Statement of Glenn M. Hackbarth, Chairman, Medicare Payment Advisory Commission

Testimony Before the Subcommittee on Health
of the House Committee on Ways and Means

May 10, 2007

Chairman Stark, Ranking Member Camp, distinguished Subcommittee members, I am
Glenn Hackbarth, Chairman of the Medicare Payment Advisory Commission (MedPAC).
I appreciate the opportunity to be here with you this morning to discuss ways that Medicare can improve its physician payment system.

Since 2000, total Medicare spending for physician services has climbed more than 9 percent per year (Figure 1). Slowing the increase in Medicare outlays is important; indeed, it is becoming urgent. Medicare’s rising costs, particularly when coupled with the projected growth in the number of beneficiaries, threaten the sustainability of the program. The Medicare Trustees’ warn that even their unrealistically constrained estimate of Part B spending growth (due to multiple years of fee reductions mandated under current law) will still significantly outpace growth in the U.S. economy. Part B and total Medicare spending growth will continue to put pressure on the federal budget. That pressure puts other national priorities, such as homeland security and education, at risk.

Figure 1.         FFS Medicare spending for physician services, 1996–2006

Note:      FFS (fee-for-service). Dollars are Medicare spending only and do not include beneficiary coinsurance.

Source:  2006 annual report of the Boards of Trustees of the Medicare trust funds.

Rapid growth in expenditures also threatens to make the program unaffordable for beneficiaries. It contributes, directly and indirectly, to higher out-of-pocket costs through increased copayments, premiums for Medicare Part B, and premiums for supplemental coverage. As beneficiaries receive more services, they are required to make more copayments. Growth in copayments, in turn, pushes up the cost of supplemental insurance. In addition, because the monthly Part B premium is determined by average Part B spending for aged beneficiaries, an increase in expenditures affects the premium directly. From 1999 to 2002, the premium grew by an average of 5.8 percent per year, but the cost-of-living increases for Social Security benefits averaged only 2.5 percent per year. Since 2002, the Part B premium has increased even faster—by 13.5 percent in 2004, 17.3 percent in 2005, and 13.2 percent in 2006 (Figure 2).  

Figure 2.         Monthly Part B premiums, 1999–2007

Note:      Beginning in 2007, monthly Part B premiums are income-adjusted. The standard premium for 2007 is $93.50.

Source:  Congressional Research Service. 2004. Medicare: Part B premiums. Washington, DC: CRS; CMS press release, dated September 12, 2006, Medicare premiums and deductibles for 2007; and CMS press release, dated September 16, 2005, Medicare premiums and deductibles for 2006.

Spending for physician services has grown largely because of increased volume—the number of services furnished and the complexity, or intensity, of those services. Some observers have hypothesized that new technology, demographic changes, and shifts in site of service spur growth in the volume of physician services. Changes in medical protocols and a rise in the prevalence of certain conditions may also play a role. But analyses by MedPAC and others suggest that much of the rise in volume is unexplained. A RAND study found that technological advances and changes in medical protocols that are specific to particular illnesses do not fully account for volume growth. Other studies suggest that, after controlling for input prices and health status, differences in the volume of physician services are driven in large part by practice patterns and physician supply and specialization. As Elliott Fisher and others described in a series of articles, in geographic areas with more health care providers and more physician specialists, beneficiaries receive more services but do not experience better quality of care or better outcomes, nor do they report greater satisfaction with their care. John Wennberg identified some discretionary services that can be overprovided as preference-sensitive care because they involve significant trade-offs and should be selected only by patients capable of making an informed decision. This suggests that some services may be unnecessary, exposing some beneficiaries to needless risk and generating unwarranted costs for beneficiaries and the program. At the same time, evidence shows that beneficiaries do not always receive the care they need, and too often the care they do receive is not high quality.

To help address Medicare’s growing financial crisis, MedPAC focuses much of its work on improving efficiency—getting more in terms of quality and outcomes for each Medicare dollar spent. Increasing the value of the program to both beneficiaries and taxpayers will require efforts to improve the incentives inherent in Medicare’s fee-for-service (FFS) physician payment system.

