Choice of birth setting among urban Bolivian women

Material Information

Choice of birth setting among urban Bolivian women
Boender, Carol Lynn
Place of Publication:
Denver, CO
University of Colorado Denver
Publication Date:
Physical Description:
viii, 74 leaves : ; 28 cm

Thesis/Dissertation Information

Master's ( Master of Arts)
Degree Grantor:
University of Colorado Denver
Degree Divisions:
Department of Anthropology, CU Denver
Degree Disciplines:
Committee Chair:
Moore, Lorna G.
Committee Co-Chair:
Brett, John
Committee Members:
Corbett, Kitty


Subjects / Keywords:
Childbirth -- Social aspects -- Bolivia ( lcsh )
Childbirth -- Economic aspects -- Bolivia ( lcsh )
Birth customs -- Bolivia ( lcsh )
Childbirth at home -- Bolivia ( lcsh )
Birth customs ( fast )
Childbirth at home ( fast )
Childbirth -- Economic aspects ( fast )
Childbirth -- Social aspects ( fast )
Bolivia ( fast )
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )


Includes bibliographical references (leaves 68-74).
General Note:
Department of Anthropology
Statement of Responsibility:
by Carol Lynn Boender.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
47120906 ( OCLC )
LD1190.L43 2000m .B63 ( lcc )


This item has the following downloads:

Full Text
Carol Lynn Boender B.A., University of Colorado, 1994
A thesis submitted to the University of Colorado at Denver in partial fulfillment of the requirements for the degree of Master of Arts Anthropology 2000

This thesis for the Master of Arts degree by
Carol Lynn Boender has been approved
l / zarrO Date

Boender, Carol Lynn (M.A., Anthropology)
Choice of Birth Setting Among Urban Bolivian Women Thesis directed by Professor Loma G. Moore
In urban Bolivia, many women give birth at home, even after receiving prenatal care. A two-phase study designed to discover the reasons why women-deliver at home included a medical records review and ethnographic interviewing at a La Paz hospital. The prenatal records of 170 women, half of whom delivered at home and half of whom delivered in the hospital, showed that home-birthing women were shorter, were more likely to be residents of El Alto (La Pazs neighboring city) and of Indian background, and had higher gravidity and parity, fewer prenatal visits, less education and income, more deceased children, fewer, salaried occupations, and less access to city water and sewage services. Logistic regression analysis showed that the lack of sewage systems in the home, El Alto residence, and low education levels were significant predictors of birth setting. The second phase used ethnographic interviewing and focused on Aymaran women and their decision-making processes about birth setting. Birth at home and in the hospital is described, including technologies and techniques, recognition of risk factors, and the role of obstetric experts. Interviewing with twenty home-birthing and seven hospital-birthing women found that, for this sample, economic factors did not explain choice of birth setting directly. All perinatal health care is free to Bolivian women, and home- and hospital-birthing women had similar access to transportation to the hospital and ability to incur the costs of missed work and household activities. A difference between home- and hospital-birthing women was that hospital-birthing women had overwhelmingly grown up in the city whereas the majority of home-birthing women had migrated from rural areas as adults. Place of origin was associated with the way a woman weighed a set of ideological and situational factors that affect birth setting, including i deologies of obstetrical risk, risk perception during a birth event, and the logistical difficulties associated with transport to the hospital. Socioeconomic differences between home- and hospital-birthing women were likely to be associated with place of origin as well, since rural-to-urban migrants tend to be poor.

This abstract accurately represents the content of the candidates thesis. I recommend its publication.

Firstly, I .wish to thank the women of Bolivia who were generous with their time and willing to speak openly about their lives to a generally clueless and sometimes, nosy stranger. The directors of Matemologico 18 de MayoRoberto Bohrt, Johnny Gonzales, and Silvia Vargasnot only opened their doors to me, but were supportive of my research in every way. Without Licenciada Lourdes Torres, the director of social work at Matemologico 18 de Mayo, my studies at the hospital would have been impossible. Lie. Torres introduced me to patients, gave me the use of her office, pointed me to literature sources, put me in contact with important hospital personnel, invited me to parties, and, most critically,,displayed amazing patience in helping me to understand the workings of the hospital and the national health care system, Other personnel in medical records and social work and the nursing and physician staff provided valuable assistance, made me feel welcome, and permitted me to look over their shoulders. I would also like to thank Mercedes Villena and Enrique Vargas of the Institute Boliviano de Biologia de Altura (TBBA) for providing institutional and personal support during my work in La Paz.
Of course, I never would have come to know any of these folks in Bolivia if Loma G. Moore had not invited me to participate in her research there and convinced IBB A and Matemologico 18 de Mayo that my project was worthwhile. Dr. Moore, John Brett, and Kitty Corbett have provided essential guidance for this research, from the funding stage through to the completion of this paper. The ideas they offered me have enriched my understanding of the issues presented here, not to mention the sometimes esoteric rales involved with writing academic papers. Other members of the Department of Anthropology assisted in important ways, including Connie Turner. Fernando Armazas assistance with data analysis was indispensable* as he provided an insiders knowledge of Bolivian society.
Funding for this project was provided by a Grant-in-Aid of Research from Sigma Xi, the. Scientific Research Society, and by the Benjamin Brown Award for International Study of the University of Colorado. Finally, I would like to thank Denes Opi Koromzay for sharing his home and his stories with me while I was writing this thesis.

Tables ....................................................viii
1. INTRODUCTION............................................ 1
2. BACKGROUND ........................................... 4
Literature Review...................................... 4
Materialist Models............................... 5
Ideological Models............................... 6
Multi-Factorial and Social Network Models.........10
The Study Site........................................ 12
Bolivia....................................... 12
La Paz/El Alto and Urban-Aymara Culture...........17
3. METHODS.................................................. 20
Research Design................................... 20
Phase I: Medical Chart Review......................... 21
Samples......................................... 21
Procedures and Analyses...........................23
Phase II: Ethnographic Interviewing and Observation... 24

4. RESULTS................................................29
Phase 1..............................................29
Phase II.............................................33
Ethnographic Description............................ 41
Hospital Birth.................................41
Homebirth.................................... 45
Balancing Ideologies and Situational Factors...55
Case Studies...................................58
5. CONCLUSION......................................... 65
WORKS CITED.............................................. 68

2.1 Percentages of homes served with basic services and consumer goods.....13
2.2 Reproduction patterns among Bolivian women, measured by averages of
various indicators...........................................................13
2.3 Maternal and child health indices for Bolivia, La Paz Department, and
El Alto......................................................................15
2.4 First-languages of La Paz and El Alto residents, as percentages of each
citys population............................................................17
3.1 Number of cases in samples and sub-samples.............................21
3.2 Variables used in medical chart review.................................22
3.3 Additional variables for interviewed groups.............................25
3.4 Topics discussed at interviews........................................ 26
4.1 Results of t-tests and chi-square tests for chart review groups........29
4.2 Odds ratios for having a home birth....................................32
4.3 Ethnic characteristics among interviewees...............................34
4.4 Tabulations of medical chart variables for home interviewed, hospital
interviewed, combined interviewed, and chart-review indian groups............34
4.5 Tabulations of additional variables of interviewed groups..............38
4.6 Place of origin/history of migration among interviewees. ..............38
4.7 Settings and primary helpers for previous births of interviewees.......53

In the summer of 1999,1 had the opportunity to assist with data collection in Bolivia for a project entitled Tnterpopulational differences in intrauterine growth restriction (IUGR) at high altitude.* Loma Grindlay Moore of the University of Colorado Health Sciences Center was the principal investigator of this project, collaborating with Enrique Vargas and Mercedes Villena of the Bolivian Institute of High-Altitude Biology in La Paz. The project sought to discover whether rates of pregnancy complications associated with high altitude, such as intrauterine growth restriction and pre^eclampsia, were less common among women of high-altitude ancestry, compared with women whose ancestors had lived at high-altitude for only a few generations. Data were collected from thousands of prenatal and delivery records in five Bolivian cities of various altitudes. Occasionally we came upon records indicating a woman had attended prenatal care but birthed the baby at home. These cases had to be thrown out of the study because very little information was available about the delivery. Licenciada Lourdes Torres, the director of social work at Hospital 18 de Mayo, the largest maternity hospital of Bolivias national health care service and the study site for La Paz, told us that homebirths were very common in Bolivia. Indeed, the National Demography and Health Survey reports that 43% of births happen at home (ENDS A 1998) Prenatal patients of Hospital 18 de Mayo who birth at home usually come to the hospital about a week after the birth for postpartum and neonatal check-ups. Lie. Torres saw homebirth as a large problem needing investigation. She had kept a list of home-birthing patients, making it possible to collect data on them. These women, then, are the focus ofthis study. Why do they deliver at home, even after attending prenatal care?
This question is important to the future of maternal and infant health in Bolivia. Rates of infant and maternal mortality are the highest in Latin America (Pan American Health Organization 1999).;Those rates are highest in rural areas, where maternity clinics and health posts are scarce and most births happen at home. Yet, even in urban areas, 24% of births happen at home (ENDSA T998).; There have not been studies comparing the outcomes of home and hospital births in Bolivia, but
' Funded by the Fogarty International Research Collaborative Award, project number PAR-99-008.

