HEALTH SERVICE NEEDS OF HISPANIC PEOPLE IN BEAVERHEAD COUNTY AS REFLECTED BY USE OF THE DISTRICT HOSPITAL EMERGENCY DEPARTMENT by Patricia Carrick A professional project submitted in partial fulfillment of the requirements for the degree of Master of Nursing MONTANA STATE UNIVERSITY-BOZEMAN Bozeman, Montana April 1996 n APPROVAL of a professional project submitted by Patricia Garrick This professional project has been read by each member of the professional project committee and has been found to be satisfactory regarding content, English usage, citations, bibliographic style, and consistency, and is ready for submission to the College of Graduate Studies. Patricia G. Butterfield, Ph.D., R.N. Approved for the Department of Nursing Kathleen H. Chafey, Ph.D., R.N. (Signature) / (j (Dat£) Approved for the College of Graduate Studies 7, Robert L. Brown, Ph.D. (Signature) (Date) STATEMENT OF PERMISSION TO USE In presenting this paper in partial fulfillment of the requirements for a master's degree at Montana State University-Bozeman, I agree that the Library shall make it available to borrowers under rules of the Library. • If I have indicated my intention to copyright this paper by including a copyright notice page, copying is allowable only for scholarly purposes, consistent with "fair use" as prescribed in the U.S. Copyright Law. Requests for permission for extended quotation from or reproduction of this paper in whole or in parts may be granted only by the copyright holder. Signature Date 1 ■ It- iv TABLE OF CONTENTS 1. INTRODUCTION 1 Purpose and Scope of the Study .. 2 Demographic Characteristics of the U.S. Hispanic Population 3 Methodologic and Sociologic Challenges in the Study of Hispanic Populations .. 5 Study Setting 6 Population of Interest............. 8 ImpUcations of Project for Health Planning and Policy 9 2. REVIEW OF THE LITERATURE 11 Health Status of Hispanic People in the United States ...... 11 Challenges in Data Gathering and Analysis 11 Major Morbidity and Mortality Indicators 13 Risk of Chronic Disease 14 Health Advantages of Hispanic Ethnicity 15 Health Risks and Barriers 16 Poverty 16 Environmental degradation 17 Special risks to children 17 Lack of health insurance 18 Lack of a primary provider 19 Educational disadvantage ...20 Practical barriers . 20 Language barriers 21 Racial and cultural prejudice 23 Lack of Hispanic Health Professionals 24 Impacts of Culture on Health Status 24 Hispanic Agricultural Laborers 27 Health Risks of Agricultural Labor .28 Health Risks of Hispanic Agricultural Workers in Southwestern Montana .... 29 Summary 31 V *• TABLE OF CONTENTS cont. 3. METHODOLOGY 32 Project Design and Data Collection Procedures 32 Record Retrieval ..... 33 Unit of Analysis 33 Sample: Inclusion and Exclusion 34 Chart Audit 36 Organization of Data for Analysis 36 Rights and Review of Human Subjects 37 Rationale for Methodologic Approach 38 Methodologic Issues in Determination of Ethnic Identity 39 4. PROJECT OUTCOME 42 Report of Research Findings 42 Demographic Characteristics 42 Gender and Age 42 Location of Residence 44 Temporal Trends 45 Occupation 46 Payment Source/Insurance Status 47 Relationship with a Primary Provider 49 Chief Complaint 50 Pediatric chief complaint .50 Adult chief complaint ... 51 Discussion of Findings 53 Population Trends 53 Insurance Status/Payment Source 55 Cultural Characteristics 56 Weaknesses and Strengths of the Study 57 Ethnic Identification 57 Accuracy of the Data Source 57 Limitations in Generalizability 58 Reflection of the Target Population 58 Implications of the Study 59 Suggestions for Program Design and Focused Intervention 60 Program Models 61 Conclusion ... 62 VI TABLE OF CONTENTS cont. REFERENCES CITED 63 APPENDIX A Map of Montana with Topographical Features 70 APPENDIX B Chart Audit Form 72 . . • \ APPENDIX C Application to the Human Subjects Review Committee 74 Vll Figure LIST OF FIGURES 1. Hispanic ED Visits, 1995 Adult/Child by Gender 43 2. Hispanic ED Visits, 1995 by Age and Gender 43 3. Hispanic ED Visits, 1995 Location of Residence • • • 44 4. Hispanic ED Visits, 1995 Comparison to Total Visits by Month .45 5. Hispanic ED Visits, 1995 by Month and Gender 46 6. Hispanic ED Visits, 1995 Payment Source by Age................. 48 , 7. Adult Hispanic ED Visits, 1995 Payment Source by Employment i 49 8. Hispanic ED Visits, 1995 Primary Provider by Month....... 1 50 9. Hispanic ED Visits, 1995 Pediatric Chief Complaint 51 10. Hispanic ED Visits, 1995 Adult Chief Complaint 53 vm ABSTRACT Hispanic people constitute an increasingly significant proportion of the population of the United States. U.S. Hispanic people tend to be younger, poorer, less educated, more transient, and less likely to be covered by health insurance than the population as a whole. U.S. Hispanic people experience numerous barriers to the acquisition of adequate health care. Poverty, environmental degradation, language and cultural differences, as well as childcare, employment responsibilities and transience all present barriers to health care access. As a result, Hispanic people tend to lack on-going relationships with primary health care providers, and have little contact with disease prevention and health promotion activities. Health care is often sought under urgent or emergent circumstances in hospital emergency departments. The purpose of the study was to examine trends in hospital emergency department (ED) visits by Hispanic clients in a rural community in southwestern Montana. Identification of Hispanic ethnicity was based on Spanish surname. Visits by Hispanic people in 1995 were identified and reviewed. Seventy-five visits were analyzed according to age and gender of clients, location of residence, temporal patterns of emergency department use, occupation, presence of an insurance payment source, relationship with a primary provider, and chief or presenting complaint. In decreasing frequency, ED visit categories included injuries and accidents (43% of total Hispanic ED visits), pediatric care (31%), miscellaneous adult medical complaints (18%), and women's health problems (8%). Hispanic males were seen more than twice as often as females. Forty-nine percent of Hispanic ED clients traveled more than 10 miles to receive care. Forty-seven percent of Hispanic ED visit records identified no available insurance payment source and 63% noted no primary healthcare provider. The frequency of ED visits by Hispanic people increased four-fold during the summer months of June, July and August. Since Hispanic people often name no alternative health care provider, emergency department visits may be their sole source of care. Thus ED visits may be seen as a reflection of the health service needs of this population. It is intended that this study contribute data in support of development of community-based outreach and interventions for medically underserved people. 1 CHAPTER 1 INTRODUCTION Hispanic people constitute an increasingly significant proportion of the population of the United States. In 1990 it was estimated that as many as 22 million persons, more than 9% of the nation's residents, were of Hispanic origin (Ruiz, 1993). By the year 2000, Hispanic people will constitute the largest ethnic minority in the country with 31 million members, 11% of all U.S. citizens (Novello & Soto-Torres, 1993). By the year 2050, it is projected that Hispanic persons will number 81 million, more than 20% of the total U.S. population (Council on Scientific Affairs, 1991; Novello & Soto-Torres, 1993; U.S. Bureau of the Census, 1995). It is estimated that as many as 1.5 to 5 million seasonal and migrant Hispanic agricultural workers live and toil in the U.S. (Ciesielski, Loomis, Mims, & Auer, 1994; Meister, 1991). This group of agricultural laborers includes both U.S. citizens and temporary immigrants who may be excluded from most demographic measures due to the transitory nature of their work and residence patterns and to their sometimes illegal or undocumented immigration status (Arias, 1986; Meister, 1991). While estimates of the actual numbers of undocumented workers are uncertain, in one study of Hispanic strawberry field-workers in California, Schlosser concluded that a "rise in the number of 2 migrant workers ... along with growth in the proportion who are illegal immigrants, reflects a national trend that has passed largely unnoticed" (1995, p. 82). This introductory chapter includes a summary of the project's purpose and scope. Demographic characteristics common to U.S. Hispanic populations are presented. Comments are offered on methodologic and sociologic challenges in the study of Hispanic populations. The setting and the population of the present study are also described. Chapter 1 closes with comments on implications of the study to community planning. Purpose and Scope of the Study The purpose of the present study is to provide data describing the health service needs of both resident and transient Hispanic people in one isolated, rural Montana community. A secondary purpose is to relate these needs to national indicators of health status in Hispanic populations. District hospital emergency department visits by Hispanic clients were analyzed to reveal patterns of use which were considered a reflection of health service needs otherwise unmet in this population. The data and interpretation generated by the project will be used to provide evidence in support of community-based efforts to design a comprehensive health service plan addressing the needs of the area's medically underserved citizens. Additionally it is intended that this study will impart information specific to a Hispanic population in a rural Montana setting which previously has not been studied. 3 Demographic Characteristics of the U.S. Hispanic Population The term "Hispanic" is a cultural rather than a racial designation. It refers to a heterogeneous group of people deriving from European, Black, and Indian heritage from several predominantly Spanish-speaking nations. The largest groups of Hispanic people in the United States are from Cuba (approximately 5% of U.S. Hispanic people), Puerto Rico (approximately 15%), and Mexico. Greater than 60% of the overall U.S. Hispanic population is of Mexican origin (Arias, 1986; Fox, 1988; Council on Scientific Affairs, 1991). Substantial populations from Central and South America (11.5%) and a group of "other Hispanic origin" (more than 8%) are included in the Hispanic designation as well (Fox, 1988; Piedra, 1992; Council on Scientific Affairs, 1991). Although diversity exists among these communitities, they are thought by many to "share common cultures, language, historical development and world views, including perspectives on health, death, and well-being" (Valdez, Giachello, Rodriguez-Trias, Gomez, & de la Rocha, 1993, p. 535). U.S. Hispanic people have many demographic characteristics in common. They are significantly younger than the overall U.S. population, with a mean age of 25 years compared with 31.4 years for the total population (Arias, 1986). In 1994 slightly more than 30% of the Hispanic population was less than 15 years old, compared with less than 20% of non-Hispanic Whites (U.S. Bureau of the Census, 1995). U.S. Hispanic people have higher birth rates than non-Hispanics, with 26.7 live births per 1000 Hispanic 4 population compared with 15.2 live births per 1000 total population in 1991 (National Center for Health Statistics, 1995). Many U.S. Hispanic people, especially Hispanic people of Mexican origin, are recently immigrated, composing what Davis refers to as "the * greatest migration of people in the history of humanity" (as cited in Smart & Smart, 1994). As a result of high rates of birth and immigration, the U.S. Hispanic population is increasing at a rate more than 5 times that of the population of the U.S. as a whole (Furino & Munoz, 1991; Torres, 1993). U.S. Hispanic people are poorer as a group than national averages (Estrada, Trevino, & Ray, 1990), with a median family income of $22,886 compared with $32,960 for non-Hispanic white families in 1993 (U.S. Bureau of the Census, 1995). Ruiz (1993) reports that Hispanic people are more than twice as likely as Anglos to be members of low-income families. In addition, U.S. Hispanic people tend to have lower levels of formal education than non-Hispanics, comparing poorly in grade level achievement, literacy rates, and representation in higher education settings (Arias, 1986; Ginzberg, 1991). Due in part to low educational levels and in part as a result of "proximity to the former homeland, the relative isolation of some predominantly Mexican-American neighborhoods, and contact with a constant flow of new immigrants who for the most part are Spanish-speaking only", Spanish monolingualism is a persistent pattern in many Hispanic communitites in the U.S. (Rodriguez, 1983, p. 136) and presents significant challenges in terms of educational integration (Rodriguez, 1982), legal advocacy, and health care access, intervention and treatment (Padgett & Barms, 1992; Kirkman-Liff & Mondragon, 1991). 