Download the full testimony (pdf)
Chairman Spratt, Ranking Member Ryan, and distinguished Members of the Committee, I thank you for the opportunity to testify on the topic of value in health care. Arguably, few other topics are more important to your work. As the Congressional Budget Office Director has testified, the long-run budget problems are largely driven by health cost growth. Medicare, Medicaid, and other health program spending comprise about one-fourth of the federal budget. Their rapid projected growth accounts for the entire long-run federal fiscal deficit. It is an economic as well as a budget issue. Health spending accounts for 16 percent of our economy—more than housing or food.Its rapid growth poses challenges to businesses and individuals whose income is increasingly devoted to paying for health care. And, despite the enormous investment in health care, the quality of that care and its outcomes fall short by most standards. As the Federal Reserve Board Chair Ben Bernanke recently said, “Improving the performance of our health-care system is without a doubt one of the most important challenges that our nation faces.”
In this testimony, I would like to suggest how value in the health care system can be improved and why the opportunity to do so is on the horizon. “Value” generally describes the perceived quality of care or benefit per dollar spent. Improving value is not necessarily synonymous with improving efficiency. Some aspects of care provision, such as its patient centeredness, are worthwhile to patients but not strictly efficient. Nonetheless, the United States spends an enormous amount on duplicative, low-utility, and even harmful health care, so that a high-value health system would be more efficient overall—and offer significant non-economic advantages as well.
Elements of a Value-Oriented System
There is no “silver bullet” for improving value in health care, precisely because of the nature of health care. The exacting rules that govern fields like engineering and physics do not apply to human health. Illnesses and therapies evolve rapidly, with new diseases and cures introduced each year. Basic economic rules also fit some aspects of health care poorly. Technological advances that typically lower labor costs have instead raised them by increasing the reliance on highly-paid health care specialists. Mass production that has revolutionized other sectors has no real foothold in health; health care jobs now outnumber manufacturing jobs. Moreover, people still trust their doctors to define their demand (i.e., diagnose it) as well as fulfill it.People believe high-cost care equals high-quality care, despite evidence to the contrary. And, they undervalue disease prevention and overvalue disease “heroics” or intense medical interventions to reverse disease—reflecting the values and beliefs that also shape our health system.
The nature of health care makes it is impossible to draft a single, perfect health care system: It would not work for all providers and people, and even if it did, it would be obsolete quickly.It also means that classical market solutions do not neatly apply to health care. Demand is complicated, providers sit on both the supply and demand sides of the equation, and suppliers—primarily insurers—have little incentive to promote value when payers and enrollees come and go.What is needed instead, in my opinion, is a strong infrastructure through which changing information, best practices, and preferences can be channeled.This infrastructure consists of standards for high-quality, cost-effective care, networks for transferring these standards throughout the system, and policies for their adoption, described below.
Standards for High-Value Health Care: In certain respects, the United States leads the world in health care. The National Institutes of Health, its universities, and its private-sector labs, have produced medical breakthroughs that have benefited millions and are used worldwide.Some insurers and payers of care have used this information to successfully shape the delivery of care. For example, most health plans use the Healthcare Effectiveness Data and Information Set that measures plan and provider performance on key quality indicators, with proven success. And, some providers have developed feedback systems to inform both the basic research as well its adaptation for daily practice. For example, Kaiser Permanente’s monitoring of its own enrollees detected the increased risk of heart attacks associated with Vioxx and dropped its coverage of the drug—contributing to its withdrawal from the market.Despite evidence of the benefits and tradeoffs for many if not most treatment options, no authoritative synthesis of such information exists. Instead, different and sometimes conflicting standards are used across the nation, propagated by specialty societies, some government programs, insurers, consumer websites, and regional coalitions.
The consequences of few standards for high-value health care and coverage are troubling. Lack of knowledge of recommended therapies likely contributed to their provision only 55 percent of the time. Another study found that only 41 percent of primary care doctors were familiar with national guidelines for blood pressure treatment, although awareness increased the probability of recommended action. Competing standards also affect performance.For example, a study by the Institute of Medicine found the six major federal health programs had different quality standards, creating unnecessary confusion and burdens for providers.A typical doctor has public and privately insured patients, making the adherence to the different coverage and quality reporting onerous.As can be seen in the studies of practice patterns, the variability is greater for services with a weaker evidence base. For example, the landmark Dartmouth study that found no better quality or outcomes in high-cost areas attributed the excess costs to greater use of testing and evaluative services as well as use of the hospital as a site of care; use of major surgical procedures and minor non-discretionary services were not significantly different. In the absence of evidence on benefits and costs, people and often providers assume that more is better even when it may be wasteful or harmful.