research suggests that less than half of Bolivian women recognize important signs of complication during labor.and delivery (MotherCare 199-8), and as few as 11% of homebirths are attended by a midwife or a trained health worker (ENDSA 1998). Identifying danger signs during the birth process is essential to preventing maternal and child death.
The question is important theoretically, as well. How does any woman, not just a Bolivian woman, choose obstetrical services? What do different women choose and why? What factors constrain their alternatives? There are few.medical anthropology inquiries into these questions. Furthermore, their answers have implications for medical decision-making beyond, obstetrics.
The majority of medical anthropology studies take place in societies that have undergone rapid social change and seen an increase in Western-derived biomedical health care services, even while pre-existing forms of health care persist. The study presented here is no exception. In Western-derived obstetrics, pregnancy and childbirth are normatively understood as medically pathological events (see Hahn 1987; Martin 1987). Yet in most.cultures, childbirth is not highly medicalized or considered dangerous, and normal pregnancy care and birth attendance by a traditional birth attendant or family member may not fall into the cultural domain of health and illness. Instead, they may belong to the realm of womens work and knowledge, family and community life, or religion and rituaLevents.
What makes this study unusual within the medical anthropology literature is that it is concerned with an urban population. It searches for the key factors relating to choice between Western-derived obstetrical services (also referred to in this paper as biomedical services) and birth care at home among urban Bolivian women, most specifically, indian women. The majority of writings on Bolivias indians limit their scope to Aymara Indians or Quechua Indians without addressing the relationships, similarities, and differences between indian groups. Based in the literature on La Paz/El Alto and the findings of the study presented here, this paper focuses on Aymara Indians. Yet, statements about Aymaras included in it may be more broadly relevant to indigenous women in Bolivia, particularly Quechua women,
I hypothesized that women who birth at home in La Paz/El Alto do so because they are less able to incur the economic, costs of hospital birth, including the costs of missed work or familial obligations. The work was conducted in two phases. The first phase was a medical at Hospital 18 de Mayo in La Paz.comparing women birthing at home with women birthing in the hospital during the same time period. Phase II consisted of ethnographic interviewing with indian women and intended to address the question, how do indian women make decisions about birth

setting? Because the differences among indian women are more complex and less apparent than differences between indian and mestizo women* this question promised to produce important data. Altogether, this paper examines differences between women who birth at home and women who birth in the hospital, traditional beliefs about childbirth and obstetrical risk, material barriers to hospital care, and differences between home and hospital birth.

Literature Review
With the advent of the field of medical anthropology in the 1960s, an early concern was medical pluralism. Biomedicine was spreading quickly around the globe due to its demonstrated success in treating acute disorders and vaccinating against communicable diseases. International health and development projects bloomed as formerly colonized nations won independence and sought assistance for their new governments. The proliferation of biomedical health clinics and hospitals brought new options to local people, adding to the pluralism of pre-existing indigenous health systems. The question for health care practitioners and development planners was, how can we get people to discover our services? While health care professionals may have believed that biomedicine would be adopted wholesale around the world due to its efficacy, medical anthropologists, knowing that health systems are in part metamedical and involve cultural domains (see Worsley 1982), questioned this assumption and began to investigate the actual utilization of indigenous and biomedical services in health seeking behavior. Of course, the therapy sought in an individual case may be multifaceted, employing more than one kind of service. This pattern may even be most common around the world:
In cultures of medical pluralism, more than in the West, with its asymmetrical attitudes to other medical systems and its professional monopolies, the patient moves back and forth in a series ofepisodes not only between agents, but also between systems (Worsley 1982:324; see also Kroeger 1982:153).
The simultaneous use of services is noted in obstetrical studies as well (Sargent 1989:46; Bradby 1999:299).
The medical anthropological literature on therapy seeking behavior displays a theoretical rift (see Kroeger 1982:157; Sargent 1989:17-22; Nations and Rebhun 1988: 1113); Some writers view materialist and political economic factors as the central influences on health care services. Others focus on ideological or

ethnomedical meaning systems as the most salient level of explanation of choice, including beliefs about religion^ illness, or treatment itself. Some models cross this theoretical dividing line and argue that beliefs and material realities interactively affect medical choice. Browner and Sargent (1996) note that childbirth studies are sorely lacking in the field of medical anthropology, due in part to the fact that most anthropologists have been male and therefore restricted from witnessing births, in many cultures. In this section, I will give an overview of materialist, ideological, and alternative models of choice of health care service, focusing on Latin America and on the small number of studies concerned with pregnancy care and the birth event. This paper does not consider the closely related topics of abortion, contraception, .neonatal care, and breastfeeding.
Materialist Models
Political economic approaches often relate individual choices about health care to national and international structures. Navarro (1974) writes that economic development has often been misguided in Latin America, resulting in poorer health for the poor. Misdevelopment, as it has been called, involves the unequal allocation of resources and leads to a widening of the gap between rich and poor (Winkelstein 1992:932; Nations and Rebhun 1996). In Brazil, the development processes that brought the nation s gross national product to the highest within Latin America also concentrated 68% of the wealth in the hands of the richest 20% of the populace, leaving 2.1% to the nations poorest20% (Nations and Rebhun 1996:7).. As the poor became poorer, their access to medical resources diminished in relation to the wealthy. National health funds went to providing the best services to the wealthiest rather than adequate services to all. Nations and Rebhun write that lack of medical care is a factor contributing to infant mortality. Scheper-Hughes (1992) formulates a somewhat different explanation for infant mortality in Northeast Brazil. Because mothers are faced with finite resources to provide for their children, only infants who appear robust will receive investments of medical attention.
Allocation of health funds in Latin America determines the density of medical centers and personnel and the level of costs to the patient associated with their use. Generally speaking, urbanites have greater access to services (Thaddeus and Maine 1994:1100). In rural areas of developing nations, centers are often sparsely dispersed. Thaddeus and Maine find that geographic distance to medical services may be a disincentive to seek care, especially where means of transportation are not widely available (1994:1102). In Cuba* when the number of health centers increased, mortality and morbi dity i mproved. Surprisingly, few studies of the utilization of medical services find that costs of the services, costs of transportation to service sites,

or opportunity costs of time spent away from normal activities were important factors in seeking care (Thaddeus and Maine 1994:1100). Yet, Goforth concludes that Mayan women choose between traditional and biomedical practitioners based on access to economic resources (1988), a rare model within the literature on obstetrical decisionmaking.
Ideological Models
Another school of analysis relates cultural beliefs and cognitive factors to health care. Cognitive factors may relate to Lewiss description of the culture of Latin American urban migrants as a culture of poverty (1959). The culture of poverty incorporates fatalistic attitudes about education, health, work, and death resulting from limited opportunities to improve quality of life Kleinmans work (1978; 1988) also centers on cognitive models (explanatory models) of disease. Patients and medical practitioners hold different models and must come to understand their differences in order that clinical treatment be effective.
Cultural beliefs, especially beliefs about illness etiologies, are found to be associated with choice of health care service in a medically pluralistic society. For example, Foster (1976) and Colson (1971) both conclude that the choice of healer differs according to whether the illness was considered to be caused by natural or supernatural forces. In her monograph of the Bariba of Benin, Sargent (1982;1988) writes that, in rural areas, the birth of a witch baby requires assistance from a traditional practitioner, who usually will help the mother end the infants life in a ritualistic manner. Yet, in urban areas, where a greater portion of births take place in the hospital and infanticide would be detected by government workers, traditional practitioners have found ways to cure the infant without producing its death.
Bolivian women, like Bariba women, succeed in maintaining a continuity of goals and strategies between traditional ways of birth and hospital birth. Bradby (1999) uses an interactionist model of knowledge that views childbirth knowledge as interwoven in the meeting of two models of birth. Many Quechua Indian women of Bolivia, who previously labored and birthed at home with the very limited aid of older female family members, are now choosing hospital birth. They evaluate it positively due to fears of bleeding too much at home and the conceptual congruity possible between traditional birth and medicalized birth. In the traditional system, the use of herbal teas was linked to beliefs about the benefits of wet substances. Therefore, intravenous fluids like oxytocin are easily accepted by women in the hospital Furthermore, women seem to prefer giving birth to the baby alone but receiving attention for the birth of the placenta. Womens stories of birth in the

hospital show that they manage to push the baby out without drawing the attention of hospital workers. The babys cries then alert workers to the birth, who attend the delivery of the placenta.
Bradbys analysis insightfully analyzes cultural models of birth in Bolivia.
But it overlooks the effects of structural elements, such as Bolivias recent public health drive to increase hospital births. The article goes no farther than noting that this is a first generation of hospital birthers, subjected to educational programs and incentives to attend hospital. Women in the study identified two strong risks of death associated with birth settingcesarean delivery, which women believe is very dangerous due to the use of metal and the opening created in the body, and excessive bleeding at home. How do they reconcile these risks when talking to a doctor? To an anthropologist based at the public clinic? To each other in their homes?
The articles most important contribution, however, is its critique of acculturation to biomedicine. When individuals from an indigenous culture begin to use biomedical services, we cannot assume it means they have accepted biomedical models of the body and/or explanations of how therapies work. Rather, biomedical obstetrical techniques are re-interpreted.
Davis ^Floyds Birth as an American Rite ofPassage (1992) provides a similar view of how women reject and accept cultural paradigms of medicine. In her unique approach to medical pluralism, individual adherence to either the wholistic or technocratic model of birth is seen to determine the choice of birth setting (home or hospital) for U.S. women. Thus, for Davis-Floyd, this choice is strictly an ideological one, but also a choice limited by an ideological hegemony of the technocratic model of birth.
Modem biomedical birth is treated as an ideological mechanism of dominant cultural metaphors. Obstetrical practices are handled as rituals. Because research shows that many biomedical obstetrical techniques fail to improve obstetrical outcomes, their continued use must lie in their ritual value. The birthing woman is a liminal individual, standing on the verge of motherhood. It is the process of birth that changes her status in society and prepares her to hand down the dominant ideologies of her culture to her children. Martin has previously described the biomedical model of birth as mechanical and technocratic (1987), and Hahns examination locates this mechanical model of the body in William s Obstetrics (1987). Birth as a rite of passage into this model entails ritualistic elements, such as repetition, to create a stage one phase of cognitive functioning, aided by the naturally stressful process of labor. The woman is inculcated in the belief that her body is a defective machine, desperately in need of biomedicine to survive (1992:12).

