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.
2 HEALTH CARE OF MEMBERS OF RACIAL AND ETHNIC MINORITY GROUPS This chapter examines the extent to which people's membership in certain racial and ethnic minority groups influences whether and from whom they obtain health care. The committee found that health problems are more common among some racial/ethnic minorities than among the rest of the population, and that this fact is not fully reflected in their use of health services. This is particularly true of dental services and nursing home care. There is also considerable evidence that race and ethnicity have a strong influence on where people obtain medical care, which hospitals they use, whether they obtain care from clinics or private physicians, and the credentials of the physicians from whom they obtain care. The social processes that produce these patterns are complex, involving residential patterns, differences in socio- economic status, ethnic differences in concepts and values associated with disease and treatment, constraints and incentives that are built into federal health programs, and, possibly, discrimination. The factor of discrimination, which is particularly important in a civil rights context, is poorly understood because it has not been examined much in research or in legal actions in health care. However, the disparities reviewed in this chapter, in the committee's view, suggest that the question of discrimination in health care deserves more attention than it has received. The focus of this review stems from questions regarding civil rights and health care. In sharp contrast to such areas as employment, housing, and education, there is little to suggest that discrimination and integration in health care have been important public policy concerns in recent years. Yet some very serious problems of possible significance under civil rights laws have been described in testimony before the committee as well as before various congressional and investigative bodies and the courts. Legal services attorneys and journalists have described incidents in which minority group patients who are seriously ill, badly injured, or in active labor have been turned away from hospitals, transferred to other (public) hospitals, or subjected to long delays before care is provided.1 In addition, accounts of segregated health facilities occasionally still come to public attention. 20
21 Most of the incidents that have been described have occurred in private, non-profit hospitals that have a substantial federal involvement through Medicare, Medicaid, or the Hill-Burton program. (See Chapter 5 for a discussion of relevant legal issues.) Problems appear to be most serious for blacks in the South and for the Mexican- American and Indian populations of the Southwest and West. The cases generally appear to involve both patients' racial/ethnic status and their poverty or inability to pay for care. (The combination of minority status and poverty appears to be much more potent than either factor taken separately, as can be seen in the statistical evidence presented later in this chapter.) The grounds for refusing to provide care vary in these cases, but involve such hospital policies as requirements for advance payment under certain circumstances, non-acceptance of Medicaid or certain other sources of third-party payment, not informing patients of the "free care" obligations that the hospital assumed in accepting Hill-Burton funds, non-admission of patients who do not have a personal physician, requirements that poor patients complete applications for Medicaid (which may effectively deter immigrant patients who are here under color of law but who are not U.S. citizens), and attempting to determine the immigration status of patients and to transfer to Mexican hospitals patients about whom some question exists.* The following descriptions, several of which come from a recent hearing of the U.S. Civil Rights Commission, are typical of these incidents, the outcomes of which have included babies being born in hospital parking lots, serious medical complications, and the death of patients. All of the facts in such anecdotes cannot be known with certainty. Nor is it known how common such occurrences are or how often non-minority patients have similar experiences. There is, unfortunately, little way of linking anecdotes with the more systematic data on patterns of use of health services that are reviewed later in this chapter. However, the anecdotes suggest possible explanations for some of the disparities that are reviewed in this chapter, and they are a source of serious concern to persons working actively on behalf of the rights and welfare of minority groups. The committee, therefore, decided that the more systematic presentation of data should be preceded by some of the anecdotes that prompted the concerns that led to this study. ^Because of the complexity of immigration laws, uneducated immigrants, even if they are in the country legally, may be intimidated and deterred from seeking care by hospitals' posting signs stating that they cooperate with the Immigration and Naturalization Service, by requirements that the patient complete forms (such as Medicaid applications) that are to be sent to the government, and by hospital efforts to determine their legal status. The latter practice may be questionnable, both from a legal standpoint and from the standpoint of the proper purpose and mission of a hospital.
22 A young black woman in Memphis, Tennessee, suffering from a ruptured ectopic pregnancy, was refused admission at one private hospital that did not take Medicaid patients and was refused by a second private facility on grounds that the hospital did not take Medicaid patients for "female problems." At a public hospital facility she was informed that she would be seen, but only after the staff had treated a number of cases they considered to be more serious in nature than hers. ... In March 1979, a 29-year-old Mexican-American woman and her baby died of a ruptured uterus in a rural part of Texas. Two hospitals had turned away this acutely ill eight-month pregnant woman for inability to pay. Similarly, Ysidro Aguinagas, an 11-month-old Hispanic baby, died in December 1978 after being denied admission to a public hospital in Dimmitt, Texas, despite the fact that the hospital was a Hill-Burton facility and publicly financed. The baby would not be admitted without a $450 deposit. Since the parents were without a $450, deposit they left the facility to seek other sources of care, but the baby died en route.3 On August 1, 1976, Mrs. Carolyn Payne, a 21-year-old black resident of Holly Springs, Mississippi, delivered her own baby in the front seat of a truck after the emergency room of the Marshall County Hospital had refused admission. The Marshall County Hospital is a 60-bed county facility in Holly Springs, built with federal Hill-Burton funds, and supported by state and county health funds.4 Spanish-speaking citizens of Mexican descent are often presumed to be illegal or undocumented persons when they arrive for services in hospitals in southern California. A November 1979 newsletter reports an incident wherein an Hispanic man, conscious and speaking Spanish, arrived at an emergency room at 7 p.m. for treatment of stab wounds suffered in an attack. No doctor arrived until 8:30. Upon arrival, the doctor inquired about insurance for the patient and whether the patient was in the country legally. The wife, also Spanish speaking and monolingual, could not satisfactorily answer these questions. By 10 p.m. that evening, three hours after his arrival, the patient died. He had been inadequately treated. He was a U.S. citizen.5 In some of the cases, questions exist both about the facts and the state of the law that can only be resolved in the appropriate legal or administrative settings. The purpose of this chapter is neither to validate specific complaints nor to recommend how the legal and financial questions that arise in these situations should be resolved, but to provide a larger context within which certain social policy questions regarding equity or fairness in health care can be considered and to suggest some means by which disparities can better be understood.
23 The primary approach used in this chapter is the review of the research and statistical literature on the health care of racial and ethnic minority groups. However, the statistical information reviewed in this chapter does not permit a direct assessment of how common unfortunate incidents like those described above are. The examples that have come to light generally have not been through the organized data-collection activities of researchers or governmental statistical agencies. It is most difficult to link individual stories with aggregate statistics. Yet, to understand the meaning of any particular incident, such as might occur between a patient and a hospital admitting office, it is essential to understand into what pattern the incident might fit. While much of this chapter is concerned with racial/ethnic patterns in health care, it should be recognized that such patterns may have a variety of causes. However, the absence of racial differences in health care, except where there are differences in need for care, would be strong evidence that no widespread patterns of discrimination exist. It is significant, therefore, that the committee found that racial and ethnic factors continue to be important in health care. This and the following chapter examine racial/ethnic aspects of several segments of the health care system, such as physician visits, hospitalization, nursing homes, and dental care. The analysis is concerned with questions regarding (1) whether racial and ethnic factors affect people's ability to obtain needed care, (2) the extent of racial separation in health care, and (3) whether race and ethnicity affect the quality of medical care received. With regard to the first of these questions, there are indications that racial/ethnic differences in the use of services have been greatly reduced since the mid-1960s, although dental and nursing home care continue to be strongly related to race. To date, most governmental concern with racial/ethnic aspects of obtaining health care has been with trying to assure the availability of medical care to people who, because of poverty or place of residence, have not found care available. As noted in this chapter, some successes have been achieved, although questions of equity continue to arise. The second set of concerns pertains to patterns of racial/ethnic separation or segregation in the health care system and the processes and factors that contribute to such patterns. Integration has not been an important consideration underlying governmental health policy. Indeed, governmental policies, such as the initial use of Hill-Burton money to construct segregated facilities or, more recently, reimbursement policies that lead to the concentration of Medicaid patients in certain institutions, may be an important cause for a tendency toward segregation in health care. In the view of the committee, any examination of racial/ethnic differences in medical care must consider differences not only in the use of medical services but also in the source of medical care. Few carefully documented historical studies exist of racial patterns in health care, and most historical accounts of the medical care of blacks in the United States focus on issues that pertain to medical
24 manpower rather than the behavior of patients. Thus, most accounts have emphasized the black physician's role, barriers faced in their obtaining hospital admitting privileges, and the development of racially segregated health care facilities by law and custom. In some locales, hospitals either refused to treat black patients or withheld admitting privileges from black physicians, which led to the establishment of black hospitals.6 Only with the landmark Simkins v. Moses H. Cone Memorial Hospital decision regarding a segregated Hill-Burton hospital in Greensboro, North Carolina, was an explicit policy of promoting segregation abandoned within the U.S. Public Health Service.7 That was in 1963, the year before the Civil Rights Act was passed with its Title VI prohibition on racial discrimination in any program receiving federal financial assistance. Even one year after the passage of the Act, a U.S. Civil Rights Commission study found no discernable pattern of compliance in two-thirds of the hospitals surveyed. The subsequent enactment of the Medicare program made most hospitals and nursing homes potential recipients of federal funds. The Department of Health, Education, and Welfare then established an office to screen hospital applicants to see that policies of discrimination were not in effect and, after a four-month review process, reported that 3,000 hospitals that had previously practiced discrimination or segregation had come into compliance with Title VI.8 Although the effectiveness of the screening procedure came under subsequent criticism by the U.S. Civil Rights Commission and others, it was clear that an era was at an end, at least with regard to medical institutions in the United States, most of which received federal funds and were covered by the Civil Rights Act. (The practice of medicine in the offices of physicians was not directly affected by Title VI because of a DHEW determination that the indirect method through which they received federal dollars from Part B of Medicare did not constitute federal involvement sufficient to bring them under the authority of the Civil Rights Act.) The third set of concerns addressed in this chapter pertains to whether there are racial/ethnic differences in quality of medical care. The research literature on quality of medical care has devoted only limited attention to the race/ethnicity of patients as a factor that might influence quality of care, although the committee identified some approaches that are potentially useful in this regard. A recurrent problem encountered in this review concerns the availability of data. In many studies, minority groups are not represented in sufficient numbers to allow separate conclusions to be drawn about them.* The smaller and more dispersed the minority, the less adequate are the data. Frequently, the problem of small numbers leads to the grouping of diverse populations. Hayes-Bautista's *lt should be noted, however, that this chapter, which emphasizes racial rather than ethnic patterns, does not fully reflect the availability of information about other ethnic groups because of limitations in time and resources.
