The Patient Protection and Affordable Care Act of 2010 (Obamacare) creates several new Medicare programs intended to improve health care quality, using “pay-for-performance” payment strategies to put financial pressure on medical providers. In such programs, reimbursement reflects provider performance on metrics based on adherence to certain care processes, scores on patient satisfaction surveys, or patient outcomes. The rationale behind pay for performance is the result of a real problem: Payment for medical services, particularly by the large government health programs, does not reflect value or benefit for patients. To address this issue, the United States should move toward a genuine market-based payment system, rather than simply perpetuating flawed financing structures.
The Patient Protection and Affordable Care Act (PPACA) of 2010 creates several new Medicare programs intended to improve health care quality, using “pay-for-performance” payment strategies to put financial pressure on medical providers. In such programs, reimbursement reflects provider performance on quality metrics based on adherence to certain care processes, scores on patient satisfaction surveys, or patient outcomes.
The rationale behind pay for performance is the result of a real problem: Payment for medical services, particularly by the large government health programs, does not reflect value or benefit for patients. To address this issue, the United States should move toward a genuine market-based payment system that would lead physician and hospital payment to reflect quality naturally, rather than simply perpetuating flawed financing structures. Current arrangements encourage doctors and hospitals to do more tests and procedures, and they often receive even higher pay when patients experience complications.
The new Medicare initiatives contained in the PPACA, or Obamacare, fail to alter the systemic factors that are responsible for quality issues in the first place. They are thus unlikely to have the desired effects on provider behavior. Meanwhile, there are many instances in which pay-for-performance initiatives can even adversely impact patient care or decrease quality. Members of Congress, who enacted these measures, should rethink this strategy and the role of the federal government in guiding health policy, with a view to realigning incentives in order to secure higher quality and better value.
Pay-for-Performance Programs in Obamacare
The Medicare pay-for-performance programs enacted in the PPACA pay individual providers based on their past performance. This strategy to improve health care quality has been tried, discussed, and debated among health policy analysts over the past decade. The concept appears to be a logical approach, and its presentation even suggests that it is rooted in free-market ideas. It is grounded in the notion that providers should compete against each other based on quality and the overall value of their services, and that payment for health care services should reflect value, not volume. Such objectives would be naturally achieved in a free market, if one existed today in health care.
In fact, however, the Medicare pay-for-performance strategy is not market-driven; it is a strategy to replace the function of a market with government management of health care delivery. This approach will not solve the problem of sluggish quality improvement; nor will it drive patients to better value care. It will, however, introduce perverse new incentives into the delivery of health care that direct resources away from real improvement and even harm quality.
1. The In-Patient Value-Based Purchasing Program. Value-based purchasing is Obamacare’s main pay-for-performance quality-improvement mechanism. The program began in October 2012 and is intended to financially incentivize hospital performance improvement by reducing Medicare’s diagnosis-related group (DRG) payments for all hospitals, then redistributing the savings according to hospital performance. Hospital performance scores reflect overall achievement compared to other hospitals, as well as improvement from year to year. The quality measures that are used to rank hospitals are drawn from Medicare’s pay-for-reporting program, which went into effect in 2004 and serves as the precursor to value-based purchasing. In its first year, value-based purchasing measured performance with 12 indicators of clinical processes that reflect adherence to treatment guidelines, as well as the results of patient-satisfaction surveys. In the second year of value-based purchasing, emphasis on adherence to process indicators will be reduced by including measures of outcomes, including mortality rates.
Improving value in hospital care is a natural imperative for health care reform, but the value-based purchasing program in the PPACA is a flawed strategy for achieving this goal. Evidence shows that the strategy is ineffective at improving outcomes, despite its moderate effect on process adherence. The program is rife with incentives for hospitals to focus on improving their performance scores without actually improving the quality of patient care, and its narrow focus will lead hospitals to direct resources to narrow areas of care, reducing the level of improvement in other areas of need.
The Premier Hospital Quality Incentive Demonstration. The best example of how value-based purchasing will impact health care is the Premier Hospital Quality Incentive Demonstration, which began in 2003. Under the demonstration, Medicare offered financial incentives for high performance on measures related to five common conditions. Of the 34 quality indicators used, 27 measured processes and seven measured outcomes. The program went through two iterations; from 2003 to 2006, only the highest-achieving hospitals received bonus payments; thereafter, hospitals were rewarded for both high achievement and improvement, a process similar to today’s value-based purchasing program.
