1 k V ' January 1992 United States General Accounting Office Report to Congressional Requesters HISPANIC ACCESS TO HEALTH CARE Significant Gaps Exist GAO/PEMD-92-6 .PUC‘ML‘ - ”7.1 is.;.\.-u_\'_“7. ; U BRfi-‘R‘t' 9 Lc.,~--f$n;-=r‘r or \, (nu-cmu/ GAO United States General Accounting Office L]; (a Washington, DC. 20548 / Program Evaluation and V f ‘ Methodology Division , ,1 a B-245673 January 15, 1992 I, .11.; The Honorable Edward R. Roybal Chairman, Select Committee on Aging House of Representatives The Honorable Solomon P. Ortiz Chairman, Congressional Hispanic Caucus House of Representatives This report contains testimony presented at the joint hearing of the Select Committee on Aging and the Congressional Hispanic Caucus on September 19, 1991. (See appendix 1.) We are publishing the statement as a report to make the information more widely available. The testi- mony responds to your request for information on Hispanic access to health care. For many Americans, the first step in accessing health care is the acqui— sition of health insurance. Hispanics, however, are much less likely than others to have health insurance coverage. Thirty-three percent of all Hispanics were without health insurance in 1989, and this problem was especially acute for the Mexican-American community, where 37 per- cent were uninsured. We found that type of employment and income are key determinants of the high rates of noninsurance among Hispanics. Although the great majority of adult Hispanic workers are employed, they often work in jobs that provide neither private health insurance nor sufficient income to make such insurance affordable to the worker. As a result, of those Hispanic family members under age 65 who are uninsured, nearly 8 out of 10 belong to families that have employed adult workers. Public health insurance—Medicaid—is one potential solution for per- sons who cannot afford private health insurance. However, Hispanics in some states—particularly Mexican-Americans—have difficulty in gaining access to Medicaid because of stringent state eligibility criteria. Our knowledge concerning the actual prevalence of disease among His- panics is limited by a shortage of comprehensive data. Nevertheless, it is clear that the high rate of noninsurance and an apparent scarcity of pri— mary care facilities together make Hispanics particularly vulnerable to adverse health outcomes. Page 1 GAO/PEMD-926 Significant Gaps in Hispanic Access to Health Care 8-245673 Initial steps toward achieving a more rational health care delivery pro- cess for the Hispanic population clearly involve more adequate health insurance coverage (both private and public), expanded neighborhood access to primary care, and major improvements in available data to allow appropriate planning and evaluation. For interested readers, we have included a bibliography of studies on Hispanic access to health care. If you have any questions or would like additional information, please call me at (202) 275-1854 or Robert L. York, Acting Director of Program Evaluation in Human Services Areas, at (202) 275-5885. Other major contributors to this report are listed in appendix II. {lo-m ' Eleanor Chelimsky Assistant Comptroller General Page 2 GAO/PEMD—92-6 Significant Gaps in Hispanic Access to Health Care \ Page 3 GAO/PEMD-92-6 Significant Gaps in Hispanic Access to Health Care Contents Letter 1 Appendix I 6 Statement of Eleanor gaCfgmxnd _ t d Wth H, h R t f N . 1615 Chelimsky, Assistant ac ors ssocia e _ 1 1g a es 0 on1nsurance Among Hispanlcs Comptroller General, Differential Rates of Coverage Between the Various 14 ‘ Hispanic Subgroups Program Evaluatlon Public Health Insurance: Medicare and Medicaid 16 and MethOdOIOgy Data on Hispanic Health 20 D1V1s1on, Before the Our Site Visits 23 H 011 S e S e1 6 Ct Summary and Conclusions 24 Committee on Aging and the Congressional Hispanic Caucus, September 19, 1991 Appendix II 26 Major Contributors to This Report Bibliography 27 Tables Table 1.1: US. Hispanic Population Distribution by State, 9 1989 Table 1.2: Estimates of Types of Health Insurance 10 Coverage by Race/Ethnicity, 1989 Table 1.3: Percent of Family Members Under Age 65 Who 11 Are Uninsured, by Race/Ethnicity and Family Employment Status, 1987 Table 1.4: Health Insurance Status of Family Members 12 Under Age 65, by Race/Ethnicity and Family Employment Status, 1987 Table 1.5: Percent of Racial/Ethnic Population Employed 13 in Industries, by Level of Insurance Coverage, 1987 Page 4 GAO/PEMD-92-6 Significant Gaps in Hispanic Access to Health Care Contents Table 1.6: Insurance Coverage of Those Hispanic Males Aged 16-64 Who Worked Year-Round, Full-Time, by Income Level, 1989 . Table 1.7: Distribution of Type of Insurance for Hispanics, by Subgroup, 1989 Table 1.8: Medicaid Eligibility Criteria for a Family of ‘ Three in Selected States, 1989 14 15 18 Figures Figure 1.1: US. Hispanic Population by Subgroup, 1990 Figure 1.2: Percent of Hispanic Persons Living in Poverty, by Subgroup, 1989 Abbreviations AFDC Aid for Families With Dependent Children HHANl-ZS Hispanic Health and Nutrition Examination Survey NHANES National Health and Nutrition Examination Survey NHIS National Health Interview Survey NMES National Medical Expenditure Survey Page 5 GAO/PEMD-92-6 Significant Gaps in Hispanic Access to Health Care J——s Appendix I Statement of ‘ eanOr Cheljrnsky, Assistant Comptroller General, Before the House Select Committee on Aging and the Congressional Hispanic Caucus, September 19, 1991 It is a pleasure to be here to share with you the results of our work on Hispanic health in the United States. As you requested, we examined the reasons for the high rate of noninsurance among Hispanics. In our testi- mony today, we will present information on (1) the reasons why a higher percentage of Hispanics than non-Hispanics lack health insur- ance, (2) the reasons why the various Hispanic subgroups have differing rates of health insurance coverage, (3) the extent to which Medicaid and Medicare serve the Hispanic population, and (4) the data sources cur- rently available to provide a profile of the health status of Hispanics in the United States. Let me first briefly summarize our major findings: - In 1989, 33 percent of all Hispanics—including 37 percent of Mexican- Americans, 16 percent of Puerto Ricans, and 20 percent of Cuban—Amer- icans—had no health insurance at all. - To a large degree, the high rate of noninsurance among Hispanics is the result of being employed in jobs that do not provide health benefits. If Hispanic workers had the same rate of health insurance coverage as white workers, the overall rate of noninsurance among members of His- panic families would have been 18 percent. - Many Hispanics are concentrated in states with the most stringent Medi- caid eligibility criteria, such as Texas and Florida; they thus have diffi- culty gaining access to this public health insurance program. . Despite recent efforts to collect data on Hispanic health, a comprehen- sive View of the morbidity and mortality trends of the different His- panic subgroups is not available at this time. To address these issues, we reviewed the literature, interviewed experts, Background and examined national sources of data on health insurance coverage. These data sources included the Census Bureau's Current Population Survey, the National Medical Expenditure Survey (NMES), and