Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 183
--> 6 Black-White Differences in the Use of Medical Care by the Elderly: A Contemporary Analysis José J. Escarce and Frank W. Puffer Introduction Reduced access to medical care among black Americans, as compared with whites, has concerned researchers and policy makers for many years. Although racial differences in access to care are important at all stages of the life cycle, such differences are especially significant when they occur among the elderly. Older persons bear a high burden of chornic disease and may be particularly vulnerable to the deleterious effects of impaired access to care. Elderly blacks have poorer health status and higher mortality rates than elderly whites (Manton et al., 1987; Gibson 1994), and the difference in life expectancy between blacks and whites who reach age 65 is increasing (Manton et al., 1987). Socioeconomic and racial differences in the elderly's use of medical care were sizable prior to the enactment of the Medicare program in 1965. For instance, in 1964, white elders had 20 percent higher rates of physician visits than black elders (Davis et al., 1981; Long and Settle, 1984), and whites were nearly twice as likely as blacks to be hospitalized (Long and Settle, 1984). By providing insurance coverage for hospital and physician services to the elderly, the Medicare program substantially improved access to medical care among previously underserved populations. Racial differences in physician visit rates largely disappeared by the late 1970s (Davis et al., 1981; Long and Settle, 1984). Indeed, some studies suggest that older blacks currently see physicians more often than older whites (Kleinman et al., 1981; Wan, 1982; Wolinsky et al., 1989; Furner, 1993). Results for inpatient hospital utilization are mixed; the most recent studies find a higher probability of hospitalization and more hospital nights among
OCR for page 184
--> blacks (Wolinsky et al., 1989), and earlier studies report persistently higher use of hospitals by whites (Wan, 1982; Long and Settle, 1984). Nonetheless, all studies agree that racial differences in hospital utilization have narrowed markedly during the past three decades. Two caveats are important in interpreting these encouraging trends. First, studies comparing rates of physician visits and hospitalization for black and white elders either have not adjusted for racial differences in health status (Davis et al., 1981; Wolinsky et al., 1989) or have used limited measures of physical health in the adjustment (Kleinman et al. 1981; Wan, 1982; Long and Settle, 1984; Furner, 1993). However, black-white differences in the health status of the elderly encompass multiple and varied dimensions of health. Older blacks not only have higher mortality rates than older whites; they also have higher rates of many common and frequently disabling chronic conditions, such as hypertension, diabetes, stroke, circulatory disease, end-stage kidney disease, arthritis and other musculoskeletal impairments, open-angle glaucoma, and certain cancers (Manton et al., 1987; Haywood, 1990; Anderson and Felson, 1988; Polednak, 1989; Leske and Rosenthal, 1979; Byrne et al., 1994). Not unexpectedly, blacks suffer from much more disability and functional impairment than whites (Manton et al., 1987). Black elders also have higher rates of mental and nervous disorders than whites (Manton et al., 1987; Polednak, 1989). Adjusting for racial differences in health status, therefore, requires more comprehensive health status measures than are generally employed. Failure to take a comprehensive approach may result in overlooking persistent and clinically important racial differences in medical care utilization. Second, most of the studies focus on racial differences in the quantity of medical care received by older persons (e.g., numbers of physician visits and hospital nights) and do not address differences in the type or quality of care. Recent research has documented racial disparities in important qualitative aspects of medical care utilization. For instance, whereas white elders are more likely than black elders to obtain their regular ambulatory care from private physicians, blacks are more likely than whites to use neighborhood health centers, hospital outpatient departments, or emergency rooms (Wan, 1982; Kotranski et al., 1987). These practice settings are characterized by long waiting times, less satisfactory patient-physician relationships, and less continuity of care than private physicians' offices (Petchers and Milligan, 1988; Dutton, 1985). In addition, elderly blacks are less likely than elderly whites to receive a wide array of specialized or high-technology medical services, including coronary angiography, angioplasty, and bypass surgery; carotid angiography and endarterectomy; cataract extraction; glaucoma surgery; hip and knee replacement; kidney transplantation; and magnetic resonance imaging (Wenneker and Epstein, 1989; Ford et al., 1989; Escarce et al., 1993; Ayanian et al., 1993; Kjellstrand, 1988; Javitt et al., 1991; Held et al., 1988; Oddone et al., 1993). Compared with white elders, hospitalized black elders receive worse processes of care, have
OCR for page 185
--> more medical instability at discharge, and are less likely to receive follow-up physician care (Kahn et al., 1994; Moy and Hogan, 1993). Blacks also receive less appropriate cancer care than whites (McWhorter and Mayer, 1987; Diehr et al., 1989; Mayer and McWhorter, 1989). Perhaps not surprisingly, blacks are less satisfied than whites with many aspects of their care (Blendon et al., 1989). In this study, we examine the current status of black-white differences in the use of medical care among elderly Americans. We focus mainly on racial differences in the quantity of medical care, although some of our analyses address qualitative aspects of care. Our study goes beyond most previous research on the elderly in two ways. First, in addition to physician visits and inpatient hospital utilization, we examine racial differences in total medical care expenditures, which serve as a rough index of the total amount of medical care received. Second, using multivariate methods, we assess the independent effect of race on utilization from two alternative perspectives. Under a ''demand-based" perspective, we adjust for all other available variables that may influence the demand for medical care, including demographic and socioeconomic characteristics, health status, and attitudes and beliefs about the efficacy of care. The demand-based analyses, therefore, address whether black and white elders of similar socioeconomic status and in similar health use the same quantity of medical care. Under a "need-based" perspective, we adjust for variables that affect individuals' need for medical care, conceptualized as health status, but exclude most demographic and socioeconomic characteristics. Thus, the need-based analyses address whether black and white elders in similar health have the same medical care utilization irrespective of socioeconomic factors. By comparing the results of the two analyses, we assess whether the current allocation of medical care services between older blacks and whites mainly reflects observed racial differences in medical need or continues to be substantially influenced by socioeconomic variables. The measures of health status used in the study are multidimensional and comprehensive. Data And Methods Data and Study Sample The source of data for this study is the Household Survey component of the 1987 National Medical Expenditure Survey conducted by the Agency for Health Care Policy and Research. The sample for the Household Survey was a stratified multistage area probability sample of about 35,000 individuals in 14,000 households, and excludes residents of nursing or personal care homes or facilities for the mentally retarded. Several population groups of particular policy interest, including blacks and the elderly, were oversampled (Cohen et al., 1991). Each family in the Household Survey received a core interview four times over a period of 16 months to obtain information about each family member's
OCR for page 186
