Paying for Quality and Doing It Right

Paying for Quality and Doing It Right

Berenson, Robert A


Scholars have amply documented the poor quality of American health care.1 There are two aspects to this lack of quality: Substandard care is tolerated and acceptable care fails to live up to its potential.2 The poor state of quality overall is mirrored in the care provided to Medicare beneficiaries. The same physicians, hospitals and other providers who serve the public generally also serve Medicare beneficiaries.3 Many scholars have engaged in a longstanding debate over whether quality should be improved by removing the bad apples from health care or by improving typical practice and thereby raising the mean or median of care.4 Improving care, however, should involve both eliminating substandard care and improving typical practice.5 Strategic reasons may exist to either “protect the floor” or “raise the ceiling” on quality in any particular initiative or activity, but stewards of the Medicare program should be committed to both efforts.

Developing a Medicare policy that protects and promotes quality would be easy if quality were a dichotomous variable, with each particular provider either supplying acceptable or subpar health care. But most quality exists on a continuum, and falls along a bell-shaped curve. Thus, the toughest challenge is to determine which part of the curve represents substandard or unacceptable quality.6 Most quality is “acceptable” by public program standards even if there is plenty of room for improvement.

Complicating things further, any particular provider’s quality is highly variable by condition. For example, the hospital that is best at treating diabetics with severe eye complications may be only mediocre at managing diabetics with serious vascular problems.7 In addition, a never ending string of wellpublicized, horrendous system errors leading to death at the “best” placesDana Farber, Mount Sinai, Sloan Kettering, and, most recently, Duke University-reminds us that system safety and the concentrated expertise necessary for providing cutting-edge quality certainly may not be congruent.8 However, the focus on purely technical aspects of care, such as whether physicians follow recommended guidelines for diagnosis and treatment, has faced complex operational challenges. Factoring in the patient and consumer perspective complicates matters even more.9 Although for purposes of policy analysis there is a common tendency to separate the issues of “access” from those of “quality,” in the world where beneficiaries live, the concepts are merged.10 Thus, closing a hospital for technical quality deficiencies should involve important considerations such as the potential loss of jobs, as well as more patient-centered aspects of quality and availability.

This complexity makes one want to throw up one’s hands and just continue to rely on or hope that physicians will take responsibility for quality as mandated by their professional ethics. By adopting this attitude, the Centers for Medicare and Medicaid Services (CMS) and its contractors might aspire to pay claims efficiently and provide good service to beneficiaries. CMS might also try to figure out how to manage costs, which again has emerged as a major issue dominating public policy attention.”

As a social insurance program, Medicare relies upon providers in the private sector to supply health care services. Other developed countries with social insurance systems tend to defer to provider self-governance or corporatism, which often includes the authority to allocate funds among guild members.12 Certainly, the tradition of deferring to professional authority over clinical matters is ingrained in social health insurance programs such as Medicare.

This Article will fundamentally challenge the notion that Medicare should maintain a passive attitude toward the state of quality provided to Medicare beneficiaries. First, this Article will explore the reasons why Medicare should be an active promoter of quality, emphasizing that its current payment system and related policies already send strong signals that influence provider behavior and performance.13 This Article next explains the difference between purchasing and buying health care and identifies some of the current impediments to Medicare becoming a strong purchaser.14 The Article then presents a purchaser model that identifies numerous interventions, in addition to paying for quality, that should be part of a comprehensive purchasing approach.15 This Article points out that many generally accepted interventions, at least conceptually, raise the same issues as those presented by an explicit policy of paying differentially for quality.16 The Article concludes by focusing specifically on conceptual and technical issues that must be dealt with when paying for quality. The Article explains how these issues might be dealt with by using two examples: end stage renal disease (ERSD) services and Medicare+Choice plan payments.17

II. Why Medicare Needs to be a Promoter of Quality

Good reasons exist as to why Medicare should not accept a passive position regarding quality that merely defers to professionalism. First, left on their own, physicians and other providers likely will not be serious about quality improvement (raising the ceiling) or even improving basic safety (protecting the floor).18 Despite the Institute of Medicine’s well-documented and enlightened description of the quality epidemic, the provider response has been limited.19 The failure to respond is not surprising because the majority of doctors do not believe that serious system problems involving safety or quality exist.20 Regarding quality and safety, most physicians subscribe to the folk wisdom, “If it ain’t broke, don’t fix it.”21