Davis-Floyd also analyzes how women receive the ideological messages imparted through the birth process. Women whose models of birth matched that of the setting in which they deliver (home or hospital) experienced no cognitive conflicts. But, women in-between who birthed in the hospital underwent conceptual fusion with the technocratic model. This happened either with cognitive ease, indicating that the woman really held the technocratic model all along, or with cognitive distress, indicating a true incompatibility of the womans own model with the technocratic model (1992:155).
Ideologies of risk have been found to be important in obstetrical decisionmaking. Kaufert and ONeils (1993) analysis of opposing ideologies of obstetrical risk uncovers how they relate politically to medicine, clinical practice, and epidemiology. In the early 1980s, the Canadian government began mandatory evacuations out of the Northern Inuit communities for pregnant women. The justification for this law was the lack of adequate hospital facilities for dealing with emergency situations in the Northern communities. At community meetings between local women and health care professionals from outside the area, Kaufert and ONeil witnessed the juxtaposition of the clinical language of risk and the Inuit language of risk. Doctors spoke clinically, placing the responsibility for the lives of patients on the physician by recounting episodes of maternal death and quoting statistical artifacts. Inuit women limited the definition of risk to local experience and accepted death as one possible natural outcome of birth. Daviss (1997) adds that clinical understandings carry more political power than do others.
In her study of amniocentesis, Rapp (1988,1993) focuses on how women pick and choose among a set of ideologies about amniocentesis to explain their decisions to undergo it or not. Rapp identifies seven ideological forces at work in the context of amniocentesis decisions (1993). Women rarely are found to draw on just one ideology in their rationales for deciding for or against amniocentesis and/or abortion following amniocentesis results. Rather, a woman may employ different explanations for her decision at different moments. For example, a Pentecostal woman who underwent amniocentesis said she would not consider abortion because of her religious beliefs. But at another time, she told Rapp that her husband had had problems with infidelity and illegal drug sales, and that religious fervor and having a baby were her way of keeping him at home (1993:64). Markens, Browner, and Press (1999) found similar results.
Rapps analysis is insightful for its demonstration of how women act in the midst of heterogeneous ideologies, how multiple discourses create both boundaries on choice and resources for personal understandings of choice. Akin to Hunt, Jordan, and Irwins critique of static explanatory models of illness (1988:945-46), Rapp also

shows how past obstetric choice is continually reconceptualized according to changing exigencies.
The concept of authoritative knowledge provides another way of examining the relationship between ideas about health and obstetrical decision-making. This concept is used to examine how a particular ideology gains precedence over other ways of thinking and knowing about childbirth, and how it therefore constrains the options that are available to women. This approach, which began with JordansBirth in Four Cultures (1978), views culture as socially distributed knowledge and/or the communication of knowledge (Duranti 1997:30-33). Therefore, conflicting cultural conceptions and ideologies of birth espouse different ways of knowing about birth. The concept of authoritative knowledge directly acknowledges the existence of multiple ways of knowing within even seemingly unified medical systems.
The central observation is that for any particular domain, several knowledge systems exist, some of which, by consensus, come to carry more weight than others, either because they explain the state of the world better for the purposes at hand (efficacy) or because they are associated with a stronger power base (structural superiority) and usually both (Jordan [1978] 1993:152).
Thus, even in the United States, where medical authority is seldom challenged, the birthing room becomes a meeting site for contradictory sources of information. The womans perceptions of her body, a kind of self-knowledge, are mostly irrelevant to the course of care provided by the obstetricians and obstetrical nurses. Rapp writes, Tn childbirth, authoritative knowledge in high-tech America takes the form of active suppression of whatever it is that women might know, think, or imagine about themselves in the birth process (cited in Jordan 1997:55). So, for example, when a woman is certain that it is time to push, the nurse will correct her with the medically sanctioned knowledge that her cervix is not yet completely dilated, or that it cannot be time to push because the doctor is not yet there to deliver the baby. The concept of authoritative knowledge has been used most often for these types of situations, in which ways of knowing are hierarchically distributed and an already established medical hegemony upholds its privileged position, naturalizing its authority through mechanisms of touch, spatial relations, linguistic phenomenon, eye-contact, and direction of speech (Jordan 1993:154-166; Jordan 1997; Kitzinger 1997).
Technology has also been recognized as a powerful tool of medically authoritative information. Koenig (1988) describes the technological imperative in biomedicine that results in the rapid routinization of a new technology, before its usefulness in treatment has been validated. Because biomedical culture deems that we

must use a technology if it is available and that the information provided by that technology is objective, those with the power to control and interpret the technology hold the key to knowledge^about health, This keyisjnore than merely withholding or failing to disclose relevant facts about a case [but is also] the providers ability to create those facts (Koenig 1988:489; italics in original). Koenig:also finds that the technological imperative becomes a moral imperative in the technologys routinization. Because physicians direct the use of medical technology, they control knowledge in obstetrical systems that are highly technologized.
Multi-Factorial and Social Network Models
Other writers have criticized theoretical approaches that are limited to political economy or belief systems* arguing that beliefs and material realities interactively affect medical choice (Rroegerl982; Sargent 1982:10; Janzen 1978). Studies employing a phenomenological framework often incorporate both material and ideological factors, which act at different levels or stages in a decision-making process. These processual models design pathways or hierarchies of health seeking behavior in temporal sequence, from recognition of symptoms to compliance with professional treatment (for example Chrisman 1977), or in cognitive steps in the process of choice of healer or therapy (Schwartz 1969; Young 1980).
Sargents work on theBenin draws a multi-factorial model for the determinants of birth setting, stressing that certain combinations of factors carry differential degrees of risk, uncertainty and benefit (1982:10; see also 1989:17-24).
It also examines how Benin women experience their own decision-making processes in case studies of pregnancy care and birth. Sargent writes, While political and economic forces may set the parameters within which individuals make decisions, it is nonetheless necessary to detail the processhy which selections are made from among available alternatives (1989:21).
Traditionally a woman of the Bariba of the Peoples Republic of Benin birthed outdoors, if possible without the emotional, physical, or technological aid of others. Midwives came only to help with complications. With the arrival of hospital-based obstetrical services, women have been able to cany religious beliefs and proverbial virtues of stoicism into hospital birth, and, as a result, women may use the same strategies and have the same goals for both home and hospital deliveries. For example, the fear of the birth of witch babies does not seem to subside with increased hospital births (1982:113). Likewise, the high esteem given to the woman who births stoically, without showing pain, goes unchallenged in poorly equipped hospitals where analgesics are rarely available (1982:180).

But, Sargent stresses, extrinsic factors must be assessed as well.
Governmental policy outlawing home delivery, the higher status associated with hospital delivery, and medical concerns about the pregnancy or the quality of a health service affect the utilization of services (1989:19). Although they are not entirely irrelevant, the geographic and economic accessibility of services are found to be less important an influence as religious beliefs and status aspirations.
Social networks have been the focus of other studies. Janzen (1987) comments on the conventional understanding of social networks, writing that users have glided over how agreement and conflict in a group affect decision-making. Janzen (1978) describes therapy management groups in the BaKongo of Central Africa as that group of family members and friends, who rally around a sick individual. The therapy, management: group assumes responsibility of not only the individuals normal duties but also of the persons health. Therefore, the group directs the handling of the illness and its treatment. Therapy management groups, coalesce in the case of pregnancy and childbirth as well as illness. Janzen makes a strong contribution to the understanding of the locus of decision-making and the specific patterns of interactions between individuals and knowledge. Is new knowledge about health interpreted individually or collectively?
In making obstetrical decisions, women must comply with the expectations of the social network in which they operate, find an acceptable way to dissent, or risk some degree of estrangement with the network. McClain (1987) and Rothman (1988) consider the role of social networks in the decisions ofNorth American women about birth setting and amniocentesis. McClain finds that the social network (exclusive of the familial network) predicts choice of birth setting better.than demographic variables such as age, education, parity, or marital status. Rothman (1988) writes that women who decide against amniocentesis. face_^trong pressures from their social networks to change their minds.
In sum, both material/structural and ideological factors have been found to relate to decisions about pregnancy and childbirth. Structural elements set boundaries on choice, and multiple systems of ideas and ways of knowing about birth have often been found to exist within one. society. In the present study, Bolivians be a medically pluralistic society, since women may choose to deliver with hospital obstetrical services, midwifeiy-services^and/or self-care at home. Both structural and ideological forces are considered, and the concept of authoritative knowledge is applied in the comparison of birth settings.

The Study Site
This study took place in La Paz, Bolivia, at a hospital of the National Health Fund, Matemologico 18 de Mayo. Here I summarize background information on the state of Bolivia, La Paz, the neighboring city of El Alto, and the life of Aymara Indians, who play a large role in this study.
General Information. Bolivias estimated population is 8.1 million, 40% of which are under 15 years old (International Planned Parenthood Federation 1999). The country covers 1,098,581 square kilometers, making an average density of 7.2 people per square kilometerone of the most sparsely populated countries in Latin America. It is also growing in population more slowly than most Latin American countries, at 2.3% per year (ENDSA). Aymara and Quechua Indians make up the two largest indigenous groups in Bolivia and comprise 24 and 34% of the national population, respectively (PAHO 2000). The Aymaran and Quechuan languages are the most spoken languages on the continent of South America after Spanish and Portuguese (Morales and Rocabado 1988:30). Quechua and Aymara Indians have lived side by side for generations. While their languages have remained distinct, today styles of dress are very similar. Quechua and Aymara surnames appear to be indistinguishable, although there has not been a good study of Indian names in Bolivia.
Bolivia is distinctly divided into three ecological and climatological zones the Andean region, the valleys, and the eastern plains. Politically it is divided into nine departments. The Altiplano, a high-altitude plateau in the Andes, covers the majority of the western portion of the country. Annual precipitation on the Altiplano is 100-500 cubic centimeters. The temperature averages 5-10 degrees Celsius* but in winter, it may drop as low as -20 degrees Celsius. Principal crops on the Altiplano are barley, potatoes, oatmeal, and quinoa: Additional subsistence needs are met through trading between regions, fishing, and herding animals like sheep and llamas. In the valleys and plains, com, wheat, fruits, rice, sugar cane, tobacco, coca* and a variety of vegetables are produced (Morales and Rocabado 1988:41). Aymaras are concentrated in the Departments of La Paz and Oruro on the Altiplano, and other