--> health, use of medical care, medical care expenditures, and insurance coverage in 1987 (Edwards and Berlin, 1989). In addition, detailed information about health status and access to medical care was gathered from special supplementary questionnaires that were administered once during the year. This study uses data from the core interviews, the health status supplement, and the access supplement. The Household Survey included 5,725 individuals who were 65 years or older in 1987. Of these, 5,573 individuals were white (n = 4,828) or black (n = 745) and substantially completed the core interviews. These individuals constitute the sample used in the descriptive analyses of utilization and demographic and socioeconomic characteristics (see below), which thus are representative of noninstitutionalized white and black elderly individuals in the United States (institutionalized elders are excluded owing to the design of the Household Survey). In addition, 4,957 individuals (4,324 whites and 633 blacks) also provided a basic set of responses to the health status supplement. These individuals constitute the sample used in the descriptive analyses of health status and attitudes. Finally, 4,288 (3,815 whites and 473 blacks) of the individuals who responded to the health status supplement answered all the questions needed to construct the complete set of variables for the multivariate analyses. Elders in the complete sample of 5,573 differed from those who responded to the health status supplement and provided enough answers for inclusion in the multivariate analyses. Specifically, elders in the complete sample were older (mean age, 74.0 vs. 73.6), less healthy (mean disability days, 25.3 vs 23.0), more likely to be black (13% vs. 11%), and more likely not have completed grade school (19% vs. 16%). Elders in the complete sample also had higher mean total medical care expenditures ($4,397 vs. $3,924) and hospital nights (2.91 vs. 2.34), although their mean number of contacts with physicians was slightly lower (6.6 vs. 6.7). Since the health status supplement was largely self-administered, non-respondents were more likely than respondents to be cognitively impaired. Therefore, the descriptive analyses of health status and attitudes and the multivariate analyses are representative of white and black noninstitutionalized elders who do not suffer from severe cognitive impairment. Variables In this study, we make use of variables measuring utilization of medical care, demographic and socioeconomic characteristics, health status, and attitudes and beliefs about the efficacy of medical care. Utilization of Medical Care We examine three main quantitative measures of medical care utilization during 1987: total medical care expenditures, physician visits, and inpatient hospital nights. Total medical care expenditures are defined as the sum of all expen-
OCR for page 187
--> ditures for medical care, irrespective of source of payment, including facility and professional fees and expenditures on pharmaceuticals and medical equipment. Physician visits are defined as face-to-face contacts with physicians in physicians' offices, group practices, or clinics; neighborhood or family health centers; free-standing surgical centers; hospital outpatient departments; emergency rooms; or patients' homes. Inpatient hospital nights are defined as nights in acute care hospitals, excluding long-term care facilities. Expenditures, physician visits, and hospital nights have frequently been used in studies of access to or demand for medical care (Blendon et al., 1989; Manning et al., 1981, 1982, 1987; Wolinsky et al., 1989; Stoller, 1982; Evashwick et al., 1984; Wan 1982). In addition, we examine telephone contacts with physicians and contacts with nonphysician providers, with the latter defined as face-to-face contacts in any setting with nonphysician medical care providers such as chiropractors, optometrists, podiatrists, audiologists, physician assistants, nurse practitioners, physical or occupational therapists, social workers, or home health aides. Finally, we examine several qualitative measures of utilization, including the site or setting of physician visits; whether individuals have a usual source of care; the type of usual source, if any; whether individuals usually see a particular physician at their usual source of care; and the specialty of this physician, if any. Demographic and Socioeconomic Characteristics The key demographic variable in the study is race, categorized as white or black according to individuals' self-identification. Additional demographic and socioeconomic characteristics used in the analyses are age, sex, employment status, educational attainment, family income, marital status, family size, insurance coverage, and location of residence. Age is categorized as 65 to 69, 70 to 74, 75 to 79, 80 to 84, or 85 and older. Educational attainment is categorized as no high school, some high school, high school graduate, or college graduate. Family income is categorized as poor, near poor (100% to 125% of poverty), low income (125% to 200%), middle income (200% to 400%), or high income (> 400%). Marital status is categorized as married; widowed more than 1 years; divorced or separated more than 1 year; widowed, divorced, or separated within the past year; or never married. Insurance coverage is categorized as Medicare only (this category also includes a small number of patients with only private insurance or Medicaid), Medicare plus Medicaid (or other public program), Medicare plus private supplementary insurance, or uninsured. Location of residence is categorized as a large metropolitan area (one of the 19 largest Metropolitan Statistical Areas), a small metropolitan area (any other Metropolitan Statistical Area), or a nonmetropolitan area. A substantial body of literature confirms the association of these variables with utilization of medical care (Blendon et al., 1989; Manning et al., 1982; Wolinsky et al., 1989; Dor and Holahan, 1990; Arling, 1985; Stoller, 1982; Evashwick et al., 1984).
OCR for page 188
--> Health Status Assessing the independent effect of race on the utilization of medical care requires comprehensive adjustment for differences in health status. The World Health Organization (1948) defines health as "a state of complete physical, mental, and social well-being and not merely the absence of disease or infirmity." This definition implies that comprehensive measures of health status are multidimensional, with distinct physical, mental, and social components (Manning et al., 1982). Further, whenever possible, the range of measured values for each component should extend beyond the absence of illness to include the degree to which positive states of health are enjoyed. Studies of access to or demand for medical care that use multivariate analyses have found that health status is the most important determinant of medical care utilization (Manning et al., 1982; Evashwick et al., 1984; Wolinsky et al., 1989; Wan, 1982; Arling, 1985). However, most such studies of the elderly have used limited measures of health status. In this study, we adopt the framework developed by Manning et al. (1982), in which measures of health status include a measure of general health in addition to measures of physical, mental, and social health. This approach has been found to substantially enhance the explanatory power of regression models for the use of medical care (Manning et al., 1982; Arling, 1985). The measure of general health status used in the study was individuals' self-rating of their current health as excellent, good, fair, or poor. Four variables were used to measure physical health: (1) a count of limitations on functional status, which assesses limits in physical activities such as walking, running, and lifting (range, 0 to 7); (2) a count of chronic conditions, which assesses the burden of chronic diseases such as diabetes, emphysema, heart disease, and arthritis (0 to 11); (3) a count of acute symptoms experienced during the preceding 30 days, such as weight loss, fatigue, abdominal pain, or shortness of breath (0 to 9); and (4) the number of disability days during the study year, defined as days in which illness or injury caused the individual to stay in bed more than half the day or otherwise cut down on usual activities. In addition, we determined whether individuals were current smokers. Although this is not a direct measure of physical health, smoking has multiple harmful effects on physical health. Mental health was measured with two variables: (1) degree of emotional instability, measured by summing the scores on three ordinal items that assess manifestations of depression and anxiety (range, 3 to 18), and (2) degree of positive emotional states, measured by summing the scores on two items that assess feelings of calmness and happiness (2 to 12). The measure of social health was the sum of four ordinal items that assess participation in voluntary organizations (e.g., church, clubs, lodges) and the frequency of social interactions with family and friends (4 to 28).