25 discussion of the diverse groups that are sometimes labeled "Hispanics" and of the confusion that is thus injected into public policy provides one example.10 However, in many studies and sources of data, the lumping together of minority groups poses greater problems than considering as one population ("Hispanics") the Raza of the Southwest, Haitians and Cubans in Miami, and Puerto Ricans in New York. In many studies data are presented only for "whites" and "others" or "non-whites." Important differences can be obscured by combining non-white groups in which poor health is relatively common with groups (e.g., Chinese or Japanese Americans) whose health status may be better than that of other "white" Americans. Sensitivity to all of these problems is reflected in recent reforms in governmental statistical policy and an effort to standardize a feasible set of racial and ethnic categories in all federal data-collection efforts. In recent years the virtual absence of health care data about many ethnic groups has begun to be addressed through procedures designed to provide representative samples of sufficient size of ethnic minorities. Such data are valuable in two senses: they are costly to obtain and there is no substitute for them. At present, however, the data needed to adequately assess many racial/ethnic differences in health status and use of medical services simply do not exist. Another problem with the available information is the difficulty of disentangling causal processes underlying patterns of use of health services. Ethnicity is always characterized by particular values, beliefs, and patterns of behavior. Behavior in matters pertaining to health is a frequently cited example. Assessing the impact of ethnically linked beliefs and values is beyond the scope of this report, although some attention is given to this matter in the next chapter. Clearly, an important challenge in health care is providing care in ways that are compatible with the needs of ethnic groups. Communication problems represent the most obvious manifestation of this problem. However, the impact of ethnic cultures on health behavior and the associated responses of health providers have not been examined in this report. HEALTH STATUS AND HEALTH CARE Racial and ethnic differences in health status have long been recognized in the United States. Apparent inadequacies in medical care have been the object of attention in governmental programs directed at poor people and residents of areas where medical resources are scanty. The Department of Health, Education, and Welfare's (DHEW) annual compendium of information, Health United States: 1979, notes that "in general, the health status of minorities has improved during recent years, and their use of health services has increased."11 In Health and the War on Poverty, Davis and Schoen observed that "poor people's access to medical care has increased remarkably [in the decade 1965-75] . . . steady progress
26 has been madeâparticularly in those kinds of poor health that are the roost prevalent among poor people and those that are most sensitive to improved medical care." Finally, in their introduction to Aday, Andersen, and Fleming's massive empirical study, Health Care in the U.S., Rogers and Aiken observe that: The findings are encouraging, and the country has made progress since the 1960s. . . . They show that we have found ways of getting more people into the health system at levels of use that seem more commensurate with their needs than heretofore. Every subpopulation group studied has better access to medical services today  than in 1963 or 1970, and in some instances the improvements have been dramatic. . . . Significant improvement for blacks is evident. The study also shows that contrary to popular opinion, most Americans (88%) seem generally satisfied with their medical care.1^ Important qualifications, however, were attached to all of these statements. The Health United States: 1979 statement about progress is followed by the observation that "many measures indicate that the health status of minorities is not as good as that of the white majority."14 Davis and Schoen warn that "much remains to be done. The gap [between the poor and others] has been narrowed, but not eliminated."15 Rogers and Aiken qualify their summary of progress with the observation that 26 million people still have difficulty obtaining appropriate medical care.16 In addition, the medical care situation of many poor people may have begun to worsen under the economic conditions of the late 1970s in ways that statistics are just beginning to show. Data from the Medicaid program, for example, show that the number of Medicaid recipients declined from 24,600,000 in 1976 to 21,600,000 in 1980.17 Without financial access to health care, it may be anticipated that the health status of people who were formerly eligible for Medicaid will decline. INDICATORS OF HEALTH STATUS The medical care system cannot be held entirely responsible for the many differences in health status within a society. The causes of poor health are complex and cannot be explained by statistics alone. However, measures of health status can provide an indicator of progress that remains to be achieved and are also an essential prerequisite to intelligent interpretation of differences in the use of services. Measures of Mortality The most recent figures from the National Center for Health Statistics continue to show the existence of substantial racial/ethnic variations in mortality. The 1970 age-adjusted death rates show whites (6.8
27 deaths per year per 1,000 population) in an intermediate position between the high rates for blacks (10.4) and "American Indians and Alaska Natives" (8.2), and the low rates for Chinese Americans (4.9) and Japanese Americans (3.3).18 There has been little change in either black or white rates since 1950, whereas the other three groups have all shown significant declines in mortality.19 Similar racial/ethnic differences are seen in life expectancy figures, which show white life expectancy to exceed black life expectancy by four to five years, whereas the life expectancy for Japanese and Chinese Americans appears to exceed white life expectancy.21 Available data for Hispanic populations, though somewhat dated and incomplete, suggests a mortality level that falls between rates for blacks and whites.22 Racial differences in infant mortality rates remain pronounced. Substantial declines occurred for all racial/ethnic groups between 1950 and 1977 in infant mortality, neonatal mortality, and post-natal mortality rates.23 Rates for Chinese Americans and Japanese Americans began and remained lower than rates for whites. The infant mortality rate for American Indians showed a dramatic reduction over this period, beginning at three times the white rate (82 per thousand live births versus 27 for whites in 1950) and moving to only slightly higher (15.6) than the white rate (12.3) in 1977. (By contrast, the post-natal mortality rate for American Indians remained at twice the white rate in 1977.) For blacks, even though substantial improvements in these mortality rates occurred in this period, rates in 1977 remained approximately twice the white rate for infant mortality (23.6 vs. 12.3), neonatal mortality (16.1 vs. 8.7), and post-natal mortality (7.6 vs. 3.6).24 Among Hispanics, there is some evidence of infant mortality rates that are elevated above rates for whites, but the data are for very limited geographic areas, and some are dated.25 A difficulty in using such gross measures of health status is that they reflect many social, economic, and cultural factors, and, thus, they cannot be interpreted as unambiguous indicators of differences in the adequacy of health care of different racial/ethnic groups. Yet the rapid changes in the infant mortality rates (particularly for American Indians) demonstrate that infant mortality is subject to dramatic improvements. Furthermore, black infant mortality rates do not stand at a uniformly high level, but show considerable variation from place to place. Figures for 1973-74 showed a range of from 17.2 in the fringe areas of large cities in the western United States to a rate of 32.7 in the non-urbanized South.26 By states, the most recent data show black infant mortality rates to vary from under 20 per 1,000 live births in Massachusetts and Washington to almost 30 per 1,000 in Illinois (and the District of Columbia).27 Thus, it is clear that the high rate of black infant mortality is not immutable. More generally, black mortality exceeds white mortality for most important causes of death in the United Statesâfor major cardiovascular diseases, for cancer, for diabetes mellitus, and for accidents and homicide,28 as well as for diseases such as influenza, pneumonia, and cirrhosis of the liver.29 Yet mortality data seldom,
28 if ever, mirror incidence data on how many people fall sick each year. Thus, such data cannot be interpreted as a sound measure of racial differences in the incidence of these diseases. Nor can such mortality data be used uncritically as an indicator of racial differences in medical care because mortality rates are affected by many other factors. However, some indications of possible differences in medical care for different populations may come from data on variations in both incidence and mortality rates for a disease. For instance, the National Cancer Institute (NCI) collects data on the incidence of diagnosed cancer, the stage of disease at diagnosis, modes of treatment used, and 5- and 10-year survival rates. The overall incidence of cancer (total of all sites) is higher for blacks (318.8 per 100,000 population) than for whites (297.7), although the relative rates of incidence vary markedly from site to site.30 (Thus, for example, breast cancer is more common in whites, while prostate cancer is more common in blacks.) The NCI data also show that for most sites cancer in whites is more likely to be localized at diagnosis than is cancer in blacks.3^ This suggests either that blacks do not obtain medical attention as early in the course of the disease as do whites or that the diagnosis of cancer is not made as early in medical evaluation of blacks as of whites. Data are not available to enable a choice between these explanations or to explain why either type of delay occurs. The NCI data contain other information that may show the effects of racial differences in health care. Five-year survival rates for most cancer sites are lower for blacks than for whites.3^ This appears to reflect more than early diagnosis among whites because the racial difference in survival persists even when the survival rates under comparison are hospitalized patients whose cancer was localized at diagnosis.33 The rate at which blacks develop cancer is 9 percent higher than that for whites, but the mortality rate for blacks is 30 percent higher than that for whites.34 Although differences in medical care are not the only possible explanation for racial differences in cancer survival, and serious questions can be raised about the representativeness of the NCI data, particularly with regard to blacks,35 the possibility that cancer survival data reflect racial differences in adequacy of medical care cannot be ignored.3** Similarly, racial differences in mortality from diabetes mellitus are much larger than racial differences in the incidence of the disease. ' Again, the extent to which differences in medical care play a role cannot be assessed with existing evidence. Measures of Morbidity Some indication of racial/ethnic differences in the need for medical care can be gained through an examination of differences in health status, both as these are reflected in data about the incidence or prevalence of various diseases and in people's self-evaluations of their own health status. Information about the incidence or
29 prevalence of different diseases is available from several sourcesâ from national studies conducted by the National Center for Health Statistics, from data on reportable diseases collected by the Center for Disease Control, and through epidemiological studies conducted within particular geographic areas. The prevalence of a large number of serious diseases is higher among blacks than among whites. Data from the U.S. National Health Survey show that, on average, blacks are much more likely than whites to report that they have diabetes, hypertension, and cerebrovascular disease, while the Center for Disease Control's data on syphilis, gonorrhea, and tuberculosis show rates much higher for blacks than whites. NCI data show that cancer is more prevalent among blacks than among whites.39 For some other diseases, such as heart disease, asthma, and some skin diseases, there is no great variation by race. For some conditions, such as bronchitis, arthritis and synovitis, eczema and dermatitis, and some digestive diseases, available statistics show that rates for whites are higher than for blacks.41 Also relevant to the need for medical care is fertility; the birthrate is substantially higher among blacks (22 live births per 1,000 population) than among whites (14 per 1,000).42 The piecemeal morbidity data that exist on Hispanics suggest that a variety of diseases may be more prevalent for them than for the majority of the white population. This was true for a variety of reportable diseases (e.g., amoebic dysentery, hepatitis, measles, mumps, syphilis, tuberculosis) in Los Angeles during the early 1970s.43 There is also some evidence of elevated morbidity and mortality associated with drug and alcohol abuse in the Puerto Rican population of New York.44 On the other hand, studies using carefully drawn samples in Alameda County, California, show the Chicano population there to have lower rates of chronic conditions, disability, and symptoms than either whites or blacks.45 Measures of Health Status Existing epidemiological evidence about the incidence or prevalence of specific diseases, though useful for many purposes, provides only a limited overall picture of racial/ethnic differences in health status. Such data are usually based either on very limited geographic areas or on reported cases that, for a variety of reasons, may not give a wholly accurate picture. (For example, people who do not seek medical care will not be counted, a factor of considerable importance if the data are to be used as indicators of whether needs for medical care are being met.) A different approach, commonly used in household surveys, is to inquire about people's own assessments of their health status. Although such data provide a useful health indicator, they also have obvious weaknesses. Considerable evidence exists that people's perceptions about their present health status are affected by many factors (including cultural factors and their own previous health
30 status).46 Furthermore, people who have obtained inadequate health care in the past may have different perceptions of when discomfort signals illness, and they may also be unaware of (and, hence, unable to report) some conditions.4' Nevertheless, self-reports of health status provide an important basis of comparison of the general health of large segments of the population. In general, self-reports of health status show that black and Hispanic populations are more likely than whites to think they have health problems. Table 1, from the National Health Interview Survey conducted by the National Center for Health Statistics, presents data on racial, ethnic, and income differences in self-reported health status, limitation of activity, restricted-activity days, and bed days.4 The relationship between health status and income is apparent in Table 1. In addition, within the two income categories, most of these measures show the health status of whites to be better than the other two groups, although the differences are neither large nor internally consistent. The differences in health status are particularly pronounced in some categories of persons who are most likely to be eligible for governmental health care programs (a category of special interest in a study prompted by civil rights concerns). Health Interview Survey data show that among elderly, low-income people, the health status of blacks is markedly worse than that of whites; the national picture for Hispanics is more mixed, with rates on some measures similar to whites.49 However, studies in cities such as New York, San Antonio, and Los Angeles suggest that the health status of such categories as "Spanish origin" or "Mexican-American" may be markedly worse than whites.50 The poor health status of elderly black persons as compared to elderly white persons has been confirmed in other national studies as well as in local studies in Los Angeles and New York.51 For children, there are some major racial/ethnic differences. Large racial differences in mortality characterize the youngest age categories. These decrease with age, and among teenagers, black mortality rates are only slightly higher than rates for whites.52 This is partially explained, however, by the greater frequency of accidental death among white teenagers; black teenagers are one-and-a-half times more likely as whites to die from disease. Parental* assessments of the health status of their children are consistent with these data. White parents are more likely than black or "Spanish" parents to assess the health of their children as "excellent," and they are less likely to assess their children's health as "fair or poor."5^ These differences are true in both families with incomes less than $10,000 and in higher-income families. Paradoxically, fewer disability days and bed days are reported for black children than for whites.54 The National Center for Health *In some instances, assessments in this household survey may have been by adults other than parents of children whose health status was being described.
31 Table 1. SELECTED HEALTH STATUS MEASURES, BY RACE/ETHNICITY AND INCOME, 1976-77 Income, Age, and Race or Ethnicity Population in Thousands Persons With Self-assessed Health Status As Fair or Poor Persons With Limitation of Activity Restricted- Activity Days* Bed Day s ^ All Incomes3 Percent of Population Number per Person per Year Black 23,066 19.1 14.6 20.7 8.9 Hispanic 11,913 12.8 9.1 16.7 7.8 White 160,129 11.0 14.0 17.6 6.6 Less Than 910,000 Black 11,961 23.5 19.2 25.1 10.5 Hispanic 5,681 17.2 12.0 21.3 9.9 White 44,555 19.5 23.8 26.2 9.8 310,000 or More Black 8,363 11.9 8.1 14.7 6.6 Hispanic 5,122 8.0 5.8 11.9 5.5 White 102,809 6.9 9.6 13.7 5.1 NOTE: The categories white, black, and Hispanic are mutually exclusive. ^Includes bed days, work-loss days, school-loss days, and other restricted-activity days. 2Bed days are a subgroup of restricted-activity days. 3Includes those for whom income was unknown. SOURCE: Division of Health Interview Statistics, National Center for Health Statistics. Data from the Health Interview Survey. Statistics, the source of these data, speculates that the apparent discrepancy between the parental assessments of children's health status, and the reported disability days and bed days, may be due to white children's greater access to medical attention (perhaps by telephone), which results in their being told more frequently than blacks to reduce or limit their activities.55 This interpretation is consistent with data on physician visits by children (reviewed later in this chapter) and on available information about the health status of children. An earlier national survey showed the reported incidence of acute conditions for persons under 17 years of age to be considerably higher for whites (270 conditions per 100 persons in 1973) than for others (169 conditions per 100 persons).56 Still, the overall picture
32 regarding racial/ethnic differences in the health of children is not completely clear. (Much more information about child health is summarized in the report of the Select Panel on Child Health Promotion. ") USE OF MEDICAL SERVICES Although the following sections examine statistics on the use of medical services, the limitations of such data should be acknowledged at the outset. Although some promising attempts have been made, it is difficult to link such statistics with measures of need for care. Thus, there is a tendency for all physician visits, for example, to be treated as equivalent, although there may be very little need for some visits, and other visits may be generated by improper patient care at the first visit. How aggregate utilization statistics are affected by such complexities is largely unknown. A different kind of problem arises because of the inevitable lags in data systems, a factor that is particularly important in times of rapid change. Racial and ethnic minorities that contain disproportionate numbers of poor people are differentially dependent upon compensatory institutions and programs that have been established to take care of the poor. The existence of compensatory institutions in times of growing budgets in public and municipal services is quite a different matter than in times of economic stress. Some governmental services affect most of the populationâroad repairs, police and fire services, and so forthâand cuts are felt across a wide segment of the population. By contrast, most governmental health programs affect relatively narrow segments of the population, making it possible to target the groups that will be affected by cuts in services. Thus, the groups that are dependent upon these programs are peculiarly vulnerable to cuts in governmental support. Since most available data on the use of health services are three to five years old, they probably do not fully reflect the current status of the health care of minority groups and, more generally, poor people. Ambulatory Care Visits to physicians provide a basic measure of the receipt of health care services and have been examined in national surveys conducted by the National Center for Health Statistics (NCHS) and the Center for Health Administration Studies (CHAS) at the University of Chicago. Data from both sources in the mid-1970s show a narrowing or elimination of earlier racial/ethnic differences in such matters as the interval since the last physician visit, having seen a physician in the previous year, and the number of physician visits in the previous year.58 NCHS and CHAS data on racial/ethnic differences regarding the latter two of these measures are shown on Table 2. At the most aggregate levels, racial/ethnic differences are relatively small. However, data from the more narrowly defined ethnic samples examined in the CHAS
33 Table 2. PHYSICIAN VISITS, BY RACE/ETHNICITY AND INCOME NCHS 1976-771 CHAS 19762 Number of Physician Visits Per Year All $10,000 Below & Above $10,000 All Above Below Poverty Poverty Level Level White3 5.0 4.8 5.6 4.1 4.1 4.1 Black 4.6 4.3 5.0 3.15 4.4* 2.95 4.56 3.25 4.36 Hispanic4 4.2 4.0 4.5 3.5 4.3 2.9 Percent of Persons with One or More Physician Visits in Past Year White3 76 77 76 77 77 73 655 675 786 645 756 Black 74 77 74 776 Hispanic4 69 71 69 â¬5 74 56 ,'-Data from National Center for Health Statistics (NCHS) . DHEW, Health United States, 1979 (Washington, D.C. : Goverment Printing Office, 1980) pp. 40-42. 2Data from Center for Health Administration Studies (CHAS), University of Chicago. Lu Ann Aday, Ronald Andersen, and Gretchen V. Fleming, Health Care in the U.S.: Equitable for Whom? (Beverly Hills, CA: Sage, 1980) pp. 102-106. 3"White" in the CHAS data does not include the "Spanish-heritage, Southwest" sample. 4CHAS data are from a sample of persons defined as "Spanish-heritage, Southwest." 5Data from sample of southern blacks not residing in Standard Metropolitan Statistical Areas (SMSAs). 6Data on "non-whites" other than those included in the sample of southern blacks not in SMSAs.
