing homes. 14 In 2003, the Home Health Quality Initiative began with a set of quality measures relating to improvements in patient functionality obtained from the Outcome and Assessment Information Set (OASIS) that is routinely collected from all Medicare home health agencies. These measures are available on the Home Health Compare Web site. Also in 2003, the Hospital Quality Initiative was started with voluntary reporting of a starter set of 10 hospital quality measures. The MMA provided hospitals with a financial incentive to report these measures, withholding 0.4 percent of their annual update in Medicare payments if they did not report them beginning in 2005—at which point compliance with the “voluntary” reporting rose to almost 100 percent. The Deficit Reduction Act of 2005 expanded the set of reportable measures required for hospitals to receive their full update to 21 and increased the amount of the withhold to 2 percent of payments beginning in 2008. These measures are available on the Hospital Compare Web site. The ESRD Quality Initiative began in 2004, with posting of quality measures for dialysis facilities on the Dialysis Facility Compare Web site. In addition, CMS implemented a strategy to improve care by setting a goal of arterial venous fistula (AVF) utilization by 65 percent of dialysis patients by 2009 (AVF is the preferred method of vascular access for patients undergoing dialysis because it provides adequate blood flow, lasts longer than alternative methods, and has a lower complication rate). The Physician-Focused Quality Initiative includes: development of measures of ambulatory care quality and their endorsement by the National Quality Forum; implementation of a Doctor’s Office Quality (DOQ) project to develop and test a comprehensive and integrated approach to measuring the quality of care for chronic disease and preventive services in doctors’ offices; and a Doctor’s Office Quality Information Technology (DOQ-IT) project to support the adoption and effective use of information technology by physicians to improve quality and safety for Medicare beneficiaries and all patients. The DOQ-IT project puts Medicare’s QIOs in the role of providing infrastructure support for disseminating appropriate health care information technology and promoting its effective use. An additional part of the Physician-Focused Quality Initiative is the Physician Voluntary Reporting Program (PVRP). This is an effort to build on existing efforts to measure and stimulate improvements in the quality of care provided by physicians, by encouraging voluntary submission of specified quality data beginning in January 2006. The Tax Relief and Health Care Act of 2006 provides physicians with a 1.5 percent incentive payment beginning in July 2007 if they report on the PVRP measures in 2007. 15 Financial Incentives to Improve Provider Performance Efforts to improve quality and efficiency have gone beyond public reporting, with stakeholders across the country focusing on quality/payment alignment. At present, the private sector has more than 100 pay-for-performance programs in place. Concurrently, Medicare demonstrations have been examining such incentive systems in the public sector. Table 3 provides a summary of these demonstrations, which are then further discussed in the paragraphs that follow. Table 3. Aligning Payment and Quality: Financial Incentives to Improve Provider Performance Initiative Description Hospital Quality Incentive More than 255 hospitals are voluntarily participating in this partnership between Demonstration CMS and Premier, Inc., which began in October 2003. It includes a bonus on top of the DRG payment for top relative performers, and failure to exceed a fixed minimum performance threshold by the third year results in penalties. Physician Group Practice This three-year project, which began in April 2005, includes 10 large multi- Demonstration specialty physician groups. Savings generated by each group relative to cost growth in its service area are shared between the group and Medicare, depending on both the amount of savings and the group’s achievement of quality targets. Medicare Care This project focuses on small to medium-sized practices in California, Arkansas, Management Performance Massachusetts, and Utah. Its goal is to promote the adoption and use of health Demonstration information technology to improve the quality of care for chronically ill patients. Nursing Home Pay-for- Under development, this initiative will provide financial rewards to nursing Performance Demonstration homes for meeting certain quality improvement standards and providing high- quality care. Home Health Agencies Beginning in 2007, CMS will begin paying for the reporting of home health quality data. Medicare Participating This project was conducted to examine the feasibility, cost-effectiveness, and Heart Bypass Center impact on quality of a negotiated all-inclusive bundled payment arrangement for Demonstration coronary bypass graft surgery. Medicare Partnerships for This project was intended to assess whether bundled payments covering both Quality Cardiovascular hospital and physician services for certain high-volume, high-cost procedures Services and Quality Total would improve coordination and quality of care while achieving savings for Joint Replacement Services Medicare. Plans to implement this demonstration were suspended, however. Demonstration Medicare