Ideally, payment systems will give providers incentives to furnish better quality of care, to coordinate care (across settings, for chronic conditions), and to use resources judiciously. However, Medicare pays its providers the same regardless of the quality of their care, which perpetuates poor care for some beneficiaries, misspends program resources, and is unfair to

providers who furnish high-quality care and use resources judiciously. Medicare’s payment system does not reward physicians for coordinating patients’ care across health care settings and providers, and it does little to encourage the provision of primary care services, even though such actions may improve the quality of care and reduce costs. Further, inaccurate prices may inappropriately affect physician decisions about whether and what services to furnish. And Medicare’s FFS method of paying for physician services contributes to volume growth by giving physicians a financial incentive to increase volume.

As discussed in our March 2007 report on Assessing Alternatives to the Sustainable Growth Rate System, Medicare needs to change the incentives of the payment system by ensuring that its prices are accurate, furnishing information to providers about how their practice styles compare with their peers’ practice styles, encouraging coordination of care and provision of primary care, and bundling and packaging services where appropriate to reduce overuse. In addition, Medicare should promote quality by instituting pay for performance, encouraging the use of comparative-effectiveness information, and, where appropriate, imposing standards for providers as a condition of payment.  If Medicare’s FFS program is to function more efficiently, the Congress needs to provide CMS with the necessary time, financial resources, and administrative flexibility. CMS will need to invest in information systems; develop, update, and improve payment systems and measures of quality and resource use; and contract for specialized services.

Ensuring accurate prices

Misvalued services can distort the price signals for physician services as well as for other health care services that physicians order, such as hospital services. Some overvalued services may be overprovided because they are more profitable than other services. Conversely, some providers may opt not to furnish undervalued services, which can threaten access to care, or they may opt to furnish other, more profitable services instead, which can be costly to Medicare and to beneficiaries.

A service can become overvalued for a number of reasons. For example, when a new service is added to the physician fee schedule, it may be assigned a relatively high value because of the time, technical skill, and psychological stress that are required to perform it. Over time, the time, skill, and stress involved may decline as physicians become more familiar with the service and more efficient at providing it. The amount of physician work needed to furnish an existing service may decrease when new technologies are incorporated. Services can also become overvalued when practice expenses decline. This can happen when the costs of equipment and supplies fall, or when equipment is used more frequently, reducing its cost per use. Likewise, services can become undervalued when physician work increases or practice expenses rise. CMS—with the assistance of the American Medical Association/Specialty Society Relative Value Scale Update Committee (RUC)—reviews the relative values assigned to some physician services every five years. But many services likely continue to be misvalued.

In recent years, per capita volume for different types of services has grown at widely disparate rates, with volume growth in imaging and non-major procedures (e.g., endoscopies) outpacing that for office visits and major procedures. Volume growth differs across services for several reasons, including variability in the extent to which demand for services is discretionary and subject to the judgment of a physician or beneficiary, as well as advances in technology that expand access and can improve patient outcomes. The Commission and others have voiced concerns, however, that differential growth in volume is due in part to differences in the profitability of furnishing services.  One reason that different services have varying opportunities for profit is their prices.  In some instances, prices for services have been set too high relative to costs. For example, MedPAC and CMS have raised issues about the equipment use rate assumptions for imaging services.  This rate may be set too low for some imaging services, meaning that Medicare’s payment rate is set too high for these services.

To the extent that the Medicare’s sustainable growth rate (SGR) system limits growth in aggregate physician spending, differences in the rate of volume increases across services mean that certain types of services—such as imaging—are capturing a growing portion of Medicare physician spending at the expense of other services. As discussed below, the Commission has expressed particular concern about the tendency of primary care services to become undervalued relative to procedural services over time. This creates disincentives to furnish primary care services and over time can affect the willingness of physicians to enter the primary care specialties. (For more discussion of this issue, see p. 13.) Based on the

RUC’s recommendation, CMS recently increased the work relative values of many evaluation and management services. Because the fee schedule changes are implemented in a budget-neutral manner, their impact is partially limited.