Creating a standard-setting process to guide health care decision making could improve value in the health system. This process could identify services and delivery system practices for which there is: (a) strong evidence for or against their use; (b) strong evidence on the tradeoffs of substitutes; or (c) weak evidence. It could also identify research gaps that should be prioritized to promote a high-performing health system. In particular, studies on the comparative clinical and cost effectiveness of different treatment options could be encouraged. These standards would neither constitute “cook-book medicine” nor the elements for a defined-benefits plan.For example, the process would not weigh in on resource allocation and who pays for care. Instead, these standards would advance a common understanding of the state-of-the-art health care practices—the basic building block for improving performance.
Information Exchange Networks: Dissemination is as critical as the development of standards. Standards can only affect performance if they reach the remote parts of our health system, and vice versa: Information and expertise from all parts of the health system are needed to set the standards. Historically, knowledge among providers and managers has been shared through annual conferences, continuing medical education programs, journals, and specialty societies. Disease registries, where information is collected on certain types of patients and treatment protocols, have proven to assist in both education and adoption. The rapid evolution of information technology has also facilitated dissemination and adaptation at all levels of the health system. “Learning networks” and Regional Health Insurance Exchanges have been created to harmonize data collection and reporting at the local level. Provider and consumer decision support tools have proliferated. And the interest in implementing a nationwide electronic medical record is strong.
Yet, rather than simplifying the system, the explosion of communication tools has sometimes increased chaos. Internet-based journals, physician and health plan resources, and consumer resources (e.g., WebMD) abound. A patchwork of registries and data bases has placed time-consuming and sometimes expensive demands on providers and organizations to participate. Entrepreneurs have entered the space, offering electronic health records, support systems, and feedback tools. Meanwhile, there is no evidence that these advances have shortened the years for a proven treatment to move from the lab bench to the bedside. A recent study found that only 4 percent of physicians used a complete electronic medical record, with an additional 13 percent using a basic system.
One key step to increasing value in health care is creating a national, health information technology infrastructure to facilitate development and dissemination of best practices. Beyond its potential administrative savings, information technology could build in prompts, reminders, and error warnings at the point of service. The Veterans’ Administration health system has used technology in this way with positive results. In addition, a national, privacy-protected electronic health record would provide data for studies on the comparative effects of clinical and delivery system interventions on a wide-scale basis. This could make the health system more efficient over time by limiting the adoption of new therapies that offer less benefit than existing ones. Technology is an essential but not the sole source for the exchange of information.The heavy reliance on judgment and experience in health care delivery supports the idea of building regional peer networks. Like specialty societies, they would provide the latest research and data feedback, but would do so with an understanding of the local context, culture, and health system resources.
Tools to Promote Adoption of High-Value Care: Arguably the greatest challenge to promoting value is ensuring that the participants in the system adhere to proven standards. Knowledge of these standards alone can help. But, evidence suggests it is not enough. Regular blood testing is a well-known standard of care for diabetics, yet only 24 percent of participants in a national study had three or more glycosylated hemoglobin tests over a two-year period. Conversely, there is little evidence supporting the use of CT scans for management of heart disease yet a recent article documented the rapid increase in their use and thus costs.
Part of this pattern can be explained by reimbursement rates. Value is rarely taken into account when determining whether and what a provider gets paid. Payment rates usually only account for a service’s cost, not its benefits—promoting high-cost health care irrespective of its merit. This may explain why there is higher adherence to standards of care for procedures (which tend to have high reimbursement) versus counseling (which tends to go unreimbursed). Similarly, the amount that patients pay in cost sharing is typically pegged to a service’s cost rather than its value.Simply stated, financial incentives for providers and patients are misaligned.
