The Promise of Pay for Performance
This chapter reviews the current health care payment systems; the strengths, weaknesses, and potential adverse consequences of pay for performance; early experiences with the approach; and the ways in which pay for performance can be used as a pathway to reform. Multiple and complex challenges confront any such effort, and monitoring and evaluation will be essential so stakeholders can learn from experience, identify unanticipated consequences, and implement midcourse adjustments as necessary.
Current payment systems are not well aligned with efforts to achieve the six quality aims set forth in the Quality Chasm report (IOM, 2001). These systems place little emphasis on achieving high clinical quality, do not reflect the value of services, frequently act to drive up costs, and do not encourage patient-centered care or the efficient use of resources. While this report does not attempt to address all shortcomings of the current systems, the committee’s analysis should be viewed within the broader framework of the need for fundamental reform of the health care payment structure.
The initiatives proposed in this report would modify current health care payment systems by using financial incentives to promote higher levels of quality across diverse health care settings. These initiatives are predicated on the assumption that the health care Americans receive could and should be of considerably greater value—better-quality care obtained at a sustainable and socially acceptable cost (see Chapter 1). Based on a review of the available evidence, the committee concluded that modest changes alone in the current systems—systems in which provider reimbursement is based largely on the quantity of health care services rendered—are unlikely to promote significant progress toward the goals of improved quality and reduced growth in costs. Rather, a profound and fundamental alignment of
incentives (financial, informational, and reputational) with desired outcomes is required to stimulate the needed transformational change in the current health care payment systems.
CURRENT PAYMENT SYSTEMS
At present, the care for 88 percent of Medicare beneficiaries, or approximately 35 million individuals, is paid for under fee-for-service systems. The remaining 12 percent of beneficiaries are enrolled in the Medicare Advantage program, under which private plans are paid monthly, risk-adjusted capitated amounts in return for providing Medicare’s benefits to those who choose to enroll in Medicare Advantage (Kaiser Family Foundation, 2005). Medicare’s fee-for-service payment rates and fees are set administratively by the Centers for Medicare and Medicaid Services (CMS) at levels intended to cover the cost of the resources typically required to provide a particular service. The service may be defined narrowly, such as a chest x-ray, or broadly to encompass a bundle of services, such as all the inpatient hospital care associated with a stay of any duration for a heart bypass operation.
Medicare’s rates and fees do not vary with the quality of the service provided. Furthermore, the fee-for-service payment structure generally does not provide reimbursement for health services that are recognized as important contributors to quality, such as comprehensive case management, care coordination, health counseling, and many preventive services that may reduce the need for hospitalization or more expensive future medical procedures. In addition, because of the way payment rates for different services are set relative to one another, new, complex, high-tech interventions tend to be better compensated than procedures involving less intensive service use, less or older technology, and more time with patients (which may be important to quality care) (Ginsburg and Grossman, 2005). Additionally, payment rates and fees do not vary according to the need for a particular service. For example, one study that examined clinical decision making under different payment systems found that expenditures for discretionary services were lower under capitated than under traditional fee-for-service arrangements (Shen et al., 2004). Providers are paid more for doing more and are not penalized when the provided services are of little or no value or, worse yet, negatively affect health outcomes. In some cases, the incentives embodied in fee-for-service payments may encourage the delivery of unnecessary or even harmful services that can raise fundamental concerns about cost and safety (Robinson, 2001).
Since fee-for-service payments offer little direct incentive to improve quality or avoid low-value services, they fall short of fostering goals in the
three critical domains identified in Chapter 1: clinical quality, patient-centeredness, and efficiency. Although fee for service is responsive to patient demand for services in the sense that the health care system is responsive to the sickest patients who require more complex and higher levels of care, this type of payment structure offers no incentives to providers or patients to improve overall health status through preventive services or lower-cost interventions that can ultimately reduce the demand for more complex clinical services. The systems pay for treating illness and injury, not for keeping people well.
Since Medicare’s inception, policy makers have been concerned about the rapid growth of health care expenditures. The program’s payment systems have been modified in response to these concerns, but these changes have not been sufficient. More recently, concern has also focused on the quality of care, and some steps to improve quality have been taken. CMS is currently conducting demonstrations to test payment systems designed to reward higher-quality care, but these initiatives have not yet been implemented on a wide scale. A brief review of Medicare payment policies follows.
Original Medicare Retrospective Payment Systems
Initially, Medicare payment policies followed the prevailing private-sector practice of the early 1960s, which was to reimburse providers for the lesser of either their usual and customary charges or their actual costs for each service delivered. The reimbursement system was retrospective because payments could not be calculated until after the service had been provided; the physician or hospital would not learn the exact payment amount until after the end of the year, when customary charges and actual costs could be audited and payment rates calculated. This payment system provided no real restraint on expenditures. The more providers spent on a service or increased charges, the more Medicare would ultimately pay. To limit growth in expenditures, Medicare began to define more narrowly which costs were acceptable and which were not, as well as to set limits on allowable increases, thereby making the payment system increasingly complex.
Prospective Payment Systems
In an attempt to gain better control over burgeoning expenditures, policy makers began in the 1980s to shift Medicare from retrospective to prospective payment systems. Prospective payment was first introduced in inpatient acute care hospitals in 1983. Since then, CMS has instituted prospective payment for other provider settings, including skilled nursing facilities in 1998, home health agencies in 2000, and
outpatient acute care hospitals in 2000 (SUNY, 2001; CMS, 2006c). Physicians are reimbursed according to the Medicare Physician Fee Schedule. The new systems set payments for various services (or bundles of services) in advance of their delivery. Thus, providers know how much they will be paid before they treat their patients and can better plan their care and resource use.
Fees and payment rates of Medicare’s prospective payment systems are set administratively to cover CMS’s estimate of the average cost of providing a service, plus a small margin. In some instances, the payment does not cover the provider’s costs; in other cases, the payment is more than sufficient. In some situations in which costs far exceed payments, Medicare provides additional “outlier payments” that cover a portion of the excess costs. Under most of Medicare’s prospective payment systems, payments are adjusted for geographic differences in labor and other costs. In general, prospective payment encourages providers to keep the costs of services below the payment amount and creates incentives to treat those with the least severe and complex conditions in any particular diagnosis, service, or risk category.
As noted above, the unit for which payments are made may be a bundle that encompasses all of the inputs necessary to provide a stay in an institution or perform a procedure, or it may be a discrete, narrowly defined item, test, or service. Under most of the payment systems, unless Congress intervenes, rates are automatically adjusted upward each year based on indexes of anticipated price increases. A major exception is the Medicare Physician Fee Schedule update, which is governed by the sustainable growth rate (SGR). This formula limits the growth in per beneficiary Medicare Physician Fee Schedule expenditures to the growth in per capita gross domestic product. Because the volume and average intensity of services paid for under the Medicare Physician Fee Schedule have been growing very rapidly, application of the SGR formula has resulted in negative updates for physicians in recent years. With the exception of 2002, Congress has acted to avert these reductions (U.S. GAO, 2005). CMS projects that the SGR will impose annual negative updates of more than 4 percent each year during 2007–2011, which may affect physicians’ willingness to consider performance-related payment changes and incentives (MedPAC, 2006).