Given the importance of accurate payment, the Commission concluded in the March 2006 report to the Congress that CMS must improve its process for reviewing the work relative values of physician services.  CMS looks to the RUC to make recommendations about which services should be revalued. But the RUC’s three reviews—completed in 1996, 2001, and 2006—recommended substantially more increases than decreases in the relative values of services, even though one might expect many services to become overvalued over time. We have noted that physician specialty societies have a financial stake in the process and therefore have little incentive to identify overvalued services. Although we recognize the valuable contribution the RUC makes, we concluded in our 2006 report that CMS relies too heavily on physician specialty societies, which tend to identify undervalued services without identifying overvalued ones. We found that CMS also relies too heavily on the societies for supporting evidence.

To maintain the integrity of the physician fee schedule, we recommended that CMS play a lead role in identifying overvalued services so that they are not overlooked in the process of revising the fee schedule’s relative weights; we also recommended that CMS establish a group of experts, separate from the RUC, to help the agency conduct these and other activities. This recommendation was intended not to supplant the RUC but to augment it. To that end, the new group should include members who do not directly benefit from changes to Medicare’s payment rates, such as physicians who are salaried, retired, or serve as carrier medical directors and experts in medical economics and technology diffusion. The Commission has also urged CMS to update the data and some of the assumptions it uses to estimate the practice expenses associated with physician services.

In addition, we recommended that the Secretary, in consultation with the expert panel, initiate reviews of services that have experienced substantial changes in volume, length of stay, site of service, and other factors that may indicate changes in physician work. For example, when a service becomes easier, quicker, or less costly to perform, physicians may be able to provide more of it. Rapid growth in volume for a specific service may therefore signal that Medicare’s payment for that service is too high relative to the time and effort needed to furnish it. The Secretary could examine services that show rapid volume increases per physician over a given period. Volume calculations would need to consider changes in the number of physicians furnishing the service to Medicare beneficiaries and in the hours those physicians work. CMS could use the results from these analyses to flag services for closer examination (by CMS or by the RUC) of their relative work values. The RUC could also conduct such volume analyses when making its work value recommendations to CMS, but its current process (every five years) may not be timely enough to capture services with rapid increases in volume.

Alternatively, the Secretary could automatically correct such misvalued services, and the RUC would review the changes during its regular five-year review.  In this scenario, CMS would identify specific service codes with volume increases exceeding a standard, such as average historical growth. The Secretary of Health and Human Services would then automatically adjust work values for these codes down.  The RUC would consider the changes as part of their next five-year review.

Corrections to the practice expense values may also be in order. MedPAC is currently studying the impact of CMS’s recent changes to the fee schedule practice expense calculation, including the use of newer practice cost data from some, but not all, specialties. We are also analyzing equipment pricing assumptions that are used to derive the practice expense values, particularly for imaging services. Ensuring that practice expense values are accurately priced reduces market distortions that make some services considerably more profitable than others, thus creating financial incentives to provide some services more than others.

Finally, revisiting the conceptual basis of the resource-based Relative Value Scale system may be in order. Some observers suggest that the pricing of individual services should account not just for time, complexity, and other resources but also for the value of the service and the price needed to ensure an adequate supply.

Measuring resource use and providing feedback

Elliott Fisher and others have found that Medicare beneficiaries in regions of the country where physicians and hospitals deliver many more health care services do not experience better quality of care or outcomes, nor do they report greater satisfaction with their care. Thus, the nation could spend less on health care, without sacrificing quality, if physicians whose practice styles are more resource intensive reduced the intensity of their practice.

In the March 2005 report to the Congress, the Commission recommended that CMS measure physicians’ resource use over time and share the results with physicians. Physicians would then be able to assess their practice styles, evaluate whether they tend to use more resources than their peers or what evidence-based research (when available) recommends, and revise their practice styles as appropriate. Moreover, when physicians are able to use this information in tandem with information on their quality of care, they will have a foundation for improving the value of care beneficiaries receive.

Private insurers increasingly measure physicians’ resource use to contain costs and improve quality. Evidence on whether measuring resource use contains private sector costs is mixed and varies depending on how the results are used. Providing feedback on use patterns to physicians alone has been shown to have a statistically significant, but small, downward effect on resource use. However, John Eisenberg found that, when feedback is paired with additional incentives, the effect on physician behavior can be considerably larger.

