Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Racial and Ethnic Disparities in Healthcare: Issues in the Design, Structure, and Administration of Federal Healthcare Financing Programs Supported Through Direct Public Funding Sara Rosenbaum J.D. Harold and Jane Hirsh Professor, Health Law and Policy INTRODUCTION An examination of the relationship between racial disparities in health- care and public healthcare financing may strike some as ironic, given the well-documented role that programs such as Medicare and Medicaid have played in reducing racial and ethnic disparities in healthcare access and health outcomes (Congressional Research Service, 1993; Committee on Ways and Means, 1996; Davis and Schoen, 1978; Moon, 1993; Starr, 1982; Smith, 1999; The Henry J. Kaiser Family Foundation, 1999). But in the face of a significant and ongoing health gap between minority and non-minority individuals, it is worth considering whether the manner in which public financing programs are administered has the potential to contribute to one of the nationâs most sobering and enduring public health problems. This paper begins with a background and overview that briefly de- scribe Medicare and Medicaid (and its companion SCHIP program) and roles in financing healthcare for minority persons. It then turns to a more extended analysis of the kinds of administrative choices made under these programs that have the potential to contribute to the problem of health disparities, either by tolerating or tacitly countenancing access, treatment and quality differentials or by failing to act affirmatively to minimize the possibility of differentials. This paper examines issues in federal and state administration of health programs rather than the legislative design of the programs them- selves. Several limitations and caveats should be noted at the outset. First, the association between race and poverty makes it difficult to disentangle 664
665 FEDERAL HEALTHCARE FINANCING PROGRAMS the two factors. Administrative choices that adversely affect poor pa- tients are also more likely to create problems for minority patients. The literature is replete with studies of race-associated healthcare disparities; regardless of payer source, income, or other characteristics unrelated to healthcare need, racially linked health disparities appear to be one of the most unfortunate constants of the American health system (U.S. Civil Rights Commission, 1999; Mayberry et al., 2000; Gaskin and Hoffman, 2000; Weinick et al., 2000). Minority patients perceive barriers and racism within the health system, and at least some research suggests that these perceptions appear to be borne out by discernible differences in how health professionals interact with minority patients (The Henry J. Kaiser Family Foundation, 1999; La Viest et al., 2000; Lillie-Blanton et al., 2000; Einbinder and Schulman, 2000; Schulman et al., 1999). Furthermore, stud- ies confirm the independent role of race in healthcare (Mayberry et al., 2000). At the same time, it is difficult to separate healthcare administra- tion choices that harm poor people from those that harm members of ra- cial and ethnic minority groups. This is particularly true in the case of programs such as Medicaid, where coverage is specifically aimed at the poor and medically indigent. Second, even if U.S. lawmakers were to enact a totally reformed sys- tem that utilizes a single payer with common coverage and payment rules, research from other nations suggests that minority and poor residents nonetheless would continue to experience reduced healthcare access and poorer health outcomes. The literature on disparities in healthcare access is replete with examples of disparate access to care and disparate utiliza- tion of health services (Mayberry et al., 2000; Lillie-Blanton et al., 2000). The most recent example can be found in a 2001 Surgeon Generalâs report examining racial and ethnic disparities in the use of mental health ser- vices by members of racial and ethnic minority groups, which speculates on the underlying causes of disparity in access, utilization and quality and concludes that the principal factors are cost, poor services in poor communities, cultural and communications barriers, fear of the health system, and general overall problems in the relationship between patients and providers (U.S. Department of Health and Human Services, 2001). This focus on provider/patient communication difficulties, fear of the sys- tem, and cultural isolation in healthcare appear to be recurrent themes throughout the literature on healthcare disparities. Furthermore, the evidence on disparate access to care even where in- surance is technically not a barrier is hardly unique to the United States. For example, studies of the apparently common practice of using govern- ment-sponsored community health clinics in nations with national health systems typically point to the need for such service delivery interventions because of access barriers related to race, ethnicity, culture, and poverty
666 UNEQUAL TREATMENT (Hawkins et al., 2000). Consequently, even aggressive efforts to reduce disparity-causing choices in the administration of healthcare financing programs could nonetheless continue to leave minority and low-income patients with unequal access. Third, even substantial incremental reforms that improve existing programs but leave them intact inevitably would leave millions of Ameri- cans dependent on sources of healthcare financing that are perceived as other than âmainstream,â and thus subject to the misapprehensions that enrollment in âlesserâ programs brings with it. The data reviewed for this study suggest that incremental reforms would leave a racially and income-identifiable group of Americans disproportionately enrolled in forms of healthcare financing that are less attractive to healthcare provid- ers because they cover less, pay less, and carry unpleasant connotations and associations, such as bureaucratic hassles and the status of a âpoor peopleâs program.â1 This lesser form of coverage in effect validates on business grounds what may be underlying prejudicial leanings on the part of members of the medical care industry. Finally, and as noted at the outset, any assessment of the limitations of directly financed public insurance programs such as Medicare and Medicaid must be read against a backdrop of their extraordinary accom- plishments over the past three and a half decades. Since their inception, Medicare and Medicaid have literally remade the American healthcare system for minority Americans, opening access that previously had been denied. At the time of their enactment, white Americans were hospital- ized 27% more frequently than African Americans and members of other minority groups, and in the case of elderly persons, the racial gap stood at 70%. By 1975, the gap had narrowed to 4% overall and 14% among the elderly (Davis and Schoen, 1978). Research also has pointed to the con- nection between the decline in U.S. infant mortality rates and the advent of Medicaid, which made pregnancy related care available and accessible to the poorest women (Davis and Schoen, 1978; Congressional Research Service, 1993). Data on access to and use of healthcare by income and insurance sta- tus suggest that Medicaid has eliminated healthcare access and utilization disparities among children and non-elderly adults, particularly when uti- lization data are adjusted for reduced health status; indeed, poor Medic- aid beneficiaries appear to use care at rates greater than the poor with 1 At one time, dependence on Medicaid was viewed by health providers and policymakers as stigmatizing. As Medicare has become increasingly complicated and overall payment has declined in relation to the overall cost of care, anecdotal evidence suggests that reliance on Medicare (either alone or in combination with Medicaid) may be producing similar reactions within the health system.
667 FEDERAL HEALTHCARE FINANCING PROGRAMS private insurance, presumably because of the programâs broad coverage rules and strict limits on cost sharing, discussed below (Congressional Research Service, 1993). Although the source of care differs, Medicaid beneficiaries have been shown to be significantly more likely to have a usual source of care (Congressional Research Service, 1993). Medicare and Medicaid have been directly credited with the desegregation of U.S. hos- pitals, nursing homes, and other healthcare institutions, as providers rap- idly moved to eliminate the techniques of discrimination in order to be able to participate in government health programs (Smith, 1999).2 Despite these limitations, it is important to understand the extent to which the administration of publicly funded health coverage has the po- tential to perpetuate, intensify, or implicitly validate differential treatment of minority Americans. Background and Overview Medicare and Medicaid represent enormous advances in American social welfare policy. The joint product of an extraordinary convergence of social, policy, and political circumstances that have been chronicled at length and in multiple dimensions by numerous experts (Fein, 1986; Marmor, 1970; Moon, 1993). Medicare and Medicaid not only opened the health system to previously uninsured persons but changed American healthcare itself by supplying the financing needed to achieve the enor- mous leaps in medicine and technology that the nation has witnessed over the past 40 years. For purposes of issues that are dealt with in this paper, it is important to remember the context in which Medicare and Medicaid were enacted. As Marilyn Moon has observed, â[t]he rules that were established to gov- ern Medicare did little to disrupt or change the way healthcare was prac- ticed or financed in the United States.â In his seminal history of the fed- eral governmentâs efforts to address race discrimination in American healthcare, David Barton Smith describes the civil rights environment in which Medicare and Medicaid were enacted, with de jure race discrimina- tion in healthcare having only recently ended, and with Southern Mem- bers of Congress threatening to derail passage of Medicare if its funding were used under Title VI as a lever to force healthcare integration (Smith, 2 These techniques included such devices as denying admission to patients without staff physicians while simultaneously denying admitting privileges to minority physicians or physicians working at publicly funded clinics located in medically underserved communi- ties, segregating the wards and wings of hospitals and nursing homes, placing strict numeri- cal limits on minority patients admitted or served, demanding insurmountable pre-admis- sion deposits (akin to a poll-tax), and refusing to participate in certain government insurance programs, particularly Medicaid (Rosenbaum et al., 2000).
668 UNEQUAL TREATMENT 1999). In its efforts to secure Medicareâs enactment, the Johnson Adminis- tration effectively promised that Medicare funding would not be used as a basis for Title VI enforcement against physicians in individual practice and thus as a means of achieving changes in the behavior among medical professionals.3 This promise had major implications for the autonomy of American medicine and was consistent with the overall hands-off ap- proach of the original Medicare legislation when it came to altering the behavior of physicians. The Administration made good on its promise by interpreting Medicare physician payments as a form of indemnity cover- age, which lacked the requisite nexus to federal funding to produce Title VI enforcement jurisdictionâwhich hinges on the receipt of federal finan- cial assistance. Despite the virtual end of Medicare as an indemnity-style program (physicians now are effectively required to accept direct, as- signed Medicare payments as a condition of participation through the use of payment penalties for those physicians who refuse assignment), suc- ceeding administrations never have issued an outright reversal of this original interpretation of the program in a Civil Rights Act context. 4 While both Medicare and Medicaid are quite complex, Medicaid is especially so because of its cash welfare assistance roots, as well as its uneasy perch atop a honeycombed federal-state system of program au- thority. It is not possible to understand how the design and administra- tion of the programs (as well as the more recent SCHIP statute) could potentially create disparities without a basic familiarity with the programs and their relationship both to other payers as well as to each other. Medicare Medicare is a federally administered social insurance program that finances a defined set of health benefits for individuals who qualify for coverage (i.e., individuals who are entitled to Social Security Old Age or Disability Insurance benefits, children and adults with end-stage renal 3 Of course no similar promise was made with respect to Medicare payments to hospitals, whose desegregation already had been forced by the courts. Smith presents a marvelous overview of the Johnson Administrationâs near-superhuman effort to achieve compliance agreements with all Medicare-participating hospitals by the date of Medicare implementa- tion, only six months following enactment. 4 The hesitation to extend civil rights statutes to office-based health professionals on the basis of their participation in federal healthcare financing programs continues today. In 2000, the federal government released guidelines clarifying existing civil rights act stan- dards as they apply to healthcare access among persons with limited English proficiency. Included in the regulations was clarification that physicians would be considered subject to these standards if they participated in Medicare and Medicaid. By the spring of 2001, nearly all of the major private medical groups were meeting with White House officials demanding the repeal of the guidance.
