MEDICARE PHYSICIAN PAYMENT: ARE WE GETTING WHAT WE PAY FOR? ARE WE PAYING FOR WHAT WE WANT? Stuart Guterman Senior Program Officer, Program on Medicare’s Future The Commonwealth Fund sxg@cmwf.org www.cmwf.org Invited testimony Energy and Commerce Committee Subcommittee on Health U.S. House of Representatives Hearing on “Medicare Physician Payment: How to Build a Payment System That Provides Quality, Efficient Care for Medicare Beneficiaries” July 25, 2006 I would like to thank Michelle Serber for her assistance in assembling the information contained in this testimony and preparing the testimony itself, and Karen Davis, Stephen Schoenbaum, and Cathy Schoen for their helpful comments and suggestions. The views presented here are those of the author and not necessarily those of The Commonwealth Fund or its directors, officers, or staff. This and other Fund publications are available online at www.cmwf.org. To learn more about new publications as they become available, visit the Fund’s Web site and register to receive e-mail alerts. Commonwealth Fund pub. no. 942. MEDICARE PHYSICIAN PAYMENT: ARE WE GETTING WHAT WE PAY FOR? ARE WE PAYING FOR WHAT WE WANT? Summary of Major Points Congress faces a challenging dilemma in considering how much to pay physicians, arising from the fact that the Sustainable Growth Rate (SGR) mechanism offers no control over the volume and intensity provided by the individual physician. Increasing physician payments would put more burden on Medicare beneficiaries—especially the most vulnerable ones—by raising the Part B premium. While it may be necessary to raise fees in the future to protect beneficiaries’ access to care, that does not seem to be a problem at present. Regardless of what we pay physicians, we must pay more attention to what we get for our money—quality and coordination of care are lacking, both absolutely and in comparison with other countries. Current pay-for-performance initiatives show promise for improving quality, but system designs and best practices for implementation will require careful thought and analysis. Both cost and quality must be considered, together rather than separately. Efficiency improvements should be encouraged and rewarded. To encourage better performance and coordination care, cost and quality should be evaluated on a broader basis than individual services or providers. Potential improvements in payment policy should be evaluated for their long-term impact and not necessarily discarded based on short-term resource requirements or lack of immediate impact. Other tools, such as information collection and dissemination, can help improve performance by securing better cooperation and coordination among providers. Supporting providers through entities like Medicare’s Quality Improvement Organizations can also enhance their ability to improve. iii In addition to serving an important role in providing access to care for aged and disabled beneficiaries, Medicare can be a useful and important platform for developing and implementing performance improvements in the health care system. Sufficient resources should be devoted to research on best practices, development and application of quality standards, and the development of other knowledge and tools to improve the performance of the health care system for Medicare beneficiaries and all Americans. iv MEDICARE PHYSICIAN PAYMENT: ARE WE GETTING WHAT WE PAY FOR? ARE WE PAYING FOR WHAT WE WANT? Thank you, Chairman Deal, Congressman Brown, and Members of the Committee, for this invitation to testify on Medicare physician payment. I am Stuart Guterman, senior program director for the Program on Medicare’s Future at The Commonwealth Fund. The Commonwealth Fund is a private foundation that aims to promote a high performing health care system that achieves better access, improved quality, and greater efficiency, particularly for society’s most vulnerable populations, including low-income people, the uninsured, minority Americans, young children, and elderly adults. The Fund carries out this mandate by supporting independent research on health care issues and making grants to improve health care practice and policy. Congress faces a challenging dilemma in considering how much to pay physicians. On the one hand, Medicare spending is rising at a rate that threatens the program’s continued ability to fulfill its mission; on the other, the sustainable growth rate (SGR) mechanism, which is intended to address that problem, produces annual reductions in physician fees that are equally difficult to accept. There is an underlying mismatch between the primary cause of rising spending—the volume and intensity of services provided by physicians—and the focus of the SGR, which is to set the fees that physicians receive for each service provided. Because the SGR offers no control over the volume and intensity provided by individual physicians—and, in fact, may create an incentive to increase volume and intensity to offset reductions in fees—it does not address the underlying cause of physician spending growth. Determining how much to pay physicians certainly is an important issue, but determining how to pay physicians so Medicare beneficiaries get the best care possible is of at least equal importance. While the payment amount may have an effect on beneficiaries’ access to physician services, the payment mechanism (as well as other tools) can be used to ensure quality and appropriateness of medical care is maximized, so that beneficiaries’ health status is enhanced and the Medicare program gets the most for its money. There is evidence that improved quality and reduced cost may both be achievable: We can have our cake and eat it, too. In this testimony, I will first discuss Medicare physician payment and some issues related to the SGR mechanism and the problems it fails to address. I then will discuss the 1 imperative for Medicare to become a better purchaser of health care, rather than remaining a payer for health services, and suggest some areas on which initiatives should focus. Finally, I will briefly discuss some promising initiatives currently underway, and offer opinions as to how they might be used to improve the Medicare program and the health care system in general. Why Physicians Are Different Than Medicare’s Other Service Providers Physicians are unique among Medicare providers in being subject to an aggregate spending adjustment. In contrast, most Medicare services are paid through prospective payment systems that set a price for a bundle of services. In these systems, the provider is free to make decisions about the volume of services provided to the patient, while the payment for the bundle remains fixed. Physicians are unique in their role in determining the volume of services they can provide. Physicians are the gatekeepers and managers of the health care system; they direct and influence the type and amount of care their patients receive. Physicians, for example, can order laboratory tests, radiological procedures, and surgery. Moreover, physicians under Medicare are frequently paid for very small units of service. Physicians may receive payment for office visits and separate