OCR for page 189
--> Attitudes and Beliefs Assessing the independent effect of race on the utilization of medical care also requires adjusting for differences in attitudes and beliefs regarding the efficacy of medical care. An important component of these beliefs is individuals' perceived "health locus of control," a construct from social learning theory that refers to whether health and health outcomes are under individuals' personal control, the control of medical care providers, or chance (Rotter, 1966; Lau and Ware, 1981; Marshall et al., 1990). In this study, beliefs regarding health locus of control were measured with two variables: (1) degree of belief in the efficacy of self-care for health problems, measured by summing the scores on three ordinal items that assess attitudes toward the relative benefits of self-care versus physician care (range, 3 to 15), and (2) degree of belief in luck as an important factor in health outcomes (1 to 5). Prior studies have found such variables to be associated with the utilization of medical care (Manning et al., 1982; Stoller, 1982). Statistical Analysis We conducted bivariate analyses to examine differences between black and white elders with regard to quantitative and qualitative measures of medical care utilization, demographic and socioeconomic characteristics, health status, and attitudes and beliefs. Quantitative utilization measures were adjusted for age and sex by using direct standardization (Kleinbaum et al., 1982). Statistical significance was assessed through the use of t tests for continuous variables and chisquare tests for proportions. To determine the independent effect of race on medical care utilization, we conducted multivariate regression analyses with total medical care expenditures, physician visits, and inpatient hospital nights as dependent variables. Our primary analyses employed a demand-based perspective. In these analyses, the explanatory variables included all demographic and socioeconomic characteristics and measures of health status and attitudes and beliefs described in the preceding section. Employment status may influence the demand for medical care through its effect on the cost of care in terms of time spent. Education is expected to influence the demand for care because better educated individuals may be more efficient producers of health (Grossman, 1972). Marital status and family size can affect the demand for care because family members may encourage older persons to seek care for symptoms that otherwise would be ignored or, alternatively, may serve as substitutes for formal medical care. Insurance coverage is expected to influence the demand for care through its effect on the out-of-pocket price of care. Metropolitan versus nonmetropolitan residence may be a proxy for certain nonmonetary costs of care (e.g., travel time). Measures of health status are expected to influence the demand for care because they reflect individuals' stock of health capital (Grossman, 1972). Finally, attitudes and
OCR for page 190
--> beliefs are expected to affect the demand for care because they shape individuals' preferences. The explanatory variables used in our demand-based analyses are similar to those used in other studies of the demand for medical care (Manning et al., 1981, 1982; Wedig, 1988). In addition, we performed secondary analyses employing a need-based perspective. These analyses included as explanatory variables only age, sex, and the measures of health status and attitudes and beliefs; other demographic and socioeconomic characteristics were excluded. Multivariate analyses were based on the two-part model of medical care utilization (Manning et al., 1981, 1987; Duan et al., 1982). The first part of the two-part model is an equation for whether or not an individual has non-zero medical care expenditures (or physician visits or hospital nights) during the year and is estimated by using logistic regression. This equation, which separates users of medical care from nonusers, assesses the factors that influence the decision to spend on care. From a statistical viewpoint, it also deals with the fact that an appreciable proportion of the population does not use any medical care during a year. The second part of the two-part model is an equation for the logarithm of medical care expenditures (or physician visits or hospital nights) conditional on non-zero expenditures and is estimated by using ordinary least squares. This equation assesses the factors that affect the level of use among those who use care. The logarithmic transformation of the dependent variable deals with the marked skewness of the distribution of medical care expenditures and use, and results in more robust and efficient estimates (Manning et al., 1981, 1987; Duan et al., 1982). Standard errors were obtained by using White's (1980) hetero-skedasticity-consistent covariance matrix estimator. All analyses were weighted with weights that reflect both the sample design of the Household Survey and complete and partial survey nonresponse (Cohen et al., 1991). A p value of .10 or less was chosen as the criterion for statistical significance. Results Bivariate Analyses Elderly whites constituted 91.3 percent of the weighted study sample and elderly blacks constituted 8.7 percent. Table 6-1 compares the demographic and socioeconomic characteristics of elderly whites and blacks. Whites and blacks were similar in age and sex distribution and had similar rates of labor force participation. However, whites had more education and higher incomes than blacks and were more likely than blacks to be married, although blacks lived in larger families. Insurance coverage differed substantially for white and black elders. Specifically, blacks were much more likely than whites to have Medicare only—that is, without either public or private supplementary coverage—and to
OCR for page 191
--> TABLE 6-1 Demographic and Socioeconomic Characteristics Variable Whites Blacks Age category (%) 65-69 33.6 36.3 70-74 26.5 25.3 75-79 20.1 18.6 80-84 11.3 11.6 ≥ 85 8.5 8.2 Sex (%) Male 41.4 40.2 Female 58.6 59.8 Employment status (%) Employed 10.4 11.4 Unemployed 89.6 88.6 Educational attainment (%)a No high school 14.2 45.6 Some high school 32.7 30.9 High school graduate 43.2 21.0 College graduate 10.0 2.6 Income category (%) Poor 9.8 33.6 Near poor 7.4 12.0 Low income 20.4 22.4 Middle income 36.1 25.0 High income 26.3 7.0 Marital status (%)a Married 54.8 42.1 Widowed > 1 year 32.7 41.4 Divorced or separated > 1 year 5.6 10.6 Widowed or divorced < 1 year 2.3 1.9 Never married 4.5 4.0 Family size (mean)a 1.86 2.25 Insurance coverage (%)a Medicare only 11.4 26.7 Medicare plus Medicaid 7.6 31.9 Medicare plus private 83.4 44.7 Uninsured 0.6 2.2 Location of residence (%)a Large metropolitan area 24.4 34.5 Small metropolitan area 48.0 39.1 Nonmetropolitan area 27.6 26.4 a p < .01 for test of difference between whites and blacks for numbers in this category. Notes: The weighted sample consisted of 5,088 whites and 485 blacks; the sample was weighted to reflect the sample design and complete and partial nonresponse. Percentages for insurance coverage add to more than 100 because some individuals had both Medicaid and private supplementary insurance in addition to Medicare.