34 study show more substantial racial/ethnic variations. Lower rates of physician visits among rural southern blacks also have been reported in community studies." There are some substantial racial/ethnic differences in the number of physician visits per patient in the previous year. Fewer physician visits are reported by persons from minority groups. These differences largely reflect differences in the proportion of people who had not seen a physician at all. That is, among persons who had seen a physician the previous year, racial/ethnic differences in the average number of physician visits were smaller in magnitude; in all groups compared, the average number of physician visits for persons who had seen a physician at all was between 4.4 and 5.8. Thus, it appears that, on average, those racial/ethnic barriers that are seen most clearly to exist operate in a way that affects initial physician visits (that is, whether people see a doctor at all) more than follow-up visits. Among children, although differences in use of physician services have narrowed considerably as a result of a variety of federal programs, small racial differences still exist, as Dutton's review of the most recent information available through the Health Interview Survey conducted by the National Health Statistics shows.61 White children had more visits to physicians (4.3 per year) than did non-white children (2.9) in 1977, and 3 to 6 percent more non-white children than white children did not see a physician at all during that year.62 In an attempt to determine whether such figures mean white physician use is too high or non-white use is too low, Kovar reanalyzed data from the Health Interview Survey to determine the extent to which children fell below a medically defined standard of adequate numbers of physician visits.* Based on 1975-76 *The same unpublished data from the 1978 Health Interview Survey also show the racial difference in source of care (within different income categories) to be more pronounced in metropolitan areas (where more than two-thirds of the white population and three-fourths of the black population reside) than in non-metropolitan areas. Among metropolitan blacks below 150 percent of the poverty level, only 44 percent (compared with 63 percent of whites in the same category) report their usual source of care to be an office-based physician, and more than 26 percent (compared with 10 percent of whites) report an outpatient department or emergency room as the usual source of care. Similar findings have been reported among the black and Puerto Rican populations of New York, and Weaver reports a 1969 study in Orange County, California, which found that, although English- speaking whites and Mexican Americans expressed similar preferences for receiving care from private physicians or hospitals, Mexican Americans received care at a public health facility four times more often. Weaver attributes this tendency to Mexican Americans' previous negative experiences with English-speaking medical personnel, the presence of Spanish-speaking personnel in the public health facility, and the perception by English-speaking patients that the facility was a "Mexican" hospital.
35 data, 13 percent of white children had inadequate numbers of physician visits, compared with 19 percent of black children and 16 percent of others.63 Thus, by a variety of measures, small but consistent differences exist in the amount of physician care received by white and non-white children. Large-scale studies that examine racial/ethnic differences in the medical care of people with equivalent levels of medical need are scarce. One approach is to examine the medical care of persons whose needs can be considered roughly equivalent because they have similar medical conditions. Data from the National Ambulatory Medical Care Survey show that, although the prevalence of diabetes mellitus among "black and all other" women is 39 percent higher than for white women, the rate of visits to physicians' offices (per 1,000 population) for this condition is 27 percent higher for the former group than for white women.^4 The prevalence of hypertension is more than 82 percent higher among "black and all other" women than among white women, whereas the rate of visits to physicians' offices for the condition is virtually identical in the two groups.65 These comparisons, however, are based only on visits to physicians' offices, not on all physician-patient encounters for these conditions and, thus, probably do not provide an accurate picture of overall racial differences in medical care for these conditions. Prenatal care is a topic for comparison of racial differences in care for groups (in this case, pregnant women) that are to some degree comparable in terms of need. There is a more than 100 percent difference between whites and blacks in infant mortality. Although the causal factors underlying low birth weight are not well understood,68 prenatal care in pregnancy has played a major role in the overall downward trend in infant mortality in the United States in recent decades. By all measures, blacks on average receive less-adequate prenatal care than whites. For example, in 1975, among blacks more than 10 percent of the live births were to women who received either late (initiated in third trimester) care or none at all, compared with 5 percent of whites.69 Racial differences are present in all educational categories and are evident as well when illigitimate births are excluded.70 (However, illegitimate births among both black and white women are preceeded by similarly low levels of prenatal care.7^ The low level of care for this category may be partially due to the fact that in 19 states poor women who will become eligible for Medicaidâby means of Aid to Dependent Childrenâafter their first baby is born cannot receive services through Nedicaid before the baby is born.) Whites averaged more than two more prenatal visits than blacks.7^ Regionally, racial differences in prenatal care are most pronounced in the urban Northeast and least pronounced in the West. However, it appears that the "Spanish heritage" population of the Southwest is a group that is relatively unlikely to see a physician during the first three months of pregnancy,74 findings that have been confirmed by data from California.75 Another approach to examining whether there are racial differences in the relationship between medical care and medical need
36 has been developed in the studies conducted through the Center for Health Administration Studies (CHAS) at the University of Chicago. Persons surveyed were asked about their medical care and about symptoms and days of disability. From this information, two indices (known as the use-disability ratio and the symptoms-response ratio) were constructed that incorporate measures of both medical care and need for medical care.76 While both of these measures have limitations, they are amenable to use in population-based surveys, and since 1963 these measures have shown that wide income and racial differences exist in medical care.78 By contrast, the 1976 CHAS survey showed that earlier racial/ethnic differences had either disappeared or that, in terms of their self-reports of disability and symptoms, blacks were receiving more medical care than whites.7 At the same time, the credence given to these measures must be tempered both by the methodological problems mentioned earlier and by other uncertainties in their use and interpretation. (One such uncertainty, as Dutton notes, is seen in the fact that different analyses of the same data from the 1976 survey have shown that large income differences exist in the use-disability ratio (number of physician visits per 100 disability days) and that no differences exist, apparently depending on what method is used to standardize for differences in the age and sex distribution of the groups under comparison.80 Preventive Services Consistent racial differences exist in preventive care in children. For example, fewer white children (9 percent) than non-white children (15 percent) have never had a physical examination.8I- Poorer preventive care among non-white children is also evident in data (displayed in Table 3) on immunizations against infectious diseases. These data, compiled by the Center for Disease Control (CDC), show that the immunization rate for white children is consistently higher than the rate for children from "other races." On the other hand, a 1976 national survey conducted by the Center for Health Administration Studies found no racial differences in polio, measles, and DPT vaccinations, as reported by parents.82 (Because this survey shows much higher percentages of children to be vaccinated than did the CDC study, there is a possibility that parents overreport the vaccination of their children.) There is some evidence of similar deficits in preventive care among Hispanic populations. In the mid-1970s, large surveys in Alameda County, California, found that although Mexican Americans reported the same number of physician visits as English-speaking whites the former were less likely to report having had a general examination or eye examinations either in the past year or ever.8-* These differences persisted, though at a reduced level, even after statistical controls for the effects of education and family income. Racial Differences in Source of Ambulatory Care Virtually the same proportions of the black (86.3 percent) and white (87.7 percent) populations reported in the 1978 Health Interview Survey that they
37 Table 3. PERCENTAGE OF CHILDREN IMMUNIZED AGAINST FIVE INFECTIOUS DISEASES, BY RACE, 1978 3+ Doses 3+ Doses Rubella1 Measles1 DPT2Polio2 Mumps1 White Non-white 69 57 73 58 74 69 49 55 46 56 Includes ages 1-14. Includes ages 0-14. SOURCE: Center for Disease Control, U.S. Inmimunization Survey: 1978. have a "usual source of care." (Among those who do not have a usual source of care, whites are much more likely than non-whites to report that they previously have had a regular source of care. ) However, there are striking racial differences in where care is obtained. Roberts and Lee found a similar pattern among blacks, whites, and Chicanos in Alameda County; although there was little difference in having a regular source of care or in physician visits, there were notable differences in the source of medical care. Such patterns are evident in national data from the Health Interview Survey (Table 4). The first two columns show that, although there is little racial difference in having a usual source of medical care, whites are somewhat more likely than blacks to report that they usually see one particular doctor. Whites are much more likely than blacks to report their usual source of care to be an office-based physician and less likely to report outpatient departments and health centers as their usual source of care. Table 4 also shows that, while both types of insurance (private and Medicare vs. Medicaid) and income level affect people's usual source of medical care, a racial difference persists even within different income groups and among people who have similar types of health insurance. Thus, it is clear that more than poverty underlies the racial disparities in where people obtain their medical care. Racial differences in the source of care also are evident in a community study of the health care of rural and "urban fringe" blacks and whites in North Carolina, where "74 percent of the rural and 64 percent of the urban whites named a private physician as their usual source of care, against 22 and 18 percent of blacks in the respective areas."86 The researchers concluded that, because so few blacks and Medicaid patients were served by community physicians (and thus served by neighborhood health centers and county health departments), "removal of legal and financial barriers has made little impact as yet on the patterns of health care delivery established before the institution of mandatory integration of health services."87
38 0) 'Jj U Â£> â¢H â¢ M p ** it ft O< eft ,-! n ,Â» ,*-, ^N nH CO VO in rH a c j^i 0 0 i0 01 i Q) * 4J rH 0i 0i i ,H CD tT) 00 ^ rH rH C E u 0) 0 u j 32 Â£ u 4J 0i E , &< Q rH n] ^"* u * A 4J 00 c 0 C nj M in a J< 'o r- rH U an ra co vo c-j oo ^ en TJ ^^ c ^. rH oo in m rj Qj 4J pa JS ^2 M a iJ *' | S -H Â§ a n 1 *H VO CTv o VO *H n rH .C CO VD 3 w 0) a ,1 tM c Â§ O D o 4C , y U 00 rfc *r in vc oo fv *~H r~ ^* *y rH at l-j 0) n rH ' C - o 0i ^ P z 4J sÂ« < i Â§ """ * i0 in rH m F u 9 H M a U g O k ^j C0 'o a CM m CM <Â»1 t 0) S C â¢H rH en in VO â¢H E5 M i0 ^ pa ,*J" CM rH |Â» D tM o H M z o â¢H c a â¢S 0 4) ^ 0 CM VO G )H s â¢HO r- * rH P Â§ * vc rH W * 4J 2 M i0 U in n 0) 0 to r- vo i0 o CM tr Â£ 0) 0 4J C rH 00 VD rH i0 a u O % * â¢H 3 -H 4J o H u a U '*"* Cft VO 01 CO rH rH â¢H 2 a. c Jg x: oo r~ ^ 4J 2 M X n CJ , H 4J M JC ra if o CO ^^ i0 VO 0 CD p*1 CO VD 'H 00 VO M rH M i-i â¢ J3 a a 4J S S fH o Â« EH â¢H 00 ^ p. ^ rH rH B u fc o .C oo r- ** k O H VM Â§ 0 C a I 1 -H O C 0 -H . S 3 S M U Â£ M U 1 u oi o 5 tk ational i | . S& w Q Q u â¢a 0 at 4J a w 01 f rH 0) rH U 0i a 3 a a n to n S I & 1 ^* 3 rH w 3 * 0) 0) i0 O u I U 4J C z Â£ S, 5 0i a o >,-n a <o â¢H r-l 4J -HO 0) â¢io S 4J Â«H -H i 0i â¢io 3 6 0i H rH tM 5 â¢" US DO. i0 O O o w a: i0 5 n Â£ Â£-1 D X CO
39 A second type of racial disparity in people's sources of medical care concerns the use of medical specialists. Table 5 shows the way in which physician visits by blacks and whites are distributed across physician specialties, as reported in household interviews. The table shows that a slightly greater percentage of physician visits by black patients than of white patients are to general practitioners. When the major settings for physician visits are examined, much larger racial differences appear. Table 6 shows the racial distribution of physician contacts by specialty, according to whether they took place in the office of a private physician, in a hospital clinic or emergency room, or over the telephone. (The small number of physician visits at other sitesâat home or at work, for exampleâare not included in this table.) Physician contacts by telephone were more than twice as common for whites as for blacks. In addition, physician visits for blacks were twice as likely as for whites to take place in hospital clinics and emergency rooms; visits with internists and pediatricians were more than three times as likely for blacks as for whites to take place in such settings. This suggests that private practitioners are more available to whites than to blacks. Although there is evidence that some organized health care settings can provide good-quality care to poverty patients,88 a large literature suggests that the use of hospital clinics and emergency rooms as a usual source of care has serious deficiencies. Nevertheless, little evidence exists that allows systematic comparison of quality of medical care across the types of sites discussed herein. Associated with the racial differences in sources of care are differences in the ease with which people obtain care. "Difficulty getting to the doctor" was second only to cost as a barrier to care cited by poor people and non-whites questioned in the 1974 Health Interview Survey, and surveys such as those conducted by the National Center for Health Statistics (NCHS) and the Center for Health Administration Studies (CHAS) have consistently found that the travel time of non-whites exceeds that of whites. For example, 1976 CHAS data show 49 percent of whites and 40 percent of non-whites travel less than 15 minutes to reach their usual source of care.90 How much of the racial difference in travel time is due to differences in the distance traveled, and how much is due to differences in modes of transportation, cannot be ascertained directly from the available data. However, in the CHAS data, the racial difference was present, though reduced, within income categories.91 NCHS data, presented in Table 7, show that black travel time exceeds white travel time no matter what income or type of insurance they have. The same unpublished data from the 1978 Health Interview Survey also show racial differences in travel time among persons who live in metropolitan areas (data not shown), notwithstanding the greater use by poor blacks of what might seem to be "local" sources of careâneighborhood health centers and hospital clinics and emergency rooms. Explanations of Racial/Ethnic Patterns in Ambulatory Care Data presented thus far show that (1) although some racial/ethnic differences in receipt of medical care have disappeared, minority groups, taken as a whole, remain at a disadvantage by some measures; (2) consistent racial/ethnic differences remain in the sources from which people obtain medical care;