payments for individual services, such as administering tests and interpreting x-rays—all of which can be provided in a single visit. Alternatively, hospitals receive payment for each discharge, with no extra payment for additional services or days, except in extremely costly cases. Further, once a physician’s practice is established, the marginal costs of providing additional services are primarily those associated with the physician’s time. That means cost estimates for physician’s services are extremely malleable because they are largely dependent on how the physician’s time is valued. Even then, there is no routinely available and auditable source of data on costs for individual physicians or practices, such as there is for hospitals via the Medicare Cost Report. Attempts to Control Spending by Adjusting for Volume In an attempt to control total spending for physicians’ services driven by volume, in the Omnibus Budget Reconciliation Act of 1989, Congress established a mechanism that set physician fees for each service and tied the annual update of those fees to the trend in total spending for physicians’ services relative to a target. Under that approach, physician fees were to be updated annually to reflect increases in physicians’ costs for providing care and adjusted by a factor that reflected the volume of services provided per 2 beneficiary. The introduction of expenditure targets to the update formula in 1992 initiated a new approach to physician payments. Known as the volume performance standard (VPS), this approach provided a mechanism for adjusting fees to try to keep total physician spending on target. The method for applying the VPS was fairly straightforward, but led to updates that were unstable. Under the VPS approach, the expenditure target was based on the historical trend in volume. Any excess spending relative to the target triggered a reduction in the update two years later. But the VPS system depended heavily on the historical volume trend, and the decline in that trend in the mid-1990s led to large increases in Medicare’s fees for physicians’ services. Congress attempted to offset the budgetary effects of those increases by making successively larger cuts in fees, which further destabilized the update mechanism. That volatility led the Congress to modify the VPS in the Balanced Budget Act of 1997, replacing it with the sustainable growth rate mechanism in place today. Like the VPS, the SGR method uses a target to adjust future payment rates and to control growth in Medicare’s total expenditures for physicians’ services. In contrast to the VPS, however, the target under the SGR mechanism is tied to growth in real, inflation-adjusted gross domestic product (GDP) per capita—a measure of growth in the resources per person that society has available. Moreover, unlike the VPS, the SGR adjusts physician payments by a factor that reflects cumulative spending relative to the target. Compared with the VPS, policymakers saw the SGR approach as having the advantages of objectivity and stability, From a budgetary standpoint, the SGR method, like the VPS, is effective in limiting total payments to physicians over time. GDP growth provides an objective benchmark, and changes in GDP from year to year have been considerably more stable and generally smaller than changes in the volume of physicians’ services. Problems with the Current Approach A key argument for switching from the VPS approach to the SGR mechanism was that, over time, the VPS would produce inherently volatile updates. But, updates under the SGR method have proven to be volatile as well. Through 2001, that volatility was to the benefit of physicians—overall, the increase in fees in the first three years of SGR was more than 70 percent higher than the Medicare Economic Index over the same period. 3 Since then, the pattern since then has been considerably different. In 2002, Medicare physician fees declined for the first time, by 3.8 percent (Figure 1). Notably, however, physician expenditures per beneficiary increased—although at the lowest rate in four years. In succeeding years, Congress has wrestled with a succession of negative updates produced by the SGR formula. Under the Medicare Modernization Act, physician fees were frozen for two years beginning in 2004—actually an increase relative to the reductions called for by the SGR formula—but physician expenditures per beneficiary continued to rise. In fact, while physician fees fell between 2001 and 2005, physician expenditures per beneficiary rose at the same rate as in the previous four years (Figure 2). Impact on Beneficiaries Decisions about how much to pay physicians under Medicare affect the program’s beneficiaries in two ways. First, rising spending for physicians’ services mean higher Part B premiums, exacerbating the financial burden of beneficiaries, particularly among the more vulnerable groups with low incomes, fragile health, disabilities, or chronic illnesses. Second, rates that are too low may affect access to needed physician care, either because physicians will refuse to treat new Medicare patients or stop treating Medicare patients or because they will refuse to take the Medicare payment rates as payment in full, meaning beneficiaries could be responsible for some additional payments. Medicare Part B, which covers physician, outpatient hospital services, and other ambulatory services, is voluntary, although beneficiaries are automatically enrolled unless they indicates a desire to opt out. Enrollment requires payment of a monthly premium (generally deducted from the beneficiary’s Social Security check), which currently amounts to $88.50, or almost 9 percent of the average Social Security check.1 Because the premium is set so that it covers 25 percent of projected Part B costs, every increase in physician payments has a proportional effect on the Part B premium. In 2006, the Part B premium increased by more than 10 percent for the third consecutive year, causing concern among beneficiaries and their advocates.2 Overall, the Part B premium has increased from $43.80 in 1998 to $88.50 in 2006—an annual rate of more than nine percent (Figure 3). By 2015, it is projected to rise to $122.40, climbing at 1 Social Security Administration, OASDI Monthly Statistics, June 2006, accessed through http://www.ssa.gov/policy/docs/statcomps/oasdi_monthly. 