|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Disparities And Quality Improvement: Federal Policy Levers
Using a quality improvement framework to address racial and ethnic disparities in health care highlights multiple opportunities for federal and state governments to exert policy leverage, particularly through their roles as purchasers and regulators. Under such a framework, federal and state governments can expand their roles in collecting race/ethnicity data; define universal and meaningful race/ethnicity categories; more broadly disseminate standards for cultural competence; and demand the reduction of disparities through leveraging their status as collectively the largest U.S. health care payer.
Several papers in this volume of Health Affairs examine approaches to reducing racial and ethnic disparities in health care, including expanding access to care and the use of legal and civil rights levers. One of the most promising strategies to emerge is the application of a quality improvement (QI) framework to promote measurable improvement on persistent patterns of unequal treatment.1 Such a framework involves better measurement of the quality of care, the widespread use of evidence-based guidelines for care, and the application and evaluation of intervention strategies. Given the level of government and private-sector interest in improving quality, incorporating disparities reduction into ongoing efforts may, in fact, be both more efficient and more sustainable than developing new initiatives targeted specifically at minority health issues. Viewing disparities in care through the lens of quality improvement also highlights the potential roles for federal health agencies in reducing them. In this paper we discuss the federal governments leadership role in the area of quality improvement and its implications for addressing disparities in care. By way of disclaimer, we do not address here the more upstream issues related to disparities in health status, including nonmedical determinants of health, because these are addressed elsewhere in the literature. Furthermore, we do not detail the array of federal initiatives aimed at reducing racial and ethnic disparities. Rather, we focus on how federal agencies can incorporate considerations for racial and ethnic disparities into their existing regulatory and purchasing functions. In doing so, we acknowledge that federal and state policies are often inextricably linked and that many of the federal policy levers we discuss are also relevant at the state level. Similarly, we recognize that nongovernmental actors in the health care system, such as large employer groups and insurers, can play important roles in addressing disparities through activation of their respective policy levers.2 We discuss some of the key components of the QI cycle: measurement, action, and evaluation.
The drive toward increased measurement of health care quality is premised on the idea that performance measures can help relevant stakeholderspurchasers, providers, consumers, patients, and policymakersunderstand what high-quality health care is and how to increase demand for it. The capacity to measure and monitor quality of care for various racial/ethnic populations rests on the ability both to measure quality of care in general and to conduct similar measurement across different population groups. One point at which the fields of quality improvement and disparities come together is with a push for better data with which to examine the quality of health care, including that received by members of different racial and ethnic groups. Quality data are widely available throughout the health care system. In contrast, while data on beneficiaries race/ethnicity are available to the Centers for Medicare and Medicaid Services (CMS) and the Department of Veterans Affairs (VA), the lack of available data on patients race/ethnicity in the private health care sector has been recognized only recently. Since most minorities receive care through private-sector insurance mechanisms, the absence of data poses a major challenge to the disparities-as-quality approach. Some are arguing persuasively that data reporting on quality of care should always be stratified by race, ethnicity, and other factors, to identify and address disparities in treatment. Two recent national reports on disparities in care explicitly call for the routine collection and availability of such data.3 Although no one believes that collecting data by race and ethnicity is a stand-alone solution to disparities, experts agree that such data are a critical first step in measurement, to identify disparities in treatment and develop targeted strategies to address them. There is legitimate debate about who is accountable for disparities in care. Regardless of the perspective, measurement is critical to accountability in any sector. The lack of data is not limited to data on race/ethnicity but may also extend to the types of quality measures that are routinely collected. Experience with standard quality measures used by health plans and the CMSsuch as the Health Plan Employer Data and Information Set (HEDIS) and the CMS Health Care Quality Improvement Program (HCQIP), stratified by race and ethnicitysuggests that these measures may be adequate as a first step for health systems to identify disparities in care.4 However, other experts call for quality measures that may be more specific to health problems that disproportionately affect various minority populations.5 The federal government, through the Department of Health and Human Services (HHS) and related agencies, such as the VA health system, oversees much of the countrys access to and quality of health care and has numerous initiatives focused on quality improvement. Yet despite the fact that half of the minority population receives care through a federally purchased program (Medicare, Medicaid VA, Defense Department, or federal employee health plans), most of these quality-related activities do not address the reduction of racial and ethnic disparities as an explicit goal, nor do they measure quality for specific racial and ethnic populations known to experience disproportionately worse care.6 The CMSs central role. We illustrate both the shortcomings and the opportunity for progress, using the CMS as an example. As the administrative agency for Medicare and the federal partner for all states with respect to Medicaid and the State Childrens Health Insurance Program (SCHIP), the CMS plays a central role in overseeing the quality of care through various measurement and certification activities. This oversight role could be greatly leveraged to address disparities. The CMSs major quality initiative centers on the Quality Improvement Organizations (QIOs), entities that are specifically charged with improving the quality of care for Medicare beneficiaries. Formerly known as Peer Review Organizations (PROs), QIOs work with providers and health care organizations to improve quality of care. A portion of their activities, including those that are done with Medicare+Choice plans (known as QAPI), are, by contract, focused on underserved populations, which may include racial/ethnic minorities. However, the CMS has been slow to collect and analyze data on whether quality indicators are being met for different racial/ethnic populations targeted by QIO programs. Because the CMS has not made such data available, it is difficult to determine whether the QIOs have been effective in reducing disparities. The CMS has also championed the Premier Hospital Quality Incentive Demonstration to test the effectiveness of financial incentives for hospitals to provide high-quality care in selected acute inpatient care areas. In addition, the CMS is exploring ways to increase transparency of hospital quality measures through the National Voluntary Hospital Reporting Initiative. In each of these endeavors, the CMS is working not only with other agencies, including the Agency for Healthcare Research and Quality (AHRQ), but with independent organizations such as the Joint Commission on Accreditation of Healthcare Organizations (JCAHO), the National Quality Forum (NQF), and the Leapfrog Group. The CMS could stimulate potentially large improvements in quality for minority populations by encouraging each of these partners to monitor quality of care for different racial/ethnic groups and to promote action on those findings, when appropriate. Moving the health system toward more regular and standardized collection of data on quality by race and ethnicity is not a simple step. Although the availability and quality of racial and ethnic data in the private sector is, in general, severely limited, data availability in the CMS is greater than in other health care systems. The CMS already has such data, obtained through the Social Security Administration and subsequently updated, for most beneficiaries, including enrollees in Medicare managed care plans. In the past few years, the CMS has made these data available to researchers, some of whom have reported on variations in the quality of care by race and ethnicity.7 More recently, the Medicare Prescription Drug, Improvement, and Modernization Act (MMA) of 2003 gave the CMS the authority to make the data available on a real-time basis to managed care organizations, a step that could accelerate the timeline for developing strategies to address identified racial and ethnic disparities for specific population groups if these organizations choose to use it. However, given the scope of MMA, implementing this provision is not a top priority for the CMS.8 Although the demonstration and managed care contexts are good beginnings, the CMS could be much more proactive in promoting increased regular reporting on quality-of-care measures for different racial/ethnic groups, in conducting analyses within Medicare (including fee-for-service Medicare), and in disseminating the findings to relevant stakeholders. A starting point might be to expand the regular state-by-state reporting of quality in general to include reporting of quality for racial and ethnic groups, by state or by region.9 Moving reporting from a broad federal level to states and regions will likely have the added benefit of increasing the salience of such information for providers and communities, because it provides data that are seemingly more directly applicable to local environments. Support for this position comes from a recent National Research Council report on HHS data collection and measurement practices, which recommended that the CMS "develop a program to collect racial, ethnic, and socioeconomic position data at the time of enrollment and for current [Medicare] enrollees." Other national organizations have made similar recommendations about the need for data.10 Private-sector efforts. It is not clear what it would take for the CMS to take these more active steps, although given the emphasis on disparities over the past several years, this would be an appropriate time to do so. We believe that several factors might inspire the CMS to exert a leadership role. First, anticipated demographic shifts and growth of nonwhite racial/ethnic populations during the coming decades create a new imperative for the CMS to be certain it can assure high-quality care to all beneficiaries. Second, there is now evidence that the private health care sector is recognizing disparities reduction as an important priority. Health plans, which are obviously positioned to monitor and influence the quality of care, are beginning to recognize that they have an important role in addressing disparities. A recent survey conducted by Americas Health Insurance Plans (AHIP) indicated that a surprisingly large portion of plans surveyed were