669 FEDERAL HEALTHCARE FINANCING PROGRAMS disease, and certain individuals who are permitted to purchase coverage). In 2000, Medicare outlays reached $216 billion (Congressional Budget Of- fice, 2001) and total program enrollment stood at over 39 million (HCFA, 2001a). Medicare is federally administered according to uniform standards, most of which are found in extensive regulations governing the program. The program consists of three parts (A, B, and C). Part A (Hospital Insur- ance) is paid for through a payroll tax-based trust fund; it covers inpatient and outpatient hospital services, post-hospital extended care services, home healthcare, and hospice benefits. Part B (Supplementary Medical Insurance) is financed through a combination of premium payments and general revenues. Part B covers physician and other medical services, outpatient hospital care, home health services, certain preventive services, clinical diagnostic laboratory services, ambulatory surgical services, and outpatient mental health services. Part C of Medicare, enacted in 1997, established the Medicare+Choice program as a means of encouraging enrollment in managed care arrange- ments and to modernize and strengthen the regulatory framework for Medicare managed care (Rosenblatt et al., 1997; 2001). Medicare contains significant cost-sharing requirements, including both deductibles and coinsurance. While Medicare is federally administered, private insurers (known as carriers and intermediaries) conduct the day-to-day business of provider enrollment, claims payment, and coverage decision-making. Qualified managed care organizations, known as Medicare+Choice providers, carry out broad contractual responsibilities for the federal government. Both the Medicare statute and implementing regulations establish conditions of participation for medical care institutions and professionals; indeed, much of the Medicare legislation is devoted to the establishment of standards of participation for health professionals, hospitals, other in- stitutions and suppliers, and the managed care industry. Medicare also specifies a range of formulas for provider compensation in the case of hospital care, physician and medical care, and payments to managed care entities. State health agencies, accreditation bodies, and peer review orga- nizations conduct provider certification and oversight activities. Medicare was founded on the notion of health system freedom for both providers and patients alike. Physicians, hospitals, health profes- sionals and suppliers have discretion over whether to participate in the program at all (although most U.S. physicians and virtually all hospitals and qualified nursing homes and home health agencies do so). Further- more, providers can decide the extent of their participation, limiting their involvement for example to patients who once were privately insured and now depend on Medicare exclusively or primarily.
670 UNEQUAL TREATMENT In a similar vein, a hallmark of Medicare from a beneficiary perspec- tive is its free-choice-of-provider guarantee. It is probably safe to say that Medicare beneficiaries are the last remaining group of insured Americans who are given a choice with respect to how they use healthcare. Benefi- ciaries can elect to enroll in a Medicare+Choice plan if one is available;5 alternatively, they can elect to remain in the âfee-for-serviceâ system, ob- taining medical and healthcare from the participating physician, health- care institution, or other health professional of their choice. While the concept of free choice among participating health providers exists in theory, in practice there are problems, although none so severe as those faced by Medicaid beneficiaries. When these problems are com- bined with various structural shortcomings in the Medicare program, they create a potential for barriers, particularly in the case of lower income beneficiaries, who are significantly more likely to be members of a racial or ethnic minority group. Medicare coverage is limited, omitting crucial services such as prescribed drugs and cost sharing is high, with monthly premiums in the case of Part B coverage and significant deductibles and coinsurance. Furthermore, because Medicare was modeled on the âmajor medicalâ health insurance plans that existed at the time of enactment (and that still dominate the insurance market), it fails to cover long-term ser- vices necessary to the management of chronic and serious physical and mental health conditions that extend beyond an initial acute phase of illness. Figures 1 through 4 illustrate the nature and extent of the dilemma facing minority Medicare beneficiaries. Figure 1 shows that members of racial and ethnic minority groups, who are at significantly greater risk of poverty, represent a sizable and growing part of the Medicare popula- tion. By 2025, minority persons will constitute 33% of the Medicare popu- lation, up from 15% in 1995. Figure 2 provides an overview of the health status of Medicare benefi- ciaries by race and ethnicity and shows that regardless of condition, mi- nority beneficiaries are more likely to experience significant limitations in health status. Latino and African-American beneficiaries are more than one-and-a-half times more likely to be in fair to poor health. They also are at significantly greater risk for one or more limitations in activities of daily living (ADLs) and cognitive impairments. This health risk profile sug- gests a higher need for services. Figure 3 shows the enormity of the poverty gap between minority and non-minority beneficiaries. In 1997, African-American and Latino 5 In recent years participation in Medicare+Choice has eroded significantly, chiefly as a result of limitations on the payment formula enacted in 1997, as well as increased regulatory participation standards (Rosenblatt et al., 2001).
671 FEDERAL HEALTHCARE FINANCING PROGRAMS FIGURE 1 Racial and ethnic minority Americans as a share of the elderly popu- lation: 1995â2025. SOURCE: Urban Institute analysis of the March 1998 Current Population Survey, prepared for The Henry J. Kaiser Family Foundation. FIGURE 2 Health problems and long-term care needs: minority and non-minor- ity individuals. SOURCE: Urban Institute analysis of the March 1998 Current Population Survey, prepared for the The Henry J. Kaiser Family Foundation.
672 UNEQUAL TREATMENT Minority and non-minority individuals FIGURE 3 Poverty rates: Minority and non-minority individuals. SOURCE: Urban Institute analysis of the March 1998 Current Population Survey, prepared for The Henry J. Kaiser Family Foundation. FIGURE 4 Supplemental insurance coverage: Minority and non-minority Medi- care beneficiaries. SOURCE: Urban Institute analysis of the March 1998 Current Population Survey, prepared for the The Henry J. Kaiser Family Foundation.
673 FEDERAL HEALTHCARE FINANCING PROGRAMS beneficiaries were over three times more likely to be poor and more than one-and-a-half times as likely to have low incomes overall. That year, two- thirds of African-American and Latino beneficiaries were poor or near poor. Figure 4 shows the healthcare financing implications of the deep pov- erty among minority beneficiaries. In 1995, white Medicare beneficiaries were one-and-a-half times more likely than African-American beneficia- ries and twice as likely as Latino beneficiaries to have additional coverage through an employer-sponsored health plan and three times as likely to have private Medigap coverage. Conversely, minority beneficiaries were twice as likely to be exclusively dependent on Medicare and about three times as likely to depend on Medicaid as a supplemental source of health coverage. These statistics suggest that minority Medicare beneficiaries are poorer and sicker, are at significantly greater risk for serious healthcare under-financing, and are far more likely to depend on sources of supple- mental financing less acceptable to providers. Minority beneficiaries are more likely to lack employer-sponsored or other private coverage, signifi- cantly more likely to depend on Medicare alone, and far more likely to depend on Medicaid as a source of supplemental coverage. Because mi- nority beneficiaries represent a rapidly growing proportion of the overall Medicare population, the consequences of these problems are more likely to become clear and pronounced. Medicaid and the State Childrenâs Health Insurance Program (SCHIP) The largest and most complex of all federal grant-in-aid programs, Medicaid is a means-tested entitlement that creates three interlocking sets of enforceable legal rights. The first is statesâ right to open-ended federal financing for their medical assistance and program administration costs. The second is an eligible individualâs right to coverage for a defined set of benefits. The third is a legally enforceable providersâ right to participate in Medicaid if qualified and to be paid for the care they furnish (Rosen- baum and Rousseau, 2001). In fundamental respects, Medicaid is the mirror image of Medicare. Medicaid is rooted in welfare principles in that its origins were as a com- panion to federal cash welfare assistance programs for certain poor families with âdependentâ children, indigent elderly and disabled persons, and cer- tain âmedically needyâ persons whose characteristics connect them to a federal welfare category (e.g., age, disability, dependent children) (Congres- sional Research Service, 1993; Schneider et al., 1998; Rosenbaum and Rousseau, 2001). These mandatory coverage categories have been ex- panded over the past 35 years to include âpoverty levelâ (i.e., low income)
674 UNEQUAL TREATMENT children and pregnant women regardless of family composition or disabil- ity status, and low-income Medicare beneficiaries whose poverty level in- comes prevent them from either purchasing supplemental Medigap cover- age or paying Medicareâs premiums, deductibles and coinsurance out of pocket. Beyond these minimum coverage groups, the law gives states the option of covering literally dozens of additional eligibility groups consist- ing of persons who bear some relationship to the mandatory groups but who are not poor enough to qualify for coverage outright. Despite the many eligibility expansions that have occurred over the past two decades, Medic- aid is still associated with coverage of the poor. After 35 years, the program remains a selective and restrictive source of coverage, reaching only ap- proximately half of all poor individuals. Despite the fact that Medicaidâs roots are in welfare, its importance as a health payer can hardly be overstated. In 1998, Medicaid was a source of health insurance for 40 million persons (Congressional Budget Office, 2001). The vast majority of individuals insured through Medicaid are persons without access to employer or other private health insurance ben- efits; they are individuals who because of age, disability, or dependency lie outside the furthest limits of the private health insurance market (Rosenbaum and Rousseau, 2001). Medicaid is an integral part of the American insurance system not only because of whom it covers, but also because of what it finances. Unencumbered by the conventions of private insurance, Medicaid is ca- pable of covering populations and services that lie outside essential struc- tural insurance limitations that flow from the problems of âavoidable riskâ and âmoral hazard,â and that are embedded in the notion of âfair dis- criminationâ (Rosenblatt et al., 1997; Rosenbaum and Rousseau, 2001). In its role as an insurer of both uninsured and uninsurable populations and services, Medicaid effectively attempts to compensate for the structural and financial limitations of the worldâs largest voluntary healthcare mar- ket. Medicaid melds state design and administration choices within a broad federal framework that contains many options and a few absolutes. The program is designed and administered by participating states in ac- cordance with broad federal standards. As a matter of federal law, state welfare agencies bear final legal responsibility for Medicaid eligibility determinations, but the law permits any state agency to act as the respon- sible âsingle state agencyâ6 for overall program accountability purposes.7 6 42 U.S.C. Â§1395a(a)(4) 7 Approximately half of all states share ongoing administrative responsibilities with county governments; even in these states however, the single state agency has a non-delegable ob- ligation to administer the program within federal requirements.
