Understanding and Addressing Racial Disparities in Health Care

David R. Williams, Ph.D., M.P.H., and Toni D. Rucker, Ph.D.

Racial disparities in medical care should ease and death rates for multiple condi- be understood within the context of racial tions. Although the role of medical care as inequities in societal institutions. a determinant of health is somewhat limit- Systematic discrimination is not the aber- ed, medical care (especially pre v e n t i v e rant behavior of a few but is often support- care, early intervention and the appropri- ed by institutional policies and unconscious ate management of chronic disease) can bias based on negative stere o t y p e s . play an important role in health (Bunker, Ef fectively addressing disparities in the Frazier, and Mosteller, 1995). Thus, racial quality of care req u i r es improved data sys- and ethnic differentials in the quantity and tems, increased reg u l a t o r y vigilance, and quality of care are a likely contributor to new initiatives to appropriately train med- racial disparities in health status. ical professionals and re c ruit more Compared with white persons, black per- p roviders from disadvantaged minority sons and other minorities have lower levels ba c k g r ounds. Identifying and implement- of access to medical care in the United ing effective strategies to eliminate racial States due to their higher rates of unem- inequities in health status and medical care ployment and under-re p resentation in should be made a national priority. good-paying jobs that include health insur- ance as part of the benefit package INTRODUCTION (Blendon et al., 1989; Trevino et al., 1991). M o re striking, and disconcerting to National data reveal that over the past 50 many is the large and growing number of years, the health of both black and white studies that find racial diffe r ences in the persons has improved in the United States receipt of major therapeutic proc e d u r es for as evidenced by increases in life expectan- a broad range of conditions even after cy and declines in infant and adult mortali- adjustment for insurance status and severi- ty (National Center for Health Statistics, ty of disease (Harris, Andrews, and 1998). However, black persons continue to E l i x h a u s e r, 1997; Wenneker and have higher rates of morbidity and mortal- Epstein,1989). Especially surprising to ity than white persons for most indicators many are the racial disparities in contexts of physical health. Hispanics and wh e r e diffe r ences in economic status and American Indians also have elevated dis- insurance coverage are minimized such as the Veterans Health Administration System The authors are with the University of Michigan. Preparation of (Whittle et al., 1993) and the Medicare pro- this article was supported by the Agency for Health Care Policy and Research (AHCPR) under Contract Number 290-95-2000, gram (McBean and Gornick, 1994). Other Grant MH-57425, from the National Institute of Mental Health re s e a rch indicates that, although physi- and by the John D. and Catherine T. MacArthur Foundation Research Network on Socioeconomic Status and Health. The cians’ ability to detect the severity of pain opinions expressed are those of the authors and do not neces- does not differ for Hispanic versus non- sarily represent the views of the University of Michigan, the Department of Health and Human Services, or the Health Care Hispanic white patients (Todd, Lee, and Financing Administration (HCFA).

HEALTH CARE FINANCING REVIEW/Summer 2000/Volume 21, Number 4 75 H o ffman, 1994), Hispanic patients are rooted in an ideology of inferiority that cat- markedly less likely than non-Hispanic egorizes, ranks and differentially allocates white patients to receive adequate analge- societal re s o u rces to human population sia (Todd et al., 1993; Cleeland et al., 1997). groups (Bonilla-Silva, 1996). It may or may Recent studies document that these diffe r - not be accompanied by prejudice at the ences in the receipt of therapeutic proc e - individual level. We will illustrate the com- du r es have adverse effects on the health of plex nature of race, racism, and discrimi- minority group members (Peterson et al., nation in society by considering access to 1997; Hannan, van Ryn, and Burke, 1999). housing and employment. How do we make sense of these diffe re n c e s First, Table 1 indicates that there have and how do we move forwa r d with an effe c - been important positive changes in the tive policy and res e a r ch agenda to elimi- racial attitudes of white persons towards nate these disparities? black persons in recent decades and broad current support for the principle of equali- RACE, RACISM, AND ty in housing and employment (Schuman DISCRIMINATION et al., 1997). In 1963, 60 percent of white persons agreed with the statement that Many observers are surprised and per- “White people have a right to keep plexed by these findings. However, we can Negroes out of their neighborhoods if they only reg a r d these findings as surprising if want to, and Negroes should respect that we take an ahistorical and decontextualized right.” In 1996, only 13 percent of white view of the data. In compliance with Article 1, persons agreed with that statement, docu- Section 2, and Paragraph 3 of the menting a substantial positive attitudinal Constitution of the United States, the very shift within the white population. Similarly, first Census in 1790 enumerated three racial in 1944, a majority of white persons (55 per- gr oups: whites, blacks as three-fifths of a cent) indicated that white people should person, and only “civilized Indians”—those have the first chance at any kind of job but, who paid taxes (Anderson, 1988). New by 1972, only 3 percent of white persons racial categories were added in the late 19th endorsed that view with 97 percent indicat- C e n t u ry and beyond (Chinese in 1870, ing that black persons should have as good Japanese in 1890, Mexican in 1930) as the a chance as white persons to get any kind need arose to track new marginalized immi- of job. grant groups (Anderson and Feinberg , Second, these positive sentiments were 1995). Race was and is a social category that given the force of law. In 1968, the Fair ca p t u r es diffe r ential access to power and Housing Act (Title VIII) made it illegal to desirable res o u r ces in society (Wil l i a m s , refuse to sell or rent a dwelling to any per- 1997). Throughout the history of the son because of race. Earlier, the Civil United States, non-dominant racial grou p s Rights Act of 1964 (Title VII) prohibited have, either by law or custom, received infe- employers from firing, refusing to hire or rior treatment in major societal institutions. promote, or in any way limiting an employ- Medical care is no exception. ee’s compensation or job conditions Thus, understanding racial disparities in because of race. medical care requires an appreciation of Third, Table 1 also indicates that there is the ways in which racism has operated and considerably less support for policies that continues to operate in society. The term would actually implement equal access to “racism” refers to an organized system, housing and jobs (Schuman et al., 1997).