OCR for page 192
--> have Medicaid or other public coverage in addition to Medicare. In contrast, whites were much more likely than blacks to have private supplementary insurance in addition to Medicare. Blacks were more likely than whites to live in large metropolitan areas, but the proportion of individuals residing in nonmetropolitan areas was similar for both races. As Table 6-2 shows, older whites perceived themselves to be in much better general health than older blacks. For instance, 9.3 percent of elderly whites assessed their current health as excellent, whereas only 5.5 percent of elderly blacks did so. Conversely, 11.0 percent of whites assessed their current health as poor, compared with 16.4 percent of blacks. White elders also had more favorable indicators of physical, mental, and social health than black elders. Compared with blacks, whites had fewer functional status limitations, chronic condi- TABLE 6-2 Health Status and Attitudes and Beliefs Variable Whites Blacks Health Status General health (%)a Excellent 9.3 5.5 Good 44.4 31.7 Fair 35.3 46.4 Poor 11.0 16.4 Physical health (mean) Functional status limitationsa 2.88 3.81 Chronic conditionsb 2.32 2.45 Acute symptoms in preceding 30 daysa 1.50 1.73 Disability daysa 22.89 32.49 Mental health (mean) Emotional instabilitya 6.18 6.93 Positive emotional statesa 8.39 8.04 Social health (mean) Social activitiesc 15.03 14.56 Current smoker (%) Yes 14.9 16.7 No 85.1 83.3 Attitudes and beliefs Health locus of control (mean) Efficacy of self-care 7.56 7.43 Luck as a factor in healtha 1.94 2.16 a p < .01 for test of difference between whites and blacks for numbers in this category. b p < .10 for test of difference between whites and blacks for numbers in this category. c p < .05 for test of difference between whites and blacks for numbers in this category.
OCR for page 193
--> tions, acute symptoms during the preceding 30 days, and disability days. Whites also had lower degrees of emotional instability and higher degrees of positive emotional states, and were more involved in social activities. Similar proportions of white and black elders were current smokers. Elderly whites and blacks were similar in their beliefs regarding the efficacy of self-care (vs. physician care) for health problems, but blacks had stronger beliefs than whites regarding the importance of luck as a factor in health outcomes. Descriptive analyses of quantitative measures of utilization uncovered interesting patterns, as Table 6-3 shows. Mean total medical care expenditures for white and black elders were not statistically significantly different. However, this finding masks racial differences in the probability of spending on medical care and in the level of use among those who use care. Thus, whereas whites were more likely than blacks to have non-zero expenditures (p < .01), among individuals who spent on care blacks had higher mean expenditures than whites (p < .10). The findings for physician visits were similar to those for total medical care expenditures. Specifically, although elderly whites and blacks had similar rates of physician visits, whites were more likely than blacks to have at least one TABLE 6-3 Medical Care Utilization Variable Whites Blacks Total medical care expenditures Mean, all individuals $4,236 $4,764 Percentage of individuals with non-zero expenditures 94.6a 88.2a Mean, individuals with non-zero expenditures $4,477b $5,401b Hospital nights Mean, all individuals 2.74 3.33 Percentage of individuals with non-zero nights 20.3 21.4 Mean, individuals with non-zero nights 13.49 15.54 Physician visits Mean, all individuals 6.64 6.41 Percentage of individuals with non-zero visits 87.9a 81.9a Mean, individuals with non-zero visits 7.55 7.83 a p < .01 for test of difference between whites and blacks. b p < .10 for test of difference between whites and blacks. Note: Adjusted for age and sex using direct standardization.
OCR for page 199
--> degree of emotional instability, was associated with more hospital nights. A stronger belief in the efficacy of self-care for health problems was associated with lower medical care expenditures and fewer physician visits and hospital nights. We also conducted secondary analyses, employing a need-based perspective, which adjusted only for age, sex, health status, and attitudes and beliefs. Table 6-8 reports the estimated coefficients of the black race indicator variable in these analyses and compares them with the estimated coefficients from the demand-based analyses (shown in Tables 6-6 and 6-7). Race had statistically significant and important independent effects on medical care expenditures and physician visits in the need-based analyses. As shown in Table 6-8, elderly whites were more likely than elderly blacks to spend on medical care after adjustment for the other explanatory variables, and the racial difference in adjusted probabilities of spending on care (whites, 94.9%; blacks, 85.0%) was larger than the difference in unadjusted probabilities (shown in Table 6-3). Moreover, among individuals who spent on care, adjusted medical care expenditures were 11.5 percent higher for whites than for