40 Table 5. PERCENTAGE OF PHYSICIAN VISITS, BY RACE OF PATIENT AND SPECIALTY OF PHYSICIAN, 1978 Race of Patient Specialty of Physician White Black General Practitioner 48.5 54.6 Dermatologist 1.7 1.5 Internist 10.5 5.5 OB, GYN 6.8 6.5 Ophthalmologist 2.5 1.5 Orthopedist 4.3 2.4 Otolaryngologist 2.3 1.3 Pediatrician 9.5 8.2 Psychiatrist 1.1 1.0 Radiologist 1.2 0.6 Surgeon 3.0 2.7 Urologist 1.3 2.0 Other Specialists 3.2 5.3 Unknown 3.6 6.5 Total 100.0% 100.0% SOURCE: National Center for Health Statistics. 1978 Health Interview Survey, unpublished data. and (3) whites seeking medical care use less travel time, on average, than do blacks. A number of factors may possibly explain at least part of these differences. These include racial/ethnic differences in preferences for various sources of care, racial/ethnic differences in ability to pay for medical care, spatial patterns in the location of populations and sources of medical care, and discrimination by providers either against minority group members or against classes (for example, Medicaid patients) of which some minority groups constitute a disproportionate share. Literature in the social sciences shows that differences in values affect virtually every kind of human behavior, including seeking medical care. However, although different values and other cultural factors undoubtedly influence whether people seek medical attention when particular symptoms are experienced, it seems unlikely that minority groups' disproportionate use of clinics and emergency rooms and their expenditure of greater amounts of travel time to obtain care are to an important degree a true expression of preferences. Not only is there little evidence to suggest this is true, but competing explanations are more plausible. A set of factors that undoubtedly affects patterns of medical care relates to cost. Disproportionate numbers of both black and Hispanic populations in the United States are found in low-income
41 Table 6. PERCENT DISTRIBUTION OF PLACE OF PHYSICIAN VISITS, BY PHYSICIAN SPECIALTY AND RACE OF PATIENT, 1978 Place of Visit Physician Specialty Race of Patient Physician's Office Hospital Clinic/ Emergency Room Tele- phone Total1 General White 70 11 13 100% Practitioner Black 60 24 4 100% Internist White 70 10 15 100% Black 55 33 * 100% OB, GYN White 76 8 13 100% Black 68 13 12 100% Pediatrician White 69 5 25 100% Black 56 18 20 100% Total2 White 69 12 13 100% Black 57 25 5 100% Too few cases for reliable percentage to be calculated. IROWS add to less than 100 percent because some visit sites are not shown and because of missing data. Table includes only specialties with sufficient numbers to allow for calculation of percentage distributions across visit sites. The total rows, however, include all physician visits. SOURCE: National Center for Health Statistics. 1978 Health Interview Survey, unpublished data. categories.92 Furthermore, disproportionate numbers of the poor have no insurance coverage for medical care; this is particularly true for the Mexican-American population of the Southwest.93 Considerable evidence exists that ability to pay (including having insurance) has marked effects on people's ability to obtain health care. 4 In addition, because of the association between racial/ethnic status and poverty, the Medicaid rolls include disproportionate numbers of minority persons. Yet it is also clear that these factors do not suffice as an explanation of racial/ethnic differences in medical care, because, as has already been noted, substantial racial/ethnic differences exist within income and insurance categories. Nevertheless, there can be no doubt that differences in income and in Medicaid status influence the racial/ethnic patterns that have been described in this chapter. This is particularly obvious in the case of income, because under the present health care system in the United States only a limited number of public facilities are available to provide care to persons unable to pay for it. The health care system is designed to concentrate persons who cannot pay in a few
42 Table 7. PERCENT TRAVELING MORE THAN 29 MINUTES TO USUAL SOURCE OF MEDICAL CARE, 1978 White Black Type of Insurance Private Insurance and Medicare 15.0 18.6 Medicaid Only 20.5 25.3 None 15.4 23.8 Income Level More than Twice Poverty Level 16.2 19.2 Less than 150% of Poverty Level 20.0 23.8 Total 15.4 20.8 SOURCE:National Center for Health Statistics.1978 Health Interview Survey, unpublished tables. facilities. (The so-called "free care" provisions of the Hill-Burton Act, described in Chapter 5, have had little effect on this because they have not been enforced and because hospitals see obvious economic disadvantages in providing care to people who cannot pay for it.) The increasingly common accounts of poor patients (including Medicaid patients in some cities) being "dumped" from the emergency rooms of voluntary or private hospitals into public hospitals (with grave economic consequences for the latter) is one example of the pervasive phenomenon of each element in the health care system (including both providers and different levels of government) seeking to shift costs elsewhere. One consequence of this trend is almost certainly the concentration of poor and minority patients into relatively few, economically unhealthy facilities. Spatial Distribution of Medical Resources Many studies have shown that the spatial distribution of health manpower does not mirror the distribution of the U.S. population.95 Most of these studies have focused on rural-urban differences and have called attention to the manpower problems of rural parts of the United States. However, because the minority population of the United States is disproportionately urban, these studies are of limited usefulness in explaining the racial patterns that have been outlined in this chapter. More useful are the studies that have been conducted at the local level. The relative scarcity of private physicians in urban neighborhoods in which ethnic minorities predominate has been documented in a variety of studies. Most of this research involves analyses of the relationship between the supply of physicians in different areas (such as census tracts) of a city and certain characteristics of the areasâmedian income, mean age, availability of
43 hospital beds, and so forth. Most of this research is directed at understanding patterns of physician location, and it is concerned less with describing the resources available to minority communities than with considering the racial or ethnic characteristics of an area as one of many factors that may "explain" physician-location patterns. These studies show that physicians' offices tend not to be located in areas where there is a predominance of black and Hispanic residents. Such research in some cities has found that the racial factor operates independently of the other factors studied (median income, supply of hospital beds, and so forth),97 although there are also indications that in more middle-class black communities the problem is much less pronounced, in part, because of the locational preferences of black physicians.98 Whatever the relative importance of class and racial factors, it is clear that the overwhelming burden of social-class differentials with regard to proximity to physician services falls on non-whites. This is seen both through correlational studies regarding the relationship between physician supply and race of residents99 and descriptive studies of the characteristics of poor, minority neighborhoods. Thus, several of the poorest neighborhoods in New York City are reported as having as few as 0.15 office-based physicians for every 1,000 residents.100 Evidence from a Chicago study showed the concentration of physicians in affluent neighborhoods to have increased between 1950 and 1970; the 10 most affluent communities saw their physician/population ratio rise from 1.78 to 2.1/1,000 in that period; within the 10 poorest communities, the ratio dropped from 0.99 to 0.26/1,000.101 Although the scarcity of private physicians is an important fact of life in neighborhoods occupied by minorities and the poor, it does not suffice as an explanation of racial differences in the source of health care. Regarding use of hospitals, several studies (reviewed later in this chapter) show that travel beyond the nearest facility seems to be a common pattern among urban blacks. Similar patterns exist with regard to ambulatory care. Data from a 1968-71 survey conducted in 10 cities shows that, even within the same general neighborhoods, large racial differences exist in where people obtain medical care, as is shown in Table 8.102 The data strongly suggest that the disproportionate use by blacks of hospitals and public clinics cannot be attributed simply to proximity, because the usual source of medical care for whites in the same neighborhoods consistently differs from that of blacks. These differences were very large in some cases; in southeast Philadelphia, for example, more than half of blacks, but fewer than 10 percent of whites, reported that a hospital or public clinic was their usual source of medical care. Furthermore, Table 8 also shows that the travel time of blacks to their medical care is consistently larger than whites who reside in the same general area of the city. (Whether this reflects a difference in mode of transportation or in distance traveled is again not clear.) The important point is that factors other than the geographic distribution of medical resources affect racial differences in where people obtain medical care. For some reason, blacks in these areas make less use of private physicians than do whites.
44 Table 8. USUAL SOURCE OF CARE AND TRAVEL TIME AMONG RESIDENTS OF 10 URBAN AREAS, BY RACE, 1968-71 Area and Survey Year Roxbury, Boston, MA. 1971 Bedford Stuyvesant-Crown Heights, Brooklyn. 1968 Red Hook, Brooklyn. 1968-69 Southeast Philadelphia, PA. 1968-69 Upper Cardozo, Washington, D.C. 1969 Southside, Atlanta, GA. 1968 Peninsula, Charleston, SC. 1969 Wayne Minor & Model Cities Area. Kansas City, MO. 1969-70 Mission, San Francisco, CA. 1970 East Palo Alto, CA. 1969 Percent Reporting Hospitals and Public Clinics as Usual Source of Care Black White 76 43 30 53 40 72 56 51 31 12 59 14 6 9 18 21 11 36 17 8 Percent Traveling 30 Minutes or More to Usual Source of Care Black White 60 58 52 44 60 78 51 60 66 28 52 40 38 35 59 58 21 51 65 26 SOURCE: Louise M. Okada and Gerald Sparer, "Access to Usual Source of Care by Race and Income in Ten Urban Areas," Journal of Community Health 1 (Spring 1976) pp. 163-174. Discrimination by Physicians Many of the data presented thus farâmost notably the findings that blacks make less use of private physicians than do whites with similar incomes and insurance coverage and that black and white residents of the same general urban areas use different sources of careâare consistent with the hypothesis that minority group use of health care is influenced by patterns of discrimination among physicians. Despite scattered reports of physician discrimination in the form of segregated waiting rooms or office hours,103 no data exist on the extent to which racial
45 discrimination exists in actually accepting patients for treatment. That a hospital that excluded black patients could still exist in the late 1970s lends plausibility to the possibility that some individual practitioners may still practice discrimination. It was initially hoped that the Medicaid program might help integrate eligible persons into the mainstream health care delivery system. Its success in doing so depends in substantial part on an adequate level of participation by providers. For various reasons, including low payment levels, a significant proportion of physicians apparently do not accept Medicaid patients, although the existing estimates of physician participation are very imprecise. In a national survey of more than 3,300 physicians conducted by the National Opinion Research Corporation in 1975-76, 77 percent of responding physicians answered affirmatively to the question "Do you participate in your state's Medicaid program; that is, do you receive payment from Medicaid?"105 This figure should be interpreted carefully, however, for three reasons. First, not all specialties were included (although the most common ones were), and because members of minorities are found in disproportionately small numbers in specialists' practices, leaving out some specialties may artificially inflate the estimate of the percentage that accepts Medicaid patients. Second, one-third of the physicians did not return their questionnaire; although the authors show that these non-respondents were similar to respondents in some regards, the possibility exists that they were different with regard to acceptance of Medicaid patients. Third, physicians who had as few as one Medicaid patient may have answered affirmatively to the question regarding acceptance of Medicaid. Thus, Mitchell and Cromwell also examined Medicaid patients as a percentage of physicians' practices, and found that, in addition to the 23 percent of physicians who had no Medicaid patients, another 27 percent had fewer than 10 percent Medicaid patients.106 Perhaps the best indicator of the availability of physicians to Medicaid patients is whether they would accept new Medicaid patients. A survey of general practitioners conducted by Mathematica Policy Research in 1975 showed as many as half were not taking new Medicaid patients, as is shown in Table 9. These rates are for general practitioners, and it is probable that they overestimate overall physician acceptance of Medicaid patients, because Medicaid participation in many specialties is much lower.107 Physician acceptance of Medicaid patients was particularly low in the South and in large cities, which coincides with the location of the bulk of minority group members in the United States. Table 9 also includes, for comparison purposes, figures on physician participation in Medicare. Physician participation is consistently higher in Medicare than in Medicaid, which may be due to the size of the market (there are more Medicare patients than Medicaid patients, and they use more care) and to higher rates of reimbursement. However, the data are also consistent with the hypothesis of racial discrimination. Differences in physician acceptance of Medicare and Medicaid patients are most pronounced in areas where racial/ethnic minority groups are concentrated. Thus, in the non-metropolitan Northeast and West, only
46 Table 9. GENERAL PRACTITIONERS' ACCEPTANCE OF MEDICARE AND MEDICAID PATIENTS, BY AREA AND REGION, 1975 Percent Percent Taking Percent Percent Taking Patients New Medicare Patients New Medicaid Region Medicare Patients Medicaid Patients Large SMSAs Northeast 25.9 79.9 12.4 56.2 North Central 25.6 78.3 12.9 52.1 South 27.3 78.3 10.9 42.8 West 29.3 84.5 17.0 49.9 Small SMSAs Northeast 30.3 84.8 19.8 73.0 North Central 26.3 80.3 12.9 53.7 South 26.0 67.5 15.8 46.0 West 21.8 83.3 17.9 61.6 Non-metropolitan Northeast 36.1 85.0 19.7 78.5 North Central 28.0 75.9 12.4 65.0 South 25.3 60.6 18.8 51.3 West 19.8 64.6 13.5 58.7 Totals Large SMSAs 26.9 79.8 12.9 49.7 Small SMSAs 26.1 76.3 16.2 55.2 Non-metropolitan 26.8 68.8 16.4 59.9 SOURCE: The Physician Capacity Utilization Surveys: Special Analyses, DHEW Publication No. (HRA) 79-30, (Washington, D.C.: DHEW, 1979) p. 225.