the His- panic Health and Nutrition Examination Survey (IIHANl-IS), a study by the National Center for Health Statistics. In addition, we Visited govern- ment officials and health care administrators in Texas, New York, and Florida to gain a focused understanding of Hispanic health insurance lssues. Before turning to the results of our work, I first will present a descrip- tion of the Hispanic population in the United States, including both dem- ographic and health profiles of this minority group. Page 6 GAO/PEMD—92-6 Significant Gaps in Hispanic Access to Health Care Appendix I Statement of Eleanor Chelimsky, Assistant Comptroller General, Before the House. Select Committee on Aging and the Congressional Hispanic Caucus, September 19, 1991 Demographic Profile Of The Hispanic population is the second largest—and the fastest Hispanics in the United growing—minority group in the United States. In 1990, an estimated 21 S t 13 million persons living in the United States (8 percent of the total US. a es population) were of Hispanic origin.1 This number represents a 40-per- cent increase since 1980, and according to a Census Bureau projection, the Hispanic population is expected to increase by an additional 200 per- : cent (40 million persons) between 1990 and 2080. The Hispanic popula- tion is also relatively youthful (with a median age of 26 compared with 34 for non-Hispanics). Further, as a result of numerous advances in medicine and public health, gains in life expectancies have been experi- enced by all Americans, including blacks and Hispanics. In fact, since 1980, the Hispanic elderly population has increased by 75 percent. Thus, the Hispanic population has been growing prodigiously in the United States. However, the same cannot be said for the group’s socio- economic standing. More than one in four Hispanic persons (26 percent) lived in poverty in 1989—about the same as in 1980—compared to only one in nine (12 percent) of non-Hispanics. Poverty rates are typically tied to low levels of educational attainment. Among persons aged 25 and over, only 51 percent of Hispanics have completed 4 years of high school or beyond, compared with 80 percent of non-Hispanics. Moreover, only 9 percent of Hispanics have completed 4 years of college, compared with 22 percent of non—Hispanics. The Hispanic population in the United States represents a diverse array of ancestry, culture, socioeconomic conditions, and needs. It is most often categorized as consisting of five main subgroups: (1) Mexican- American, (2) Puerto Rican, (3) Cuban-American, (4) Central and South American, and (5) persons of other Hispanic origin. As figure 1.1 indi- cates, Mexican-Americans are by far the largest of these groups, repre— senting nearly two thirds of the entire Hispanic population of the United States. 1All data presented in this testimony on the Hispanic population in the United States exclude residents of Pucrto Rico. Page 7 GAO/PEMD-92-6 Significant Gaps in Hispanic Access to Health Care .—_— , Appendix I Statement of Eleanor Chelimsky, Assistant Comptroller General, Before the House Select Committee on Aging and the Congressional Hispanic Caucus, September 19, 1991 Figure M: US. Hispanic Population by Subgroup, 1990 5% Cuban-American 7% Other Hispanic Puerto Rican Central and South American Mexican-American Note: Figures do not total 100 because of rounding. Source: US Bureau of the Census, Current Population Survey. 1990 Although Hispanics reside in all parts of the United States, nearly 90 percent of the Hispanic population live in eight states. (See table 1.1.) Moreover, the different Hispanic subgroups tend to inhabit different regions of the country. For instance, almost three fourths of the Mex- ican-American population live in California and Texas (42 percent and 32 percent, respectively). Nearly one half of the Puerto Ricans who reside within the fifty states live in New York (49 percent), with another 12 percent living in New Jersey. Similarly, 58 percent of the Cuban-American population reside in Florida. Page 8 GAO/PEMD-92-6 Significant Gaps in Hispanic Access to Health Care Appendix I Statement of Eleanor Chelimsky, Assistant Comptroller General, Before the House Select Committee on Aging and the Congressional Hispanic Caucus, September 19, 1991 Table l.1: U.S. Hispanic Population Distribution by State, 1989 — Percent of total Hispanic Hispanic State population population California 6,762,000 34 Texas 4,313,000 22 New York 1,982,000 10 Florida 1,586,000 8 Illinois 855,000 4 Arizona 725,000 7 4 New Jersey 638,000 3 New Mexico 549,000 3 Total 17,410,000 88 Source: US. Bureau of the Census, Current Population Survey, 1989 Health Profile of Hispanics in the United States The health profile of the Hispanic population compared to that of non- Hispanics again reflects the socioeconomic disparities discussed previ— ously. The sparse data on Hispanic mortality indicate that while His- panics live about as long as non-Hispanic whites, they tend to die from different causes. Among the major causes of death, accidents, diabetes, cirrhosis of theJLver, and other liver diseases kill proportionally more Hispanics than non-Hispanic whites. Moreover, the top ten killers of His- panics includey'homicide, ‘AIDS, and perinatal conditions, whereas none of these conditions is among the major killers of non-Hispanic whites. Hispanics are more likely than non-Hispanics to suffer from such ail- ments as hypertension, hyperlipidemia, hyperglycemia, cardiopulmo- nary problems, stroke, cirrhosis, obesity, and certain types of cancer. Among selected groups of Hispanics who are heavily addicted to intra- venous drug use, the rising trend of AIDS cases represents a serious con— cern. Hispanics are also two to three times more likely than non- Hispanics to have diabetes, and such complications as blindness and amputation are likely to occur in the absence of treatment. For instance, a study of Texas border counties foundthat of all cases of diabetes— caused blindness identified in the study, 60 percent could have been pre- vented with proper treatment, as could 51 percent of kidney failures and 67 percent of diabetes-related amputations of feet and legs. In View of these data on the morbidity and mortality among Hispanics, their access to the health care system, and especially to health insurance coverage, becomes extremely important. Lack of health insurance is, in the United States, a primary barrier to the receipt of adequate and Page 9 GAO/PEMD—92-6 Significant Gaps in Hispanic Access to Health Care .