5 Methodologic and Sociologic Challenges in the Study of Hispanic Populations The lives of both permanent and migrant Hispanic people in the United States are often highly transient. As noted above, many permanent Hispanic residents are recent immigrants. Migrant or seasonal laborers may maintain primary residence at sites distant from their constantly changing worksites, and both migrant and undocumented laborers may continue to maintain residence in their countries of origin. Of the vast majority of Hispanic persons who maintain permanent residence in the U.S., over 85% live in only nine states. Hispanic populations are densest in California, Texas, New York, Florida, Illinois, Arizona, New Jersey, New Mexico, and Colorado where they tend to be heavily concentrated in metropolitan areas (Council on Scientific Affairs, 1991; Ginzberg, 1991). Recognition of the enormous impacts of rapid Hispanic population growth during past and coming decades has directed considerable national attention toward the health status of Hispanic Americans. The Hispanic Health and Nutrition Examination Survey (HHANES) conducted in 1982-84 was the first special population survey ever undertaken by the National Center for Health Statistics. This survey collected data on over 16,000 U.S. Hispanic people, focusing on geographic centers noted to have high Hispanic population density (Delgado, Johnson, Roy, & Trevino, 1990). These data have been widely analyzed, reviewed, and augmented by numerous studies published throughout the second half of the 1980s and early 1990s. Major causes of morbidity and mortality in Hispanic people and comparisons between Hispanic and non-Hispanic populations derived 6 from such studies are discussed in Chapter 2 of this document. Several factors have contributed to incompleteness and inaccuracy in descriptions and therefore understanding of U.S. Hispanic populations. Vague or conflicting definitions of Hispanic ethnicity have confounded the usefulness of some studies (Sacco, 1995; Culebras, 1995). Multiple obstacles to the quantification and characterization of seasonal and migrant Hispanic populations have contributed to difficulties in evaluating health status and health care access specific to these groups on either national or community levels. Because of the continuing emphasis in most research on areas of high population density, little data has been compiled relative to the circumstances of Hispanic people in more remote, rural settings. However, as the U.S. Hispanic population has expanded in recent years, Hispanic people have found their way well beyond the southwestern border towns and metropolitan centers which have been the focus of research until now into even the most rural communities of the northern Rocky Mountain states where Hispanic populations previously have not been identified or described. Study, Setting Beaverhead County is the largest county in Montana, encompassing 5,529 square miles (Montana Department of Health and Environmental Sciences, 1993). It is located in the geographically remote, southwestern comer of the State where precipitous mountains along its southern and western borders form a natural boundary with the State of Idaho (see Appendix A). Nestled among the County's several mountain ranges lie a series of valleys which lie at 5,000 to 7,000 feet elevation. Beaverhead County is considered high 7 mountain desert, with average annual precipitation on valley floors ranging from 16 plus inches per year maximum in the Big Hole Valley to as little as 8 to 10 inches per year around most of the Beaverhead Valley. Drought years in the 1980s and 90s have seen precipitation levels even lower than these (J. Maki, Beaverhead County Extension Service, personal communication, September 19, 1994). Average temperatures in the County range from highs of 80°F in July to lows of 6°F in January; summer temperatures above 100°F and winter temperatures of less than -SOT are recorded (J. Maki, personal communication, October 17, 1994). The local economy historically has been based on natural resource industries which include mining, timber and agriculture. Of these, agriculture is dominant in importance. Cattle ranching is the primary agricultural industry with agricultural receipts and cattle numbers both ranking highest among all counties in the State of Montana (J. Maki, personal communication, September 24, 1994). Beaverhead County is served by a single, public district hospital, 31-bed Barrett Memorial Hospital. It is the only acute care facility between Idaho Falls, Idaho to the south and Butte, Montana to the north, a distance of 210 miles, and between Salmon, Idaho to the west and Sheridan, Montana to the northeast, a distance of more than 170 miles. There are no alternative urgent-care or walk-in clinics within the BMH catchment area of over 5,000 square miles. Barrett Memorial Hospital's emergency department receives more than 3,100 visits per year and forms a busy interface with the community. All who present for care, regardless of residence or financial status, are served within the capabilities of the facility and its staff. In some cases total care is provided on-site; in 8 others cases, clients are stabilized and transferred as appropriate by ground or air to larger, more distant facilities and specialty services. Because of its geographic isolation, the Barrett Memorial Hospital emergency department serves most of the urgent and emergent health needs that arise within the County. For many of those who have no established relationship with primary health care providers in the community, the ED provides the initial point of contact with non-urgent health services as well. Population of Interest Like demographic data related to Hispanic people nationally, data for Hispanic people of Beaverhead County are sketchy and probably incomplete. Although the 1990 census notes only 133 County residents of Hispanic origin (Montana Census of Population and Housing, 1990), the likelihood is that transient Hispanic agricultural laborers swell the population to at least two to three times reported numbers during the summer season (Carrick, 1994a). One Hispanic resident estimates that summertime Hispanic populations in recent years have exceeded 400 individuals (anonymous, personal communication, September 18, 1994), or almost 5% of the County’s total population. Data which are available suggest that Hispanic people of Beaverhead County are, like their counterparts nationwide, a youthful population. No Hispanic individual older than 49 years of age was noted in the sample count of the 1990 census (Montana Census of Population and Housing, 1990). 9 Resident Hispanic families in Beaverhead County are, like Hispanic families throughout the United States, poorer than the population as a whole. The 1990 census estimates that Hispanic families in Beaverhead County have a mean annual household income of $17,878, barely 70% of the $25,201 which constitutes the mean income for the County’s household population as a whole, and less than 50% of the national mean household income $41,428 estimated for 1993 (U.S. Bureau of the Census, 1995). The overwhelming majority of Hispanic people in Beaverhead County are primarily Spanish¬ speaking; about one-third reported feeling that they are "linguistically isolated" (Montana Census of Population and Housing, 1990). Like Hispanic populations in other western regions, more than 80% of the Hispanic population in Beaverhead County is of Mexican derivation (Montana Census of Population and Housing, 1990). Implications of Pro ject for Health Planning and Policy This previously unrecognized, apparently growing Hispanic population presents significant challenges to a rural community with limited professional, institutional and 1 fiscal resources. Funding for the development of appropriate public health services is a daunting issue. Effective outreach, disease prevention, health promotion and health education activities will demand the development of culturally appropriate materials and services based on an understanding of the needs of service recipients. Planning efforts will require recognition of present and projected impacts on local institutions and providers. It is intended that the following analysis of emergency department use will provide initial data articulating the needs of Hispanic people which may help to guide local health 10 policy makers and primary care providers in the design and development of community- based health services which are culturally and practically appropriate to the needs of this target population. 11 CHAPTER 2 REVIEW OF THE LITERATURE In order to create a background for the presentation of data which is compiled by this study, the literature has been explored for a description of the current health status of Hispanic people nationwide. Health risks and barriers to U.S. Hispanic populations are identified. A discussion of demographic characteristics and health risks of Hispanic agricultural workers in the United States is conducted. The chapter concludes with a projection of potential health risks specific to Hispanic agricultural workers in rural Montana. It is intended that this overview will supply a context within which the reader may perceive the patterns of emergency room use by Hispanic people in southwestern Montana which are revealed by the study. It is further intended that the overview will provide a rationale for understanding use of the emergency department by Hispanic people as a reflection of otherwise unmet, underlying health service needs. Health Status of Hispanic People in the United States Challenges in Data Gathering and Analysis Despite considerable analysis and supplementation of the HHANES data by numerous studies published during the last ten years, valid health indicators of morbidity 12 and mortality have been difficult to ascertain for Hispanic people on either national or local levels (Delgado & Estrada, 1993). Neither national death certificate forms nor most hospital records contain reliable Hispanic identifiers. Even when identifiers exist, designation tends to be subjective and imprecise, depending heavily on the disposition of the individual completing the records (Council on Scientific Affairs, 1991). In 1990 a congressional mandate, The Disadvantaged Minority Health Improvement Act, was established which has improved identification of ethnicity and increased Hispanic representation in national health data bases (Delgado & Estrada, 1993). Other studies have had some success using Hispanic surnames and self-designation by study subjects to clarify Hispanic ethnicity (Council on Scientific Affairs, 1991). It is reasonable to question the accuracy of ethnic identification based on voluntary revelation in the context of a U.S. political climate characterized by anti-immigration sentiment. The mid-1990's have seen federal and state legislation which increases constraints on eligibility for legal immigration status, especially for Mexican and Central American citizens, which tightens U.S.-Mexican border immigration controls, increases documentation requirements for immigrant persons, and threatens improperly documented immigrants with loss not only of their own access to health and social services but which, in addition, prohibits access to public education, health and social service opportunities for their children (Morning Edition. February, 1996). With so much at stake, reluctance to reveal ethnicity non-indigenous to the U.S. would certainly be understandable and could have a significant impact on statistical analyses related to ethnic identification. In light of these and other methodological and political factors, contemporary statistics describing 13 major health indicators for Hispanic populations tend to be approximate and sometimes conflicting. Major Morbidity and Mortality Indicators Studies which are available related to Hispanic health indicators have yielded findings that, overall, Hispanic people die of the same major causes of mortality common to the population as a whole - heart disease, cancer, and stroke - and at rates which are approximately equivalent (Council on Scientific Affairs, 1991; Winkleby, Fortmann, & Rockhill, 1993). However, Hispanic people have been found to be at increased risk of several specific causes of morbidity and mortality in the United States. It has been found that Hispanic people suffer an incidence of esophageal, stomach, liver and pancreatic cancers which is increased above that of non-Hispanics; deaths due to stomach cancer are twice as high for Hispanics (Markides & Coreil, 1986). Hispanic women have been found to have a rate of cervical cancer which is twice to three times that of non-Hispanic white women (Council on Scientific Affairs, 1991; Giuliano & Alberts, 1994). Alcoholism and cirrhosis are more prevalent as well, particularly among Mexican and Puerto Rican men. One study conducted in Southern California concluded that, of all deaths of Mexican- American men aged 30 to 60 years between 1918 and 1970, 52% were due to alcohol compared with 24% for Anglo men in the same examination period. Mortality due to narcotic addiction and rates of violent death due to accidents, suicides, and homicides are disproportionately high among Hispanic, young adult males (Ruiz, 1993; Council on Scientific Affairs, 1991). A study of young southwestern Hispanic men of Mexican origin 14 revealed homicide rates more than three times that for Anglo men of comparable ages (Smith, Mercy, & Rosenberg, 1986). Risk of Chronic Disease Hispanic people experience higher rates of some chronic diseases, particularly non¬ insulin-dependent diabetes mellitus (NIDDM ), for which Mexican Americans are at more than three times greater risk than the general population (Hanis, Hewett-Emmett, Bertin, & Schull, 1991; Furino & Munoz, 1991; Flegal et al, 1991). Additionally, Hispanic people afflicted by NIDDM may have higher rates of serious microvascular complications including retinopathy, nephropathy, and neuropathy (Furino & Munoz, 1991). In studies conducted in Texas and Los Angeles, rates of end-stage renal disease secondary to diabetes were found to be as much as twice as high among Hispanics as among non- Hispanic whites (Chiapella & Feldman, 1995); controversy exists as to whether this reflects a genetic predisposition, variations in access to and quality of medical care, or other environmental factors (Hamman et al., 1991). Risk factors related to blood pressure indicators were similar in Hispanics and Anglos in one study in which subjects were matched for socioeconomic status as reflected by level of education (Winkleby et al., 1993). However, as noted earlier, Hispanic people are at greatly increased risk of poverty, and Hispanic people who are poor are particularly vulnerable to an increased risk of hypertension which is more apt to go unrecognized and untreated (Council on Scientific Affairs, 1991; Novello, Wise, & Kleinman, 1991). 