675 FEDERAL HEALTHCARE FINANCING PROGRAMS In many states, agencies other than welfare agencies, such as public health agencies, public healthcare financing agencies, and social welfare agen- cies, are the named single state agency (Commerce Clearinghouse, 2001). The federal Medicaid statute, which has been described by courts as one of the most complex social welfare statutes ever conceived,8 contains minimum standards regarding eligibility, benefits and coverage, patient cost sharing, and program administration (Congressional Research Ser- vice, 2001). Minimum provider participation and payment rules also are specified; since the repeal of the so-called âBoren Amendmentâ in 1997,9 federal Medicaid provider payment formula standards have been dimin- ished, but certain standards remain in place. 10 As noted, states that elect to participate in Medicaid must extend cov- erage to certain categories of individuals, known as the mandatory cat- egorically needy. These groups consist of âAFDC-relatedâ families with children,11 certain former welfare recipients, elderly and disabled recipi- ents of Supplemental Security Income, certain low-income Medicare ben- eficiaries (for Medicare cost sharing only), and âpoverty levelâ pregnant women and children. States have numerous expansion options, of which the most important for purposes of this paper is the option to extend cov- erage to any adult with a Medicaid-eligible child (Rosenbaum and Maloy, 1999). In the absence of special federal demonstration authority under Â§1115 of the Social Security Act, states do not have the option to extend coverage to non-elderly, non-disabled adults without Medicaid-eligible children (Rosenbaum et al., 1999b). Persons who are eligible for Medicaid are legally entitled to apply for assistance and have the right to a prompt determination of eligibility.12 8 See, e.g., Lewis v City of New York, 2001 WL 540657 (2d Cir., May 22), in which the court began its decision regarding Medicaid eligibility among undocumented pregnant women and by noting that Medicaidâs complexity required that it offer a âregrettably detailed re- view of the lengthy history of this action as well as the evolution of Medicaid law as it pertains to prenatal care and aliens.â 9 The Boren Amendment guaranteed hospitals and nursing facilities payment in accor- dance with a specific cost-related payment methodology. This methodology created an en- forceable legal right and was thus highly unpopular among states. Wilder v Virginia Hospital Association, 496 U.S. 498 (1990). In 1997 Congress repealed the Boren Amendment as part of the Balanced Budget Act, Pub. L. 105-33. 10 For example, federal law continues to set minimum, cost-based payment rules for feder- ally qualified health centers, rural health clinics, and hospices. 11 In 1996 when the Aid to Families with Dependent Children program was repealed, Medicaid was amended to preserve this âAFDC-relatedâ category in order to avoid the loss of Medicaid coverage among individuals no longer entitled to cash welfare. Despite this effort, Medicaid enrollment declined steeply, as welfare recipients were pushed off the pro- gram and eligible persons, caught in welfare diversion efforts, were denied the opportunity to enroll (Ellwood and Ku, 1998; Rosenbaum and Maloy, 1999). 12 42 U.S.C. Â§1396a(a)(8).
676 UNEQUAL TREATMENT States must make provision for application at certain âoutstationedâ loca- tions in addition to local welfare offices.13 If found eligible, individuals must be furnished with medical assistance with âreasonable prompt- ness.â14 This âreasonable promptnessâ requirement has in recent years been interpreted by courts to apply not only to evidence of coverage but to medical care itself.15 Because Medicaid creates an individual legal en- titlement in eligible persons, the law conditions the denial, reduction or termination of benefits and coverage on compliance with relatively rigor- ous due process requirements (Rosenblatt et al., 1997; 2001). These pro- tections apply not only to persons receiving care in ambulatory fee-for- service settings but, under legal agency theory, to residents of private long-term care institutions as well as managed care organization enroll- ees (Rosenblatt et al., 1997; 2001). Medicaid benefits are perhaps the broadest ever granted through an insurance program. This is particularly true in the case of children, who are entitled to all forms of federally funded medical assistance, special preventive benefits, vision, dental and hearing care, and whose coverage must be furnished in accordance with a âpreventive standardâ of medical necessity (Rosenbaum and Rousseau, 2001). For beneficiaries other than those medically needy individuals who âspend downâ to eligibility, cost sharing can be nominal at most, and certain groups are entirely exempted (Schneider and Garfield, 2000). Ostensibly, Medicaid beneficiaries are entitled to freedom of choice in the selection of their medical provider.16 The Medicaid freedom of choice provision was added in 1967, after initial state implementation of Medic- aid suggested that states were relying exclusively on publicly operated health systems to furnish care (Starr, 1982; Stevens and Stevens, 1974). However, notoriously low provider participation has been a hallmark of the program for decades (Stevens and Stevens, 1974; Congressional Re- search Service, 1993). As a result, Medicaid beneficiaries continue to rely disproportionately on publicly supported systems of care such as public hospitals, health centers, and public health agencies. 13 42 U.S.C. Â§1396a(a)(55). 14 42 U.S.C. Â§1396a (a)(8). 15 See, e.g., Boulet v Celluci 2000WL 1030398 ( MA, 1st Cir.) Cramer v Chiles 33 F. Supp. 2d 342 (FL. 11th Cir., 1999), Doe v Chiles, 136 F. 3d 709 (11th Cir, 1998), Lewis v New Mexico Dept of Health 94 F. Supp. 2d 1217 (NM, 10th Cir, 2000) Benjamin v Ohl (WV, 4th Cir., 1999) King v Sullivan 776 F. Supp. 645 (RI, 1st Cir., 1991) King v Fallon 801 F. Supp. 925 (RI, 1st. Cir., 1992). McMillan v McCrimon 807 F. Supp. 475 (IL, 7th Cir., 1992) Sobky v Smoley 855 F. Supp. 1123 (CA, 9th Cir., 1994). 16 42 U.S.C. Â§1396a(a)(23).
677 FEDERAL HEALTHCARE FINANCING PROGRAMS In recent years, this so-called âfreedom of choiceâ guarantee has been significantly curtailed. Since 1997, states have been given the express de- sign option of requiring enrollment in some form of managed care ar- rangement as a condition of coverage for most beneficiaries.17 As of 2000, total Medicaid enrollment in managed care stood at 18.8 million persons, 55.8% of the Medicaid population and a 40% increase since 1996 (HCFA, 2000).18 The addition of this state option to mandate managed care enroll- ment came after two decades of state experimentation with mandatory managed care, an effort that was actively encouraged by the Clinton Ad- ministration (Rosenbaum et al., 1999a). As of 1999, virtually all states did so for some or most of their beneficiary populations, using either risk style comprehensive managed care service arrangements (akin to commercial HMOs) or more modest primary care case management systems. Although freedom of choice has never been a reality for beneficiaries and now is limited as a matter of law as a result of managed care, federal law historically has provided an additional protection. This protection, known in the law as the âequal accessâ requirement, requires states to ensure that Medicaid beneficiaries have basic equality in access to care. Specifically, the equal access provision requires that payments be âsuffi- cient to enlist enough providers so that care and services are available under the plan at least to the extent . . . as to the general public.â 19 Read literally, this provision requires equality in access, even though the actual source of care may be different because of differences in provider partici- pation. Federal Medicaid law affords states considerable discretion in the de- sign and operation of their state programs. As a result, state plans varyâ at times dramaticallyâin who qualifies for coverage, the level and range of services to which individuals are entitled; participation rules for pro- viders and provider compensation; and methods of administration, such as enrollment procedures, service delivery, and quality oversight. In re- cent years, Congress has enacted numerous reforms to give states even more flexibility with respect to who is covered, for what services and ben- efits, and through which type of arrangement. State responses to these flexibility options are quite literally all over the map, with certain states maintaining extremely broad programs and others doing only slightly more than the mandatory minimum. It is fair to say that because of the 17 The Balanced Budget Act of 1997, Pub. L. 105-33, Â§4701 et seq. 18 HCFA, Penetration Rates from 1996-2000: National Summary Table; accessed June 2, 2001, at http://www.hcfa.gov. 19 42 U.S.C. Â§1396a(a)(30). The leading case in the field is Clark v Kizer 758 F. Supp. 572 (1990), holding Californiaâs payment structure unlawful as a matter of law for its failure to produce a sufficient number of participating dentists.