blacks. Similarly, white elders were more likely than black elders to see a physician, and the racial difference in adjusted probabilities of having at least one physician visit (whites, 88.2%; blacks, 77.8%) was nearly twice as large as the difference in unadjusted probabilities (shown in Table 6-3). Among individuals who saw a physician, whites made 4.4 percent more physician visits than blacks. Race did not have a statistically significant impact on the utilization of inpatient hospital care. Finally, because the effect of race on the utilization of medical care may differ among population subgroups, we repeated our demand-based and need-based analyses after including interaction terms between race and sex, race and income, and race and self-rated health in the regression models. The only statistically significant interactions in these analyses were between race and self-rated health in the equations for the number of hospital nights among elders who were hospitalized. Specifically, hospitalized older blacks who reported excellent health spent fewer nights in the hospital than their white counterparts, whereas there were no differences in the level of hospital use between hospitalized older blacks and whites who reported good, fair, or poor health. Owing to small cell sizes, however, the lack of significant interactions should be interpreted with caution. Discussion This study indicates that racial differences in the quantity of medical care received by elderly persons in the United States have largely disappeared. In particular, bivariate analyses found that in 1987, white and black elders had similar mean annual total medical care expenditures, physician visits, and inpa-
OCR for page 200
--> TABLE 6-7 Regression Results on Expenditures (or Physician Visits or Hospital Nights) on Non-zero Use During the Year: Part II of a Two-Part Model Using a Demand-Based Perspectivea Variableb Level of Medical Care Expenditures Number of Physician Visits Number of Hospital Nights Demographic and socioeconomic characteristics Black -0.032 (0.80) 0.001 (0.03) 0.068 (1.22) Age 70-74 -0.016 (0.65) 0.004 (0.25) 0.077c (1.78) Age 75-79 0.031 (1.09) 0.021 (1.17) 0.036 (0.78) Age 80-84 0.067c (1.78) 0.001 (0.05) 0.124d (2.36) Age ≥ 85 -0.001 (0.01) -0.047 (1.60) 0.113 (1.53) Female -0.079e (3.39) -0.028c (1.94) -0.051 (1.44) Employed -0.003 (0.08) -0.007 (0.33) -0.011 (0.18) Some high school 0.052 (1.50) 0.008 (0.39) 0.063 (1.30) High school graduate 0.148e (4.22) 0.044d (2.08) 0.081c (1.70) College graduate 0.200e (4.41) 0.055c (1.77) 0.014 (0.21) Near poor 0.124d (2.41) -0.013 (0.42) 0.073 (1.08) Low income 0.118e (2.87) 0.013 (0.51) 0.009 (0.15) Middle income 0.118e (2.84) 0.021 (0.84) 0.049 (0.75) High income 0.180e (4.15) 0.056d (2.10) 0.035 (0.54) Widowed > 1 year -0.0002 (0.01) -0.017 (0.86) 0.058 (1.24) Divorced > 1 year 0.084 (1.52) 0.024 (0.84) 0.069 (0.82) Separated > 1 year -0.057 (0.54) -0.135d (2.02) -0.068 (0.40) Widowed or divorced < 1 year -0.017 (0.26) -0.041 (0.98) 0.104 (1.02) Never married -0.010 (0.16) -0.001 (0.04) 0.121 (1.47) Family sizef -0.189d (2.48) -0.111d (2.43) 0.025 (0.23) Medicare plus Medicaid 0.171e (4.23) 0.040c (1.77) -0.028 (0.49) Medicare plus private 0.077d (2.38) 0.044d (2.30) 0.052 (1.14) Uninsured -0.240 (0.75) -0.069 (0.62) 0.504e (4.75)
OCR for page 201
--> Variableb Level of Medical Care Expenditures Number of Physician Visits Number of Hospital Nights Small metropolitan area -0.041 (1.58) -0.014 (0.89) -0.089d (2.12) Nonmetropolitan area -0.094e (3.22) -0.016 (0.93) -0.164e (3.61) Health status Good general health 0.035 (0.90) 0.035 (1.42) -0.074 (0.87) Fair general health 0.086c (1.86) 0.053c (1.88) -0.058 (0.61) Poor general health 0.146d (2.36) 0.094d (2.52) 0.030 (0.27) Functional limitationsg 0.260e (5.54) 0.117e (4.11) 0.145c (1.82) Chronic conditions 0.060e (8.49) 0.038e (8.43) -0.014 (1.36) Acute symptomsg -0.027 (0.56) 0.056c (1.87) -0.0001 (0.001) Disability daysf 0.297e (19.99) 0.129e (14.43) 0.114e (5.66) Emotional instability 0.007 (1.29) 0.001 (0.42) 0.019d (2.28) Positive emotional states -0.005 (0.80) -0.001 (0.34) 0.007 (0.57) Social activities -0.002 (0.79) 0.003c (1.83) -0.002 (0.69) Current smoker -0.090e (3.00) -0.056e (2.90) 0.056 (1.09) Attitudes and beliefs Efficacy of self-care -0.024e (6.80) -0.014e (6.32) -0.018e (2.89) Luck as a factor in health 0.002 (0.23) 0.005 (0.75) 0.029c (1.95) Intercept 2.579e (21.21) 0.438e (5.95) 0.530d (2.47) R2 0.30 0.22 0.18 N 4,055 3,782 792 a The logarithmic equation for the second part of the model is estimated using ordinary least squares regression; t ratios are in parentheses. b The omitted category for age is 65 to 69; for educational level, no high school; for income category, poor; for marital status, married; for insurance coverage, Medicare only; for location of residence, large metropolitan area; and for general health, excellent. c p < .10. d p < .05. e p < .01. f In the regressions, the variable was replaced by its natural logarithm to reduce the effect of skewness. g In the regressions, the variable was replaced by the natural logarithm of 1 plus the variable to reduce the effect of skewness.