Appendix A presents more detailed descriptions of payment systems and their incentives for in- and outpatient hospital care, skilled nursing facilities, home health care, outpatient dialysis services, physicians, and Medicare Advantage plans. The discussion there is intended to give a broad overview of payment methodologies, not a detailed picture of all the complexities of each method, to provide a context for the consideration of payment incentives. Table 2-1 presents an overall picture of spending in the Medicare program by provider setting.
TABLE 2-1 Medicare Program Spending by Provider Setting, 2003
PAY FOR PERFORMANCE AS A PATHWAY TO REFORM
Pay for performance has emerged as a promising strategy to address the inadequacies of the current payment system outlined above, and has attracted considerable attention in the private marketplace. Additionally, CMS has begun to invest resources in pay for performance as a reform strategy (CMS, 2006a). These pay-for-performance initiatives must be implemented successfully, both to achieve their particular goals—improved quality of care and cost containment—and to prompt the fundamental changes needed in the health care system overall.
In other sectors of the American economy, a reform of this magnitude would be based not only on sound theory, but also on pertinent practical experience. While the database on which to base the design and evaluation of pay-for-performance programs is growing steadily, it remains incomplete and without substantial validation. Despite this lack of a definitive evidence base, both private- and public-sector decision makers would like to move forward aggressively with pay-for-performance programs. However, experience with other health care initiatives suggests that the rapid implementation of new payment strategies based on theory and preliminary results does not always achieve the desired goals. In fact, it can prove to be counterproductive, exacerbating current problems and creating new ones.
The Theory Behind Pay for Performance
In essence, pay for performance represents an attempt to align incentives in the payment system so that rewards are given to providers who foster the six quality aims set forth in the Quality Chasm report (IOM,
2001) (see Chapter 1) and improve health outcomes while using resources parsimoniously. At the most basic level, improving care requires changes in the behavior of providers. Paying providers for improving performance or achieving superior levels of performance should motivate them to focus on doing so in measured areas. Pay for performance also has the potential to achieve change by influencing the environment in which providers practice. For instance, performance-based payment could make it attractive for both providers and provider organizations to invest in improved systems for tracking and enhancing the quality of care, making them better able to manage the health of the populations they serve. Ideally, pay for performance would encourage certain changes in structural and organizational practices, such as a new emphasis on comprehensive and coordinated care and collaboration across individual settings of care, and stimulate consumers to pay attention to quality practices.
Effects of Medicare Payment Systems on Provider Behavior
Evidence that providers have responded to changes in Medicare payment policies in the past suggests that health care providers will likely change their behavior in response to Medicare payment incentives to improve quality. The implementation of various Medicare prospective payment systems has been associated with significant changes in provider behavior. All of these behavior changes cannot be attributed conclusively to the new payment systems because those systems did not emerge in isolation, and because research on their effects often examined varying aspects of change, used different data, and focused on different types of providers. Nonetheless, the literature attests to dramatic shifts in the way health care is delivered since the new systems were instituted. For example:
In the 1980s, hospital discharges and average lengths of stay were slowly decreasing among those under age 65, while both rates were increasing for the Medicare population. This trend reversed in the Medicare population after a prospective payment system was implemented in acute care hospitals in 1983. Also, utilization rates dropped dramatically in 1984 and 1985, while those rates among the rest of the population continued to decrease at a more gradual pace, although utilization increased somewhat for both populations later in the 1980s (Hodgkin and McGuire, 1994).
Controlled studies of the responses of hospitals to Medicare payment changes showed that their behavior was related directly to Medicare’s portion of their volume. Thus hospitals that were more dependent on Medicare patients were likely to show a larger change in the observed behavior relative to hospitals with a smaller proportion of Medicare patients (Hodgkin and McGuire, 1994).
One study that compared cost-based and flat-rate Medicaid payments for skilled nursing facilities and intermediate care facilities found that facilities in states with cost-based reimbursement tended to have more registered nurses and fewer licensed practical nurses per resident than facilities in flat-rate states, regardless of the ownership status of the facility (Cohen and Spector, 1996).
A nursing home study based on Medicare claims data and other administrative datasets examined the charges of skilled nursing facilities for rehabilitation therapy (physical, occupational, and speech therapies) (White, 2003). When a prospective payment system was implemented, there was a dramatic drop in the percentage of skilled nursing facility residents that received a high level of therapy (more than $200/day), and residents of these facilities were more likely to receive a moderate level of daily therapy than either extremely high levels or no therapy. The changes observed were consistent with the incentives of the prospective payment system, which offers relatively generous payments for moderate levels of rehabilitation therapy and stops paying for therapy above a specified level per week. The changes were observed between 1997 and 2000. Transition to the new payment system was gradual, beginning in 1998.
Early Experiences with Pay for Performance
CMS has undertaken several Medicare pay-for-performance initiatives for different provider settings. Some of these initiatives are in the planning phase; others have recently been implemented. One example of the latter is CMS’s demonstration project with Premier, Inc. (the Premier Hospital Quality Incentive Demonstration), in which hospitals among the top 20 percent of performers receive bonus payments (CMS, 2006b; Premier Inc., 2006). Year 1 of the project yielded positive results among the 262 participating hospitals; data showed statistically significant improvement in all five clinical areas examined, with an overall improvement of 6.6 percent (Remus, 2005). Another such project is Medicare’s Physician Group Practice Demonstration, which rewards group practices for their performance on quality-of-care metrics, but only after the practices achieve savings of at least 2 percent of projected expenditures (Kautter et al., 2004).
Other CMS projects in development focus on promoting the adoption and use of information technologies among physicians and the use of disease management models to improve the quality of care. Bonuses (or administrative fees) are often contingent upon demonstration of net savings to Medicare. The 3-year Medicare Management Performance Demonstration
was mandated by the Medicare Prescription Drug, Improvement, and Modernization Act of 2003. The project focuses on small and medium-sized physician practices, and promotes the use of health information technologies to improve care for the chronically ill. The same act also mandated several disease management projects that often make vendor fees dependent on proven savings.
CMS is currently running several other demonstrations that include projects related to dialysis facilities, nursing homes, and chronic care (CMS, 2006a). While the demonstrations are impressive in number and variety and may yield valuable insights, demonstration projects in and of themselves may not generate large changes in the health care system since they tend to be short-lived, end with isolated reports to Congress, and do not necessarily lead to specific follow-up activities.
Aside from these specific efforts, it is important to note that CMS is actively collaborating with the Ambulatory care Quality Alliance and the Hospital Quality Alliance on pay-for-performance strategies. This type of collaboration is key to the success of pay for performance, not only for the improvement of individual programs through shared learning, but also for the success of pay for performance on a larger level through alignment of incentives, decreased confusion associated with multiple requirements, and synergism among the multiple efforts under way.