Health Care This five-year demonstration will test major changes in system design aimed both Quality Demonstration at improving quality of care and increasing efficiency. Physician Hospital This project will assess the effects of gainsharing on the quality of care in a health Collaboration delivery system. Demonstration Medicare Hospital This demonstration will examine the use of gainsharing arrangements between Gainsharing Demonstration hospitals and physicians to improve quality and efficiency of care. Source: CMS Web site, http://www.cms.hhs.gov. 16 CMS is testing models for rewarding hospitals that demonstrate high-quality performance. The Hospital Quality Incentive Demonstration, which started in October 2003, is a partnership between CMS and Premier, Inc., a nationwide purchasing alliance that includes some 1,500 not-for-profit hospitals. Participation in this demonstration is voluntary; as of January 2006, it included more than 255 hospitals. Rewards (and potential penalties) under the demonstration are based on 34 process and outcome measures that describe the quality of care for inpatients with five conditions: heart attack, heart failure, pneumonia, coronary artery bypass graft, and hip and knee replacements. Data on these measures are posted on the Medicare Web site, and top-performing hospitals are rewarded financially: hospitals ranking in the top 10 percent among the participating facilities for each condition receive a bonus equal to 2 percent of their Medicare payments for patients with that condition, while hospitals in the second decile for each condition are paid a 1 percent bonus. By the third year of the demonstration, hospitals that are performing relatively poorly (i.e., they fail to exceed performance requirements based on the distribution of scores in the first year) may be subject to decreased payments. For year one, CMS paid out a total of approximately $8.9 million in bonuses under this program, with incentive payments for individual hospitals of as much as $847,000.47 Results from the first year indicate that hospital performance improved in every category, and this improvement appears to have taken place not only among the top hospitals but all the hospitals in the demonstration.48 Moreover, hospitals that were high performers on the quality measures also tended to have lower rates of readmissions, indicating that high quality and lower costs are compatible (Figure 5).49 17 Figure 5. Medicare Premier Hospital Demonstration: Higher Quality Hospitals Have Fewer Readmissions Readmission rates by pneumonia quality ranking (percent) 20 15.4 14.8 15 13.6 13.1 11.6 10 5 0 Bottom 26–50% 51–75% 76–90% Top quality quality decile quartile THE © 2005 Premier, Inc. COMMONWEALTH FUND Source: Stephanie Alexander, “CMS/Premier Hospital Quality Incentive Demonstration Project: 1st Year Results,” Presentation at Institute of Medicine Pay-for-Performance Subcommittee Meeting, Nov. 30, 2005. CMS is also testing financial incentives for physicians. The Physician Group Practice Demonstration, which began in April 2005, provides incentives for large multi-specialty group practices to improve the coordination of care for their Medicare fee- for-service beneficiaries. The demonstration’s goals are to promote coordination of all Medicare services, encourage investment in administrative structure and process to increase efficiency, and reward physicians for improving health outcomes.50 The 10 sites participating in the demonstration represent more than 5,000 physicians and 200,000 fee- for-service Medicare beneficiaries (Figure 6). 18 Figure 6. Medicare Physician Group Practice Demonstration THE COMMONWEALTH FUND Source: M. Trisolini, G. Pope, J. Kautter, and J. Aggarwal, Medicare Physician Group Practices: Innovations in Quality and Efficiency (New York: The Commonwealth Fund, Dec. 2006). Each of the participating practices shares in any Medicare savings resulting from improved coordination and efficiency, with the size of the bonus depending on the difference between total Medicare spending for the patients assigned to the practice and a target amount. The target, in turn, is calculated from base-year spending for the patients assigned to the practice, inflated by growth in case-mix-adjusted per-capita spending for other beneficiaries in the practice’s service area. Medicare retains the first 2 percent of savings, and 80 percent of the remaining savings go into a bonus pool. Each practice will receive 70 percent of the amount in its bonus pool directly as a cost performance payment (this share falls to 60 percent in the second year and 50 percent in the third year). The other 30 percent of the practice’s bonus pool is distributed in accordance with the practice’s performance on a set of quality measures (this share rises to 40 percent in the second year and 50 percent in the third year). Across sites there has been a focus on improving care management and coordination of care; expanding palliative and hospice care; modifying physician practice patterns and behavior; and enhancing information technology.51 Initial results have been promising, but the bonus amounts for the first year have not yet been calculated and distributed (as of December 2006). 