678 UNEQUAL TREATMENT long-term care financing pressures that have dominated Medicaid for de- cades, states that maintain limited programs with only modest adoption of coverage options have tended to adopt those options that relate to the provision of long-term institutional care, rather than the coverage options that are designed to extend access to community residents. States also have broad flexibility in the area of coverage design, at least in the case of adults. State plans can adopt significant âamount, duration, and scopeâ limitations on classes of covered benefits. So other than in the case of children, states are relatively free to adopt any defini- tion of medical necessity they choose as long as it meets minimum tests of reasonableness.20 However, the basic Medicaid entitlement has been in- terpreted to mean that coverage decisions conducted by state agencies and their contractors (such as managed care entities, home health agen- cies, and long-term care institutions) must be based on individualized determinations grounded in relevant and reliable evidence. Unlike em- ployer benefits, decisions cannot turn on the irrebuttable application of generalized practice guidelines.21 With respect to provider participation and compensation, states have broad discretion. State agencies can fashion their own participation stan- dards and as long as the standards do not discriminate against an entire category of practitioners whose practice is authorized under state law, states have near total control over provider rules. The same is true with provider compensation: State agencies generally are free to adopt what- ever payment methodologies they choose. With the above-noted excep- tion of payments to certain types of clinics and to hospices, states have broad discretion over both payment formulas and provider participation standards. As of 2000, all states22 participated in Medicaid. Federal Medicaid con- tributions to approved state plans accounted for nearly 60% of all funds spent on medical assistance. Medicaidâs role in the financing of American healthcare is key overall and particularly strong in certain respects. In 1998 Medicaid covered approximately 20% of all children and financed one-third of all U.S. births. The program accounted for nearly 50% of all 20 See, e.g., Cowan v Myers 232 Cal. Rptr. 299 (Cal. App., 1986), cert. den. 484 U.S. 846 (1987) in which the California Supreme Court upheld a medical necessity standard under the California Medicaid program (known as MediCal) that limited coverage to services âto protect life, to prevent significant disabilities or illness, or to alleviate severe pain.â 21 Massachusetts Eye and Ear Infirmary v Commissioner of the Division of Medical Assistance 705 N.E. 2d 592 (Mass., 1999) in which the Supreme Judicial Court struck down denials of cov- erage and payment to providers that were based on irrebuttable treatment guidelines that prohibited individualized decisions as to the medical necessity of in-patient hospital care in particular patientsâ cases. 22 In this context the term state includes the District of Columbia, as well as the common- wealths of Puerto Rico and the Virgin Islands and the trust territories.
679 FEDERAL HEALTHCARE FINANCING PROGRAMS 100% 45% 54% 55% 69% 79% P riv a te /O t h e r 50% M e d ic a id 19% U n in s u re d 19% 22% 9% 8% 36% 27% 23% 22% 14% 0% W h it e , n o n - Af r ic a n Am e r ic a n L a t in o As ia n /P a c if ic N a t iv e Am e r ic a n H is p a n ic Is la n d e r FIGURE 5 Health insurance status, by race and ethnicity, 1997: Total Nonelderly Population. SOURCE: Urban Institute analysis of the March 1998 Current Popula- tion Survey, prepared for The Henry J. Kaiser Family Foundation. nursing home care, and over a third of all funds received by such âcoreâ safety net providers as public hospitals and federally funded health cen- ters (Rosenbaum and Rousseau, 2001; IOM, 2000). In general, Medicaid is a principal source of healthcare financing for minority Americans. Figure 5 shows that in 1997, the program covered one in five non-elderly African Americans, Latinos, and Native Americans compared to fewer than one in ten non-elderly white Americans (The Henry J. Kaiser Family Foundation, 1999). Figure 6 shows, the reliance on Medicaid versus private insurance is equally pronounced when only low- income beneficiaries are considered. In 1997, nearly half of all white, non- Latino beneficiaries had access to private insurance compared with only 32% and 26% of African-American and Latino beneficiaries, respectively. Concurrently, dependence on Medicaid was higher among minority populations, even when only low-income persons were considered. Unlike Medicare, the federal government does not systematically col- lect or analyze Medicaid recipient and expenditure data on the basis of race, nor does it require states to do so; so far, efforts to force such collection as a means of enforcing Title VI of the 1964 Civil Rights Act have failed23 (Rosenbaum et al., 2000). However, the limited data that are available sug- gest significant racial and ethnic disparities in the level of care received by recipients. These disparities, illustrated in Table 1 below, have been noted in the program virtually since its inception (Davis and Schoen, 1978). 23 In Madison Hughes v Shalala, 80 F. 3d 1121 (6th Cir., 1996), plaintiffs unsuccessfully sought to require the Secretary of HHS to compile Medicaid recipient and payment data on the basis of race and ethnicity. Their argument was that Title VI enforcement was impos- sible without such data.
680 UNEQUAL TREATMENT 100% 26% 27% 32% 39% 46% 29% P riv a te /O t h e r 36% M e d ic a id 50% 39% 27% 25% U n in s u re d 45% 37% 34% 29% 29% 0% W h it e , n o n - Af r ic a n Am e r ic a n L a t in o As ia n /P a c if ic N a t iv e A m e r ic a n H is p a n ic Is la n d e r FIGURE 6 Health insurance status, by race and ethnicity, 1997: Low-income nonelderly population. Low-income is defined as 200% below the federal poverty level. SOURCE: Urban Institute analysis of the March 1998 Current Population Survey, prepared for The Henry J. Kaiser Family Foundation. SCHIP: In 1997, Congress amended the Social Security Act to add a new Title XXI program known as the State Childrenâs Health Insurance Program (SCHIP). SCHIP is a federal grant-in-aid statute that specifically creates no legal entitlement in individuals. Instead, SCHIP entitles states to a defined sum (subject to aggregate annual national and state-specific limits) that can be used to furnish âchild health assistanceâ to âtargeted low-income children.â The term âtargeted low-income childrenâ applies to children under age 19 who are neither eligible for Medicaid nor cov- ered by any other form of âcreditableâ health insurance coverage as the term is used under the Public Health Service Act (HCFA, 2001b). States may use SCHIP allotments to expand Medicaid coverage (in which case all Medicaid rules apply). Alternatively, SCHIP permits states TABLE 1. Medicaid Payments per Recipient, by Race (1998) Race Dollar Payment Levels White 4609 (100%) American Indian/Alaskan Native 3297 (71%) Black 2836 (62%) Asian/Pacific Islander 1924 (42%) Hispanic 1842 (40%) SOURCE: Healthcare Financing Administration, unpublished data, published in Rosenbaum, et al., 2000.
681 FEDERAL HEALTHCARE FINANCING PROGRAMS to establish separate programs that operate directly under the authority of the SCHIP statute, which in structural design bears only passing resem- blance to Medicaid. States have near total discretion over the benefits they furnish and the manner in which they furnish coverage. Although coverage options are broad, the minimum coverage rules need only satisfy a bench- mark that is tied to the commercial insurance market rather than the broad coverage rules of Medicaid. Certain preventive benefits also must be cov- ered and cost sharing (up to a statutory maximum) is permitted. For purposes of this paper, perhaps the most notable aspect of SCHIP is that states may elect to use their funds to expand Medicaid, establish a wholly separate program for âtargetedâ children, or adopt a hybrid of the two. As of 2000, 34 states had adopted separately administered SCHIP programs that operate directly under Title XXI authority. The remaining states used their funds to expand Medicaid, thereby retaining a single form of health insurance coverage for lower-income children. Federal law permits states to establish separate SCHIP programs that stand apart from Medicaid not only in terms of their benefit packages and cost sharing rules, but also in terms of their organization and service de- livery arrangements. SCHIP contains no minimum standards regarding the relationship between the administration of a Medicaid plan and a separately administered SCHIP plan. States are free to contract with com- pletely separate providers and health plans, cover different benefits, pay different rates, and so forth. Preliminary research beginning to emerge from state SCHIP programs indicates that in states with separately ad- ministered SCHIP programs (i.e., not a Medicaid expansion) the Medic- aid child population is likely to be a minority one, while children enrolled in SCHIP are somewhat more likely to be white (Edwards, 2001). Health Disparities and Medicare and Medicaid Administration This section of the paper considers various issues in the administra- tion of Medicare, Medicaid and SCHIP that potentially affect the issue of health disparities in access, quality and outcome. In reviewing these ma- terials, it is important to bear in mind that there can be acts of either com- mission or omission where healthcare financing programs are involved. Acts of commission involve affirmative choices on the part of federal agen- cies (or state agencies, in the case of Medicaid) to adopt operating rules that have the potential to create or perpetuate disparities. An act of omis- sion, on the other hand, entails the failure to elect an option related to program structure or administration that could conceivably reduce the potential for disparities. There is virtually no evidence regarding the relationship, if any, be- tween the existence of health disparities and acts of either regulatory com-
682 UNEQUAL TREATMENT mission or omission in Medicare and Medicaid.24 Nonetheless, even a passing acquaintance with the realities of the health system and the dy- namics of healthcare allows one to hypothesize regarding the types of administration practices that could potentially contribute to the problem of disparities. The systemic choices reviewed here are ones that arise from the ad- ministration of existing programs rather than from their basic legislative framework. Put another way, the issues addressed here focus on how programs are administered, not how they are designed by Congress. Even the act of identifying systemic choices in public program administration that have disparity-creating potential probably can be viewed as highly controversial. This is because many of the types of discretionary conduct discussed here fall within the administrative discretion of agencies. But this analysis is predicated on the notion that quite apart from the legal requirements of the non-discrimination provisions of the Title VI statute and regulations, there is reason to be sensitive to those structure and de- sign choices in government financing that, while perhaps not violative of Title VI (even under a broad effects test), nonetheless give pause because of their possible consequences. Seven distinct problems in program administration are identified and discussed: 1) Medicare policies regarding conditions of participation and their impact on low income Medicare beneficiaries; 2) Medicare policies that potentially underlie the problem of racially disparate health outcomes among similarly insured Medicare beneficiaries; 3) issues related to the design of Medicaid eligibility standards and enrollment arrangements; 4) Medicaid provider compensation; 5) administration choices related to the administration of separate SCHIP programs; 6) issues related to qual- ity management and improvement; and 7) contractually sanctioned pro- vider discrimination in Medicaid managed care. 24 The Hospital Survey and Construction Act of 1946 (the Hill Burton Program) presents an analogous situation to this type of examination of the potentially racially adverse impact of public healthcare financing regulatory and administration choices. In 1979, federal Hill Burton regulations were revised to restructure program operating standards in order to eliminate the potentially racially disparate impact of previous rules. Prior rules had permit- ted Hill Burton hospitals to engage in practices that could lead to adverse regulatory conse- quences. Examples of these practices were vividly illustrated in the seminal case, Cook v Ochsner Foundation Hospital et al. 61 F.R.D. 354 (1972), which documented practices such as refusal to participate in Medicaid, the use of pre-admission deposits, and the refusal to ad- mit patients who did not have a private physician with staff privileges. All of these practices were viewed by most facilities as lawful under existing Hill Burton regulations. Following extensive hearings, the Department of Health and Human Services revised the rules because of their potential to cause both exclusion and segregation (Rosenblatt et al., 1997).