76 HEALTH CARE FINANCING REVIEW/Summer 2000/Volume 21, Number 4 Table 1 Racial Attitudes of White Persons in the United States: Selected Years, 1944-1996 Year of Survey Question 1944 1963 1964 1972 1973 1980 1983 1990 1996 Percent in Agreement Access to Housing Principle Item White persons have a right to keep black persons out of neighborhood. — 60 — 41 — 34 — 24 13

Implementation Item They would support a law to let homeowners discriminate if they want to. ———— 67 60 54 47 33

Access to Employment Principle Item White persons should have the first chance at any job. 55 15 — 3 — — — — —

Implementation Item Government should ensure no discrimination in jobs. —— 38 40 ———— 28

No interest in issue. —— 13 18 ———— 36 SOURCE: (Schuman, Steeh, Bobo, and Krysan, 1997.) In 1973, 67 percent of a national sample of black persons prefer to live off welfare, 51 white persons indicated that they would p e rcent believe that black persons are s u p p o rt a law that would guarantee a prone to violence, 29 percent view black homeowner the right to decide for himself persons as unintelligent, and 44 percent whom to sell his house to even if he pre- view them as lazy (Davis and Smith, 1990). ferred not to sell to black persons. In 1996, Comparatively, white persons believe that one-third of white persons would still grant only 4 percent of white persons prefer to a homeowner that right. In a similar vein, live off welfare, 16 percent are prone to vio- there is only weak support for policies to lence, 6 percent are unintelligent, and 5 eradicate employment discrimination. In percent are lazy. Instructively, white per- 1964, 38 percent of white persons indicated sons viewed black persons, Hispanics, and that the government in Washington should Asians more negatively than themselves, see to it that black people get fair treatment but black persons were viewed more nega- in jobs, and 13 percent indicated that they tively than all other groups, and Hispanics lacked enough interest in the question to twice as negatively as Asians. It is possible favor one side over another. In 1996, the that these re p o rted levels of negative percentage of white persons supporting stereotypes of minority groups are under- Federal intervention to ensure fair treat- stated due to social desirability concerns. ment in jobs declined to 28 percent, while Such high levels of the acceptance of the percentage expressing no interest in negative stereotypes of minority groups is the question increased to 36 percent. an ominous harbinger of widespread soci- F o u rth, national data on stere o t y p e s etal discrimination. A large body of psy- reveal that white persons continue to view chological re s e a rch indicates that the black persons negatively, which presum- endorsement of negative racial stereotypes ably would make them undesirable as leads to discrimination against minority neighbors and employees. For example, g roups (Devine, 1995; Hilton and von 56 percent of white persons believe that Hippel, 1996). More o v e r, well-learn e d

HEALTH CARE FINANCING REVIEW/Summer 2000/Volume 21, Number 4 77 stereotypes are resistant to disconfirma- gains of 71,100 for white persons, 55,100 tion (Stangnor and McMillan, 1992), and for Asians, and 60,000 for persons of Latin their activation is an automatic process extraction (Sharpe, 1993). These job loss- with individuals spontaneously becoming es reflected the relocation of employment aware of relevant stereotypes after encoun- facilities to areas of lower black concentra- tering someone to whom the stereotypes tion. Audit studies of employment dis- are applicable (Devine, 1989; Hilton and crimination also find racial differences in von Hippel, 1996). being allowed to submit an application, in In the United States, racial stereo t y p e s obtaining interviews, and in being offered have real-life consequences for minority a job. In these studies, when trained black gr oups’ access to housing and employment. and white job applicants with identical Based on negative stereotypes of black per- qualifications applied for jobs, discrimina- sons, the majority of white persons expres s tion favored the white over the black appli- a strong pref e r ence for living in racially cants in one in every five audits (Fix and se g r egated neighborhoods (Williams et al., Struyk, 1993). 1999; Bobo and Zubrinsky, 1996), and black The bottom line is that the policies persons in search of housing are still sys- designed to eliminate racial discrimination tematically steered toward neighborho o d s in housing and employment have failed. having a greater number of minorities, The degree of residential racial segrega- lower home values, and lower median tion in 1990 was virtually identical to what income (Fix and Struyk, 1993). A review of it was when Congress passed the Fair the data on the persistence of housing dis- Housing Act in 1968 (Massey, 1996). crimination in the United States concluded Similarly, the unemployment rate for black that, “On any given encounter between a persons has been consistently about twice black home-seeker and a rea l t o r , the odds that of white persons from 1950 to the pre- ar e at least 60 percent that something will sent (Economic Report of the President, happen to limit that black renter or buyer’s 1998). Thus, the advent of civil rights leg- access to housing units that are available to islation and changes in the racial attitudes white persons” (Massey, Gross, and of white persons have not been sufficient Shibuya, 1994). to eradicate discrimination. In spite of Studies of white employers reveal that these changes, there has been remarkable racial stereotypes are used to deny employ- stability over time on multiple dimensions ment opportunities to black applicants of racial inequality. For example, the medi- (Kirschenman and Neckerman, 1991; an income of black persons was 59 cents N e c k e rman and Kirschenman, 1991). for every dollar earned by white persons in Additionally, both U.S.-based and foreign 1996—identical to what it was in 1978 companies explicitly use the racial compo- (Economic Report of the President, 1998). sition of labor market areas in their deci- sionmaking process regarding where to Lessons for Racial Disparities in locate new plants (Cole and Deskins, Medical Care 1988). Not surprisingly, a Wall Stre e t J o u rn a l analysis of the employment The larger literature on societal discrim- records of more than 35,000 U.S. compa- ination suggests that, although racism has nies found that black persons had a net job changed over time from a blatant “Jim loss of 59,000 jobs during the 1990-1991 Crow racism” to a more subtle “laissez- economic downturn, compared with net faire racism” (Bobo, Kluegel and Smith,

78 HEALTH CARE FINANCING REVIEW/Summer 2000/Volume 21, Number 4 1997), it persists in contemporary America. scious, unthinking, and unintentional As painful as it may be to acknowledge, we (Allen, 1995; Johnson, 1988; Lawre n c e , must begin with the recognition that dis- 1987; Oppenheimer, 1993). As noted earli- crimination is routine and commonplace in er, biases based on racial stereotypes occur society and likely to be similarly prevalent automatically and without conscious in medicine. With few exceptions (Smith, awareness even by persons who do not 1998; Geiger. 1996; Council on Ethical and endorse racist beliefs (Devine, 1989). Judicial Affairs, 1990), the literature on Recent psychological re s e a rch indicates racial disparities in medical care is reluc- that persons who do not see themselves as tant to admit and address racial bias among prejudiced will make health care allocation providers as a critical causal factor. In con- decisions that adversely affect black per- trast, the evidence is abundant and clear sons when other negative characteristics that racial discrimination is not the aber- a re also present (Murphy-Berm a n , rant behavior of a few “bad apples” but a Berman, and Campbell, 1998). For exam- widespread societal problem. ple, respondents expressed greater resent- It is unlikely that personal discrimination ment, gave lower health care priority on the part of providers is the sole cause of scores and were more reluctant to make a disparities in health care. In any area of financial contribution to the health care societal evaluation, the causes of racial dif- costs of patients presented as black and ferences are complex and multi-dimension- unemployed than as white and unem- al, with discrimination being only one of ployed. In real-life medical encounters, the them. Moreover, institutional discrimina- automatic activation of stereotypes may tion is often at least as important as indi- provide the negative characterization that vidual discrimination. In the case of racial triggers unconscious bias. Most legisla- disparities in medical care, other potential tion, intervention programs, and policy explanations include the geographic mal- regarding discrimination have been inef- distribution of medical resources, racial fective because of their focus on purposeful differences in patient preferences, patho- or intentional discrimination (Allen, 1995). p h y s i o l o g y, economic status, insurance Relatedly, our review also suggests that coverage, as well as in trust, knowledge, one cannot rely on the stated racial atti- and familiarity with medical procedures tudes of white persons or the mere exis- ( H o rn e r, Oddone, and Matchar, 1995; tence of laws prohibiting discrimination to Smith, 1998). On the surface, patient pref- ensure that it does not occur in health care erences would be the alternative explana- settings. tion that would be most consistent with all of the available evidence. However, recent POLICY AND RESEARCH re s e a rch indicating that patient pre f e r- DIRECTIONS ences do not account for these disparities (Hannan, van Ryn, and Burke, 1999) sug- Improving Equity in Access to gests that discrimination remains as a cen- Medical Care tral plausible explanation. Probably most important, much discrim- Although this article centrally addresses ination today occurs through behaviors racial differences in the quality of care, a that the perpetrator does not subjectively c o m p rehensive approach to addre s s experience as intentional. Much contem- inequities must begin by ensuring parity in porary discriminatory behavior is uncon- access to care. Effectively addre s s i n g