19 CMS is also expanding its pay-for-performance demonstrations to smaller physician practices. The Medicare Care Management Performance Demonstration, which will focus on solo and small to medium-sized practices, was recently announced. The goal of this demonstration, which was mandated in the MMA, is to promote the adoption and use of health information technology to improve the quality of care for chronically ill beneficiaries. CMS will implement this demonstration in Arkansas, California, Massachusetts, and Utah, and physicians who meet or exceed clinical performance standards will receive a bonus payment for managing the care of eligible Medicare beneficiaries. The demonstration also will allocate rewards based on how well practices do in the provision of preventive services. Payments under the demonstration include an initial payment for reporting baseline clinical quality measures; an annual payment based on the practice’s score on the clinical measures; and an additional annual bonus for reporting some or all of the measures electronically through an electronic health record system that meets the standards of the Certification Commission for Healthcare Information Technology.52 One interesting aspect of the Medicare Care Management Performance Demonstration is that it is an expansion of the DOQ-IT initiative described earlier, with the Medicare QIOs providing technical assistance to participating practices so that they can meet the standards required for bonus payments. Another notable feature is that it is an attempt to adapt the model used in the Bridges to Excellence private sector initiative developed by General Electric, Verizon, FedEx, and several other large employers at several sites around the country; one of the Bridges to Excellence sites is the Boston area, which provides the opportunity to observe and evaluate the overlap and synergy between essentially similar (but not identical) private and public initiatives. In addition to the hospital and physician demonstrations described above, CMS is developing demonstrations that offer financial incentives to other types of providers. Under a Nursing Home Pay-for-Performance Demonstration, financial rewards would be provided to nursing homes that meet certain standards for delivering high- quality care and also for quality improvement, facilitating the sharing of best practices.53 For home health agencies, CMS will begin paying for the reporting of home health quality data in 2007, as mandated by the Deficit Reduction Act of 2005. Medicare also has implemented or is planning several initiatives that encourage collaboration between hospitals and physicians. For example, the Medicare Participating Heart Bypass Center Demonstration was conducted to assess the feasibility, cost-effectiveness, and impact on quality of a negotiated all-inclusive bundled 20 payment arrangement for coronary bypass graft surgery. The demonstration began with four sites in 1991, and was expanded to include three more sites in 1993.54 The evaluation indicated that all-inclusive bundled payment can provide an incentive to physicians and the hospital to work together to provide services more efficiently, improve quality, and reduce costs. The demonstration involved major changes in reimbursement arrangements, which, however, benefited some parties (such as patients, who had a single copayment amount) but were seen as a burden by others (such as hospitals, which were responsible for billing and collection).55 Medicare developed the Medicare Partnerships for Quality Cardiovascular Services and Quality Total Joint Replacement Services Demonstration to test whether bundled payments covering both hospital and physician services for certain high- volume, high-cost procedures would improve the coordination and quality of care provided to beneficiaries while achieving savings for the program. This demonstration had been conceived in the mid-1990s, but it was put on hold because of extensive system requirements mandated in the Balanced Budget Act of 1997 and additional Y2K-related changes; it was restarted in 2000. Applications were accepted from facilities in Michigan, Illinois, and Ohio, and the selection process was initiated, but plans to implement the demonstration were suspended in the fall of 2002, before final selections could be made, as the combination of delays and the difficulties in sorting out responsibilities among the various parties prevented the project from regaining its original momentum. A project with a very broad scope, the Medicare Health Care Quality Demonstration Program, was mandated by the MMA. This is a five-year project under which CMS will test major changes in system design aimed at improving quality of care while increasing efficiency. Unlike most other demonstrations, which are relatively limited in scope and intended to test specific types of changes in Medicare rules, the provision that mandated this program gives CMS broad flexibility to consider a range of payment systems designed to support significant changes in the organization of health care delivery. Organizations eligible to participate in the demonstrations are integrated delivery systems and regional consortia of providers. One round of applications for this program was due by the end of January 2006; CMS is (as of December 2006) in the process of finalizing the selection of several sites from that round. Another round of applications was due at the end of September 2006, and CMS has begun evaluating those as well. CMS has stated that it intends to choose a total of 8 to 12 sites to focus on improving the