Both independent studies and the Congressional Budget Office (CBO) concluded that the Premier demonstration showed no evidence of improving outcomes, as measured by 30-day mortality rates, and had no significant impact on Medicare spending. In addition, the changes to the program made in 2006 were intended to encourage low performers to improve, even though they might not be able to surpass already high-performing institutions. Value-based purchasing is modeled after this second iteration, even though studies show it did not have its intended effect.
While participating hospitals did improve their performance on process measures, by the end of the demonstration, the rest of the nation’s hospitals had also improved following the introduction of required Medicare hospital quality reporting in 2004. After five years, one study shows, performance at participating and non-participating hospitals was “virtually identical.” Regarding the Premier demonstration, the CBO concluded, “The best available evidence indicates that the demonstration was responsible for small increases in quality of care and that most of the increases in quality that occurred at the participating hospitals would have occurred in the absence of the demonstration.” Ironically, then, a program designed to improve quality using evidence of what works is grounded in little evidence that it itself will work.
Payment reflects patient characteristics. Beyond the likelihood that it will be ineffective, value-based purchasing has the potential to actually harm the quality of patient care. Performance scores are calculated based on both overall achievement, relative to other hospitals, and improvement relative to the same hospital’s score in previous years. This is done to avoid penalizing safety-net hospitals, teaching hospitals, and other providers that may score low on quality metrics because of the complex cases and demographics of their patients. Punitive reimbursement for these low-ranked hospitals could reduce the resources available for investment in quality improvement, reducing quality and increasing outcome disparities among different races and ethnic groups.
Nevertheless, the methodology for calculating value-based purchasing performance weights achievement higher than improvement, putting low-performance hospitals and those serving certain patients at a distinct disadvantage. For hospital achievement, the program measures all participants equally; in other words, a certain level of achievement at one hospital would be scored equal to the same level of achievement at a different hospital. The same is not true for scoring improvement. As Drs. William Borden and Jan Blustein explain, hospital improvement on quality metrics is measured according to an “elastic ruler,” such that “initial low-performing hospitals have a wider improvement range and, thus, need a greater absolute score increase to achieve the same improvement score as an initial high-performing hospital.”
If low performance on quality metrics is influenced by patient demographics and case complexity, then, logically, the reverse is also true. Hospitals serving healthier, wealthier patients and providing less complicated care may perform better and receive bonus payments, not due to better quality, but due to “better” patients. Even in its first year, a disproportionate number of physician-owned and specialty hospitals were among the top-performing hospitals across the nation. Critics of these care delivery models blamed their high performance scores on the patient population they serve and have called for their exclusion from the program as a result of their high performance.
But the problem is not physician-owned and specialty hospitals. These institutions often do deliver better value due to their ability to streamline care, increasing efficiency, and offering quality care at lower prices than general hospitals. The problem is value-based purchasing itself; if the program penalizes hospitals for taking care of sicker, poorer patients and rewards those serving the opposite, its measures do not solely reflect the quality of care provided by hospitals; they reflect patient characteristics. Paying hospitals differentially will not improve value unless hospitals assuming responsibility for more complicated cases receive more payment, and hospitals that care for less complicated patients are paid less to reflect proportionate gains in efficiency. As Harvard researchers explain, “If the business model of general hospitals today can be separated into its component value propositions with distinct business models of care delivery, and the payment system properly rewards each for their work, what seems to be cherry picking today will in reality be recognized as the efficient distribution of resources.” This is impossible under a system of administrative pricing, such as the traditional Medicare program, even with value-based modifications.
Perverse incentives detract from real quality. The measures used in the first year of value-based purchasing reflected areas of care where performance was already high due to years of quality reporting by Medicare. Under the new program, measures are to be removed once they become “topped out”—meaning when there is little room left for significant improvement. Officials at the Centers for Medicare and Medicaid Services (CMS) explain that these measures are to be removed in order to avoid unintended consequences, including “inappropriate delivery of a service to some patients (such as delivery of antibiotics to patients without a confirmed diagnosis of pneumonia), unduly conservative decisions on whether to exclude some patients from the measure denominator, and a focus on meeting the benchmark at the expense of actual improvements in quality or patient outcomes.”