Medicare’s feedback on resource use has the potential to be more successful than previous experience in the private sector. As Medicare is the single largest purchaser of health care, its reports should command greater attention. In addition, because Medicare’s reports would be based on more patients than private plan reports, they might have greater statistical validity and acceptance from physicians. Confidential feedback of the results to physicians might induce some change. Many physicians are highly motivated individuals who strive for excellence and peer approval. If identified by CMS as having an unusually resource-intensive style of practice, some physicians may respond by reducing the intensity of their practice. However, confidential information alone may not have a sustained, large-scale impact on physician behavior.

Using results for physician education would provide CMS with experience using the measurement tool and allow the agency to explore the need for refinements. Similarly, physicians could review the results, make changes to their practice as they deem appropriate, and help shape the measurement tool. Once greater experience and confidence were gained, Medicare could use the results for payment—for example, as a component of a pay-for-performance program (which rewards both quality and efficiency). Alternatively, Medicare could use the results to create other financial incentives for greater efficiency or could make the results public to enable beneficiaries to identify physicians with high-quality care and more conservative practice styles. Eventually, collaboration between the program and private plans could result in the development of a standard report card.

MedPAC has been conducting research using episode grouping tools for the past two years and has found that they may be a promising tool for measuring resource use among physicians. We have found that the vast majority of Medicare claims can be assigned to an episode, and that most episodes can be attributed to a responsible physician. Once episodes are assigned to a responsible physician, each physician’s spending for a given episode can be compared to that of his or her peers and the results aggregated into an overall “score.” Episode groupers also permit analysis of the reasons for higher or lower resource use: Each episode can be subdivided into its component costs (e.g., hospital inpatient admissions, diagnostic testing, physician visits, post-acute care).

Additional research remains, however, to ensure that resource use measurement consistently groups claims into episodes and attributes episodes to physicians in a manner that correctly classifies physicians as high, average, or low users of resources. We also want to integrate quality measures into our comparisons of resource use. Adequate risk adjustment is crucial to ensure that episode grouping tools are measuring actual variation in resource use rather than variation in the health status of the beneficiaries being treated. Further, we and others have found significant variations in practice patterns for some conditions across the nation. As a first step it may be prudent to hold physicians to a local standard (e.g., metropolitan statistical area or state) rather than a national one and to compare physicians only to others in the same specialty. For example, in our March 1 report to the Congress on the SGR, we

compare a selected cardiologist in Boston to his local peers for his treatment of a specific condition (Table 1).  In this way, we control for some of the differences in practice patterns and patient health status that can drive resource use.

Table 1.           Hypertension episode resource use and scores by type of service

Note:      E&M (evaluation and management). Stage indicates the progression of the disease, with 1 being the mildest form. Resource use score is the ratio of the cardiologist’s resource use to the average for cardiologists in Boston.

Source:  MedPAC analysis of 100 percent sample of 2001–2003 Medicare claims using the Medstat Episode Group grouper from Thomson Medstat.

Encouraging coordination of care and the use of care management processes

The Commission has explored multiple strategies to provide incentives for high-quality, low-cost care and thus improve value in the Medicare program. However, even if individual providers are efficient, a beneficiary may still receive less-than-optimal care if providers do not communicate well with each other or if they do not monitor patient progress over time. To address this problem, we have considered ways to promote care coordination and care management by creating incentives for providers to share clinical information with other providers, monitor patient status between visits, and fully communicate with patients about how they should care for themselves between physician visits.

While many patients could benefit from better coordination of care and care management, the patients most in need are those with multiple chronic conditions and other complex needs. Gerard Anderson found that, in 2001, 23 percent of Medicare beneficiaries had five or more chronic conditions and accounted for 68 percent of program spending. But according to researchers at RAND, beneficiaries with chronic conditions do not receive recommended care and may have hospitalizations that could have been avoided with better primary care. Studies attribute this problem to poor monitoring of treatment—especially between visits—for all beneficiaries and to a general lack of communication among providers. Physician offices, on their own, struggle to find time to provide this type of care, and few practices have invested in the necessary tools—namely, clinical information technology (IT) systems and care manager staff. At the same time, beneficiaries may not be educated about steps they can take to monitor and improve their conditions. Coordinated care may improve patients’ understanding of their conditions and compliance with medical advice and, in turn, reduce the use of high-cost settings such as emergency rooms and inpatient care. Ideally, better care coordination and care management will improve communication among providers, eliminating redundancy and improving quality.