2 See W. Novelli, Statement by AARP CEO Bill Novelli on the 2006 Social Security COLA, accessed through http://www.aarp.org/research/press-center/presscurrentnews/2006_social_security_cola.html. 4 a much slower rate than in the past few years, but still almost 40 percent higher than its current level.3 These increases can put burdensome financial pressure on Medicare beneficiaries, particularly those who are most vulnerable because of low incomes or other economic or health-related factors. Medicare beneficiaries tend to be particularly vulnerable to the financial pressures of health care costs. Seventy-eight percent of the Medicare aged are in fair or poor health or have a chronic condition or disability, compared with 31 percent of the population under 65 with employer coverage; and 46 percent have incomes below 200 percent of the federal poverty level, compared with 21 percent of the younger population with employer coverage (Figure 4). In fact, these twin problems of low income and poor health—two-thirds of beneficiaries have one or the other—are the major reason Medicare was enacted in the first place. Even typical aged beneficiaries had out-of-pocket costs that were more than 20 percent of their incomes, on average (Figure 5). That burden was projected to rise to almost 30 percent by 2025—although it may be somewhat reduced by the availability of prescription drug coverage under Medicare Part D. Beneficiaries with physical or cognitive health problems and no other health insurance were paying 44 percent of their incomes on average for their health care costs out-of-pocket, with that burden projected to grow to more than 60 percent by 2025. Although the availability of Medicare Part D may reduce that number somewhat, beneficiaries are clearly in a precarious position. So far, access to physicians does not seem to be a problem. Telephone surveys conducted for the Medicare Payment Advisory Commission (MedPAC) indicate that 74 percent of beneficiaries never had a delay in getting an appointment for routine care, and 83 percent never had a delay getting an appointment in cases of illness or injury (Figure 6). These percentages were about the same as in the previous two years—and somewhat higher than for people who were privately insured. Similarly, the vast majority of beneficiaries reported no problems finding a new physician—either primary care or specialist—with the numbers being about the same across years and sources of insurance coverage. MedPAC also reports that, although Medicare physician payments overall amount to only 83 percent of the rates paid by private insurers in 2004, that ratio has been fairly 3 Beginning in 2007, the premium will be higher for beneficiaries with incomes above a certain threshold. 5 stable over the past five years and, if anything, has increased slightly.4 Moreover, 99 percent of allowed charges for physician services were assigned in 2002, which means that essentially all physicians accept the Medicare payment rates as full payment for services.5 Nonetheless, given the cuts scheduled in every year from 2007 through 2011, MedPAC concludes that: “We are concerned that such consecutive annual cuts would threaten access to physician services over time, particularly primary care services.”6 In addition, it states: “The Commission considers the SGR formula a flawed, inequitable mechanism for volume control and plans to examine alternative approaches to it in the coming year.”7 Congress must evaluate these alternatives in light of three potentially conflicting requirements: to control the growth of Medicare spending, to provide a fair rate of payment to physicians and preserve access for Medicare beneficiaries, and to keep the financial burden on the most vulnerable beneficiaries from becoming worse. What Are We Getting for Our Money? Regardless of how physicians are paid under Medicare, there is a basic issue that must be addressed for the good of the Medicare program, its beneficiaries, and the rest of the health care system. Adult patients in the U.S. receive only 55 percent of recommended care overall, with even lower proportions for patients with some conditions—such as hip fracture, with only 23 percent (Figure 7). This poor performance is particularly striking given the fact that the United States devotes 16 percent of its GDP to health services—by far, the highest in the world.8,9 Not surprisingly, the poor performance of the health care sector has implications for Medicare. Life expectancy at age 65 in the U.S. is among the lowest among industrialized countries (Figure 8). This poor performance is the product of many specific aspects of the way health care is structured and provided in the U.S. These include the complexity and 4 Medicare Payment Advisory Commission, Report to the Congress: Medicare Payment Policy (Washington, D.C.: MedPAC, Mar. 2006), p. 91. 5 Ibid., p. 90. 6 Ibid., p. 99. 7 Ibid., p. 99. 8 S. Heffler, S. Smith, S. Keehan et al., “U.S. Health Spending Projections for 2004–2014,” Health Affairs Web Exclusive (Feb. 23, 2005):W5-74–W5-85. 9 G. F. Anderson, B. K. Frogner, R. A. Johns et al., “Health Care Spending and Use of Information Technology in OECD Countries,” Health Affairs, May/June 2006 25(3):819–31. 6 fragmentation of our health care system, specialization of physicians, intensive use of medications, and poor coordination of care. The Commonwealth Fund has found that 34 percent of U.S. patients surveyed in 2005 reported a medical mistake, medication error, or test error in the past two years, compared with 22 percent in the United Kingdom, which has the lowest rate among the survey countries, and 30 percent in Canada, which has the next highest rate (Figure 9). Interpersonal aspects of health care also are lacking. Thirty-five percent of community-dwelling adults age 65 and older reported that health providers did not always listen carefully to them and 41 percent reported that health providers did not always explain things clearly (Figure 10). In addition, 31 percent of sicker adults in the U.S. surveyed in 2002 reported that they had left a doctor’s office in the past two years without getting important questions answered, compared with 19 percent in the U.K. (Figure 11). Coordination is an important dimension of health care delivery, with a rising proportion of the population—especially seniors—having multiple chronic conditions and multiple doctors. More than 20 percent of Medicare beneficiaries have five or more chronic conditions, and they are treated by an average of almost 14 different doctors in a given year.10 In our current payment system, there is nothing to encourage physicians to communicate with each other about patients they have in common. Although there have been some efforts to change this, fee-for-service Medicare is still largely based on the acute care model, in which a patient becomes ill and is treated by a doctor in the office or hospital until the discrete episode is over and the patient can resume normal life.11 Moreover, until recently, there were substantial barriers to the appropriate coordination of care even in the Medicare+Choice program.12 Difficulties in care coordination are evident around the world, but not as much as in the U.S. Thirty-three percent of adults with health problems reported that in the past two years a doctor had ordered tests that had already been done or that test results or records were not available to their doctor at the time of their appointment, compared with 19 percent in the U.K. and Australia, which boasted the lowest proportions, and 26 percent in Germany, which had the next highest rate after the U.S. (Figure 12). Although 10 Partnership for Solutions, “Medicare: Cost and Prevalence of Chronic Conditions,” Fact Sheet, July 2002. 