HEALTH CARE FINANCING REVIEW/Summer 2000/Volume 21, Number 4 79 health care disparities will require compre- M c L a ff e rt y, 1982). Second, and more hensive efforts by multiple sectors of soci- important, the movement from a fee-for- ety to address larger inequities in major service (FFS) system to a managed care societal institutions. There is clearly a system is likely to adversely impact access need for concerted societal-wide efforts to to medical care for minorities and other confront and eliminate discrimination in vulnerable populations (Schlesinger, 1987; education, employment, housing, criminal Randall, 1994; McClellan, 1999; Harvar d justice, and other areas of society which Law Review, 1995). will improve the socioeconomic status Critics argue that managed care is likely (SES) of disadvantaged minority popula- to exacerbate current racial inequalities in tions and indirectly provide them greater access to medical care in multiple ways. access to medical care. The United States The new competitive pre s s u res in the also needs to make the moral and political financing and delivery of medical care can commitment to guarantee access to med- reduce profits generated by treating pri- ical care as a fundamental right of citizen- vately insured patients, leading to cutbacks ship. in the provision of uncompensated care H o w e v e r, eff o rts to ensure equitable (Schlesinger, 1987). In addition, managed access to care must go beyond the elimina- care plans often limit access to minority tion of financial barriers. A broad range of physicians and other doctors who primari- system barriers such as long waiting time, ly care for disadvantaged populations complex bureaucratic procedures, and the (Harvard Law Review, 1995). Managed failure to treat patients with dignity and care plans look for cost effective physi- respect can lead to patient alienation and cians who order few procedures, write lim- the avoidance of contact with the health ited prescriptions, and make limited refer- care system unless absolutely necessary. rals. In contrast, given the high morbidity, Research has long indicated that poor per- comorbidity and greater severity of disease sons and racial minorities are not viewed at time of diagnosis in minority and low as desirable patients and health care income populations, these patients require providers deliver inferior care to persons of intense management of chronic illness low SES (Duff and Hollingshead, 1968; van which can involve more medical services Ryn and Burke, 2000). Problems of and thus higher costs. Accordingly, man- patient-provider communication are exac- aged care plans are likely to view physi- erbated among persons of low SES, with cians who work in minority communities higher SES patients receiving better tech- as undesirable and may also limit the nical and interpersonal care, and more pos- physicians that their enrollees, including itive communication than their lower SES poor patients and residents of poor com- peers (Hall, Roter, and Katz, 1988). munities, can consult. The net result could Efforts to ensure equitable access to be that many physicians who work in care are urgently needed to counteract two minority communities may have a smaller f o rces that could potentially exacerbate patient load and some may be even unable racial disparities in access to care. First, to maintain a financially viable practice there are closures of a growing number of (Harvard Law Review, 1995). health care facilities with hospitals located However, it is not clear if these fears in low income and minority communities have materialized. A 1994 survey of black being more likely to close than those locat- physicians at a national conference found ed in other areas (Whiteis, 1992; that 92 percent believed that managed care

80 HEALTH CARE FINANCING REVIEW/Summer 2000/Volume 21, Number 4 plans terminated the contracts of black g roups: Mexicans, mainland Puert o doctors more often than those of white Ricans, and Cubans. Given the central role doctors (Lavizzo-Mourey et al., 1996). In of SES in health and the strong relation- fact, 88 percent had been refused a con- ship between race/ethnicity, and SES, it is tract by a managed care org a n i z a t i o n important that indicators of SES are also (MCO) and 71 percent had lost patients to included in any uniform data set (Krieger, a MCO with which they were not affiliated. Williams, and Moss, 1997; Williams, 1996). At the same time, 71 percent had at least The pervasiveness of discrimination one managed care contract and 75 percent suggests that racial data should be avail- indicated that their practice had grown or able for every medical encounter. The remained stable in the previous year. National Committee of Vital and Health Some observers also note that it is not Statistics, a public advisory body to the inevitable that access to medical care will D e p a rtment of Health and Human worsen for minority populations under Services has called for the assessment of managed care arrangements. Incre a s e d race/ethnicity and SES (National Commit- competition could provide a financial tee of Vital and Health Statistics, 1993) in incentive for some health care providers to an enrollment database which could then provide treatment to segments of the com- be linked to data on medical encounters. munity that they may have earlier viewed Years of formal education was suggested as undesirable (Schlesinger, 1987). as the most practical and convenient SES indicator in this context. The presence of Improved Data Systems and these identifiers in an enrollment database Monitoring the Quality of Medical would eliminate the re s o u rce intensive Care ordeal of attempting to request racial data at every medical encounter. Stringent Any concerted effort to address racial efforts to ensure patient privacy and confi- bias in the medical arena requires system- dentiality would have to be implemented atic and routine data of its occurrence. As and the necessary training would have to noted, important changes are taking place be provided to health care workers to in the organization and delivery of health ensure the uniform collection of racial/eth- care services in the United States and it is nic and SES data. The analyses of racial critical to monitor the impact of these disparities in the HCFA data files illustrate changes on the health care access and that administrative databases can be very quality for vulnerable populations. Major helpful in providing findings that shed light efforts are currently under way to identify on the nature and magnitude of the prob- the data elements that should be included lem. However, these analyses have focused in national uniform standardized data sets. only on black-white differences because It is essential that racial and ethnic “black, white and other” were the only status are uniformly and comprehensively racial identifiers present. In the early assessed in these minimum core data sets. 1990s, HCFA and the Social Security C o m p rehensive assessment means that Administration (SSA) went to considerable data systems should include identifiers for effort to improve racial/ethnic identifiers major ethnic subgroups within the stan- in the SSA and Medicare data systems. d a rd racial/ethnic categories (Wi l l i a m s , What is needed now is the routine analysis 1996). For example, among Hispanics, it is and reporting of data for all racial/ethnic necessary to distinguish the three largest groups.