already collecting data on membership by race and ethnicity and using this information for targeted QI initiatives.11 The Health Research and Educational Trust (HRET) is also involved in this issue from the hospital perspective and has issued a report on data collection.12 Two large collaboratives of insurers, one representing mostly commercial plans and one consisting of Medicaid plans, are now working toward the goal of reducing disparities, by obtaining data on enrollees race/ethnicity, stratifying HEDIS quality measures, and using the results to develop and test QI interventions.13 Several participating plans report that large national employers have strongly encouraged their disparities activities and in some cases are working in active partnership with them.14 Elsewhere in this volume, Audrietta Izlar describes the experiences of Verizon from the employer perspective.15 Although it is too early to know whether these efforts will be successful, it is clear that the CMS will not be alone if it moves in this direction. Furthermore, its entry into this arena may stimulate others that are not part of current initiatives to take similar action and may spark additional consumer interest in information on quality of care. Public and private funders can complement federal efforts by supporting research to develop tools that help promote a greater drive toward disparity-focused quality measurement and reporting. AHRQ has also done a great deal to advance the quality measurement and reporting field, and expansion of its research agenda with respect to disparities has been proposed in pending federal legislation.16 Other federal activity. The CMS is not the only federal agency with an important role to play in policy surrounding data collection and its connection to quality improvement. The Health Resources and Services Administration (HRSA) has been actively involved in this area, focusing on disparities for a set of clinical conditions, such as asthma, HIV, and diabetes, through its community clinic disparities collaboratives. Further, the National Academies report highlighted the fact that HHS already has a set of policy recommendations (issued in 1999) with respect to data collection on race and ethnicity.17 These policies pertain to "all data collection or reporting systems required under [HHSs] programs or activities relating to the collection of race, ethnicity, or socioeconomic position."18 The committee suggested ways that these recommendations could be implemented and highlighted the role for other federal agencies that interact with HHS around this issue, such as Social Security.
Defining workable categories. Not all of the challenges related to better data collection are associated with the absence of a coordinated federal approach to obtaining these data. One area of confusion, especially in the private sector, that would benefit from more federal action is defining workable categories for race/ethnicity that allow stakeholders to capture meaningful racial or ethnic variations in the quality of care. The federal Office of Management and Budget (OMB) already has standards in place for this, but they are not widely known or understood, and several federal agencies are out of compliance with them. Many are concerned that these standards are insufficient in view of the substantial variation in ethnicity within groups, particularly for Asian and Hispanic subgroups and non-U.S.-born African Americansas well as the increasing number of people who identify with two or more race/ethnicity categories. Although the OMB standards do not preclude the identification of subgroups (provided they can be aggregated back to the "major" race/ethnicity categories), it is increasingly clear that the field would benefit from clearer guidance for how ethnic subgroup data should be collected by various entities. This could be provided either through the OMB or through a guidance issued by HHS. Such guidance should address when categories beyond the standard "African American," "Asian American," "Caucasian," and "Hispanic" should be used, an issue of particular importance in demographically diverse states and regions. A logical approach might be to continue to require the current data collection categories and to tailor data collection about subgroups to individual communities. For example, guidance could suggest that whenever one ethnic group (such as Vietnamese or Cuban American) exceeds a certain size threshold (for example, 5 percent of the population in a community or health plan), data to identify members in that subgroup be collected, recognizing that the data can always be aggregated back into the broader categories. Such data are available from the U.S. Census and intercensal surveys. In some states, Medicaid collects data on beneficiaries primary language and provides these data to plans. Standards for data collection and use. One of the most formidable challenges to making data collection by race and ethnicity routine in the private sector is the prevailing confusion among health care providers and insurers about whether such data collection would expose them to greater risks in terms of potential liability or consumer backlash. What is lacking is not the activity of data collection by race and ethnicity itselffederal agencies collect these data on a regular basisbut information about and uniformity in standards for collecting such data and how to use the data to improve quality for specific racial and ethnic population groups. Although recent analyses have determined that collecting such data is legal under Title VI, program-specific statutes, and regulations for data collection and reporting, this determination is in some ways beside the point.19 If they are required to do so, the discussion can shift to how such data can be used to supplement the existing federal initiatives to improve the quality of care.