In the past several years, numerous employers, purchasing coalitions, and health plans have announced new efforts to reward health care quality. Estimates suggest that more than 100 individual pay-for-performance efforts are currently under way (Med-Vantage Inc., 2006). These programs vary in the number and types (e.g., process or outcome) of quality indicators used, the clinical conditions targeted, the magnitude of the incentives and how they are structured (e.g., as a competition in which only the top providers receive a bonus or as an award based on performance relative to a common benchmark), and whether the program applies to a large or small share of a provider’s patients. Several examples are described below to illustrate the diversity of approaches.
In California, seven health plans are coordinating pay-for-performance programs under the auspices of the Integrated Healthcare Association (IHA), a multistakeholder coalition (www.iha.org). The seven plans consist of 225 physician groups representing about 35,000 physicians treating 6.2 million patients (IHA, 2006a). Bonuses are awarded to large, multispecialty physician groups based on clinical process measures from the Health Plan Employer Data and Information Set (HEDIS), patient experiences of care,
TABLE 2-2 IHA Pay for Performance 2006 Measurement Set
(measurement year 2006, reporting year 2007)
Information Technology Investment
(Each activity is worth 5%. The maximum 20% credit requires four activities, at least two of which must be from measure 2.)
SOURCE: IHA, 2006a.
and investments in technology and infrastructure. Table 2-2 lists the specific measures targeted for IHA’s 2006 program (the fourth year of the program) in three domains and the weights used to determine the share of the total possible bonus that is allocated to each domain. While performance measures are common across the seven plans, the structure of the bonus varies; most plans have opted to reward only the top performers (e.g., the top deciles or quartiles) using a bonus that is proportional to the number of the plan’s patients cared for by the group. Additionally, for the 2006 measurement/2007 reporting year, IHA encourages health plans to reward year-to-year improvement (IHA, 2006b).
Some large employers, through coalitions, are also beginning to offer direct rewards for physician performance on health care quality measures. One example is Bridges to Excellence (www.bridgestoexcellence.org), which operates in four markets and involves a collaborative effort among several large employers, including General Electric and Verizon Communications. Bridges to Excellence offers physicians who become certified by
the Diabetes Physician Recognition Program1 $100 for each diabetic patient in their panel. This program requires physicians to document performance on a number of process and outcome measures through medical record review. Similarly, the Heart/Stroke Physician Recognition Program2 has been launched in selected markets. Finally, through Physician Practice Connections,3 doctors can receive up to $55 per patient for establishing clinical information systems in their offices that aid in regular follow-up for chronically ill patients and for implementing patient education programs.
Pay-for-performance programs are being used in both the health maintenance organization (HMO) and preferred provider organization (PPO) settings. Since 1999, the Hawaii Medical Service Association, the local Blue Cross and Blue Shield affiliated plan and the largest health plan in the state, has rewarded the physicians in its PPO network based on quality measures. In 2003, individual bonuses ranged from $500 to $20,000 (Landro, 2004). These bonuses represent about 5.5 percent of the physician’s overall salary (Rosenthal et al., 2004). Anthem Blue Cross and Blue Shield of New Hampshire’s plan pays bonuses based on a variety of measures that assess appropriate primary and secondary prevention, including screening for breast, cervical, and prostate cancer; screening of patients with coronary artery disease for high cholesterol; and provision of retinal exams for diabetic patients. Anthem’s performance bonus was $20 per patient per year for the top quartile of physicians (about 5 percent of compensation) and about half of that amount for physicians ranked between the 50th and 75th percentiles. Physicians were also eligible for an additional payment of $20 per patient for participating in the plan’s disease management program (Rosenthal et al., 2004).
Some medical groups and independent practice associations incorporate incentive programs into their payment methods. For example, the Hill Physicians Medical Group, one of the nation’s largest independent practice associations, puts up to 15 percent of physician compensation at risk based on quality performance (PBGH, 2005). The program looks at clinical measures (including IHA measures plus other HEDIS measures), information technology functionality, and patient experience. In 2005, Hill Physicians received $5.9 million in funds under the IHA program, but actually distributed $26 million in performance rewards (Hill Physicians, 2005).
The Experience of the United Kingdom
Standing in contrast to the pay-for-performance programs in the U.S. commercial insurance market is the recent General Practitioner (GP) contract with Britain’s National Health Service (NHS) (Roland, 2004; Smith and York, 2004). The arrangement instituted under this contract awards a substantial portion of compensation according to performance on 146 quality indicators. The program targets physician practices, which generally have fewer than five physicians, rather than individual physicians, and pays according to overall performance using a balanced scorecard (i.e., points are awarded for each of the 146 indicators, and the total score is then used as the basis for payment). In addition to performance bonuses, practices are provided with subsidies for infrastructure improvements, as well as additional staffing. The plan is to put approximately 18 percent of GP income at risk, to be distributed subsequently on the basis of performance measures. Financial penalties for persistent low performance are also planned for future years.
Of note are several important differences between the NHS and Medicare that relate to the ease of implementation and effectiveness of a pay-for-performance program. Every NHS patient must register with an individual practitioner who assumes responsibility for that patient’s care. Additionally, the United Kingdom is in the process of instituting a uniform national computerized information system that will include capabilities to automate reporting on the specific measures employed. Most physicians in the United States contract with multiple private and public payers, and many plans with pay-for-performance programs do not account for a large portion of a physician’s income. Moreover, the majority of physician award programs in the United States do not put more than 5 percent of compensation at risk (Rosenthal et al., 2004).
Previous programs in the United Kingdom showed positive responses to financial incentives (Smith and York, 2004). A “fundholding experiment” from 1991 to 1998 that gave practitioners fixed budgets for providing secondary care and pharmaceuticals to their patients ultimately resulted in fewer inpatient procedures and reduced patient waiting times. A program in East Kent from 1998 to 2000 defined disease management targets that practitioners had to meet or repay funds. Both of these programs required new money initially; however, the first created incentives for efficiency savings, while the second relied on a reverse withhold to encourage quality improvement.
Common Themes Among Pay-for-Performance Programs
The majority of incentive arrangements target a mix of population-based measures of clinical quality and patient experience measures (Rosenthal
et al., 2004). According to a recent survey, payers are also increasingly providing direct incentives for the adoption of information technology and for performance on cost-efficiency metrics (AIS, 2004).
In almost all cases, pay-for-performance sponsors reward all physicians whose performance exceeds an absolute threshold (e.g., at least 80 percent of patients with coronary artery disease undergo appropriate cholesterol screening) or all physicians above a given percentile rank based on the level of performance. Thus, quality improvement is not explicitly required for the receipt of a bonus, and the incentives to improve vary with baseline performance. With an absolute performance threshold, physicians whose baseline performance is high need only maintain the status quo to receive payment. For physicians with the lowest performance, the award may not be sufficient to balance the cost of making the required dramatic improvement.
Most of the early pay-for-performance programs for physicians targeted primary care domains (although payments were often to multispecialty groups). More recently, payers also appear to be measuring and rewarding the quality of care delivered by specialists. According to a private survey, more than two-thirds of current pay-for-performance programs now cover specialists, including cardiologists, obstetrician/gynecologists, orthopedists, gastroenterologists, otolaryngologists, and general surgeons (AIS, 2004).