Research suggests that, without the support of IT and nonphysician staff, physicians can only do so much to improve care coordination. Individual physicians may not have the time or be well suited to provide the necessary evaluation, education, and coordination to help beneficiaries, especially those with multiple chronic conditions. One study found that older patients with select conditions that require time-consuming processes, such as history taking and counseling, are at risk for worse quality of care. Further, physicians may lack training or resources that would allow them to educate patients about self-care or to set up systems for monitoring between visits. Physicians’ use of basic care management tools is low, even in group practices where building the infrastructure for care coordination, including the use of clinical IT, may be more feasible.

Care coordination is difficult to accomplish in the FFS program because it requires managing patients across settings and over time, neither of which is supported by current payment methods or organizational structures. Further, because patients have the freedom to go to any willing physician or other provider, it is difficult to identify the practitioner most responsible for the patient’s care, especially if the patient chooses to see multiple providers. The challenge is to find ways to create incentives in the FFS system to better coordinate and manage care.

In our June 2006 report to the Congress, the Commission outlined two illustrative care coordination models for complex patients in the FFS program: (1) Medicare could contract with providers in large or small groups that are capable of integrating the IT and care manager infrastructure into patient clinical care, and (2) CMS could contract with stand-alone care management organizations that would work with individual physicians. In the second model, the care management organization would have the IT and care manager capacity.

In either model, payment for services to coordinate care would depend on negotiated levels of performance in cost savings and quality improvements. Given that Medicare faces long-term sustainability problems and needs to learn more about the most cost-effective interventions, the entities furnishing the care managers and information systems should initially be required to produce some savings as a condition of payment. However, demonstrating continued savings may not be necessary or feasible once strategies for coordinating care are broadly used.

To encourage individual physicians to work with care coordination programs, Medicare might pay a small monthly fee to a beneficiary’s personal physician or medical group for time spent coordinating with the program. As with other fee schedule services, these expenditures would be accommodated by reallocating dollars among all services in the fee schedule.

In either model, patients would volunteer to see a specific physician or care provider (e.g., a medical group or other entity) for their care. CMS could help beneficiaries identify the physician or physicians who provide most of their care. Beneficiaries could then designate the practitioner they wanted to oversee most aspects of their care to be the contact with the care management program. The physician and the beneficiary would agree that the beneficiary would consult first with that physician but would not be restricted to seeing only that physician. The physician, or the medical group on behalf of the practitioner in the case of a provider-based program, would receive the monthly fee when the beneficiary enrolls in the care management program. This designated physician (which need not be a primary care physician, because a specialist might be the appropriate person for patients with certain conditions) would serve as a sort of medical home.

These models do not represent the only ways care coordination might work in Medicare. The American College of Physicians recently advocated using advanced medical homes. In addition, other strategies, such as pay for performance, complement care coordination models by focusing on improving care. In addition, adjusting Medicare’s compensation to physicians to reflect the longer time spent caring for patients with complex issues may be warranted if the current fees do not compensate for this extra time. (For example, CMS could apply a multiplier to the relative value of certain services for identified patients with multiple chronic conditions.) Medicare could also establish billing codes to enhance payments for chronic care patients for services such as case management. The Medicare Health Care Quality Demonstration, which tests the ability of innovative payment arrangements for providers in integrated delivery systems to improve quality, may provide further models for improving coordination of care.