15 Tuberculosis is more prevalent in Hispanic populations with rates as much as 4.3 times greater than that for non-Hispanics (Council on Scientific Affairs, 1991). The increased rate of tuberculosis is associated with an incidence of acquired immunodeficiency syndrome (AIDS) which is significantly higher in Hispanic than in non- Hispanic white adults. Although Hispanic people composed around 9% of the U.S. population in 1991, they were thought to account for 14% of all reported acquired immunodeficiency syndrome (AIDS) cases, more than 20% of AIDS cases among women, and 22% of all pediatric AIDS cases in the U.S. (Ruiz, 1993). Other serious communicable and parasitic diseases, including influenza, pneumonia, and sexually transmitted diseases are problematical among adult Hispanics, especially in inner city and U.S.-Mexican border settings (Markides & Coreil, 1986; Warner, 1991). Hispanic children suffer disproportionately high rates of lead poisoning and measles (Novello et al., 1991), otitis, conjunctivitis, influenza, pneumonia, rheumatic fever, diarrhea and parasitic infestations including pediculosis and scabies (Rodriguez, 1983). Health Advantages of Hispanic Ethnicity Hispanic ethnicity appears to confer some advantages in terms of certain major health indicators. Despite an increased incidence of cervical cancer and a more advanced stage at diagnosis, 5-year cervical cancer survival rates are slightly higher for Hispanic women compared with women overall (Council on Scientific Affairs, 1991). Hispanic women are 30% less likely to develop breast cancer than non-Hispanic women (Giuliano & Alberts, 1994), and there are some indications that Hispanic people have lower than 16 average rates of colon and prostate cancer as well (Markides & Cored, 1986; Ginzberg, 1991). Although Hispanic women have decreased access to prenatal and postnatal care, rates of premature delivery and low birth weight are significantly lower, especially for Mexican Americans, than would be predicted based on socioeconomic status alone (Ginzberg, 1991; Council on Scientific Alfairs, 1991; Novello et al., 1991). Neonatal mortality among Mexican American infants is less than half that among Anglos (Torres, 1993). Health Risks and Barriers Poverty. The extent to which increased health risks experienced by Hispanic Americans are attributable to depressed socioeconomic status is uncertain, since no studies have been successful in controlling for the multiplicity of its effects (Markides & Cored, 1986). Blane (1995) observes that the distribution in society of morbidity and mortality indicators is predictably graded along dimensions of socioeconomic status and dubs this the "social patterning of health" (p. 903): "parental disadvantage is associated with... social disadvantage during chddhood and adolescence. The accumulation of disadvantage and risk during the period from birth to early adulthood is associated with disease, risk factors, and health behavior during middle age ... [and] with several prevalent chronic diseases in late middle age" (p. 904). Funkhouser and Moser (1990) agree that poverty and poor health are mutually synergistic throughout life. Pamies and Woodard (1992) assert that "the poor in the United States have always been at a disadvantage for health care" (p. 443) and suggest that "some combination of income, education, and employment are accountable for most of the 17 differences in survival rates" (p. 444). Environmental degradation. In many cases, the effects of poverty on the circumstances in which large numbers of U.S. Hispanic people reside are altogether too graphic. Hispanic residents along both sides of the U.S.-Mexico border are subject to wide-spread water and air pollution and exposed to illegal hazardous waste dumping (Ginzberg, 1991). Overcrowding and substandard housing with lack of septic facilities and an absence of potable water have been found to be unconscionably prevalent in locations where poor Hispanic people reside: in border areas, in migrant labor settings, and in inner city barrios (Rodriguez, 1983; Warner, 1991). Ironically, conditions in all of these settings may be so extreme as to mimic the same third world conditions many immigrants have sought to escape by leaving their countries of origin (Schlosser, 1995). Such conditions predispose Hispanic residents to the transmission of infectious and parasitic diseases, and contribute to their increased vulnerability to adverse outcomes from both acute and chronic health problems as described above. Special risks to children. Vulnerability to the effects of poverty and environmental degradation are intensified for Hispanic children. Children in inner city settings are exposed to increased risks related to lead poisoning and violence. Children along the southwest border may suffer long term effects from exposure to contaminants flowing from maquiladoras and sewage discharged directly into surface water or underground aquifers (Warner, 1991). Children of migrant laborers suffer multiply increased health risks including direct exposure to pesticides, lack of sanitation, and crowded living conditions 18 by performing field work themselves, by accompanying parents to the fields, by living close to the fields, and indirectly by having contact with adults at increased risk for exposure to environmental contaminants and disease (Waldman, 1994; Mobed & Schenker, 1992). Lack of health insurance. Of the many factors which contribute to increased health risk among Hispanic people, lack of health insurance may be the single most important factor of all. Pamies and Woodard declare in no uncertain terms that "lack of health insurance is ... the greatest barrier to health care access facing the growing number of disadvantaged people in the United States" (1992, p. 446). Although Hispanic people are among the most actively employed participants in the labor force, a large and increasing number have neither public nor private medical insurance coverage (Valdez et al., 1993). Below average levels of educational achievement contribute to the employment of a large proportion of Hispanic people in low-wage sectors of the economy where they do not receive employer-provided health insurance (Ginzberg, 1991). Despite low-income work and a frequent lack of employer-provided insurance, Hispanic people may be unaware of or ineligible for government-sponsored insurance programs such as Medicaid (Solis, Marks, Garcia, & Shelton, 1990). Even when they are technically eligible, Hispanic people may be linguistically unable to negotiate the complicated forms and procedures required for application. In the face of prohibitive private insurance costs, Hispanic people are more likely than any other racial or ethnic group to be un- or underinsured (Anderson, Giachello, & Aday, 1986; Ruiz, 1993; Novell© & Soto-Torres, 1993; Valdez et al., 1993). 19 It is estimated that, in 1993, 31.6% of Hispanics were uninsured nationally, compared with 20.5% of blacks and 14.2% of white non-Hispanics (U.S. Bureau of the Census, 1995). Lack of a primary provider. For those who do not have insurance, exorbitant health care costs tend to prohibit access to primary caregivers and preventive care, diminish the likelihood of continuity of care, and result in high rates of hospitalization for potentially preventable exaccerbations of such conditions as asthma and diabetes (Valdez et al., 1993; Ruiz, 1993). High deductibles and copayments, and providers who refiise to provide care to Medicaid recipients may effectively reduce access even for those who are insured (Funkhouser & Moser, 1990). In fact, only about half as many Hispanics as Anglos report a regular source of primary health care (Council on Scientific Affairs, 1991). In their report from HHANES, Solis et al. (1990) describe the situation thus: Hispanics are less likely to have health insurance coverage and less likely to have a routine place for obtaining health services. This reflects their low income, low education, and employment in positions that do not provide insurance benefits. Because of these barriers, they are more likely to rely on public health facilities, hospital outpatient clinics, and emergency rooms (p.l 1). Compared with non-Hispanics, twice as many Hispanic persons report using the emergency department as a source of primary care (Council on Scientific Affairs, 1991). In the face of few alternatives, emergency room care often "comes about because they have delayed treatment until the condition has reached emergency proportions" (Cordes, Doeksen, & Shaffer, 1994, p. 29). Pamies and Woodard note that among the "numerous reasons why low socioeconomic status negatively affects health in general... [one factor is the] lack of continuity of health care because of overdependence on emergency room 20 care owing to lack of insurance or underinsurance” (1992, pp. 444-445). Although preferable to no care at all, use of the emergency room as a source of primary care precludes continuity, follow-up and implementation of most preventive health interventions (Aguirre-Molina, Ramirez, & Ramirez, 1993). Educational disadvantage. In 1994, 46.7%, almost half of the U.S. Hispanic population older than 25 years of age, had not completed high school as compared with only 16.6% of non-Hispanic persons (U.S. Bureau of the Census, 1995). This relative lack of education affects health status and health practices in a complex manner. Hispanic people are predisposed to employment in low-wage job sectors in which employers are unlikely to provide health insurance, while wages are inadequate to purchase expensive private health insurance. Low levels of education have been shown to be related to retention of traditional health beliefs and health practices. A concomitant tendency to distrust the techniques and philosophy of modem medicine may contribute to avoidance of contact with established medical providers. Low levels of education resulting in decreased literacy rates are related to a general lack of information about available services, and to low utilization of preventive health services (Andersen et al., 1986). Practical barriers. Both insured and uninsured Hispanics encounter other barriers both to access and to appropriate health service utilization. Few providers or urgent care clinics locate practices in Hispanic neighborhoods and communities, creating predictable transportation dilemmas. In urban settings public transportation services to and from barrio neighborhoods may be infrequent, dangerous, or otherwise difficult to access. In 21 rural settings distances may be great making private travel expensive even if available, while public transportation may be inconvenient or entirely absent. Neither private providers nor community clinics commonly offer hours which accommodate the double shifts and multiple jobs required to support families on low-wage incomes. Modem urgent care clinics which tend to offer expanded hours are almost universally fee-for-service, and are seldom available in rural settings. Hispanic patients report that they experience longer delays between appointments and longer delays in medical facilities waiting for scheduled appointments and services than non-Hispanic patients (Valdez et al., 1993). These circumstances create many problems with scheduling of childcare and employment responsibilities (Estrada et al., 1990; Valdez et al., 1993). In addition, when "a day off to see the doctor means a day of lost wages" (Rodriguez, 1983, p. 137), it is the rare minimum or low wage-earner who can afford time off for any but the most urgent health matter. Language barriers. Language barriers also discourage access and create discrepancies which undermine quality care (Andersen et al, 1986; Padgett & Barms, 1992). Language barriers cause problems for Hispanics attempting to negotiate entry into complex health systems which can be prohibitively confusing even to native English- speakers (Rodriguez, 1983). Incomprehension, embarassment, and fears of inadequacy may prevent eligible individuals from initiating the often complicated applications for government insurance benefits and many public health programs. Language barriers result in decreased use of acute care and, perhaps to a greater extent, decreased use of 22 preventive and mental health services, which tend to be even more language-dependent (Ruiz, 1993). Language barriers may impair communication from patient to provider, thus complicating the acquisition of the health history, the development of an accurate diagnosis and prohibiting the participation of the Spanish-speaking patient in the treatment plan. Language barriers may equally impair communication from provider to patient, significantly inhibiting the creation of trust and rapport, and curtailing health education and illness prevention activities and information. Language barriers may preclude the use of written English-language materials such as information sheets, discharge instructions and treatment plans. At the same time, low levels of educational achievement may imply illiteracy in Spanish as well, effectively eliminating the usefulness of written Spanish- language materials as well. In many health care settings professional interpreters are not readily available. Funds to pay for the services of professional interpreters in health care settings are limited. Frequently patients and clinicians are forced to utilize suboptimal mechanisms including dependence on their own halting language skills, reliance on the skills of family or friends, or reliance on ad hoc interpreters who may be bilingual strangers from the waiting room, or employees called away from unrelated job responsibilities to act as interpreters. Inadequate interpretation based on questionable or unverified language skills raises ethical problems related to the medicolegal obligation of informed consent. Interpretation performed inappropriately by children or other family members may carry