683 FEDERAL HEALTHCARE FINANCING PROGRAMS 1. Medicare conditions of participation applicable to physicians and their impact on low-income Medicare beneficiaries Federal Medicare regulations permit Medicare-participating physi- cians to select their patients at will. As a result, the federal power to set conditions of Medicare participation has failed to address the issue of low- income Medicare beneficiaries, who risk serious access barriers unless they can locate providers that also participate in and will accept Medicaid as a supplemental source of payment for uncovered deductibles and coin- surance and necessary but uncovered services. Only a small proportion of all Medicare-participating physicians and managed care organizations also participate in Medicaid. The refusal of Medicare-participating hospitals and nursing homes to participate (or participate fully) in Medicaid has been identified as both an actual and potential violation of Title VI of the 1964 Civil Rights Act (Rosenbaum et al., 2000). Despite the existence of a parallel problem in the case of other health providers, the federal government has not attempted to systemati- cally lessen the problem of Medicaid non-participation that occurs among Medicare-enrolled physicians25 and health plans. The refusal by physicians, managed care companies and other pro- viders to participate in Medicaid, at least in the case of low-income Medi- care patients who need Medicaid to âperfectâ their Medicare coverage, has a disproportionate adverse impact on minority beneficiaries because of their higher representation among poor beneficiaries and their lower level of private coverage. Without the ability to use supplemental Medic- aid coverage, patients would face the choice of foregoing treatment by a non-Medicaid-participating physician or paying out-of-pocket uncovered premiums, deductibles and coinsurance. As a practical matter of course, this second option is out of the question. The problem of non-participation in Medicaid, at least in the case of low-income Medicare beneficiaries, is exacerbated by state Medicaid pay- ment choices in response to the Balanced Budget Act of 1997. This gave state Medicaid programs the option of limiting payments for services cov- ered under both programs to the nearly always lower Medicaid rate rather than the higher Medicare rate. Previously, a physician who chose to par- 25 Indeed, as noted above, despite the transformation of Medicare from an indemnity pay- ment system to one in which most physicians participate directly, there are no formal regu- lations regarding the status of direct Medicare payments to physicians as a source of federal financing for Title VI enforcement purposes, although references to this effect are contained in 2000 OCR guidance on services to individuals with limited English proficiency. Regard- less of whether Medicare participation by physicians is deemed to be the receipt of federal assistance for Title VI purposes, it establishes a basis to regulate their conduct toward poor patients who depend on Medicaid supplemental insurance.
684 UNEQUAL TREATMENT ticipate in both programs could be assured that Medicaid would cover at least those deductibles and coinsurance requirements imposed under Medicare (i.e., that Medicaid would pay up to the Medicare payment rate). This is no longer the case; state agencies now have the option of capping their payments at the normal Medicaid rate which, given the low pay- ment levels under Medicaid, are depressed even in relation to already low Medicare physician payment levels. 2. Medicare administration issues underlying racial disparities in health outcomes among similarly insured Medicare beneficiaries Much of the literature on race-based health disparities focuses on dis- parate access to certain types of treatment among similarly situated hospitalized beneficiaries and other beneficiaries undergoing medical treatment (U.S. Civil Rights Commission, 1999). In considering how Medicareâs structural dimensions might contribute to these disparities, it may be important to focus on the utilization review process that is used by Medicare (and virtually all other insurers) to control access to advanced inpatient care. In the modern health system, utilization review approval is a prerequisite to virtually all advanced care. While the literature on the actual dynamics of utilization review is slender, case law suggests that a basic element of utilization review is the extent to which a physician is willing to aggressively advocate for the needs of a patient as part of the treatment approval process.26 As a result, successful navigation of the utilization review process in the case of complex treatments may hinge to a significant degree on the existence of a âcommitted sponsorâ relation- ship of the type explored by Duff and Hollingshead in their seminal study of physician/patient relationships (Duff and Hollingshead, 1968). Low-income beneficiaries may be far less likely to have access to a health specialist who is a âcommitted sponsorâ and thus may fare less well within Medicare utilization review. Low-income Medicare patients, who are disproportionately minority, face barriers to physician acceptance for the reasons explored above. Furthermore, those primary care physi- cians who do treat lower-income beneficiaries in ambulatory settings and who may be able to help link a patient to specialty care may lack the colle- gial and referral relationships with medical specialists and specialized treatment centers that committed sponsorship in the specialized inpatient 26 In Wickline v State of California 239 Cal. Rptr., 810 (Cal. App., 1986), petition for review dismissed, 741 P. 2d 613 (Cal., 1987) the court specifically addressed the liability of physi- cians who fail to act as committed sponsors for their patients in response to an adverse utilization review decision.
685 FEDERAL HEALTHCARE FINANCING PROGRAMS setting demands.27 To the extent that a committed sponsor relationship is lacking, so too might be the aggressive advocacy integral to better ensur- ing that complex treatments ultimately are judged to be medically neces- sary and appropriate under prospective and concurrent review. In sum, in a world in which successful navigation of utilization re- view acts as a precondition to access to highly specialized in-patient pro- cedures, patients without committed sponsors may fare less well in the utilization review process. Inadequate direct access to specialists because of low acceptance rate coupled with dependence on primary care physi- cians who lack significant collegial networks may reduce the likelihood that once admitted, a patient will be under the care of a âcommitted sponsorâ specialist who can help pave the path through the utilization review process. If âcommitted sponsorshipâ of the type observed by Duff and Hollingshead a generation ago continues to play an equally vital role in the modern world of healthcare quality, particularly given the more aggressive efforts to control access to care found in utilization manage- ment and prospective case review, then the inability of lower income patients to attach to a committed specialist (or be attached to one through a collegial referral) may have a major impact on the outcome of Medicare utilization review. 3. Administrative choices in setting Medicaid eligibility standards and enrollment arrangements Studies suggest that individuals and families place great value in the Medicaid program (Stuber et al., 2000; Kaiser Commission, 2000). At the same time, research also suggests that the manner in which states admin- ister their programs creates serious problems of stigma. The stigma per- ceived by families can be traced to how they are treated by two distinct groups: The individuals who enroll them in the program (and the settings in which they work) and healthcare providers (discussed below). To the extent that statesâ choices in eligibility and enrollment cause stigma, it is probably safe to assume that this consequence falls disproportionately on minority beneficiaries, given their disproportionate dependence on the program and the actual or perceived problem of prejudicial attitudes and beliefs within the underlying health system. Although Medicaid contains a number of specific requirements re- lated to program administration, states also have broad latitude in how 27 Indeed, in discussing specialist referrals with physicians in practice at federally funded community health centers, the author has heard on innumerable occasions about the diffi- culties these physicians have in locating specialists who will accept and aggressively man- age their referrals as a result of financial and other considerations.