—— Appendix I Statement of Eleanor Chelimsky, Assistant Comptroller General, Before the House Select Committee on Aging and the Congressional Hispanic Caucus, September 19, 1991 timely health care. For instance, uninsured persons are less likely to have a regular source of health care, less likely to have an ambulatory visit during the year, more likely to use an emergency room or hospital clinic as their usual source of care, and less likely to use preventive ser- vices such as pap smears, blood pressure checks, and breast examina- tions. Moreover, among persons with chronic and serious illnesses, the uninsured make fewer visits to a physician than the insured. One of the primary concerns of health care providers is that when uninsured patients finally receive care, their physical complications are more advanced and therefore more difficult and costly to treat. In 1989, over 33 million Americans (14 percent of the population) were not covered by any type of health insurance (public or private) at any time during the year. Although lack of health insurance is a problem that cuts across every demographic category, it is especially prevalent among Hispanics. For instance, according to the Current Population Survey, 33 percent of Hispanics (over 6 million persons) were uninsured during all or part of 1989, compared with about 19 percent of blacks and about 12 percent of whites. (See table 1.2.) The percentage of His- panics not insured varies substantially across states. In addition, data from the Current Population Survey indicate that about 2.1 million undocumented aliens were living in the United States as of November 1989, with 1.6 million of these persons having been born in Mexico. We do not know the extent to which this population is included in the Cur- rent Population Survey’s estimate that we are using. However, it seems reasonable to assume that these undocumented aliens are undercounted and less likely than other Hispanics to have insurance coverage. Table l.2: Estimates of Types of Health Insurance Coverage by Race/Ethnicity, 1989 Percent of population covereda Type of Insurance White Black Hispanicb Private ”7 78 54 50 Nileidicare 7 W194 10 7—6 M‘eéficaid ’7 ' ""i ”5' 23 15 Other publiclnsurance 4 4 V 3 inflamed ’ ' i3 19 7 ‘35 aFigures do not total 100 because persons may have more than one type of Insurance coverage. bHispanic persons may be of any race Source. US Bureau of the Census, Current Population Survey, 1990 Now let me turn to the results of our study. The first question you asked us to examine was why the Hispanic population has a higher rate of Page 10 GAO/PEMD-92-6 Significant Gaps in Hispanic Access to Health Care Appendix I Statement of Eleanor Chelimsky, Assistant Comptroller General, Before the House Select Committee on Aging and the Congressional Hispanic Caucus, September 19, 1991 Factors Associated With High Rates of Noninsurance Among Hispanics noninsurance for health care than other racial/ethnic groups in the United States. We have looked at this question in terms of several fac- tors: (1) employment, (2) type of employer, and (3) income. Employment Table 1.3 shows that Hispanic families are more likely to be uninsured than either white or black families, and this is true regardless of whether there is an adult worker in the family. While whites are much more likely to be uninsured if there is no adult worker in the family, this is less true for blacks and essentially not true at all for Hispanics. Table L3: Percent of Family Members Under Age 65 Who Are Uninsured, by Race/Ethnicity and Family Employment Status, 1987 Family employment status 7__ White Black Hispanic An adult worker employed __ ,, 13% 24% 35% No adult worker employed 26 29 36 Source National Center for Health Services Research and Health Care Technology Assessment. NMES. Round 1, 1987. Employed persons can be insured either through private insurance—— offered by employers or purchased by the worker—or by public insur- ance. Table 1.4 shows that the big difference between employed persons in the different racial and ethnic groups is in the extent of private insur- ance coverage. In families with adult workers, only 57 percent of His— panics compared with 69 percent of blacks and 84 percent of whites have private insurance coverage. If Hispanic families with adult workers had the same rate of coverage through private and public insur- ance that whites have, the overall rate of noninsurance for persons in Hispanic families would have been 18 instead of 35 percent. (The com- parable figure for persons in white families is 14 percent.) Page 1 l GAO/PEMD-92-6 Significant Gaps in Hispanic Access to Health Care Appendix I Statement of Eleanor Chelimsky, Assistant Comptroller General, Before the House Select Committee on Aging and the Congressional Hispanic Caucus, September 19, 1991 Table La: Health Insurance Status of Family Members Under Age 65, by Race/ Ethnicity and Family Employment Status, 1987‘ Family employment status White Black Hispanic An adult worker employed Any private insurance 84% 69% 57% ‘ r Public insurance 3 7 8 Uninsured 13 24 35 \ No adult worker employed Any private insurance 47 12 10 Public insurance 27 59 54 ' Uninsured 26 29 36 in all families Any private insurance 82 49 47 Public insurance 8 25 18 Uninsured 14 26 35 3Figures may not total 100 because of rounding. Source: National Center for Health Servrces Research and Health Care Technology Assessment, NMES, Round 1. 1987. Notably, 78 percent of Hispanic family members under age 65 who are uninsured are in families with an adult worker. If a way could be found to increase private insurance through employers without causing nega- tive effects on employment, this would be a highly effective single mea- sure for reducing the large number of uninsured persons. Table 1.4 also examines insurance status for families with no adult worker employed. In this situation, coverage by private insurers is much lower, reflecting the fact that most such coverage occurs as an employee benefit. Very few Hispanic families with no adult worker employed have private insurance, which is not the case for whites, and Medicaid and other public insurance make up for only part of this difference. Industry Type of employment is also associated with health insurance coverage in that certain industries are more likely to provide (1) health insurance benefits to employees and/or (2) higher wages with which to purchase insurance. In comparison with whites and blacks, a greater proportion of Hispanics are employed in industries that are less likely to provide health insurance coverage (for example, personal services and agricul- ture). (See table 1.5.) Conversely, Hispanics are less likely than whites or blacks to be employed in industries that routinely provide such coverage Page 12 GAO/PEMD-92-6 Significant Gaps in Hispanic Access to Health Care o Appendix I Statement of Eleanor Chelimsky, Assistant Comptroller General, Before the House Select Conunittee on Aging and the Congressional Hispanic Caucus, September 19, 1991 (for example, manufacturing, professional services, and public administration). Table I5: Percent of Racial/Ethnic Population Employed in Industries, by Level of Insurance Coverage — Percent of workers employeda Category White Black Hispanic Industries with high rates of uninsured employees (30-32%)b 15 12 20 Industries with moderate rates of uninsured employees (21-22%)C 29 23 29 Industries with low rates of uninsured employees (7-1 1%)d 58 65 50 8Figures may not total 100 because of rounding, bPersonal services, construction, agriculture, and entertainment cRepair services and sales dProfessronal services, manufacturing, mining, transportatIon/communication/utilities, financral servuces, and public administration Source: National Medical Expenditure Survey, 1987, and US. Bureau of Labor Statistics, 1991 Income Income is, of course, also strongly related to lack of health insurance coverage in the United States. Among Hispanics, we have seen this reflected, in part, in the higher rate of private insurance coverage among those families that had an adult worker compared with those that did not. An additional linkage of income to insurance can be seen by examining