Second, even when physicians and other providers are motivated to improve quality, the need to sustain their financial well-being often impedes quality improvement activities. In short, investing resources and effort in quality improvement often does not pay for itself and therefore meets resistance.22 The perverse reimbursement incentives used by both public and private purchasers of care are a major barrier to provider-initiated efforts in quality.23 Thus, if CMS takes no new action, Medicare’s policies will continue to contribute to provider complacency and frustrate well-intentioned quality improvement efforts.

Third, after spending decades accepting the current system, some employers and purchasing groups have tried to improve quality by becoming more active purchasers. Indeed, these active purchasers have demonstrated the potential for quality improvement through paying-for-quality programs. For example, the Business Roundtable founded an effort known as the Leapfrog Group, which uses its contracting authority to promote specific quality/safety activities related to mostly urban hospitals.24

Additionally, private purchasers in a few geographic areas actually pay differentially for quality. For example, Empire Blue Cross Blue Shield, a New York-based health insurer, worked with four large employers to design a system that rewards hospitals that meet the Leapfrog criteria. Private purchasers have established other payment-for-quality initiatives in Cincinnati, Louisville,25 and California.26 So far, these efforts have had little effect. Even the committed self-funded employers often lack the requisite market clout to affect provider behavior. By the time Medicare, Medicaid, the small and medium group insurance markets, the uninsured, workman’s compensation programs, and patient self-pay obligations are eliminated, the remaining large employer market share of expenditures may only be 10 to 20% in a typical market. Further, private plans ordinarily put no more than 10% of their total payments in such a quality bonus pool.27 In real terms, perhaps 1 to 2% of overall provider revenues might be affected through any particular private payment-for-quality contract. Perhaps this percentage is enough to get providers’ attention and motivate them to take the steps necessary to achieve available bonuses, but it is unlikely to be sufficient to change business as usual fundamentally. Thus, although some of these private sector pay-for-performance initiatives are wellintentioned and often well-designed, private purchasers have little ability to move markets and overcome physician and hospitals’ resistance to change.28

Other structural problems influence the well-meaning efforts of private employers and certain health plans to improve quality, including free rider problems, lack of first mover advantage, and lack of economies of scale.29 In short, although some scholars think that the private sector should take the lead in promoting quality in general, and in paying for quality in particular,30 such a position leads us nowhere but to the status quo-providers still will not improve quality on their own and private purchasers, even when motivated, have a limited ability to affect provider behavior.

In fact, Medicare’s payment and associated policies drive most provider behavior, for better or worse. For example, when Medicare pays generously to physicians and hospitals to perform invasive cardiac procedures but pays little for secondary prevention and disease management, cardiologists and cardiac surgeons structure their professional activities around performing the invasive procedures.31 Thus, by not explicitly focusing on how its coverage and payment policies affect quality, Medicare implicitly is rewarding certain behaviors that have a major effect on the quality of care that beneficiaries are receiving.

III. The Difference Between Purchasing and Buying Health Care

Traditionally, public purchasers in international social health insurance and national health systems have expressed concern about the total amount they spend, the general boundaries of the benefit packages that they fund, and what care is considered mainstream and what is considered alternative care. These purchasers, however, generally are less concerned about trying to influence the nature of the care they finance.32 Recently, however, purchasers have placed an increased emphasis on the performance of health systems. This attention has developed a consensus that the broad goals of health systems should include improvement in population health, responsiveness to legitimate public expectations, and fairness in financing.33 In order to achieve improvements in population health, purchasers need to be more actively concerned about how the health care delivery system operates. In systems based on social insurance principles, this need provides a rationale for providers to move from the passive payment of claims to the active purchase of services.