parts of the Altiplano and the valleys are predominantly Quechua (Morales and Rocabado 1988:43,51).
The largest portions of Bolivias gross domestic product are in manufacturing, agriculture, and mining. But the greatest growth in recent years has been in electricity, gas, water, transportation, and communications services, as well as in construction and public projects, reflecting improvements in the infrastructure of basic services in the country (ENDSA 1998:2). Despite these investments^ rates of basic utility services and consumer goods in the home remain low, especially in rural areas (see Table 2.1). Seven of the nine departmental capitals have waste collection and disposal services, but only four of them have wastewater treatment plants (PAHO 2000).
Rural Urban Total
Electricity 29 96 71
Water taps inside or outside the home, at a neighbors or public taps Human waste disposal 44 93 75
Plumbing 2 45- 74
Latrine 34 37. 36
Housing with dirt floors 33 10 71
Telephone 1 34 23
Refrigerator 7 50 34
Radio 71 93 85
Television 17 89 62
Table 2.1: Percentages of homes served with basic services and consumer goods (Source: ENDSA 1998)
The irregular topography of the Andes and the low population density partly explain the difficulty in developing these services.
Patterns of marriage and reproduction among Bolivian women are summarized in Table 2.2. Note that marriage refers to both civil marriages and household partnerships that may precede marriage. On the Altiplano, it is normal for a couple to try living together before they marry, and, even if a child is produced, the union is not yet binding (Morales and Rocabado 1988:163).
National Department of La Paz City of El Alto
Age at marriage 20.9 21.3 20.7
Age at first birth 21.5 21.9 21.4
Fecundity 4.2 3.7 3.6
Birth interval, in months 30.5 31.4 30.5
Table 2.2: Reproduction patterns among Bolivian women, measured by averages of various indicators (Source: ENDSA 1998)

Historically, on the Altiplano, residents were organized in ayllus, a kind of clan system. The goal of the ayllu was to acquire communal and individual lands at diverse altitudes and climatological, variations, apparently to optimize production and trade. A related system, called the qyw/r provided ^cooperative labor (Morales and Rocabado 1988:56). During colonization, these production systems were disrupted in many areas.
Bolivia achieved independence from Spain in 1825. For the next 127 years, until the Revolution of 1952, most land continued to belong to the hacienda system, although some independent communities existed as well. After the Revolution, agrarian reform broke up the haciendas and distributed agricultural land individually to families and pasture land communally to communities, after the ayllu system.
The reform also dictated that inheritances be divided equally among children, male and female. This law has interacted increase in population to result in the diminishment of plot size (Crandon-Malamud 1991:56). It has also increased the geographic dispersal of plots owned by a family* since women and men own plots in different locations when they marry. Much of the land in the Altiplano is not cultivable, due to sandy soil and irregular topography. Plots must lie fellow for five to seven years to prevent extreme degradation (Crandon-Malamud 1991:75).
Because of these problems, Bolivia has the lowest calorie production per capita in Latin America2088 calories per day. This figure includes coffee and sugar,: which are predominantly export products. Nutritional studies estimate the average daily consumption of a Bolivian, excluding those in the highest social strata, at around: 1800 calories per day, .a portion of which comes from legaL; black market, and donated food imports. Eighteen hundred calories per day is 16% below National Research Council recommendations, adjusted for climate and body size. Compared to residents of other nations, Bolivians consume very little animal protein. Other nutritional deficiencies include calcium, niacin, thiamin, and vitamin A. Urban life has worsened this situation. As processed foods have become more available in the city, sugar, bread, and noodles have begun to replace potatoes, barley, and quinoa (Morales and Rocabado 1988:200-210). The biggest nutritional deficit is found on the Altiplano (Morales and Rocabado 1988:204).
Diminishing per capita production in the Altiplano is one push factor towards out-migration, and pressure is often put on young adults to leave the home in order to reduce the number of mouths to feed (Albo, Greaves, and Sandoval 1981:63; Morales and Rocabado 1988:160; Crandon-Malamud 1991:25); Migration is one of the most salient elements in Bolivian society. It is most often rural-to-urban, but rural-to-rural movements are common as well. Among push factors, we can add the closing of several mines in 1985, which displaced 21,000 workers (Morales and Rocabado

1988:36). Many migrants head to the coca fields or the city to find work (see Mclver Weatherford 1987 for the detrimental consequences of migration to the coca fields). Pull factors to the city include increased access to services, economic opportunities, and the desire to become more cultured or modem (Albo* Greaves, and Sandoval 1981:68-72).
Health in Bolivia. Rates of maternal, neonatal, infant, and child mortality have decreased in recent years, due to improvements in delivery services, vaccination programs, increased urban migration, and improved education levels. .Table 2.3 displays rates reported in 1998. Yet both infant and maternal mortality indices remain the highest in Latin America (PAHO 2000). It is also likely that infant mortality rates are underestimated, due to the tendency to round the childs age at death upward (ENDSA 1998:121). Maternal mortality rates have not been tracked regularly. For decades, it was reported to be 480/100,000 live births, although no one knew where that estimate originated (Morales and Rocabado 1988:183). Finally in the 1994 National Survey on Demography and Health (ENDSA), the index was figured methodologically, but it was not recalculated in the 1998 ENDSA. Causes of maternal mortality include hemorrhage, eclampsia, infection, obstructed labor, and abortion (1PPF 1999).
National Department of La Paz City of El Alto
Rural Urban Total
Maternal mortality (associated with pregnancy, delivery, and the postpartum period) (per 100,00 live births)' 524 274 390
Neonatal mortality (in the first month) (per 1000 live births) 52 24 36 42 35
Infant mortality (in the first year) (per 1000 live births) 100 53 73 82 89
Child mortality (in the first five years) (per 1000 live births) 134 72 99 106 120
Table 2.3: Maternal and child health indices for Bolivia, La Paz Department, and El Alto (Source:
1998 and 1994 ENDSA)
Maternal mortality estimates are from die 1994 ENDSA and were only available at the national level.
Bolivias public health and social security systems are decentralized, with funds distributed by the national government to local governmental centers. Local officials are then responsible for managing and operating health care services, according to the particular needs of the locality. Overall, population per doctor is 2564, and 69% of the population have access to health services (3PPF 1999). National public spending on health is US$23 per person per year, or 2.5% of the gross domestic product. Public institutions are organized into three care levels. On the first

level are 896 health centers and 1210 health posts, which hold 2276 beds and attend normal births and general medical emergencies. The second level, basic hospitalization services, is represented by 63 hospitals. The third level is made up of hospitals able to provide specialized services, and there are 81 general hospitals and 29 specialized hospitals on this level (PAHO 2000).
Social security programs provide, care for 20% of Bolivias population. The largest fund within the social security system is the Caja National de Salud (CNS), or the National Health Fund. From the above mentioned facilities, 131 health centers, 60 health posts, and 26 hospitals belong to the CNS, for a total of2524 beds. The institution where this study, was conducted^ Matemologico 18 de Mayo (sometimes referred to as Hospital 18 de Mayo), is a tertiary-level, specialized facility for obstetrical and gynecological attention, the largest such hospital in the CNS. Located in the northern part of the city of La Paz, it has 69 beds and averages 74% occupation. In 1998, it attended 5094 births, 27% of which were cesarean deliveries. It also reported 98.9% live births and 1.8% stillbirths, although, obviously, these numbers do not add up to 100%. Postnatal infant deaths occurred in 0.4% of cases, and, in the year, there was one maternal death. A new facility is in construction, which will hold 500 beds, as administrators hope to attract private patients (data from internal hospital documents).
Under public health laws, uninsured pregnant women are eligible for two prenatal visits, attention during labor and delivery, and emergency care with signs of problems free of charge. This policy is intended to decrease homebirths and increase the use of hospital services. Twenty-seven percent of patients attended at Matemologico 18 deMayo are of this type. In practice, the policy does not work as easily as it sounds, because women must have with them a set of identification documents in order to receive service. Women do not always carry such documents or do not realize that they are required.
The private health sector serves around 10% of the population (Morales and Rocabado 1988:215). Non-govemmental and religious organizations are connected to both public and private branches of health care, but nearly all their funding is international (PAHO 2000). Traditional medicine in Bolivia includes numerous kinds of practitioners, who use techniques including spiritual intervention, herbal therapy, bonesetting, guinea pig divination, and coca leaf reading. Recently, it has become common for traditional healers to incorporate aspects of biomedicine into their practices, such as antibiotics and aspirin. It is estimated that there is a traditional practitioner for every 500 population in Bolivia and that they capture 60% of outpatient visits in the country. Thus, they by far outnumber physicians, and their services are used more (Morales and Rocabado 1988:224-226).

La Paz/El Alto and Urban- Avmara Culture
La Paz is the administrative center of Bolivia. It and its neighboring eity of El Alto lie on the Altiplano between 3200 and 4100 meters altitude. On June 6, 2000, the city announced that the one-millionth citizen of La Paz would be bora at 9:45, later that morning. A competition between the maternity hospitals ensued, each one calling the radio stations and promising there was a woman in labor who was going to birth at 9:45. In the end, a planned cesarean delivered twins, giving the city its one-millionth and one-millionth and one citizens. Thus, approximately one-eighth of the nations populace resides in La Paz. Approximately 24% of these residents are first-language Aymara speakers, and 8% speak Quechua as a first-language (see Table 2.4).
Aymara Quechua Spanish and other
La Paz 24 8 -68
El Alto 50 6 44
Both cities 32 7 61
Table 2.4: First-languages of La Paz and El Alto residents, as percentages of each citys population (Source: INE 2000)
La Paz is built into a canyon and one of the canyons branches, descending from the Altiplano. From the central avenue, which runs down the middle of the canyon, streets angle upward on both sides, and houses are built on slopes. One hundred and eighty-five rivers and ravines break the city into neighborhoods and necessitate windy roads. The distance from the center of La Paz to the center of El Alto is only 3.1 kilometers, but the road that takes you there covers 10 kilometers. Due to the difficult topography, the city has grown irregularly, and many streets in marginal neighborhoods do not have names (Albo, Greaves and Sandoval 1981:85)
El Alto sits above La Paz, on the Altiplano. The flatness of its topography makes it the home of the areas airport, which lies in the middle of town. El Alto is less developed than La Paz--only the biggest roads are paved, shops are very scarce outside the central zone, and the outskirts of El Alto look more like planned rural settlements than like city neighborhoods. The population of El Alto is 405,492, 50% of which speak Aymara as a first-language, and 6% of which speak Quechua as a first-language (see Table 2.4).