The HHS Office of Minority Health (OMH) has issued national standards for Culturally and Linguistically Appropriate Services (CLAS) in health care. The CLAS standards, which are primarily directed at health care organizations, contain recommendations at different levels. Some CLAS standards are mandates for all recipients of federal funds; some are activities recommended for adoption by accreditation agencies for high-quality health care; and others are guidelines for voluntary adoption by health care organizations.20 Although enforcement of any of these standards, mandatory or otherwise, has been problematic for the OMH and for the HHS Office for Civil Rights (OCR), the standards at least set expectations for what this dimension of quality for diverse patient populations ought to be, and they should be more widely disseminated by the OMH and other HHS agencies, such as the CMS and HRSA, that have a stake in improving care for minority populations. Some federal agencies are trying to incorporate aspects of CLAS standards into ongoing quality oversight activities. For example, the CMS, through its QIO activities and the national quality assessment and performance improvement (QAPI) initiative, required that Medicare+Choice plans shift their underserved population emphasis slightly in 2003 to explicitly include a project pertaining either to clinical health care disparities or the provision of services that complied with CLAS standards. The CMS worked with AHRQ to develop guides for managed care plans that would help them meet CLAS standards. In addition, national quality organizations such as JCAHO and the National Committee for Quality Assurance (NCQA) are working to align CLAS principles with systems of measurement that are already in use, such as HEDIS and HCQIP.21 The CLAS standards and what they call for could lead the way for the private-plan sector to take up the challenge of clarifying what "high-quality health care" for diverse patients really means. Many health plans and other organizations have proposed cultural competence training for providers as a way to address disparities. However, few understand what it encompasses; more importantly, it is not clear whether providers who receive this training begin delivering care differently, in a way that improves the quality of care for minorities. Some organizations, such as Kaiser and HealthPartners in Minnesota, are experimenting with larger-scale organizational change efforts in cultural competence. Research is under way to determine whether and how cultural competence training affects the quality of health care. One recent study among Medicaid managed care plans suggests that plans language and cultural competence practices are positively related to childhood asthma outcomes.22 If further research confirms the value of such practices, AHRQ should work much more closely with the OMH to assure that only evidence-based practices become incorporated into the CLAS standards and that related agencies within HHSin particular HRSA and the CMSare responsible for disseminating these standards and providing technical assistance to health plans interested in meeting them. Meanwhile, the BBA requires states to "establish a methodology for identifying the prevalent non-English languages spoken by enrollees and potential enrollees." However, the definition of prevalent differs from state to state. For example, Connecticut requires plans to provide translation services if the language group exceeds 5 percent of the overall population, while California sets the threshold at 3,000 eligible beneficiaries residing in a county, 1,000 in a single ZIP code, or 1,500 in two contiguous ZIP codes. New Jersey has a general threshold of 5 percent of the Medicaid line of business or 200 enrollees per plan.23
In the QI field, organizations such as Leapfrog and the National Business Group on Health (NBGH) have made strides in leveraging purchasers to demand higher-quality health care for employees. A recent paper from the NBGH outlined the business case for purchasers to engage in disparities reduction and provided guidance for purchasers on how to begin.24 Several large employers, such as Verizon and Marriot, are working with health plans in activities to reduce disparities.25 However, most believe that purchaser leverage exerted for better-quality care could start with the largest payer of health care services, the federal government. The Institute of Medicine report on this topic, part of its Quality Chasm series, recommended that the government take up the quality measurement issue to a much greater degree than it has thus far.26 Although the CMS has made progress in this area by releasing performance measures on nursing homes and dialysis units, it has yet to extend this work to common chronic conditions or to incorporate measures of race/ethnicity to the existing reporting areas. The CMS is not the only federal purchaser that has underused its leverage with respect to quality improvement and disparities reduction. For example, the Federal Employees Health Benefits Program (FEHBP) covers more than eight million federal employees. In this purchaser role, the Office of Personnel Management (OPM) already monitors quality-of-care indicators for enrollees, and it has the authority to ask plans to report on quality indicators by race and ethnicity. Similarly, the Department of Defense could require plans that provide care to military and dependents to do the same. Certainly, the federal government as purchaser could provide leadership for both states and the private sector, by adding requirements for data collection and promoting the reporting of quality indicators by race and ethnicity.