Beyond the goal of improved population health, the nature of health care provides an additional justification for public payors to assume an active purchaser role. As demonstrated in Nobel Laureate Kenneth Arrow’s seminal article, the relationship between patients and health care professionals is characterized by an asymmetry of information, certainly with regard to the technical aspects of care.34 The stress on the traditional principal-agent relationship that results because of this asymmetry of information emphasizes the need for purchasers to act on behalf of their patients in certain areas of care.35

It is important to understand that active purchasing involves many strategies beyond those considered in this Article. Indeed, in a comprehensive purchaser strategy, paying differentially for quality likely would be one of the least-used purchasing techniques. In brief, a strategic purchasing framework involves a cyclical set of activities.36 The first step in the framework is to assess health needs and, in particular, those needs that are less likely to be voiced by consumers and patients themselves. The second step is to determine how best to meet the identified needs by drawing on evidence of effectiveness, not just in relation to interventions on specific clinical problems, but also to the organizational structures and incentive systems that are most likely to result in delivery of effective care. The third step is to purchase care that complies with this specification. The fourth and final step in the cycle is to monitor the impact of this process, seeking to assure that effective care is firmly in place. Because health needs are changing and the delivery system is evolving, the assessment of health needs has to be a continuous process.37

Before discussing the tools that Medicare could use as an active, strategic purchaser to encourage quality improvement for Medicare beneficiaries, it is important to realize that the Centers for Medicare and Medicaid Services (CMS), the agency with primary responsibility for administering Medicare, must confront major barriers before becoming a purchaser. From its inception as a political compromise, the Medicare program has been prohibited from interfering in the practice of medicine and from limiting beneficiaries’ access to all participating providers in the program.38

Perhaps the most important policy concern about Medicare as an active purchaser relates to the government’s exercise of market power. In the case of purchasing for improved quality, this exercise runs the risk of distorting the market based on which provider or providers may be relatively favored.39 Medicare is subject to Section 553 of the Administrative Procedures Act,40 which Congress established to limit agency discretion and provide the opportunity for public review and comment-certainly a constraint in relation to the relative freedom that private sector purchasers enjoy.41

Due to procedural restrictions that limit discretion and the national nature of the program and its defined, uniform benefits, CMS generally has not been an active health care purchaser.42 When it purchases, CMS relies on the more traditional tools it has used since the inception of the program: regulated entry into the program and statutory coverage and payment policies. In recent years, CMS has begun to broaden its approach, for example, by providing information and education to providers and beneficiaries through the Quality Improvement Program43 and the National Beneficiary Education Program.44 These tentative efforts at purchasing have not yet included a commitment to paying for performance, of which paying differentially for quality is a component.45

Over the years, some individuals within CMS have attempted to develop a comprehensive purchasing model, which they have recently expanded.46 For the purposes of this discussion, it is important to summarize the different tools or levers that CMS potentially could use as part of a strategic-purchasing initiative. These categories of tools, with examples, include:

* Regulatory requirements that govern provider participation, e.g., hospital conditions of participation;47

* Benefit design, e.g., reduced cost-sharing for specified prevention services;48

* Coverage policy, e.g., covering organ transplants only in designated centers;49

* Payment policy, e.g., paying physicians using a fee schedule based on resource costs of production;50

* Technical assistance to providers, e.g., the Quality Improvement Organizations (previously called Peer Review Organizations);51

* Consumer information and education, e.g. the National Beneficiary Education Program;52

* Paying for performance, e.g., bonuses for Medicare+Choice plans for performance on national standards for congestive heart failure processes;53

* Collaboration among purchasers to achieve common objectives, e.g. participation in the National Quality Forum;

* Intervention directly in health care delivery, usually through contracting arrangements with organizations doing similar work for managed care organizations, e.g., disease management demonstrations.