only those persons who were employed year-round, full-time, with incomes above and below the poverty level. Table 1.6 shows that employed Hispanic males with incomes above the poverty level had much higher rates of private insurance than did employed Hispanic males with incomes below the poverty level (67 versus 31 percent). Con- versely, those Hispanic males with lower incomes were twice as likely to be uninsured (64 versus 30 percent). The higher income Hispanic males were probably both more likely to have insurance coverage through their employers and more likely to be able to afford private health insurance when it was not offered by their employers. Page 13 GAO/PEMD-926 Significant Gaps in Hispanic Access to Health Care Appendix I Statement of Eleanor Chelimsky, Assistant Comptroller General, Before the House Select Conunittee on Aging and the Congressional Hispanic Caucus, September 19, 1991 Table I.6: Insurance Coverage of Those Hispanic Males Aged 16-64 Who Worked Year-Round, Full-Time, by Income Level, 1989 — Above poverty line Below poverty line Coverage status Number“ Percent” Number‘ Percentb insured _ Private insurance 2,310 67 81 31 Public insurance 125 4 13 5 Uninsured 1,031 30 169 64 Total 3,466 263 aIn thousands 0Figures may not total 100 because of rounding. Source: US. Bureau of the Census. Current Population Survey, 1990. Experts whom we interviewed told us that it is not uncommon for per— sons to receive employer-related health benefits for themselves but not their families. The problem here is that, because of their low incomes, Hispanic persons are often less able to purchase additional coverage for their families. This problem is further exacerbated by the fact that, on average, Hispanics have larger families than non-Hispanics and, there— fore, more persons for whom to purchase extended coverage. Summary Differential Rates of Coverage Between the Various Hispanic Subgroups While older Hispanics are almost all covered by Medicare or other health insurance, we found that about 35 percent of younger Hispanics are without insurance coverage. About 78 percent of this uninsured group are employed. We found that Hispanics are somewhat more likely to have jobs that are less likely to offer insurance coverage. Also, working Hispanics with low incomes—those below the poverty line—were much more likely to be uninsured than those with higher incomes. This situa- tion probably reflects income differences among employment sectors, differential coverage between occupational levels, and differing capacity to afford the purchase of individual health insurance policies. The disparities that exist between the Hispanic and non-Hispanic popu- lations are also evident within the Hispanic community—a population that is by no means homogeneous. Although the data are not always clear, it is generally believed that more than one third of the Mexican- American population (37 percent) was not insured in 1989 compared with 16 percent of the Puerto Rican and 20 percent of the Cuban-Amer- ican population. As table 1.7 also indicates, Cuban-Americans are more likely than Mexican-Americans and Puerto Ricans to have private health insurance. However, because Puerto Ricans are more than twice as Page 14 GAO/PEMD-92-6 Significant Gaps in Hispanic Access to Health Care o Appendix 1 Statement of Eleanor Chelimsky, Assistant Comptroller General, Before the House Select Committee on Aging and the Congressional Hispanic Caucus, September 19, 1991 likely as the other subgroups to have Medicaid coverage, they have the lowest rate of noninsurance. Table |.7: Distribution of Type of Insurance for Hispanics, by Subgroup, 1989“ Type of Mexican- insurance American Puerto Rican Cuban- American Private 44% 44% 56% Medicare , A 37 7 4 10 fidicaid ' 14 33 V 12 Othermalic 3’ 5 é finsured 7' ' 37 16 ’ 20 aFigures do not total 100 because persons may have more than one type of insurance coverage Source Fernando Trevuno et at, “Health Insurance Coverage and Utilization of Health Servrces by Mex- Ican Americans, Mainland Puerto Ricans, and Cuban Americans." Journal of the American Medical Association. 265:2 (1991), 233737 We found that, as it did for the Hispanic population as a whole, type of employment also plays a role in the health insurance disparities among Hispanic subgroups. In 1990, Mexican-Americans were less likely to hold managerial and professional positions (10 percent) than either Puerto Ricans (16 percent) or Cuban—Americans (23 percent). Moreover, 8 per- cent of Mexican-Americans held farming, forestry, or fishing jobs, com- pared with only 2 percent of Puerto Ricans and 1 percent of Cuban- Americans. It is likely that these differences in occupation have contrib— uted to the disparities in private health insurance coverage. (See table 1.7.) Income disparities are also evident among Hispanic subgroups. In 1989, of all persons with earnings, 44 percent of Mexican-Americans earned less than $10,000, compared with 32 percent of Puerto Ricans and 32 percent of Cuban-Americans. In addition, 28 percent of Mexican-Ameri— cans and 33 percent of Puerto Ricans lived in poverty, compared with 15 percent of Cuban-Americans. (See figure 1.2.) Thus, it is clear that, as a group, Cuban-Americans are more economically advantaged—and thus more able to purchase health insurance—than either Mexican- Americans or Puerto Ricans. Page 15 GAO/PEMD-92-6 Significant Gaps in Hispanic Access to Health Care Appendix I Statement of Eleanor Chelimsky, Assistant Comptroller General, Before the House Select Committee on Aging and the Congressional Hispanic Caucus, September 19, 1991 Figure I.2: Percent of Hispanic Persons Living in Poverty, by Subgroup, 1989 40 Percent 35 30 25 20 15 10 f i i 4, :gi a" ‘ 0: a e? .9 ° s f e5 is? f é, Source' US. Bureau of the Census, Current Population Survey. 1990 Public Health Insurance: Medicare and Medicaid You asked us to address the extent to which Medicare and Medicaid serve the Hispanic population. I will first present our findings for Medi— care, and then discuss Medicaid. Medicare We found that Medicare coverage for the elderly is nearly universal in the United States, with 96 percent of persons aged 65 or over having coverage. The racial/ethnic breakdown is similar: 96 percent of elderly whites, 95 percent of elderly blacks, and 91 percent of elderly Hispanics are covered by Medicare. The reason for this widespread coverage is that Medicare eligibility is relatively straightforward. Persons over the age of 65 who are eligible to receive monthly Social Security benefits are automatically entitled to receive Medicare. However, although coverage is nearly universal, it remains true that 4 percent of elderly Hispanics (about 42,000 people) are not covered by Medicare or any other health insurance. Page 16 GAO/PEMD-92-6 Significant Gaps in Hispanic Access to Health Care 7 ‘ Appendix I Statement of Eleanor Chelimsky, Assistant Comptroller General, Before the House Select Committee on Aging and the Congressional Hispanic Caucus, September 19, 1991 Medicaid In explaining the lack of access to Medicaid among Hispanics, both the literature and the experts we interviewed point to the stringent eligi- bility criteria in a number of states with high concentrations of His— panics. Since eligibility criteria for Medicaid are determined, within federal guidelines, by each state, the criteria vary