11 See 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 Ibid. 7 most U.S. adults with health problems reported having a regular doctor (84%), only half of had been with that doctor for five years or more (Figure 13). The number of doctors treating a patient, not surprisingly, is correlated with coordination problems. In the U.S., 22 percent of patients with one doctor had experienced at least one of these problems, while 43 percent of patients with four or more doctors had experienced them—almost twice as many (Figure 14). This pattern held in all the countries in which the survey was conducted. Addressing the lack of care coordination in the U.S. is not just a quality issue—as mentioned previously, about 20 percent of Medicare beneficiaries have five or more chronic conditions, but this group also accounts for two-thirds of Medicare spending each year (Figure 15). About $300 billion is going to be spent for this group of people in the next year, and it could be spent much more productively than it is now.13 The Role of Health Information Technology Health information technology is frequently cited as a tool for improving both the quality and coordination of care. It is also widely recognized that its diffusion of health across the health care sector has been much slower than desired. Researchers at RAND found that only 20 to 25 percent of hospitals—and only 15 to 20 percent of physicians’ offices—across the country have adopted electronic medical records (EMRs).14 The use of electronic technology in physicians’ offices is fairly common, but applications may fall short of the comprehensive, quality-enhancing EMR that proponents of health information technology envision. In a 2003 survey of physicians, The Commonwealth Fund found that almost 80 percent of all physicians used electronic billing in their offices, and almost 60 percent used health information technology to access to test results (Figure 16). Only 27 percent, however, used the technology for electronic ordering and about the same proportion had EMRS. In most instances, larger practices make more use of health information technology. In 2004 and 2005, The Commonwealth Fund supported a study of solo and small group practices, to investigate the business case for technology adoption in those settings. It found that the costs of adopting, installing, and using electronic health records could be substantial (Figure 17). In addition to the initial costs, which averaged almost $44,000 per provider, there were ongoing costs of almost $8,500 per provider per year. 13 See Board of Trustees, Federal HI and Federal SMI Trust Funds, 2006 Annual Report. 14 See K. Fonkych and R. Taylor, The State and Pattern of Health Information Technology Adoption (Santa Monica, Calif.: RAND, 2005). 8 There were also substantial financial benefits, and the average practice recouped its costs in about two-and-a-half years. It is important to note that the financial benefits of adoption, which averaged about $33,000 per provider per year, came from two main sources: increased efficiency, which accounted for almost $16,000 per provider per year; and increased coding levels, which accounted for almost $17,000 per provider per year (Figure 18).Of the 14 practices in the study, only two reported any quality performance rewards and those were nominal. Some quality improvement activities were implemented at almost all of the practices, but varied in focus and intensity.15 Can We Get More for What We Spend? The Dartmouth Atlas has produced a chart that is by now well-known, which shows the wide variation in Medicare spending per beneficiary among different areas in the U.S. (Figure 19). In 1996, the 20 percent of areas with the highest spending were about 60 percent higher than their counterparts at the low end; by 2000, that ratio had not changed much, and it is the same as of 2003. In fact, these numbers conceal the tremendous amount of variation in spending across individual regions: in 2003, spending in Miami, Fla.—the area with the highest Medicare spending per beneficiary—was more than two and a half times that in Salem, Ore.—the area with the lowest spending. Similar variation in spending was found in data recently analyzed by The Commonwealth Fund on Medicare spending for beneficiaries with all three of the following conditions: diabetes, chronic obstructive pulmonary disease, and congestive heart failure. Using the same area definitions used by the Dartmouth Atlas, the Fund found that median spending per patient across all areas was almost $30,000, but the variation ranged from less than $15,000 to almost $80,000 (Figure 20). Those costs were compared to a composite quality measure relevant to the three study conditions, and found no obvious correlation between cost and quality across areas. Some areas with high-quality scores had low costs, while some had high costs; in addition, some areas with high costs had lower-than-average quality scores. While the quality measures represented in the previous figure are process measures—that is, measures that represent what doctors do—the same relationship appears to hold between spending and outcomes. Data from the Dartmouth Atlas show that Medicare beneficiaries in states with higher Medicare spending per beneficiary do 15 See R. Miller, C. West, T. M. Brown et al., “The Value of Electronic Health Records in Solo or Small Group Practices,” Health Affairs, Sept./Oct. 2005 24(5):1127–37. 9 not appear to have lower overall mortality rates than in states with lower spending (Figure 21). These data are aggregated at the area level, but the decisions that determine both cost and quality are made by individual provider. They should not be taken as an indication that costs can be easily be reduced at an aggregate level without harming quality or access to care, or that quality improvements that save money in the long term may not cost more in the short term. But they do indicate that there appear to be patterns in how health care decisions are made that are not necessarily driven by factors that improve quality, and that we should be able to figure out how to use our resources more effectively to provide higher quality care at the same or even lower costs than we currently face. Do Efforts to Improve Quality Work? One of the underlying problems with our health care financing mechanism is that providers are paid for providing more care and more intensive care, but not necessarily better care. This problem is particularly evident in the way that Medicare pays physicians—in fact, it is the real issue that confronts the Congress in discussing how to “fix” the SGR. What we should discuss, instead, is how to restructure the payment system to get what we want for the tremendous amount of money that we spend. Over the past several years, both private and public payers, purchasers, and providers have been developing efforts to address this problem. The Leapfrog Group Incentive & Reward Compendium lists 97 programs around the country aimed at providing financial incentives to improve quality.16 Several are already beginning to produce results, and they indicate that there is some promise to this approach. In a pay-for-performance program run by the Integrated Healthcare Association in California—involving about 35,000 physicians in more than 200 physician organizations—participants reported that they screened about 60,000 more women for cervical cancer, tested nearly 12,000 more individuals for diabetes, and administered about 30,000 more childhood immunizations in 2005 than they had in 2004.17 Earlier findings indicated that the use of information technology in various clinical applications had also increased substantially under the initiative (Figure 22). 