An active strategic-purchasing regime, however, would employ a mix of the various tools. Yet, at any time, a particular purchasing tool may receive the greatest emphasis and generate the greatest controversy, often because of the views of the particular CMS administrator in place. For example, William Roper, who was administrator at the end of the second Reagan Administration, achieved national attention with his emphasis on publishing hospital mortality data to help inform Medicare beneficiaries’ choice of hospitals.54 After Roper’s departure, clinical outcomes received much less attention. In Bruce Vladeck’s tenure at CMS, he withdrew the hospital mortality data because of his view that the data was insufficiently case-mix adjusted and therefore not valid.55 More than a decade after Roper’s initial foray into clinical outcomes reporting, Thomas Scully, the current administrator, has again focused attention on quality reporting. Scully initially has focused on nursing home quality reporting, but plans to systematically provide quality data for many Medicare providers.56

IV. Distinctions Without a Difference

As noted earlier, Congress has permitted and even required CMS to engage in some purchasing activities, though not always using purchasing terminology, in the areas of regulatory entry barriers, coverage and payment policy, technical assistance to providers, and consumer education.57 Indeed, the recent trend is to advocate that Medicare should pay differentially for quality, and some experts have explored the technical issues of this idea to some degree.58 In contrast, conceptual concerns of the kind raised by Bruce Vladeck’s article in this volume have received less examination.59

One of Vladeck’s more troubling arguments is that quality is not randomly distributed among providers but rather is correlated with the relative affluence of the communities that physicians and hospitals serve.60 Thus, he argues, a straightforward incentive for achieving a threshold score on national quality measures61 would send extra Medicare dollars to Vermont and Minnesota but not to Louisiana or Texas, even if the quality in the former states does not change.62 As a result, the rich would get richer, and because Medicare works in a budget-constrained world, the poor would become poorer.63

As discussed below, CMS can mitigate the “rich get richer” argument that Vladeck presents. However, for the purpose of argument, let us accept that paying for quality provides disproportionate rewards for those providers whose financial positions place them in the best position to receive those rewards. If the desire not to advantage further those already doing a good job is a compelling constraint on paying for quality, then some broadly accepted purchasing approaches should be criticized because they would likely have similar effects on the relative financial impact on providers.

For example, despite the vacillating commitment of CMS to providing comparative quality information, most people accept that one of CMS’s legitimate and important roles is to provide valid, objective data to help inform beneficiaries’ choices. This choice exists in both the selection of an insurance vehicle, that is, traditional Medicare or a Medicare+Choice plan and, in the selection of particular providers at the point of service. The objections to using this purchasing tool tend to be “technical”-whether the quality data is appropriately case-mix adjusted so as to be valid, whether the data is reliable, and whether the data is understandable and useful. These are important and difficult technical issues, but for the most part, there is broad agreement on the desirability of providing such information to beneficiaries if the technical hurdles can be overcome.

For purposes of this discussion, assume CMS is able to provide beneficiaries with data that successfully overcomes these technical hurdles. One example of this type of data is the comparative information regarding the clinical outcomes of coronary artery bypass surgery (CABG), in terms of both morbidity and mortality.64 One goal of providing clinical outcome data is that some physicians who refer patients to hospitals for CABG surgery may act on this information and alter their normal referral patterns, while some beneficiaries may alter their care-seeking patterns. Accordingly, if hospital A has better outcomes data, it will treat more CABG patients. Hospitals B, C, and D, with less than average outcomes data, will treat marginally fewer patients.

Under the Medicare payment system, hospital A would get more average diagnosis related group (DRG) payments, while spending marginal costs. Hospital A would become a financial “winner” and hospitals B, C, and D would be financial “losers,” with the dollar amounts at stake probably numbering far in excess of the kinds of marginal financial incentives typically proposed in paying-for-quality initiatives.65

On this point, it is possible to move from the theoretical to the real by looking at what many consider to be one of the CMS’s most successful demonstrations.66 This demonstration approach used to be called the “Centers of Excellence” demonstration, but CMS changed the name to “Medicare Partnerships for Quality Services,”67 because hospitals excluded from the list did not want the public to view them as less than excellent.68

The original CABG demonstration, which took place between 1991 and 1996, bundled all hospital and physician services that apply to two related bypass surgery DRGs. CMS selected seven hospitals based on price, quality, service, and other criteria. In addition to receiving specific information about the documented outcome differences, centers were allowed to waive some or all patient cost-sharing obligations,69 and could pay for transportation costs-even for family members-given the immunity from the anti-kickback prohibitions.70 Importantly, beneficiaries retained the choice of seeking care at any participating hospital according to standard statutory cost-shanng and other terms.71