HEALTH CARE FINANCING REVIEW/Summer 2000/Volume 21, Number 4 81 Routine rep o r ting of data would identify because it usually requires the compilation which health care institutions, if any, and analysis of large quantities of data to demonstrate racial parity in terms of the prove a discriminatory effect. However, de l i v e r y of medical care. It is likely that routine collection of racial data would facil- th e r e is considerable variation in racial dis- itate the assessment of racial group disad- parities across multiple settings. This kind vantage in the medical arena. of data can be used to establish benchmarks Smith (1998) notes that the emphases on —levels of excellence achieved by industry monitoring both clinical and financial infor- leaders that could enhance our understand- mation in the managed care environment ing of these best practices and facilitate also provides new opportunities for their replication (Weissman et al., 1999). enhanced civil rights monitoring of health Research is also needed to identify the care delivery. He indicates that the addi- optimal specific strategies that health care tion of racial identifiers to existing data sys- institutions can implement, at least on a tems would facilitate the creation of report periodic basis if not continuously, to detect cards that could be used to monitor dispar- and respond to patterns of discrimination ities in health plans, health care institu- in medical treatment. It has been suggest- tions, and communities. These re p o rt ed that hospitals could develop an anony- cards would include broadly accepted indi- mous reporting system to facilitate the cators of health and health care delivery detection of incidents of systematic pat- (such as mortality rates and the use of spe- terns of biased medical decisionmaking cific screening tests) that have been rec- (Noah, 1998). However, it will be crucial to ommended by various standards organiza- c reate an environment that encourages tions for the comparative evaluation of reporting as part of a strategy of construc- medical care. Smith (1998) indicates that tive problem-solving (Leape, 1997). In gen- similar re p o rting re q u i rements in the eral, the five strategies for detecting inap- banking industry have led to a dramatic propriate medical treatment, in order of increase in the number of loans approved yield and intensity of eff o rt are dire c t for racial minorities. However, in order for o b s e rvation, chart re v i e w, computer report cards to work, they must be a part s c reening, focus groups and voluntary of a larger system. reporting (Leape, 1997). The availability of data on racial differ- Renewed Regulatory Vigilance ences in medical care would also facilitate at least some civil rights enforc e m e n t Another policy strategy to address the efforts. One of the limitations of current p roblem of discrimination in medicine antidiscrimination laws is that they often would be increased regulatory vigilance. rely on proving subjective discriminatory The history of overt discrimination in med- intent (Allen, 1995). Given that the vast ical care indicates that legal mandates and majority of white Americans favor non-dis- Federal regulations were ineffective until crimination in principle, it is often difficult the institutional commitment and capacity to prove discriminatory intent. In the legal to enforce them was created (Smith, 1998). arena, disparate impact claims require sta- Some legal scholars argue that there are tistical data to document the differential existing statutes that are not now being effect of policies on racial groups. It is not enforced. For example, Title VI of the Civil a frequently used strategy by individuals Rights Act of 1964 is a promising statutory seeking relief under civil rights statutes avenue for dealing with discrimination in

82 HEALTH CARE FINANCING REVIEW/Summer 2000/Volume 21, Number 4 health care delivery (Noah, 1998). Title VI The Joint Commission on Accreditation prohibits any entity that receives Federal of Health Care Organizations (JCAHO) is financial assistance from discriminating on another regulatory organization that could the basis of race in providing goods or ser- play a larger role in addressing the ques- vices to the beneficiaries of that Federal tion of racial inequities in health care. It program. Since Federal financial assis- could, for example, insist that health care tance includes Medicare and Medicaid entities establish procedures for the moni- funds, this prohibition against discrimina- toring and correction of unexplained dis- tion applies to virtually all hospitals, nurs- parities in the delivery of services as one of ing homes, and other health care facilities the requirements for continued accredita- in the United States. Given that the courts tion (Noah, 1998). Historically, the JCAHO have held that Title VI prohibits both inten- has played an important role in encourag- tional discrimination and disproportionate ing progressive change in the health care adverse impact (Noah, 1998), the docu- industry. A recent example is the efforts to mentation of adverse impact would provide encourage hospitals to develop and imple- a powerful strategy for addressing and cor- ment policies to address communication recting discrimination in care. problems with patients who do not speak Noah (1998) also notes that many neu- English. HCFA might also review its cur- tral policies that may have a disproportion- rent utilization review methods to see if ate impact on racial and ethnic minorities better methods can be developed to identi- receiving medical treatment could also be fy and correct observed patterns of racial addressed if disparate impact were docu- inequalities and delivery of services to ben- mented. For example, some hospitals eficiaries of Medicare and Medicaid (Noah admit only those patients whose physicians 1998). have staff privileges at the hospital, require substantial deposits before admission for Monitoring Managed Care inpatient care, refuse to deliver babies if their mothers had not received a certain Increased regulatory efforts must also amount of prenatal care, and create other specifically focus on managed care plans b a rriers to the admission of Medicaid given that managed care creates an envi- patients. Minority patients are more likely ronment conducive to discrimination and to be disadvantaged by these policies, some have argued that there will be more although the policies are not specifically widespread discrimination under managed racial in content or intent. Thus, the appli- care (Harvard Law Review, 1995). Unlike cation of a disparate impact analysis of the FFS system where a physician’s these policies could be an effective avenue income increases as the number of ser- for attacking them. At the same time, vices provided increases, in managed care Noah (1998) warns that, because Title VI there are often pressures to limit the num- covers institutional policies that are caus- ber and cost of services delivered to mem- ing disproportionate impact, they would bers of the prepaid health plan. Many man- not cover the behavior of individual physi- aged health care plans offer financial cians who either consciously or unthink- incentives to providers for limiting refer- ingly discriminate as long as the health rals to specialists. Often a pool of money is care entity can prove that it is not its insti- re s e rved for specialists and diagnostic tutional policy. tests, with the provider sharing in the