safety, effectiveness, efficiency, patient-centeredness, timeliness, and equity of the health care system.56 21 Further, under the umbrella of the Medicare Health Care Quality Demonstration CMS has announced that it is willing to consider the incorporation of financial and quality measurement and reporting models used in the Physician Group Practice Demonstration in proposals received from applicants that meet all of the goals of the Medicare Health Care Quality Demonstration and also have 150 or more physicians. In addition, CMS has, under the authority provided for the Medicare Health Care Quality Demonstration, developed a Physician Hospital Collaboration Demonstration—a three-year project to examine the effects of gainsharing aimed at improving the quality of care in a health delivery system. CMS has announced that it is particularly interested in demonstration designs that track patients well beyond a hospital episode. This will permit evaluation of the impact of hospital–physician collaborations on preventing short- and longer-term complications, avoiding duplication of services, coordinating care across settings, and making other improvements that eliminate preventable complications and reduce unnecessary costs. The Deficit Reduction Act of 2005 mandated a Medicare Hospital Gainsharing Demonstration to test different types of arrangements between hospitals and physicians designed to improve quality and efficiency of care provided to beneficiaries. This three-year demonstration will allow hospitals to provide physicians with gainsharing payments that represent solely a share of the savings incurred as a result of collaborative efforts to improve overall quality and efficiency. These arrangements are otherwise restricted by the civil monetary penalty law, which prohibits hospitals from rewarding physicians for reducing services to patients, even if such reductions are limited to duplicative services or otherwise represent improvements in quality.57 MOVING FROM DEMONSTRATIONS AND PILOTS TO PROGRAM IMPROVEMENTS All of the demonstrations, pilot projects, and other initiatives described above are intended to find ways to change the way in which Medicare does business—to help Medicare move from its traditional role as a passive payer of health care bills to a more active purchaser of effective and efficient health care for its beneficiaries. With pressure mounting to find ways to improve quality while simultaneously controlling the growth of Medicare spending, it is increasingly important to be clear on what these initiatives have to tell us about whether and how policy should be changed. However, it is also increasingly difficult to find the resources needed to carefully study the results these initiatives produce. The availability of more resources—for research to help 22 identify potential improvements and assess their likelihood of success; for the development of appropriate design and implementation strategies; and for the evaluation of results— would facilitate the translation of those initiatives into better policy. Moreover, the process of identifying, developing, testing, evaluating, and implementing Medicare policy improvements includes several hurdles. Here we discuss a few of those hurdles and offer some suggestions for lowering or eliminating them. One major problem in testing new and potentially productive ideas is the inflexibility of the waiver approval process. Franklin Roosevelt has been quoted as saying: “The country needs, and unless I mistake its temper, the country demands, bold, persistent, experimentation. It is common sense to take a method and try it. If it fails, admit it frankly and try another. But above all, try something.”58 With Medicare facing insolvency and quality of care in dire need of improvement, it is important to try new ideas that have a reasonable prospect of success in improving quality and efficiency. Another impediment to broad thinking in this regard is the requirement of budget neutrality for approval of new initiatives. This requirement and the way it is applied by the Office of Management and Budget (which has final approval of demonstration waivers) presents at least three difficulties. One problem is that budget neutrality is defined so strictly with regard to time that even a policy change that saves money in the intermediate or long run risks rejection because it may require some initial increase in federal spending. Another issue is that budget neutrality is defined separately for each federal program. Initiatives that reduce overall spending (improving coordination of Medicare and Medicaid for dual eligibles, for example) may not be considered budget neutral because one program may spend more while the other spends much less; the impact on other affected parties, such as states, is yet another complicating consideration. Moreover, the current application of the budget neutrality requirement does not allow for initiatives that substantially improve value even if they involve some extra spending. This shortsighted and overly restrictive interpretation of what in general is a reasonable criterion—but not the only reasonable criterion—for evaluating the potential merits of a new policy initiative hinders Medicare’s ability to develop and test improvements. It also fosters the view that CMS is more interested in reducing spending than improving the program. It is important to remember that Medicare (and