As a result, the CABG demonstration cut program costs by 10% for the 10,000 CABG surgeries performed, reduced expected mortality, and received high beneficiary satisfaction.72 Although most of the savings resulted from negotiated discounts, about 9% of the savings resulted from a shift in market share to lower-cost-demonstration hospitals.73 Hospital and physician trade associations have raised major objections to a government program which designates some providers as higher quality care providers than others and pays for services on a differential scale. Consequently, the objectors have effectively prevented a follow-up Centers of Excellence demonstration in Illinois, Ohio, and Michigan, and have consistently opposed giving CMS statutory authority to apply Centers of Excellence principles to Medicare more broadly.74

Clearly, providers are concerned about the loss of business which would result from a decreased volume of services, and not the differential payments.75 This experience highlights the political difficulties CMS faces in trying to become a strategic purchaser of health care. The focus of this discussion is that opposition to paying for quality should logically extend to other purchasing activities that on first glance may appear more straightforward and less controversial, such as informing beneficiaries to assist them in making better choices.76 Paying for quality may seem more radical than providing robust consumer information. The two approaches are, in fact, complementary purchasing tools that have different appropriate applications. The theoretical and actual examples also re-emphasize that a comprehensive purchasing strategy would be a more effective way to use purchasing tools than a concentration on the merits of paying for quality might suggest.

V. Paying for Quality Can Reward Different Kinds of Performance

Paying for quality typically is translated as paying for achieving specified quality outcomes. As noted earlier, this particular approach must confront technical barriers. These barriers include the need to do case-mix adjustment, which raise issues of validity, reliability, and usefulness.77 Some quality experts have argued that quality processes, as well as outcomes, should be the appropriate locus of quality attention for a number of reasons, including the fact that measures need to provide a clear link to the concrete actions that organizations can make to improve their quality.78 For purposes of Medicare’s consideration of payment for quality, rewards might be considered not only for measures of process and outcomes, but also for structural components of quality.79

As noted earlier, Vladeck and other experts are concerned that paying for quality will reward already well-funded, elite providers at the expense of the rest in a zero-sum game. The approach suggested in this discussion, however, should at least partly address this serious concern about tiering and equity in any paying-for-quality activities. First, because paying for quality can include both structural and process elements of quality, and not just outcomes, some of the specific measures might be designed so that institutions serving low-income or other disadvantaged populations are better positioned to do well and receive quality bonuses.

For example, under Title VI of the Civil Rights Act,80 all hospitals, health plans, and other health care institutions must provide translation and interpreter services for anyone with “limited English proficiency” (LEP), in order to avoid national origin discrimination.81 The burden of complying with these LEPrelated obligations naturally falls disproportionately on providers and plans based in geographic locations with large populations of non-English speakers. Despite this correlation, current Medicare provider payment rules do not take into account the significant financial burden that this difference creates for providers and plans that must provide translation, interpreter, and related services.

One strategic-purchasing approach to this inequity might include differential payments as part of the basic administrative pricing systems which recognize unique categories of provider organizations that meet certain criteria for providing burdensome LEP-related services. These differential payments would be similar to the rules that allow for differential payment to specific categories of rural hospitals, such as “sole community hospitals” and “critical access hospitals.”82 Under this formulation, extra payments would be made regardless of performance in implementing LEP services.

An alternative approach would be to construct quality structure and process measures that recognize and reward organizations that provide exemplary LEP services to non-English speaking populations. Arguably, this approach would generate more interest and would stimulate the exchange of the most effective practices. This approach could also be constructed to provide relatively small additional payments in aggregate to exemplary institutions, thereby avoiding the inevitable political fights that would occur if additional payments were proposed for all institutions falling into a new payment category. Under this approach to strategic purchasing, all organizations would be eligible to receive quality-related bonuses. However, a hospital in Westchester County, New York, which serves a predominantly affluent population, would likely be less interested in making the commitment to meet specified language-related measures than a hospital in East Harlem, which serves a lower income, multi-ethnic population.