Thus, the indian population of La Paz/El Alto is predominantly Aymaran (INE 2000; Albo, Greaves and Sandoval 1981). Aymaras have been present in the La Paz area since the 16th century, but it is only in recent decades that the city has experienced a high influx of rural migrants from the Altiplano. In their four-volume account of migration, Albo, Greaves and Sandoval describe the position of Aymara migrants in La Paz/El Alto (Albo, Greaves and Sandoval 1981 and 1983; Sandoval, Albo, and Greaves 1987). Migrants occupy two worlds, that of Aymara culture and that of urban culture. The urban- Aymara culture emerges from this meeting of worlds, creating a culture with unique aspects.
Upon arrival in the city, the migrant finds a new social position, not only in relation to the citys mestizos, but also in relation to other Aymaras. Ethnic categories are closely tied to socioeconomic categories, and are therefore fluid (see Crandon-Malamud 1991). In the countryside, a campesino is an indian who works the land for subsistence. Moving to the city, such a term becomes derogatory, as it connotes poverty. Cholo is a term used in many different ways throughout Latin America. In La Paz, on the tongue of a mestizo, it is a pejorative term for Indians. But among indians, it signals an individual who has joined the cash economy, usually through a move to the city (Crandon-Malamud 1991:224).
Cholita is a term applied to indian women. In La Paz/El Alto, its use has become acceptable andeven embraced by cholitas themselves. The predominant occupational roles for cholitas outside the home are that of comerciantes, vendedoras, and tejedoras. Comerciantes and vendedoras work in the kiosks of the citys markets and on the sidewalks of busy streets, selling anything from fresh produce to underwear to herbal remedies. Largely this type of work belongs, to the informal sector of the economy, which offers easy entry and requires little investment. Therefore, it is attractive to rural migrants who have come with little capital but a desire to improve their earnings. It is particularly attractive to women, who have few other options for work, due to both their lower rates of education in Bolivia (compared to men) and a sexism that deems them unfit for most kinds of employment outside the home (see Browner 1989:462). Tejedoras practice the traditional indigenous craft of weaving mantas, shawls that are tied over the shoulders for warmth, fashion, or carrying children and market goods.
Urban Aymaras are less economically and socially powerful than their mestizo neighbors, but levels of social stratification exist within the Aymara population as well. For example, at the festival of Gran Poder, celebrated in early June in La Paz, an Aymara beauty queen is selected. She must be a participant in one
' Although in the last decade, there has been a surge of cultural pride associated with the term

of the social organizations that dance in the parade, which coststhe individual approximately US$4000 a year. This figure is about seven times the National Minimum Salary, alevel ofipay which is usually higher than that received by unsalaried workers.
Among women, social and/or economic position can be indicated by variations on the traditional style of dress. The traditional outfit consists of the pollera, a tiered skirt, the awayu (or manta), a type of shawl worn oyer the shoulders, and plastic shoes elaborated with lace patterns and small bows. Hair is worn long, in two braids attached at the ends, and covered by a derby hat Women may indicate a higher economic standing by wearing two or more layered poller as, mantas of finer materials, gold teeth, imported hats, and more expensive earrings; The woman who adopts Western dress and hairstyles is said to be de vestido, although she may continue to use the awayu to carry children on her back. As a fife vestido woman, she can hide her rural origins and enter educational and work institutions that she would not be permitted to participate in if she were de pollera. But, among Aymaras, she is subordinate to women de pollera. Social stratification is tied to skin color as well, as it is said that ones skin gets lighter in the city (Albo, Greaves, and Sandoval 1983:3-39).
Although urban migrants wish to place some social distance between themselves and the countryside, they maintain contact with their communities of origin, especially those that own land there. Those who lack land holdings express the desire to acquire them, through inheritance or purchase. As noted, under the traditional ayllu system, Altiplano residents were accustomed to holding scattered pieces of land, each of which may have a small home built next to it. To divide time between a home in the city and one in the countryside indicates success in both arenas. These possibilities are kept open by return visits to rural communities, especially at fiesta times (Sandoval, Albo and Greaves 1987:6-13, 56-61).
In sum, Bolivia is a struggling underdeveloped nation. Its indigenous population continues to maintain a strong presence, in urban areas as much as in rural areas. Maternal and child health is the worst in Latin America. Geographic divisions, economic crisis, land degradation, cultural diversity, and bureaucratic complexities create barriers to improvements in health.

Research Design
During data collection in La Paz, Bolivia, for a NIH project entitled, Interpopulational differences in intrauterine growth restriction (IUGR) at high altitude, investigators found that a number of women eligible for the services of the national health service of Bolivia attend prenatal and postpartum appointments, but do not report to the hospital when labor has begun. Rather, they remain in their homes to labor and birth, typically coming to the hospital with the baby in the first week for postpartum and infant outpatient visits. This studys overall research question was, why do some women in La Paz/El Alto birth at home and others birth in the hospital? I hypothesized that women who birth at home do so because they are less able to incur the economic costs of hospital birth, including the costs of missed work or familial obligations. I also asked, how do indian women in La Paz/El Alto make decisions about birth setting?
The study took place in two phases. Phase I was a medical records review and Phase II used ethnographic interviewing with Bolivian women. Table 3.1 presents the number of cases in both phases of the study. Because interviews were obtained opportunistically and all interviewees were indian (explained below), the interviewed groups may not be representative of the populations of home- and hospital-birthing women in La Paz. To evaluate whether the interviewed groups were representative of indian women who were patients at the hospital, they were compared with chart-review cases (combined home- and hospital-birthing cases) with at least one indian surname. These cases are referred to as the chart-review indian group. The use of surnames as a proxy measure for ethnic background is discussed in a later section. Sample sizes in the interviewed groups were too small to allow statistical comparisons between interviewed and chart-review home-birthing women, on the one hand, and interviewed and chart-review hospital-birthing women, on the other hand.

Chart review Interviewed
Home Hospital Combined Home Hospital Combined
Indian 71 48 119 20 7 27
Non-indian 14 37 51 0 0 0
Combined 85 85 170 20 7 27
Table 3.1: Number of cases in samples and sub-samples
Note that throughout this paper, women who birthed at home are occasionally called home-birthers, and women who birthed in the hospital may be called hospital-birthers, in order to facilitate the flow of discussion. The pregnancy and delivery that occurred during January to March 1998, for chart review cases, or the pregnancy and delivery that occurred most recently, for interviewed cases, is referred to as the study pregnancy and delivery.
Phase I: Medical Chart Review
In July of 1999, medical, obstetrical, demographic, and socioeconomic data on all cases of homebirth during January, February, and March 1998 were collected from medical records at Hospital 18 de Mayo in La Paz. Cases from these three months totaled 85. Similar data had already been gathered by Loma G. Moore and Fernando Armazathe principal investigator and co-collaborator in the larger project on IUGR, mentioned aboveon women who birthed at the hospital during the same time period. From these latter data, 85 cases were randomly chosen. Cases of cesarean delivery were not excluded. Thus, among women who had been obstetrical patients at Hospital 18 de Mayo and who had delivered their babies in the first three months of 1998, two groups were createdwomen who birthed at home versus women who birthed in the hospital. Throughout this paper, these samples are referred to as chart-review groups.
The variables recorded are listed in Table 3.2. Biological health was measured by weight gain during the course of prenatal care (in kilograms), hemoglobin, maximum mean arterial pressure (diastolic pressure x 2 + systolic pressure / 3), and height in centimeters. Because pre-pregnancy weight was rarely recorded, it was not

possible to measure total weight gain during pregnancy. Obstetrical history included gravidity (number of pregnancies), parity (number of births after 24 weeks pregnancy), previous miscarriages (before 24 weeks pregnancy), stillbirths (after 24 weeks pregnancy), birth interval in months (time past since the end of the last pregnancy), number of prenatal visits during the study pregnancy, and weeks pregnancy at the first prenatal visit. Outcome of the study pregnancy was measured by length of gestation in weeks, stillbirth, and birthweight in kilograms. Demographic indicators included age, marital status (married, single, or living together), and city of residence (El Alto or La Paz).
The only information available in medical charts suggesting ethnic background (Indian or mestizo) were surnames. Chakraborty et al (1989) have demonstrated that Aymaran surnames are reliable indicators of indigenous genetic traits and, by extension, Aymaran ancestry Aymara and Quechua surnames are identical. But, because Aymara-speakers make up a much greater portion of La PazM Altos population than Quechua-speakers (see Table 2.4), it is likely that most indian surnames in the sample are Aymaran names. Therefore, the number of indian surnames is used as a proxy measure for ethnicity, and, for the purposes of this study, women with at least one indian surname are considered indian. Dr.
Fernando Armaza, of Hospital 18 de Mayo, collaborated in determining which names are indian. A limitation to this scheme is that, in contemporary Bolivia, ethnic terms are applied according to economic and social position rather than ancestry, as discussed in the background section. Thus, many women with indian surnames may not be considered indian by themselves or by other Bolivians. Another limitation is that occasionally individuals choose to change their indian surname to a mestizo name that is either similar in sound or has the same meaning in Spanish as the original name in Aymara or Quechua. In these cases,
Weight gain during prenatal care, kgs
Height, eras
Maximum mean arterial pressure
Obstetrical history:
Gravidity (number of pregnancies) Parity (number of births)
Previous miscarriages
Previous stillbirths
Birth interval, in months
Number of prenatal visits
Weeks pregnant at first prenatal visit -
Outcomes of the study pregnancy:
Birthweight, kg?
Length of gestation, weeks
Marital status
City of residence
Ethnicity, by indian surnames
Husbands ethnicity, by indian surnames
Relation to insured
Level of education Income, in Bolivianos Ocorpation Husbands occupation City water services in home City sewage services in home Number of deceased children
Table 3.2: Variables used in medical chart review