Medicaid, for that matter) is not just a line item in the federal (or state) budget, but a social program that exists to provide access to necessary care to the population it serves. Ironically, this 23 narrow interpretation prevents CMS from implementing new policies that might both improve its programs and reduce overall public spending. Another problem that hinders the development of policy improvements is the limitations of the available methodology for evaluating their impact. Theoretically, demonstrations and pilots are carefully designed and implemented to adhere to strict methodological criteria in a carefully controlled environment. These types of social experiments, however, are not conducted in laboratories, but in a world in which the policy environment is constantly changing; the ability to maintain complete control over all aspects of such trials is limited. Consequently, although formal evaluations are conducted in most cases, those evaluations must deal with imperfect controls and incomplete data with which to control for mitigating factors. Moreover, in many cases some of the major objectives of the policy change being tested are difficult to measure, either because they are qualitative in nature or because no baseline data exist to determine whether the policy in question has had the hoped-for effect. New ways of evaluating the results of demonstrations and pilots and identifying their implications for the design of potential policy improvements would be useful for maintaining the appropriate balance between scientific rigor and policy usefulness. The timeliness of evaluations is also a chronic issue. Careful evaluation requires accurate and complete data, but the process of collecting, cleaning, and analyzing those data is not only inherently time-consuming but frequently either cannot begin until the initiative is over or has to start before the full effects of the trial have occurred. This can result in failure to act on potentially useful policy initiatives at the opportune moment or premature enactment of incompletely informed policy decisions. Mechanisms must be put in place that allow for continuous monitoring of evaluations, or perhaps bellwether measures can be developed that allow preliminary evaluations to help indicate directions not only for the development of new policies but also for changes in the trial itself. In any case, these issues call for a reexamination of the methodological tools at our disposal so that we can attain the best balance of scientific rigor and policy usefulness. Finally, the program would benefit from a more explicit mechanism for translating what we learn from demonstrations into new policy. In some cases—the Medicare Health Support pilot is the most recent example—the Congress gives Medicare (through the Secretary of Health and Human Services) the authority to continue or expand a trial, but most often additional action is needed in order to effect the policy changes suggested by demonstration results. Even when independent evaluations of a mandated demonstration’s 24 results are required by law, it can take years for the evaluation report to be cleared by the executive branch and transmitted to Congress. Again, making the process more transparent would help considerably, as this would allow more open discussion of policy changes of interest and their potential impacts. The designation of a regular vehicle for reporting the findings from demonstrations and other initiatives, such as in an issue of the Health Care Financing Review, would provide a visible mechanism for reviewing those results. Another such mechanism might be the annual reports produced by MedPAC. Those reports are widely distributed and consulted, and their release is generally followed by Congressional hearings to review their recommendations. CONCLUSIONS Medicare, being much in need of good ideas for policies that address the evolving needs of its beneficiaries and the health care system overall, has undertaken an array of initiatives— and will pursue others in the future—to help it do that. In this paper, we suggest several ways of improving the process used to identify the policy issues that need to be addressed, design and implement the appropriate initiatives, evaluate their results, and then disseminate the indicated improvements to additional sites around the nation. To carry out these changes, more resources are required and several impediments to progress should be modified or eliminated. One needed improvement is increased clarity and flexibility in the demonstration approval process. The purpose of Medicare demonstrations should be to enhance the program’s value, and the process of identifying, developing, approving, and implementing them should be more transparent and explicitly based on criteria consistent with that purpose. A major requirement for approval of Medicare demonstrations is budget neutrality, which was rightly intended to protect the fiscal integrity of the program; however, budget neutrality as currently applied has become a strict and narrowly defined rule that can hinder the development of important new initiatives. The application of this requirement must take into account the fact that some policy changes may involve short-term costs but long-term benefits. Cross-program savings and costs must be considered as well, so that policy changes that reduce overall spending can be tested, even if they