686 UNEQUAL TREATMENT they design their plans. A state can extend Medicaid to the entire low- income population regardless of disability, age, work status, or the pres- ence of dependent children. Federal financial participation is available for all low-income individuals other than non-elderly, non-disabled adults without children. In other words, with the exception of childless, work- ing-age adults, federal financial participation can be claimed for all low- and moderate-income children and adults, thereby removing the percep- tion that Medicaid is only for welfare recipients and families that choose not to work (Rosenbaum and Rousseau, 2000; Schneider et al., 1998). State flexibility does not stop at the point of eligibility design. Even though only welfare agencies can by law render a formal eligibility deter- mination, these agenciesâ involvement in the actual eligibility determina- tion process can be rendered virtually invisible (Cohen Ross, 2000). En- rollment arrangements can be established in out-stationed locations (e.g., health clinics, physiciansâ offices, the work place, churches, supermarkets, etc.). Applications can be extremely short in any state that elects to deter- mine eligibility on the basis of income alone. Other than proof of legal status or citizenship, program rules require virtually no written verifica- tion or documentation of eligibility factors (i.e., affirmance of eligibility is sufficient). Enrollment periods can be set to last for a duration of 12 months, with full federal financial participation available for children re- gardless of any change in eligibility status during the 12-month period. Redeterminations of eligibility can be accomplished completely by mail, using âpassiveâ procedures that automatically renew coverage in the ab- sence of evidence to the contrary. In short, a state can design its program to virtually eliminate all of the stigma-producing events that arise from the eligibility, eligibility deter- mination, and enrollment process itself. Federal financial participation is available for most groups of individuals. Furthermore, the federal gov- ernment in recent years has been relatively generous about granting states the authority to conduct federally funded demonstrations that de-catego- rize program eligibility (Rosenbaum et al., 1999). The leader in this effort to transform Medicaid from a âwelfare piggy- backâ program to a true public insurance program has been Medicaid, whose program now serves individuals without access to employer cov- erage, as well as persons who need subsidization in order to secure ben- efits available through their employers. States that have taken active steps designed to produce destigmatizing program structures that move Med- icaid away from welfare and toward a more neutral public insurance stance are Oregon, Tennessee, Arizona, and Rhode Island. At the same time, many states retain a dated version of Medicaid, strictly limiting cov- erage and enrollment options and forcing the program to run like welfare. This choice helps perpetuate attitudes and beliefs on the part of provid-
687 FEDERAL HEALTHCARE FINANCING PROGRAMS ers, persons in need of assistance with insurance (and who does not need assistance with insurance costs today?), policy makers, and the public at large that while it is acceptable to subsidize individual health benefits through deep tax breaks, a direct subsidy is a mark of shame. 4. Administrative choices in payment of Medicaid providers It is perhaps safe to say that the best-known problem plaguing the Medicaid program is its notoriously low payment rates. A recent GAO study documented the practice on the part of business consultants of en- couraging physicians to âration âcertain lower paid patientsâ and âtell some of the higher paid patients to come right on inââ (GAO, 2001). According to taped consultant interviews that were played at a Senate hearing at which the results of the GAO investigation were presented, the following conversation occurred: Referring to the advice a consultant gave a physician practice, the con- sultant joked, âSo what we said about to do there, you have to ration your Medicaid, and if anyone calls from Blue Cross/Blue Shield, you say, âWhen do you want to come in? Weâll come and get you.â â The consultant said that one way of discouraging Medicaid patients while welcoming private pay patients whose insurance policies often reimburse at higher rates, is to give Medicaid patients the most inconvenient ap- pointment times while saving the most popular appointment slots for private pay patients. 28 Low rates, whether set for physicians, pharmacies, managed care or- ganizations, nursing homes, home health agencies, or other health suppli- ers have several potential effects, all of which fall with disproportionate impact on minority patients. First, low rates make it impossible for any provider but those that are heavily dependent on Medicaid revenues (e.g., core safety net providers such as public hospitals and health centers, childrenâs hospitals, and nursing facilities) to participate in the program. The loss of revenues as a result of steep contractual allowances is simply so steep that any significant level of participation becomes economically out of the question for other than small classes of providers. Second, as the GAO study illustrates, low rates also induce fraudulent practices aimed at rationing care, discouraging access, discriminating in the provi- sion of services, and excluding patients from practices. Low payment rates in effect encourage and validate bad provider con- duct. To the extent that the attitudes and beliefs regarding minority pa- 28 BNA, 2001. Unscrupulous Consultants Tell Providers how to Cheat Third Party Payers, GAO Finds, Health Care Policy Report 9:26 (July 2, 2001), pp. 1037-1038.
688 UNEQUAL TREATMENT tients remain significant matters within the provider community, grossly low payment rates offer a ready-made excuse for non-participation that rests entirely on neutral economic grounds. This excuse is particularly available to urban healthcare providers who are plentiful in number to the point of saturation in more affluent communities. As Table 2 illus- trates, because of the high concentration of minority individuals in the poorest large city neighborhoods, this refusal to participate also has its greatest adverse impact on minority beneficiaries who ultimately are starved for access in the midst of plenty. Table 2 shows the marked racial and ethnic patterns of urban poverty in the largest cities. While the concentration of urban poverty increased for all races between 1970 and 1990, by 1990, 83% of all urban African- American poor persons resided in census tracts that could be labeled as poor, while nearly 42% resided in high-poverty census tracts. Latino pov- erty concentrations were somewhat less pronounced but decidedly el- evated above overall rates. In 1990, 68% of all poor persons residing in the nationâs 100 largest cities were either African American or Latino, and an astonishing 87% of all poor persons residing in these citiesâ highest TABLE 2 Concentration of Povertyâ100 Largest Cities, 1970-1990 1970 1980 1990 Total # poor 7,542,479 8,133,277 9,392,953 % poor 14.5% 16.7% 18.3% Poor in poverty (>20%) tracts 4,156,543 5,178,509 6,466,097 % poor in poverty tracts 55.1% 63.8% 68.8% Poor in high poverty 1,240,855 1,828,576 2,650,142 (> 40% poor) tracts % in high poverty tracts 16.5% 22.5% 28.2% African American # poor 3,182,881 3,428,593 4,002,094 % population poor 27.7% 27.2% 29.9% Poor in poverty (>20%) tracts 2,567,429 2,837,386 3,328,652 % poor in poverty tracts 80.7% 82.5% 83.2% Poor in high poverty 895,920 1,157,537 1,664,872 (> 40% poor) tracts % in high poverty tracts 28.1% 33.8% 41.6% # poor 966,413 1,575,569 2,394,890 Latino % population poor 23.2% 26.2% 21.7% Poor in poverty (>20%) tracts 664,375 1,162,367 1,842,990 % poor in poverty tracts 68.8% 73.8% 77.0% Poor in high poverty 196,202 378,832 650,747 (> 40% poor) tracts % in high poverty tracts 20.3% 24.0% 27.2% SOURCE: Fossett and Perloff, 1999.
689 FEDERAL HEALTHCARE FINANCING PROGRAMS poverty urban census tracts were members of these racial and ethnic groups. These statistics underscore the particularly serious impact that provider non-participation in Medicaid and depressed Medicaid payment levels potentially could have on minority beneficiariesâ access to care. Low provider compensation rates have other pernicious effects as well. The failure of Medicaid programs to include capital payments in their compensation rates to safety net providers that so frequently anchor poor communities seriously limits their ability to engage in the level of renovation, facility and practice improvement, and overall technical up- keep that is essential to maintaining a safe and good quality healthcare environment. Equipment cannot be updated or replaced as needed. Building space essential to the expansion of capacity (along with greater employment opportunities in poorer neighborhoods) cannot be added. Supplies cannot be maintained. And finally, the recruitment of personnel and health professionals becomes even more difficult because of the de- pressed working conditions. The federal Medicaid equal access requirement described previously does not guarantee precisely the same pattern of access that privately in- sured persons have. But it does require that states maintain payment lev- els that are reasonable to enlist sufficient providers to achieve an equal overall level of access to care. Related to this equal access requirement is the requirement that medical assistance be furnished promptly, a basic operating rule that in recent years, as noted earlier, has been used to ex- pand access to care, particularly in the case of persons with disabilities. Other than isolated litigation efforts designed to challenge grossly low provider payment levels, this basic requirement of the program has at- tracted no attention other than from the nationâs governors who have pe- riodically called for its repeal.29 The federal government has done virtu- ally nothing with the provision, and there are virtually no guidelines that interpret how to apply the equal access requirement or what is expected in terms of state implementation (e.g., specific data collection to measure levels of access where disparities in health outcomes are pronounced, af- firmative efforts to increase rates, or affirmative efforts aimed at attract- ing healthcare providers in high need communities). 5. Administration of separate SCHIP programs As noted earlier, SCHIP permits states to use their allotments to es- tablish and operate separate SCHIP programs. Research on SCHIP is just 29 Under pressure from the governors, the Balanced Budget Act of 1997 repealed specific equal access provisions related to obstetrical and gynecological care but left the overall re- quirement in tact.