OCR for page 202
--> tient hospital nights, although whites were slightly more likely than blacks to spend on medical care and to see a physician during the year. These data generally agree with other recent studies of medical care utilization by the elderly (Kleinman et al., 1981; Wolinsky et al., 1989; Wan, 1982; Long and Settle, 1984; Furner, 1993). Interpreting descriptive findings, however, requires careful consideration of differences between older whites and blacks in demographic and socioeconomic characteristics, health status, and attitudes and beliefs regarding medical care. Consistent with prior research (Manton et al., 1987; Polednak, 1989), we found that elderly blacks fared worse than elderly whites on multiple and varied indicators of health status, including individuals' self-rating of their current health as well as measures of physical, mental, and social health. Worse health would be expected to lead to higher use of medical care among blacks. On the other hand, compared with black elders, white elders had more education, higher incomes, and much higher rates of private insurance coverage to supplement Medicare, all of which could result in higher utilization among whites. We assessed the independent effect of race on the utilization of medical care using multivariate regression analysis. Our primary analyses employed a demand-based perspective that adjusted for all available variables that may influence the demand for care. The explanatory variables in these analyses thus included a full complement of demographic and socioeconomic characteristics, comprehensive measures of health status, and measures of attitudes and beliefs about the efficacy of medical care. These analyses found that race generally did not have statistically significant independent effects on medical care expenditures, physician visits, or hospital nights. The only exception was that black elders were slightly less likely than white elders to spend on medical care after adjustment for other demand factors, but the racial disparity in the adjusted probabilities of non-zero expenditures was very small. The finding that race has little independent influence on the demand for medical care among the elderly, however, does not necessarily imply that the current allocation of medical care services between older whites and older blacks is socially desirable. Researchers on the equity of access to and delivery of medical care have long maintained that an equitable and desirable allocation of medical care services occurs when these services are distributed among individuals according to their ''need" for care, which is usually conceptualized as health status (Aday and Andersen, 1981; Andersen and Newman, 1973; Aday, 1975; Culyer et al., 1992; van Doorslaer and Wagstaff, 1992). Policy makers in many other industrialized countries also accept this view (van Doorslaer and Wagstaff, 1992). But allocation of services according to the demand for care may fall short of this criterion. In particular, features of the medical care delivery system may result in individuals of high socioeconomic status receiving more services than their lower socioeconomic status counterparts. To address this issue, we performed additional multivariate analyses em-
OCR for page 203
--> TABLE 6-8 Medical Care Expenditures, Physician Visits, and Hospital Nights: Coefficients of Black Race in Multivariate Regression Analyses Employing Need-Based and Demand-Based Perspectivesa Part of modelb Utilization Measure and Perspective Part I Part II Total medical care expenditures Need-based perspective -1.227c (5.59) -0.109c (3.09) Demand-based perspectivee -0.419d (1.73) -0.032 (0.80) Physician visits Need-based perspective -0.704c (3.87) -0.044d (1.93) Demand-based perspectivee -0.178 (0.89) 0.001 (0.03) Hospital nights Need-based perspective -0.182 (1.07) 0.016 (0.31) Demand-based perspectivee -0.077 (0.42) 0.068 (1.22) a Parentheses contain t ratios. b Part I of the model is an equation for whether an individual has non-zero expenditures (or physician visits, or hospital nights) and is estimated with logistic regression. Part II of the model is an equation for the logarithm of expenditures (or physician visits, or hospital nights) and is estimated with an ordinary least squares regression. The demand-based findings of the model are presented in detail in Tables 6-6 and 6-7. c p < .010. d p < .01. e From Table 6-6 and 6-7. ploying a need-based perspective. The explanatory variables in these analyses included only age, sex, and the measures of health status and attitudes and beliefs; educational attainment, income, insurance coverage, and indicators of family structure were excluded. Thus, the need-based analyses addressed the question of whether older whites and blacks of similar health status have the same medical care utilization irrespective of socioeconomic factors. The results of the need-based analyses differed strikingly from the findings of the demand-based analyses. Specifically, the need-based analyses revealed that, with adjustment made for the other explanatory variables, elderly blacks were considerably less likely than elderly whites to spend on medical care or to see physicians. Moreover, among users of medical care, black elders had lower adjusted medical care expenditures than white elders, and among individuals who saw a physician, blacks had fewer physician visits than whites. These findings indicate that elderly blacks, as compared with whites, received less
OCR for page 204
--> medical care and saw physicians less often than would be expected given their health status. Why are white and black older persons in equal need of medical care treated differently? Comparison of our demand-based and need-based analyses suggests that a large portion of the racial discrepancy in the use of medical care found in the need-based analyses is due to racial differences in socioeconomic variables. Socioeconomic status may influence the allocation of medical care among elderly Americans through several mechanisms. Socioeconomic status is positively associated with elders' likelihood of having private insurance to supplement Medicare (Rice and McCall, 1985; Nelson et al., 1989). Private supplementary coverage may influence patient behavior by reducing the out-of-pocket price of care and may influence provider behavior through more generous reimbursement. Other financial and nonfinancial barriers to care may also differ across socioeconomic groups. For instance, individuals of low socioeconomic status are more likely than those of high socioeconomic status to travel long distances to receive care, rely on public transportation, and reside in areas where medical care providers—especially private physicians—are scarce (LeGrand, 1982; Aday, 1975; Dutton, 1985; Ernst and Yett, 1985). High socioeconomic status may also be associated with increased ability to navigate successfully within the complex health care delivery system in the United States. A potential objection to our analyses is that socioeconomic status, not race, is the important factor and that, consequently, our need-based analyses are misleading. However, there are at least three reasons why it is meaningful to ask whether the allocation of medical care services between elderly blacks and whites corresponds to observed racial differences in need. First, socioeconomic status and race are inextricably linked in American society, and race directly affects educational and economic opportunities through societal mechanisms such as stratification, segregation, and discrimination (Wallace, 1990). Second, it is likely that a portion of the racial disparity in medical care utilization observed in the need-based analyses is due to race per se rather than to socioeconomic status (Wallace, 1990). Physicians may tend to avoid areas with large minority populations when establishing private practices (Ernst and Yett, 1985). Some observers believe that patient race may also have a direct influence on clinical decision making by physicians (Eisenberg, 1979; Maynard et al., 1986; Yergan et al., 1987; Escarce et al., 1993). Third, racial inequalities in the use of medical care have long interested both researchers and policy makers. Our comparison of demand-based and need-based analyses makes a novel contribution to the literature on this topic. Another potential objection to our analyses is that they overlook the full richness and complexity of the relationships among race, socioeconomic status, and health status. A comprehensive model of these relationships, for instance, would explicitly acknowledge the impact of race on socioeconomic opportunities as well as the effects of socioeconomic factors on health and vice versa (Feinstein, 1993; Williams et al., 1994). Such a comprehensive model is best formulated and
OCR for page 205