47 about 7 percent more physicians accept new Medicare patients than Medicaid patients. On the other hand, more than one-third more physicians in large cities of the South and West will accept new Medicare patients than will accept new Medicaid patients. (The non-metropolitan South; it should be noted, is not notably different from other non-metropolitan areas of the country regarding the difference in acceptance of new Medicare and Medicaid patients.) Quality of Ambulatory Care In discussing the results of a study they conducted in the rural South in the mid-1970s, Davis and Marshall make the following observations about what they learned about racial differences in the quality of medical care: Cursory, inadequate physical examinations are frequently given to minority patients. In some places, rural blacks are unaware that it is customary to undress for medical examinations while this procedure is common among whites in the same area. Blood pressure readings are taken through the clothing of black patients, thus increasing the risk of inadequate measurements, a particularly serious problem for blacks with a high incidence of hypertension. Minority women are less likely to receive professional preventive services such as Pap smears and breast examinations. High rates of hysterectomies are also seen in some areas. Â°^ Unfortunately, no description has been published of the methods used in this study, and no data are presented on the frequency of these shortcomings in the medical care of blacks and whites. However, echoes of these findings can be heard in Senator Moss's account of three "Medicaid mills" that he visited in New York City while posing as a patient. Moss writes of dirty facilities, impersonal care, unnecessary tests and prescriptions, and blood pressure and pulse readings being taken through clothing.110 It is difficult to define the role played by the racial/ethnic characteristics of patients in the patterns of care that Davis and Marshall and Moss describe. However, when such accounts are considered in light of the history of racial discrimination in the United States, the association between race/ethnicity and income, and the segregated patterns described earlier in this chapter, it is reasonable to ask whether there are racial differences in the quality of medical care provided. Great interest has arisen about the quality of medical care, and an active research literature has developed. This literature, which suggests that important deficiencies occur in the medical care of Americans, provides relatively little systematic information about racial differences in the quality of medical care.112 Indications that a racial difference exists in the quality of care is provided by some of the information already presented in this chapter; however, more direct measures are also available. One determinant of quality of medical care is the competence of the physician providing care. A number of studies have shown that the
48 quality of care in a medical setting is influenced by factors such as the percentage of specialists practicing there and the length of their training.113 As was noted earlier in this chapter, more of the care provided to minority groups than to whites is provided by non-specialists. Furthermore, there are indications that less than fully qualified foreign medical graduates provide care in state- financed medical institutions, particularly state mental hospitals, where the patients are disproportionately poor and black.11 Studies of physicians providing care to Medicaid patients provides some useful, inferential material about the care of poor members of minority groups, although many Medicaid patients are not from minority groups. Relatively few physicians provide care to relatively large numbers of Medicaid patients; estimates from a national survey suggest that 5 percent of the physicians in the country may provide care to one-third of the Medicaid patients. Physicians who provide care to relatively large numbers of Medicaid patients include disproportionate numbers of general practitioners, and, because of the negative association between age and specialty training, they tend to be older than the average physician.116 Foreign medical graduates also provide a disproportionate amount of the care to Medicaid patients.117 Kavaler's study of 126 physicians participating in the Medicaid program in the black and Puerto Rican slums of New York City found that 35 percent had no access to hospital beds and 42 percent had only limited privileges at proprietary hospitals. 18 (Alers also reviews data that raises questions about the qualifications of some physicians providing care in Puerto Rican neighborhoods in New York.)**' Similar results were reported from a study in Chicago.120 Because of such characteristics of physicians who treat relatively large numbers of Medicaid patients, Mitchell and Cromwell suggest that the Medicaid program and its beneficiaries constitute a "secondary, residual market" for medical care.121 That is, the physicians least able to compete in the medical marketâbecause of foreign training, lack of specialty credentials, or lack of hospital privilegesâend up providing much of the care for Medicaid patients. Thus, "a primary goal of the public benefits programs to integrate the poor into mainstream medicine is thereby thwarted."122 Patient satisfaction provides another aspect of possible differences in the quality of medical care. Patients' assessments of the care they receive are influenced by a variety of factors, including waiting time and time spent with physicians,123 and cannot be considered a measure of quality in a strict medical sense. Nevertheless, patients are the only persons who are in a position to judge certain aspects of the care that they receive, and their perceptions of that care should be taken seriously. The best available evidence shows higher levels of dissatisfaction with various aspects of medical care among both blacks and Hispanics than among whites. Table 10 shows such racial/ethnic (as well as income) differences in patients' evaluations of several aspects of the care that they receiveâits convenience and availability, the financing of care, the humaneness of doctors, the quality of care, and
49 CN CN VO ro f^ vo r^ en m en v in ^ in oh m ,5f CO rH â¢v r~- rH 00 rsi o en i0 co o vo co ro hi ro vo ro CM r- o Â£ 00 m rH CM f CM H rH g o 3 12 " in S Â£ â¢H 4j o JS CW rH ro 9 Â§ * M a S-n 4J oo in Â«r m in vo in in 00 VO 3 00 oo ro CM m o vo vo vo vo in 0i il S Â«a a m i0 I h s a o a iM i Q O . 1 Â£ 4J 4J i â¢H n m â¢H rH O o en in T VC rH rÂ» in ft i0 CD VO rH Oh m O I 1-8 .Â» m in Â» in in in in in ^ in m t a Â« W is it D ra Â£ tJ o m VO CO 0 0 n m ^ o 00 ^ rH Oh O rH 4J Hj C CO in ^ ^3" ^* in in in m in io in in m in m 8 tD 2 a B _ i SS in vo â¢r^ l 0i a a> M M rH rH Â« f^ 0 i0 U '* ra 3 0 ) Â§ ^ Is 8*5 4J '8 83 < 0i Z U rH K ," â¢ ,â Oh Oh in rH oh r-* Oi â M i0 Oh CM ^ Oh o 0i U In -HO .w m vo m <Â» Â» m m to VO in in m ^ in 4) "H c n fM -â¢ h B CO * I 4J 8 Fleming i M â¢ 2 Â£Â» Q -H 0i â¢-â¢HO in 4J 0i 0, rH U en ro .H â¢ Â£ -H -H > IH r, JD > 00 0 in vo oo m en H v0 iO vo f~^ oo ** o a IH n Â« 5 o! ,* >fi Â«r in Â«r â¢* M r- 00 00 oo vo vo vo m rH 0i 0i S â¢ â¢ I 5 jj -H W o1 o fa 0 i C * s s-s C "S 3 4J 2 O n â¢H a C9 4J kl tJ) 1 M nil rH JQ -H Ol 3 -H CO ,H > oo ,Â» in ^ co m CM r~- to CM oh in ro oh o Â»1\ e EH vo 13 ,ir m in in i0 vo tÂ» rÂ» r- vo vo vo in S *^C t-~ > CO M CO aÂ» c ^J 1^ CO rH 0 iÂ« 3 JC M *o M u o '0 C7i 0 c â¢ Q - n i0 -H vu < ro in H H Southwe â¢ i -o " 3 U 1? Q w rH rH rH rH r-H 3 H H rH rH â¢ 2 2 I vw J3 vV ^ r QJ a o> V 9 o 4J 2 o â¢ 0) 0 0i 0i 0i 0i Â» 00 S H ttl 0i rH rH H r-i rH rH rH rH W 0) 4J !: 7 I 51 W * ij* Â£ I? .M â¢HuO (2 o Â£ 4J kl kl hi U Jtf hi kl -H kl kl a *o *S k> fH vl j, ,; 0i > 8 o ! â¢ I f I S HI C o< 0) S as. jj O a H 8. c R R c a â¢H & < U) â¢ o ) -U Ul 3 â¢â¢ 0 White Spanish ii Below 0i I C2| ,""Percent 2In paren numbers , SOURCE: L Hills, CA ! Q rH *4 > rH kl O rH Is JS CO 0) 0) o 0) 3 O i I I pa -c < ca rH S 1
50 overall dissatisfaction. The data, taken from the most recent national survey conducted by the Center for Health Administration Studies, show generally higher levels of dissatisfaction among both the "Spanish heritage, Southwest" sample and among blacks than among whites, and appear to reflect, at least in part, differences in the source of care.124 The racial/ethnic differences were generally more pronounced on the measures of convenience and availability than on the measures of performance of physicians (humaneness and quality), although that has not been found in all studies.125 A majority of persons expressed satisfaction with the various dimensions of medical care,12** and although perceptions of the same objective reality may differ, the pattern seems clear that whites are more satisfied with their medical care than are blacks and the Hispanic population of the Southwest. Hospital Care The most recent national surveys do not show consistent or striking differences among whites, blacks, and Hispanics in the rate of hospitalization.127 However, this may not indicate that blacks and whites have the same access to hospital care when it is needed. Since a variety of serious health problems are more common among blacks than among whites, equivalent access to care might be expected to produce higher rates of hospitalization among blacks. (Indeed, data for enlisted Naval personnel, whose access to care is presumably only minimally affected by race, show higher rates of hospitalization for blacks than for whites.128) Data from both the Health Interview Survey and the Hospital Discharge Survey show the length of stay for blacks to be higher than for whites.129 Whether this indicates a racial difference in patients' condition upon admission, as is sometimes suggested, is speculative. There are some indications that hospitalized blacks are slightly less likely than whites to have surgery, although assessment of the meaning of this difference is most uncertain. Hospital Discharge Survey data also show great variation in the ratio of whites to non-whites in the incidence of different surgical procedures.131 However, given both the level of aggregation in the published data and the amount of missing racial data in the Hospital Discharge Survey, no conclusions can be drawn about racial trends. Regarding a set of surgical procedures about which particular concern has been expressed over the yearsâsterilizationâavailable data show earlier racial differences in incidence to have largely disappeared. Data from the Hospital Discharge Survey, for example, show that, in 1971, black women were undergoing tubal sterilization at a rate of 12 per 1,000 women aged 15-44, while the comparable rate for white women was 5.3; by 1975, the rate of tubal sterilization among blacks was still 12 per 1,000, while the rate for whites was 11.6 per 1,000.132 Similar trends are evident in survey data published by the National Center for Health Statistics. As of 1976, among women
51 aged 15-45 who had ever been married, blacks (13.1 percent) were more likely than whites (10.5 percent) and Hispanics (9.4 percent) to have ever undergone a tubal ligation.*133 However, in the years between 1973 and 1976, the incidence of tubal ligation was slightly higher among whites (5.7 percent) than among blacks (5.5 percent) or Hispanics (4.7 percent). Data on hysterectomies showed little difference between blacks and whites in ever having the surgery (approximately 8 percent of both groups) or having had it in the previous three years (approximately 4 percent of both groups). The rate of hysterectomies among Hispanic women aged 15-45 was somewhat lower. The topic of sterilization of minority groups has also been linked to concerns about informed consent for several years, having received considerable attention in Senate hearings held in 1973.13* Following those hearings, regulatory changes were made in Medicaid that attempted to set conditions more conducive to informed consent and to limit the use of hysterectomy for sterilization purposes. No systematic studies of the effects of those regulations have been published. However, there is an imperfect indicator of racial differences in the incidence of sterilization under conditions that may be questionable from the standpoint of informed consent: the percentage of tubal ligations performed on women who were pregnant when hospitalized. Because the hospitalization is for a purpose other than sterilization, the possibility is increased that the woman might not understand that a sterilization procedure is involved. Informed consent procedures undertaken as part of the process of labor and childbirth can easily go awry.13** Therefore, consent is better obtained prior to hospitalization. Although no data are available regarding when consent is obtained for sterilization, both the regulations and the concern about the issue may have been responsible for a large decrease between 1970 and 1975 in the proportion of tubal ligations performed on women who were pregnant when hospitalized. However, in the most recent year for which data are available (1975) , 59.7 percent of the black women who underwent tubal ligation were pregnant when hospitalized, compared with 41.5 percent of white women. While this difference may be a reflection of racial differences in fertility,137 it points to a potential source of consent problems that is more common among blacks than among whites. Patterns of Hospital Use The racial patterns in physician visits have parallels in patterns of hospitalization, although no national data exist that are comparable to the data reviewed above on racial *When this report was written, all available data from the NCHS interview survey were for women who had been married. However, never married women make up a significant proportion (22.1 percent) of black women who underwent tubal sterilization in 1975, compared with only 2.3 percent of white women.134 Thus, data that are limited to the incidence of sterilization among ever married women may possibly obscure continuing racial differences.
52 patterns in the use of physicians. Studies conducted in several cities, however, describe the elements of a dual track system. This is true of all cities for which information is available on racial patterns in hospital care. Comprehensive studies of the workings of entire urban medical care systems are scarce. One of the best studies was of Chicago over a 15-year period ending in 1965. It found that blacks were hospitalized in a very small subset of the more than 150 hospitals in the metropolitan area.138 Fifty percent of all black patients in Chicago traveled to one hospital, Cook County General Hospital, where 85 percent of the patients were black.139 Another 30 percent of black patients were served by five university-affiliated teaching hospitals and one (of three) traditionally "black" hospital. Writing in the late 1960s, DeVise described the racial trends in hospitalization as follows: The dual system of Negro indigent patient hospitals and white private patient hospitals has persisted even though extensive Medicare and Medicaid programs now reimburse private hospitals and physicians for the care of indigents; even though OEO and Children's Bureau now pay private hospitals to set up free neighborhood health centers and pediatric clinics; even though there has been a sevenfold increase in the number of Negro physicians admitted to practice in private white hospitals; even though the Negro ghetto has more than doubled in area, absorbing in the process six more white hospitals; even though the average distance from Negro homes to Cook County Hospital has increased from five to eight miles, while the average distance from Negro homes to white hospitals stayed under one mile.140 A vivid measure of the nature of the travel patterns involved in the use of Cook County Hospital comes from the calculation that the 500,000 patient miles per month that were traveled to Cook County Hospital would be reduced to 50,000 miles if patients used the hospital nearest their homes.141 The average distance traveled by all black patients to the various Chicago hospitals where they were admitted was six miles; the average trip for a white patient was three miles.142 Travel time was also used as an indicator of a racial dual track system in a study of hospital use in Cleveland.143 Seventy-four percent of the blacks surveyed, compared with 59 percent of whites, traveled beyond the hospital that was second nearest to their home. The authors noted that personal and cultural preferences can lead to travel to a hospital. Thus, for example, Jews were particularly likely to travel beyond the second nearest hospital (92 percent did so) in order to make use of one particular hospitalâMt. Sinai. The authors noted, however, that "the concentration of blacks at Metropolitan General Hospital suggests a different set of constraints featuring poverty and discrimination."144