emotional impacts for patient and interpreter alike which can prevent accurate understanding and be potentially damaging to one or both. Interpretation by informally recruited strangers or 23 facility employees may challenge the obligation to confidentiality inherent in every patient/provider relationship (Haffner, 1992; Woloshin, Bickell, Schwartz, Gany, & Welch, 1995). Racial and cultural prejudice. More elusive, but perhaps as influential in access to care as the impacts of such relatively tangible factors as language, transportation, and income, may be the effects of racial and cultural prejudice. Such postulated effects are difficult to identify and still more difficult to measure. They are so difficult, in fact, that few authors in the field of health care research include direct reference to prejudice in their discussions of health care access. The social scientists, who have tended to be less reticent to entertain a discussion of such issues, have developed an entire literature on the subjects of prejudice and racism. Moore states unequivocally that "institutionalized racism... operates in education, the job market, the housing market, and the criminal justice system" (1985, p.l). It would require only a reasonable logical extension to include health care systems in this list. Among those from the health sciences who have dared to address issues related to prejudice, Funkhouser and Moser struggle to distinguish the extent to which inequalities in health care access and delivery are attributable to racism or to poverty and educational disadvantage. They conclude that "the problem of inadequate health care .. .is largely due to socioeconomic constraints" (1990, p. 54). Despite their conclusion, Funkhouser and Moser state that inequalities in health care "exist along racial lines" (1990, p. 47). In terms of investigation, analysis, and 24 planning and prevention, it may be more useful to recognize the fact that the racial stratification of health care exists than to attempt to identify conclusively the reason why this is so. Lack of Hispanic Health Professionals. One solution to the problem of multiple barriers to health care which has been proposed is to increase the proportion of bilingual, Hispanic health care workers. However, there is already a scarcity of Hispanic providers in the health professions. Less than 3% of dentists, registered nurses, and pharmacists, less than 5.5% of physicians, and only 4.6% of psychiatrists were Hispanic in 1989 (Ruiz, 1993). This condition is expected to worsen because of the severe underenrollment of Hispanics in professional schools (Andersen et al., 1986). As the Hispanic population reaches 10 and then 20% of the overall population, most health facilities will be hard- pressed to provide even a portion of the appropriate ratios of bilingually or culturally proficient staff (Woloshin et al., 1995). Impacts of Culture on Health Status Several aspects of Hispanic culture may impact the health status of Hispanic people in the U.S. Even when an effort is made to control for the influence of access related to socioeconomic status, a traditional gender modesty appears to inhibit participation by Hispanic women in cancer screening programs based on mammography and pap testing. This factor may contribute to the identification of breast and uterine cancers at later stages in Hispanic women (Elder et al., 1991; Stein, Fox, & Murata, 1991). 25 The diets of Hispanic people include higher levels of both fats and calories than those of Anglos, predisposing them to increased risk of both cancer and heart disease. This risk is somewhat counterbalanced by the higher rate of fiber servings in Hispanic diets and the finding that Hispanic people are less likely to smoke (Elder et al., 1991). The tradition of machismo, or male pride, may present special stressors to Hispanic men who are often employed in low wage positions and hold little prestige or social status in a nation dominated by Anglo culture. Some researchers contend that people of Mexican origin are more likely than Anglos to experience their lives as directed by chance or by an external locus of control (fatalismo), and that this perception predisposes them to higher levels of depression (Gonzalez, Atwood, Garcia, & Meyskens, 1989). Whether perception of an external locus of control is an inherent cultural characteristic, or an artifact of life in a culture in which the locus of control actually is largely outside the domain of a socioeconomically disadvantaged population, depression is a common complaint, although seldom the presenting complaint, among Hispanic people who interact with health care providers. Ruiz (1993) contends that psychiatric illnesses often are understood in terms of folk concepts and practices, and rarely receive diagnosis or treatment within the medical/health care system. Several studies refer to systems of Hispanic folk medicine guided by healers such as curanderos. A question regarding the extent to which folk medicine forms a primary health practice among Hispanic people in the United States has been raised repeatedly. In their report from the HHANES, Higginbotham, Trevino, and Ray (1990) discovered that only a very small percentage of their study population (less than 5%) consulted folk 26 medicine practitioners. Suarez (1992) reports consistent findings that use of lay consultation and implementation of folk remedies were associated with barriers to health service access rather than the resource of primary preference and did not preclude seeking formal medical care. Rodriguez (1983) concurs with Ruiz in concluding that folk medicine is predominantly an interventional system used in psychiatric rather than physiologic illnesses. Researchers agree that Hispanic culture is characterized by a strong, highly interdependent family structure within which cooperation and shared decision-making prevail (Rodriguez, 1983; Ruiz, 1993; Torres, 1993). Rehance on the family is suggested as contributing to low utilization of health services by Hispanic people since family members are encouraged first to seek health advice within the family itself (Andersen et al., 1986). The implication is made that this pattern results in adverse health events. Conversely, Torres (1993) and Novello et al. (1991) submit that favorable family and community systems may create a protective effect which supports the positive birth outcomes observed especially in Mexican American populations. Dedication and interdependence within families may be positively motivating in terms of health risk prevention behaviors as well. Vaughan (1993) reports that concern about potentially harmful effects of pesticide exposure on children is greater than concern for self among farmworkers exposed to pesticides, and may have implications for risk management in employment settings. Garrick (1994a) reports that, while extensive discomfort and morbidity is endured by rural Hispanic adults who tend to avoid the expense of consultation with available health service providers, the health needs of children much 27 more readily stimulate parents to overcome barriers and interact with local health services. Hispanic Agricultural Laborers Although no one really knows how many migrant and seasonal agricultural laborers there are in the United States, projections range from 1.5 to 5 million farmworkers and their families (Ciesielski et al., 1994; Meister, 1991; Mobed, Gold, & Schenker, 1992). The vast majority of these are thought to be of Hispanic, specifically of Mexican-origin Hispanic derivation (Guamaccia, Angel, & Angel, 1992; Waldman, 1994). Vaughan reports that "more than 80% of the hired farm labor in California and many other western agricultural states consists of individuals who have recently migrated, primarily from Mexico" (1993, p. 675). Moses (1989) suggests that Hispanic people of Mexican origin may comprise as much as 90% of the farm labor force nationwide. In a review article summarizing the years 1966-89, Rust (1990) was not able to discover any published reports measuring mortality or survival data specific to farmworkers. Lack of a uniform definition for farmworkers, including distinctions among permanent, seasonal, and migrant individuals, contributes to this scarcity of epidemiological data (Mobed et al., 1992; Zahm & Blair, 1993), as do the transience of the workers themselves and the conditions of the workplace (Ciesielski et al., 1994). To these factors may be added the reluctance of many immigrant workers to engage in contact with government agencies or investigators who could be perceived as authority figures. With few reliable denominators, such basic health indicators as maternal and infant mortality rates, incidence of common disability, and disease prevalence are impossible to 28 calculate reliably (Mobed et al, 1992). Nonetheless, certain aspects of the lives and the health status of Hispanic agricultural workers can be characterized. A discussion of health risks to farmworkers follows. Health Risks of Agricultural Labor Since 1987, agriculture has had the dubious distinction of being the most hazardous occupation in the country, with 49 (Mobed et al., 1992) to 52 (Rust, 1990) work-related deaths per 100,000 workers per year compared with approximately 11 deaths per 100,000 workers for all other occupations. Rates of injury and illness are estimated at 12.7 cases per 100 workers per year (Rust, 1990). Occupational hazards of agricultural labor have been studied extensively, especially in row-crop settings in California and the Southwest. Hazards which have been identified are varied and include acute and chronic pesticide exposure; dermatoses related to a variety of chemical and irritant exposures; lack of potable water and toilets in the workplace; inadequate housing with similarly inadequate sanitation facilities; urinary tract infections and kidney disorders; infectious and non-infectious respiratory conditions; reproductive health problems; acute injuries related to farm machinery and falls; soft tissue and musculoskeletal injuries, particularly chronic back and joint trauma; and heat and cold exposure (Guamaccia et al., 1992; Meister, 1991; Mobed, et al, 1992; Rust, 1990; Shaver & Tong, 1991). Most labor laws and protective agencies fail to address the plight of farmworkers. Meister (1991) reports that only 30 states provide workers’ compensation benefits to 29 farmworkers. Occupational Safety and Health Act (OSHA) regulations pertain to farms with over 10 employees, effectively excluding more than 85 percent of farmworkers. Lack of unionization (Meister, 1991) and political powerlessness in local decision-making arising from patterns of transience (Bechtel, Shepherd, & Rogers, 1995) may both be associated with increased health problems. It is clear that farmworkers, who are most likely to be Hispanic, and therefore already in jeopardy of diminished health status because of the health risks and barriers discussed earlier in this paper, are further threatened by the myriad dangers associated with their work. The fact that many Hispanic farmworkers are migrant exacerbates health problems related to poverty, access, language, and culture still more (Meister, 1991). Undocumented workers are at greatest risk of all for compromised health. They are most likely to be homeless, living in the fields where they work, and fears of deportation may lead them to avoid contact with health care providers. Illegal farmworkers "live under more stress, in the worst living conditions, and with the least health care of all migrant workers" (Meister, 1991, p. 506). Health Risks of Hispanic Agricultural Workers in Southwestern Montana Agriculture is an industry as diverse as the locations in which it is conducted and as varied as the outputs of its production (Shaver & Tong, 1991). Many of the studies cited in this paper have described hazards specific to field labor. However, Hispanic ranchworkers in the Northwestern Rockies engage in work-related activities which differ drastically from row-crop labor and imply previously undefined dangers. Ranch duties may 30 include such unique health threats as: operation of heavy machinery with unshielded, moving parts; work on horseback; contact with large, range-raised (and therefore sometimes dangerously wild) livestock; management of irrigation water delivered over long distances under tremendous pressures; long hours of outdoor work in frigid climactic conditions; work with both animals and machinery in precipitous terrain; and solitary chores in remote locations with no provision for communication in case of emergency. The circumstances of Hispanic people who live and work in sparsely populated regions of the Northern Rockies entail other, perhaps more subtle health challenges also not previously discussed in the literature. It has been stated that life in remotely rural settings involves a potential for geographic and social isolation which may threaten the psychological health and well-being of any rural resident (Dietz, 1991; McKinley, 1991). Meister (1991) mentions the potential for social and physical isolation in the circumstances of non-working wives of farmworkers. The potential for social isolation inherent in rural settings is compounded for Hispanic agricultural laborers in the Northwestern Rocky Mountains by decreased access not only to the human community in general, but to the cultural community of origin, specifically. On many family-owned ranches, Spanish¬ speaking men are employed individually or in small groups of two or three. The nearest Spanish-speaking neighbor may be miles, even towns away. Shops and services may be at a considerable distance; public transportation is infrequently available; acquisition of a drivers' license, and therefore unconstrained mobility, may be dependent on English language literacy seldom acquired by the newly immigrant or migrant worker (Garrick, 1994b). Spanish-speaking interpreters may be distant or nonexistent, while those who are 31 available maybe inappropriate or of questionable competence (Haffiier, 1992). Summary In the preceding pages, literature concerning the overall health status of Hispanic people in the United States has been reviewed with an emphasis on those major health indicators for which data is available. Health risks and barriers to health care for all Hispanic Americans have been identified, and health risks related to agricultural labor have been discussed. Finally, health risks specific to Hispanic agricultural laborers in the Rocky Mountains of southwestern Montana have been suggested. Chapter 3 describes the methodology of a study which sought to identify the health service needs of one rural Hispanic population. 