Beyond financing, the lack of organization of the delivery system diffuses the accountability for producing value. Studies have found that having an organizational culture that promotes quality results in high performance ratings for providers. Yet, most doctors still practice alone or in small groups and lack the critical mass to implement and connect to larger systems to improve the value of the care for their patients.Ideas to remedy this range from linking all providers to a hospital to forging “interdependent practice organizations” that assume responsibility for members’ performance. Beyond their ability to invest in system supports, organizations could also have a social network effect on provider behavior, which has recently been found to be powerful in reducing obesity and tobacco use.
Lastly, an often-overlooked tool in improving value is making it the path of least resistance. A growing literature suggests that making the desired behavior the default improves the odds of achieving it. For example, the use of beta blockers after a heart attack is the standard of care. Research has found greater use of this drug among patients to whom it was prescribed in the hospital discharge orders—not leaving it to the patient to fill the prescription independently later. Moreover, some hospitals automatically prescribe beta blockers on the discharge order, allowing the doctor to take it off the order, but asking for an explanation why. This system reduces the required steps needed to achieve the desired result. Across the board, payment and delivery systems could be designed so that high-value care is the easiest choice for individuals and providers.
Policies to Create the Infrastructure for a Value-Oriented System
This infrastructure—standards for value, information exchange networks, and tools for its use—could undergird different mixes of public and private insurance.It does not depend on either an exclusively public or private insurance system to work. As such, it could be incorporated into a number of different health reform plans. However, a key to achieving a high-value health system is seamless coverage for all Americans: High-value care cannot be initially or consistently applied when one in three individuals falls out of the system for at least a month over a two-year period. Similarly, inadequate coverage—a problem for 25 million insured Americans according to a recent study—results in cost-related barriers to care and coordination and communications problems which interfere with value-oriented care. As Henry Aaron has put it, ensuring adequate coverage for all Americans is, “a precondition for effective measures to limit overall health care spending.”
That said, some of the infrastructure for a value-oriented health system could be put into place in the context of incremental reform. These components are described below.
Investing in Comparative Effectiveness Research: A prerequisite to assigning value in health care is knowing various services’ relative impact. “Comparative effectiveness research” is the rigorous assessment of the relative safety, effectiveness, and cost of treatments or approaches for addressing the same condition. This type of research has been funded by the Agency for Healthcare Research and Quality’s appropriations, but at a fraction of the amount authorized in the Medicare drug law enacted in 2003. Proposals to significantly increase comparative effectiveness research funding and ensure its independence have support from a wide range of businesses, consumer groups, and experts, including the health advisor to George H.W. Bush. Bipartisan legislation has been introduced by Reps. Allen and Emerson and a version of it was included in the Children’s Health and Medicare Protections Act of 2007 that passed the House but was vetoed by President Bush. The Congressional Budget Office estimated that this provision, which created a trust fund seeded by public and private funding, would save the system $6 billion over 10 years and reduce federal spending by the 10th year. The 111th Congress should enact this legislation since this information is essential to setting standards for value.
Creating a Federal Reserve-Like Board to Set Standards: Another policy to consider is the creation of an independent board to promote high-value health care. Composed of experts with long terms, this board would be modeled on the Federal Reserve Board which has succeeded in making crucial decisions with greater credibility than most federal agencies. This board would be an authoritative source of information on the value and tradeoffs of health care services and delivery mechanisms. Because of the breadth of health care, the board would focus on new and high-cost services. To ensure it complements rather than replaces existing efforts, it could give its imprimatur to publicly and privately developed standards (for example, U.S. Preventive Services Task Force guidelines; the AHRQ’s Evidence-Based Practice Centers; the National Quality Forum; specialty societies’ protocols). Its assessments of high-value health care would be accessible to payers, providers, patients, and the public.
The board would also assess the optimal mode for delivering high-value care. This function may be best carried out regionally. The Federal Reserve has 12 district banks whose governance includes key stakeholders as well as experts. They are responsive to their regional resources and climate—features that could be valuable in promoting value given the geographic variation in health care. Regional “health value” boards could tap into medical leadership to tailor their work to region-specific problems. They could gather data, analyze it for patterns, and feed the results back to providers and facilities with comparisons to local, regional, and national process and outcome measures. They may be better able to gain the trust and change the behavior of local providers than a national board. Several states have already developed regional consortia to promote quality and efficiency. These boards could also be built from the current Quality Improvement Organizations in Medicare. Medicare could support regional boards by providing data as well as incentives (or requirements) that providers participate in them.
Accelerating the Use of Health Information Technology: This Congress may succeed in enacting legislation that creates standards, privacy protections, and funding for the implementation of electronic health records. The bipartisan legislation is necessary but probably not sufficient to yield rapid adoption of EHRs in a short period of time. Physicians may still be resistant given their inability to capture the return on the investment; private plans may worry about losing a competitive edge; and the benefits that result may make it more of a public good than private commodity. Congress should consider making the president’s aspiration that most Americans have an EHR by 2014 a deadline. It could enforce this requirement in a number of ways, including lower or no Medicare payments to providers who do not comply. Loans and grants would likely be needed to assist in meeting this deadline. The ongoing activity to set standards for interoperability and privacy and create data exchanges to support EHRs would need to be stepped up. Other nations have already made the switch from paper-based to electronic systems; it is feasible as well as essential to optimizing health system performance.
Allowing Medicare to Align Policies with Value: Even though it funds less than 20 percent of the health system, Medicare’s policies have often set the standard for the private sector. The shift to a new standard based on value could be led by Medicare as well. Congress could delegate authority to Medicare to adopt payment policies that the Medicare Payment Advisory Commission recommends based on the value-oriented standards set by the new board. These changes could include adopting successful “pay for performance” models, creating bundled payments across providers and/or services, and adjusting patient cost sharing to promote high-value care and discourage low-value care. Such changes could be allowed within boundaries; for example, the authority could be limited to modifications that reduce spending within the budget window according to the Congressional Budget Office. The Medicare Trustees might also take a bigger role in program operation, having to approve the policies recommended by the program administrator. Congress could always override the changes, but the default would be flipped: instead of having to wait for Congress to align payments with value, Medicare would do so unless Congress blocked it.
In addition, Medicare payment systems build in some funding for capital improvements; this funding could be directed toward system design to facilitate high-value care. For example, Medicare could incentivize hospitals to develop or adopt computer-assisted reminder or default order systems that have proven effective at improving adherence and outcomes. It could also add the use of effective, simplifying systems as a condition of accreditation; arguably, they are as important to safety and the system as a facility’s cleanliness or doctor attendance at medical staff meetings.
Prioritizing Prevention: Lastly, the gravity of the problem of preventable disease, coupled with the inadequacy of the existing system, suggests that a new model is needed to prioritize wellness. To be effective, it should strive to make preventive services valued by individuals and providers, available, and affordable. It should elevate wellness within the health system and complement it with new delivery systems. Payment for prevention should be designed to leverage behavioral change and widespread use. Finally, it should be universal, providing recommended prevention services irrespective of individuals’ insurance status.
A Wellness Trust is one approach for structuring an effective prevention system. Under this model, preventive services would be carved out of the health insurance system and financed through a new independent agency. The Wellness Trust would set national priorities for prevention, employ unconventional systems for delivering services, use payment policy to drive results, and integrate prevention with the health care system through information technology. Congress could lay the groundwork for this approach by creating the Trust, assessing prevention spending, reviewing priorities, and developing a prevention workforce.
The imperative for improving value in the health system is strong, and the opportunity to do so may be near. The next Congress and president face inescapable tax and budget – as well as health policy – decisions. A number of expiring policies will be waiting on the doorstep. These include: the 2001 and 2003 tax cuts, the escalating taxes due to the Alternative Minimum Tax problem, and a budget that will likely be unresolved in 2008. Tax and budget reform represents an opportunity for health reform. Responsible tax policy to replace the expiring Bush tax cuts could build in revenue to fund upfront health system changes. Modifying while maintaining the tax break for employer health benefits could redirect high-income tax breaks to low-income tax credits. And, the cost savings inherent in health reform are essential to long-run budget stability.
No doubt, enacting health—and budget and tax—reform is hard. Yet the only thing harder may be turning a blind eye while our nation’s health and economic prospects fade. Incremental reform can lay the groundwork for a high-quality, efficient, equitable health system; the policies described here take steps toward it. But small changes may take as much political capital as big ones. A strong infrastructure must be combined with coverage in a seamless system supported by sustainable financing to achieve the potential of a high-value health system.
Download the full testimony (pdf)