The second response to the concern that paying for quality would accentuate tiering relates to the assumption of a zero-sum game, which means that paying some hospitals more for quality would result in other hospitals getting less. Some scholars that make the “business case for quality” suggest that improving the quality of health care would actually save substantial sums, and suggest that purchasers can improve quality and save money at the same time.83 Although reducing the quality problems of overuse and misuse may generate savings, one must acknowledge that substantial underuse of necessary services exists in areas of care where patients are not receiving interventions that would improve the quality of their care.84 Correcting underuse would often result in a net cost to the system, although in some cases, providing an underused preventive service might allow a patient to avoid a more expensive complication. At least some of the business case for quality analysis tends to discount or ignore the costs of correcting underuse.85 Also, for the most part, the recent studies examining geographic variations in quality received by Medicare beneficiaries document underuse, rather than misuse and overuse.86 This finding implies that the immediate action items to improve these documented quality problems could address underuse deficiencies, and as a result, paying for quality may cost money and support concerns that paying additionally for some underused procedures may require reductions for other procedures in order to maintain budget neutrality.

While much of the focus of quality measurement has been on documenting underuse, easily identifiable areas of overuse and misuse exist in the care provided to the Medicare population, areas where quality improvement would save substantial dollars. Indeed, if overall budgetary constraints were of compelling concern, the initial paying-for-quality initiatives might concentrate on those clinical conditions and geographic areas where improvement in quality could be expected to save money. One of the advantages to being a strategic purchaser, rather than a passive buyer, is that one can be selective with one’s targets and goals. As noted earlier, CMS used the authority of establishing a risk adjustment mechanism in the Medicare+Choice program to reward plans that achieved a national threshold of performance care for patients with congestive heart failure. These plans included the performance of tests to measure the left ventricular function to help assess the nature of the heart failure. These tests were followed, where appropriate, based on the results of left ventricular testing, with prescription of an angiotensin 1-converting enzyme inhibitors (ACEI) as therapy.87 In this case, greater adherence to these measures should not only reduce medical complications and improve quality of life, but also should reduce the number of costly hospital admissions.88

Consistent with the broad strategic-purchasing model described earlier, CMS currently is engaged in a variety of disease management demonstrations, including a recently announced full capitation demonstration of disease management which included congestive heart failure.89 The agency should also consider separate contracts with independent vendors that can assist traditional Medicare providers and suppliers-including physicians and hospitals-to deliver quality services. However, rewarding those traditional providers directly for achieving specified targets of quality that are associated with reduced hospitalization rates for CHF is an alternative approach that might be easier to administer or might be preferable in geographic areas that do not have the volume to support payment to a separately contracted disease management organization. Regardless of which approach is used, CMS could choose to emphasize clinical problems where enhanced quality should result in decreased program spending, thereby mitigating concerns that the program would contribute to tiering by rewarding some hospitals at the expense of others.

VI. Opportunities for Paying for Quality

Other examples exist in Medicare where paying for quality might result in improved quality provided to Medicare beneficiaries and where the extra payments made would not have to assume a zero sum, with winners and losers. That is, the initiative might result in net savings for Medicare. The End Stage Renal Disease (ESRD) Program is one of the most well-developed areas of quality measurement. It is now known with reasonable assurance that achieving a defined threshold of adequacy of hemodialysis sessions predicts lower rates of mortality for ESRD patients and is associated with lower rates of hospitalization.90 This particular measure-Kt/V or its functional equivalent URE.91-does not need to be case -mix adjusted.92 While measures of mortality rates for beneficiaries with ESRD would have to be case-mix adjusted, this intermediary outcome measure related to the adequacy of individual hemodialysis sessions does not.