Evidence shows that care coordination programs improve quality, particularly as measured by the provision of necessary care. Evidence on cost savings is less clear and may depend on how well the target population is chosen. When cost savings are shown, they are often limited to a specific type of patient, the intervention used, or the time frame for the intervention. Indeed, researchers at Mathematica have suggested that cost and quality improvements are more likely to be achieved if programs are specifically targeted and the interventions are carefully chosen to benefit the targeted patient group. If care coordination programs work, annual spending may decrease, but beneficiaries may live longer with a better quality of life—a positive outcome for Medicare beneficiaries, but the Medicare program may not spend less than it otherwise would have. This possibility argues for assessing programs on the basis of whether they provide the interventions known to be effective or achieve certain quality improvements rather than on the basis of cost savings.

Promoting the use of primary care

Research shows that geographic areas with more specialist-oriented patterns of care are not associated with improved access to care, higher quality, better outcomes, or greater patient satisfaction. Cross-national comparisons of primary care infrastructures and health status have demonstrated that nations with greater reliance on primary care have lower rates of premature deaths and deaths from treatable conditions, even after accounting for differences in demographics and gross domestic product. Increasing the use of primary care in the United States, therefore, and reducing reliance on specialty care, could improve the efficiency of health care delivery without compromising quality.

But many observers worry that the United States is not training enough primary care physicians. Indeed, the growth in the supply of physicians in recent decades has occurred almost solely due to growth in the supply of specialists, while the supply of generalists—family physicians, general practitioners, general internists, and pediatricians—has remained relatively constant. A study by Perry Pugno and others found that the share of U.S. medical graduates choosing family medicine fell from 14 percent in 2000 to 8 percent in 2005. A 2006 study by Colin West and others found that 75 percent of internal medicine residents become subspecialists or hospitalists. There are many reasons why an increasing number of physicians choose to specialize, but one factor may be differences in the profitability of services.

Historically, Medicare’s payment system has valued primary care services less highly than other types of services. For example, according to a recent Annals of Internal Medicine article by Thomas Bodenheimer and others, the 2005 fee for a typical 30-minute physician office visit in Chicago was $90 while the fee for an outpatient colonoscopy, also about 30 minutes, was $227. In addition, primary care services also may be more likely than other services to become undervalued over time. While other types of services become more productive with the development of new techniques and technology, primary care services do not lend themselves as easily to these gains. Primary care is largely composed of cognitive services that require that the physician spend time with the beneficiary. In addition, many beneficiaries have multiple chronic conditions and a compromised ability to communicate with and understand their physician, both of which increase the time required for visits. It is difficult to reduce the length of these visits without reducing quality. (For that reason, physicians also find it difficult to increase the volume of primary care services furnished in a work day.) Over time, the specialties that perform those services may become less financially attractive.

Some Commissioners have argued that the relative value units of the physician fee schedule should be at least partly based on a service’s value to Medicare. Such an approach would focus on primary care services as well as other valuable services. For example, if analysis of clinical effectiveness for a given condition were to show that one service were superior to an alternative service for a given condition, then Medicare’s process of setting relative values might reflect that. This process would be a significant departure from the established method of setting relative values based only on the time, mental effort, technical skill and effort, psychological stress, and risk of performing the service.

In the longer term, the Commission is concerned that the nation’s medical schools and residency programs are not adequately training physicians to be leaders in shaping and implementing needed changes in the health care system. Physician training programs must emphasize a new set of skills and knowledge. For example, programs need to train residents to measure their performance against quality benchmarks, use patient registries and evidence-based care guidelines, work in multidisciplinary teams, manage the hand-off of patients, and initiate improvements in the process of caring for patients to reduce medication and other costly errors. Policymakers may want to consider tying a portion of the medical education subsidy to specific programs or curriculum characteristics that promote such educational improvements. In addition, policymakers may want to consider policies that promote the education of primary care providers and geriatricians. Bear in mind that physicians’ motivations to enter certain specialties go beyond income, including lifestyle concerns and professional interests.

Medicare’s cost-sharing requirements provide no encouragement for beneficiaries to seek services, when appropriate, from primary care practitioners instead of specialists, unlike most cost sharing in the under-65 market, where primary care copayments are often lower than those for specialists. Medicare’s payment policies and cost-sharing structure need to be aligned to encourage the use of primary care. The Commission’s pay-for-performance and care coordination recommendations could also encourage the use of primary care.