individuals may or may not continue to follow a way of life that is deemed to be Indian within Bolivian society (Albo, Greaves, and Sandoval 1983:19, 39).
Workers and employers pay into the National Health fund, which serves them and their families. Accordingly, a womans relation to the insured party (self or beneficiary) was also recorded. Socioeconomic status was measured by level of education, income (reported inBolivias currency, the Boliviano, and estimated in US dollars), occupation (salaried or unsalaried), husbands occupation (salaried or unsalaried), and the presence of city water and/or sewage services in the home. The number of deceased children was expressed both as a continuous and as a binary variable (either a woman has suffered the death of children or not), limiting cases to those with previous li ve births. It was also reported as a fraction of live births among women with previous live births.
Procedures.and Analyses
Records at Hospital 18 de Mayo are not electronic; therefore the data was copied by hand from prenatal records onto a form designed by Loma G. Moore and Fernando Armaza for the parent study. To make statistical comparisons, the data were entered into SPSS files. Two-tailed two samples t-tests were used to analyze continuous variables, with a confidence interval of 95% and a significance level of p<05, except that two-factorial ANOVA was used to compare number of deceased children controlling for number of live births. Chi-square and Cramers V measure of association were calculated for ordered and nominal variables. Cases with missing values were excluded from analysis on a test by test basis in these tests.
In a logistic regression analysis, cases with missing values were excluded on a listwise basis. Odds ratios with 95% confidence intervals estimated the increase in odds of having a home birth with the presence of or one-unit increase in included predictors. Independent variables entered into the model were those found to be significantly different between the two groups, with exceptions noted in the results section. Ethnicity was dichotomized to form a group with no Indian surnames and a group with at least one indian surname. Education was dichotomized into one group with no or primary level education and another with secondary or university education. History of deceased children was also entered as a binary variable. Predictors were entered into the model in the standard fashion (as a block). In order to assess whether results varied with different selection methods, the analysis was subsequently run with forward and backward stepwise methods.

Phase II: Ethnographic Interviewing and Observation
To gather information not available in medical records, ethnographic interviewing with patients of Hospital 18 de Mayo, some of whom had birthed at home and some of whom had birthed in the hospital, was undertaken in May-July 2000. The interviewed groups are sometimes referred to as the interviewed home group and the interviewed hospital group.
Interviews with home-birthing women were obtained opportunistically. Each morning during the study period; I waited at the hospital, hoping that a woman would arrive seeking a postpartum check-up after a homebirth. Every woman who reports she has had a recent homebirth is required to visit the hospitals social worker, Licenciada Lourdes Torres. Lie. Torres would introduce me to a woman as a foreign student interested in homebirth. After obtaining informed consent (discussed below),
I accomplished twenty interviews in this manner. In actuality, interviews in the homebirthing group included one interview with the birthing womans husband, one interview with the birthing womans mother, who was also the attending midwife at the birth, three interviews in which both the birthing woman and her husband were present, and two interviews with women who were at the hospital that day for prenatal visits but who had delivered all of their children at home and said they would most likely deliver their next babies at home.
There were seven interviews in the hospital-birthing group. Working with the information that was consistently available in charts and the characteristics that identify women as Indian, I reviewed medical charts in the postpartum wards each day, looking for the following criteria:
1) indian surnames, based on my familiarity with them, a list reported by
Chakraborty et al (1989), and a bit of help from the nurses;
2) husbands indian surnames,
3) informal sector occupations, such as tejedora and comercicmte; and
4) residence in El Alto, since 50% of the El Alto population speak primarily
Aymara, compared with 24% of La Paz residents (Table 2.4).

It was difficult to find a good time to interview these women in the postpartum wards. When women were not feeling well or appeared to be resting* I did not disturb them. Limited by these circumstances, seven interviews were accomplished. In two of these interviews, husbands were present but did not choose to participate. One of the interviews in the hospital-birthing group was with a woman who was at the hospital that day for a prenatal visit. She had delivered her first two babies at home and her next two in the hospital, and she planned to deliver the baby of her current pregnancy in the hospital.
Interviews focused on indian women. Data on the variables in Table 3.2 were
gathered from interviewees medical records. Additional quantified variables in this phase (Table 3.3) included place of origin (rural or urban) and migration history, expressed as bom in La Paz/El Alto, migrated as child (younger than fourteen), migrated as adult (fourteen or older), or currently lives in rural area part-time. Fourteen was chosen as the cutoff point between childhood and adult status because I was interested in knowing where a woman reached reproductive age. Cultural characteristics included style of dress {de pollera versus transitional or de vestido) and non-Spanish languages. Factors related to potential barriers to hospital care were measured by the availability of vehicles and telephones to womens households and the number of days a woman reported resting from her normal activities after the birth. Information on reproductive history beyond what was available in medical records included contraception history, operationalized as past use of the IUD and/or the rhythm method (these were the only two methods reported) and whether or not the last pregnancy was planned. Age at first birth was also calculated from reproductive histories.
Values for these variables were drawn from the interview topics listed in Table 3.4. The first eight items were meant to address the hypothesis that women who birth at home are less able to incur costs related to hospital birth.
Place of origin/pattem of migration Non-Spanish languages Style of dress
Availability of vehicle to household Telephone in home or cell phone available
Number of days of rest postpartum
Planned pregnancy
Past use of IUD
Past use of rhythm method
Age at first birth
Table 3.3: Additional quantified variables for interviewed groups
* This is an estimate, as no data on age at menarche among this population is available.

Factors related to costs of hospital birth:
1. Occupations of the woman, the domestic partner, if present, and children who contribute to the household income, including both formal and informal sector work
2. The scheduling of wages (hourly, according to productivity, self-employed, etc.) earned by all family members
3. The extent of flexibility in the womans work schedule, according to a supervisors expectations or according to the demands of the market, in the case of self-employment
4. Total household income
5. The locus of responsibility for childcare, meal preparation, and other household duties
6. Means and costs of transportation (city bus, taxi, personal vehicle, drivers, etc.) available to the woman for prenatal visits and transport to the hospital at the onset of labor
7. Any money given for hospital and auxiliary care including lab tests, ultrasound exams, etc.
8. Number of individuals both inside and outside the household who have contributed to the household in the past either monetarily or through the temporary adoption of work and/or familial responsibilities of the household
Other topics:
9. Place of birth and history of migration to the city (as a child or adult)
10. History of migration, including seasonal movements, of the domestic partner
11. Reproductive history, including past births, use of birth control, success in family planning, and age at first birth
12. Knowledge of traditional birth attendants and history of use of their services
13. Perceptions of the degree of difficulty of the last labor and delivery and complications experienced
14. Participants and the nature of their participation during homebirths
15. Techniques used during a homebirth, including placental disposal
16. Reasons for attending prenatal care
17. Speculation on why other women in the interviewees society might birth at home or in the hospital
18. Precautions women must take during pregnancy, labor, delivery, and
Table 3.4: Topics discussed at interviews

Interviews took place at Hospital 18 de Mayo, in empty offices, for home-birthers, and in the postpartum wards, for hospital-birthers. All interviews were conducted in Spanish. As it turned out, the nature of each interview (semi-structured versus open-ended) was determined more by my Spanish skills than by methodology. I was able to let those women whom I found easy to understand in Spanish speak narratively. Interviews with women whose accents were more difficult for me or who were less motivated to speak clearly were more structured, simply because I could not follow lengthy stories. Interviews lasted from a half-hour to one hour and a half. All but the first three were taped, with the interviewees permission, and later reviewed on a micro-cassette recorder, which allowed me to write down important quotes, monitor my interviewing style, and study local idioms. As my Spanish improved, so did the quality and length of interviews. Tapes were reused and copied over within a week, after I had listened to them.
In addition to interviews, observed labors and births at Hospital 18 de Mayo for a total of twenty-five hours on two different days. During this time, I witnessed five vaginal births and two cesareans, as well as the labors of several women.
Informed consent and confidentiality. Before each interview, I explained, among other things, the following items:
1. Self-introduction: I was a student at a university in the United States, and the information the woman shared with me would help me finish my degree. I had previously studied and attended homebirths in my own country.
2. The goals of my project: to understand why some women choose to birth at home when they could birth in the hospital.
3. Confidentiality: I stressed to women that nothing they said to me would be recorded in their medical charts or shared with hospital workers. I asked permission to use the tape recorder, and explained that the tape of the intervi ew would be erased within a few days. Every effort was made to ensure that interviews took place in private rooms, a difficult task due to the extreme limitations on space at Hospital 18 de Mayo. In only two cases were hospital personnel present for a part of the intervi ew.
4. Content of the interview: The questions I asked would be about the womans previous births, prenatal care, her occupation and the occupations of family members.