53 Descriptive material on racial patterns in hospitalization are also available for New Orleans, where data were assembled in connection with a race discrimination suit (Cook v. Ochsner Foundation Hospital, et al.) brought by the Department of Health, Education, and Welfare in 1970. These data, which are included in a summary of the Cook case that is presented in Appendix E of this report, showed blacks to be concentrated in two hospitals; of blacks that had been hospitalized in 1974-77, 75 percent had gone to 2 of the 16 hospitals in metropolitan New Orleans. 4^ Conversely, blacks were underrepresented in other New Orleans hospitals. Identifiably "black" hospitals and associated patterns of racial segregation continue to exist to some degree in hospitals in many other cities, although no systematic analysis has been done on the topic. However, applicable research methodologies have been developed in studies of segregation in other areas, such as education and residence,146 and data that are potentially useful for describing racial patterns of where people obtain medical care exist from programs such as Medicare and Medicaid. In principle, studies could be done of the degree of segregation in hospital use in various cities, the extent of its variation from city to city, and whether it is increasing or decreasing in response to factors such as economic trends and civil rights enforcement activity. Quality of Care in Hospitals Existing studies of quality have generally not examined racial and ethnic issues. Nevertheless, despite the complexities of defining and measuring quality, several existing approaches can be used to detect at least gross disparities that may deserve closer examination. These approaches include studies of resources (especially personnel), treatment processes, and outcomes. Within medical institutions, there are indications that care is provided to minority groups by less well trained physicians. Studies of this matter are not common, perhaps because of its sensitivity. Duff and Hollingshead showed that the social class of patients had a pervasive impact on their care at a large university hospital.147 The class position of patients, for example, influenced whether a patient was managed by a "committee," with no one clearly identifiable to the patient as responsible for his or her case, or whether the patient had a "committed" sponsor in the hospital. Although this study provided extensive documentation of the effects of class in hospital care, it was confined to whites. More direct evidence about the racial factor in hospitals comes from Egbert and Rothman's study of the relationship between patient characteristics and the training of their surgeon at a teaching hospital.148 Blacks were much more likely than whites to be under the care of surgeons in training (that is, a resident surgeon) rather than a staff physician. For example, among patients who were "paying directly or with commercial insurance," 34 percent of blacks and 7 percent of whites were treated by a resident surgeon. However, among Medicaid patients, no statistically significant racial difference was found; more than 40 percent of both white and black Medicaid patients were treated by residents. Egbert and Rothman also found that among
54 emergency patients, blacks were twice as likely as whites to be treated by a resident. Egbert and Rothman's study was based on medical records and could easily be duplicated at other institutions and for other types of care. However, no other well-documented accounts appear in the literature in which racial differences within institutions are examined. Similar racial/ethnic sorting processes have been described in mental health settings. For example, Flaherty and Meagher, in a study of 66 black and 36 white male schizophrenic inpatients, found blacks to be more likely than whites to have been given medications on an "as needed" basis, less likely to have received recreation therapy and occupational therapy, and more likely to have been put in seclusion and to have had restraints used. These differences appeared to have been due to subtle racial stereotyping among staff members and their greater familiarity with white patients, rather than to racial differences in pathology. In a study of racial differences in the treatment of children in five mental health clinics, Jackson, Berkowitz, and Farley found that black children were less likely than white children to be accepted for treatment, less likely to receive individual treatment, and (at two clinics) to be seen for a lesser length of time.150 In his study of services provided to members of different ethnic groups in 17 centers in Seattle, Sue found significant differences in the types of personnel seen both at intake and during therapy.151 At intake and during therapy, blacks saw significantly fewer psychiatrists, psychololgists, social workers, and nurses, and more "other professionals," non-professionals, and "other personnel," than did whites.152 This was true even after demographic differences were controlled statistically. However, no consistent pattern of differences from whites were found among the ethnic groups studied. Thus, for example, American Indians saw more social workers and fewer non-professionals than did whites, Asian Americans saw fewer professionals than did whites, and Chicanos did not differ significantly from whites in the kind of personnel seen. Sue also examined ethnic differences in diagnoses, the type of program and the number of sessions in which patients became involved, and their rates of premature dropping out (an indicator of the effectiveness of the program). All of the results are summarized in Table 11, which shows different patterns of treatment among the different ethnic groups, with blacks differing from whites on all the variables examined. When statistical controls were introduced to eliminate the effects of ethnic differences in age, sex, education, income, and marital status, members of all of the ethnic groups were still significantly more likely than whites to terminate prematurely their course of treatment. Over the years, the psychiatric literature has shown more evidence of concern about possible racial bias in treatment than has the literature of any other area of health care. Considerable evidence exists of serious concern about racism and mental health. The problem of cultural differences is a particularly difficult one in the field of mental health because of the nature of both the problems that are addressed and the theoretical conceptions that have been
55 Table 11. SUMMARY OF FINDINGS REGARDING ETHNIC-WHITE DIFFERENCES IN PATIENTS AT 17 COMMUNITY MENTAL HEALTH FACILITIES Native Asian Blacks Amerleans Americans Chicanos Utilization Rates + + - - No. of Demographic Differences 5 3 3 3 Significant Diagnosis Differences Yes No No No Type Staff Seen at Intake Yes Yes Yes No Type Staff Seen in Therapy Yes No Yes NO Type of Program Yes No No No Type of Service Yes No No NO Number of Sessions Yes Yes Yes Yes Premature Termination Yes Yes Yes Yes SOURCE: Stanley Sue, "Community Mental Health Services to Minority Groups: Some Optimism, Some Pessimism," American Psychologist (August 1977), pp. 616-624. dominant. These problems have not been addressed in detail in this report and perhaps cannot be examined within the broad focus that it takes. That serious efforts have been made within the field to assess the operation of racial/ethnic biases in treatment is evidence of concern about these problems. The results of this research suggest that concern by those outside the field is also warranted. Another approach makes use of existing information to examine differences in the process by which care is provided. An example is Shaw's study using 1968 hospital discharge information from the Commission on Professional and Hospital Activities.153 He found differences between whites and blacks in the rates that various diagnostic and therapeutic techniques were documented in medical records, as is shown in Table 12. Furthermore, the data show that these racial differences were less likely to occur in small hospitals than in large hospitals, and were less likely to occur in the western section of the United States than in the remainder of the country. The presentation of Shaw's data does not allow judgments to be made regarding inequities in medical care. The operational details used were not described, and the nature of the racial differences was not indicated. Nevertheless, the methodology itself appears to be powerful and worth pursuing, at least in an exploratory or research mode. Its potential for better documenting differences in care, many of which have strong quality implications, appears promising. Studies of Outcomes of Hospital Care Differences in treatment cause concern, at least in part, because of suspicions that such will produce differences in the health of people. However, researchers have found it difficult to link process and outcome measures of quality of care. Although differences in process are of themselves
56 a Â£ CO to CTi * Â§ X I â¢ u â¢ Ml â¢H 0) X O . 0) >i E-â¢ ' 'â¢ rH 0 i0 3 . 4J TJ CO CM rH I 4J C0 4J M CO U II: S B PS I SB O O B B -^ -rl O â¢H Q. 4J -rl M O ID 4J >, O 4J O C 3 s2 51 S2 rH N PI ,Â» B O a Â£ O o i0 jj a, -H N I 5 CQ at S a -n Â« 0i M U 0i O Som Sim 'W II te â¢ â¢ tH 4J a B x Â§ Â§ H Â« 0) -H 0) X X * Â«H Ot B 0) 0i hi E hi >i EH 0i -H O rH fj rH o *. X i0 B â¢> -H B - U E X. *Â» > 0) a B -H (0 O rH hi O â¢H hi U W M a O Q D a Â« fi o* a Â§Â» â¢rl 0i C Â§ I ? â¢ U <0 I-) D) SÂ«g0g"g W 09 X 4 i0 fH fH rH "rl O* **H 0) X X Qj 4J *0 4J m w 0iWoaMa - o *Â» 0) O O B rH 4) -H .0 SS 3 g a 'D " -2 " * B hi >i EH i0 rH Â§ Â§ Â£ Â£ 5 S^T: o Q a h a 0) O fiO Â»-5 Â« ? m B 0) >i M >. EH B Â«) rH CJ rH 0.x i0 Â£ a ! 1 M*J" S 8 E *j > n u 4J iW 0) Â« i*a â¢ a 4J 5. JJ iw -H E -H <fl 9 $ 4J CO E 4J O E U B 0) rH 14-i ,
57 important, because of the possibility that minority groups may be underserved or treated differentially, some approaches that focus more heavily on outcome are also potentially useful. Rutstein et al. have suggested a method that might be applied to examining the outcomes of care as a measure of quality.154 They postulate that "sentinel events" can often be identified that represent preventable disease, disability, or untimely deaths. Under the approach they describe, the occurrence of these events triggers an investigation in which the responsibility may be determined and corrective action taken. They point to the example of child and maternal health, in which every maternal or perinatal death would lead to an investigation of the prior events to determine what, if anything, went wrong and what steps should be taken to avoid such outcomes. It is clear that not all such deaths are the result of inappropriate or unacceptable care, but enough have been documented and enough changes have been made to significantly lower maternal and perinatal death rates. By the same logic, documented racial/ethnic differences in the occurrence of "sentinel events" that represent preventable disease, disability, and untimely deaths would be important evidence that unequal, inequitable, and possibly discriminatory care was being given. It is not difficult to identify a number of sentinel events that might be investigated. Decubitus ulcers, infection associated with intrauterine devices, morbidity and mortality associated with various forms of treatable cancers, and the complications of poorly managed intravenous techniques are just a few examples. Most hospitals already monitor such events as post-surgical complications and transfusion reactions, and professional standards review organizations (PSROs) sometimes conduct studies in which racial comparisons could be made in measures of quality of care. Apparently, no research on racial/ethnic differences has been published based on such an approach. However, some "sentinel events" could be studied using currently available data systems, such as those that exist at the Health Care Financing Administration for Medicare patients and in certain states for Medicaid patients and those at PSROs for both types of patients. If racial/ethnic differences are small, multi-institutional studies will be required. If they are great, differences might be identified within individual institutions. Data from the National Center for Health Statistics Hospital Discharge Survey might also prove amenable to such research. Three points should be made concerning this approach. First, the study design must account for differences other than race and ethnicity, such as age, sex, income, and severity of illness. Second, differences in the rates of occurrence of sentinel events may be based on racial factors rather than on the medical care that is provided. Even so, such differences may point to differing medical care requirements. Third, as technology advances and practices change, the "sentinel events" (the criteria) may require change, as may the acceptable level of occurrence (the standard). Hypertension provides an excellent illustration of the last two points. Recent studies clearly show that we have developed a
58 technology to control blood pressure, and, further, that the successful application of this technology results in a significant decrease in morbidity and mortality, even for people with minimally elevated levels, and that these benefits were most pronounced among blacks. It is reasonable to expect that the application of this knowledge will lead to a decrease in strokes and hypertensive cardiovascular disease. It is also known that the incidence and severity of hypertension is greater among middle-aged black males than among others of similar age. This racial difference has more to do with race and sex per se than with differences in medical care. However, recognition of this fact should lead to the application of more resources to this unmet need. Thus, monitoring of the impact of this new technology should show a greater impact on morbidity and mortality among blacks than among whites. Dental Care Some of the best-documented racial differences in health status are in dental health. Data on treated and untreated dental disease are available through the dental examinations that have been conducted as part of the National Health and Nutritional Survey by the National Center for Health Statistics and are based on sound, representative samples of the U.S. population. As Table 13 shows, consistent racial differences exist in needs for dental treatment (that is, in untreated dental disease). The differences are smallest for the youngest children, but even in the category of age 1-5, black children (5 percent) were more than twice as likely as white children (2 percent) to need dental care to "remove debris and calculus." Among older children and adults, the rates of untreated disease are much higher, and the percentage differences by race are much larger. For example, among blacks aged 12-17, 75 percent had untreated tooth decay, compared to 48 percent of whites. Among blacks aged 65-74, 20 percent needed extractions because of periodontal disease, compared with 7 percent of whites. Despite evidence showing greater needs for dental treatment among blacks than among whites, national data from the Health Interview Survey show that whites use the services of dentists more than do "other races."157 As Table 14 shows, this can be partially explained by differences in income, particularly since public funding of dental care is very limited, but racial differences exist even within income categories. Although fewer data are available, there is also evidence of low levels of dental care among the Chicano population.158 The explanation of the striking racial discrepancies regarding need and use of service in the dental area is not clear. There are some indications that "discretionary" health services in general are less likely to be used by blacks than by whites; thus, use of dentists in the black population is more confined to the treatment of relatively serious conditions than is true in the white population.159 However, in at least some parts of the country, the explanation may involve
59 TT JC CO CM in o r^ in rH CM T 7 JJ i cn â¢ r- en r- CM i o ^ CM rH in rH rH â¢ r- CM rH 1-i C*1 i CM rH rH t~ m H 2 O 0) n en m iH i0 co o i0 r- in H 4J in â¢ m ro "^ vo i vo co m in â¢H Â» en VM Jj) vo Â§ rH in rH CM rH rH rH i O S f^ 3 *z* r- 4J 5s ^ rH â¢* rH vo en in in o r- 4J Â§ en . r^cMvoinl co voco a W ^ â¢ O vo Â« CM vo i m c C5 vo rH rH On M *â¢ O pa rH a Â», S f% u C-| 0i E en in m m rH r~- O m oo jj fi CO â¢H en â¢ â¢ â¢ â¢ â¢ â¢ â¢ â¢ a B rH r^ â¢ 0O o rH ro en i m ro cn H ^ CM VO ^ CM rH m i CM 4J < O rH S i JX rH 3 . U i0 00 f- o r- co m r- H in H o ft rH rH m â¢ ^ ^ in in rH i en o 0i 3 CQ m oo vo CM r- i X M 0 Â£ 0 S i8 u S Â« fi 4J vo in vo i0 Â» ^ in o i0 Â»H < rH g o â¢ oo ro rH co m i in o Oi . CQ * CM i*1 rH ^ '! c P W u Â« rH Q JC 0O rH ft g u in in ^ 0O CM VO .Â» rH PH C â¢â¢ 5 i0 O o rH rH Â» rH vo m o o CM i o o â¢rt 00 O ro r- m ^* ^* I *Â» en g S JS <0 rH H fl r- - -C p 4J O CM 00 vO rH CO r* CM in ", 4J i VO i i** â¢ i i Â» CM oo rH o QO en | o o EH en vo CM CM m ^ Jj! EH â cn W C u rH en rH rH rH 00 rH u S 3 3 Â° U in m Â» en m o o o en i i i Hi ii Q O CM rH ia u C 4J 0i g S V o O 4J ,rl 0 S CM en O rH ^ rH 4J m s . rH i CM â¢ Â» in CM | i (3 in i i0 g ,Â» rH ii rH i ! ,rH O 1Â° O rH & C S in C -H O S r- H t~cMOenvo c~ oo O 4J Â« - -<T *0 r- o i0 vo CM vom -H n Â« 8 e rH VO m rH rH m rH â¢ -n o en a u > rH 'O .no 1 Â£ * S H . C 3 -H 4J 4J C 0i -C 03 to O C HI 0) 4J i0 -H 0i 0i i EH 0i OO)4JM g Z2 S l2 HÂ« a! a e -^ oijjnuo Â£ a u M C i0 â¢H 8 Â« 4) fl -Ht04J JJ .H 0. O cn oi n e u > u .H fr. 0 ij 10 o) c WQP.C J3 10 0i u 0) >i a \0i S3 u EH c u *n rH uj a -C rH C *J JQ T* rH 3 tm â¢S 0i 0i H EH i0 g <0 0i C 0i a p, Treatment 0i c m o Z â¢2 VMU a i0 M -H cCTjrH ^,Populatio SOURCE: H Medicine, H V 4J 4J 0 ) QrH -H 4J 0i ^ OO -H -Q i0 4J C Cki t^ *rl*O klQ rHU-HO 4JOOa> 0 kH C M 4J 0 ) 0) i0 U i0 >IO â¢, "* u -H O >TJ -H 0 >, >i tob -DE M rH 1 ESS QC cn -^ a *o LJ o) oioi oi B*0 c M u u 4Jdt x ^ Â£ a I s. 0) ,H0)$0)K -H 4J K O M Q M H DM O EH
60 Table 14. PERCENT OF PERSONS WITH A DENTAL VISIT WITHIN A YEAR, BY RACE AND INCOME, 1977 Income Level White Other Races Under $5,000 35.4 29.9 $5,000-9,999 38.9 34.3 $10,000 & Over 58.6 41.9 SOURCE: Health Interview Survey, National Center for Health Statistics (Institute of Medicine, Public Policy Options for Better Dental Care, 1980, p. 21). racial discrimination. The authors of a study of dental care in a rural North Carolina county (where whites reported twice as many dental visits as blacks) attribute part of the large racial difference to the "slow to change established pattern of private health care delivery in Durham Countyâa certain reluctance on the part of both providers and patients to expand service patterns between like race to those of opposite race when resources are scarce."160 The extent to which this explanation may account for the national racial patterns in dental care is unknown. CONCLUSION From the information described in this chapter, the committee drew the following general conclusions. First, there is considerable evidence that racial/ethnic factors continue to influence patterns of health care in ways that are not in the interests of the groups that are affected. These patterns are consistent with the belief that minority groups are still exposed to discrimination in this country, although little direct evidence is available. Second, racial/ethnic patterns in health care deserve much more serious and systematic attention than they have received to date from researchers and governmental statistical agencies. More studies are needed that empirically examine the factors that influence the medical care decisions of minority group members. In recent years, the National Center for Health Statistics and the National Center for Health Services Research have shown more concern with collecting data that will be statistically valid for members of minority groups.161 Such data are expensive to collect because of sampling problems, yet they are of great importance if equity questions in American health care are to be assessed. The present Health Care Financing Administration (HCFA) data are so inadequate that it is virtually impossible to draw meaningful conclusions about racial/ethnic equity in the Medicaid, Medicare, and Title V programs. HCFA should extend strong efforts to improve the quality of the racial/ethnic information