HEALTH CARE FINANCING REVIEW/Summer 2000/Volume 21, Number 4 83 unexpended portion of this pool toring of fairly rapid changes in health care (McClellan, 1999). Some managed care delivery. There is also the need to develop systems hold individual physicians finan- new legislation or regulations that can cially liable when their patients use a high- ensure the appropriate inclusion of physi- er than average amount of hospital care cians who practice in minority communi- i rrespective of whether the costs were ties in managed care plans. For example, medically justified (Hillman, 1987). These managed care plans that re c ruit fro m financial incentives have been shown to minority communities could be required to shape decisionmaking by providers (Pauly, h i re a certain pro p o rtion of medical Hillman, and Kerstein, 1990). providers who previously practiced there. Currently available research does not S i m i l a r l y, laws could be established to clearly indicate whether medical care over- ensure that the categorical exclusion of all is better under the FFS system or under providers in poor communities does not managed care plans. Some evidence sug- occur. For example, they could prohibit the gests that enrollees in managed care plans exclusion of providers from managed care are more likely to receive cancer screening plans based on the health, racial composi- tests than persons in FFS plans (Potosky et tion, or SES of their patients. Finally, sanc- al., 1998). Recent national data from 35,000 tions can be utilized to prevent MCOs, and non-elderly persons revealed that there health care entities more generally, from were no differences between health main- engaging in discrimination. Systematic tenance organizations (HMOs) and other evidence of ongoing discrimination could types of insurance in the use of inpatient be met with substantial fines or even the c a re, emergency room and surg e r i e s threat of the loss of the right to practice. (Reschovsky et al., 2000). At the same Wallace, et al. (1998) emphasize that time, patients in HMOs reported less satis- managed care has the potential to improv e faction, less trust in physicians, lower rat- the quality of medical care for minority pop- ings of physician visits and more unmet ulations. Given the greater burea u c r a t i c medical need. This study found that, while oversight present in MCOs, there is the HMOs provided more primary and preven- potential for greater coordination of care tive services, they provided less specialist that could ensure that individuals rec e i v e d care and had higher administrative barri- ap p r opriate medical care. That is, if the ers to care. Other evidence suggests that ap p r opriate incentives were built into man- health care for vulnerable populations, aged care plans they could help reverse the such as the elderly and the chronically ill, tr ends of minorities receiving less aggres - is worse in managed care plans sive treatment and fewer medical proc e - (McClellan, 1999; Wallace, Enriquez- du r es across a broad range of illnesses. Haass, and Markides, 1998). S i m i l a r l y, organizations such as the Several solutions have been proposed to National Committee for Quality Assurance the potential threat to minority health (NCQA) that currently provides voluntary posed by managed care org a n i z a t i o n s a c c reditation to almost one-half of the (Harvard Law Review, 1995). First, there HMOs in the United States could add racial is need for more systematic auditing of the inequities in care to the more than 60 stan- services of managed care providers by the dards that currently provide the basis for Government, consumer groups, and pro- the evaluation of health plans. NCQA has fessional medical organizations. Currently, also managed the development of the there is little Federal oversight and moni- Health plans Employer Data and Information

84 HEALTH CARE FINANCING REVIEW/Summer 2000/Volume 21, Number 4 Set (HEDIS) which is the principal perfor- ination. Cultural sensitivity programs may mance assessment for managed care. It enhance and accentuate negative stereo- uses a set of standardized measures to types. Some cultural sensitivity training compare health plans. Currently HEDIS focuses on the distinctive behavioral pat- evaluates what a health plan actually does terns of subgroups in the population and in key areas of care such as immunization appears to focus primarily on the “strange” rates and member satisfaction. The extent behavior of patients. More research is also of racial/ethnic disparities should be needed on provider attitudes and behavior added as a performance indicator. (King, 1996) and especially the identifica- tion of the strategies that may be most Education and Training e ffective in identifying and re d u c i n g unconscious discrimination among med- Th e r e is also a need for intensive and sys- ical professionals. tematic educational campaigns about the Another critical goal of medical educa- pr oblem of racial inequities in health care. tion should be to increase the number of The awareness levels of the public and pro- minority professionals. Research clearly fessional community, especially the med- indicates that black and Hispanic physi- ical community, must be raised. Research cians are much more likely than other is needed to identify strategies that are physicians to care for the uninsured and most effective to raise awareness of and those with Medicaid and to practice in in c r ease sensitivity to the issues of race in urban and rural underserved areas where medical practice. Although education has the percentage of residents of their its limits, it is also instructive to know that racial/ethnic group is high (Komaromy et educational campaigns can accomplish al., 1996). A recent study reported that in much. For example, in the case of tobacco order to reach racial and ethnic population th e r e has been a per capita decline in tobac- parity, the United States needs to double co consumption in the United States over the number of black and Hispanic first- the course of the last century whenever year residents and triple the number of th e r e was a major media campaign on the Native American residents (Libby, Zhou, negative effects of cigarette smoking and Kindig, 1997). White first-year resi- (Wa rn e r , 1985). Effo r ts are clearly needed dents would need to be reduced by two- to impact the medical school curriculum to fifths and Asians by two-thirds. en s u r e that issues of race and sensitivity Cu r rent trends do not suggest that these t o w a rds these issues are adequately goals are likely to be reached. There has a d d ressed. As Geiger (1996) indicates, been relatively little increase in the prop o r - “ a w a reness of the dilemmas associated tion of physicians from underrep re s e n t e d with race and health care should be a part minority backgrounds in medicine in the of every physician’s training.” last 30 years. For example, black physicians It is not clear that the current emphases have increased from 2.5 percent of all U.S. and approaches to cultural sensitivity will physicians prior to 1968 (Carlisle, Gardn e r , address the systemic problem of discrimi- and Liu, 1998) to 2.9 percent curren t l y nation. Even model cultural sensitivity pro- (Editorial, 1999). The number of underrep - grams for medical students and residents resented minorities (black Americans, (Robins et al., 1998; Zweifler and Gonzalez, Mexican Americans, mainland Puert o 1998) do not address unconscious discrim- Ricans, and American Indians) increased in