Although this paper has mapped out some clear directions with respect to quality improvement, reporting, and data collection that the federal government can take to help reduce racial and ethnic disparities in care, the one question that is often left unresolved is the potential role of the consumer. Are consumers themselves levers for change that federal and state policies can help activate? Should the drive to collect data by race and ethnicity for the purposes of reducing disparities also include a drive to report on variations in quality that are found, and, if so, at what level of aggregation? Reporting data on quality of care for various racial or ethnic groups for the CMS overall, or at a state and regional level, has different implications than public release of such data at lower levels of aggregation, such as a hospital or health plan. Data describing quality at a site of care or a geographic region may engage consumers differently than nationally aggregated data. However, evidence to date has shown little use of measures by consumers regarding quality in general, so it will be important to determine what it will take to accelerate use. Data from Commonwealth Fund surveys indicate that minority populations in general are less likely to ask questions of their providers or to understand information received from them.27 This suggests that promoting the use of quality information by some minority populations may require special attention. The CMS and Leapfrog, as well as other organizations such as the NQF, have begun thinking about what consumer education around measures and quality improvement might look like.28 The NQF has suggested a set of strategies to engage minority patients in using health quality information, such as involving community-based intermediaries.29 In the meantime, it seems that some large employers, particularly those that employ sizable minority populations, are poised to use race- and ethnicity-specific information, combined with their purchasing power, to press for improvements in the quality of their employees health care. The extent to which they will work to engage their employees in this issue is not known.
In addition to opportunities provided through its regulatory roles, the federal government has numerous other opportunities to address disparities. For example, developing and diversifying the health care workforce at all levels is key to realizing the goals of culturally competent health care organizations. Strengthening programs involving the HRSA Bureau of Health Professions, as well as collaboration with the Departments of Labor and Education, will be crucial. The use of civil rights enforcement tools is another potential area for policy exploration; it is discussed elsewhere in this volume.30 The Office for Civil Rights can also step up its efforts to educate health care organizations that the regular, voluntary collection of race/ethnicity data is not illegal. Just as guidance is needed regarding race/ethnicity categories and how to collect data, guidance that clarifies when disparate patterns of care represent serious quality problems that need to be addressed and when these patterns represent clear discriminatory practices that are in violation of Title VI could go a long way toward encouraging others to work on measuring quality and reducing disparities without fear of litigation. Finally, the development of new knowledgewhether related to measurement, reporting, interpretation and use of information, and other aspects of QI or increasing workforce diversityis always an important facet of quality improvement and should receive increased support from federal agencies. Federal and state governments already play important roles in overseeing the quality of health care, especially in their dual roles as purchasers and regulators of health care. Any efforts to support these QI initiatives should automatically incorporate the reduction of racial and ethnic disparities as part of the QI agenda. Straightforward measures such as requiring the collection of racial and ethnic identifiers for patient populations and setting standards for the usage and collection of such data are actions that the federal government can take with the cooperation of several agencies and with the help of the private sector, the research community, and philanthropy. In addition, the federal government already has initiatives such as the CLAS standards to ensure that culturally diverse populations receive appropriate health care; ensuring that these initiatives are enforced and best practices are disseminated is one area where the federal government is well positioned to act.
Nicole Lurie (lurie{at}rand.org) is senior natural scientist and the Paul ONeill Alcoa Professor, at RAND in Arlington, Virginia. Minna Jung is a communications officer at the Robert Wood Johnson Foundation (RWJF) in Princeton, New Jersey. Risa Lavizzo-Mourey is the RWJFs president and chief executive officer. The authors appreciate the thoughtful contributions of Anne Weiss and the help of Rebecca McLaughlin in preparing the manuscript.
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||