Pay for Performance in Medicare
While measurement systems have provided an impetus for continuing improvement in the quality of care in the Medicare program, overall change has been slow. To date, CMS has invested heavily in the collection and reporting of data on the quality of care of health plans, hospitals, and other institutional providers. While information-based approaches continue to evolve, reliance on benchmarking, subtle pressure from purchasers, and the market impact of individual patient choice are unlikely to eliminate the gap between optimal, evidence-based medicine and actual practice. However, these efforts do lay the groundwork for an effective pay-for-performance program by generating critical baseline information and the infrastructure that will serve as the base for the reward system.
A broad policy rationale for a Medicare pay-for-performance program is the opportunity to improve not only the overall quality of care for Medicare enrollees, but also the care provided to other populations. Many quality improvement investments involve fixed costs, such as those for information technology or training, whose benefits will accrue to all patients. In addition, the added market power of Medicare will magnify the importance of the existing pay-for-performance programs of health plans and may have further positive spillover effects if other payers follow the lead of CMS in payment reform, as was the case with prospective payment systems.
The objectives of pay for performance are to:
Encourage the most rapid feasible performance improvement by all providers.
Support innovation and constructive change throughout the health care system to achieve clinical quality improvements, patient-centered care, and efficient use of health resources.
Promote better outcomes of care, especially through coordination of care across provider settings and time, especially in the treatment of chronic disease.
Pay for performance is not simply a mechanism to reward those who perform well; rather, its purpose is to encourage redesign and transformation of the health care system to ensure high-quality care for all. In such a system, all participants—providers, purchasers, and beneficiaries—can potentially benefit.
As pay-for-performance programs go forward, it will be crucial to develop a strong learning system within the Medicare enterprise to ensure successful implementation and ongoing improvement (see Chapter 6). The evidence base to support pay for performance is still emerging (see below) and implementation efforts should encompass extensive testing and evaluation to assess the effects of the new system. While pay for performance appears to induce change in some health care environments, it cannot by itself create either the high-performing health care system or the payment reform envisioned in the previous reports of the Institute of Medicine’s (IOM’s) Pathways to Quality Health Care series. Ideally, the contributions of payment reform should be compared with the outcomes that could be produced by other mechanisms, such as continuing medical education, accreditation, and consumer activation, which may also be linked to financial incentives. Such an assessment was beyond the scope of this study. Other nonfinancial mechanisms, such as public reporting, benefit redesign, and professional and public education, are also critical components of a far-reaching quality improvement strategy. All of these efforts should be aligned with pay-for-performance programs in order to ensure a common goal and synergistic effects.
In designing a pay-for-performance program for Medicare, financial rewards could be directed at providers, beneficiaries, or both. For example, mechanisms could be devised to allow those consumers who improved their lifestyles (to promote better health outcomes) to share with providers in the savings that resulted from the prevention of consequent
and more expensive treatments, such as avoidable hospitalizations. Copays, deductibles, or premiums could be reduced for those beneficiaries who used designated high-performance providers. However, not all beneficiaries have equal access to those providers deemed high performers, and many hospitals that have excellent reputations for certain medical procedures are not necessarily high performers on all dimensions of quality (Jha and Epstein, 2006). While many possible designs for rewarding beneficiaries could be considered, the evidence base is not yet robust enough to be used to determine how consumer rewards should be structured on a broad level. In interpreting its charge and in an effort to set parameters for what could reasonably and competently be accomplished in Medicare in the short term, the committee did not evaluate beneficiary-oriented approaches, important as they may be. Rather, this report focuses on the provider side of pay for performance.
Pay for Performance and Care Coordination
The deficiencies of the current payment system, as previously described, include its inability to recognize, encourage, or even merely pay for the intentional coordination of patient care across settings and time. The failure to measure these transitions was discussed extensively in the first report in the Pathways to Quality Health Care series, Performance Measurement: Accelerating Improvement (IOM, 2006b). In the current health care system, care is often fragmented and not well coordinated. The need for measures for use in evaluating, and ultimately rewarding, the coordination of care is necessary to quality improvement with regard to both monitoring gains in clinical quality and reducing inefficiencies. For example, improved coordination of care management could potentially result in a reduction in hospital admissions (Rich et al., 1995; Bodenheimer, 1999; Bodenheimer and Fernandez, 2005). Care coordination measures that might help achieve this end have been developed by Eric Coleman at the University of Colorado (IOM, 2006b). One measure of care coordination in the treatment of congestive heart failure patients is part of CMS’s Hospital Compare initiative, although it is not among the first 10 measures that hospitals are being encouraged to report. And the Institute for Healthcare Improvement’s 100,000 Lives Campaign includes medication reconciliation for patients being discharged from hospitals (IHI, 2006; Manno and Hayes, 2006). Pay for performance has the potential to act as a catalyst for improved care coordination through program design, making it possible to reward performance based on outcomes by disease, instead of rewarding providers on the basis of individual services at a single point in time. This aspect of pay for performance is elaborated upon further in Chapters 3 through 5.
THE RESEARCH BASE ON PAY FOR PERFORMANCE
The theory behind pay for performance is derived from the basic economic principle that when one pays more for a certain attribute or dimension of a good or service, more of that attribute or dimension is supplied. While this principle tends to hold true in most markets, the introduction of new incentives linked to performance in health care delivery is relatively new. More than 100 pay-for-performance programs have been initiated in the health care arena in the past decade, yet very few studies have assessed their impact empirically (Petersen et al., 2006; Rosenthal and Frank, 2006). Hence, little evidence for the efficacy of pay for performance in the health care setting exists at this time.
Research in the Health Care Sector
In synthesizing the available research literature, the committee identified pay-for-performance studies that demonstrated both positive effects on processes of care (Beaulieu and Horrigan, 2005; Rosenthal et al., 2005b) and negative (but not statistically significant) effects (Beaulieu and Horrigan, 2005). Most studies have failed to demonstrate any significant effects on processes of care (Rosenthal et al., 2005b; Rosenthal and Frank, 2006). However, many of these studies focused on incentives that affected only a small portion of provider income. Early results of experimental projects have also shown that pay for performance can influence positive changes in nontargeted care practices. For example, a physician targeted improvement in immunization rates; as a consequence, documentation of immunizations improved more than the rates themselves (Fairbrother et al., 1999). In general, however, as noted above, a robust literature demonstrating that pay-for-performance strategies lead to improved health outcomes does not yet exist (although this connection may be implicit, as when a measure is evidence-based, such as use of beta-blockers after myocardial infarction). Indeed, one study that examined clinical quality of hospital care for acute myocardial infarction found that performance on process measures accounted for only 6 percent of the variation in 30-day mortality rates (Bradley et al., 2006). The researchers concluded that performance on process measures could not, in this case, reliably predict mortality outcomes. Therefore, the relationship between payment incentives and health outcomes remains uncertain. Overall, however, fewer than 20 studies designed to demonstrate the effectiveness of pay-for-performance programs in improving the quality of care have been conducted (see Appendix B) (Petersen et al., 2006).