690 UNEQUAL TREATMENT underway, but several issues are becoming apparent. First, in many states with separate programs, the SCHIP population is whiter than the Medic- aid population because of racially identifiable poverty distribution. Mi- nority children simply are more likely to be extremely poor, and thus any state with a separately administered SCHIP program is more likely to have a SCHIP population that is white in relation to its Medicaid popula- tion. As the inner-city data presented earlier underscore, SCHIP and Med- icaid children also are likely to reside in different communities, with Med- icaid children more concentrated in inner-city poor neighborhoods, and SCHIP children (those with incomes about twice the federal poverty level), throughout a metropolitan area. Thus, even without taking any step other than deciding to set up a separate SCHIP program, a state that does so likely faces a situation in which its Medicaid children are more likely to be minority children, and its SCHIP children are more likely to be white. This racial skewing of children receiving public insurance into two sub-groups may pose problems in and of itself, since these patterns create racial imprimaturs for the programs. Added to this problem however, is the fact that early research conducted by the George Washington Univer- sity Center for Health Services Research and Policy, as well as anecdotes from around the country, indicate that states with separate programs are permitting their physicians to participate in SCHIP but not Medicaid, per- mitting managed care organizations and insurers to sell to SCHIP agen- cies but not to Medicaid, and even paying better rates under SCHIP and not Medicaid. These choices in design and administration obviously have the potential to take a bad situation and make it far worse, labeling minor- ity children as members of substandard health coverage arrangements reserved for minorities and outside of the healthcare mainstream. No federal regulations address this problem. 6. Administrative choices in establishing conditions of participation and quality of care measurement; self examination by federal and state governments The final problem is one that affects Medicare, Medicaid, and SCHIP. There is very little in the conditions of participation under these three programs that requires or finances the efforts of healthcare providers to take systematic steps to examine enrollment and utilization patterns in relation to the demographics of the communities in which they serve, and undertake affirmative steps to improve access to their services. Limited conditions of participation under federal Medicare and Medicaid man- aged care regulations do require that participating managed care organi- zations make certain efforts to address access to care in their service ar-
691 FEDERAL HEALTHCARE FINANCING PROGRAMS eas.30 But there is no affirmative obligation on the part of either providers as a condition of participation or participating states in the case of Medic- aid and SCHIP to collect and analyze health data on access and utilization by race, examine health outcomes by race, examine patterns of healthcare administration that conceivably could contribute to racially identifiable outcomes, or take affirmative steps to attempt to remedy these problems through restructuring of healthcare delivery arrangements. Notably, state Medicaid programs are far ahead of the federal gov- ernment in the case of managed care organizations; their contracts with managed care organizations typically contain extensive access require- ments related not only to networks but to hours, locations of services, cultural competency and translation services, and other steps designed to remove barriers that disproportionately could affect minority enrollment and utilization (Rosenbaum et al., 1997; 1998; 1999a). Only in the case of the federally funded community health centers program does one find federal policies aimed at engendering this type of careful self-examina- tion by health providers on an ongoing basis to determine whether access to care is appropriate. Not since the federal government undertook such an effort in the mid-1970s as part of its revision of Hill Burton hospital regulations31 has there been this type of careful self-assessment of the fed- eral governmentâs (or state governmentsâ) payment practices or the prac- tices of federally assisted providers. Indeed, in the case of Medicaid, the federal government neither collects nor requires much racial data. 7. Administrative choices in the design of managed care systems: contractually sanctioned discrimination in provider networks As noted previously, federal law provides states with broad leeway in the design of their managed care arrangements and selection of man- aged care contractors. Despite the âequal accessâ provisions in the Med- icaid statute (noted above), most state contracts with managed care orga- nizations do not expressly prohibit contractorsâ provider networks from engaging in what can be termed âcontractually sanctioned discrimina- tion,â i.e., permitting network providers under a general duty of care to all plan members to nonetheless refuse to treat the Medicaid sponsored members of the plan (Rosenbaum et al., 1997). The issue of segregated provider networks (i.e., networks operated by Medicaid participating 30 The Medicaid managed care regulations are now in suspense by the Bush Administra- tion. Medicare+Choice regulations require merely that a provider network be âsufficient to provide adequate access to covered services to meet the needs of the population served.â 42 C.F.R. Â§422.112()(1). 31 See note 25, supra.
692 UNEQUAL TREATMENT managed care organizations that subdivide members based on sponsor- ship status) has been the subject of both speculation and actual legal chal- lenges (Rosenbaum et al., 2000). Anecdotal discussions with physicians and other health professionals who participate in managed care plans sug- gest a great unwillingness on their part to accept Medicaid-sponsored members into their practices, either because of payment differentials or because of discomfort with Medicaid-sponsored members.32 As serious as network provider discrimination might be at the pri- mary care stage, in communities with health centers, safety net provider clinics, and other sources of primary care for medically underserved popu- lations, there may be enough primary care access to overcome the worst tangible effects of internally sanctioned discrimination (although the long- lasting tangible and intangible effects of contractually sanctioned discrimi- nation could never fairly be calculated). At the point of specialty care how- ever, the real and immediate impact of sanctioned network discrimination could be enormous, since permitting specialists to refuse to accept or treat referred Medicaid-sponsored plan members is tantamount to the denial of specialty care. Unless a managed care organization was to literally run two entire specialty networks, contractually sanctioned discrimination against Medicaid beneficiaries could have an incalculable effect on access to specialized services. Because of disproportionate minority representa- tion within the Medicaid-sponsored managed care enrollee population, the impact would be felt most heavily by minority patients. Regulations issued by the Clinton Administration in January 2001 and applicable to Medicaid managed care systems prohibit contractually sanc- tioned provider discrimination against Medicaid patients within Medic- aid-participating managed care organizations.33 On August 20, 2001, the Bush Administration suspended these rules.34 The Administration simul- taneously proposed new regulations that seek to relax certain of the re- quirements imposed on state agencies and managed care organizations 32 This position on the part of providers serves to at least informally dispel any notions that managed care would somehow erase healthcare access differentials based on sponsor- ship. Medicaid-only managed care plans are the norm in many communities, and in com- munities in which MCOs that do business across sponsors are in the market, separate Med- icaid-only subsidiary operations may be common. There are many reasons to maintain a special Medicaid subsidiary, because the specifications of a Medicaid contract offer differ enormously from those found in a commercial agreement. Furthermore, the geographic isolation in which beneficiaries may live (particularly in the case of inner city residents) may justify enhanced provider networks in order to ensure adequate access in underserved com- munities. These affirmative reasons for maintaining a separate Medicaid business are a different matter from doing so in order to isolate and separate Medicaid customers. 33 42 C.F.R. 438.206(d)(7). 34 66 Fed. Reg. (August 17, 2001).
693 FEDERAL HEALTHCARE FINANCING PROGRAMS under the January 19 rule.35 One of the rules eliminated in its entirety is the regulation that explicitly prohibits discrimination against Medicaid beneficiaries. The implicit message sent by the repeal of this express anti- discrimination provision is that while general compliance with civil rights laws remains a requirement,36 contractually sanctioned discrimination based on payer status is no longer specifically prohibited. It is conceivable of course that the Health and Human Services Office for Civil Rights (OCR) could conclude upon investigation that federal civil rights regulations are violated by contractually sanctioned discrimination against Medicaid patients by network providers.37 To date however, OCR does not appear to have taken such a position, nor has it developed stan- dards to clarify the legality of this practice under Title VI. Furthermore, by relying on general Title VI sanctions rather than expressly prohibiting patient âredliningâ by member sponsorship status, the Administration essentially foregoes an opportunity to set an explicit standard designed to directly address an identified problem issue in Medicaid managed care that has the potential to hurt not only all Medicaid beneficiaries but dis- proportionately harm minority patients. Putting aside Title VI, the repeal of such a regulation appears to have direct implications for the enforce- ability of the equal access provisions of the Medicaid statute themselves. Discussion There is no question regarding the contribution that public financing programs have made to improving health and healthcare for minority Americans. At the same time, the data presented in this paper underscore the disproportionate dependence on these programs that minority indi- viduals maintain, as well as their vulnerability to heightened health risks and reduced access to healthcare because of their poverty and where they live. Together Medicare, Medicaid, and SCHIP pumped better than $350 billion into the American healthcare system in 2000. At the same time, a review of key issues in program administration, such as treatment of low- income Medicare beneficiaries, Medicaid eligibility and enrollment prac- tices, provider recruitment and payment, conditions of participation for health providers, and overall program management by both agencies and 35 66 Fed. Reg. 454564 (August 20, 2001). 36 The NPRM does not repeal the January 19 rule requiring compliance by MCOs with Title VI as well as other applicable federal civil rights statutes. 37 In light of the recent Supreme Court decision in Alexander v Sandoval, 121 S.Ct. 1511 (2001), the ability of beneficiaries to directly challenge such discrimination as a violation of the Title VI regulations is thrown into doubt, as are the discriminatory effects regulations themselves.
694 UNEQUAL TREATMENT participating providers reveals that little has been done to make affirma- tive use of this vast purchasing leverage to both promote and finance customization of the healthcare to better meet the needs of minority pa- tients and blunt or minimize prejudicial attitudes. Indeed, if anything, stigmatizing enrollment arrangements and dramatically low payment lev- els have not only tacitly sanctioned provider and system aversion but have affirmatively encouraged the rejection of lower income patients. Even in SCHIP, where the entire population is lower income, emerging informa- tion suggests that states are unwittingly creating a dynamic that encour- ages the racially identifiable treatment of children within the publicly fi- nanced health system. What will it take to fix these problems? Two issuesâfinancial and politicalâneed to be addressed. First, making the programs more ori- ented to minority patients will require significant financial investments. Destigmatizing Medicaidâs eligibility and enrollment arrangements not only requires funds to underwrite eligibility expansion but would also, if past reform efforts are any gauge, result in greater enrollment rates and would thus push program costs up. Medicaid payments are so depressed that even modest proportional increases in payments necessitate major outlays, and Medicaid provider participation research suggests that mod- est rate increases in fact result in little change in the system. While money is always a problem, and can be expected to become increasingly so in an era of declining rates of government revenues, the financial problems actually pale in comparison to the two awesome po- litical problems that arise in any restructuring discussion: the healthcare industry and state governments. Restructuring Medicare and Medicaid administration to emphasize orientation toward minority patients and beneficiaries as a condition of federal financial participation means confronting the fundamental char- acter of both programs. As Marilyn Moon has observed, the context for enactment of the programs was that they would require nothing of health providers (Moon, 1993). The promises made at the birth of Medicare and Medicaid (and once again at the birth of SCHIP) were that participation would be voluntary and that few if any conditions of participation would be imposed. Despite the presence of direct government financing, pro- viders could continue to select their patients and their markets, and to a greater or lesser degree could continue accountable and open to minori- ties. The act of literally pulling physician payments out of Title VI enforce- ment authority in 1965 is emblematic of the delicacy with which govern- ment payments were overlaid on the healthcare system. To the extent that anyone believes that provider attitudes regarding government regu- lation of their health practices have softened, one need only look at the
695 FEDERAL HEALTHCARE FINANCING PROGRAMS recent dust-up over the Office for Civil Rights Limited English Proficiency Guidelines, when dozens of prominent provider groups in April 2001 joined together to actively protest to the White House the application of these guidelines to healthcare. It is true that much has changed since 1965. Providers now are in managed care networks and are obligated to accept as patients the mem- bers who select or are assigned to them.38 Payment arrangements are now direct rather than indemnity in nature. Old racial barriers and atti- tudes certainly have softened if not dissipated. But the notion of telling Medicare-participating physicians that they must participate in Medicaid at least to the extent that they serve low-income Medicare beneficiaries would strike most persons as an utterly radical idea and one that lies be- yond the furthest reaches of permissible payer leverage over the health system. The other political behemoth is state governments. States always have operated Medicaid with substantial levels of autonomy; this autonomy has grown over the years, as the power of governors has increased and as succeeding administrations have lessened their regulatory enforcement of the Medicaid statute. Were a federal agency to suddenly impose a series of regulatory requirements related to eligibility and enrollment de- sign, provider compensation, and assessment of program impact on mi- nority families, the federal officials prescribing such changes probably would be regarded as daft and most likely would lose their jobs. Where does that leave policy in this area? The enormous difficulty of achieving changes of the magnitude described here means that the effort to make changes will only work if they are constantly placed in front of policy makers and program administrators and if the changes that are identified as potentially beneficial are tied to incentives. Congress might consider extending additional levels of compensation to both state agen- cies and providers that take steps to orient programs toward minority patients and away from practices that result in segregation, exclusion, and denial of care. Also necessary is sufficient health services research to sup- port the claim that certain healthcare financing decisions and service ar- rangements are at least associated with better (or poorer) access to health- care and health outcomes among minority beneficiaries. There are certain practices that appear to create so much dispar- ity that careful consideration should be given to how to stop them. 38 In this regard, it is important to note that provider challenges to âall productsâ clauses have meant that in practice, managed care companies that sell their products to both public and private payers may in fact set up separate subsidiaries that do Medicaid business with separate networks or else may maintain separate networks for their Medicaid customers (Rosenblatt et al., 1997; 2001).