32 CHAPTER 3 METHODOLOGY An overview of project design and data collection procedures for the current study is outlined in the following paragraphs. The unit of analysis is defined. The sample is described and inclusion and exclusion criteria are delineated. The data collection tool, data analysis techniques and procedures are discussed. Processes for the protection of human subjects are described. The chapter closes with discussion of a rationale for the methodologic approach chosen for the study and final comments on methodologic issues related to the designation of ethnic identity. Project Design and Data Collection Procedures The study was a retrospective, descriptive investigation which examined use of the emergency department of a rural hospital in a geographically isolated agricultural community by individuals of Hispanic ethnicity. Emergency department records were analyzed to discern patterns in the demographic characteristics of Hispanic clients, identify the availability of payment sources and relationships with primary providers outside the emergency department setting, and examine trends in the nature of the acute health care services sought by Hispanic clients. Informed by indicators in the review of literature, the 33 investigator analyzed instances of ED use with the intent of perceiving through them a reflection of otherwise unmet health service needs of Hispanic clients in this catchment area. All ED visits by clients of Hispanic ethnicity from January 1, 1995 through December 31, 1995 were included in the analysis. Record Retrieval - Emergency department log books and client medical records provided the data sources which were accessed for the study. Data for the study were collected using a multi-step process. First, entries in emergency department log books, which contain an itemization of every ED visit, were reviewed by the investigator for the calendar year, 1995. In order to keep log books accessible to staff, this initial review was conducted in the hospital emergency department setting. A comprehensive list of the names of clients with Spanish surnames and the dates on which ED services were delivered was compiled from the log book. The list included those clients whose names were judged by the investigator to be ethnically inconclusive. The list was submitted to the Director of the medical records (MR) department. Medical records staff retrieved from the files hard copies of each emergency department visit record and these were assembled in the MR Department. Ninety-three ED visit records were pulled by MR staff for review. Unit of Analysis The basic unit of analysis was defined as one visit by one Spanish surname client. Each contact was considered a single visit, including repeat and follow-up visits by the 34 same client within the year regardless of the interval between instances of service. Sample: Inclusion and Exclusion The sample population identified by the investigator consisted of those persons of Hispanic ethnicity who presented for service to the BMH ED in 1995. Hispanic ethnicity was determined initially by Spanish surname. Chents with traditional Spanish surnames were included if there were no specific indicators in the medical record contradicting Hispanic ethnicity. Chents with non-traditional surnames not otherwise indicated to be Hispanic were excluded from the chart review. Chents with Spanish-sounding but non- traditional surnames and chents with Spanish surnames and simultaneously with indicators suggesting non-Hispanic ethnicity such as an Anglo given name or indication of residence on a reservation were judged to be questionably appropriate for inclusion. Because of concerns regarding potential inaccuracies, these cases were checked using additional procedures. Records of these chents were reviewed in detail for other confirmation of ethnicity. Notes written by ED nursing staff, physician or physician's assistant identifying Hispanic ethnicity, Spanish as language of preference, use of translators, or concerns related to communication were taken into account as possible vahdation for inclusion. Additional medical records, if available, were reviewed for further indication of ethnicity. In some cases, interviews were conducted with staff. Anecdotal information from both ED and medical records department staffs was weighed in the balance when identification of ethnicity was thought to be questionable. The distinctly rural nature of the environment in which the study was conducted, the small number of visits under consideration, and the 35 access of the investigator to staff during the study period allowed the conduct of this informal inquiry as data was compiled. An example may clarify the problems of inclusion and exclusion encountered by the investigator. One male client with an unusual but Spanish-sounding surname and a distinctly Anglo given name presented three times during the study period. On none of the three records was Hispanic ethnicity or Spanish as language of preference documented by any of the attending staff. Interviews with two staff members who had provided care to the patient failed to elicit any confirmation of Hispanic ethnicity. Based on the cumulative weight of the lack of indicators, it was determined that this client should be excluded from the sample. Of the 93 ED records retrieved by MR staff, 75 visits by 52 separate individuals were included in the study. Eighteen visits were excluded for the following reasons: 1 had been pulled in error; 4 were visits by the same individual, known by a member of the MR staff to be of non-Hispanic ethnicity. Investigatory interviews revealed that 3 records represented visits by members of a family who had discussed their Italian heritage with ED staff; 10 records were excluded, including the 3 in the example mentioned above, which represented visits by persons whose surnames were not based on Spanish root sounds familiar to the investigator, and for whom no additional indication of Hispanic ethnicity could be found. 36 Chart Audit In an effort to protect the confidentiality of client records, all chart audit was conducted by the investigator in the medical records department during non-office hours during the months of December, 1995 and January, 1996. Records included in the study were assigned a case code. A code directory linking the names of clients with their case codes was retained in the MR department, which is locked at all times except when staffed by MR department employees. Data were compiled using manual transcription of relevant information to a chart audit form developed by the investigator. A sample of this instrument is included in Appendix B. The chart audit form itemized: case code numerical identifier; date, time, and duration of service; age and gender of client; location of residence; employment status; presence of insurance coverage or alternate payment source; primary provider designation, if noted; nature and circumstances of presenting complaint; medical diagnosis; management or treatment received; and providers of service. The audit process required an average of 10 to 15 minutes per chart. Organization of Data for Analysis Chart audit forms were reviewed repeatedly. Extracted data were grouped and considered from several perspectives in order to characterize the sample. Another goal was to discover any relationships which might exist between demographic characteristics of the sample (variables such as age, gender, and location of residence) and those variables demonstrated in the literature to be related to health service use (variables including employment status, existence of a payment source, preexisting relationship with a primary 37 , ; provider, and nature of the presenting complaint). The data collected in the study were reported without statistical manipulation. Numerical frequencies were compiled and comparisons among them were considered. Mary Snepenger of the Montana State University (MSU) College of Nursing Office of Applied Research Services (OARS) was retained to develop the graphical representations of these numerical descriptors which may be found in Chapter 4. Rights and Review of Human Subjects Because this study involved the investigation of confidential medical records of a potentially vulnerable population, procedures addressing the protection of human subjects were drafted and submitted for University review. Application was made to the MSU College of Nursing (CON) Human Subjects Review Committee (HSRC) in October, 1995 for permission to undertake the study. The application contained a request for exemption from the more extensive aspects of the human subjects review procedures in light of the retrospective chart review study design. This design entailed neither direct encounters with study subjects nor the necessity of revealing study participation to employers, insurance sources, or other entities with possible relationships with subjects. In response to requirements by the Committee, mechanisms for anonymous case coding and retention of the case code directory in the secure Medical Records Department were developed and implemented. Since Barrett Memorial Hospital did not have its own research or human subjects review committee, permission was sought from the BMH Chief Executive Officer (CEO) 38 to conduct the study within the facility. A letter of understanding was submitted to and approved by the CEO, recognizing the umbrella of the MSU CON Human Subjects Review Committee. Copies of submissions to the MSU CON Human Subjects review committee and the letter of approval from the BMH CEO are included in Appendix C. Rationale for Methodologic Approach The development of a process appropriate to describe the health service needs of a transient, rural, Hispanic population of uncertain dimensions presents a challenging problem. The comprehensive, multipronged approach of the HHANES study, which combined interviews and physical exams with laboratory analyses and diagnostic testing on over 16,000 participants (Delgado et al, 1990), is beyond the scope of small-scale investigations. However, in a review of the health status of migrant farmworkers. Rust (1990) suggests that there may be alternative approaches to the study of elusive populations which could be taken by the independent researcher. In a discussion of the difficulties of estimating the parameters of a fluid, in some senses invisible population whose size is utterly uncertain, a population with a "missing denominator” (p. 1214), the author states that "one study of emergency room visits for farm equipment injuries suggests a methodology for future studies" (p.1215). This statement serves as a reminder that, as noted previously, for U.S. Hispanic people in general, the most frequent source of primary care is the emergency department (Juarbe, 1995). This service utilization pattern has been found to be also true for Hispanic agricultural workers, who seek health care almost exclusively in emergency rooms (Guamaccia et al., 1992). Bechtel et al. noted that, 39 for their study population of agricultural laborers in Arizona, "the emergency room at the community hospital was considered the most accessible [source of primary care]" (1995, p. 19). Even if we cannot accurately estimate overall population size, we may be able to glean important information by concentrating on instances of contact with the health care establishment which bring into graphic relief a population which is usually inaccessible and only diffusely defined. In light of these conclusions, instances of emergency department use may be perceived as a reflection of the absence of otherwise accessible, alternative health care services. Coincident with the hope to perceive through the data a reflection of health needs experienced by this Hispanic population outside the ED environment, the ED visit rather than the individual Hispanic ED client was identified as the unit of analysis. When no relationship with a primary provider and no alternative clinic service exist, routine, episodic, acute and follow-up care must all take place within the ED context, if they are to occur at all. It was anticipated that an examination of every visit, including repeat visits by the same clients, might more readily reveal repetitive patterns that could reflect the need for routine, non-emergent services which were not otherwise accessible to the study population. Metho do logic Issues in Determination of Ethnic Identity Information regarding ethnicity was not routinely included on any part of Barrett Memorial Hospital's emergency department form. Therefore, it was necessary to develop an operational definintion of Hispanic ethnicity which would allow reasonably accurate 40 retrospective retrieval of data. Reference to reports of other studies revealed that this is not a unique dilemma in transcultural research in the United States, and suggested that one acceptable method uses Spanish surname as an indicator of Hispanic ethnicity. In the Stanford Five-City Project, Winkleby et al. (1993) report that, when information on self- identified ethnicity was unavailable, ethnicity was coded by Spanish surname. Chiapella and Feldman (1995) discuss the use of a truncated list of six-character, Spanish surname roots to identify Hispanic ethnicity in U.S. males with end-stage renal disease. Markides and Coreil (1986) report results from several studies in which Hispanic ethnicity was identified by Spanish surname. However, certain errors both of exclusion and inclusion must be recognized with implementation of this method. First, it is clear that the Hispanic individual with a non-traditional surname would elude recognition. Hispanic persons who have assumed the names of spouses who are not Hispanic, as well as those who have adapted Hispanic names to sound more "American” or who have elected to adopt less ethnically identifiable names would be lost to the study in this manner, as might their children. By the same token, persons of non-Hispanic background who have Spanish surnames would be incorrectly included. Again, this might include the non-Hispanic spouses of Hispanic individuals or non-Hispanic individuals whose names simply "sound Hispanic". In light of the recent immigration status and transient residence patterns of many Hispanic laborers described earlier in this paper, it was thought that factors related to name alterations and marriage would have minimal impact as sources of inaccuracy for the particular population under scrutiny. 