16 See The Leapfrog Group Compendium at http://ir.leapfroggroup.org/compendium/. 17 IHA News Release, “Continued Quality Improvement in California Healthcare Announced by Integrated Healthcare Association” July 13, 2006. 10 In an analysis of a natural experiment in pay-for-performance, PacifiCare Health Systems paid its medical groups in California bonuses according to performance on a set of quality measures. Performance on cervical cancer screening improved significantly (Figure 23). There was no significant increase, however, in mammography screening or hemoglobin A1c testing.18 The National Committee for Quality Assurance, with the American Diabetes Association, has developed a Diabetes Physician Recognition Program that awards recognition to physicians who demonstrate they provide high-quality care to patients with diabetes.19 Although no financial incentive is provided under this program—in fact, there is a fee to participate—there have been several areas of improvement. The proportion of patients with hemoglobin A1c counts below 7 percent rose from 25 to 46 percent between 1997 and 2003 and the proportion of patients with low-density Lipoprotein cholesterol levels below 100 milligrams per deciliter rose from 17 to 45 percent (Figure 24). Can We Get Better Care at Lower Cost? All the pay-for-performance initiatives described have focused primarily on quality improvement. In addition, some ongoing initiatives are producing data that indicate it is possible to achieve a higher level of quality at a lower cost. The Hospital Quality Incentive demonstration is being conducted by the Centers for Medicare and Medicaid Services (CMS) with Premier, Inc., at about 255 hospitals. Under this demonstration, hospitals are awarded bonus payments based on their performance on discharges in each of five clinical conditions, based on a total of 34 measures. In the first year, a total of almost $9 million in bonuses was paid, and quality improved in each of the five performance domains.20 Premier, Inc. also found that better performance along several dimensions related to efficiency correlated with better performance on quality. For example, the readmission rates for pneumonia were 25 percent lower for the 10 percent of the hospitals in the top quality group than for the hospitals in the bottom quartile (Figure 25). A study sponsored by The Commonwealth Fund has found that coordination across sites of care was correlated with factors that could indicate more appropriate use of health care providers. Upon leaving the hospital, patients who said they had a good 18 M. B. Rosenthal, R. G. Frank, Z. Li et al., “Early Experience with Pay-for-Performance: From Concept to Practice,” Journal of the American Medical Association, Oct. 12, 2005 294(14):1788–93. 19 See http://www.ncqa.org/dprp. 20 Centers for Medicare & Medicaid Services Press Release, “Medicare Demonstration Shows Hospital Quality of Care Improves with Payments Tied to Quality” Nov. 14, 2005. 11 understanding of what they were responsible for in terms of managing their health had significantly lower rates of subsequent emergency department use and hospital readmissions (Figure 26). In another study sponsored by The Commonwealth Fund, the application of advanced practice nurse care for congestive heart failure patients reduced the total cost per patient from $9,618 to $6,152 (Figure 27). It is important to note that this decrease was composed of a 45 percent increase in the cost of ambulatory care and a 44 percent decrease in the cost of inpatient care—because inpatient care is much more expensive, the decrease in inpatient costs more than offset the increase in ambulatory care costs. However, with our current fragmented health care financing and delivery systems, it is difficult to implement programs that shift resources across providers, even if they could both improve the quality of care and save money overall. Challenges in Aligning Financial Incentives with Better Performance Although pay-for-performance mechanisms may be promising in encouraging improved health care, careful attention must be paid to the design of the payment systems intended to elicit these improvements. Systems designed with the best of intentions can have unintended consequences. For example, in the previously mentioned evaluation of the PacifiCare pay-for-performance initiative in California, although cervical cancer screening rates improved, the greatest improvement was among the doctors who initially were in the lowest-performing group (Figure 28). However, the vast majority of the bonus money went to the doctors initially in the highest performing group who showed the smallest improvement. As MedPAC has recommended, a balance must be struck between rewarding the level of performance and improvement in performance.21 It should also be noted that, despite the scores of pay-for-performance initiatives implemented, most physicians have not been involved in any sort of collaborative effort to improve quality of care (Figure 29). Although these data are several years old, they probably are not very different from the current situation. Perhaps these results are not surprising, given the small number of physicians who are financially affected by quality considerations—only 19 percent of physicians surveyed in 2003 indicated that quality bonuses or incentive payments were a major factor affecting their compensation (Figure 30). These data indicate that the involvement of Medicare on a nationwide basis is needed to draw physicians into coordinated efforts to improve quality and efficiency. 21 Medicare Payment Advisory Commission, Report to the Congress: Medicare Payment Policy (Washington, D.C.: MedPAC, Mar. 2005). 12 Conclusions As Congress considers Medicare physician payments for the remainder of this session and beyond, several points must be kept in mind. First, the current SGR mechanism for updating physician fees does not work—it produced inappropriately large increases in fees in its early years and untenable reductions for the past several years and foreseeable future. Because the updates produced by the SGR formula are incorporated in the budget baseline, which is used to “score” the budgetary effects of new legislation, even freezing physician fees for the next 10 years would be “scored” as “costing” the Medicare program billions of dollars, making it difficult for the Congress to appropriately address the problem without appearing to exacerbate the federal deficit. Moreover, it does not appear that the current mechanism has been effective in controlling the growth in Medicare spending, which is produced primarily by increased volume and intensity, rather than fees. Second, the Medicare program is more than a line item in the federal budget or a source of income for providers—it is one of the most popular social programs in history that