HEALTH CARE FINANCING REVIEW/Summer 2000/Volume 21, Number 4 85 medical schools starting in the late 1960s. in the recruitment of competent physicians Minority enrollment peaked in 1974 at 10 (Tekian, 1997). In addition, despite their pe r cent of total enrollment but subsequent- c u rrent unpopularity, aff i rmative action ly declined in the wake of reverse discrimi- p rograms are defensible on multiple nation lawsuits (Carlisle, Gardn e r , and Liu, grounds including the societal obligation 1998; Nickens and Cohen, 1996). To to meet the health care needs of all seg- reverse this trend the Association of ments of the population (Nickens and American Medical Colleges launched a Cohen, 1996). Moreover, recent research new campaign to have 3000 first-year documents that racial prejudice, especially minority medical students by the year 2000. contemporary subtle prejudice, is the sin- This project was initially successful with gle most important source of opposition to en r ollment reaching a high of 2014 under- affirmative action (Williams et al., 1999). rep r esented minorities (12 percent of all These fingings emphasize the critical need new entrants) in 1994 (Carlisle, Gardn e r , to confront and eradicate societal racism in and Liu, 1998). In 1996, there was a large o rder to develop effective strategies to dr op in applications from underrep re s e n t - o v e rcome America’s painful history of ed minorities to medical schools (Editorial, exclusion and discrimination. 1999) with more than one-half of all U.S. However, McClellan (1999) notes that medical schools experiencing a decline in simply matching physician race with client minority enrollment (Carlisle, Gardn e r , and race is not enough. He cites an uncom- Liu, 1998). The decline was largest in pub- fortable but insightful example from the lic medical schools and in the four States Tuskegee Syphilis Study. In 1969, the ( C a l i f o rnia, Texas, Louisiana, and Macon County Medical Society endorsed Mississippi) where affi r mative action pro- the continuation of the Tuskegee Syphilis grams have been banned. Study. At that time, the Society was virtu- M o re re s e a rch is needed to identify ally all-black. McClellan (1999) argues that what are the most effective strategies for social class provides an explanation of the recruitment and retention of physi- those black physicians’ behavior. That is, cians from disadvantaged backgro u n d s . because most of the patients in the However, we should also capitalize on the Tuskegee study were poor and illiterate, currently available evidence on effective the study was not a threat to middle class strategies. Aff i rmative action pro g r a m s physicians, their families, or friends. He (Federal initiatives that allowed schools argues that educational and professional and employers to take into consideration a socialization may lead health care profes- qualified applicant’s race, sex, national ori- sionals to distance themselves in terms of gin, or disability) have been successful. It emotional attachment and self-intere s t is estimated that aff i rmative action is from their groups of origin. Unthinking responsible for 40 percent of all U.S.- discrimination is likely to occur whenever trained physicians from underrepresented medical professionals, ir respective of race, minority backgrounds (Editorial, 1999). endorse negative societal stereotypes of Although underrepresented minority stu- their patients. Research is needed to iden- dents tend to have lower test scores than tify the extent to which these processes other medical students, they do not differ operate and what may be the most effec- on clinical performance suggesting that tive strategies to counter them. other non-cognitive variables are essential Rather than focusing only on individual

86 HEALTH CARE FINANCING REVIEW/Summer 2000/Volume 21, Number 4 characteristics, efforts to reduce racial/ Commission on Race. Relatedly, in 1994, ethnic disparities in the quality of medical P resident Clinton issued an executive care should seek to reform the systems for order on environmental justice that could the delivery of care. Is it possible to design be a model of the kind of effort that is and implement systems for the delivery of needed (Noah, 1998). This executive medical care that can ensure appropriate o rder created an interagency working behavior irrespective of effect and uncon- group to address questions of environmen- scious stereotypes? A definitive answer to tal justice and to provide guidelines to this question awaits the necessary Federal agencies for coord i n a t i n g re s e a rch, but Leape (1997) emphasizes research, collecting data, and developing that in medicine, as in other occupations, effective strategies to address inequities. conditions of work can be managed, tasks A similar working group on racial dispari- and processes can be designed, and work- ties in medical care could monitor and ers can be trained in ways that minimize coordinate data from multiple sources and undesirable outcomes. provide a coordinated picture of the nature and prevalence of inequities in multiple CONCLUSION health care settings. A commitment to the eradication of racial disparities also It is a national embarrassment that there requires the creation of the institutional are large and persisting racial differences capacity to effectively monitor and enforce in health. National data reveal, for exam- all existing laws and regulations. ple, that black persons had an overall mor- Our review indicates that racism appears tality rate that was 1.6 times higher than to be a technological hazard in the practice white persons in 1995—identical to the of medicine. Much of it may be unthinking black/white mortality ratio in 1950 and careless and not deliberately hateful (Williams, 1999). Moreover, for multiple behavior. However, regardless of motive, causes of death (heart disease, cancer, dia- there are pervasive adverse consequences. betes, and cirrhosis of the liver) the racial As a society, we lack data on effective discrepancy was larger in 1995 than in strategies to reduce racism at both the indi- 1950. These inequities fly in the face of vidual and institutional levels. While there cherished American principles given the are many books published on the topic and public’s commitment to principles of equal many programs on cultural diversity and treatment in society. As a society, we need tolerance, there is little systematic data to make it a national priority to build on the available about the conditions under which cultural support for egalitarian principles particular strategies are more or less effec- and develop strategies to eradicate racial tive. Given the growing body of evidence inequities in medical care. that indicates that racism adversely affects The United States President and other health in multiple other ways (Krieger, leaders in the executive branch of 1999; Williams, 1999), more systematic Government should use the “bully pulpit” efforts to develop and assess the impact of to place this issue on the national agenda. various strategies to reduce racism is war- President Clinton has drawn much nation- ranted. Well-funded research centers for al and media attention to the problem of excellence should be established to foster health disparities by making it a topic of interdisciplinary research on understand- one of his Saturday morning radio address- ing and eliminating racial/ethnic dispari- es and by the establishment of his ties in medical care. Courageous moral and