--> empirically tested using a life-cycle perspective, however, and is beyond the scope of our study and our cross-sectional data. Rather, our study focuses on the more modest goal of assessing the influence of race on medical care utilization, with health status and socioeconomic status being taken as given. Our study also is limited by imperfect measurement of health status, since it is unlikely that even our comprehensive approach captured all of the relevant racial differences in health. But if omitted health status variables are correlated with socioeconomic characteristics, then the difference between our demand-based and need-based analyses may be partly explained by omitted variable bias. To conclude, our study indicates that the evidence that the Medicare program has substantially improved access to medical care for older blacks in the United States is subject to strong qualification. Simple descriptive analyses comparing whites and blacks and multivariate analyses employing a demand-based perspective suggest that racial differences in elders' use of medical care are small. On the other hand, multivariate analyses employing a need-based perspective reveal that elderly blacks and whites in similar health do not receive the same amount of medical care. In particular, blacks spend substantially less on care than would be expected given their health status. Our findings are consistent with those of van Doorslaer and Wagstaff (1992), who performed an international comparison of equity in the delivery of medical care among industrialized nations. Their study, which also conceptualized equity as equal treatment for persons with equal medical need, found that the United States was characterized by inequity, with elderly persons of higher socioeconomic status being favored. The findings of our study have two implications. First, the poorer health status of elderly blacks, as compared with whites, may be partially related to inadequate medical care. Older blacks do not receive the quantity of care that would be expected based on their medical need. There are also qualitative differences in care between black and whites. Although the contribution of racial differences in medical care utilization to differences in health may be small relative to the contribution of other influences, such as socioeconomic factors, elimination of the disparity in the use of care would be a positive step. Second, additional policies to efface the relationship between socioeconomic status and the use of medical care in the United States may be helpful. Potential policies include extension of insurance coverage to low-income elderly individuals, incentives or programs to locate private physicians in undeserved areas, and development of clinical practice guidelines to mitigate the undue influence of patient race (or other demographic factors) on clinical decision making by physicians. Two recent policies that may have salutary effects on the use of care by older blacks with Medicare insurance are the extension of public supplementary coverage to a higher proportion of the low-income elderly, which is expected to decrease out-of-pocket payments for this group, and the implementation of the resource-based Medicare fee schedule (Health Care Financing Administration, 1991), which is likely to reduce the difference in physician reimbursement be-
OCR for page 206
--> tween Medicare patients with private and with public supplementary coverage. We must acknowledge, however, that health policy has only limited ability to reverse the effects of societal mechanisms and structures that constrain black Americans' economic opportunities and adversely affect their health. References Aday, L.A. 1975 Economic and noneconomic barriers to the use of needed medical services. Medical Care 13:447-456. Aday, L.A., and R.M. Andersen 1981 Equity of access to medical care: A conceptual and empirical overview. Medical Care 19(suppl.):4-27. Andersen, R., and J.F. Newman 1973 Societal and individual determinants of medical care utilization in the United States. Milbank Memorial Fund Quarterly 51:95-124. Anderson, J.J., and D.T. Felson 1988 Factors associated with osteoarthritis of the knee in the first National Health and Nutrition Examination Survey (HANES I). American Journal of Epidemiology 128:179-189. Arling, G. 1985 Interaction effects in a multivariate model of physician visits by older people. Medical Care 23:361-371. Ayanian, J.Z., S. Udvarhelyi, C.A. Gatsonis, C.L. Pashos, and A.M. Epstein 1993 Racial differences in the use of revascularization procedures after coronary angiography. Journal of the American Medical Association 269:2642-2646. Blendon, R.J., L.H. Aiken, H.E. Freeman, and C.R. Corey 1989 Access to medical care for black and white Americans: A matter of continuing concern. Journal of the American Medical Association 261:278-281. Byrne, C., J. Nedelman, and R.G. Luke 1994 Race, socioeconomic status, and the development of end-stage renal disease. American Journal of Kidney Diseases 23:16-22. Cohen, S.B., R. DiGaetano, and J. Waksberg 1991 Sample Design of the 1987 Household Survey. National Medical Expenditure Survey Methods 3, Agency for Health Care Policy and Research. AHCPR publication 91-0037. Rockville, MD: Public Health Service. Culyer, A.J., E. van Doorslaer, and A. Wagstaff 1992 Comment: Utilization as a measure of equity. Journal of Health Economics 11:43-98. Davis, K., M. Gold, and D. Makuc 1981 Access to health care for the poor: Does the gap remain? Annual Review of Public Health 1:159-182. Diehr, P., J. Yergan, J. Chu, P. Feigl, G. Glaefke, R. Moe, M. Bergner, and J. Rodenbaugh 1989 Treatment modality differences for black and white breast-cancer patients treated in community hospitals. Medical Care 27:942-958. Dor, A., and J. Holahan 1990 Urban-rural differences in Medicare physician expenditures. Inquiry 27:307-318. Duan, N., W.G. Manning, C.N. Morris, and J.P. Newhouse 1982 A Comparison of Alternative Models for the Demand for Medical Care. R-2754-HHS. Santa Monica, CA: RAND Corporation. Dutton, D. 1985 Financial, organizational and professional factors affecting health care utilization. Social Science and Medicine 20:721-735.
OCR for page 207
--> Edwards, W.S., and M. Berlin 1989 Questionnaires and Data Collection Methods for the Household Survey and the Survey of American Indians and Alaska Natives. National Medical Expenditure Survey Methods 3, National Center for Health Services Research and Health Care Technology Assessment. DHHS publication 89-3450. Rockville, MD: Public Health Service. Eisenberg, J.M. 1979 Sociologic influences on decision making by clinicians. Annals of Internal Medicine 90:957-964. Ernst, R.L., and D.E. Yett 1985 Econometric and statistical studies of the geographic distribution of physicians. Pp. 179-226 in Physician Location and Specialty Choice , R.L. Ernst and D.E. Yett, eds. Ann Arbor, MI: Health Administration Press. Escarce, J.J., K.R. Epstein, D.C. Colby, and J.S. Schwartz 1993 Racial differences in the elderly's use of medical procedures and diagnostic tests. American Journal of Public Health 83:948-954. Evashwick, C., G. Rowe, P. Diehr, and L. Branch 1984 Factors explaining the use of health care services by the elderly. Health Services Research 19:357-382. Feinstein, J.S. 1993 The relationship between socioeconomic status and health: A review of the literature. Milbank Quarterly 71:279-322. Ford, E., R. Cooper, A. Castaner, B. Simmons, and M. Mar 1989 Coronary arteriography and coronary bypass surgery among whites and other racial groups relative to hospital-based incidence rates for coronary artery disease: Findings from the NHDS. American Journal of Public Health 79:437-440. Furner, S.E. 1993 Chartbook on health data on older Americans: United States, 1992. Health care use and its cost. Vital and Health Statistics 3(29):21-36. Gibson, R.C. 1994 The age-by-race gap in health and mortality in the older population: A social science research agenda. Gerontologist 34:454-462. Grossman, M. 1972 On the concept of health capital and the demand for health. Journal of Political Economy 80:223-255. Haywood, L.J. 1990 Hypertension in minority populations. American Journal of Medicine 88(suppl. 3B):17S-20S. Health Care Financing Administration, U.S. Department of Health and Human Services 1991 Medicare program: Fee schedule for physicians' services: Final rule. Federal Register 56:59502-59811, November 25. Held, P.J., M.V. Pauly, R.R. Bovbjerg, J. Newmann, and O. Salvatierra 1988 Access to kidney transplantation: Has the United States eliminated income and racial differences? Archives of Internal Medicine 148:2594-2600. Javitt, J.C., A.M. McBean, G.A. Nicholson, J.D. Babish, J.L., Warren, and H. Krakauer 1991 Undertreatment of glaucoma among black Americans. New England Journal of Medicine 325:1418-1422. Kahn, K.L., M.L. Pearson, E.R. Harrison, K.A. Desmond, W.H. Rogers, L.V. Rubenstein, R.H. Brook, and E.B. Keeler 1994 Health care for black and poor hospitalized Medicare patients. Journal of the American Medical Association 271:1169-1174.