Even so, for 11 of the 12 clinical care process indicators used in 2012, the achievement threshold was greater than 90 percent, meaning that hospitals must adhere to the indicator 90 percent of the time to receive any bonus at all. Since payment depends on hospital scores for this narrow subset of quality metrics, the program encourages large investments to achieve relatively insignificant improvements, solely to obtain a moderately improved score. The small amount of improvement possible from this kind of endeavor would not likely have a significant benefit for patients. According to Tufts Medical Center researchers, “A hospital with 97.5% compliance may be penalized, and it may take a significant financial expenditure and use of staff resources to increase that compliance from 97.5% to 98.5% with minimal or unclear gains to patients.”
For many hospitals, improving in the measured areas of care reduces time and resources that could be invested in areas of care in greater need of attention. Conversely, removing topped-out measures might allow high performance to diminish over time, as attention moves to performance on new measures. The Kaiser Permanente health system experienced performance decline for diabetic retinopathy and cervical cancer screenings between 1997 and 2007, following the removal of financial incentives for performance on these measures of care.
Finally, a narrow focus on quality measures can harm quality by incentivizing care that is not appropriate for certain patients. A number of current measures were initially introduced under the Surgical Care Improvement Project (SCIP) in 2006. Under this and programs like it, according to Tufts researchers,
measures are rolled out before their full impact is assessed, using live hospitals as the testing ground and relying on the individuals trying to comply with these measures to troubleshoot. When issues do arise that require the measures to be changed, response times are invariably at least 6 months; meanwhile patients may be at risk, and measures are consistently failed…. At our institution, even a small number of misses can result in major losses in compensation.
Under SCIP, the initial measures were too rigid to dictate sensible decision making in many clinical situations. For example, one quality measure indicated use of one class of antibiotic to prevent surgical infection and another if the patient was allergic to the first-line choice. The alternative drug, however, had the potential for severe side effects. A third, unlisted class of antibiotic might have been more appropriate, pitting physicians’ choices among one drug that would cause an allergic reaction, another with potential for severe side effects, or a third that may work better for the patient but would lower the quality score. While the specific measures used in SCIP have since been refined, similar experiences are to be expected with greater reliance on quality metrics for defining and rewarding quality in medicine.
Insufficient financial gain to drive quality improvement. As one study predicted, the financial amounts of even the largest value-based purchasing bonuses and penalties are fairly insubstantial, amounting in most cases to less than 1 percent of Medicare payment. This makes it doubtful that value-based purchasing will be able to drive the change needed in care delivery. The balance between the cost of investing in quality improvement and the financial benefit is likely to be even less attractive to hospitals where performance on quality metrics is already low and improvement would require significant investment. While the size of the incentive payment will increase as the program moves forward, so will the breadth and complexity of the quality measures used to assess performance. As a result, despite the potential for perverse and unintended consequences, value-based purchasing will still prove an insufficient driver of change.
2. The Hospital Readmissions Reduction Program. The Hospital Readmissions Reduction Program (HRRP) is a variation of the pay-for-performance strategy. Rather than offering incentive payments, it penalizes hospitals with high 30-day readmission rates for three conditions. Penalties are determined based on a comparison of a hospital’s performance to the national average, adjusting for clinically relevant factors, such as patient demographics, comorbidities, and patient frailty. Readmission rates are calculated using discharge data for each hospital from the three years prior to the year in which the penalty is assessed. In the first year of the program, which began in October 2012, the maximum penalty was 1 percent of total Medicare reimbursement; in 2013, it will increase to 2 percent, and in 2014, to 3 percent. In the first year, roughly two-thirds of hospitals were penalized.
Readmission rates reflect external factors. The goal of the HRRP is to reduce readmissions, which are considered an indicator of poor quality of care. However, like many other outcome measures, the quality of hospital care is not solely responsible for high readmission rates, which also reflect a patient’s socioeconomic status, complexity of illness, and the availability of other health resources in the community. Despite the attempt to adjust for factors that fall outside hospitals’ control, concerns remain that providers caring for sicker, poorer patients are disproportionately penalized under the program. One study has already shown that in the first year of the program, a higher percentage of the penalized hospitals were large hospitals, teaching hospitals, and safety-net hospitals. As with value-based purchasing, reducing payment based on factors that reflect patient profile could decrease the availability of funding for quality improvement investments, making it more difficult for hospitals that care for patients with more complicated needs to show improvement.
Hospital readmissions also reflect the level of care patients receive outside the hospital. It thus seems odd that hospitals should assume responsibility for keeping discharged patients out of the hospital, when a stronger role for others involved in a patient’s care—including primary care physicians, case managers, and insurers—might have more of an impact. Availability of these resources clearly also influences readmission rates. For example, calculated 30-day readmission rates might be higher in an area if patients have access to better outpatient services, which would keep healthier patients out of the hospital in the first place and leave only the sickest patients, who ended up in the hospital, in the denominator of the calculation of a hospital’s readmission rate.