require spending increases in some individual programs; the same reasoning should be applied across levels of government. In addition, there should be procedures for considering policy changes that increase value, even if they do not reduce spending per se. 25 The methodology for evaluating demonstration results needs to be reexamined. We need to balance the imperfect controls and incomplete data available in the real world of policy analysis with the need for rigorous testing of potential policy improvements. This means developing ways of providing timely results that meet the needs of policymakers while maintaining research standards that ensure scientific validity. Finally, a more explicit and transparent mechanism for moving from pilot to policy must be established. The designation of a regular vehicle for reporting evaluation results would provide a visible source of information and a platform for the open discussion of their policy implications. Such a strategy of dissemination would enable potential policy improvements to be considered and implemented and their benefits realized in a timely manner. New initiatives to improve Medicare should build on the lessons learned from demonstrations, allowing interventions to be reshaped as they proceed to make them more effective. These changes will help improve the quality and effectiveness of health care while controlling the precipitous increases in cost that threaten the program’s fiscal viability. 26 NOTES 1 K. Davis and S. R. Collins, “Medicare at Forty,” Health Care Financing Review, Winter 2005–2006 27(2):53–62. 2 The Commonwealth Fund Commission on a High Performance Health System, Why Not the Best? Results from a National Scorecard on U.S. Health System Performance (The Commonwealth Fund, Sept. 2006). 3 Boards of Trustees, Federal Hospital Insurance and Federal Supplementary Medical Insurance Trust Funds, 2006 Annual Report (Washington, D.C.: U.S. Government Printing Office, May 1, 2006). 4 G. F. Anderson, “Medicare and Chronic Conditions,” New England Journal of Medicine, July 21, 2005 353(3):305–09. 5 G. F. Anderson and P. S. Hussey, “Population Aging: A Comparison Among Industrialized Countries,” Health Affairs, May/June 2000 19(3):191–203. 6 S. R. Collins, K. Davis, C. Schoen, M. M. Doty, S. K. H. How, and A. L. Holmgren, Will You Still Need Me? The Health and Financial Security of Older Americans (New York: The Commonwealth Fund, June 2005). 7 Partnership for Solutions, “Medicare: Cost and Prevalence of Chronic Conditions,” Fact Sheet (Baltimore: Johns Hopkins University and the Robert Wood Johnson Foundation, July 2002). 8 J. L. Wolff, B. Starfield, and G. F. Anderson, “Prevalence, Expenditure, and Complications of Multiple Chronic Conditions and the Elderly,” Archives of Internal Medicine, Nov. 11, 2002 162(20):2269–76. 9 U.S. Congressional Budget Office, “High-Cost Medicare Beneficiaries” (Washington, D.C.: CBO, May 2005). 10 U.S. Congressional Budget Office, “Fact Sheet for CBO’s March 2006 Baseline: Medicare” (Washington, D.C.: CBO, corrected Mar. 15, 2006). 11 S. Guterman, “U.S. and German Case Studies in Chronic Care Management: An Overview,” Health Care Financing Review, Fall 2005 27(1):1–8. 12 B. Biles, L. H. Nicholas, and S. Guterman, Medicare Beneficiary Out-of-Pocket Costs: Are Medicare Advantage Plans a Better Deal? (New York: The Commonwealth Fund, May 2006). 13 Centers for Medicare & Medicaid Services, 2006 Annual Report of the Boards of Trustees of the Federal Hospital Insurance and Federal Supplementary Medical Insurance Trust Funds (Washington, D.C.: CMS, 2006), p. 30, http://www.cms.hhs.gov/ReportsTrustFunds/downloads/tr2006.pdf, accessed Oct. 2006. 14 Henry J. Kaiser Family Foundation, “Medicare at a Glance,” fact sheet (Washington, D.C: Kaiser Family Foundation, July 2006). As specified in the Medicare Modernization Act of 2003, the “Medicare funding warning” is generated when the Medicare Trustees’ Reports in two consecutive years indicate that the proportion of Medicare spending from general revenues will exceed 45 percent within seven years. 15 Commonwealth Fund Commission, Why Not the Best? 2006. 16 G. F. Anderson, B. K. Frogner, R. A. Johns, and U. E. Reinhardt, “Health Care Spending and Use of Information Technology in OECD Countries,” Health Affairs, May/June 2006 25(3):819–31. 17 Commonwealth Fund Commission, Why Not the Best? 2006. 27 18 C. Schoen, K. Davis, S. K. H. How, and S. C. Schoenbaum, “U.S. Health System Performance: A National Scorecard,” Health Affairs Web Exclusive (Sept. 20, 2006):w457–w475. 19 Ibid. 20 Commonwealth Fund Commission, Why Not the Best? 2006. 21 Ibid. 22 Schoen et al., “National Scorecard,” 2006. 23 Commonwealth Fund Commission, Why Not the Best? 2006. 24 Medicare Payment Advisory Commission, “National Health Care and Medicare Spending,” In A Data Book: Health Care Spending and the Medicare Program (Washington, D.C.: MedPAC, June 2006). http://www.medpac.gov. 25 The descriptions of the projects in this section are based on information at the Centers for Medicare & Medicaid Services Web site, http://www.cms.hhs.gov/DemoProjectsEvalRpts/MD/list.asp#, accessed Oct. 24, 2006. 26 S. Guterman, “Eliminating Barriers to Chronic Care Management in Medicare,” testimony before the U.S. House of Representatives, Committee on Ways and Means, Subcommittee on Health, Feb. 25, 2006. 