696 UNEQUAL TREATMENT Above all is the refusal of providers to participate in Medicaid in the case of low-income Medicare beneficiaries and the tendency of sepa- rately administered state SCHIP programs to segregate white, near-poor children from minority poor children in access and coverage. Both of these practices, as well as the practice of permitting Medicaid managed care plans to treat enrollees through separate networks, appear to di- rectly countenance a form of payer segregation that comes close, at least in principle, to segregated waiting rooms and hospital wings. The elimi- nation of these practices should be matters of first priority for federal and state policy makers. REFERENCES Cohen Ross, D. 2000. Making it Simple (Kaiser Commission on Medicaid and the Unin- sured, Washington DC). Commerce Clearinghouse, 2001. Medicare and Medicaid Guide. Chicago, IL: Commerce Clear- inghouse. Committee on Ways and Means, United States House of Representatives. 1996. Green Book. Washington DC, GPO 1996. Congressional Budget Office. 2001. The Budget and Economic Outlook: 2001-2011. Washington DC. Accessed June 15, 2001, at www.cbo.gov. Congressional Research Service. 1993. Medicaid Source Book: Background Data and Analysis. Washington DC, GPO. Davis, K. and Schoen, C. 1978. Health and the War on Poverty. Washington DC: Brookings Institution Press. Duff, RS. and Hollingshead, AB. 1968. Sickness and Society. New York: Basic Books. Edwards, J. 2001. Presentation at 2001 annual meeting of the Association for Health Services Research, June 10-13 (Atlanta, GA). Einbinder, LC. and Schulman, KA. 2000. The Effect of Race on the Referral Process for Inva- sive Cardiac Procedures. Medical Care Research and Review 57:Supplement 1:162-181. Ellwood, M. and Ku, L. 1998. Welfare and Immigration Reforms: Unintended Side Effects for Medicaid Health Affairs 17:3 (May/June) pp. 137-152. Fein, R. 1986. Medical Care, Medical Costs. Cambridge, MA: Harvard University Press Fossett, J. and Perloff, J. 1999. The âNewâ Health Reform and Access to Care: The Problem of the Inner City Access to Health Care: Promises and Prospects for Low Income Americans (Lillie-Blanton, M., Martinez, RM, Lyons, B. and Rowland, D., eds.). Washington DC: The Henry J. Kaiser Family Foundation. Gaskin, D. and Hoffman, C. 2000. Racial and Ethnic Differences in Preventable Hospitaliza- tions Across 10 States Medical Care Research and Review 57 (Supplement 1):85-108. General Accounting Office (GAO). 2001. Consultantsâ Billing Advice May Lead to Improperly Paid Insurance Claims (GAO-01-818). Hawkins, D., Gevorgyan, M., Lopes, A. and Rosenbaum, S. 2000. The Use of Community Health Centers In Countries with National Health Insurance: Evidence from the Literature. Washington DC: National Association of Community Health Centers. Health Care Financing Administration (HCFA). 2001a. Medicare web site, accessed July 3, 2001, at www.hcfa.gov/medicare. Health Care Financing Administration (HCFA). 2001b. Medicare web site, accessed July 3, 2001, at www.hcfa.gov/schip.
697 FEDERAL HEALTHCARE FINANCING PROGRAMS Health Care Financing Administration (HCFA). Managed Care Penetration Rates from 1996- 2000: National Summary Table. Accessed June 2, 2001, at http://www.hcfa.gov. The Henry J. Kaiser Family Foundation. 1999. Key Facts: Race, Ethnicity and Medical Care. Accessed June 20, 2001, at www.kff.org/minority. The Henry J. Kaiser Family Foundation. Medicare and Minority Americans. Accessed June 20, 2001, at www.kff.org/medicare. Institute of Medicine (IOM). 2000. Americaâs Health Care Safety Net: Intact but Endangered. Washington DC: National Academy Press. La Viest, TA., Nickerson, K., and Bowie, J. 2000. Attitudes about Racism, Medical Mistrust, and Satisfaction with Care Among African American and White Cardiac Patients. Medical Care Research and Review 57 (Supplement 1):146-162. Lillie-Blanton, M., Brodie, M., Rowland, D., Altman, D. and McIntosh, M. 2000. Race, Ethnicity and the Health Care System: Public Perceptions and Experiences. Medical Care Research and Review 57 (Supplement 1):218-236. Marmor, T. 1970, 1973, 2000. The Politics of Medicare (2d Ed.). New York: Aldine De Gruyter. Mayberry, R., Mili, F., and Ofili, E., 2000. Racial and Ethnic Differences in Access to Medical Care. Medical Care Research and Review 57 (Supplement 1):108-146. Moon, Marilyn. 1993. Medicare Now and In the Future. Washington DC: Urban Institute Press. Rosenbaum, S. and Rousseau, D. 2001. Medicaid at Thirty-Five. St. Louis University Law Jour- nal 45(1):7-53. Rosenbaum, S. and Maloy, KA. 1999. The Law of Unintended Consequences. Ohio State Law Journal 60(4) 1443-1479. Rosenbaum S., Smith, B., Sonosky, C., Shaw, K., Shin, P., Zakheim, M. 1997, 1998, 1999a. Negotiating the New Health System: A Nationwide Study of Medicaid Managed Care Con- tracts. Washington DC: The George Washington University School of Public Health and Health Services. Rosenbaum, S., Darnell, J., Keenan, PS. and Simon, L. 1999b. Section 1115 Medicaid Waiv- ers: Charting A Path for Medicaid Managed Care Reform. Access to Health Care: Prom- ises and Prospects for Low Income Americans (Lillie-Blanton, M., Martinez, RM., Lyons, B. and Rowland, D., eds.). Washington DC: The Henry J. Kaiser Family Foundation. Rosenbaum, S., Markus, A., and Darnell, J. 2000. U.S. Civil Rights Policy and Access to Health Care by Minority Americans: Implications for a Changing Health Care System. Medical Care Research and Review 57 (Supplement 1):236-260. Rosenblatt, R., Law, S. and Rosenbaum, S. 1997. Rosenblatt R., Rosenbaum S., and Frankford, D. 2001. Law and the American Health Care System. New York, NY: Foundation Press. Schneider, A., Fennel, K., and Long, P. 1998. Medicaid Eligibility for Families and Children. Washington DC: Kaiser Commission on Medicaid and the Uninsured. Schneider, A. and Garfield, R. 2000. Medicaid Benefits. Washington DC: Kaiser Commission on Medicaid and the Uninsured. Schulman, K., Berlin, J., Harless, W., Kerner J., Sistrunk, S., Gersh, B., Dube, R., Taleghani, C., Burke, J., Williams, S., Eisenberg, J. and Escarce, J. 1999. The Effect of Race and Sex on Physiciansâ Recommendations for Cardiac Catheterization. New England Journal of Medicine 340(8):618-626. Smith, D. 1999. Health Care Divided Ann Arbor, MI: The University of Michigan Press. Starr, P. 1982. The Social Transformation of American Medicine. New York, NY: Basic Books. Stevens, R. and Stevens R. 1974. Welfare Medicine in America: A Case Study of Medicaid. New York, NY: Free Press. Stuber, J., Maloy, K., and Rosenbaum S. 2000. Stigma and the Medicaid Program. Washington, DC: The George Washington University School of Public Health and Health Services, Center for Health Services Research and Policy. www.gwu.edu\~chsrp\whatâsnew.
698 UNEQUAL TREATMENT U.S. Commission on Civil Rights. 1999. The Health Care Challenge: Acknowledging Disparity, Confronting Discrimination, and Ensuring Equality. Washington DC. U.S. Department of Health and Human Services. 2001. Mental Health: Culture, Race and EthnicityâA Supplement to Mental Health: A Report of the Surgeon General. Rockville, MD: U.S. Department of Health and Human Services. Weinick, R., Zuvekas, S. and Cohen, J. 2000. Racial and Ethnic Differences in Access to and Use of Health Care Services, 1977 to 1996. Medical Care Research and Review 57 (Supple- ment 1):36-55.