Readmission rates do not always signal low quality. Another issue is that readmissions do not necessarily signal poor-quality care. A disease management program conducted by the Brisbane Cardiac Consortium for inpatient and post-discharge congestive heart failure patients sought to increase use of evidence-based guidelines to improve processes of care; the program was successful in reducing mortality rates, but, unexpectedly, readmission rates actually increased. This same paradox was further noted by Cleveland Clinic clinicians in a study that showed that while the Cleveland Clinic has lower mortality rates for heart failure than the rest of the nation, its readmission rates are higher, indicating that taking better care of more patients and preventing deaths may increase readmission. In another Cleveland Clinic study, no evidence was found of a strong association between a hospital’s performance on mortality and readmission rate for acute myocardial infarction or pneumonia, although there was a modest inverse relationship for heart failure. Whether or not low mortality rates cause high readmission rates, it seems clear that they are not necessarily tied to low quality.
Readmission rates are not always preventable. Even if a hospital does succeed in providing the highest quality of care, some readmissions simply are not preventable. Researchers estimate that 23.1 percent of 30-day unplanned readmissions are potentially unavoidable. Meanwhile, the CMS goal for the Hospital Readmissions Reduction Program is to reduce 30-day readmission rates by 20 percent by the end of 2013. This would require a 91 percent reduction among those readmissions that are avoidable, which may be unrealistic.
Reducing preventable hospital readmissions through better care coordination, discharge planning, medication adherence, and increased use of outpatient services is important to achieving cost-control and quality-improvement goals. However, a federal program to achieve this mission through strict management of hospital care is misguided. Comparisons of the quality of care experienced by patients covered by traditional Medicare versus Medicare Advantage shows that Advantage enrollees experience lower readmission rates, despite using the same hospitals and physicians as other Medicare patients. This indicates that the best way to achieve this goal may be through a new insurance model and engagement of stakeholders other than hospitals in readmission reduction.
3. The Physician Value-Based Payment Modifier. Under the PPACA, Medicare will administer another pay-for-performance program for physicians through a modification of the existing Medicare fee schedule: the value-based payment modifier. This new fee adjustment will be applied to Medicare physician reimbursement beginning in 2015. Medicare physician payment will be adjusted to reflect performance using quality data from the Physician Quality Reporting System, and cost data from Medicare fee-for-service claims. In 2015, the value-based payment modifier will be applied to group practices with 100 or more “eligible professionals,” based on quality data reporting from 2013. In 2017, the modifier will apply to individual and small group practices, using quality and cost data from 2015.
While nurse practitioners, physician assistants, and other health care professionals are included in determining the number of eligible professionals in a practice, only Medicare payment to physicians is subject to adjustment by the modifier. The program is budget neutral for the federal government, like value-based purchasing, which means there will, by necessity, be winners and losers. At the onset, practices that meet the reporting requirements, either by reporting on one or more measures from the PQRS group practice reporting option or by electing administrative claims reporting, can choose to receive no pay adjustment, or a pay adjustment based on a composite score reflecting quality and cost data. Eligible practices that do not meet the reporting requirements will face a penalty of 1 percent of Medicare reimbursement.
Evidence that pay for performance improves the value of care offered by physician groups or individual physicians is even weaker than it is for hospital-based programs. Dr. Robert Berenson of the Urban Institute points out that process indicators used to measure physician performance reflect a small portion of a physicians’ professional activities, and large measurement gaps exist that are unlikely to be filled. For example, it is not possible to measure a physician’s ability to make a correct diagnosis or choose an appropriate intervention, taking into account each patient’s clinical condition and personal preferences, using only administrative claims data. In short, “the numerator of the value equation—quality—captures too little of any physician’s performance on quality, while the denominator—cost—cannot be accurately attributed to an individual physician.”
A Better Way to Foster Quality Improvement
Incentives matter in health care, and subjecting health care providers to financial pressure is necessary to encourage meaningful, widespread, and lasting quality improvement. Real market competition can achieve these goals, allowing providers of medical care to secure a larger patient base by offering the best value. With transparency in outcomes and other quality indicators that matter to patients, individuals can take advantage of the best that medical professionals have to offer before an episode of care takes place. Within a system of administrative pricing, pay-for-performance instead offers financial incentives to improve quality through slight alterations to reimbursement after care has already been provided. Rather than making quality improvement imperative to a provider’s existence—building a strong “business case” for quality—this strategy provides a slight slap on the wrist for those who fail to keep up.