HEALTH CARE FINANCING REVIEW/Summer 2000/Volume 21, Number 4 87 political leadership is also needed to take Site Location and Employment Patterns of Japanese the necessary steps to apply all of the Auto Firms in America. California Management Review 31(1):9-22, 1988. knowledge that we currently have for Council on Ethical and Judicial Affairs: Black-White reducing racial/ethnic disparities in Disparities in Health Care. Journal of the American health, and more generally. Medical Association263(17):2344-2346, 1990. Davis, J.A., and Smith, T.W.: General Social Surveys, REFERENCES 1972-1990. Chicago. National Opinion Research Center, 1990. Allen, J.: A Remedy for Unthinking Discrimination. Devine. P.G.: Prejudice and Out-Group Perception. Brooklyn Law Review61:1299-1345, Winter, 1995. In A. Tesser. Advanced Social Psychology. New York. Anderson, M.J.: The American Census: A Social McGraw-Hill, 1995. History. New Haven, CT. Yale University Press, Devine, P.G.: Stereotypes and Prejudice: Their 1988. Automatic and Controlled Components. Journal of Anderson, M., and Feinberg, S.E.: Black, White, Personality and Social Psychology56:5-18, 1989. and Shades of Gray (and Brown and Ye l l o w ) . Duff, R., and Hollingshead, A.: Sickness and Society. Chance8(1):15-18, 1995. New York. Harper and Row, 1968. Blendon, R., Aiken, L., Freeman, H., and Corey, C.: Economic Report of the President. Washington, DC. Access to Medical Care for Black and White U. S. Government Printing Office; 1998. Americans. J o u rnal of the American Medical Editorial. Affirmative Action. Lancet 353(9146):1, Association261(2):278-281, 1989. 1999. Bobo, L., Kluegel, J.R., and Smith, R.A.: Laissez Fix, M., and Struyk, R.J.: Clear and Convincing Faire Racism: The Crystallization of a “Kinder, Evidence: Measurement of Discrimination in Gentler” Anti-Black Ideology. In Tuch, S.A., and America. Washington, DC. Urban Institute Press, Martin, J.K., eds.: Racial Attitudes in the 1990s: 1993. Continuity and Change. Westport, CT. Praeger; 1997. Geiger, H.J.: Race and Health Care - An American Dilemma? New England Journal of Medicine Bobo, L., and Zubrinsky, C.L.: Attitudes on 335(11):815-816, 1996. Residential Integration: Perceived Status Differences, Mere In-Group Preference, or Racial Hall, J.A., Roter, D.L., and Katz, N.R.: Meta-Analysis Prejudice? Social Forces74(3):883-909, 1996. of Correlates of Provider Behavior in Medical Encounters. Medical Care 26(7):212-229, 1988. Bonilla-Silva, E.: Rethinking Racism: To w a rd a S t ructural Interpretation. American Sociological Hannan, E.L., van Ryn, M., Burke, J., et al.: Access Review 62(3):465-480, 1996. to Coro n a ry Art e ry Bypass Surg e ry by Race/Ethnicity and Gender Among Patients Who Bunker, J. P., Frazier, H.F., and Mosteller, F.: The Are Appropriate for Surgery. Medical Care 37(1):68- Role of Medical Care in Determining Health: 77, 1999. Creating an Inventory of Benefits. In Amick, B.C.I., Levine, S., Tarlov, A.R., and Walsh, D.C., eds.: Harris, D.R., Andrews, R., and Elixhauser, A.: Society and Health. New York. Oxford University Racial and Gender Diff e rences in Use of Press, 1995. P ro c e d u res for Black and White Hospitalized Adults. Ethnicity and Disease7(2):91-105, 1997. Carlisle, D.M., Gardner, J.E., and Liu, H.: The Entry of Underrepresented Minority Students Into Harvard Law Review: The Impact of Managed Care U.S. Medical Schools: An Evaluation of Recent on Doctors Who Serve Poor and Minority Patients. Trends. American Journal of Public Health 108:1625-1642, May 1995. 88(9):1314-1318, 1998. Hillman, A.: Financial Incentives for Physicians in Cleeland, C.S., Gonin, R., Baez, L., et al.: Pain and HMOs. Is There a Conflict of Interest? New England Treatment of Pain in Minority Patients with Cancer: Jo u r nal of Medicine317(27):1743-1748, 1987. The Eastern Cooperative Oncology Group Minority Hilton, J.L., and von Hippel, W.: Stere o t y p e s . Outpatient Pain Study. Annals of Internal Medicine Annual Review of Psychology47(250):237-271, 1996. 127(9):813-816, 1997. Cole, R.E., and Deskins, D.R., Jr.: Racial Factors in

88 HEALTH CARE FINANCING REVIEW/Summer 2000/Volume 21, Number 4 Horner, R.D., Oddone, E.Z., and Matchar, D.B.: McBean, A.M., and Gornick, M.: Differences by Theories Explaining Racial Diff e rences in the Race in the Rates of Procedures Performed in Utilization of Diagnostic and Therapeutic Hospitals for Medicare Beneficiaries. Health Care Procedures for Cerebrovascular Disease. Milbank Financing Review 15(4):77-90, 1994. Quarterly 73(3):443-462, 1995. McClellan, F.M.: Is Managed Care Good for What Johnson, S.L.: Unconscious Racism and the Ails You? Ruminations on Race, Age and Class. Criminal Law. Cornell Law Review 73:1016-1037, Villanova Law Review 44:227-255, 1999. July 1988. McLafferty, S.: Neighborhood Characteristics and King. G.: Institutional Racism and the Hospital Closures: A Comparison of the Public, Medical/Health Complex: A Conceptual Analysis. Private, and Voluntary Hospital Systems. Social Ethnicity and Disease6(1,2):30-46, 1996. Science Medicine16(19):1667-1674, 1982. Kirschenman, J., and Neckerman, K.M.: “We ’ d M u r p h y - B e rman, V.A., Berman, J.J., and Love to Hire Them, But...”: the Meaning of Race for Campbell, E.: Factors Affecting Health-Care Employers. In Jencks, C., and Peterson, P.E., eds. Allocation Decisions: A Case of Aversive Racism? The Urban Underc l a s s. Washington, DC.: The Journal of Applied Social Psychology28(24):2239- Brookings Institution. Pp 203-232, 1991. 2253, 1998. Komaromy, M., Grumbach, K., Drake, M., et al.: National Center for Health Statistics: Health, United The Role Of Black And Hispanic Physicians In States and, Socioeconomic Status and Health P roviding Health Care For Underserv e d Chartbook. Hyattsville, MD. U.S. Department of Populations. Black and Hispanic Physicians and Health and Human Services. Washington, DC. Underserved Populations334(20):1305-1310, 1996. 1998. Krieger, N.: Embodying Inequality: A Review of National Committee on Vital and Health Statistics: Concepts, Measures, and Methods for Studying 1. Diversity and Health Care Data Base: A Health Consequences of Discrimination. P reamble for Policymakers. DHHS Publication International Journal of Health Services 29(2):295- PHS 93-1205:3-19, 1993. 352, 1999. Neckerman, K.M., and Kirschenman, J.: Hiring Krieger, N., Williams, D.R., Moss, N.: Measuring Strategies, Racial Bias, and Inner-City Workers. Social Class in U.S. Public Health Researc h : Social Problems38(4):433-447, 1991. Concepts, Methodologies, and Guidelines. Annual Nickens, H.W., and Cohen, J.J.: On Affirmative Review of Public Health18:341-378, 1997. Action [Policy Perspectives]. J o u rnal of the Lavizzo-Mourey, R., Clayton, L.A., Byrd, M., et al.: American Medical Association275(7):572-574, 1996. The Perceptions of African-American Physicians Noah, B.A.: Racial Disparities in the Delivery of Concerning Their Treatment By Managed Care Health Care. San Diego Law Review 35:135-178, O rganizations. Jo u r nal of the National Medical Winter 1998. Association88(4):210-214, 1996. O p p e n h e i m e r, D.B.: Negligent Discrimination. Lawrence, C.R.I.: The Id, the Ego, and Equal University of Pennsylvania Law Review141:899-972, Protection: Reckoning with Unconscious Racism. January 1993. Stanford Law Review 39:317-38, January 1987. P a u l y, M.V., Hillman, A.L., and Kerstein, J.: Leape, L.L.: A Systems Analysis Approach to Managing Physician Incentives in Managed Care: Medical Error. Journal of Evaluation in Clinical The Role of For-Profit Ownership. Medical Care Practice 3(3)213-222, 1997. 28(11):1013-1024, 1990. Libby, D.L., Zhou, Z., and Kindig, D.A.: Will Peterson, E.D., Shaw, L.K., DeLong, E.R., et al: Minority Physician Supply Meet U. S. Needs? Racial Variation in the Use of Coro n a ry - Health Affairs (Millwood) 16(4):205-214, 1997. Revascularization Procedures--Are the Differences Massey, D.A., Gross, A.B., and Shibuya, K.: Real? Do They Matter? New England Journal of Migration, Segregation, and the Geographic Medicine336(7):480-486, 1997. Concentration of Poverty. American Sociological P o t o s k y, A.L., Breen, N. Graubard, B. I., and Review 59(3):425-445, 1994. Parsons, P.E.: The Association Between Health Massey, D.S.: The Age of Extremes: Concentrated Care Coverage and the Use of Cancer Screening Affluence and Poverty in the Twenty-First Century. Tests. Results from the 1992 National Health Demography33(4):395-428, 1996. Interview Survey. Medical Care 36(3):257-70, 1998.