provides access to care for 43 million aged and disabled beneficiaries, who tend to be sicker and poorer than other Americans. As Congress considers changes to Medicare physician payment, it must weigh the effects of those changes on the Part B premium that beneficiaries must pay. Increases in physician payments proportionately raise the premium and put more financial pressure on the most vulnerable groups of beneficiaries. At the same time, the sharp cuts in fees projected for the next several years are a potential threat to beneficiaries’ access to care. These issues must be put in the context of a health care system that has the highest costs in the world, but fails to yield commensurate results in terms of the quality and appropriateness of care it provides. This failure cannot—and should not—be tolerated any longer. Fragmentation, lack of communication among physicians and between physicians and patients, medical errors and duplication of tests and other services, and the absence of a mechanism that encourages—or even allows—care coordination across sites of care are attributes of a health care system that is not a system at all. There are many efforts in both the private and public sectors aimed at addressing at least some of these problems. Many are still in their early stages, but early evidence indicates the promise of some success. Both CMS and Congress have expressed the desire to move toward pay-for-performance in Medicare, starting with hospitals and physicians, as well as nursing homes. Efforts should ensure the systems put in place are 13 appropriate and will encourage broadly improved care, and not just focus on narrow activities to meet specific quality goals. Progress is being enhanced by several CMS demonstration and pilot projects currently in operation, such as the Hospital Quality Incentive demonstration mentioned earlier, the Physician Group Practice demonstration, and the Medicare Health Support pilot, as well as several that are being developed, such as the Medicare Care Management Performance demonstration, the Nursing Home Quality-Based Purchasing demonstration, the Medicare Hospital Gain-Sharing demonstration—and particularly the Medicare Health Care Quality demonstration, which will test different approaches to broader system redesign. Resources must be made available for continued efforts to develop appropriate measures of quality and the means to apply them. Some improvements may require high initial costs. This is particularly problematic in Medicare, where demonstration projects intended to produce higher quality must meet a budget-neutrality requirement that may be applied so strictly as to hinder the development of some potentially beneficial projects. To be sure, the projected spending impact of proposed demonstration projects is extremely important, but that issue needs to be considered more broadly. An especially difficult situation that needs to be addressed is accounting for the overall effects on Medicare and Medicaid—rather than the effects on each of the two programs separately—of projects that might enhance the quality—and overall efficiency—of care provided to the almost eight million beneficiaries who are eligible for both programs. Pay-for-performance must also be considered in the context of other tools available to improve quality and efficiency. The primary objective of paying for performance should not be merely to reward good providers and punish bad ones, but to align the health care financing mechanism with good results. Prices are messages to producers. Currently, the message we are sending to health care providers is that we want more services and procedures, but we do not care very much about how well those services are provided or how much they help patients achieve better health. There are several additional tools that can be used to achieve the desired objectives—each should be pursued to get where we want to be: • Public information on quality and cost should be made available in a format that can be understood by patients and their advocates and acted upon by providers. Patients with particular medical needs should be able to identify providers that are best able to give them appropriate and efficient care and providers should be able 14 to use that information to improve their quality and efficiency. Public reporting has been shown o be an effective tool in spurring quality improvement efforts.22 • We must find ways to encourage more productive and beneficial interaction between patients and providers. In addition to rewarding physicians for producing units of care in an effective and efficient way, they must be encouraged to provide that care in a way that is effective and efficient in a broader sense. Examples of these types of incentives would be payments to specific providers for serving as the patient’s “medical home”—that is, taking responsibility for obtaining and coordinating all the care needed by the patient across settings, including at home. Other ways to provide more coordination of care across sites—such as follow-up by hospitals for patients discharged with on-going conditions—should be developed. • Making extra payment available for achieving certain quality and efficiency goals helps to align the incentives of the financing and delivery systems, but some providers may face other barriers to achieving the goals established for them. Additional resources must be available to establish an infrastructure that enables providers to improve their performance. Medicare’s Quality Improvement Organizations currently are tasked with that function, but relatively little is known about its priority in their list of requirements and their effectiveness in fulfilling that role. All of these approaches hold promise in improving provider performance, not only for Medicare but for all patients. Finally, payment reform to reward excellence and efficiency would be greatly facilitated by a major enhancement of health services research funding that includes research on best practices, performance of different forms of health care delivery organization, diffusion of innovation, quality standards, evidence-based medicine, cost- effectiveness and comparative effectiveness, and the development and application of quality standards. This would require some effort and perhaps a substantial amount of resources, but it is the only way to avoid the seemingly endless spiral of spending that we face and improve the value of what we spend. 22 See J. H. Hibbard, J. Stockard, and M. Tusler, “Does Publicizing Hospital Performance Stimulate Quality Improvement Efforts?” Health Affairs, Mar./Apr. 2003 22(2):84–94. 