5. Further explanations: Occasionally women had additional questions about my project and the content of interviews. These questions were answered before the interview proceeded.
6. Consent: Consent was given verbally, in response to the question, Ts it okay if I ask you these questions? Only one woman refused.
Compensation and building rapport. At first I tried to compensate women for their time by feeding them during the interview. I brought drinks, bread, and other food and set them out on napkins. But each time, the woman would not eat with me. Rather, the bread would disappear into her bag while my head was turned away. It seemed that bread was a welcome gift, but that consuming it together was either undesirable or inappropriate for the circumstances. So, I started offering loaves of bread at the end of interviews. These gifts were usually greatly appreciated and often sparked renewed interest in interview topics. After I noticed this, I tried to create rapport earlier in the interview by giving the bread at the beginningbut it did not work the same way. Well, I thought, at least by giving a gift at the end of an interview, I would be well situated to seek a second visit with a participant. Unfortunately, I never attained a second interview because women did not often return to the hospital and, although they never said so outright, they clearly did not want me coming to their homes. At times they were very vague about their addresses and tried to escape me when I offered to pay their cab fare home if I could come along and learn how to get to their house. I quickly quit asking for permission to visit women in their homes. Distrust of outsiders is probably an adaptive cultural element of the Aymarans, considering their long history of exploitation by Quechua Indians, the Incas, and the Spanish.
Notes from interviews, tapes, and observations were coded by hand. Codes focused on key rationales for and considerations involved in the decision of where to give birth, as well as the mention of various cultural practices and beliefs related to birth. Themes were then reviewed and compared between home- and hospitalbirthing women. Quantified data from interviews (Table 3.3) were entered into an Excel file for tabulation and an SPSS file for statistical comparison with the chart-review indian group. These statistical analyses were the same as those described for Phase I.

Phase I
Table 4.1 presents results of two-tailed two samples t-tests, chi-square tests and Cramers V measures of association on chart review data for home- and hospitalbirthing women.
Mean or N (%)
Home Hospital Combined Cramers V
BIOLOGICAL Maximum mesa arterial pressure 90 91 90
Weight gain during prenatal care, kgs 5.6 6.5 6.1
Hemoglobin 13.4 13.6 13.5
Height, cms** 149.6 152.0 150.7
OUTCOME OF STUDY PREGNANCY Birthweight, kgs 3390 (N=8) 3130 (N=85) 3150 ...
Stillbirth 0 (0%) 0 (0%) 0 (0%)
length of gestation, weeks 39.2 39.0 39.1
OBSTETRICAL HISTORY Birth interval, months 44.1 38.5 41.8
Gravidity** 4.1 2.7 3.4
Parity** 3.9 2.5 3.2
Previous miscarriages 0 64(75%) 64(75%) 128 (75%)
1 or more 21 (25%) 21(25%) 42 (25%)
Previous stillbirths 0 84(99%) 83 (98%) 167(98%)
1 or more 1 (1%) 2 (2%) 3 (2%)
Numbs- of prenatal visits* 5.6 6.4 6.0
Weeks pregnant at 1st visit 20.2 19.8 20.0
DEMOGRAPHIC Age 28.6 28.7 28.7
Ethnicity, by number of indian surnames** 0 14(17%) 37 (44%) 51 (30%) .30
1 34 (40%) 22 (28%) 58 (34%)
2 37(43%) 24 (28%) 61 (36%)
Husbands ethnicity, by number of indian surnames** 0 20 (24%) 33 (49%) 53 (35%) .26
1 27(32%) 15 (22%) 42 (28%)
2 37(44%) 20 (29%) 57(37%)
Marital status** Married 78 (94%) 59 (70%) 137(82%) .31
Single 4(5%) 12 (15%) 16 (10%)
Living together 1 (1%) 12 (15%) 13 (8%)

Table 4.1 (COnt)
Mean or N (%\
Home Hospital Combined Cramers V
City of residence** El Alto 61 (74%) 29 (35%) 90 (54%) .39
La Paz or other 22 (26%) 54(65%) 76 (46%)
Relation to insured** Self 13(15%) 31(39%) 44(27%) .27
Beneficiary 72 (85%) 49 (61%) 121 (73%)
Monthly income, in Bolivianos** 723 (US$120) 1103 (US$184) 914 (US$152)
Numbs of deceased children among women with .17
previous live births***
0 57(71%) 53 (86%) 110(77%)
1 or more 23 (29%) 9 (14%) 32 (23%)
Number of deceased daildrcn as a fraction of live .09 .05 .07
births among woman with previous live births
Education** None or primary 60 (73%) 14 (24%) 73 (44%) .59
Secondary 21 (26%) 42 (49%) 63 (38%)
University 1(1%) 23 (27%) 24(14%)
City water in home** Yes 69(83%) 85(100%) 154(92%) .31
No 14 (17%) 0 (0%) 14(8%)
City sewage in home** Yes 50 (60%) 80 (94%) 130 (77%) .41
No 33 (40%) 5 (6%) 38 (23%)
Occupation** Unsalaried 78 (92%) 52 (64) 130 (78%) .34
Salaried 7(8) 30 (37%) 37 (22%)
Husbands occupation Unsalaried 44(56) 31(46) 75 (51%)
Salaried 34(44) 37 (54%) 71 (49%)
Table 4.1: Results of t-tests and chi-square tests for chart review groups. (Note that the number of cases does not always equal 170, due to missing values for some items.)
***As a continuous variable; the number of deceased children is not significant when number of live births is controlled for using two-factorial ANOVA.
The only significant biological/medical difference between home-birthing and hospital-birthing women was in height, which was greater among hospital-birthing women. This finding may suggest that home-birthing women experienced poorer nutrition during their fetal periods or childhoods. Women were similar in maximum mean arterial pressures during the prenatal period, weight gain during prenatal care, and hemoglobin.
Pregnancy outcome, as measured by length of gestation and birthweight, did not differ between home- and hospital-birthing women. However, it should be kept in mind that birthweight was recorded for only eight cases in the home-birthing group, since only eight women reported to the hospital the same day as the birth. These eight cases suffered immediate postpartum complications such as placental retention and hemorrhage. Unfortunately, information on postpartum complications was not

recorded for the hospital-birthing group. There were no stillbirths recorded for the study delivery. Twenty percent of the hospital-birthing women delivered by cesarean, slightly less than Hospital IB de Mayos average rate of 27%. Of course, all the homebirths were vaginal Together, these results suggest that home-ftirthing women and hospital-birthing women, all of whom have received prenatal care, experience similar birth outcomes. However, the possibility remains that home-birthing women suffer greater postpartum complications.
Obstetrical history varied between groups in gravidity, parity, and number of prenatal visits. Home-birthing women had higher gravidity than hospital-birthers. Hospital-birthers attended the most prenatal visits. Birth interval, previous miscarriages and stillbirths, and weeks of pregnancy at first prenatal visit were similar between groups. Induced abortion is illegal in Bolivia, but seemingly not uncommon. As many as two-thirds of urban Aymaran women in one study had intentionally aborted pregnancies or attempted to induce abortion (Schuler, Choque, and Ranee 1994). Hospital staff have a wealth of stories about women treated for complications after an attempted abortion. An unfortunate limitation of health services in the country is that staff are required to determine whether an abortion was induced or spontaneous, so that the woman can be charged for services normally covered by the national health care service, if it is decided that the abortion was induced. In this atmosphere, women are very unlikely to reveal a history of induced abortion, and such information is not included in medical charts, or in the National Health and Demography Survey (ENDSA 1998).
Both groups were similar in age. Because age and birth intervals were similar between groups, but gravidity was higher among home-birthers, the data suggest that home-birthers began having children at younger ages. Home^birthing women were more likely to be married than hospital-birthing women, probably because they were also more likely to be beneficiaries of their husbands insurance, rather than insured workers. We can conclude that women who work in the formal sector of the economy are not very likely to have a homebirth. Of those insured workers who chose a homebirth, half were schoolteachers. Schoolteachers are often stationed in rural areas without health care centers, coming to the city only occasionally for medical services such as prenatal care. Birth in the hospital may be much more geographically inaccessible for them than for women who live in the city. Women in the hospitalbirthing group and their husbands had fewer indian surnames than other groups. Most home-birthers lived in El Alto, while most hospital-birthers lived in La Paz.
Home-birthing and hospital-birthing women differed in nearly all socioeconomic variables. Home-birthers were less educated, with most individuals completing only grade school; the greatest portion of hospital birthers attended

secondary school. Household income was lower among home-birthing women. In both groups, occupations were most often unsalaried (such as homemaker, domestic employee, shop helper, etc.), but a greater portion of hospital-birthers had jobs with salaries. To compare the likelihoods of experiencing the death of children, cases were limited to those with at least one live birth prior to the study pregnancy. Women birthing at home were more likely to experience the death of children. However, controlling for number of live births, there was no difference between home- and hospital-birthers. Basic utility services (water and sewage) were missing from the homes of home-birthers more often than they were from the homes of hospital-birthing women.
Cramers V measure of association for categorical variables showed that level of education, city of residence, presence of city sewage systems, occupation, presence of city water services, marital status and ethnicity as measured by surnames were all moderately associated ( 40-.69) with birth setting. Husbands ethnicity, whether or not the woman had experienced child death, and relation to the insured were weakly associated (.10-39) with birth setting.
Results of the logistic regression analysis are presented in Table 4.2. Marital status and presence of city water services could not be included in the model because, after excluding cases listwise, they became constants for one or both groups. Parity and relation to the insured were excluded for their high coefficient correlations with gravidity (-.944) and maternal occupation (-.824), respectively. Height was excluded after the initial analysis because its correlation with the constant in the model was so strong (-.978) as to suggest that it did not contribute to the model. Results were similar when variables were entered on a forward or backward stepwise basis.
Odds Ratios 95% confidence interval
Indian surnames 1.16 0.28, 4.82
Husbands Indian surnames 0.76 0.23, 2.52
El Alto residence 4.42** 1.65, 11.81
Monthly income 0.99 0.99, 1.00
Deceased children 1.78 0.42, 7.56
Secondary or university education 0.26* * 0.09, 0.74
Presence of city sewage services 0.25* 0.07, 0.91
Salaried occupation 0.84 0.21, 3.46
Gravidity 1.25 0.90, 1.74
Number of praiatal visits 0.91 0.73, 1.15
Table 4.2: Odds ratios for having a home birth
*p< .05
**p< .01
The Hosmer and Lemeshow goodness-of-fit test was found to have a significance level of .04, demonstrating that, collectively, the variables in the regression analysis contributed significantly to prediction of birth setting. Yet, the