27 By “disease management” we mean programs that are aimed at improving the quality and coordination of care for patients with single or multiple chronic conditions, in an effort to provide more effective care, eliminate avoidable acute care episodes, and improve outcomes. 28 S. Guterman, “Disease Management in Traditional Medicare: A Square Peg in a Round Hole?” testimony before a Special Forum of the U.S. Senate, Special Committee on Aging, Nov. 4, 2003. 29 Participating sites included AdminiStar Solutions, Iowa Foundation for Medical Care (IFMC), and Providence Hospital. AdminiStar Solutions recruited Medicare CHF patients throughout the state of Indiana; IFMC recruited Medicare CHF and COPD patients seen at any of 10 participating hospitals in Des Moines, western Iowa, and eastern Nebraska; and Providence Hospital (in Southfield, Mich.) took Medicare beneficiaries with CHF, COPD, or a range of other chronic problems who were patients of the hospital’s staff and resided in the Detroit metropolitan area. 30 J. L. Schore, R. S. Brown, and V. A. Cheh, “Case Management for High-Cost Beneficiaries,” Health Care Financing Review, Summer 1999 20(4):87–101. 31 The organizations originally participating in this demonstration were: Avera McKennan Hospital of Sioux Falls, S.D.; Carle Foundation Hospital of Urbana, Ill.; CenVaNet of Richmond, Va.; CorSolutions Medical, Inc. of Buffalo Grove, Ill. (site in Texas); Erickson Retirement Communities of Baltimore, Md.; Georgetown University Medical Center of Washington, D.C.; Hospice of the Valley of Phoenix, Ariz.; Jewish Home and Hospital of New York, NY; Mercy Medical Center of Mason City, Iowa; Medical Care Developments of Augusta, Maine; PennCARE of Allentown, Pa.; Quality Oncology, Inc., of McLean, Va. (site in Broward County, Fla.); QMED, Inc., of Laurence Harbor, N.J. (site in Northern Calif.); University of Maryland at Baltimore; and Washington University of St. Louis, Mo., with StatusOne Health of Hopkinton, Mass. (site in St. Louis, Mo.). 32 The three participating sites were CorSolutions of Rosemont, Ill. (site in the Shreveport- New Orleans corridor of La.); XLHealth of Baltimore, Md. (site in Texas); and HeartPartners of Santa Ana, Calif. (site in Calif. and Ariz.). 33 As of December 2006, the number of beneficiaries participating in the demonstration was about 120,000, but one site—LifeMasters Supported SelfCare, Inc., operating in Oklahoma—was scheduled to drop out at the end of the month. 28 34 The organizations participating in Phase I of Medicare Health Support are LifeMasters Supported SelfCare, Inc. (site in Okla.); Health Dialog Services Corp. (site in Western Pa.); American Healthways, Inc. (site in Washington, D.C., and Md.); McKesson Health Solutions, LLC (site in Miss.); CIGNA Health Support (site in Northwest Ga.); Aetna Health Management, LLC (site in Chicago, Ill.); Green Ribbon Health (site in Central Fla.); and XLHealth Corp. (site in Tenn.). 35 The organizations participating in this demonstration are ACCENT (site in Ore. and Wash.); Care Level Management (sites in Calif., Texas, and Fla.); Massachusetts General Hospital and Massachusetts General Physicians Organization (site in Boston, Mass.); Montefiore Medical Center (site in the Bronx, N.Y.); RMS Disease Management, LLC (site in Nassau and Suffolk Counties in N.Y.), and Texas Senior Trails (site in Texas panhandle area). 36 U.S. Department of Health and Human Services, “Special Needs Plan—Fact Sheet & Data Summary” (Washington, D.C.: DHHS, Feb. 14, 2006). 37 Health Care Financing Review Medicare and Medicaid Statistical Supplement, 2005 (July 2006), p. 56. 38 Lewin Group and University Renal Research and Education Association, Final Report on the Evaluation of CMS’s ESRD Managed Care Demonstration (Falls Church, Va.: Lewin Group, June 2002). 39 Centers for Medicare & Medicaid Services, “CMS Announces Demonstration Sites for Beneficiaries with End-Stage Renal Disease,” press release (Washington, D.C.: CMS, Oct. 31, 2005). 40 The organizations participating in this demonstration are DaVita, with SCAN Health Plan (which is offering an MA SNP in parts of San Bernardino and Riverside Counties, Calif.); Fresenius Medical Care North America, with Sterling Life Insurance Co. (which is offering an MA private fee-for-service plan in Philadelphia and Pittsburgh, Pa., and Dallas, Houston, and San Antonio, Texas); and Fresenius Medical Care North America, with American Progressive Life and Health Insurance Co. (which is offering an MA private fee-for-service plan in Boston and Springfield, Mass.). 41 Institute of Medicine, Leadership by Example: Coordinating Government Roles in Improving Health Care Quality (Washington, D.C.: National Academies Press, 2002); Medicare Payment Advisory Commission, Report to the Congress: Medicare Payment Policy (Washington, D.C.: MedPAC, Mar. 2004). 42 D. M. Berwick, N. A. DeParle, D. M. Eddy, “Paying for Performance: Medicare Should Lead,” Health Affairs, Nov./Dec. 2003 22(6):8–10. 43 Institute of Medicine, Rewarding Provider Performance: Aligning Incentives in Medicare (Washington, D.C.: National Academies Press, 2006), p. 1. 44 S. C. Schoenbaum and A. L. Holmgren, The National Committee for Quality Assurance’s The State of Health Care Quality 2006 (New York: The Commonwealth Fund, Nov. 2006). 45 S. M. Shortell, J. Schmittdiel, M. C. Wang et al., “An Empirical Assessment of High- Performing Medical Groups: Results from a National Study,” Medical Care Research and Review, Aug. 2005 62(4):407–34. 