Bundling to reduce overuse

A larger unit of payment puts physicians at greater financial risk for the services provided and thus gives them an incentive to furnish and order services judiciously. Medicare already bundles preoperative and follow-up physician visits into global payments for surgical services. Candidates for further bundling include services typically provided during the same episode of care, particularly those episodes for conditions with clear guidelines but large variations in actual use of services, such as diabetes treatment.

Bundled payments could lead to fewer unnecessary services, but they could also lead to stinting or unbundling (e.g., referring patients to other providers for services that should be included in a bundle). Medicare should explore options for increasing the size of the unit of payment to include bundles of services that physicians often furnish together or during the same episode of care, similar to the approach used in the hospital inpatient prospective payment system.

The Medicare Prescription Drug, Improvement, and Modernization Act of 2003 (MMA) changed the way Medicare pays for dialysis treatments and dialysis drugs.  However, the MMA did not change the two-part structure of the outpatient dialysis payment system. One part is a prospective payment called the composite rate that covers the bundle of services routinely required for dialysis treatment; the other part includes separate payments for certain dialysis drugs, such as erythropoietin, iron, and vitamin D analogs that were not available when Medicare implemented the composite rate. Providers receive the composite rate for each dialysis treatment provided in dialysis facilities (in-center) or in patients’ homes.

The Commission has recommended that the Congress broaden the payment bundle to modernize this payment system. Medicare could provide incentives for controlling costs and promoting quality care by broadening the payment bundle to include drugs, laboratory services, and other commonly furnished items that providers currently bill separately and by linking payment to quality.

A bundled rate would create incentives for providers to furnish services more efficiently. For example, a bundled rate would remove the financial incentive for facilities to overuse separately billable drugs under the current payment method. In addition to an expanded bundle, changing the unit of payment to a week or a month might give providers more flexibility in furnishing care and better enable Medicare to include services that patients do not receive during each dialysis treatment.

MedPAC is examining bundling the hospital and physician payments for a selected set of diagnosis related groups (DRGs), which could increase efficiency and improve coordination of care. This approach to bundling could be expanded in the future to capture periods of time (e.g., one or two weeks) after the admission but likely to include care (e.g., post-acute care, physician services) strongly related to the admission, further boosting efficiency and coordination across sites of care. Bundled payments could be adjusted to provide incentives for hospitals and physicians to avoid unnecessary readmissions. Bundling services could be structured so that savings go to the providers, the program, or both. The Commission is also examining bundling physician payments with payments for other providers, such as hospital outpatient departments and clinical laboratories. In addition, MedPAC plans to examine the physician services furnished to patients before, during, and after inpatient hospitalizations for medical DRGs to assess whether a global fee should be applied to these services, as it is for surgical DRGs.

Hospital readmissions are sometimes indicators of poor care or missed opportunities to better coordinate care. Research shows that specific hospital-based initiatives to improve communication with beneficiaries and their other caregivers, coordinate care after discharge, and improve the quality of care during the initial admission can avert many readmissions. Medicare does not reward these efforts.  In fact, the program generally pays for readmissions, creating a disincentive to avoid them. To encourage hospitals to adopt strategies to reduce readmissions, policymakers could consider requiring public reporting of hospital-specific readmission rates for a subset of conditions and adjusting the underlying payment method to financially encourage lower readmission rates.

Episode grouper software, which is used to measure physician resource use and was discussed earlier, could also serve as a platform for bundling services for selected conditions.

Linking payment to quality

Medicare, the single largest payer in the U.S. health care system, pays all health care providers without differentiating on the basis of quality. Those providers who improve quality are not rewarded for their efforts. In fact, Medicare often pays more when poor care results in complications that require additional treatment.

To rectify this situation, MedPAC has recommended that Medicare change the incentives of the system by basing a portion of provider payment on performance. We recommended that CMS start by collecting information on structural measures associated with use of IT, such as whether a physician’s office tracks whether patients receive appropriate follow-up care, and claims-based process measures for a broad set of conditions important to Medicare beneficiaries. At the outset, CMS should base rewards only on the IT structural measures, with claims-based process measures being added to the pay-for-performance program within two to three years. Two other structural measures—certification and education—could become part of a measure set, but the link with improved care would need to be clear. The program should be funded initially by setting aside a small portion of budgeted payments—for example, 1 percent to 2 percent. The program should be budget neutral; all monies set aside would be redistributed to those providers who perform as required.