41 Chiapella and Feldman (1995) point out the additional possibility of confusion and overlap with names of individuals of Italian nationality. In Beaverhead County, 4.3% of residents claim at least partial Italian ancestry (Montana Census of Population and Housing, 1990). Although all of these individuals would not be likely to have Italian surnames, consideration to this potential confounding factor was given in the record review and did constitute the basis for excluding three charts. In the end, however, the investigator agreed with Chiapella and Feldman who concluded that, "despite its limitations ... [identification of Hispanic ethnicity based on Spanish surname] provides a means of generating demographic and health data for Hispanics that are otherwise unobtainable" (1995, p. 1001). 42 CHAPTER 4 PROJECT OUTCOME The present chapter includes a report and discussion of the research findings and consideration of the strengths and limitations of the project. The chapter closes with a discussion of the study’s implications for nursing practice and health service delivery. Report of Research Findings Demographic Characteristics Gender and Age. Fifty-three Hispanic individuals made 75 emergency department (ED) visits (n=75) in 1995. Twenty-three of 75 visits (31%) were made by Hispanic females; 52 of 75 visits (69%) were made by Hispanic males. Twenty-three ED visits (23/75, 31%) by Hispanic people were made by children aged 15 years or less. Five pediatric visits were made by 5 Hispanic girls and 18 visits were made by 10 Hispanic boys. Fifty-two ED visits (52/75, 69%) by Hispanic people were made by adults aged 16 years and older. Eighteeen visits (18/75, 24%) were made by 14 Hispanic female adults, 34 visits (34/75, 45%) were made by 24 Hispanic male adults. The frequency of Hispanic ED visits in 1995 by age and gender is presented in Figure 1. 43 Figure 1. Hispanic ED Visits, 1995 Adult/Child by Gender 40 35 30 25 20 15 10 5 0 I Females ■ Males <=15 years >=16 years The mean age of Hispanic ED clients based on all Hispanic ED visits (n=75) for 1995 was 22.3 years. The youngest Hispanic client was a 2-week old female infant. The oldest client was a 65-year-old female. Hispanic males from 16 to 40 years of age accounted for 27 of 75 (36%) ED visits by Hispanic people. The frequency of Hispanic ED visits in 1995 by age (5 to 10 year increments) and gender is presented in Figure 2. Figure 2. Hispanic ED Visits, 1995 by Age and Gender less than two 2-5 6- 10 11 -15 16-20 21 -30 31 -40 41 -50 51 -60 61 -70 0 5 10 15 20 25 30 44 Location of Residence. Fifty of all ED visits by Hispanic clients (50/75, 67%) were made by individuals who resided within Beaverhead County. Twenty-four records (24/75, 32%) noted residences within 10 miles of the hospital; 26 (26/75, 35%) were more than 10 miles away. An additional 9 visits (9/75, 12%) were made by individuals who resided in adjacent counties more than 10 miles from the hospital. In total, 35 visits (35/75, 47%) were made by clients who resided within the hospital's potential catchment area but had to travel more than 10 miles for service. Fourteen visits (14/75, 19%) were made by individuals who claimed residence outside the State. Two charts lacked documentation of residence. The relative frequencies of locations of residence of Hispanic ED clients in 1995 is presented in Figure 3. Figure 3. Hispanic ED Visits, 1995 Location of Residence 45 Temporal Trends A pattern of increased ED visits by Hispanic clients during the months of June, July and August was clearly demonstrated by the data. This rise peaked decisively in July and was proportionately much sharper than the modest rise observed in the number ED visits by all clients during July, August and September. Bar graphs comparing the frequencies of all ED visits in 1995 (i.e., all races and ethnic designations) by month with frequencies of ED visits by Hispanic clients in 1995 by month are presented in Figure 4. Figure 4. Hispanic ED Visits, 1995 Comparison to Total Visits by Month Hispanics Total January February March April May June July August September October November December 0 100 200 300 400 More than twice as many ED visits were made by Hispanic males than by Hispanic females in 1995. However, visits by females almost equalled those by males in the busiest month, July, and exceeded those by males in the months of November and December. Hispanic ED visits by month and gender are presented in Figure 5. 46 Figure 5. Hispanic ED Visits, 1995 by Month and Gender 0 5 10 15 20 25 Occupation. Employment was noted in 31 Hispanic adult (16 years and older) ED visit records (31/52, 60%). Four of eighteen female adult visit records identified employment (4/18, 22%). One individual noted employment on a ranch, one noted self- employment, 2 noted employment by service businesses in town. There were no repeat visits among employed female adults. Fourteen visits were made by 10 unemployed female adults: 1 unemployed female adult made 4 ED visits in 1995; another made 2 visits. Employment was declared in 27 of 34 total male adult visit records (27/34, 79%). Twenty-one records noted employment on ranches and farms; 4 noted employment by service businesses in town; 2 records noted the military as employer. Eighteen employed males made 27 ED visits in 1995. Seven visits (7/34, 21%) were made by unemployed male adults; none of these were repeat visits. 47 Payment Source/Insurance Status An insurance payment source (including Medicaid, Medicare, Workers Compensation, Champus, Canadian National Health Plan, or private insurance) was claimed for 40 ED visits (40/75, 53%). For 35 visits (35/75,47%), no available insurance source was identified. Ten of 23 pediatric visits (10/23, 44%) were not covered by any insurance payment source, 10 indicated Medicaid as primary payor (10/23, 44%), and 3 (3/23, 13%) were covered by private, group or family insurance plans. Fourteen of 23 visits by females of all ages were not covered by any insurance payment source (14/23, 61%). Four indicated Medicaid as payor (4/23, 17%), and 5 designated private, group or family insurance plans (5/23, 22%). Twenty-one of 52 visits by males of all ages were uninsured (21/52,40%), 1 indicated Medicare and 8 indicated Medicaid as primary payors (9/52, 17%), 15 indicated Workers Compensation (15/52,29%), and 7 designated private, group or family insurance plans (7/52, 14%). Among visits by employed adults, 20 visit records (20/31, 65%) noted an insurance payment source. Eleven visits by employed adults (11/31,35%) were not covered by insurance of any kind despite employment. Twenty-two records of visits by employed adults indicated agricultural employment (22/31, 71%). Fourteen of these (14/22, 64%) indicated coverage by Workers Compensation, while 8 (8/22, 36%) lacked indication of coverage by any insurance payment source. 48 Six of 9 records of Hispanic adults who declared non-agricultural employment designated an insurance payment source (6/9, 67%). Three records declaring non- agricultural employment (3/9, 33%) failed to identity any insurance payment source. Of 21 visits by unemployed adults, 13 records (13/21, 62%) indicated no payment source, 2 indicated Medicaid, 1 indicated Medicare, and 5 indicated private, group or family insurance plans. Health care payment source by age of Hispanic ED clients is presented in Figure 6. Payment source by employment status is presented in Figure 7. Figure 6. Hispanic ED Visits, 1995 Payment Source by Age Medicare/Medicaid Other Insurance Workers Comp Uninsured I less than 16 ■ adult 0 10 20 30 40 49 Figure 7. Adult Hispanic ED Visits, 1995 Payment Source by Employment Uninsured Medicare/Medicaid Workers Comp Other Insurance 0 2 4 6 8 10 12 14 16 Relationship with a Primary Provider Hispanic clients named primary health care providers outside the ED setting in 28 records (28/75, 37%). Forty-seven records (47/75, 63%) contained no reference to a primary provider. During most of the year, the number of visit records claiming prior relationship with a primary provider exceeded visits in which no primary provider was identified. However, the proportion of records in which no primary provider was designated increased notably during the months of highest ED use in June, July and August. Frequencies of visits in which a primary provider was designated by Hispanic ED clients by month are presented in Figure 8. 50 Figure 8. Hispanic ED Visits, 1995 Primary Provider by Month 0 5 10 15 20 25 Chief Complaint Emergency Department visits by Hispanic clients were stimulated by chief complaints which fell into several distinct categories. Pediatric problems, women's health issues, motor vehicle accidents, occupational injuries, and miscellaneous adult health problems each accounted for numerous visits. Pediatric chief complaint. Pediatric problems accounted for 23 instances of ED service to Hispanic people in 1995 (23/75, 31%). Thirteen of 23 pediatric visits (13/23, 57%) involved illnesses of the respiratory tract: 7 of these were cases of otitis media; 5 visits were for asthma and pneumonitis in the same child. Four pediatric visits were accident-related: 2 involved splashes to the eyes, 1 involved removal of a large splinter from the leg and 1 resulted from a motor vehicle accident. Five pediatric visits were 51 initiated for abdominal complaints; 2 of these led to emergency appendectomies. One pediatric client was diagnosed with allergic conjunctivitis. In all, 15 children made 23 ED visits: 2 children were seen twice, 1 child was seen 3 times and, as noted above, 1 child was seen 5 times in 1995. The relative frequency of chief complaints among Hispanic pediatric clients who presented to the ED in 1995 is presented in Figure 9. Figure 9. Hispanic ED Visits, 1995 Pediatric Chief Complaint Adult chief complaint. Women's health concerns accounted for 6 ED visits by 4 Hispanic adult females (6/52, 12% of total adult ED visits). Chief complaints included abdominal pain and cramping, vaginal bleeding, questionable preterm labor, and a blood pressure check for pregnancy-induced hypertension. 52 Injuries and accidents accounted for 32 ED visits by 23 Hispanic adults (32/52, 61%). In 17 instances, the presenting or chief complaint arose from injuries sustained in a work setting (17/32, 53%); 15 of these were agriculturally-related. All Hispanic clients who presented with work-related injuries were males. Of 17 visits for work-related injuries, 16 occurred between March 21 and August 28, 1995; 1 visit occurred in October. Visits were initiated for work-related accidental injuries which included lacerations and puncture wounds (8 visits for 5 injuries), foreign bodies in and abrasion to the eye (3 visits for 3 injuries), blows and falls (2 visits for 2 injuries), trampling by a bull (2 visits for 1 injury), electrocution by lightning strike incurred while moving irrigation pipe (1 visit), and 1 motorcycle accident also incurred while moving irrigation pipe. Nine non-work-related motor vehicle accidents involving adult clients accounted for 10 individual ED visits (10/32, 31% of visits for injuries and accidents). A stab wound, hand injuries and lifting strain accounted for the remaining 5 adult ED visits arising from non-work-related accidental injuries (5/32, 16% of visits for injuries and accidents). Miscellaneous medical complaints accounted for 14 ED visits by 11 Hispanic adults in 1995 (14/52,27% of total adult ED visits). Five visits were initiated for problems of the upper respiratory tract (sinusitis, tonsilitis, bronchitis), 4 for abdominal discomfort (peptic ulcer disease and gastroenteritis). Chest pain, shortness of breath and suicidality constituted chief complaints for the remaining 5 adult visits. The relative frequency of chief complaints for adult Hispanic ED visits in 1995 are presented in Figure 10. 