dramatically across 1 states. Two of the most restrictive states are Texas and Florida, in which about 3 of every 10 Hispanics in the United States reside. There are two broad classes of eligibility under Medicaid: categorically needy and medically needy. Categorically needy persons are generally those who qualify for assistance under the Aid for Families With Depen- dent Children (AFDC) or the Supplemental Security Income programs. They are automatically eligible for Medicaid. At the option of each state, Medicaid eligibility may be extended to medically needy persons, including certain groups (for example, the aged, the blind, families with dependent children, and so on) whose income or resources are in excess of the qualification standards for the categorically needy. Table 1.8 illustrates the eligibility criteria for enrolling in Medicaid through AFDC and the medically needy program in those states in Which most Hispanics reside. For instance, in California, a family of three must earn less than 79.1 percent of the federal poverty-line income to qualify for Medicaid through AFDC. In contrast, a family of three in Texas must earn less than 21.9 percent of the federal poverty-line income to qualify through the same program. Thus, in 1989, a family of three that earned $6,500 (61 percent of the poverty-line income of $10,600) would have qualified for Medicaid through AFDC in California but not in Texas. Page 17 GAO/PEMD-92-6 Significant Gaps in Hispanic Access to Health Care Appendix I Statement of Eleanor Chelimsky, Assistant Comptroller General, Before the House Select Committee on Aging and the Congressional Hispanic Caucus, September 19, 1991 Table l.8: Medicaid Eligibility Criteria for a Family of Three in Selected States, 1989 AFDC eligibility Medically needy eligibility As a percent As a percent State of poverty Flanka of poverty Rankb 4 Arizona fl 35.0 38 if C C ”California 79.1 ’ 1 106.4 “1 Florida 34.2 41 745.7 2?; Illinois 740.8 33 ' 54.6 ’27) New Jersey ' 50.6 17 67.5 12 New York 64.3 7 84.6 6 X New Mexico 31.5 W 44 c " iTexas 21.9 49 31.8 33 aFlank is based on the eliglbility criteria of the 50 states. Higher rank indicates less stringent criteria. bOut of a total of 35 states that have a program for the medically needy CArizona and New Mexico have no medically needy program. Source: National Coalition of Hispanic Health and Human Services Organizations (COSSMHO), . . And Access for All: Medicaid and Hispanics (Washington, DC ' 1990) Under medically needy programs, Medicaid eligibility may be extended to persons whose incomes are in excess of the income limit applicable under the categorically needy program. However, these persons must either “spend down” or accumulate enough medical expenses to deduct from their income to meet eligibility criteria. As shown in table 1.8, the income criteria vary across those states where many Hispanics reside, with California having the least stringent criteria and Arizona and New Mexico not having a medically needy program at all. The differences in eligibility criteria across states largely explain the discrepancy in Medicaid coverage across the Hispanic subgroups. For instance, as I noted previously, Mexican-Americans and Puerto Ricans both have high rates of poverty and low median incomes. However, Puerto Ricans (concentrated in New York and New Jersey) are much more likely than Mexican-Americans (with a substantial population in Texas) to meet the Medicaid eligibility criteria. As a result, a higher pro— portion of Puerto Ricans than Mexican-Americans are eligible for Medicaid. I should point out that, although the greatest proportion of Mexican— Americans (41.6 percent) reside in California, and although California has the least stringent eligibility criteria in the nation, nevertheless, with over 30 percent of the Mexican-American population residing in Texas, the Texas Medicaid policies play a major role in restricting health care coverage for that group. Page 18 GAO/PEMD-92-6 Significant Gaps in Hispanic Access to Health Care Appendix I Statement of Eleanor Chelimsky, Assistant Comptroller General, Before the House Select Committee on Aging and the Congressional Hispanic Caucus, September 19, 1991 It is also the case that making the eligibility criteria less restrictive in order to include all persons below the poverty line would still leave many working people—who earn more than the poverty-line income, but not enough to afford health insurance—without coverage. The situ- ation in California is a perfect example of this hole in the health insur- ance safety net. Despite California’s less restrictive criteria, 23 percent (6 million persons) of the nonelderly population of California were unin- sured in 1989. Only Texas, the District of Columbia, New Mexico, and Oklahoma had higher rates of noninsurance. Being employed in low- wage jobs that do not provide health insurance has been identified as one of the factors that contribute to California’s noninsurance rate. Thus, raising the Medicaid thresholds closer to the poverty line would still leave many working persons uninsured. In sum, Hispanics in general, and Mexican-Americans in particular, have difficulty gaining access to Medicaid because of the stringency in state Medicaid policies. While they suffer from similar economic disadvan- tages, Puerto Ricans are better able to gain access to this public health insurance program by virtue of their place of residence, thus ensuring some measure of health insurance coverage for this population. How- ever, even states with less restrictive Medicaid policies, such as Cali— fornia, have difficulties providing health insurance coverage to the large proportion of people who do not meet the criteria and still cannot afford health insurance. Nonfinancial Barriers to Medicaid Access We discussed the issues involved in gaining access to Medicaid with Medicaid officials in Texas, New York, and Florida. In addition to the low income-eligibility criteria in Texas and Florida noted earlier, the officials cited the complexity of the Medicaid system as a principal bar- rier to Medicaid access. There are numerous avenues to Medicaid enroll- ment that may involve such factors as family income, financial assets, family composition, age of children, and medical need. For instance, in Texas, there are nearly 10 different programs for Medicaid enrollment, each with its own criteria for eligibility (for example, pregnant women with incomes up to 133 percent of the poverty line; children born before January 2, 1982, who are eligible for AFDC; children born before October 1, 1983, with incomes between the AFDC and medically needy criteria; and so on). Officials in Texas told us that the Medicaid caseworkers engage in 4 weeks of training just to learn the eligibility criteria and how to communicate them to potential recipients. Page 19 GAO/PEMD-92-6 Significant Gaps in Hispanic Access to Health Care Appendix I Statement of Eleanor Chelimsky, Assistant Comptroller General, Before the House Select Committee on Aging and the Congressional Hispanic Caucus, September 19, 1991 Data on Hispanic Health Medicaid officials stated that they find it difficult to “market” Medicaid because of its complexity. For instance, officials in Texas said that it is difficult to enroll people in the medically needy program because people do not want to accumulate medical bills before