The research literature identifies key considerations that require attention when a pay-for-performance program is being designed. Most important is the level of reward necessary to stimulate significant changes in pro-
vider behavior or processes of care. In this connection, it must be noted that, as the literature remains silent on how much additional payment is needed to drive change in the domains of clinical quality, patient-centeredness, and efficiency, it is unclear whether a business case can be made for any and all stakeholders in the health care system for pay-for-performance programs. Second is whether financial rewards by themselves can change practice, or other quality improvement initiatives must be implemented as well to achieve positive results (Beaulieu and Horrigan, 2005; Rosenthal et al., 2005b). Third is the issue of designing a system in which low-performing providers can achieve at least the same rate of improvement as high performers. Further experimentation and research are needed to answer these and related questions (Dudley et al., 2004; Rosenthal et al., 2005a, 2005b).
Research in Other Sectors
Outside of the health care sector, a variety of studies of analogous incentive programs have yielded results that may be instructive (Rosenthal and Frank, 2006). Pay for performance has been widely introduced in the education field, and several recent experiments have documented improvements in test scores and other outcomes under these programs (Grumbach et al., 1999; Lavy, 2002). One of these studies also demonstrated that pay for performance was more cost-effective (produced a larger impact for the same expenditure) than direct subsidies for new programs and additional staff time (Lavy, 2002). Pay for performance has also been incorporated into federal contracts for job training programs. Studies examining these programs have found that pay for performance had a positive impact on the rate of job placement and average earnings of trainees, even after accounting for gaming on the part of contractors. Finally, the most commonly observed use of pay for performance is for executive compensation. Corporate executives are frequently awarded performance bonuses based on measures of profitability or market value. According to recent reviews of this extensive literature, results of studies of executive compensation suggest that pay-for-performance programs, typically in the form of stock-based compensation, cause executives to improve firm value (Murphy, 1999; Mishra et al., 2000). It is also worth noting that many observers have raised concern that these types of incentive programs may potentially inhibit beneficial long-term investments that do not provide returns in the short term.
It is difficult to draw conclusions for the health care sector based on the existing evidence on pay for performance. Findings based on observational
data are suspect, and suitable natural experiments are lacking. The negative studies in the health care literature used small rewards and incentives based on performance relative to peers, so physicians had no way to know what performance levels would ensure a bonus. Overall, past studies have yielded no clear guidance on the appropriate magnitude of performance-based compensation.
MONITORING FOR UNINTENDED CONSEQUENCES
As noted above, more than 100 pay-for-performance programs have been implemented in the health care sector (Med-Vantage Inc., 2006). These initiatives constitute a rich source of experience regarding the impact of this innovation in health care financing, experience that can help answer questions about what works that are asked by all stakeholders. Concrete data with which to assess the benefits and identify the unintended adverse consequences of the approach are increasingly becoming available; quantification of the impact of pay-for-performance programs is possible, however, only if they are evaluated thoughtfully and systematically. Such evaluation requires careful planning.
Evidence for unintended or unexpected consequences of pay for performance outside of the health care arena, such as gaming in return-to-work and school programs, is relatively well established (Burgess and Ratto, 2003; Courty and Marschke, 2004). In health care, if providers are paid based on performance according to outcome criteria, they may attempt to select healthier patients to maximize net revenues. Other possible negative effects of targeted incentives, such as reductions in various dimensions of quality of care in areas not targeted for financial rewards (which may be a particular concern in primary care because of the broad scope of practice), have not been evaluated empirically. While providers for the most part have the best interests of their patients in mind, such unintended adverse consequences may be a real concern. Table 2-3 is a nonexhaustive listing of some of these potential unintended adverse consequences, each of which is reviewed below. Further experience may identify additional concerns.
Improved quality of care overall is a highly desirable goal, but it should not be achieved at the expense of decreased access to care. Access to necessary services forms the foundation for high-quality care. A meaningful decrease in access to care resulting from the implementation of a pay-for-performance program constitutes an unacceptable outcome.
In their efforts to reach performance thresholds that will result in augmented payment, providers may exclude patients from their practices who
TABLE 2-3 Potential Unintended Adverse Consequences of Pay for Performance
Marginalized Comprehensive Integrated Care
Impeded Knowledge Transfer and Innovation
Forestalled Reform Efforts
Denial of high-risk or noncompliant patients
Creation of incentives for tiering
Promotion of management to the measure or condition
Decreased sharing of best practices and misadventures
Withdrawal of providers from Medicare
Ignorance of other possible reform efforts
Shift of costs to the private sector
Creation of unmet demand for “successful” providers
Disadvantage to undercapitalized practices
Diversion of resources from nonmeasured areas of care
Slowed uptake of nonmeasured practices
Stalled progress in quality agenda
Stalled progress in quality agenda
Shift of costs to consumers
are known to be at high risk for adverse clinical outcomes. As the evidence base continues to grow, providers will be better able to identify prospectively those patients likely to respond poorly to their care. Process-based performance measures may exert a similar adverse selection pressure. Noncompliant patients constitute a particularly frustrating group of patients to manage, putting health care providers at risk for poor performance based on measures of both process and outcome. If providers react by avoiding these patients to keep their performance scores high, the result could be restricted access to care and worsened health. This is the case especially for the old or chronically ill; initially at higher risk, their health status is more likely to deteriorate and at a faster rate if their access to care is limited. Therefore, researchers must make the investigation of risk adjustment for performance measures a high priority. For example, pay-for-performance programs might be structured to give greater rewards to providers who treat high-risk patients (see Chapter 4). If pay for performance is to realize its full potential for change, it will be necessary to engage providers in the care of these challenging patients.
The public reporting of provider performance may also contribute to decreased access. As emphasized throughout this report, public reporting is a cardinal feature of health care reform as it enhances transparency. It can be a powerful motivator to guide change in provider behavior and provide consumers with key data on which to base good decisions (Shaller et al., 2003). Both of these effects are thought to result in higher-quality health care, which in turn represents better value (Marshall et al., 2000; Mason and Street, 2006). At the same time, however, there is concern that the public reporting of provider performance could have unintended adverse consequences. Health care consumers, both individual patients and payers for health care services, would likely seek out the high-quality providers. Providers shown to perform at lower levels might opt to reduce their Medicare caseloads in favor of participants in private plans. As a result, some consumers could be denied access to the care they desire.
Previous IOM reports have highlighted disparities in quality of care that occur along many specific dimensions, including geographic region; provider type; and patient age, sex, and ethnicity (IOM, 2002, 2005). Disparate care is, by definition, low-quality care, and pay for performance could exacerbate such disparities. Populations most affected by disparities in health care are cared for disproportionately by undercapitalized providers who are likely to lack the resources necessary to invest in the infrastructure (such as health information technology) needed to facilitate participation in pay for performance. Nevertheless, the health care services they offer
constitute a critical safety net. The same market forces that will operate to improve or eliminate the cohort of providers who perform poorly may leave populations subject to disparities in care with fewer provider options than they had before. Pay-for-performance programs must therefore be carefully designed to identify relationships that exist between populations subject to disparate care and poorly performing providers. Objective assessment will help limit cultural bias in performance measurement.