61 that is collected on beneficiaries of these crucially important federal financing programs. The agreement that was signed between HCFA and OCR in 1980 may be the first step in increasing HCFA's attention to racial and ethnic issues in Medicare and Medicaid.162 Third, the question of racial separation in health care and associated questions regarding quality of health care also deserve much more serious attention than they have received to date. At present, many federal policies do not encourage racial/ethnic integration in health care. Since members of some minority groups are disproportionately dependent upon Medicaid, policies that encourage the segregation of Medicaid patients also encourage racial/ethnic segregation. There is little indication that racial separation and the question of separate-but-equal have received serious consideration. Fourth, specific attention should be given to the factors that may explain the striking racial patterns regarding dental care. The data clearly show that the need for dental care (as defined by untreated disease) is much greater among blacks than among whites, while the use of dental services is much greater among whites than among blacks. Existing data suggest that these trends are due to more than socioeconomic differences. REFERENCES 1. Hattie Mae Campbell et al. v. R. J. Mincey, Appeal filed with the U.S. Court of Appeals for the Fifth Circuit, No. 76-1325; Geraldine Dallek, "Summary: Health Care for California's Poor: Separate and Unequal," (Santa Monica, CA: National Health Law Program, 1979); Donald Donati, Sylvia Drew Ivie, and Beth Lief, Letter to Harry P. Cain, Bureau of Health Planning and Resources Development, DHEW, October 7, 1977; Carmen Estrada, (Mexican American Legal Defense Fund, San Francisco) Testimony before the Institute of Medicine Committee on Health Care of Minority Groups and Handicapped Persons, Los Angeles, CA, May 9, 1980; Georgia Legal Services et al., "Testimony on Behalf of Legal Services Clients before the Subcommittee on Health and the Environment of the House Committee on Interstate and Foreign Commerce Regarding the Health Planning and Resources Development Amendments of 1978 (H.R. 10460) and Title III of the Hospital Cost Containment Act of 1977 (H.R. 9717)," February 1, 1978; Sylvia Drew Ivie, Ben Thomas Cole, and Beth Lief, Letter to Harry P. Cain, Bureau of Health Planning and Resource Development, DHEW, and to David Tatel, Director, Office for Civil Rights, DHEW, November 7, 1977; Sylvia Drew Ivie and Howard Newman, Joint memorandum from the Director, Office for Civil Rights, and the Administrator, Health Care Financing Administration on "Civil Rights Responsibilities," August 28, 1980; Dorothy T. Lang, Letter to Rep. Charles Rangel, Subcommittee on Health, Committee on Ways
62 and Means, U.S. House of Representatives, March 24, 1980; Mexican American Legal Defense and Educational Fund v. Madera Community Hospital, Administrative complaint filed with the U.S. Department of Health, Education, and Welfare by Carmen Estrada for the Mexican American Legal Defense and Educational Fund, San Francisco, CA, September 5, 1979; Maria Carmen Ordonez et al. v. Mercy Hospital of Laredo, Administrative complaint filed with the U.S. Department of Health, Education, and Welfare by Lee J. Train for the Texas Rural Legal Aid, Inc., Laredo, TX, April 8, 1980; "Patient Dumping," Investigative series published in the Long Beach Independent Press-Telegram, July 6-16, 1980; Anne Ronan, Letter to Mr. Peter Schey, Director, Alien Rights Center, Legal Aid Foundation of Los Angeles, March 21, 1980; State of California Department of Health, Transcript of Civil Rights Hearing held in Los Angeles, CA, November 14, 1976. 2. Carmen Estrada (Mexican American Legal Defense Fund, San Francisco, CA), Testimony before the Institute of Medicine Committee on Health Care of Minority Groups and Handicapped Persons, Los Angeles, CA, May 9, 1980. 3. Sylvia Drew Ivie, "Ending Discrimination in Health Care: A Dream Deferred," Presentation before the U.S. Civil Rights Commission, April 15, 1980. 4. Sylvia Drew Ivie, Ben Thomas Cole, and Beth Lief, Letter to Harry P. Cain, Bureau of Health Planning and Resource Development, DHEW, and to David Tatel, Director, Office for Civil Rights, DHEW, November 7, 1977. 5. Ivie, pp. 29-30. 6. Bonnie Bullough and Vern L. Bullough, Poverty, Ethnic Identity and Health Care (New York: Appleton, 1972); Paul B. Comely, "Segregation and Discrimination in Medical Care in the United States," American Journal of Public Health 46 (September 1956) pp. 1074-1081; Herbert M. Morals, The History of the Negro in Medicine (New York: Publishers Co., 1971); Dietrich C. Rietzes, Negroes and Medicine (Cambridge, MA: Harvard University Press, 1958). 7. Kenneth Wing, "Title VI and Health Facilities: Forms Without Substance," Hastings Law Journal 30 (September 1978) p. 145. 8. Bullough and Bullough, p. 162. 9. Wing, pp. 154-161. 10. David E. Hayes-Bautista, "Identifying 'Hispanic' Populations: The Influence of Research Methodology Upon Public Policy," American Journal of Public Health 70 (April 1980) pp. 353-356. 11. Department of Health, Education and Welfare, Health United States: 1979, DHEW Publication No. (PHS) 80-1232 (Washington, D.C.: Government Printing Office, 1980) p. 3. 12. Karen Davis and Cathy Schoen, Health and the War on Poverty: A Ten-Year Appraisal (Washington, D.C.: The Brookings Institution, 1978) p. 1. 13. Lu Ann Aday, Ronald Andersen, and Gretchen V. Fleming, Health Care in the U.S.: Equitable for Whom? (Beverly Hills, CA: Sage, 1980) p. 18.
63 14. U.S. Department of Health, Education, and Welfare, Health United States: 1979, p. 3. 15. Davis and Schoen, p. 1. 16. Aday, Andersen, and Fleming, p. 18. 17. Health Care Financing Administration, "Medicaid State Tables, 1977" (Unpublished data, 1980). 18. Health United States: 1979, p. 15. 19. Ibid., p. 17. 20. Melvin H. Rudov and Nancy Santangelo, Health Status of Minorities and Low-Income Groups, DREW Publication No. (HRA) 79-627 (Washington, D.C.: Government Printing Office, 1979) p. 36. 21. Lester Breslow and Bonnie Klein, "Health and Race in California," Journal of the American Public Health Association 61 (April 1971) p. 769. 22. Health Resources Administration, Health of the Disadvantaged: Chartbook, DHEW Publication No. (HRA) 77-628 (Washington, D.C. : DHEW, 1977); K. S. Markides, H. W. Martin, and E. Gomez, "Older Mexican Americans: A Study in an Urban Barrio," Chapter 6: Health Status and Health Care Utilization, Unpublished manuscript; Chicano Health Institute of Students, Professors, and Alumni, The California Raza Health Plan: An Action Guide for the Promotion of Raza Health in California (Berkeley, CA, CHISPA: 1979). 23. Health United States: 1979, p. 21. 24. Ibid. 25. Jerry L. Weaver, National Health Policy and the Underserved: Ethnic Minorities, Women, and the Elderly (St. Louis, MO: Mosby, 1976) p. 60; Ian Canino, Brian Barley, and Lloyd Rogler, The Puerto Rican Child in New York City: Stress and Mental Health. (Bronx, NY: Fordham University Hispanic Research Center, 1980) . 26. Mary Grace Kovar, "Mortality of Black Infants in the United States," Phylon 38 (December 1977) p. 389. 27. Health United States: 1979, pp. 140-141. 28. Ibid., p. 25. 29. Rudov and Santangelo, pp. 91-92. 30. Lillian M. Axtell, Ardyce J. Asire, and Max H. Myers (eds.), Cancer Patient Survival, Report No. 5 from the Cancer Surveillance, Epidemiology and End Results (SEER) Program, DHEW Publication No. (NIH) 77-992 (Washington, D.C.: Government Printing Office, 1977) p. 4. 31. Ibid., p. 5. 32. Ibid., p. 9. 33. Ibid. 34. Jack E. White et al., "Cancer Among Blacks in the United StatesâRecognizing the Problem," To be published in the Proceedings of the International Conference on Cancer among Black Populations, held at Roswell Park Memorial Institute, Buffalo, NY, May 5 and 6, 1980, p. 10. 35. Earl S. Pollock and John W. Horm, "Trends in Cancer Incidence and Mortality in the United States, 1969-76," Journal of the National Cancer Institute 64 (May 1980) p. 1101.
64 36. Donald Fredrickson, Testimony before the House of Representatives Subcommittee on Health and the Environment, Committee on Interstate and Foreign Commerce, June 18, 1979. 37. Health United States: 1979, p. 25; National Diabetes Data Group "Selected Statistics on Health and Medical Care of Diabetics, 1980," forthcoming. 38. National Diabetes Data Group; Rudov and Santangelo, pp. 114 and 118; Department of Health, Education, and Welfare, "VD Fact Sheet - 1971"; Center for Disease Control, 1978 Tuberculosis Statistics: States and Cities, DHEW Publication No. (CDC) 80-8249 (Washington, D.C.: DHEW, 1979). 39. Pollock and Horm, pp. 1093-1094. 40. Rudov and Santangelo, p. 110; National Center for Health Statistics, Prevalence of Selected Chronic Respiratory Conditions, United States-1970, Vital and Health Statistics, Series 10, No. 84, (Washington, D.C.: Government Printing Office, 1973); National Center for Health Statistics, Prevalence of Chronic Skin and Musculoskeletal Conditions, United States-1969, Vital and Health Statistics, Series 10, No. 92, (Washington, D.C.: DHEW, 1974). 41. National Center for Health Statistics, Prevalence of Selected Chronic Respiratory Conditions, United States-1970, Vital and Health Statistics, Series 10, No. 84. (Washington, D.C.: Government Printing Office, 1973) p. 15; National Center for Health Statistics, Prevalence of Chronic Skin and Musculoskeletal Conditions, United States-1969, Vital and Health Statistics, Series 10, No. 92, (Washington, D.C.: DHEW, 1974); National Center for Health Statistics, Prevalence of Selected Chronic Digestive Conditions: United States, 1975, Vital and Health Statistics, Series 10, No. 123. (Washington, D.C.: DHEW, 1979) . 42. Health United States: 1979, p. 127. 43. Health Resources Administration, Health of the Disadvantaged: Chartbook, p. 33; California Raza Health Plan, pp. 35 and 41. 44. Jose Oscar Alers, Puerto Ricans and Health: Findings from New York City (Bronx, NY: Fordham University Hispanic Research Center, 1978). 45. Robert E. Roberts and Eun Sul Lee, "The Health of Mexican Americans: Evidence from the Human Population Laboratory Studies," American Journal of Public Health 70 (April 1980) pp. 375-384. 46. David Mechanic, Medical Sociology, Second Edition (New York: Free Press, 1978) pp. 273-287. 47. Sandra Green, Eva Salber, and Jacob J. Feldman, "Distribution of Illness and Its Implications in a Rural Community," Medical Care 14 (December 1978) p. 872. 48. Health United States: 1979, pp. 37-38. 49. Ibid., pp. 37-39. 50. Marjorie Cantor and Mary Mayer, "Health and the Inner City Elderly," The Gerontologist 16 (1976) p. 19; Deborah Newquist et al., Prescription for Neglect: Experience of Older Blacks and
65 Mexican-Americans with the American Health Care System, Final Report on Administration on Aging Grant No. 90-A-1297 (Los Angeles, CA: Andrus Gerontology Center, 1979); J. J. Dowd and V. L. Benston, "Aging in Minority Populations: An Examination of the Double Jeopardy Hypothesis," Journal of Gerontology 33 (1978) pp. 417-436; Markides et al., pp. 8-14. 51. Ethel Shanas, National Survey of the Aged, Final Report of Social Security Administration Contract No. 10P-57823, (no date); Newguist et al.; Cantor and Mayer, p. 19. 52. Dorothy Rice and Kathleen M. Danchik, "Changing Needs of Children: Disease, Disability and Access to Care," Presented at the Annual Meeting of the Institute of Medicine, October 25, 1979. 53. Rice and Danchik, Table 7; Health United States: 1979, p. 37. 54. Health United States: 1979, p. 37. 55. Ibid., p. 9. 56. Rudov and Santangelo, p. 87. 57. Select Panel for the Promotion of Child Health, Better Health for Our Children: A National Strategy, 4 vols., (Washington, D.C.: Government Printing Office, 1980). 58. Health United States: 1979, p. 187. 59. Eva S. Salber et al., "Access to Health Care in a Southern Rural Community," Medical Care 14 (December 1976) pp. 971-986. 60. Aday, Andersen, and Fleming, p. 106. 61. Diana Dutton, "Children's Health Care: The Myth of Equal Access," Background paper prepared for the Select Panel for the Promotion of Child Health, July 1980. 62. Ibid., Tables 7 and 8. 63. Ibid. 64. National Center for Health Statistics, Office Visits by Women. The National Ambulatory Medical Care Survey, United States, 1977, Vital and Health Statistics, Series 13, No. 45. (Washington, D.C.: DHEW, 1980) p. 10. 65. Ibid. 66. National Center for Health Statistics, unpublished tables from the 1978 Health Interview Survey. 67. Alers, pp. 31-55; Weaver, p. 63. 68. Research evidence leaves some doubt about the exact role that prenatal care may play in influencing infant mortality figures. Although groups that receive little or no prenatal care have higher infant mortality rates, questions of causation remain. Low birth weight is generally acknowledged to be the major risk factor for infant mortality, and the higher incidence of low birth weight infants among blacks is amply documented and is true regardless of age, marital status, and receipt of prenatal care (NCHS, Series 21, Number 37). Furthermore, evidence that prenatal care influences birth weight is ambiguous, at best. The incidence of low birth weight hardly varied no matter when, in the course of pregnancy, prenatal care was undertaken (NCHS, Series 21, No. 37, 1980: 16). That is, there is little difference in rates of low birth weight between women who
66 initiated prenatal care in the first or second month of pregnancy and women who initiated prenatal care in the seventh to ninth months of pregnancy. This is true regardless of factors such as race, age, and marital status. However, the incidence of low birth weight infants is greatly elevated among women who receive no prenatal care (a category that includes disproportionate numbers of blacks), although causation is not clear. In many cases it appears that both the low birth weight and the lack of prenatal care are a result of the prematurity of the birth (NCHS, Series 31, No. 33: 21). 69. National Center for Health Statistics, Prenatal Care; United States, 1969-1975, Vital and Health Statistics, Series 31, No. 33, (Washington, D.C.: DREW, 1978). 70. Ibid., pp. 14-17. 71. Ibid., p. 14. 72. Ibid., p. 19. 73. Mary Grace Kovar, "Mortality of Black Infants in the United States," Phylon 38 (December 1977) pp. 370-397. 74. Aday, Andersen, and Fleming, p. 118. 75. Antonio S. Medina, "Hispanic Reproductive Health in California, 1976-77," Paper presented at annual meeting of the American Public Health Association, 1979. 76. The "use-disability" ratio is: â?â X 100, D where P is the mean number of physician visits for those with one or more disability days, and D is the mean number of disability days for those with one or more disability days. The "symptoms-response" ratio reflects a degree of medical judgment about the need for medical care in the presence of various symptoms. The actual ratio is: A-E IT X 100, where A is the actual number of persons who contact a doctor at least once for given symptoms, and E is the estimate of a panel of physicians of the number of persons who should contact a doctor for those symptoms (Aday, Andersen, and Fleming, 1980, pp. 186, 192-193). 77. Problems with the symptoms-response ratio include its disregard of preventive care, its dependence on respondent recognition and reports of symptoms, and its disregard of the severity of and interaction among symptoms. See Dutton, p. 22; Ronald Andersen, "Health Status Indices and Access to Medical Care," American Journal of Public Health 68 (May 1978) p. 461; Martin K. Chen, "Comment on 'Health Status Indices and Access to Medical Care,' " American Journal of Public Health 68 (October 1978)
67 p. 1027. Some believe that these problems lead to overestimates of the poor's use of medical care in relation to symptoms. Some of the same problems (for example, the disregard of prevention) and analagous problems (there may be class or ethnic differences in what is defined as disability) also apply to the use-disability ratio. The problems with these measures are acknowledged and discussed in reports of the most recent research (Aday, Andersen, and Fleming, 1980, pp. 192-196). 78. Andersen, p. 461; Aday, Andersen, and Fleming, p. 195. 79. Aday, Andersen, and Fleming, p. 197. 80. Andersen; Dutton, p. 24; Aday, Andersen and Fleming, p. 983. 81. National Center for Health Statistics, Use of Selected Medical Procedures Associated with Preventive Care: United Statesâ1973, Vital and Health Statistics, Series 10, No. 110. (Washington, D.C.: DHEW, 1977) p. 27. 82. Aday, Andersen, and Fleming, p. 116. 83. Roberts and Lee. 84. Aday, Andersen, and Fleming, p. 76. 85. Roberts and Lee. 86. Salber et al., p. 190. 87. Stephen Thacker et al., "Primary Health Care in an Academic Medical Center," American Journal of Public Health 68 (September 1978) p. 69. 88. Mildred A. Morehead, "Evaluating Quality of Medical Care in the Neighborhood Health Center Program of the Office of Economic Opportunity," Medical Care 8 (March-April 1970) p. 130. 89. National Center for Health Statistics, "Advance Data, Health Interview Survey 1974," DHEW Publication No. (PHS) 78-1250, 1978. 90. Aday, Andersen, and Fleming, p. 59. 91. Ibid. 92. Health United States: 1979, p. 7. 93. Aday, Andersen, and Fleming, p. 89. 94. Davis and Schoen; Aday, Andersen, and Fleming; Rudov and Santangelo; E. A. Skinner et al., "Use of Ambulatory Health Services by the Near Poor," American Journal of Public Health 68 (1978) pp. 1195-1201. 95. S. W. Shannon and G. E. A. Dever, Health Care Delivery: Spatial Perspectives (New York: McGraw Hill, 1974) . 96. Alers. 97. David Elesh and Paul T. Schollaert, "Race and Urban Medicine: Factors Affecting the Distribution of Physicians in Chicago," Journal of Health and Social Behavior 13 (September 1972) pp. 236-250; Robert S. Kaplan and Samuel Leinhardt, "Determinants of Physician Office Location," Medical Care 11 (September 1973) pp. 406-415. 98. Elesh and Schollaert; Guzick and Jahiel, "Distribution of Private Practice Offices of Physicians with Specified Characteristics Among Urban Neighborhoods," Medical Care 14 (June 1976) pp. 419-488. 99. Guzick and Jahiel.