being reimbursed. They also noted that it is difficult to effectively communicate to people that they may not be eligible for Medicaid at the present time, but that they may be eligible in the future (for example, if a woman becomes preg- nant). New York state officials noted that the process of enrolling people is complicated and burdensome, and that standing in line for a full day at the Medicaid office does not compete favorably with the practical alternative of receiving free care in an emergency room or a community health center. Finally, let me turn to the issue of the data that are available, and are currently being collected, and how adequate a profile they provide of the health status of Hispanics in the United States. The health status of any group, including Hispanics, can be examined by looking at data on the mortality and morbidity of that group. Our dis- cussions with experts revealed their general View that there is a lack of comprehensive data to assess the health status and needs of the His- panic population. Nevertheless, they identified the following sources of data that do provide some useful information on the health status of the Hispanic population: (1) the census and the Current Population Survey, (2) vital statistics, (3) Medicare and Medicaid, and (4) special surveys conducted by the Centers for Disease Control and the National Center for Health Statistics, such as the Hispanic Health and Nutrition Exami- nation Survey (HHANES), the National Health and Nutrition Examination Survey (NHANES), the National Medical Expenditure Survey (NMES), and the National Health Interview Survey (NHIS). Census data are necessary for an accurate description of the population, since they are the principal source for determining the total number of people residing in the United States. This information is critical for cal- culating the rates of disease and death in a given population. However, the census has been criticized for undercounting minorities in general and Hispanics in particular. Further, the census is a decennial count and cannot provide the accurate, up-to—date information needed to keep abreast of conditions in a growing, highly mobile segment of the popula- tion. This limitation is partly overcome by a monthly survey, called the Current Population Survey, which investigates about 60,000 households. Page 20 GAO/PEMD-92-6 Significant Gaps in Hispanic Access to Health Care Appendix I Statement of Eleanor Chelimsky, Assistant Comptroller General, Before the House Select Committee on Aging and the Congressional Hispanic Caucus, September 19, 1991 The Current Population Survey depends on a sampling procedure. How- ever, because Hispanics are a small proportion of that sample, subgroup data often represent a very small number of actual interviews. As a result, such information often remains unpublished because the sample sizes are too small to provide reliable estimates for Hispanic subgroups. Mortality Statistics Mortality statistics provide a readily available indicator of the fre— quency of occurrence of diseases leading to death in the population. They are routinely collected through the vital statistics registration system of the National Center for Health Statistics. As of 1988, only 30 states included an Hispanic identifier on their death certificate form. However, even among those states, poor reporting on the Hispanic origin item on the death certificate precludes the use of vital statistics for ana- lytical purposes or for a general health assessment of the Hispanic population. Morbidity Statistics Morbidity statistics are another source for determining the health status of the population. There are two primary ways such information is col- lected: (l) administrative records of Medicare and Medicaid, and (2) spe- cial surveys. Medicare and Medicaid require providers of health care to submit infor- mation on the morbidity of the population groups they serve for pro- gram management and reimbursement purposes. These data bases provide information on episodes of individual patient care such as the patient’s demographic information, diagnosis, procedures performed, and charges. To the extent that such files contain information on the use and content of health services, they are useful in developing health profiles. However, they frequently lack critical patient level informa- tion. There is often an absence of validation of many data elements. In addition, while coverage of the elderly in the Medicare program is nearly universal, the Medicaid data base is limited to those individuals who meet the eligibility criteria. These data cannot provide health profiles of those who are not covered. Moreover, data on Hispanic origin are not reliably reported by all states. Thus, their use may actually be misleading, and may point to erroneous conclusions about the health status of the Hispanic population. Special Surveys In recent years, several surveys have been conducted to gather informa- tion on the health status of the Hispanic population. The Hispanic Page 21 GAO/PEMD-92-6 Significant Gaps in Hispanic Access to Health Care Appendix I Statement of Eleanor Chelimsky, Assistant Comptroller General, Before the House Select Committee on Aging and the Congressional Hispanic Caucus, September 19, 1991 Health and Nutrition Examination Survey (HHANES) is probably the most comprehensive survey of the status of Hispanic health in the United States. Conducted for the first time by the National Center for Health Statistics from 1982 through 1984, HllANES provides information on demographics, economic conditions, health insurance coverage, health services use and satisfaction, acculturation, and assessments of health. However, there has been no follow-up to the original survey. The National Health and Nutrition Examination Survey (NIIANES), con- ducted every 10 years by the National Center for Health Statistics, pro- vides comprehensive information on the health and nutrition status of the US. population that includes extensive medical and nutritional examinations. The sample sizes of Hispanic respondents, however, were never large enough to provide separate estimates for the Hispanic popu- lation as a whole, or for the subgroups. The National Medical Expenditure Survey (NMBS) is conducted every 8 to 10 years by what is now known as the Agency for Health Care Policy and Research, with the most recent survey having been conducted in 1987. It collects information on health services utilization, expenditures, insurance coverage, and estimates of persons with functional disabilities and impairments. NMES oversamples the Hispanic population, allowing . for national estimates of Hispanics, as well as of the Mexican—American, Puerto Rican, and Cuban-American subgroups. The National Health Interview Survey (NHIS) is conducted annually by the National Center for Health Statistics and contains information about the prevalence, distribution, and effects of illness and disability in the United States. The most recent NHIS, conducted in 1989, oversampled Mexican—Americans, allowing for more precise national estimates con- cerning this Hispanic subgroup. It did not, however, oversample the remaining subgroups. The Centers for Disease Control collect and maintain records on abor- tion, AIDS, congenital anomalies, rubella, nosocomial infection, tubercu- losis, and other conditions that may have a