Admittedly, reducing mortality rates would result in net increased costs for the Medicare program. Achieving the recommended threshold of dialysis adequacy, however, also significantly reduces hospitalization rates for these patients overall and for specific dialysis-related conditions, such as cardiac and infectious complications.93 One could attempt to estimate the net expenditure difference associated with both decreasing mortality rates and decreasing hospitalization rates for beneficiaries with ESRD. Regardless, an intervention to improve performance of dialysis offers promise as a high priority strategic purchasing activity. Indeed, using the purchasing tool of technical assistance to providers, CMS already has had a significant impact on dialysis unit performance.94

The ESRD example illustrates an important point about quality. The strategy that CMS has been using in this case has not focused on weeding out substandard performance. Instead, the program has focused on the quality continuum, where some units do marginally better than others on some measures of performance.95 The misguided notion that quality is a dichotomous variable-that is, a provider either provides acceptable quality or not-is not supported by the data from hemodialysis centers. As such, a proposal that Medicare should adopt a policy of including “quality” providers and excluding “non-quality” providers would not work.96 CMS has been trying to improve the quality of acceptable performance and thereby to move the quality curve. As a result, CMS seems to have been successful in improving the adequacy of hemodialysis.

Nevertheless, a recent study concludes that because all methods of raising dialysis dose were associated with higher costs, the existing payment system may be hindering ongoing efforts by the federal government and professional societies to implement guidelines calling for an increase in dose in order to reduce the mortality rate.97 The study also argues that a dose-adjusted payment system would be more complex than the existing system, because it involves standardizing the methods of measurement and instituting audit capabilities sufficient to deter fraudulent reporting. The study further suggests that the complexities of adjusting administered prices to achieve the objective of increasing dialysis dose for certain patients might make the effort infeasible.98

An alternative purchasing strategy to adjusting administered prices would involve paying for quality. This alternative would require marginally higher payment to dialysis units that achieve a threshold performance, measured by the percent of patients whose Kt/V is greater than the clinical threshold of 1.2. The Kt/V measure already is the gold standard and is being reliably measured and reported by scholars and practitioners.”

The broader point that the ESRD example demonstrates is that in an opportunistic way, CMS already uses strategic purchasing approaches to obtain better quality for Medicare beneficiaries and to enable beneficiaries to access high quality providers. Paying differentially for quality is merely one set of available purchasing tools. Although differential payment raises particular technical issues that would have to be confronted, paying for quality in this case does not seem to raise fundamental philosophical or ideological problems, as critics suggest.100

VII. Paying for Quality in Medicare+Choice

Recently, the author joined with two other policy analysts to propose a specific paying-for-quality initiative in the Medicare+Choice program.101 In this case, the proposal is straightforward. Instead of giving all Medicare+Choice plans an increase above that provided in the controversial, statutory administrative pricing formula,102 any increases should take the form of bonuses. These bonuses would be between 1% and 3% above current payment rates, and would be given to Medicare+Choice plans that have outstanding performance records, as demonstrated by score achievement on well-recognized quality measures.103

A logical rationale exists for launching a payment-for-quality initiative based on Medicare+Choice plans.104 Medicare+Choice plans have accountable management structures and are responsible for populations of enrollees. These features provide the denominator that often is lacking in the traditional program and is needed to calculate many quality outcome measures.105 In addition, the National Committee for Quality Assurance has pioneered well-accepted quality measures that are routinely used to compare HMO performance. These measures are based on the Health Plan Employer Data and Information Set (HEDIS), which includes a patient satisfaction component called the Consumer Assessment of Health Plan Satisfaction (CAHPS).106

Regarding the concern that paying for quality would result in inequitable tiering, it could hardly be a more inequitable payment system than the one currently utilized in Medicare+Choice programs. Currently, beneficiaries in places such as Dade County, Florida, and Los Angeles County are able to get generous prescription drug benefits because Medicare+Choice plans are overpaid. Beneficiaries in Minneapolis and Portland, Oregon, however, get no prescription drug benefits because their plans are underpaid.107 When determining how much additional funding Medicare+Choice plans should receive, it makes greater policy sense for CMS to distribute the additional funds to high-quality medical service providers than to distribute uneven payments arbitrarily under plans based on the local payment levels found in the traditional Medicare program.