HEALTH CARE FINANCING REVIEW/Summer 2000/Volume 21, Number 4 89 Randall, V.R.: Impact of Managed Care Orga n i z a t i o n s van Ryn, M., and Burke, J.: The Effect of Patient on Ethnic Americans and Under Serv e d Race and Socio-Economic Status on Physicians’ Populations. Journal of Health Care for the Poor Perceptions of Patients. Social Science & Medicine Underserved5(3):224-226, 1994. 50(6):813-828, 2000. Reschovsky, J.D., Kemper, P., Tu, H.T., et al.: Do Wallace, S.P., Enriquez-Haass, V., and Markides, K.: HMOs Make a Difference?: Comparing Access, The Consequences of Color-Blind Health Policy for Service Use and Satisfaction Between Consumers Older Racial and Ethnic Minorities. Stanford Law in HMOs and Non-HMOs. Issue Brief: Findings and Policy Review9(2):329-346, 1998. From Health System Change28:1-8, March 2000. Wa rn e r, K.: Cigarette Advertising and Media Robins, L.S., Fantone, J.C., Hermann, J., et al.: Coverage of Smoking and Health. New England C u l t u re, Communication, and the Inform a l Journal of Medicine312(6):384-388, 1985. C u rriculum: Improving Cultural Aw a reness and Weissman, N.W., Allison, J.J., Kiefe, C.I., et al.: Sensitivity Training in Medical School. Academic Achievable Benchmarks of Care: the ABC’s of Medicine73(10):S31-S34, 1998. Benchmarking. Journal of Evaluation in Clinical Schlesinger, M.: Paying the Price: Medical Care, Practice 5(3):269-281, 1999. Minorities, and the Newly Competitive Health Care We n n e k e r, M.B., and Epstein, A.M.: Racial System. Milbank Quarterly65(2):270-296, 1987. Inequalities in the Use of Procedures for Patients Schuman, H., Steeh, C., Bobo, L., and Krysan, M.: with Ischemic Heart Disease in Massachusetts. Racial Attitudes in America: Trends and J o u rnal of the American Medical Association I n t e r p retations, Rev. Edition. Cambridge, 261:253-257, 1989. MA.Harvard University Press, 1997. Whiteis, D.G.: Hospital and Community Character- Sharpe, R.: In Latest Recession, Only Blacks istics in Closures of Urban Hospitals, 1980-87. S u ff e red Net Employment Loss. Wall Stre e t Public Health Reports107(4):409-416, 1992. Journal. p. A-1, September 14, 1993. Whittle, J., Conigliaro, J., Good, C.B., and Lofgren, Smith, D.B.: Addressing Racial Inequities in Health R . P.: Racial Diff e rences in the Use of Invasive Care: Civil Rights Monitoring and Report Cards. Cardiovasular Procedures in the Department of Health Politics, Policy, and Law 23(1):75-105, 1998. Veterans Affairs. New England Journal of Medicine S t a n g n o r, C., and McMillan, D.: Memory for 329(9):621-626, 1993. E x p e c t a n c y - C o n g ruent and Expectancy- Williams, D.R.: Race and Health: Basic Questions, Incongruent Information: A Review of the Social E m e rging Directions. Annals of Epidemiology and Social Development Literatures. Psychological 7(5):322-333, 1997. Bulletin 111(1):42-61, 1992. Williams, D.R.: Race/Ethnicity and Socioeconomic Tekian, A.: A Thematic Review of the Literature on Status: Measurement and Methodological Issues. Under Represented Minorities and Medical International Journal of Health Services 26(3):483- Training, 1981-1995: Securing the Foundations of 505, 1996. the Bridge to Diversity. Academic Medicine Williams, D.R.: Race, SES, and Health: The Added 72(10):S140-S146, 1997. Effects of Racism and Discrimination. Annals of Todd, K.H,, Lee, T., Hoffman, J.R.: The Effect of New York Academy of Sciences896:173-188, 1999. Ethnicity on Physician Estimates of Pain Severity in Williams, D.R., Jackson, J.S., Brown, T.N., et al.: Patients With Isolated Extremity Trauma. Journal Traditional and Contemporary Prejudice and Urban of the American Medical Association271(12):925- Whites’ Support for Af f i rmative Action and 928, 1994. Government Help. Social Problems 46(4):503-527, Todd, K.H., Samaroo, N., and Hoffman, J.R.: 1999. Ethnicity as a Risk Factor for Inadequate Z w e i f l e r, J., and Gonzalez, A.M.: Te a c h i n g Emergency Department Analgesia. Journal of the Residents to Care for Culturally Diverse American Medical Association 269(12)1537-9, 1993. Populations. Academic Medicine73(10):1056-1061, Trevino, F.M., Moyer, M.E., Valdez, R.B., and 1998. Stroup-Benham, C.A.: Health Insurance Coverage and Utilization of Health Services by Mexican Reprint Requests: David R. Williams, Ph.D., Institute for Social Americans, Mainland Puerto Ricans, and Cuban Research, University of Michigan, P.O. Box 1248, Ann Arbor, MI Americans. J o u rnal of the American Medical 48106-1248. E-Mail: [email protected] Association265(2):233-237, 1991.

90 HEALTH CARE FINANCING REVIEW/Summer 2000/Volume 21, Number 4