68 100. Alfred E. Miller, "The Changing Structure of the Medical Profession in Urban and Suburban Settings," Social Science and Medicine 11 (1977) pp. 233-243. 101. Pierre Devise, Misused and Misplaced Hospitals and Doctors! A Locational Analysis of the Urban Health Care Crisis (Washington, D.C.: Association of American Geographers, 1973). 102. Louise M. Okada and Gerald Sparer, "Access to Usual Source of Care by Race and Income in Ten Urban Areas," Journal of Community Health 1 (Spring 1976) pp. 163-174. 103. Ed Bradley, "Blacks in America: With All Deliberate Speed?" CBS Television, (July 24, 1979); Karen Davis and Ray Marshall, "Section 1502(1). Primary Health Care Services for Medically Underserved Populations," in Papers on the National Health Guidelines: The Priorities of Section 1502 (Washington, D.C. : DHEW, 1977) pp. 14-23; Aaron Shirley, Presentation before the U.S. Civil Rights Commission on the federal role in rural health care delivery, April 15, 1980, p. 3. 104. Theodore D. Wood, Office of Civilian Health and Medical Programs of the Uniformed Services, Department of Defense, Letter of finding to William Baxter, Administrator, Newnan Hospital, Newnan, Georgia, November 19, 1979. 105. Janet B. Mitchell and Jerry Cromwell, Large Medicaid Practices: Are They Medicaid Mills? Health Care Financing Grants and Contracts Report Series (Washington, D.C.: DHEW, 1980) p. 41. 106. Ibid., pp. 33 and 38. 107. Ibid., p. 43. 108. The Physician Capacity Utilization Surveys: Health Manpower References, DHEW Publication No. (HRA) 79-30, (Washington, D.C.: DHEW, 1979). 109. Davis and Marshall, p. 14. 110. Frank E. Moss, "Through the Medicaid Mills," The Journal of Legal Medicine 5 (May 1977) pp. 6-11, reprinted in Alan D. Spiegel (ed.), The Medicaid Experience (Germantown, MD: Aspen, 1979) pp. 387-392. 111. Institute of Medicine, Assessing Quality in Health Care: An Evaluation (Washington, D.C.: National Academy of Sciences, 1976). 112. Robert H. Brook and Kathleen N. Williams, "Quality of Health Care for the Disadvantaged," Journal of Community Medicine 1 (Winter 1975) pp. 132-156. 113. Rodney M. Coe and H. P. Brehm, "Preventive Health Services and Physician Error," Social Science and Medicine 7 (1973) pp. 303-305. 114. B. A. Lockett and K. N. Williams, Foreign Medical Graduates and Physician Manpower in the United States, DHEW Publication No. (HRA) 74-30 (Washington, D.C.: Government Printing Office, 1974). 115. Mitchell and Cromwell, p. 42. 116. Frank Sloan, Janet Mitchell, and Jerry Cromwell, "Physician Participation in State Medicaid Programs," The Journal of Human Resources 13 (Supplement, 1978) pp. 211-245; Mitchell and Cromwell; F. Kavaler, "Medicaid in New York: Utopianism and Bare
69 Knuckles in Public Health. IV. People, Providers and Payment Telling How It Is," American Journal of Public Health 59 (May 1969) pp. 820-825; Michael W. Jones and Bette Hamburger, "A Survey of Physician Participation in and Dissatisfaction with the Medi-Cal Program," The Western Journal of Medicine 124 (January 1976) pp. 75-83; reprinted in Alan D. Spiegel (ed.), The Medicaid Experience, (Germantown, MD: Aspen, 1979) pp. 277-289. 117. Sloan, Mitchell, and Cromwell, James Studnicki, Robert M. Saywell, Walter Wiechetek, "Foreign Medical Graduates and Maryland Medicaid," New England Journal of Medicine 294 (1976) pp. 1153-1157, reprinted in Alan D. Spiegel (ed.), The Medicaid Experience (Germantown, MD: Aspen, 1979). 118. Kavaler; Alers. 119. Alers. 120. Lawrence S. Bloom et al., "Medicaid in Cook County: Present Status and Future Prospects," Inquiry 5 (June 1968) pp. 13-23. 121. Mitchell and Cromwell, p. 113. 122. Ibid. 123. Aday, Andersen, and Fleming, pp. 157, 160. 124. Ibid. 125. Barbara S. Hulka et al., "Correlates of Satisfaction and Dissatisfaction with Medical Care: A Community Perspective," Medical Care 8 (August 1975) pp. 648-658. 126. Aday, Andersen, and Fleming, pp. 144-148. 127. Ibid., p. 122; Health United States: 1979, pp. 40-42. 128. Health United States: 1979, p. 40; Rudov and Santangelo, p. 239; National Center for Health Statistics, Utilization of Short-Stay Hospitals: Annual Summary for the United States, 1978, Vital and Health Statistics, Series 13, No 46. (Washington, D.C.: DHEW, 1980) p. 46. 129. Rudov and Santangelo, p. 244. 130. National Center for Health Statistics, Utilization of Short-Stay Hospitals: Annual Summary for the United states, 1978, Vital and Health Statistics, Series 13, No. 46. (Washington, D.C.: DHEW, 1980) p. 46. 131. Center for Disease Control, Surgical Sterilization Surveillance: Tubal Sterilization 1970-1975, DHEW Publication No. (CDC) 79-8378 (Atlanta, GA: CDC, 1979) p. 13. 132. Ibid. 133. National Center for Health Statistics, National Survey of Family Growth (Unpublished data). 134. Center for Disease Control, Surgical Sterilization Surveillance: Tubal Sterilization 1970-1975, p. 13. 135. Subcommittee on Health of the Committee on Labor and Public Welfare, United States Senate, Hearings on Quality of Health CareâHuman Experimentation, 1973, Part 4. (Washington, D.C.: Government Printing Office, 1973) . 136. Bradford H. Gray, Human Subjects in Medical Experimentation (New York: Wiley-Interscience, 1975). ~"
70 137. Peter M. Layde et al. "Demographic Trends of Tubal Sterilization in the United States, 1970-1975" (Unpublished) p. 5. 138. Devise. 139. Ibid., p. 25 140. Pierre Devise, "Persistence of Chicago's Dual Hospital System," in Pierre Devise (ed.) Slum Medicine: Chicago's Apartheid Health System (Chicago: Community and Family Study Center, University of Chicago, 1969) p. 19. 141. Ibid., p.24. 142. Richard L. Merrill, Robert S. Earickson, and Philip Rees, "Factors Influencing Distance Traveled to Hospitals," Economic Geography 46 (April 1970) pp. 161-170. 143. R. L. Bashur, G. W. Shannon, and C. M. Metzner,"Some Ecological Differentials in the Use of Medical Services, "Health Services Research 6 (Spring 1971). 144. Ibid., p. 69. 145. General Counsel of DHEW et al., Brief submitted in the matters of: Mercy Hospital, Southern Baptist Hospital, Hotel Dieu Hospital et al., DHEW Docket Numbers 78-VI-7, 78-VI-8, and 78-VI-9 (1978). 146. Numerous indices or summary measures of spatial or organizational segregation have been proposed and evaluated, especially within the sociological literature. These measures have often been applied as indices of the extent to which a residential or school system population is racially segregated. Most widely used has been the "index of dissimilarity," which provides a measure of the extent to which a "system" as a whole is segregated (or integrated) based upon the extent to which the racial composition of each unit of measurement (school, census tract, hospital, etc.) deviates from a theoretical norm at which the percent black and percent white of each unit would be the same as the overall percentage of each group within the system as a whole. For interpretative purposes, this index "D" can also be considered as the percentage of the minority (or non-minority) population that would have to be redistributed were each unit were to have the identical racial makeup. See Otis and Beverly Duncan, "A Methodological Analysis of Segregation Indices," American Sociological Review 20 (April 1955) pp. 210-217. 147. Raymond S. Duff and August B. Hollingshead, Sickness and Society (New York: Harper and Row, 1968). 148. Lawrence D. Egbert and Ilene L. Rothman. "Relation Between the Race and Economic Status of Patients and Who Performs Their Surgery," New England Journal of Medicine 297 (July 14, 1977) pp. 89-91. 149. Joseph A. Flaherty and Robert Meagher, "Measuring Racial Bias in Inpatient Treatment," American Journal of Psychiatry 137 (June 1980) pp. 679-682. 150. Anna M. Jackson, Hershel Berkowitz, and Gordon K. Farley, "Race as a Variable Affecting the Treatment Involvement of Children," Journal of the American Academy of Child Psychiatry 13 (1974) pp. 20-31.
71 151. Stanley Sue, "Community Mental Health Services to Minority Groups," American Psychologist (August 1977) pp. 616-624. 152. Ibid., pp. 620-621. 153. Clayton T. Shaw, "A Detailed Examination of Treatment Procedures of Whites and Blacks in Hospitals," Social Science and Medicine 5 (1971) pp. 251-256. 154. David D. Rutstein et al., "Measuring the Quality of Medical Care: A Clinical Method," New England Journal of Medicine 294 (March 11, 1976) pp. 582-588. 155. Hypertension Detection and Follow-up Program Cooperative Group, "Five-Year Findings of the Hypertension Detection and Follow-up Program. I. Reduction in Mortality of Persons With High Blood Pressure, Including Mild Hypertension. II. Mortality by Race-Sex and Age," Journal of the American Medical Association 242 (December 1979) pp. 2562-2577. 156. Institute of Medicine, Public Policy Options for Better Dental Health, (Washington, D.C.: National Academy Press, 1980). 157. Ibid. 158. Roberts and Lee, p. 266. 159. Joanna Kravits and John Schneider, "The Relationship of Attitudes to Discretionary Physician and Dentist Use by Race and Income" in Ronald Anderson, Joanna Kravits, and Odin Anderson (eds.), Equity in Health Services: Empirical Analyses in Social Policy (Cambridge, MA: Ballinger, 1975) pp. 169-187. 160. Eva J. Salber et al., "Utilization of Services for Preventable Disease: A Case Study of Dental Care in a Southern Rural Area of the United States," International Journal of Epidemiology 7 (1978) pp. 163-173; Eva J. Salber et al., "Access to Health Care in a Southern Rural Community," Medical Care 14 (December 1976) p. 983. 161. Eva J. Salber and Angel1 Beza, "The Health Interview Survey and Minority Health," Medical Care 18 (March 1980) pp. 319-326; Dorothy P. Rice, Thomas F. Drury, and Robert H. Mugge, "Household Health Interviews and Minority Health," Medical Care 18 (March 1980) pp. 327-335. 162. Ivie and Newman.