The main structural flaw of the federal pay-for-performance approach is that it does not involve patients in the drive for better value. The programs allow patients to continue behaving as they normally would when choosing providers, since patients do not share in any of the savings from finding and using high-value health care providers. The goal of these programs is for the payer—Medicare—to achieve better value by retroactively paying more when, on the whole, high-quality care was received and less when low-quality care was received. A more effective, patient-centered way to promote improved quality is to revamp the financing and delivery of American health care such that each individual experiences incentives to seek out value in the system, and has both the necessary assistance and proper tools to do so.
Congress can advance these goals by first acknowledging that pay-for-performance programs are not the best option for improving value in Medicare or driving improvement in health care quality. Congress should instead begin to build a more workable system on the basis of real consumer choice and genuine free-market competition.
Congress can encourage the creation of a system that incentivizes patients to actively seek out value by reforming the third-party payment system that today isolates patients from the true cost of health coverage, as well as from health care goods and services. When patients are unaware of the true cost of their insurance, they are also unaware of the cost of health care delivery; not to mention that many health plans, especially those sponsored by employers and the government, tend to downplay the role of cost sharing.
Obamacare uses Medicare to push for quality improvement because of the large program’s potential to impact provider behavior. Congress can alternatively leverage this aspect of the program by reforming Medicare to put patients in charge of their insurance decision making and enable them to seek out and reward value in the health care delivery system. Under several reform proposals, a new “defined-contribution” model would replace the current “defined-benefit” program with financial assistance to allow seniors to choose an approved health plan that offers both dependable coverage and affordability. In this kind of system, insurers would see incentives to guarantee high-quality care and lower costs, or risk losing their customer base. This model has proved successful in Medicare Advantage, and lessons learned from this existing alternative to traditional Medicare form an excellent basis for reforming the rest of the program.
Other changes, including removing the existing preferential tax treatment for employer-sponsored coverage and encouraging defined-contribution financing models for the working, non-elderly population, would further advance the alignment of incentives among insurers and providers with what is best for the patient.
When insurers sell coverage directly to patients and compete for their business, they will face new incentives to offer value, which can be best accomplished by directing patients to medical professionals who have demonstrated that they can and do offer proven high-quality care at lower cost. This is something some insurers have already begun to do. Unlike unpopular health maintenance organization (HMO) models, which controlled costs by delaying or even refusing care, such a market-based system would hold insurers accountable to patients for guaranteeing that they receive appropriate care at a competitive price. Insurers would also have incentives to develop ways to assist patients in identifying and accessing high-value care.
Finally, patients will need tools to identify value, including consumer-friendly information on quality, and true price transparency. Professor Regina Herzlinger of the Harvard Business School, one of the nation’s top health care economists, has suggested the creation of a consumer information center similar to the Securities and Exchange Commission (SEC) to give patients access to unbiased, consolidated information. Some of this information is available now, but is often incomplete or has questionable reliability. Instead, “the health care version of the SEC could collect the data and make it public while the private sector can analyze the data, producing reports and information that are meaningful to consumers.”
While Obamacare’s pay-for-performance programs for Medicare are either in the early stages of implementation or have already gone into effect, policymakers will continue to debate whether this approach can create the degree of quality improvement needed in the delivery of American health care. In addition to the major issues highlighted above, still others have been noted, including the complexity of the program structure, use of quality metrics that do not reflect aspects of care that matter to patients, and the simple fact that, though measures undergo rigorous review and alteration, they cannot take into account the real-world variability and dynamism of medical practice.
Solving these shortcomings through tweaks to existing programs is a temporary, palliative measure. At some point, no matter how many alterations are made to the program structure, it should be clear that the pay-for-performance method is not the best one for achieving value. Creating a true, quality-driving health care system requires starting at the beginning and addressing the root of the problem, not its symptoms. Doing so will require massive change and new direction for health care reform, but is necessary for bringing about the quality improvement and strides toward better value that Americans need. Lawmakers should address the underlying systemic factors—the absence of robust market forces—that contribute to slow improvement in health care quality, and encourage market-based reforms that enable patients to discern high-quality, lower-cost providers, and entrust them with their care.
—Kathryn Nix is a second-year medical student at the Medical College of Wisconsin, and a former Heritage Foundation graduate fellow in health policy.