Marginalized Comprehensive Integrated Care
The application of performance measures in the evaluation of health care for a particular condition (e.g., diabetes mellitus) or preventive service (e.g., breast cancer screening) poses the risk of decreasing performance and thereby compromising the quality of care being provided in areas that are not the focus of pay for performance. Pay for performance could encourage this tendency to manage to the measures, focusing efforts excessively on those measures that yield the greatest financial return. At the same time, however, this could be beneficial by focusing efforts on areas with the greatest need for improvement, such as the treatment of chronic diseases.
Additionally, measures may conflict with one another, ultimately causing harm to patients. This concern reinforces the need, articulated in the Performance Measurement report (IOM, 2006b), to develop a comprehensive set of performance measures as rapidly as possible. The present report articulates the need for measures that reward three key domains of care: clinical quality, patient-centeredness, and efficiency. As noted earlier, a single-minded focus on clinical quality can lead to increased health care costs through overuse of services. A similar narrow focus on efficiency could compromise clinical quality and raise at least the appearance of a fundamental conflict of interest. And performance measures that place undue emphasis on clinical quality or efficiency are unlikely to be patient-centered.
A comprehensive portfolio of performance measures must reflect consensus around the vision of a reformed and integrated health care system designed to achieve the goals articulated in the Quality Chasm report (IOM, 2001). For example, prompt, understandable, and empathetic communication to the patient of the results of a magnetic resonance scan is as important as the technical quality and value of the imaging study itself; ideally, financial incentives should be restructured, based on valid and robust measures of performance, to encourage both.
Impeded Knowledge Transfer and Innovation
In the health care sector, best practices are adopted at a surprisingly and disconcertingly slow rate (Lomas et al., 1993; Bates et al., 2003). While
health care presents a unique set of challenges for practice improvement, innovations that are evidence based and have been demonstrated to improve the quality of care can take in excess of 17 years to become common practice (Balas and Boren, 2000). Delay in the development or implementation of best practices has substantial human and financial costs. Open dissemination of experience is necessary to harness the capacity of the health care industry to improve. Pay-for-performance programs could unintentionally subordinate collaboration to competition. Providers following a more economically directed model of care might hesitate to share successful practice improvement strategies with their competitors, fearing that doing so would put at risk not only the financial incentives offered through pay for performance, but also the competitive advantage that these successful innovations would offer in negotiating with patients and insurers.
It is difficult to know how best to prevent this from occurring. Clearly the business case for cooperation must be made as solidly and quickly as possible so that providers will be motivated to share both successful strategies and barriers to implementation they may identify. Government is limited in its ability to bring about this type of interchange. Entities such as Medicare’s Quality Improvement Organizations might provide a forum for exchange of such information, fostering the creation of a culture of quality improvement (IOM, 2006a).
A separate compelling concern is that pay for performance could inadvertently stifle long-term innovation by shifting the focus of quality improvement exclusively to the achievement of short-term goals. While it is important to reward interventions that result in short-term improvements, it is essential as well not to suppress the experimentation and innovation that can lead to new procedures, applications, and approaches that can generate long-term continuous improvement in quality. Successful pay-for-performance programs must not foster the development of a new status quo that is better, but incomplete.
Pay for performance must be structured to promote higher-quality care and cost control, but not at the expense of driving providers from the health care arena. Provider acceptance will be a large point of contention in any pay-for-performance initiative. If payment under such a program is perceived by providers as unfair, they may become increasingly demoralized. Additionally, if the burden on providers of participating in a Medicare pay-for-performance program is too overwhelming (relative to the potential rewards), providers may withdraw from participation in Medicare, causing serious access issues in some geographic regions in addition to those discussed above. For example, fewer physicians are choosing primary care as
their field, in part because of lower income and increased workload (Moore and Showstack, 2003). An even further decrease in payments could exacerbate this problem. The problem may be compounded if decreased payments lead those already in primary care practices to leave the profession (Sox, 2003). Lack of acceptance may increase if providers are not consulted during the process of program implementation in order to allay some of these concerns.
The committee concludes that most providers have the capacity and desire to improve the care they deliver—an essential component of a successful pay-for-performance program. A lack of acceptance by providers would not only disrupt the management of a pay-for-performance program, but potentially hinder the entire quality agenda. If payers rely on pay for performance as a pathway to improving care and health outcomes, progress along that pathway will be forestalled if a majority of providers refuse to participate. Fundamentally, providers must believe in and accept the system not just because of the rewards they may receive, but also because they believe in its ability to advance the quality agenda. That goal will not be realized if action to achieve the system is delayed by a lack of acceptance of program terms among providers.
Forestalled Reform Efforts
If payers or policy makers focus too intensely on pay for performance as a major solution to the current inadequacies of the payment system, they may fail to recognize other mechanisms that might work as well or better. In the context of the need for pay for performance to be a learning system, as emphasized throughout this report, payers and policy makers should remain aware of other options that could enhance or replace pay-for-per-formance strategies. On the other hand, policy makers could decelerate progress by focusing too much on potential unintended negative consequences, diverting attention from the intended positive consequences of rewarding higher quality and better outcomes to improve the quality of care received by all Medicare beneficiaries. This is not meant to imply that the implementation of pay for performance should proceed without caution, but to emphasize that the possible unintended adverse consequences should not hinder progress.
Assuming the pay-for-performance program will involve a reduction in base payments (see Chapter 3 on use of existing funds), when Medicare pays less for a service, providers could try to shift those unreimbursed costs to the private sector. Cost shifting results when decreased reimbursements
by one payer leads providers (by force of market power) to demand higher rates of payment from another payer (Lee et al., 2003). For example, researchers examining data from the late 1980s through the early 1990s found that lower Medicare payments to hospitals were associated with statistically significant increases in payment rates to hospitals by private payers (Zwanziger et al., 2000). As a result, the burden of some costs may eventually shift directly to the consumer (Gabel et al., 2002; Lee and Tollen, 2002). For example, increased costs to private payers may lead to increased premiums and copays for the consumer, which in turn could contribute to other potential adverse consequences already described, such as decreased access. Thus decreased payments might save costs for Medicare, but lead to an undue burden for private payers and consumers. However, if the main private payers were to follow Medicare’s lead on pay for performance, the opportunity for such cost shifting would be reduced.
Quality improvement is a continuous and dynamic process; caution in the design of pay-for-performance programs is necessary to ensure that successful programs do not foster the development of a new status quo—one that is better, but incomplete. Overall, any pay-for-performance program must be designed as a learning system that will allow for modifications in response to feedback obtained, including unintended positive consequences. Additionally, the program must incorporate mechanisms designed to monitor for unintended negative consequences and allow for rapid correction to prevent any resulting harm. These issues are discussed in more detail in Chapter 6.