One of the difficult design issues in this proposal relates to what kind of outstanding quality performance should be rewarded. In other words, should bonuses be based on national leadership, market area leadership, and/or a health plan’s individual improvement?108 This design issue certainly would be relevant to any approach to paying for quality care in the traditional program as well as in Medicare+Choice, given the well-documented variations in quality by state and region of the country.109 The use of national standards would reward prototype health plans that “offer national models, benchmarks and best practices for the rest of the Medicare program.”110 Certainly, targeting quality bonuses to high, national performance levels would seem most consistent with the notion of paying for quality.

It is likely, however, that these national-level standards would be beyond the immediate reach of many health plans that inherit as a baseline the practice patterns of the providers they contract with or employ. Thus, one might include a strategy for targeting quality incentives for regional leadership, even if the incentives do not meet national performance standards.111 “Area-based bonuses would provide incentives for health plans to offer Medicare enrollees better quality of care than would be otherwise available, and they could establish a competitive learning and performance dynamic among health plans competing in regional or local markets.”112

Another approach for providing quality incentives would be to reward Medicare+Choice plans for improvements in their own quality scores from previous years.113 A bonus system that considers levels of improvement might strengthen the incentives even for poorly performing plans to do better. This bonus system would eliminate the concern that quality bonuses only allow the rich to get richer.

Building on the paying for quality approach for Medicare+Choice plans, the Alliance of Community Health Plans-a trade association representing many non-profit health maintenance organizations-in conjunction with Group Health Cooperative of Puget Sound, a Seattle-based health plan, recently developed a proposal that would allocate 75% of the incentive payment pool to the top 25% of plans nationally and grant the remaining 25% to plans based on state level performance, as long as the plan achieved the 60th percentile nationally.114

VIII. Conclusion

Bruce Vladeck correctly argues that the characteristic of payment systems that has the greatest effect on providers is the absolute level of payments, and that adding an additional signal relative to quality may not provide any additional incentives.115 Put another way, the base payment system may often convey basic performance signals that overwhelm marginal incentives, such as providing modest bonuses for achieving specified quality measures.

Thus, if the basic payment signal rewards physicians and hospitals for admitting patients to the hospital and performing procedures on them, it is that signal-in combination with other considerations such as professional judgment, patient preferences, and ability to pay and the nature of provider competition-that will determine provider behavior. In the face of that basic financial incentive, providing a small, marginal incentive to keep patients out of the hospital may not have much effect.

This reality has two basic implications for paying-for-quality initiatives. First, specific paying-for-quality incentives must make sense in the context of the other forces that are affecting provider behavior and will be meaningless if those other forces dominate. Put another way, paying for quality would have influence “all other things being equal.” As part of a comprehensive purchasing strategy, paying for quality programs must be aligned with other program policies.

The corollary is that, as currently conceptualized and tentatively implemented, paying for quality does not raise the troublesome concerns that are specific to this particular payment approach. The concerns raised by Vladeck and others would apply equally or even more to other approaches to improving quality-such as giving consumers more robust information on which to make choices, and selecting centers of excellence that are given a degree of program preference. Those who oppose paying differentially for quality should logically also oppose these other approaches to influencing the delivery system and how beneficiaries make choices.

Yet, to do so would be to adopt a nihilistic policy that inevitably would result in support for the status quo of mediocre care provided to Medicare beneficiaries. Private sector efforts to influence quality are limited by their lack of expertise and the requisite market share. If allowed to do so, Medicare has a unique opportunity to take the lead in creating quality incentives.

Legitimate concerns exist about some of the design issues in paying for quality. If done poorly, there may be untoward side effects such as exacerbated hospital tiering. However, Medicare, as a broadly supported social insurance program, is likely to be more sensitive to some of the difficult design issues that will determine who benefits and who loses under a regime of paying for quality.

Robert A. Berenson*

* Senior Fellow, the Urban Institute; Adjunct Professor, School of Public Health, University of North Carolina; and Adjunct Professor, Fuqua School of Business, Duke University. I would like to thank Jack Ebeler, Lynn Etheredge, and Tim Jost for their helpful suggestions in connection with this Article and Bruce Vladeck for proposing that this topic be debated.

Copyright Washington & Lee University, School of Law Fall 2003

Provided by ProQuest Information and Learning Company. All rights Reserved