preventable component. As with the collection and recording of vital statistics, the responsibility for reporting disease is legally vested in the individual states. The list of diseases that must be reported and exact procedures for reporting vary somewhat from state to state. While there is some underreporting by physicians of certain diseases that carry a social stigma, such data are important for an understanding of community health. For example, the Page 22 GAO/PEMD-92-6 Significant Gaps in Hispanic Access to Health Care Appendix I Statement of Eleanor Chelimsky, Assistant Comptroller General, Before the House Select Committee on Aging and the Congressional Hispanic Caucus, September 19, 1991 Our Site Visits Centers for Disease Control data have been very useful in determining the incidence of AIDS in various groups, including Hispanics. In summary, no existing data base currently provides accurate, com- plete, and available data on the entire Hispanic population, including subgroups, residing in the United States. As noted previously, we visited El Paso, New York City, and Miami, to develop a more in-depth knowledge of the issues involved in health insurance coverage. In each of these cities, we met with officials from the local Medicaid offices, as well as officials from a public hospital and a community health center. Although officials from each city noted con- cerns particular to that location, there was a high degree of consensus regarding the health care needs of Hispanics and barriers to health care for this population. All stated that a primary barrier was the shortage of physicians serving Hispanic communities. This problem was especially acute in El Paso, where only 30 of the city’s 800 physicians (4 percent) maintain practices in the poorest part of the city—an area that houses 170,000 people (32 percent of the El Paso population). This shortage of physicians is also related to the dearth of primary care facilities available to the Hispanic community. Again, this problem was especially acute in El Paso, where there are only two federally funded community health centers to serve the entire county. This shortage of primary care facilities results in a situation in which patients go to the community health centers or to the hospital emergency rooms in advanced stages of illness. This, in turn, makes treatment both more dif— ficult and more expensive. Of particular concern is the high rate of dia- betes among Hispanics, as well as the secondary complications that arise because of delayed treatment. Officials also noted the high degree of complications for women and children because of inadequate prenatal care. Finally, I should note the special health concerns of the colonias in Texas. Colonias are unincorporated subdivisions on the outskirts of El Paso, along the Mexican border. The colonias present a special health concern because many inhabitants of these communities do not have running water, sewers, or septic systems. Hand-dug wells may supply water to the families, but the wells are often too shallow and too close to outhouses. In addition, it is not uncommon for families to dump their waste into nearby irrigation ditches. As a result, families in the colonias Page 23 GAO/PEMD-92-6 Significant Gaps in Hispanic Access to Health Care Appendix I Statement of Eleanor Chelimsky, Assistant Comptroller General, Before the House Select Committee on Aging and the Congressional Hispanic Caucus, September 19, 1991 are at a high risk of infection from drinking water that they or their neighbors have contaminated. In summary, despite the vast differences between the Hispanic commu- nities in Texas, New York, and Florida, we found that they share many similar health problems. Of particular concern is the shortage of health care providers who offer primary care to Hispanics. All the officials we interviewed noted that timely and adequate care in the early stages of sickness and disease would not only alleviate many medical problems but would also reduce the fiscal problems currently draining public health care systems. Health care delayed exacerbates both medical problems and their costs. Summary and Conclusions Based on our review of the literature, interviews with experts in the area of Hispanic health, and site visits to locations with large concentra- tions of Hispanics, we have reported several critical findings. First, many Hispanics, and particularly Mexican-Americans and Puerto Ricans, do not have private health insurance coverage because they are employed in jobs that do not provide such coverage and because their incomes are too low to allow them to purchase private health insurance coverage. If Hispanic workers had the same rate of health insurance coverage as white workers, the overall rate of noninsurance among members of Hispanic families would have been 18 instead of 35 percent in 1989. Second, because many Hispanics are concentrated in states with strin- gent Medicaid eligibility criteria, they have difficulty gaining access to this public health-insurance program. This problem is particularly acute for Mexican—Americans, who experience approximately the same eco- nomic disadvantages as Puerto Ricans but whose rate of Medicaid cov- erage is much lower. Third, despite recent efforts to collect data on Hispanic health, experts agree that the health status of Hispanics, especially Hispanic subgroups, is imprecisely known and has thus far been insufficiently analyzed. As a result, a comprehensive View of the morbidity and mortality trends of different Hispanic subgroups is not available at this time. The high rate of noninsurance among Hispanics—and the paucity of data on Hispanic health—are especially troublesome in light of the information about Hispanics that i_s available. It is known, for instance, Page 24 GAO/PEMD-92-6 Signifith Gaps in Hispanic Access to Health Care Appendix I Statement of Eleanor Chelimsky, Assistant Comptroller General, Before the House Select Committee on Aging and the Congressional Hispanic Caucus, September 19, 1991 that Hispanics experience a high degree of morbidity. Experts agree that diseases such as type II diabetes—an illness that strikes a younger age group and often leads to secondary complications such as blindness and amputation—can be ameliorated if treatment is accessible and timely. Initial steps toward achieving a more rational health care delivery pro- cess for the Hispanic population clearly involve more adequate health insurance coverage (both private and public), expanded neighborhood access to primary care, better sanitation in specific locations, and major improvements in available data to allow appropriate planning and evaluation. Mr. Chairmen, this concludes my remarks. I would be happy to answer any questions you may have. Page 25 GAO/PEMD-92-6 Significant Gaps in Hispanic Access to Health Care Appendix 11 Major Contributors to This Report - Patrick G. Grasso, Assistant Director Program Evaluatlon Sushil K. Sharma, Assistant Director and Methodology Alan D. Stein-Seroussi, Project Manager Division Page 26 (iAO/PEMD—QZ-fi Significant Gaps in Hispanic Access to Health Care Bibliography Andersen, Ronald M., et a1. Ambulatory Care and Insurance Coverage in an Era of Constraint. Chicago: Pluribus Press, 1987. Andersen, Ronald M., Aida L. Giachello, and Lu Ann Aday. “Access of Hispanics to Health Care and Cuts in Services: A