A FIRST STEP TOWARD PAYMENT SYSTEM REFORM
The many pay-for-performance programs now planned or in place have been supported with considerable resources and enthusiastic commitment over a remarkably short period of time. Many public policy makers share the enthusiasm of their colleagues in the private sector, and state and federal programs are poised to follow the lead of private insurers. Unfortunately, clear goals, the best intentions, and a substantial investment of human and financial capital do not guarantee success in the implementation of a pay-for-performance program.
The urgency of the quality problem in the environment of the current payment system demands that steps be taken now to align payment with the six quality aims of the Quality Chasm report (IOM, 2001) and improve health outcomes. The committee recognizes that no perfect payment strategy has yet been identified to advance these goals. However, the economic
rationale behind and early experience with pay for performance suggests that it offers an initial and interim pathway to change, although more fundamental restructuring of the payment system may be necessary in the future. Therefore, the committee makes the following recommendation:
Recommendation 1: The Secretary of the Department of Health and Human Services (DHHS) should implement pay for performance in Medicare using a phased approach as a stimulus to foster comprehensive and systemwide improvements in the quality of health care.
The committee concludes that as pay for performance is implemented, the aim should be to move toward rewarding comprehensive care as soon as possible, instead of focusing payments and rewards on individual episodes of care in isolated settings. In this sense, the committee envisions that as pay for performance evolves, shifts should occur from rewarding process measures toward rewarding outcome measures, and from rewarding by setting toward rewarding by health condition. Initial pay-for-performance programs will be limited by the availability of reliable measures and the structure of the current payment system. Therefore, for example, while the availability and reliability of measures may necessitate an initial focus on rewarding process more than outcome measures and rewarding by setting instead of by condition, the committee foresees this balance shifting over time. It will be important to use research and evaluation techniques to identify milestones by which this shift should be encouraged. The details of a phased implementation are discussed in more detail in Chapter 5.
This chapter has examined the promise of pay for performance as a lever in the redesign of the current health care payment system. The current system does not reward high-quality care, nor does it provide incentives for providers to improve their performance in care delivery. Medicare expenditures have continued to grow rapidly, while improvements in the overall quality of care have not kept pace. Pay for performance holds promise as one component of a redesigned approach to create incentives for improving the quality of care. Many pay-for-performance initiatives have been undertaken in both the public and private sectors, offering some initial feedback, but a larger effort is needed to create changes on the scale necessary to improve the nation’s health care system. Medicare wields sufficient power to act as a leader in this effort. While pay for performance appears to hold much promise, the committee cautions that it alone cannot reform the health care system. Additionally, a pay-for-performance
program of this scale must be an evolving learning system that can adapt to knowledge gained and monitor for unintended negative consequences. The following chapters describe how pay for performance in Medicare could be designed. Chapter 3 addresses funding alternatives, Chapter 4 issues surrounding the distribution of those funds, and Chapter 5 specific details of program implementation.
AIS. 2004. Case Studies in Health Plan Pay-for-Performance Program. Washington, DC: Atlantic Information Services, Inc.
Balas EA, Boren SA. 2000. Managing clinical knowledge for health care improvement. Year-book of Medical Informatics 2000: Patient-Centered Systems. Stuttgart: Schattaver Verlasgesellschaft. Pp. 65–70.
Bates DW, Kuperman GJ, Wang S, Gandhi T, Kittler A, Volk L, Spurr C, Khorasani R, Tanasijevic M, Middleton B. 2003. Ten commandments for effective clinical decision support: Making the practice of evidence-based medicine a reality. Journal of Medical Informatics Association 10(6):523–530.
Beaulieu ND, Horrigan DR. 2005. Putting smart money to work for quality improvement. 40(5 I):1318–1334.
Bodenheimer T. 1999. Disease management—Promises and pitfalls. New England Journal of Medicine 340(15):1202–1205.
Bodenheimer T, Fernandez A. 2005. High and rising health care costs. Part 4: Can costs be controlled while preserving quality? Annals of Internal Medicine 143(1):26–31.
Bradley EH, Herrin J, Elbel B, McNamara RL, Magid DJ, Nallamothu BK, Wang Y, Normand S-LT, Spertus, JA, Krumholz HZ. 2006. Hospital quality for acute myocardial infarction: Correlation among process measures and relationship with short-term mortality. Journal of the American Medical Association 296(1):72–78.
Burgess S, Ratto M. 2003. The role of incentives in the public sector: Issues and evidence. Oxford Review of Economic Policy 19(2):285–300.
CMS (Centers for Medicare and Medicaid Services). 2006a. National Health Expenditures by Type of Service and Source of Funds. [Online]. Available: http://www.cms.hhs.gov/NationalHealthExpendData/02_NationalHealthAccountsHistorical.asp#TopofPage [accessed April, 18, 2006].
CMS. 2006b. Rewarding Superior Quality Care: The Premier Hospital Quality Incentive Demonstration. [Online]. Available: http://www.cms.hhs.gov/HospitalQualityInits/downsloads/HospitalPremierFS200602.pdf [accessed June 9, 2006].
CMS. 2006c. Prospective Payment Systems—General Information. [Online]. Available: http://www.cms.hhs.gov/ProspMedicareFeeSvcPmtGen/ [accessed May 31, 2006].
Cohen JW, Spector WD. 1996. The effect of Medicaid reimbursement on quality of care in nursing homes. Journal of Health Economics 15(1):23–48.
Courty P, Marschke G. 2004. An empirical investigation of gaming responses to explicit performance incentives. Journal of Labor Economics 22(1):23–56.
Dudley R, Frolich A, Robinowitz DL, Talavera JA, Broadhead P, Luft HS. 2004. Strategies to Support Quality-Based Purchasing: A Review of the Evidence. Rockville, MD: Agency for Healthcare Research and Quality.
Fairbrother G, Hanson KL, Friedman S, Butts GC. 1999. The impact of physician bonuses, enhanced fees, and feedback on childhood immunization coverage rates. American Journal of Public Health 89(2):171–175.
Gabel J, Levitt L, Holve E, Pickreigh J, Whitmore H, Dhont K, Hawkins S, Rowland D. 2002. Job-based health benefits in 2002: Some important trends. Health Affairs 21(5): 143–151.
Ginsburg PB, Grossman JM. 2005. When the price isn’t right: How inadvertent payment incentives drive medical care. Health Affairs w5.376.
Grumbach K, Selby JV, Damberg C, Bindman AB, Quesenberry C Jr., Truman A, Uratsu C. 1999. Resolving the gatekeeper conundrum: What patients value in primary care and referrals to specialists. Journal of the American Medical Association 282(3):261–266.
Hill Physicians. 2005. Hill Physicians Medical Group: 2005 Annual Report. [Online]. Available: http://www.hillphysicians.com/HPMG_AR2005.pdf [accessed August 3, 2006].
Hodgkin D, McGuire TG. 1994. Payment levels and hospital response to prospective payment. Journal of Health Economics 13(1):1–29.