53 Figure 10. Hispanic ED Visits, 1995 Adult Chief Complaint 35 30 25 20 15 10 5 0 Respiratory tract Abdominal discomfort Accidents/ injuries Women's health Miscellaneous medical Discussion of Findings This study was undertaken with the hope that an improved understanding of the characteristics and health needs of Hispanic people in Beaverhead County might be derived from an analysis of their interface with the local hospital's emergency department. A rationale for this goal is presented in the first two chapters of this paper. Population Trends Fifty-three Hispanic individuals made 75 emergency department visits in 1995. Of these, 5 were girls less than 16 years of age, 10 were boys less than 16 years of age, 14 were adult females 16 or older, and 24 were adult males 16 or older. Twelve Hispanic individuals who presented for service claimed residence outside the State. Forty-one claimed residence in Beaverhead or adjacent counties including Madison and Siver Bow. 54 Perhaps the most striking feature of the data is the large increase in utilization of emergency department services by Hispanic people in the summer months of June, July and August. During those months, the average number of visits by Hispanic clients was 14.3 per month. This visit frequency is 4 times greater than the visit frequency over the other 9 months of the year (14.3 during summer months, 3.5 during non-summer months). There are several possible explanations for this increase. It may be that Hispanic people, like many others in the United States, are more inclined to travel longer distances for work and recreation during the summer season, and are therefore at greater risk for highway accidents. This would seem to be supported by the fact that six of 10 ED visits involving Hispanic victims of motor vehicle accidents occurred during the month of July, and all of those clients claimed residence outside the State. It may also be that members of a stable target population unchanged in size become more actively involved in agricultural labor, which is known to be dangerous, and therefore incur a greater risk of injury during the summer season. It should be noted that all 15 visits arising from agricultural work-related injuries occurred in the spring and summer months between late March and late August. However, it is also possible that the increased incidence of ED visits reflects the arrival of a transient migrant population of Hispanic agricultural laborers that swells the stable local population to several times its usual numbers. The 1994 estimate by one permanent local Hispanic resident of many years noted above suggested that the seasonal population of Hispanic laborers in Beaverhead County was three to four times greater than the 1990 census had reported (Carrick, 1994a). Evidence of a similar four-fold increase in ED use by Hispanic people during summer months lends further credibility to this 55 estimate. Recently reported figures from the U.S. Census Bureau describe a 9% increase in Montana's overall population, with growth in 49 of 56 counties including Beaverhead County (The Missoulian, March, 1996). Sources suggest that Hispanic populations are growing at rates up to 5 times national averages for the overall population. If national and state population trends are extrapolated to the local setting, increases may be anticipated in the numbers of both seasonal migrant Hispanic laborers and permanent Hispanic residents. Insurance Status/Pavment Source Fifty-three percent of Hispanic ED visits in 1995 (40/75) were covered by Workers Compensation, Medicare or Medicaid, and private, group, or family insurance plans. Forty-seven percent of Hispanic ED visits (35/75) were not covered by health insurance of any kind. This exceeds national estimates which indicate that 31.6% of the nation's overall Hispanic population lacked health insurance in 1993, and is in stark contrast to an estimate that only 15.3% of the population of the State of Montana as a whole was uninsured in 1993 (U.S. Bureau of the Census, 1995). The difference may provide another piece of evidence to suggest that the study sample includes a large proportion of migrant agricultural laborers who are not eligible for public health insurance (Medicaid) because of transience or lack of immigration documentation, or who are unable to negotiate the system due to language barriers. The ED form does not collect information about the nature of employment. However, the possibility that the study sample contains large numbers of migrant workers is further supported by the increase in 56 agricultural work-related accidents during spring and summer months as noted above. It is worth noting that, since it is a completely incident-related form of insurance. Workers Compensation provides only partial health coverage. Clients claiming coverage by Workers Compensation would not have been covered had injury been incurred under any circumstances other than in the context of the work environment. Furthermore, non- work-related illness or disability is not covered under Workers Compensation programs. Cultural Characteristics Commitment to family and especially to the well-being of children has been noted previously as characteristic of Hispanic culture. Unfortunately, local census data do not accurately reflect demographic characteristics of transient Hispanic people any more accurately than do the national statistics which have been discussed already in this paper. It has been observed that in Beaverhead County, similarly to patterns in other migrant agricultural settings, many seasonal male laborers do not bring family members with them during the working months, but send home earnings and return to familes at the end of the agricultural production season (D. Hagenbarth, self-employed rancher, personal communication, January, 1996; Kerr & Ritchey, 1990). As a result, it is possible that the proportion of children less than 15 years of age in the Hispanic population of the County is less than the 31% of ED visits for pediatric services, but that cultural traditions of nurturance and concern for children motivate parents to seek care more readily for their children than do most Hispanic adults for themselves (Carrick, 1994b). It is also possible, however, that use of pediatric ED services may be due to a disproportionate vulnerability 57 to troubling illnesses in children living under conditions of transience and poverty (Rodriguez, 1983; Waldman, 1994). An alternative explanation for this finding may be that higher birth rates characteristic of Hispanic populations result in the presence of more children for fewer families, despite a higher proportion of single men in the population. Finally, the potential exists that basic population patterns are changing and that previously transient laborers are establishing local residence and bringing family members to join them. Weaknesses and Strengths of the Study Ethnic Identification The lack of identification of ethnicity on the emergency department record is the most significant limitation of the study. Difficulties in distinguishing ethnicity based on surname, with resultant errors of inclusion and exclusion, have been discussed in the previous chapter. It is important to reiterate, however, that inaccuracies may have resulted from this method of identifying study subjects, and that any inaccuracies that did result could have been magnified by the small sample size. Accuracy of the Data Source Since the ED form provided the primary data source for the study, errors in its completion could have had important ramifications for the study's internal validity as well. At the study site, ED forms are completed in most cases by clinical staff who must balance priorities on physical care of patients with the need for data entry. Forms were not altered, 58 nor were staff trained or provided with additional support to ensure precision during the study period. The thoroughness with which patients were questioned was likely to have depended on the urgency of patient presentation and the abilities of both ED staff and patients to communicate. The presence or absence of an interpreter with adequate language skills would have impacted the completion of the form as would the abilities, motivation, and experience of staff members, as well as other unanticipated factors. Errors and inconsistencies in completion of the ED form were possible and may not have been recognized by the investigator. Limitations in Generalizabilitv This study is presented with the conviction that there is value in a focused analysis of a specific population in a defined locale. However, it is important to acknowledge that the unique setting and small sample size have created a study which may not be representative of any other population or environment. The potential generalizability of study findings to populations outside Beaverhead County is therefore distinctly limited. Reflection of the Target Population In any study, questions arise as to whether or not the study sample adequately represents its target population. While the rurality and the remote geography of this study setting may have contributed to modest case numbers, these factors plus the lack of alternative sources of care also increase the probability that Hispanic people in the area who had reason to seek urgent or emergent health care during the study period did so at the Barrett Memorial Hospital ED. No other reasonably accessible options existed: no 59 urgent care centers, no private providers with extended hours, no other acute care facilities within less than 50 to 60 miles in any direction. These conditions maximize the potential that most appropriate sample cases were retrieved and that, however limited the study numbers, an accurate reflection of ED use by this population during the study period is reflected. Implications of the Study It is de rigeur for a research study to conclude with recommendations for further study. Certainly, it could not be disadvantageous to increase our observations and understanding of any population or phenomenon. Since we have observed that there is a dearth of information on the health status of Hispanic people in remote, rural settings, perhaps we could only benefit from more numerous, more detailed investigations addressing these settings. However, it has been reported time and time again that health and mental health status are affected positively by health service utilization (Ruiz, 1993), health service utilization is strongly related to access (Solis et al., 1990), an essential key to access is provided by adequate health insurance (Andersen et al., 1986), and the primary barrier to adequate health insurance is poverty (Council on Scientific Affairs, 1991). There can be no justification in waiting to say that one appropriate focus for our efforts on behalf of medically underserved Hispanic populations is to work to make health services and health insurance more affordable, accessible and equitable. This is not different from the priority we should share on behalf of all medically underserved populations, on either local or 60 national levels. The large proportion of Hispanic people who were found in the current study to be totally uninsured adds urgency to this mandate. It is unfortunate, perhaps, for the relative importance of the current study that, more significant even than those who came to the ED, may be those who did not come. The thought that there exist a number of health needs marginally less than urgent or emergent for which proud, hard-working, but poor people do not seek care further supports the importance of undertaking activities to create and fund community-based programs. Suggestions for Program Design and Focused Intervention Several authors have noted that episodic use of the ED as primary provider tends to disallow the focus and continuity of care necessary for disease prevention, health and mental health promotion activities (Ruiz, 1993; Solis et al., 1990). The frequency of work- related injury in adult men, especially during summer months, is a notable pattern revealed by the study. This pattern suggests that occupational health education and accident prevention could be appropriate interventions to address the health status of Hispanic people in Beaverhead County. Access to the target population and creation of trust and rapport are major factors essential to the success of any proposed interventions. Men working 10 to 14 hours a day during spring and summer months are not likely to take or to be given time off for non-urgent clinic visits related to accident or disease prevention or health promotion education. However, admission might be gained to worksites through a collaborative approach with employers concerned about workers condensation rates, especially if effective programs could be designed which directly addressed the promotion 61 of safety in the context of actual work tasks. A substantial number of ED visits were for reasons of pediatric illness; a smaller but still notable number were for chief complaints related to women’s health. In light of an understanding that Hispanic families are culturally predisposed to make every effort to promote infant and child health and well-being, perinatal and pediatric health also provide logical targets for program development and funding. Institution of a program of services for Hispanic mothers and children might help to develop the rapport which could provide subsequent entree into the world of adult and occupational health as well. The largest proportion of ED visits observed in the current study were made by Hispanic people residing within the catchment area but more than 10 miles from the hospital, some as many as 45 minutes to an hour’s drive away. Thus any program that would attempt to reach a Hispanic working population in such a rural setting must address transportation difficulties faced by non-English speaking workers on remote ranches. A system of bus or taxi transportation to a bi-cultural health service center could constitute a useful model, but might be only cost-effective in more heavily populated areas. A mobile clinic with a bilingual clinicians or assistants could provide another option for delivery of care. Program Models The Migrant Health Outreach Team in Arizona offers one successful model of a mobile, rotating outreach clinic staffed by advanced practice nurses (LaPlante Stein, 1993). The outreach clinic is designed to overcome distance and transportation barriers. t> 62 and to decrease practical problems related to child-care and employment responsibilities in the delivery of primary care. The Adams County Migrant Clinic in Pennsylvania demonstrates expanded hours of clinic service for "migrants [who] work even if they are sick, and get care after dark" (McGreevy, 1993, p.16). Other models demonstrate the education and support of community-based peer or lay health educators who provide a resource to Hispanic communities and create a liason between Hispanic communities and health care providers (Meister, Warrick, de Zapien, & Wood, 1992; Warrick, Wood, Meister, & de Zapien, 1992). The Cornell Migrant Program has developed several innovative programs including a child-health voucher system, a nutrition education program, and development of a data base on migrant health needs (Henderson, 1992). It is encouraging that several models exist which seek to overcome the numerous barriers to improved health care for Hispanic people in the United States. Successful features of many of these and other models may prove to be adaptable to the needs of previously underserved areas and populations. Conclusion This study provides one of few analyses of the health of Hispanic people in a remotely rural United States setting. To the extent that attention is focused on the unique social and health-related circumstances of a minority population in a relatively little- investigated environment, this study may serve to expand understanding and appreciation for the variety and complexity of Hispanic lifeways, and to encourage recognition of and planning for these expanding populations in unexpected milieus. 