15 Figure 1. Annual Increases in Physician Fees and SGR-Related Expenditures Per Fee-for-Service Beneficiary, SGR- Fee-for- 1998-2005 1998- Fees SGR-related expenditures per fee-for-service beneficiary 12 10.8 10.0 10 9.3 7.7 8.2 Percent change 8 5.9 6 4.9 4.5 3.8 4.0 4 2.2 2.1 1.4 2 0.1 0 -2 -0.6 -4 -3.8 -6 1998 1999 2000 2001 2002 2003 2004 2005 Year THE COMMONWEALTH FUND Source: Letter to Glenn M. Hackbarth, Chair, Medicare Payment Advisory Commission, from Herb B. Kuhn, Director, Center for Medicare Management, Centers for Medicare and Medicaid Services, dated April 7, 2006. Figure 2. Annual Rates of Increase in Physician Fees and SGR-Related Expenditures Per Fee-for-Service Beneficiary, SGR- Fee-for- 1997-2001 and 2001-2005 1997 - 2001- Fees SGR-related expenditures per fee-for-service beneficiary 8 7.4 7.4 7 Annual Percent Change 6 5 4 3.4 3 2 1 0 -1 -0.7 -2 1997-2001 2001-2005 Period THE COMMONWEALTH FUND Source: Letter to Glenn M. Hackbarth, Chair, Medicare Payment Advisory Commission, from Herb B. Kuhn, Director, Center for Medicare Management, Centers for Medicare and Medicaid Services, dated April 7, 2006. 16 Figure 3. Medicare Part B Premium (Monthly), 1998-2006 (Actual) and 2007-2015 (Projected) 1998- 2007- Part B Premium 140 122.40 120 100 88.50 80 60 43.80 40 20 0 98 9 0 1 2 3 04 05 06 7 8 9 0 1 12 13 14 5 9 0 0 00 00 0 00 00 01 01 01 19 19 20 20 20 20 20 20 20 20 20 2 2 2 2 2 2 2 Year THE COMMONWEALTH FUND Source: Board of Trustees, Federal HI and Federal SMI Trust Funds, 2006 Annual Report. Figure 4. Profile of Medicare Elderly Beneficiaries and Employer Coverage Nonelderly, Nonelderly, by Poverty and Health Status, 2003 Health problems, No health problems, lower income higher income Health problems, 7% 15% lower income 38% No health problems, Health problems, lower income higher income 8% 24% No health problems, Health problems, lower income higher income No health problems, 14% 40% higher income 56% Medicare, Ages 65+ Employer, Ages 19–64 Note: Respondents with undesignated poverty were not included; lower income defined as <200% of poverty; health problems defined as fair or poor health, any chronic condition (cancer, diabetes, heart attack/disease, and arthritis), THE COMMONWEALTH or disability . FUND Source: The Commonwealth Fund Biennial Health Insurance Survey (2003). 17 Figure 5. Projected Out-of-Pocket Spending As a Share of Out-of- Income Among Groups of Medicare Beneficiaries, 2000 and 2005 Out-of-pocket spending as percent of income 80 2000 2025 71.8 63.3 60 51.6 44.0 41.1 40 29.9 29.1 21.7 20 8.9 7.8 0 Beneficiaries age 65+ Beneficiairies with Disabled beneficiaries Beneficiaries ages Female beneficiaries physicial or cognitive ages 45–65 65–74 with high age 85+ with physical health problems and incomes* or cognitive health no other health problems and low insurance incomes^ * Annual household incomes of $50,000 or more. ^ Annual household incomes of $5,000 to $20,000. THE Source: S. Maxwell, M. Moon, and M. Segal, Growth in Medicare and Out-of-Pocket Spending: Impact on Vulnerable COMMONWEALTH FUND Beneficiaries, The Commonwealth Fund, January 2001 as reported in R. Friedland and L. Summer, Demography Is Not Destiny, Revisited, The Commonwealth Fund, March 2005. Figure 6. Access to Physicians for Medicare Beneficiaries and Privately Insured People, 2005 Medicare Privately Insured Percent 100 89 86 90 83 74 75 75 75 80 67 70 60 50 40 30 20 10 0 Routine Care Illness or Primary care Specialist Injury Never had a delay to No problem finding appointment physician THE COMMONWEALTH FUND Source: MedPAC Report to the Congress: Medicare Payment Policy, March 2006, p. 85. 18 Figure 7. Proportion of Recommended Care Received by U.S. Adults, by Selected Conditions Percent of recommended care received 100 76 80 65 55 54 60 39 40 23 20 0 Overall Breast Cancer Hypertension Asthma Pneumonia Hip Fracture THE Source: McGlynn et al., “The Quality of Health Care Delivered to Adults in the United COMMONWEALTH FUND States,” The New England Journal of Medicine (June 26, 2003): 2635–2645. Figure 8. Life Expectancy at Age 65 Female Male 25 23.0 21.3 21.0 20.6 19.6 20.0 19.6 19.5 19.1 20 18.0 16.9 17.6 17.2 16.7 16.1 16.0 16.6 16.1 15 10 5 0 Ja Fr A C O G U U N us an E er ni ni ew an pa C te te m tr ad D ce n Z al an d d a M ea (2 ia S K (2 y ed (2 ta 00 in la 0 (2 (2 0 ia gd t 01 nd 3 02 es 00 00 ) n om ) ) (2 1 3) (2 (2 ) 0 00 00 (2 02 2 00 2) ) ) )2 THE COMMONWEALTH FUND Source: OECD Health Data, 2005. 19 Figure 9. Patient Reported Medical Mistake, Medication Error, or Test Error in Past 2 Years Percent 50 34 30 27 25 23 22 25 0 AUS CAN GER NZ UK US THE COMMONWEALTH FUND Source: 2005 Commonwealth Fund International Health Policy Survey. Figure 10. Interpersonal Quality of Care Is Lacking Percent of community-dwelling adults in 2001 who visited doctor's office in past year 100 Ages 45–64 Age 65+ 80 65 66 56 59 59 59 60 54 46 40 20 0 Health providers Health providers Health providers Health providers always listened always explained always showed always spent carefully things clearly respect enough time Source: S. Leatherman and D. McCarthy, Quality of Health Care for Medicare Beneficiaries: A THE Chartbook, 2005, The Commonwealth Fund. www.cmwf.org; Medical Expenditure Panel Survey COMMONWEALTH FUND (AHRQ 2005). 20 Figure 11. Communication With Physicians Views of Sicker Adults In the past 2 years: AUS CAN NZ UK US Left a doctor’s office without 21 25 20 19 31 getting important questions answered Did not follow a 31 31 27 21 39 doctor’s advice THE COMMONWEALTH FUND Source: 2002 Commonwealth Fund International Health Policy Survey. Figure 12. Deficiencies in Care Coordination Percent saying in the AUS CAN GER NZ UK US past 2 years: Test results or records not available at time of 12 19 11 16 16 23 appointment Duplicate tests: doctor ordered test that had 11 10 20 9 6 18 already been done Percent who experienced either 19 24 26 21 19 33 coordination problem THE COMMONWEALTH FUND Source: 2005 Commonwealth Fund International Health Policy Survey. Adults with Health Problems. 21 Figure 13. Continuity of Care with Same Physician Percent: AUS CAN GER NZ UK US Has regular doctor 92 92 97 94 96 84 --5 years or more 56 60 76 57 66 42 No regular doctor 8 8 3 6 4 16 THE COMMONWEALTH FUND Source: 2005 Commonwealth Fund International Health Policy Survey. Adults with Health Problems. Figure 14. Coordination Problems by Number of Doctors Percent 75 1 Doctor 4 or more Doctors 50 43 31 30 30 27 26 22 23 25 16 15 11 7 0 AUS CAN NZ UK US GER *Either records/results did not reach doctors office in time for appointment or doctors ordered a duplicate medical test THE Source: C. Schoen et al., “Taking the Pulse: Experiences of Patients with Health Problems in Six COMMONWEALTH FUND Countries,” Health Affairs Web Exclusive (November 3, 2005). Based on the 2005 Commonwealth Fund International Health Policy Survey. 22 Figure 15. Two-Thirds of Medicare Spending is for People With Five or More Chronic Conditions No chronic conditions 1% 1-2 chronic conditions 10% 5+ chronic conditions 3 chronic 66% conditions 10% 4 chronic conditions 13% THE COMMONWEALTH Source: G. Anderson and J. Horvath, Chronic Conditions: Making the Case for Ongoing Care (Baltimore, FUND MD: Partnership for Solutions, December 2002) Figure 16. Physician Use of Electronic Technology Varies by Application Percent indicating "routine” or “occasional" use All Physicians 1 Physician 87% 85% 84% 2-9 Physicians 79% 77% 10-49 Physicians 68% 66% 59% 61% 50+ Physicians 57% 46% 36% 37% 35% 27% 25% 27% 23% 14% 13% Electronic billing Access to test Ordering* Electronic medical results records * Electronic ordering of tests, procedures, or drugs. THE COMMONWEALTH FUND Source: Commonwealth Fund 2003 National Survey of Physicians and Quality of Care. 23 Figure 17. Electronic Health Records (EHR) in Solo or Small Group Practices: A Case Study EHR Financial Costs Per FTE Provider For 14 Practices, 2004-2005 Initial costs Ongoing costs per provider per year Dollars 70,000 63,600 60,000 50,000 43,826 40,000 30,000 20,000 14,462 11,867 8,412 5,957 10,000 0 Average per FTE Minimum Maximum provider THE COMMONWEALTH Source: R. Miller, et al. “The Value of Electronic Health Records in Solo or Small Group FUND Practices. Health Affairs. 