The Institute of Medicine (IOM) and MedPAC have stated that, ideally, pay-for-performance measures should be developed and used for all physician service providers to create incentives to provide better quality care. However, currently we do not have well-established measures for all providers of physician services. Thus, initially, policymakers might consider prioritizing the implementation of some pay-for-performance measures over others. Focusing measures on high-cost, widespread, chronic conditions (e.g., congestive heart failure) might be a good short-term strategy that will maximize benefits to the Medicare program and to beneficiaries. Further, measures that reflect coordination between health sectors will encourage and reward communication between providers, which may improve patient outcomes and reduce Medicare costs. The Commission considers that pay-for-performance initiatives would be implemented in a budget-neutral manner.

IOM and MedPAC assessments of the current state of quality measurement are similar. The indicators that are available now could form a starter quality measurement set. However, the measures that are currently available are fragmented across different users for different purposes and cannot be tied explicitly to the overarching, national goals laid out by IOM. Composite scores that could bring together multiple measures of different aspects of quality into a meaningful summary are needed, but judging the relative value of competing goals that would underpin such a summary is a challenge.

Both IOM and MedPAC have recommended that a national entity is needed to:

·        set and prioritize the goals of the health care system;

·        monitor the nation’s progress toward these goals;

·        ensure the implementation of data collection, validation, and aggregation;

·        coordinate public and private efforts at local, state, and national levels;

·        establish public reporting methods;

·        identify and fund development of the measures; and

·        evaluate the impact of quality improvement initiatives.

Encouraging the use of comparative-effectiveness information

Increasing the value of the Medicare program to beneficiaries and taxpayers requires knowledge about the costs and health outcomes of services. Comparative-effectiveness information, which compares the outcomes associated with different therapies for the same condition, could help Medicare use its resources more efficiently. Comparative effectiveness has the potential to identify medical services that are more likely to improve patient outcomes and discourage the use of services with fewer benefits. CMS already assesses the clinical effectiveness of services when making decisions about national coverage and paying for certain services. But to date FFS Medicare has not routinely used comparative information on the costs of services, although Medicare Part D plans and other payers and providers, such as the Veterans Health Administration, do use comparative information (e.g., in drug formulary decision-making processes).

Medicare could use comparative-effectiveness information in a number of ways to improve the quality of care beneficiaries receive. Medicare could use such information to inform providers and patients about the value of services, since there is some evidence that both might consider comparative-effectiveness information when weighing treatment options. Medicare might also use the information to prioritize pay-for-performance measures, target screening programs, or prioritize disease management initiatives. In addition, Medicare could use comparative-effectiveness information in its rate-setting process or in coverage decisions.

Given the potential utility of comparative-effectiveness information to the Medicare program, an increased role of the federal government in sponsoring the research is warranted. In our forthcoming June report, MedPAC will recommend that the Congress should establish an independent entity whose sole mission is to produce and provide information about the comparative effectiveness of health care services. The entity should set priorities and standards for new clinical- and cost-effectiveness research, examine comparative effectiveness of interventions over time and disseminate information to providers, patients, and federal and private health plans. The entity could be funded jointly by the federal government and the private sector, with an independent board of experts overseeing the development of research agendas and ensuring that research is objective and methodologically rigorous.

Using standards to ensure quality

CMS has set standards to ensure minimum qualifications for various types of providers (e.g., hospitals and skilled nursing facilities), but there are few examples of federal standards that apply to physician offices. The Commission has recommended that such standards be implemented for physicians who perform and interpret imaging studies. This recommendation was motivated by rapid growth in the volume of imaging. This growth was driven in part by imaging being increasingly provided in physician offices rather than in facility settings. (The growth is not fully offset with a corresponding decrease in imaging use in facilities.)  The lack of quality standards for imaging conducted in physician offices raises a number of quality concerns.  Therefore, the Commission recommended standards for physicians, facilities, and technicians that perform imaging studies. In the future, other types of services may be candidates for such standards.