State-of-the-Art Over- view.” Public Health Reports, 101:3 (1986), 238-52. Andersen, Ronald, et al. “Access to Medical Care Among the Hispanic Population of the Southwestern United States.” Journal of Health and Social Behavior, 22 (March 1981), 78-89. Andrews, Jane. Poverty and Poor Health Among Elderly Hispanic Amer- icans. Baltimore, Md.: The Commonwealth Fund Commission on Elderly People Living Alone, September 1989. Becerra, Jose E., et a1. “Infant Mortality Among Hispanics: A Portrait of Heterogeneity.” Journal of the American Medical Association, 265:2 (1991), 217-21. Brown, E. Richard, et al. Health Insurance Coverage of Californians in 1989. Berkeley, Calif .: California Policy Seminar, University of Cali- fornia, 1991. Council on Scientific Affairs. “Hispanic Health in the United States." Journal of the American Medical Association, 265:2 (1991), 248-52. De La Rosa, Mario. “Health Care Needs of Hispanic Americans and the Responsiveness of the Health Care System.” Health and Social Work, 14:2 (1989), 104-13. Delgado, Jane L., et al. “Hispanic Health and Nutrition Examination Survey: Methodological Considerations.” American Journal of Public Health, Supplement, 80 (1990), 6-10. Espino, David V., and David Maldonado. “Hypertension and Accultura- tion in Elderly Mexican Americans: Results From 1982-1984 Hispanic HANES." Journal of Gerontology, 45:6 (1990), M209-13. Estrada, Antonio L., Fernando M. Trevino, and Laura A. Ray. “Health Care Utilization Barriers Among Mexican Americans: Evidence From IIIIANES 1982-1984." American Journal of Public Health, Supplement, 80 (1991),27-31. l Page 27 GAO/PEMD-92—6 Significant Gaps in Hispanic Access to Health Care Bibliography Freeman, Howard E., et al. “Uninsured Working-Age Adults: Character— istics and Consequences.” Health Services Research, 24:6 (1990), 811-23. Giachello, Aida L., et a1. “Uses of the 1980 Census for Hispanic Health Services Research.” American Journal of Public Health, 73:3 (1983), 266-73. Geronimus, Arline T., Lisa J. Neidert, and John Bound. “A Note on the Measurement of Hypertension in IIHANES.” American Journal of Public Health, 80:12 (1990), 1437—42. Ginzberg, Eli. “Access to Health Care for Hispanics.” Journal of the American Medical Association, 265:2 (1991), 238-41. Lacayo, Carmela G., and Jean K. Crawford. A National Study to Assess the Service Needs of the Hispanic Elderly, Final Report. Los Angeles, Calif.: Asociacion Nacional Pro Personas Mayores, December 1980. Lewin/ICF. Health Care For Hispanic Individuals. Washington, DC: Health Resources and Services Administration, 1991. Markides, Kyriakos S., and Jeannine Coreil. “The Health of Hispanics in the Southwestern United States: An Epidemiological Paradox.” Public Health Reports, 101:3 (1986), 253-65. Mendoza, Fernando S., et al. “Selected Measures of Health Status for Mexican-American, Mainland Puerto Rican, and Cuban-American Chil- dren.” Journal of American Medical Association, 265:2 (1991), 227-32. National Coalition of Hispanic Health and Human Services Organizations (COSSMHO). ...And Access for All: Medicaid and Hispanics. Washington, DC: 1990. Schur, Claudia L., Amy B. Bernstein, and Marc L. Berk. “The Importance of Distinguishing Hispanic Subpopulations in the Use of Medical Care." Medical Care, 25:7 (1987), 627-41. Short, Pamela F. Estimates of the Uninsured Population, Calendar Year 1987. Rockville, Md.: Public Health Service, 1990. Short, Pamela F., Llewellyn J. Cornelius, and Donald E. Goldstone. “Health Insurance of Minorities in the United States.” Journal of Health Care for the Poor and Underserved, 1:1 (1990), 9-24. Page 28 GAO/PEMD-92-6 Significant Gaps in Hispanic Access to Health Care Bibliography Short, Pamela F., Alan C. Monheit, and Karen Beauregard. A Profile of Uninsured Americans. Rockville, Md.: Public Health Service, 1989. Solis, Julia M., et a1. “Acculturation, Access to Care, and Use of Preven- tive Services by Hispanics: Findings from HHANES 1982-1984." American Journal of Public Health, Supplement, 80 (1990), 11-19. Stroup-Benham, Christine A., and Fernando M. Trevino. “Reproductive Characteristics of Mexican-American, Mainland Puerto Rican, and Cuban—American Women.” Journal of American Medical Association, 265:2 (1991), 222-26. Trevino, Fernando M., J. G. Bruhn, and Harold Bunce. “Utilization of Community Mental Health Services in a Texas-Mexico Border City. Social Science and Medicine, 13a (1979), 331-34. Trevino, Fernando M., and Abigail J. Moss. ”Health Insurance Coverage and Physician Visits Among Hispanic and Non-Hispanic People.“ Health, United States. Washington, DC: National Center for Health Statistics, 1983. Trevino, Fernando M., et a1. ”Health Insurance Coverage and Utilization of Health Services by Mexican Americans, Mainland Puerto Ricans, and Cuban Americans.“ Journal of the American Medical Association, 265:2 (1991), 233—37. US. Bureau of the Census. The Hispanic Population in the United States: March 1990. Series P-20, No. 449. Washington, DC: 1990. US. General Accounting Office. Health and Nutrition: Collection of Vital Statistical Data on Hispanics, GAO/HRD—88-688R. Washington, DC: March 1988. US. General Accounting Office. Health Care: Availability in the Texas- Mexico Border Area, GAO/HRD-89-12. Washington, DC: October 1988. US. General Accounting Office. Health Insurance Coverage: A Profile of the Uninsured in Selected States, GAO/HRD—91-31FS. Washington, DC: Feb- ruary 1991. Page 29 GAO/PEMD-92-6 Significant Gaps in Hispanic Access to Health Care Bibliography (973727) U. S. Library of Congress, Congressional Research Service. Selected Demographic Characteristics of the US. Hispanic Population and of His- panic Subgroups: A Report for the Subcommittee on Census and Popula- tion of the Committee on Post Office and Civil Service, US. House of Representatives. Washington, DC: US. Government Printing Office, 1986. U.S. Library of Congress, Congressional Research Service. Issues in the 100th Congress of Special Interest to Hispanics: An Overview of Major Issues Identified by the Congressional Hispanic Caucus, Report No. 87- 755 GOV. Washington, DC: 1987. U .S. Library of Congress, Congressional Research Service. US. His- panics: A Demographic Profile, Report No. 89-469 GOV. Washington, DC: 1989. Warner, David C. ”Health Issues at the U.S.-Mexican Border.“ Journal of the American Medical Association, 265:2 (1991), 242-47. Wells, Kenneth B., et a1. ”Acculturation and the Probability of Use of Health Services by Mexican Americans.“ Health Services Research, 2422 (1989), 237-57. Wells, Kenneth B., et a1. "Which Mexican Americans Underutilize Health Services?“ American Journal of Psychiatry, 144:7 (1987), 918-22. Page 30 GAO/PEMD-92~6 Significant Gaps in Hispanic Access to Health Care Ordering Information I I 1 The first copy of each GAO report is free. Additional copies are $2 1 each. Orders should be sent to the following address, accompanied 1 by a check or money order made out to the Superintendent of Docu- ' ‘ ments, when necessary. Orders for 100 or more copies to be mailed E to a single address are discounted 25 percent. US. General Accounting Office P.O. Box 6015 Gaithersburg, MD 20877 Orders may also be placed by calling (202) 275-624]. United States . General Accounting Office Washington, DC. 20548 Official Business Penalty for Private Use $300 First-Class Mail Postage & Fees Paid GAO Permit No. Gr_'.100~