24(5). (September/October 2005): 1127. Figure 18. EHR Financial Benefits Per FTE Provider, For 14 Solo/Small Group Practices, 2004-2005 2004- Average per FTE provider ($) 35,000 32,737 30,000 25,000 20,000 16,929 15,808 15,000 10,000 5,000 0 Total benefits Savings from Efficiency per provider increased coding savings levels THE COMMONWEALTH Source: R. Miller, et al. “The Value of Electronic Health Records in Solo or Small Group FUND Practices. Health Affairs. 24(5). (September/October 2005): 1127. 24 Figure 19. Variation in Per Capita Medicare Spending by Hospital Referral Region, 2000 THE COMMONWEALTH FUND Source: Eliot Fisher, presentation at AcademyHealth Annual Research Meeting, June 2006. Figure 20. Variation in Annual Total Cost and Quality for Chronic Disease Patients Quality of Care* and Medicare Spending for Beneficiaries with Three Chronic Conditions, by Hospital Referral Region 1.60 Best Practice Curve Ft. Lauderdale, FL 1.40 A East Long Island, NY Orange County, CA Greenville, NC Manhattan, NY Average Quality of Care Score 1.20 1.00 D 0.80 Boston, MA Newark, NJ 0.60 B C Saginaw, MI Melrose Park, IL Median Amount Spent per Patient 0.40 per HRR = $28,694 0.20 0.00 $- $10,000 $20,000 $30,000 $40,000 $50,000 $60,000 $70,000 $80,000 $90,000 Ave rage Annua l Re imburse ment per Be ne ficiary (W age -Inde x Adjuste d) * Based on percent of beneficiaries with three conditions (diabetes, chronic obstructive pulmonary disease, and THE COMMONWEALTH congestive heart failure) who had a doctor’s visit four weeks after hospitalization, a doctor’s visit every six months, FUND annual cholesterol test, annual flu shot, annual eye exam, annual HbA1C test, and annual nephrology test. Source: G. Anderson and R. Herbert for The Commonwealth Fund, Medicare Standard Analytical File 5% 2001 data. 25 Figure 21. Medicare Spending Per Enrollee and Mortality Rate by State, 2003 Medicare Spending per Enrollee Mortality Rate of Medicare Enrollees $4,500 $5,500 $6,500 $7,500 $8,500 2.5% HI 3.5% MN 4.5% OR UT CO AK ID WY NH AZ DC 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 Figure 22. IHA Trends in Point-of-Care Technology Point- of- Percent 2003 measurement year 2004 measurement year 70 60 50 40 30 20 10 0 Electronic Electronic Electronic Electronic Electronic prescribing check of Rx retrieval of access of retrieval of interaction lab results clinical patient notes reminders THE COMMONWEALTH FUND Source: Tom Williams, “California Pay for Performance (P4P): A Case Study.” 26 Figure 23. Evaluation of PacifiCare Pay for Performance: Improvement in Cervical Cancer Screening Percent improvement in cervical cancer screening rates among physician groups 20 15 10 5.3 5 1.7 0 California Pacific Northwest (Intervention group) (Control group) THE COMMONWEALTH FUND Source: M.B. Rosenthal et al., “Early Experience with Pay-for-Performance: From Concept to Practice,” JAMA 294, no. 14 (October 12, 2005): 1788-93. Figure 24. Physicians Participating in the Diabetic Care Program From 1997 to 2003 Showed Significant Improvement in Performance Percent of patients reaching quality target 100 46 45 50 25 17 0 1997 2003 1997 2003 HgA1c < 7% LDL/Chol < 100mg/dL THE COMMONWEALTH FUND Source: National Committee for Quality Assurance web site, www.ncqa.org/dprp. 27 Figure 25. 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 COMMONWEALTH © 2005 Premier, Inc. FUND Source: Stephanie Alexander, “CMS/Premier Hospital Quality Incentive Demonstration Project:1st Year Results,” Presentation at IOM P4P Subcommittee Meeting, November 30, 2005 Figure 26. Coordination Across Sites of Care: Care Transition Measure Scores,* Emergency Department Use, and Hospital Readmissions 100 Emergency Department Use 100 Hospital Readmissions 90 90 80 p=0.01 80 p=0.04 70 70 60 60 50 50 40 40 30 30 20 20 10 10 0 0 No Yes No Yes * When I left the hospital, I had a good understanding of the things I was responsible for in managing my health; when I left the hospital, I clearly understood the purpose for taking each of my medications; The hospital staff took my preferences and those of my family or caregiver THE into account in deciding what my health care needs would be when I left the hospital. COMMONWEALTH FUND Source: E.A. Coleman, “Windows of Opportunity for Improving Transitional Care,” Presentation to The Commonwealth Fund Commission on a High Performance Health System, March 30, 2006. 28 Figure 27. Improving Care Coordination and Reducing Cost Effect of Advanced Practice Nurse Care on Congestive Heart Failure Patients’ Average Per • Importance of improving Capita Expenditures transitions in care, doctor to doctor, and Dollars Visits Inpatient Care post-hospital $12,000 $9,618 • Follow-up care following $10,000 hospital discharge could $8,000 $6,152 reduce rehospitalization $6,000 • High cost care $8,809 management could $4,000 $4,977 reduce errors and lower $2,000 costs $0 $809 $1,175 • Will require restructuring Medicare benefits and Control Intervention incentives THE COMMONWEALTH FUND Source: M.D. Naylor, Making the Bridge from Hospital to Home, The Commonwealth Fund, Fall 2003. Figure 28. Improvement in Doctors’ Cervical Cancer Doctors’ Screening Rates Compared to Bonus Payments After Implementation of Quality Incentive Program Improvement in Bonuses Received Screening Rates Percent Thousands of dollars 25 500 437 20 400 15 11.1 300 10 7.4 200 128 5 2.5 100 27 0 0 High Middle Low High Middle Low performing performing performing performing performing performing group group group group group group THE COMMONWEALTH FUND Source: M.B. Rosenthal et al., “Early Experience with Pay-for-Performance: From Concept to Practice,” JAMA 294, no. 14 (October 7, 2004): 1788-93. 29 Figure 29. Most Physicians Have Not Been Involved in Collaborative Efforts to Improve Quality of Care Percent indicating involvement in any collaborative efforts in past two years* 100 No, have not been involved Yes, a LOCAL effort Yes, a REGIONAL effort Yes, a NATIONAL effort 67 50 Involved in at least one effort (32%) 23 8 6 0 THE COMMONWEALTH * Multiple answers possible. FUND Source: The Commonwealth Fund National Survey of Physicians and Quality of Care. Figure 30. Current Factors Affecting Physicians’ Compensation Physicians’ Major Factor Minor Factor Not a Factor Productivity/ Billing 58% 14% 27% 72% Board Re-Certification Status 11% 28% 60% 39% Measures of Clinical Care 8% 19% 72% 27% Patient Surveys/ Experience 8% 19% 72% 27% Quality Bonus/Incentive Payments from 4% 15% 80% Insurance Plans 19% THE COMMONWEALTH FUND Source: The Commonwealth Fund 2003 National Survey of Physicians and Quality of Care. 30