OCR for page 208
--> Kjellstrand, C.M. 1988 Age, sex, and race inequality in renal transplantation. Archives of Internal Medicine 148:1305-1309. Kleinbaum, D.G., L.L. Kupper, and H. Morgenstern 1982 Epidemiologic Research: Principles and Quantitative Methods. New York: Van Nostrand Reinhold. Kleinman, J.C., M. Gold, and D. Makuc 1981 Use of ambulatory medical care by the poor: Another look at equity. Medical Care 19:1011-1029. Kotranski, L., J. Bolick, and J. Halbert 1987 Neighborhood variations in the use of city-supported primary health care services by an elderly population. Journal of Community Health 12:231-245. Lau, R.R., and J.E. Ware 1981 Refinements in the measurement of health-specific locus-of-control beliefs. Medical Care 19:1147-1157. LeGrand, J. 1982 The Strategy of Equality: Redistribution and the Social Services . London, England: Allen and Unwin. Leske, M.C., and J. Rosenthal 1979 Epidemiologic aspects of open-angle glaucoma. American Journal of Epidemiology 109:250-272. Long, S.H., and R.F. Settle 1984 Medicare and the disadvantaged elderly: Objectives and outcomes. Milbank Quarterly 62:609-656. Manning, W.G., C.N. Morris, J.P. Newhouse, L.L. Orr, N. Duan, E.B. Keller, A. Leibowitz, K.H. Marquis, M.S. Marquis, and C.E. Phelps 1981 A two-part model of the demand for medical care: Preliminary results from the Health Insurance Study. Pp. 103-123 in Health, Economics, and Health Economics, J. van der Gaag and M. Perlman, eds. Amsterdam, Netherlands: North Holland. Manning, W.G., J.P. Newhouse, N. Duan, E.B. Keeler, A. Leibowitz, and M.S. Marquis 1987 Health insurance and the demand for medical care: Evidence from a randomized experiment. American Economic Review 77:251-278. Manning, W.G., J.P. Newhouse, and J.E. Ware 1982 The status of health in demand estimation: Or, beyond excellent, good, fair, and poor. Pp. 143-184 in Economic Aspects of Health, V.R. Fuchs, ed. Chicago: University of Chicago Press. Manton, K.G., C.H. Patrick, and K.W. Johnson 1987 Health differentials between blacks and whites: Recent trends in mortality and morbidity. Milbank Quarterly 65(suppl. 1):129-199. Marshall, G.N., B.E. Collins, and V.C. Crooks 1990 A comparison of two multidimensional health locus of control instruments. Journal of Personality Assessment 54:181-190. Mayer, W.J., and W.P. McWhorter 1989 Black/white differences in non-treatment of bladder cancer patients and implications for survival. American Journal of Public Health 79:772-774. Maynard, C., L.D. Fisher, E.R. Passamani, and T. Pullum 1986 Blacks in the Coronary Artery Surgery Study (CASS): Race and clinical decision making. American Journal of Public Health 76:1446-1448. McWhorter, W.P., and W.J. Mayer 1987 Black/white differences in type of initial breast cancer treatment and implications for survival. American Journal of Public Health 77:1515-1517.
OCR for page 209
--> Moy, E., and C. Hogan 1993 Access to needed follow-up services: Variations among different Medicare populations. Archives of Internal Medicine 153:1815-1823. Nelson, L., A. Ciemnecki, N. Carlton, and K. Longwell 1989 Assignment and the Participating Physician Program: An Analysis of Beneficiary Awareness, Understanding, and Experience. Background paper 89-1. Washington, DC: Physician Payment Review Commission. Oddone, E.Z., R.D. Horner, M.E. Monger, and D.B. Matchar 1993 Racial variations in the rates of carotid angiography and endarterectomy in patients with stroke and transient ischemic attack. Archives of Internal Medicine 153:2781-2786. Petchers, M.K., and S.E. Milligan 1988 Access to health care in a black urban elderly population. Gerontologist 28:213-217. Polednak, A.P., ed. 1989 Racial and Ethnic Differences in Disease. New York: Oxford University Press. Rice, T., and N. McCall 1985 The extent of ownership and the characteristics of Medicare supplemental policies. Inquiry 22:188-200. Rotter, J.B. 1966 Generalized expectancies for internal versus external control of reinforcement. Psychological Monographs 80:1-28. Stoller, E.P. 1982 Patterns of physician utilization by the elderly: A multivariate analysis. Medical Care 20:1080-1089. van Doorslaer, E., and A. Wagstaff 1992 Equity in the delivery of health care: Some international comparisons. Journal of Health Economics 11:389-411. Wallace, S.P. 1990 The political economy of health care for elderly blacks. International Journal of Health Services 20:665-680. Wan, T.T.H. 1982 Use of health services by the elderly in low-income communities. Milbank Quarterly 60:82-107. Wedig, G.J. 1988 Health status and the demand for health. Journal of Health Economics 7:151-163. Wenneker, M.B., and A.M. Epstein 1989 Racial inequalities in the use of procedures for patients with ischemic heart disease in Massachusetts. Journal of the American Medical Association 261:253-257. White, H. 1980 A heteroskedasticity consistent covariance matrix and a direct test for heteroskedasticity. Econometrica 48:817-838. Williams, D.R., R. Lavizzo-Mourey, and R.C. Warren 1994 The concept of race and health status in America. Public Health Reports 109:26-41. Wolinsky, F.D., B.E. Aguirre, L-J. Fann, V.M. Keith, C.J. Arnold, J.C. Niederhauer, and K. Dietrich 1989 Ethnic differences in the demand for physician and hospital utilization among older adults in major American cities: Conspicuous evidence of considerable inequalities. Milbank Quarterly 67:412-449. World Health Organization 1948 Constitution. Basic Documents. Geneva, Switzerland: World Health Organization. Yergan, J., A.B. Flood, J.P. LoGerfo, and P. Diehr 1987 Relationship between patient race and the intensity of hospitalservices . Medical Care 25:592-603.
Representative terms from entire chapter: