|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Improving Quality Through Public Disclosure Of Performance Information
Despite a growing consensus that serious quality problems afflict U.S. health care, state and federal governments have done little to improve the quality of care. Proposed health insurance reforms, including a Medicare prescription drug benefit and the use of tax credits for insurance expansion, could create a mechanism for stimulating and then monitoring improvements in quality. We propose legislative requirements that any new expenditure of federal funds for health benefits be accompanied by public disclosure of performance information regarding quality, effectiveness, and safety. Such disclosure could yield diverse public and institutional benefits.
A series of major national reports have focused attention on numerous pervasive and serious quality problems in the U.S. health care system.1 At the same time, state and federal governments face continuing pressure to raise payment rates, expand coverage, and increase investment in the system. How can public officials responsibly manage and even expand federally financed health care coverage if as much as one-fourth of covered services may be wasteful or even dangerous?2 There have been few systematic national proposals to address these quality problems. A recent Institute of Medicine (IOM) report encouraged health care organizations and purchasers to implement policies that increase the likelihood that health care is "safe, effective, patient-centered, timely, efficient, and equitable."3 The IOM exhorted federal purchasers and others to develop financing and payment mechanisms that supported these goals, but it did not offer specific, practical approaches for doing so. Congress has not devoted substantial resources to improving the quality of U.S. health care services. The annual budget for the Agency for Healthcare Research and Quality (AHRQ) represents only 1.3 percent of the National Institutes of Health (NIH) budget, for example. The Centers for Medicare and Medicaid Services (CMS) contracts with peer review organizations (PROs), allocating approximately $235 million per year (in FY 2000) for quality oversight and improvement in support of the $218 billion Medicare program (one-tenth of 1 percent), and the inspector general for health and human services (HHS) recently indicated that the PRO oversight process is dangerously inadequate.4 There appear to be few mechanisms available for public policy to influence quality of care. The most common intervention in national health policy since the mid-1990s has been the incremental expansion of publicly financed health insurance. The 1996 Temporary Assistance for Needy Families (TANF) welfare reform legislation, the creation of the State Childrens Health Insurance Program (SCHIP) in 1997, and the Health Insurance Portability and Accountability Act (HIPAA) of 1996 demonstrated the political feasibility of increasing coverage but did not explicitly address quality problems. The political consensus that changes to our health care system will occur through modest reforms in coverage policy has created a paradox for advocates of safer, more effective, more patient-centered care. In a highly fragmented, decentralized marketplace with minimal federal quality oversight, how can we encourage (or compel) better-quality care? In an era when each congressional initiative is narrowly crafted to avoid provoking opposition from major institutions and associations, what will prompt those institutions to change how they deliver care? By wrestling only with narrowly defined problems in financing, Congressand by extension society as a wholehas been unable to promote systemic change in health care quality. These incremental reforms can be made to serve the interests of improved quality through a simple device: Any major change in federally funded health benefits should require the appropriate agency to receive and publish standardized reports regarding the quality, safety, and effectiveness of care delivered with any public funds authorized by that legislation. Such reports should be developed through a public-private collaboration that includes participation by consumers and other likely users of the information. Such a requirement could trigger a thorough public-private process for developing measures, mechanisms, and reports. The governments role would be merely to demand accountability; it could look to industry and public stakeholders to develop appropriate ways to satisfy that demand.
Theoretical underpinnings. The theoretical basis for using disclosure as a means of improving quality is long-standing. Justice Louis Brandeis famously wrote that "publicity is justly commended as a remedy for social and industrial diseases. Sunlight is said to be the best of disinfectants; electric light the most efficient policeman."5 More recently, Justice Stephen Breyer argued that disclosure provides an intermediate means of addressing public risks without creating impossibly complex regulatory systems or reducing the beneficial effects of choice in the marketplace.6 William Sage carefully analyzed how "regulating through information disclosure" might apply to contemporary health care.7 He noted four possible benefits of this approach: (1) to enable both group and individual purchasers to reward superior performance in a competitive market; (2) to inform consumers and policymakers about the policies and practices of their agentsemployers, health plans, doctors, and othersthat may affect their care (for example, gag rules, informed consent, financial incentives, licensing, and accreditation status); (3) to inform providers and plans about how their performance compares with that of others so that they may plan and focus improvement activities; and (4) to inform the public and its representatives about the performance and equity of a vital element of national life and a major national expense. Sage concluded that a disclosure strategy dovetailed nicely with the expectations and behavior of the baby-boom generation and the prevailing model of government regulation. Practical applications. The basic idea that quality measurement is fundamental to building more effective programs has gained widespread report. In education policy, President George W. Bush has insisted on annual academic assessments, coupled with a National Assessment of Educational Progress in fourth and eighth grades, as part of his No Child Left Behind program.8 In the private sector, performance measurement systems are used in many areas under the auspices of such organizations as the American National Standards Institute and the Financial Accounting Standards Board.9 The large body of theoretical work on behalf of federal disclosure requirements has been made practical in numerous recent examples including the Emergency Planning and Community Right to Know Act (EPCRA) and the Ryan White CARE Act of 1990.10 Similarly, Congress has required that the Indian Health Service and the Maternal and Child Health Block Grant programs use the Healthy People 2010 indicators to monitor the effectiveness of federally funded programs.11 The most thoroughly evaluated reporting effort is the New York State cardiac surgery program, which has published provider-specific outcome reports since 1989. Analysts note that the program has been associated with reductions in operative mortality at almost twice the rate of the national trend, that providers (both hospitals and physicians) with better outcomes had higher rates of growth in market share, that poorly performing surgeons have left the state, and that more than a third of New York cardiologists say that the data have affected their referrals to surgeons "very much" or "somewhat."12 The New York program has had its skeptics, and some of its consequencessuch as some surgeons greater reluctance to take on high-risk casesare controversial.13 But the very nature of these debates and the remarkable attention given to the data by cardiologists, surgeons, and administrators demonstrate the value of the initiative. Until recently, these disclosure initiatives were conjured up by experts and regulators to serve the interests of oversight bodies. Only secondarily was their value for the public discovered. In creating SCHIP, Congress created a new and powerful model that engages both the expert and consumer communities in a joint undertaking to assess and improve the quality of care provided to children.
SCHIP was created by the Balanced Budget Act (BBA) of 1997. The main focus of the law was the provision of funds to the states to increase the enrollment of low-income children in comprehensive health insurance plans. One simple requirement in the enabling legislation has stimulated a broad, national initiative to measure and improve the quality of care for children and adolescents. The law requires each participating state to "submit to the Secretary an evaluation that includes...a description and analysis of the effectiveness of...the quality of health coverage provided including the types of benefits provided."14 The CMS (then HCFA) subsequently issued regulations specifying the reporting requirements, allowing much state flexibility while also outlining detailed measurement guidelines: Sec. 457.710(d) provides that the State plan must describe how performance under the plan will be measured through objective, independently verifiable means and compared against performance goals. For purposes of measurement, States may find it helpful to conceptualize performance in two broad categories: quality of care measures and program operations measures. Quality of care measures focus on access to care, health status and delivery of clinical services. A measure of performance in either category must be valid (that is, reflect the concept it is intended to capture) and reliable (that is, yield results that are reproducible in repeated analyses).15 These regulations then guided the private-sector efforts to develop an accountability reporting system. At the instigation of AHRQ, several private foundations and nongovernmental agencies came together to form a national consortium to develop tools to measure the quality of childrens health care. The consortium, known as the Child and Adolescent Health Measurement Initiative (CAHMI), rapidly grew to comprise more than eighty national organizations, including medical societies, consumer organizations, federal and state government agencies, and research centers.16 States put considerable pressure on the CAHMI consortium to develop simple, affordable measurement tools in time to meet their federal reporting obligations. A CAHMI executive committee, comprising academics, consumer advocates, and industry leaders, established broad priorities and criteria. A technical staff worked closely with about twenty-five clinical and measurement scientists to develop three sets of quality measures based on published, validated research instruments. The initial measurement sets addressed the quality of pediatric care for young children (ages 03), care for children with chronic illness, and preventive services provided to adolescents (ages 1217). The intended usersMedicaid-enrolled parents of young or chronically ill children and teenage enrolleesreviewed the proposed surveys and measures in detail. Each measurement set was then fielded and validated in several states, in multiple organizational settings (such as Medicaid and commercial managed care, primary care case management, communitywide), using multiple data sources (parent survey, chart abstraction, claims data). These measurement systems ultimately were published in the peer-reviewed medical literature and industry measurement guides.17 States were interested in CAHMI largely as an efficient means of solving a burdensome technical problemhow to provide the HHS secretary with evidence of program effectivenessbut quickly recognized the additional benefits of using the new quality information. In just the first year following publication of the quality measurement tools, twenty-three states began to use one or more of the recommended tools (1) to conduct policy analysis and respond to their executives or legislatures concern about the rapid shift of enrollees from Medicaid fee-for-service (FFS) into managed care, (2) to support health plan contracting negotiations, (3) to publish consumer "report cards" to begin the move to consumer-choice benefit models, (4) to screen and identify families appropriate for targeted services, and (5) to provide feedback to providers (individual doctors and clinics) on areas where their performance was below norms or established targets. Most major national health care surveys such as the National Health Interview Survey and the Medical Expenditure Panel Survey have adopted the CAHMI surveys as a de facto standard. One measurement setaddressing care for children with chronic illnesseswas adopted by the National Committee for Quality Assurance (NCQA) as part of the Health Plan Employer Data and Information Set (HEDIS) national managed care reporting system. In 20022003 quality reports are being provided to consumer advocates in local communities to be shared directly with individual families and physicians. Each of these uses of the child health quality measurement system helps to foster a uniform national definition and assessment of high-quality care for children. Federal officials, state legislators, providers, health plans, families, and researchers are all using common words and measures to describe and evaluate care. The nations policymakers now have a consistent tracking system for monitoring how changes in the health care system are affecting the quality of care delivered to and even the health of Americas youth.
Americans spent an estimated $132 billion on prescription medications in 2000, yet we have little evidence regarding the value or overall impact of that expense. For example, Americans spent $8.2 billion in 2000 on cholesterol-lowering medications; in one recent study only 15 percent of patients on such drugs had achieved the target cholesterol level.18 A Medicare prescription drug program would bring substantial new funds into a medication distribution system fraught with serious inefficiencies and errors. Errors occur in 324 percent of all prescriptions dispensed.19 Approximately 40 percent of all patients do not complete necessary treatment programs (for example, by taking pills for only half as long as necessary to achieve the therapeutic effect, effectively wasting the medicine already dispensed and taken).20 Physicians fail to prescribe necessary medications or prescribe inappropriate medications as much as half of the time, for common situations such as cholesterol reduction and care for heart attack survivors.21 Any new prescription drug program should incorporate mechanisms to inform Congress and the public about the benefits or harms associated with the program and create incentives to prescription drug plans, providers, and consumers to make effective use of the new benefit.22 Given the likely inflationary trends in pharmacy costs, the new program must anticipate a continuing debate about how to ensure that the new benefit delivers value and is well managed over time. Public disclosure of plan performance. Under a Medicare drug benefit program, public disclosure of information about the performance of prescription drug plans could serve multiple objectives.23 Performance data could assist Medicare beneficiaries in selecting the plan most likely to meet their needs; help the government determine which plans should be made available to the public (meeting minimum standards); permit Congress to determine if beneficiaries are receiving health benefits that justify the new federal expenditures; and create incentives for drug benefit plans to invest in and focus management effort on delivering safe, effective, and high-quality service. These objectives can be realized if the drug benefit legislation requires prescription drug plans to report to HHS on the quality, effectiveness, and safety of benefits provided. The HHS secretary should be required to establish reporting specifications within twelve months of enactment, based on consultation with appropriate private- and public-sector experts and organizations. The statute could direct the secretary to consider specific criteria in establishing reporting requirements, including that the information be made available to the public, that it be relevant and understandable to beneficiaries, that the experiences and concerns of beneficiaries be reflected in the program design, that the information address both the effective operation of the prescription drug benefit (prescribing, dispensing, payment) and the achievement of intended health benefits, and that the reporting system impose the least possible administrative burden on prescription drug plans and providers. The development of this reporting program could be assigned to a technical agency such as AHRQ or the Food and Drug Administration (FDA), which would, in turn, convene public and private stakeholders to recommend the standards. As was the case with SCHIP, an advisory commission might recommend two sets of plan performance reports: one regarding the outcomes achieved by the covered benefit and one evaluating the operation of the plan. Public disclosure of health outcomes. Political pressures might prompt Congress to allocate substantial public funds to pay for a drug benefit. Such expenditures would be authorized in the hope of improving the health and functioning of Medicare beneficiaries. Is it inappropriate to ask those administering the benefit to assess whether beneficiaries are achieving the intended health outcomes? For example, approximately five million older Americans suffer from congestive heart failure (CHF), a condition associated with a weakening heart muscle. CHF is the leading cause of death in persons over age sixty-five. Appropriate and sustained use of medicines can dramatically improve the quality of life for CHF patients while also reducing their rate of hospitalization and associated costs. Yet inappropriate medications are now often prescribed, approximately one-fourth of CHF patients receive medications in inadequate doses, about one-third of patients fail to continue taking their medication, and up to 60 percent of all CHF patients are not receiving the necessary medications at all.24 A qualifying drug benefit plan could be required to report on (1) the number of patients with a diagnosis of CHF who are prescribed and receive angiotensin converting enzyme (ACE) inhibitors, (2) the number who continue to receive therapeutic dosage levels of ACE inhibitors over time, and (3) their functional health status. The infrastructure and information systems prevailing today would not permit drug plans to be held accountable for outcomes. But prescription drug legislation would provide a key opportunity to improve that infrastructure. Tracking of outcomes and persistence of appropriate medication use would require that the initial prescription reflect the diagnosis or indication the medications are intended to address. With this single additional data item, policymakers, managers, and care coordinators could evaluate the numbers and types of patients failing to get full benefits from the medications being provided and paid for from public funds. Public disclosure of dispensing accuracy. An advisory commission would also be likely to recommend measures of operational performance, especially in areas of direct concern to beneficiaries. For example, studies estimate that as many as 1524 percent of prescriptions filled at retail pharmacies involve errors.25 These errors include dispensing the wrong medication, or the wrong dose strength, or the wrong number of pills or using the wrong instructional label on the packaging. One easily understood operational measure is "dispensing accuracy": the percentage of prescriptions filled that exactly reflect the prescribing physicians instructions. The advisory commission might recommend that all dispensing pharmacies participating in the drug plan report annually on their dispensing accuracy, using a standard audit methodology. Use of this reporting requirement would favor those pharmacies that invest in computerized or other systems of safety verification and allow patients to seek service from drug plans that help their pharmacy network adopt safer practices. This simple standard is likely to reduce medication errors, thereby reducing morbidity and costs in the Medicare population.
There are certainly numerous challenges to the successful implementation of a quality-information strategy. The prescribing physician has the greatest control over selecting medications and educating patients in their proper use, yet the probable drug coverage mechanisms are unlikely to have contractual or regulatory reach to individual doctors. Some measures would require that results be risk-adjusted or stratified for different populations. Any new data collection process entails a new obligation and cost to those subject to its mandates and should only be imposed where the public benefits and system efficiencies ultimately gained outweigh the implementation costs. Yet we have learned much about how to negotiate these hurdles over the past decade. The New York surgery data, the HEDIS measures, and the CAHMI process have each developed technical and political techniques that have adequately addressed such problems. Previous initiatives, including the publication of Medicare hospital mortality data, managed care process data, and patient satisfaction results, have led to only modest changes in behavior by consumers, purchasers, and providers.26 One tracking study noted an 18 percent increase from 1996 to 2000 in the number of Americans who would choose their surgeon based on quality ratings and a 7 percent increase in the number who would choose a hospital based on ratings.27 Fifteen percent of corporate purchasers consult HEDIS data when selecting a managed care plan.28 Several studies have found that physician and managed care organizations have changed clinical practices in response to published performance data.29 These evaluations illustrate the potential of public data to change behavior but also point to the limitations of the early efforts. In almost all cases, the value of public reporting was reduced because the information was not considered relevant to the quality concerns of the expected users.30 Future performance measurement initiatives must fully involve consumers and other users in the design and specification of the information requirements, the reporting approach, and the communications campaign that supports the publication program. Expanded public awareness. The managed competition theorists expectation that disclosure would prompt direct shifts in provider or health plan market sharethe "vote with your feet" scenariowas simplistic. Implementation of a public reporting system on prescription drug quality may modestly affect the immediate business success of some drug plans or pharmacies, but it will have a second, more powerful result: raising the awareness of consumers that they are not always getting the quality of care they are paying for and may be assuming. Expanded public awareness may lead to greater demand for consistent, evidence-based, and safe health care.31 Independent trends, such as defined-contribution health plans, online provider directories, direct-to-consumer drug advertising, and disease management or pharmaceutical care programs may all lead to the same result: a growing number of informed health care consumers expecting and demanding high-quality care. A changed and less complacent public climate, in turn, might provide political cover for both public agencies and private purchasers to become more stringent in their regulatory or contractual handling of quality issues. Role of Congress. Legislators do not need to define, or even anticipate in any detail, the possible scope and form of reporting requirements associated with a Medicare prescription drug plan. Consumers, researchers, and industry representatives can do that. But only Congress can establish the importance of providing public information on the effectiveness of the program and hold new federal programs accountable for achieving national health objectives. Such an effort will certainly help Medicare beneficiaries to identify the best-performing drug plans available to them and help Congress to evaluate the impact of the program as a whole. But just as important, the statutory responsibility to report on the effectiveness of a federally funded benefit will set a precedent for other major public insurance expansion efforts. In the coming years Congress will consider other incremental health reforms, such as the creation of tax credits for individual insurance coverage and reauthorization of the TANF legislation. Any large expenditure of public funds should carry with it the responsibility to inform the public of that investments benefits. Americans health care quality problems show no signs of fixing themselves. At each opportunity, Congress must establish the expectation and the infrastructure that can lead to a more informed public and a more accountable health care system.
David Lansky is president of the Foundation for Accountability (FACCT), in Portland, Oregon. This paper was developed with the generous support of the Commonwealth Fund. My thanks to AnneMarie Audet and Ann Monroe for encouragement to develop these ideas; to Ken Barker, Lucinda Maine, Jerry Avorn, Stuart Lancer, Alan Wright, and Les Paul for their expert introduction to contemporary pharmacy management; to Jessica Lind, Dan Roble, Tom Susman, and Blaire Osgood for legal research and context; and to Lynn Etheredge for his helpful critique of an earlier draft.
This article has been cited by other articles:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||