46 See “Quality Initiatives General Information” on the CMS Web site at http://www.cms.hhs.gov/qualityinitiativesgeninfo. 47 Centers for Medicare & Medicaid Services, “Rewarding Superior Quality Care: The Premier Hospital Quality Incentive Demonstration,” fact sheet (Washington, D.C.: CMS, Jan. 2006). 48 Premier, Inc., CMS/Premier Hospital Quality Incentive Demonstration Project: Findings from Year One (Charlotte, N.C.: Premier, Inc., 2006). 29 49 S. Alexander, “CMS/Premier Hospital Quality Incentive Demonstration Project: 1st Year Results,” Presentation at Institute of Medicine Pay-for-Performance Subcommittee Meeting, Nov. 30, 2005. 50 Centers for Medicare & Medicaid Services, “Medicare Physician Group Practice Demonstration,” fact sheet (Washington, D.C.: CMS, Jan. 31, 2005). 51 M. Trisolini, G. Pope, J. Kautter et al., “Medicare Physician Group Practice Site Meeting: Quality and Efficiency Innovations,” site meeting summary (New York: The Commonwealth Fund, forthcoming). 52 Centers for Medicare & Medicaid Services, “Medicare Care Management Performance Demonstration,” demonstration summary (Washington, D.C.: CMS, 2006). 53 Abt Associates, “Nursing Home Pay-for-Performance Demonstration Draft Design,” presentation at Centers for Medicare & Medicaid Services’ Open Door Forum, Sept. 20, 2005. 54 The organizations originally participating in the demonstration were St. Joseph’s Hospital of Atlanta, Ga.; St. Joseph Mercy Hospital of Ann Arbor, Mich.; the Ohio State University Hospitals of Columbus, Ohio; and University Hospital of Boston, Mass. The three sites added in spring 1993 were St. Luke’s Episcopal Hospital of Houston, Texas; St. Vincent’s Hospital of Portland, Ore.; and Methodist Hospital of Indianapolis, Ind. 55 J. Cromwell, D. A. Dayhoff, N. T. McCall et al., “Medicare Participating Heart Bypass Center Demonstration,” final report (Washington, D.C.: Health Care Financing Administration, July 1998). 56 Centers for Medicare & Medicaid Services, “Medicare Modernization Act, Section 646: Medicare Health Care Quality Demonstration Program,” fact sheet (Washington, D.C.: CMS, 2005). 57 Centers for Medicare & Medicaid Services, “Medicare Hospital Gainsharing Demonstration,” fact sheet (Washington, D.C.: CMS, 2006). 58 As quoted in A. Dobson, D. Moran, and G. Young, “The Role of Federal Waivers in the Health Policy Process,” Health Affairs Winter 1992 11(4):72–94. 30 RELATED PUBLICATIONS Publications listed below can be found on The Commonwealth Fund’s Web site at www.cmwf.org. State Strategies to Expand Health Insurance Coverage: Trends and Lessons for Policymakers (January 2007). Alice Burton, Isabel Friedenzohn, and Enrique Martinez-Vidal. Public Reporting and Transparency (January 2007). John M. Colmers. Slowing the Growth of U.S. Health Care Expenditures: What Are the Options? (January 2007). Karen Davis, Cathy Schoen, Stuart Guterman, Tony Shih, Stephen C. Schoenbaum, and Ilana Weinbaum. The State Children’s Health Insurance Program: Past, Present, and Future (January 2007). Jeanne M. Lambrew. Medicare Physician Group Practices: Innovations in Quality and Efficiency (December 2006). Michael Trisolini, Gregory Pope, John Kautter, and Jyoti Aggarwal. Developing a Center for Comparative Effectiveness Information (November 7, 2006). Gail R. Wilensky, Health Affairs Web Exclusive (In the Literature summary). The Cost of Privatization: Extra Payments to Medicare Advantage Plans—Updated and Revised (November 2006). Brian Biles, Lauren Hersch Nicholas, Barbara S. Cooper, Emily Adrion, and Stuart Guterman. The National Committee for Quality Assurance’s The State of Health Care Quality 2006 (November 2006). Stephen C. Schoenbaum and Alyssa L. Holmgren. Why Not the Best? Results from a National Scorecard on U.S. Health System Performance (September 2006). The Commonwealth Fund Commission on a High Performance Health System. Assessing Medicare Prescription Drug Plans in Four States: Balancing Cost and Access (August 2006). Erika Heaton, Tanisha Carino, and Heidi Dix. Medicare Physician Payment: Are We Getting What We Pay For? Are We Paying for What We Want? (July 25, 2006). Stuart Guterman. Invited testimony before the U.S. House of Representatives, Energy and Commerce Committee, Subcommittee on Health. Health Care Spending and Use of Information Technology in OECD Countries (May/June 2006). Gerard F. Anderson, Bianca K. Frogner, Roger A. Johns, and Uwe E. Reinhardt. Health Affairs, vol. 25, no. 3 (In the Literature summary). Medicare Beneficiary Out-of-Pocket Costs: Are Medicare Advantage Plans a Better Deal? (May 2006). Brian Biles, Lauren Hersch Nicholas, and Stuart Guterman. Medicare at Forty (Winter 2005/2006). Karen Davis and Sara R. Collins. Health Care Financing Review, vol. 27, no. 2 (In the Literature summary). Medicare Extra: A Comprehensive Benefit Option for Medicare Beneficiaries (October 4, 2005). Karen Davis, Marilyn Moon, Barbara S. Cooper, and Cathy Schoen. Health Affairs Web Exclusive (In the Literature summary). Will You Still Need Me? The Health and Financial Security of Older Americans (June 2005). Sara R. Collins, Karen Davis, Cathy Schoen, Michelle M. Doty, Sabrina K. H. How, and Alyssa L. Holmgren. Prescription Drug Coverage and Seniors: Findings from a 2003 National Survey (April 19, 2005). Dana Gelb Safran, Tricia Neuman, Cathy Schoen et al. Health Affairs Web Exclusive (In the Literature summary). 31