63 REFERENCES CITED Andersen, R.M., Giachello, A.L., & Aday, L.A. (1986). Access of Hispanics to health care and cuts in services: A state-of-the-art overview. Public Health Reports. 101(31 239-252. Aguirre-Molina, M., Ramirez, A., & Ramirez, M. (1993). Health promotion and disease prevention strategies. Public Health Reports. 108(51. 559-564. Arias, M.B. (1986). The context of education for Hispanic students: An overview. American Journal of Education. 95(11. 26-57. Bechtel, G.A., Shepherd, M.A., & Rogers, P.W. (1995). Family, culture, and health practices among migrant farmworkers. Journal of Community Health Nursing. 12(1), 15-22. Blane, D. (1995). Editorial: Social determinants of health - socioeconomic status, social class, and ethnicity. American Journal of Public Health. 85(7). 903-904. Carrick, P. (1994a). Hispanic farmworkers in rural southwestern Montana. Unpublished manuscript, Montana State University. Carrick, P. (1994b). Isolation: A chameleon concept. Unpublished manuscript, Montana State University. Chiapella, A.P., & Feldman, H.I. (1995). Renal failure among male Hispanics in the United States. American Journal of Public Health. 85(7). 1001-1004. Ciesielski, S., Loomis, D.P., Mims, S.R., & Auer, A. (1994). Pesticide exposures, cholinesterase depression, and symptoms among North Carolina migrant farmworkers. American Journal of Public Health. 84(3L 446-451. Cordes, S., Doeksen, G.A., & Shaffer, R. (1994). Rural economic development and health services. In J.E. Beaulieu & D.E. Berry (Eds.), Rural health services: A management perspective (pp. 27-56). Ann Arbor, MI: AUPHA/Health Administration. 64 Council on Scientific Affairs. (1991). Hispanic health in the United States. Journal of the American Medical Association. 265(2). 248-252. Culebras, A. (1995). Controversies in neurology: Hispanic: An epidemiologically meaningless term. Archives of Neurology. 52f5J. 533-534. Dietz, Sr. M. (1991). Stressors and coping mechanisms of older rural women. In A. Bushy (Ed.), Rural Nursing. Vol. 1. (pp. 267-280). Newbury Park, CA: Sage. Delgado, J.L. & Estrada, L. (1993). Improving data collection strategies. Public Health Reports. 108f5k 540-545. Delgado, J.L., Johnson, C.L., Roy, I., & Trevino, R.M. (1990). Hispanic Health and Nutrition Examination Survey: Methodological considerations. American Journal of Public Health. SOfSuppB. 6-10. Elder, J.P., Castro, F.G., de Moor, C., Mayer, J., Candelaria, J.I., Campbell, N., Talavera, G., & Ware, L.M. (1991). Differences in cancer-risk-related behaviors in Latino and Anglo adults. Preventive Medicine. 20(6). 751-763. Estrada, A.L., Trevino, F.M., & Ray, L.A. (1990). Health care utilization barriers among Mexican Americans: Evidence from HHANES 1982-84. American Journal of Public Health. 80fSuppO. 27-31. Flegal, K.M., Ezzati, T.M., Harris, M.I., Haynes, S.G., Juarez, R.Z., Knowler, W.C., Perez-Stable, E.J., & Stem, M.P. (1991). Prevalence of diabetes in Mexican Americans, Cubans, and Puerto Ricans from the Hispanic Health and Nutrition Examination Survey, 1982-1984. Diabetes Care. 14(7. Suppl. 3), 628-638. Fox, G. (1988). Hispanic communities in the United States. Latin American Research Review. 23f3L 227-236. Funkhouser, S.W. & Moser, D.K. (1990). Is health care racist? Advances in Nursing Science. 12(2). 47-55. Furino, A., & Munoz, E. (1991). Health status among Hispanics: Major themes and new priorities. Journal of the American Medical Association. 265(2). 255-257. Ginzberg, E. (1991). Access to health care for Hispanics. Journal of the American Medical Association. 265(21. 238-241. 65 Giuliano, A. & Alberts, D. (1994). Cancer prevention among US Hispanics [Editorial]. Archives of Internal Medicine. 154fMav 23). 1057-1058. Gonzalez, J.T., Atwood, J., Garcia, J.A., & Meyskens, F.L. (1989). Hispanics and cancer preventive behavior: The development of a behavioral model and its policy implications. Journal of Health & Social Policy. 1(2), 55-73. ’ Guamaccia, P.J., Angel, J.L., & Angel, R. (1992). The impacts of farm work on health: Analysis of the Hispanic Health and Nutrition Examination Survey. International Migration Review. 26(1 T 111-132. Haffner, L. (1992). Cross-cultural medicine a decade later: Translation is not enough. Western Journal of Medicine. 153(3T 255-259. Hamman, R.F., Franklin, G.A., Mayer, E.J., Marshall, S.M., Marshall, J.A., Baxter, J., & Kahn, L.B. (1991). Microvascular complications of NIDDM in Hispanics and Non-Hispanic Whites: San Luis Valley diabetes study. Diabetes Care. 14(7. Suppl. 3), 655-664. Hanis, C.L., Hewett-Emmett, D., Bertin, T.K., & Schull, W.J. (1991). Origins of U.S. Hispanics: Implications for diabetes. Diabetes Care. 14(7. suppl. 3), 618-627. Henderson, Z.P. (1992). Human development: Migrant program breaks down social barriers. Human Ecology Forum. 20(3). 29-31. Higginbotham, J.C., Trevino, F.M., & Ray, L.A. (1990). Utilization of curanderos by Mexican Americans: Prevalence and Predictors: Findings from HHANES 1982-84. American Journal of Public Health. 80 (Suppl), 32-35. Juarbe, T.C. (1995). Access to health care for Hispanic women: A primary health care perspective. Nursing Outlook. 430). 23-28. Kerr, M.J. & Ritchey, D.A. (1990). Health-promoting lifestyles of English- speaking and Spanish-speaking Mexican-American migrant farm workers. Public Health NursingJZ(2), 80-87. Kirkman-Liff, B. & Mondragon, D. (1991). Language of interview: Relevance for research of Southwest Hispanics. American Journal of Public Health. 81(11), 1399-1404. LaPlante Stein, L.M. (1993). Health care delivery to farmworkers in the southwest: An innovative nursing clinic. Journal of the American Academy of Nurse Practitioners. 5(3). 119-124. 66 Markides, K.S. & Coreil, J. (1986). The health of Hispanics in the Southwestern United States: An epidemiologic paradox. Public Health Reports. 10U3T 253-265. McGreevy, J.C. (1993). Physicians in public health: Caring for the health and culture of migrant workers. Pennsylvania Medicine. 16-17. McKinley, A. (1991). Suicide among the rural elderly: An Arizona perspective. In A. Bushy (Ed.), Rural nursing. Vol. 1. (pp. 297-308). Newbury Park, CA: Sage. Meister, J.S. (1991). The health of migrant farm workers. Occupational Medicine. 6(3), 503-518. Meister, J.S., Warrick, L.H., de Zapien, J.G., & Wood, A.H. (1992). Using health workers: Case study of a community-based prenatal intervention. Journal of Community Health, 17(0. 37-51. Mobed, K., Gold, E.B., & Schenker, M.B. (1992). Occupational health problems among migrant and seasonal farm workers. In Cross-cultural medicine - a decade later [Special Issue]. The Western Journal of Medicine. 157(31. 367-373. Montana Census of Population and Housing: 1990. (Summary File 1), [Electronic data tape]. (1990). Helena, MT. Montana Department of Health and Environmental Sciences (MDHES). (1993). Statewide countv-based primary health care plans (Vol. 1). Helena, MT. Morning Edition. (1996). The current state of immigration legislation: A one-week series. New York and Washington, DC: National Public Radio. Moore, J.W. (1985). Isolation and stigmatization in the development of an underclass: The case of Chicano gangs in East Los Angeles. Social Problems. 33(0.1-12. Moses, M. (1989). Pesticide-related health problems and farmworkers. AAOHN Journal. 37(3L 115-130. National Center for Health Statistics. (1995). Vital Statistics of the United States, 1991: Vol. 1. Natality. Washington, DC: Department of Health and Human Services. Novello, A.C., & Soto-Torres, L.E. (1993). One voice, one vision - uniting to improve Hispanic-Latino health [Editorial]. Public Health Reports. 108(5L 529-533. Novello, A.D., Wise, P.H., & Kleinman, D.V. (1991). Hispanic health: Time for data, time for action. Journal of the American Medical Association. 265(2L 253-255. 67 Padgett, R., & Barms, A.G. (1992). Registered nurses' perceptions of their communication with Spanish-speaking migrant farmworkers in North Carolina: An exploratory study. Public Health Nursing. 9(31. 193rl99. Pamies, R.J. & Woodard, L.J. (1992). Cancer in socioeconomically disadvantaged populations. Primary Care: Clinics in Office Practice. j£(3), 443-450. Piedra, J. (1992). The conglomerate world of Hispanics [Introduction]. Human Ecology Forum. 20(31 5. Rodriguez, J. (1983). Mexican Americans: Factors influencing health practices. Journal of School Health. 53(2T 136-139. Rodriguez, R. (1982). Hunger of memory: The education of Richard Rodriguez: An autobiography. Boston: David R. Godine. Ruiz, P. (1993). Access to health care for uninsured Hispanics: Policy recommendations. Hospital and Community Psychiatry. 4400). 958-962. Rust, G.S. (1990). Health status of migrant farmworkers: A literature review and commentary. American Journal of Public Health. 80(101. 1213-1217. Sacco, R.L. (1995). Controversies in neurology: Hispanic: An epidemiologic meaningful term needed for public health planning. Archives of Neurology. 52f5). 531- 532. Schlosser, E. (1995). In the strawberry fields. Atlantic Monthly. Nov.. 80, 82, 84- 87, 90, 92, 94-96, 98-99,102, 104,106-108. Shaver, C.S., & Tong, T. (1991). Chemical hazards to agricultural workers. Occupational Medicine. 6(31. 391-413. Smart, J.F., & Smart, D.W. (1994). The rehabilitation of Hispanics experiencing acculturative stress: implications for practice. The Journal of Rehabilitation. 60(4). 8-13. Smith, J.C., Mercy, J.A., & Rosenberg, M.L. (1986). Suicide and homicide among Hispanics in the Southwest. Public Health Reports. 101(3). 265-270. Solis, J.M., Marks, G., Garcia, M., & Shelton, D. (1990). Acculturation, access to care, and use of preventive services by Hispanics: Findings from HHANES 1982-84. American Journal of Public Health. 80(Suppl). 11-19. 68 Stein, J.A., Fox, S.A., & Murata, P.J. (1991). The influence of ethnicity, socioeconomic status, and psychological barriers on use of mammography. Journal of Health and Social Behavior. 32(2). 101-113. Suarez, Z.E. (1992). Use of self-care by Hispanics: Culture, access or need? Journal of Health & Social Policy. 4(2). 31-45. Torres, S. (1993). Nursing care of low-income battered Hispanic pregnant women. AWHONN's Clinical Issues in Perinatal and Women's Health Nursing. 4(31 416-423. United States Bureau of the Census. (1995). Statistical abstract of the United States: 1995 (115th ed.). Washington, DC: Department of Health and Human Services. Valdez, R.B., Giachello, A., Rodriguez-Trias, H., Gomez, P., & de la Rocha, C. (1993). Improving access to health care in Latino communities. Public Health Reports. 108(5), 534-539. Vaughan, E. (1993). Individual and cultural diflerences in adaptation to environmental risks. American Psychologist. 48f6L 673-680. Waldman, H.B. (1994). Invisible children: The children of migrant farm workers. Journal of Dentistry for Children. 61(3), 218-221. Warner, D.C. (1991). Health issues at the US-Mexican border. Journal of the American Medical Association. 265(2L 242-247. Warrick, L.H., Wood, A.H., Meister, J.S., & de Zapien, J.G. (1992). Evaluation of a peer health worker prenatal outreach and education program for Hispanic farmworker families. Journal of Community Health. 12(1), 13-26. Winkleby, M.A., Fortmann, S.P., & Rockhill, B. (1993). Health-related risk factors in a sample of Hispanics and Whites matched on sociodemographic characteristics: The Stanford five-city project. American Journal of Epidemiology. 137. 1365-75. Woloshin, S., Bickell, N.A., Schwartz, L.M., Gany, F., & Welch, H.G. (1995). Language barriers in medicine in the United States. Journal of the American Medical Association. 273(9). 724-728. Zahm, S.H., & Blair, A. (1993). Cancer among migrant and seasonal farmworkers: An epidemiologic review and research agenda. American Journal of Industrial Medicine. 24, 753-766. 69 APPENDICES 70 APPENDIX A MAP OF MONTANA WITH TOPOGRAPHICAL FEATURES BEAVERHEAD COUNTY HIGHLIGHTED LOCATION OF ACUTE CARE FACILITIES NOTED 71 Potomac FOREST Sjv I Canyon VCreeky i Ghost fcwn i Ur Baw, HELENA NATIONAL FOREST ^ Smith — MONTANA NATION* £or?fHELENAY ii£tr Lqke ' -^jp^Canyon Ferry Ej/400] MulUn Pit? - , El. 5902 ^ ENIUpn Drummond a r-ia 394a 6/ “ Hall // Golc NATIONAL n^fsiaU White Sulphur ^rings iSprsigs' Maxville DEERLODG* i Winston Hidden Peak if '^Canyon ' Ferry Lake FOREST J HELENAJ uofferaonCity' NATIONAL Pinesdale [Corvallis f Philipsburg fichu I-aM* BITTERROOT • "S ipy Man: - \ v i ^HaIrn'!.ton D Porters | *1) Corner Hint Creek Dam ^ 'AGeorgetowi / \\Townsend f \\l1 f 3813 B^R JO A\D W A T E 1 Toston Southern jrantsdale BITTERROOT Radersburg ML EVMJ Anaconda EL 10635 * Mt. Hajgin . Uoston Dam ^Lombard Opportunity i ’ Walhervill. NATIONAL FOREST FOREST1 ViSSOlilj L o/D G E stjTT3j ^ [. DEEBlWt ikL V T ^NATIONAL ->sThi»y .dgr’i OVERHEAD/,' i MS and Clark Caverns a. BEAVERHEAD i?: ; >»• . /* ns AHc