IHA (Integrated Healthcare Association). 2006a. IHA News Release. [Online]. Available: http://www.iha.org/020606.htm [accessed August 3, 2006].
IHA. 2006b. IHA P4P Measurement Set. [Online]. Available: http://www.iha.org/p4pyr4/fmy4.pdf [accessed August 3, 2006].
IHI (Institute for Healthcare Improvement). 2006. Getting Started Kit: Prevent Adverse Drug Events (Medication Reconciliation). [Online]. Available: http://www.ihi.org/NR/rdonlyres/47D5AE1C-0B29-4A59-8D58-BABF8F4E829F/0/ADEHowtoGuideFINAL5_25.pdf [accessed June 8, 2006].
IOM (Institute of Medicine). 2001. Crossing the Quality Chasm: A New Health System for the 21st Century. Washington, DC: National Academy Press.
IOM 2002. Insuring Health—Care Without Coverage: Too Little, Too Late. Committee on the Consequences of Uninsurance, eds. Washington, DC: The National Academies Press.
IOM. 2005. Quality Through Collaboration: The Future of Rural Health. Washington, DC: The National Academies Press.
IOM 2006a. Medicare’s Quality Improvement Organization Program: Maximizing Potential. Washington, DC: The National Academies Press.
IOM. 2006b. Performance Measurement: Accelerating Improvement. Washington, DC: The National Academies Press.
Jha AK, Epstein AM. 2006. The predictive accuracy of the New York state coronary artery bypass surgery report-card system. Health Affairs 25(3):844–855.
Kaiser Family Foundation. 2005. Medicare Advantage Fact Sheet. Washington, DC: Henry J. Kaiser Family Foundation.
Kautter J, Pope GC, Trisolini M, Bapat B, Olmsted E, and Klosterman M. 2004. Physician Group Practice Demonstration Bonus Methodology Specifications. Unpublished.
Landro L. 2004, September 17. Booster shot: To get doctors to do better, health plans try cash bonuses. The Wall Street Journal.
Lavy V. 2002. Evaluating the effect of teachers’ group performance incentives on pupil achievement. Journal of Political Economy 110(6):1286–1317.
Lee JS, Tollen L. 2002. How low can you go? The impact of reduced benefits and increased cost sharing. Health Affairs w290–241.
Lee JS, Berenson RA, Mayes R, Gauthier AK. 2003. Medicare payment policy: Does cost shifting matter? Health Affairs w3, 480–488.
Lomas J, Sisk JE, Stocking B. 1993. From evidence to practice in the United States, the United Kingdom, and Canada. The Milbank Quarterly 71(3):405–410.
Manno MS, Hayes DD. 2006. Best-practice interventions: How medication reconciliation saves lives. Nursing 36(3):63–64.
Marshall MN, Shekelle PG, Leatherman S, Brook RH. 2000. The release of performance data: What do we expect to gain? A review of the evidence. Journal of the American Medical Association 283(14):1866–1874.
Mason A, Street A. 2006. Publishing outcome data: Is it an effective approach? Journal of Evaluation in Clinical Practice 12(1):37–48.
Med-Vantage Inc. 2006. Pay for Performance. [Online]. Available: http://medvantageinc.com/Content/solutions.p4p.php4 [accessed June 9, 2006].
MedPAC (Medicare Payment Advisory Commission). 2005. MedPAC Data Runs. Washington, DC: MedPAC.
MedPAC. 2006. Report to the Congress: Medicare Payment Policy. Washington, DC: MedPAC.
Mishra CS, McConaughy DL, Gobeli DH. 2000. Effectiveness of CEO pay-for-performance. Review of Financial Economics 9(1):1–13.
Moore G, Showstack JA. 2003. Primary care medicine in crisis: Toward reconstruction and renewal. Annals of Internal Medicine 138(3):244–247.
Murphy K. 1999. Executive Compensation. Handbook of Labor Economics. New York: Elsevier Science Pub. Co.
PBGH (Pacific Business Group on Health). 2005. Aligning Physician Incentives: Lessons and Perspectives from California. [Online]. Available: http://www.pbgh.org/programs/documents/PhysIncentivesReport_09-01-05final.pdf [accessed August 2, 2006].
Petersen LA, Woodard LD, Urech T, Daw C, Sookanan S. 2006. Does pay-for-performance improve the quality of health care? Annals of Internal Medicine 145(4):265–272.
Premier Inc. 2006. HQI Demonstration. [Online]. Available: http://www.premierinc.com/all/quality/hqi/ [accessed June 9, 2006].
Remus D. December 2005. Pay for Performance—CMS/Premier Hospital Quality Incentive Demonstration Project (Year 1 Results). [Online]. Available: http://www.premierinc.com/all/quality/hqi/results/year-1-results-participant-calll.pdf [accessed June 9, 2006].
Rich MW, Beckham V, Wittenberg C, Leven CL, Freedland KE, Carney RM. 1995. A multidisciplinary intervention to prevent the readmission of elderly patients with congestive heart failure. New England Journal of Meddicine 333(18):1190–1195.
Robinson JC. 2001. Theory and practice in the design of physician payment incentives. The Milbank Quarterly 79(2):149–177.
Roland M. 2004. Linking physicians’ pay to the quality of care: A major experiment in the United Kingdon. New England Journal of Medicine 351(14):1448–1454.
Rosenthal MB, Frank RG. 2006. What is the empirical basis for paying for quality in health care? Medical Care Research and Review 63(2):135–157.
Rosenthal MB, Fernandopulle R, Song HR, Landon B. 2004. Paying for quality: Providers’ incentives for quality improvement. Health Affairs 23(2):127–141.
Rosenthal M, Hsuan C, Milstein A. 2005a. A report card on the freshman class of consumer-directed health plans. Health Affairs 24(6):1592–1600.
Rosenthal MB, Frank RG, Li Z, Epstein AM. 2005b. Early experience with pay-for-performance: From concept to practice. Journal of the American Medical Association 294(14):1788–1793.
Shaller D, Sofaer S, Findlay SD, Hibbard JH, Delbanco S. 2003. Perspective: Consumers and quality-driven health care: A call to action. Health Affairs 22(2):95–101.
Shen J, Andersen R, Brook R, Kominski G, Albert PS, Wenger N. 2004. The effects of payment method on clinical decision-making: Physician responses to clinical scenarios. Medical Care 42(3):297–302.
Smith PC, York N. 2004. Quality incentives: The case of U.K. general practitioners—An ambitious U.K. quality improvement initiative offers the potential for enormous gains in the quality of primary health care. Health Affairs 23(3):112–118.
Sox HC. 2003. The future of primary care. Annals of Internal Medicine 138(3):230–232.
SUNY. May 2001. Fact Sheet: Medicare Basics. [Online]. Available: http://www.canton.edu/can/can_start.taf?page=ota_medicare_facts [accessed May 31, 2006].
U.S. GAO (United States Government Accountability Office). 2005. Medicare Physician Payments: Considerations for Reforming the Sustainable Growth Rate System. Washington, DC: U.S. Government Printing Office.