odds of having a home birth were significantly associated with only three covariates. Residents of El Alto were 4.42 times as likely to birth at home. Women with secondary or university education levels were 0.26 times as likely to be home-birthers, and women with sewage services in their homes were 0.25 times as likely to birth at home. The finding that indian surnames do not significantly contribute to the prediction of birth setting should be interpreted with reservation, since they are a weak proxy measure of ethnicity.
In sum, the chart review data paint a picture of home-birthing women who are of poorer economic standing and education level and more often residents of El Alto than other women. El Alto residence may create barriers related to greater distance from the hospital, or it may be important because Indians have concentrated there (see Table 2.4), and indians have cultural reasons for choosing homebirth. Although El Alto is less developed than La Paz, the presence of sewage systems in the home is independently associated with birth setting, indicating residence in the least developed neighborhoods of both La Paz and El Alto, and, by extension, greater poverty among home-birthers. Lower education rates among home-birthers may restrict their knowledge of medical issues about birth, or they may be associated with other factors related to low socioeconomic standing that lead to homebirth, such as unfamiliarity or discomfort with state-associated institutions. Ethnographic interviewing with women intended to clarify these questions by providing rich detail about home and hospital birth and womens processes of deci sion-making on birth setting.
Phase II
As explained, interviews with home-birthing women were accomplished opportunistically, according to their arrival at the hospital in search of postpartum and infant care. Because all interviewed home-birthing women were indian, indian women became the focus of the second phase of the study. The differences among indian women are more complex and less apparent than differences between indian and mestizo women. Thus, this focus promised to produce important data.
Face-to-face contact with interviewees allowed a better evaluation of ethnicity than was possible with medical chart data. See Table 4.3 for tabulations of ethnic characteristics among interviewees. Seventy-one percent of home-birthers and 57 % of hospital-birthers were de poller a, while others were dressed transitionally or de vestido (see the background section for an explanation of these styles of dress). Unfortunately, my untrained eye could not distinguish the more subtle forms of dress

Aymaran women use to give cues about their relative social standing within urban Aymara culture. All but one home-birther (94%) and one hospital-birther (86%) spoke Aymara. The remaining home-birther spoke Quechua as a first-language and was the only participant who had moved to the area from outside of La Paz Department. The remaining hospital-birther spoke only Spanish. However, this interviewee was de pollera, indicating that she moved in indian social circles and was likely to at least understand Aymara and/or Quechua. Only one interviewee (4%), a home-birther, did not have indian surnames. Her family names may have been changed to mestizo names at some point in the past, since she spoke Aymara and was de pollera.
N (%)
Home Hospital Combined
Aymara 17(94%) 6 (86%) 23 (92%)
Quechua 1 (6%) 0 (0%) 1 (4%)
Spanish only 0 (0%) 1 (14%) 1 (4%)
Style of dress
Traditional (de pollera) 12 (71%) 4(57%) 16(67%)
Transitional or de vestido 5 (29%) 3 (43%) 8 (33%)
Indian surnames
0 1 (5%) 0 (0%) 1 (4%)
1 7 (35%) 3 (43%) 10 (37%)
2 12 (60%) 4(57%) 16 (59%)
Table 4.3: Ethnic characteristics among interviewees
To assess differences between interviewed and chart review groups, interviewees were compared with chart-review cases with at least one indian surname. These cases are referred to as the chart-review indian group. As discussed, surnames are not the best indicators of ethnicity and cultural identity in Bolivia, and we cannot expect that women with one or more indian surname will be similar in all other respects. Yet, for the chart review data, surnames were the only available indicators of ethnicity. Comparison of the chart review indian group and the combined interviewees found no statistically significant differences on any variable (Table 4.4). Due to small sample sizes, quantified data on interviewees were not compared statistically between home-birthers and hospital-birthers, or between interviewed and chart-review home-birthing women, on the one hand, and interviewed and chart-review hospital-birthing women, on the other hand.
Mam or N (%)
Home Interviewed Hospital Interviewed Combined Interviewed Chart Review Indian
BIOLOGICAL Maximum mean arterial pressure 87 89 88 90
Weight gain, kgs 6.7 5.3 6.4 5.8
Hemoglobin 14.3 13.6 14.1 13.6
Height, cms 150.2 148.4 149.6 150.0
OUTCOME OF STUDY PREGNANCY Birthweigbt, kgs 3290 3190

Table 4,4 (Cont.)
Mean o r N (%'
Home Interviewed Hospital Interviewed Combined Interviewed Chart Review Indian
Stillbirth 1 (5%) 0 (0%) 1 (4%) 0 (0%)
Length of gestation, weeks 40.3 39.5 40.0 39.2
Birth interval, months 43.6 60.8 48.1 41.6
Gravidity 4.2 3.1 3.9 3.5
Party 4.0 2.7 3.6 3.3
Previous miscarriages
0 16 (80%) 4(57%) 20(74%) 97(82%)
1 or more 4 (20%) 3 (43%) 7(26%) 22(18%)
Previous stillbirths 0 17 (85%) 7(100%) 24 (89%) 119(100%)
1 or more 3 (15%) 0 (0%) 3(11%) 0 (0%)
Number of prenatal visits 5.7 4.7 5.5 5.9
Weeks.pregnant at. 1 visit 19.7 24.4 21.0. 20.4
Age 28.4 28.6 28.5 28.5
Husbands ethnicity, by number of indian surnames
0 2(10%) 0(0%) 2 (8%) : -22(20%)
1 . 12(60%) 2 (50%) .14(58%) 37(34%)
2 6 (30%) 2 (50%) 8 (33%) 50 (46%)
Marital status Married or livingtogether 20(100%) 7(100%) 27(100%) 107 (92%)
Single 0.(0%) 0 (Q%) .0 (0%) . .9 (8%)
City of residence
El Alto 15 (75%) 5 (71%) 20 (74%) 74 (63%)
La Paz or other 5 (25%) 2 (29%) 7 (26%) 43 (37%)
Relation to insured
Self 4(20%). 1 (L4%) . 5 (19%) 20 (17%)
Beneficiary 16 (80%) 4(57%) 20 (74%) .96(83%)
Uninsured 0 (0%) 2 (29%) 2 (7%) 0 (0%)
SOCIOECONOMIC Monthly income, in Bolivianos 767 (US$128) 762 (US$127) 766 (US$128) 800 (US$133)
Number of deceased children among women with. previous live births
0 13 (68%) 3 (60%) 16 (67%) 79(78%)
1 or more 6(32%)- 2 (40%) 8 (33%) 23 (22%)
Number of deceased children as a fraction, of live .12. .14 .12 ,05
births among women with previous live births Education
None or primary 10 (53%) 2 (29%) 12 (47%) 64(55%)
Secondary . 6(32%) 5 (71%). 11(42%) 42(36%)
University. 3(15%). 0(0%) 3(11%) 10 (9%)
City water in home Yes 19 (95.0%) 5 (100.0%) 24(96%) 107(90%)
No 1 (5.0%) 0 (0%) . I (4%) 12 (10%)
City sewage in home Yes 17(85:0%) 3 (60%) 20 (80%) 86 (72%)
No 3 (15.0%) 2(40%) 5 (20%) 33 (28%)
Occupation Unsalaried 17 (85%) 6 (86%) 23 (85%) 103 (87%)
Salaried 3 (15%) I (14%) 4(15%) 15 (13 %)
Husbands occupation
Unsalaried 11(58%) 6(86%) 17(65%) 63 (59%)
Salaried 8(42%) 1 (14%) 9 (35%) 43 (41%)
Table 4.4: Tabulations of medical chart. variables for home interviewed, hospital interviewed, combined interviewed, and chart-review Indian groups. There are no statistically significant differences between the combined interviewees and the chart-review indians. (Note that sample sizes vary because of missing data for some items.)

The home interviewed and hospital interviewed groups appear similar on most variables. Differences that may be important will be mentioned briefly here. The only recorded stillbirths occurred in the home-birthing interviewed groupthree women had had previous stillbirths, and another experienced a stillbirth with the study delivery. A greater percentage of hospital-birthers had had previous miscarriages. Birth interval was somewhat longer among hospital-birthers.. The greatest portion of home-birthers were educated to the primary level, while the greatest portion of hospitahbirthers were educated to the secondary level.
Home- and hospital-birthers were also similar in all of the following factors related to household economics and access to health services (numbers refer to the items in Table 3.4):
#1 and 4 Occupation and income: Both groups were similar in terms of occupation. Fifty-five and 57% of the home-birthers and the hospital-birthers, respectively, were homemakers. Another.fbur (20%) of the home-birthers were comerciantes, although none of the hospital-birthers were. There was no difference in monthly income, which averaged 161 Bolivianos (US$128) per month for home-birthers and 762 Bolivianos (US$127) for hospital-birthers.
#2,5, and 5 Opportunity costs of missed work or household duties: The number of days rest postpartum before returning to normal duties was meant to measure the possibility that women who birth at home do so because the costs of missed work during a postpartum interim in the hospital were too great. Women who are insured workers receive maternity leave for 45 days prior to and 45 days after a birth. However, all women suggested that they would be returning to household work before any work outside the home. Women who birthed at home most often reported that eight days (36%) or one month (36%) from the birth would pass before they went back to their normal duties. In the meanwhile, someone else in the family would assume the household chores, and gradually the woman would re-adopt them. The most common response from hospital-birthers (57%) was that they would return to household work immediately after their postpartum stay in the hospital, normally two days, because there was no one to help. I can think of two possible explanations for this difference.
1) Women with little support at home choose hospital birth because they will be guaranteed at least two days of rest.
2) Women who birth at home are more likely to report a cultural ideal of how long a woman should rest after giving birth. This possibility seems likely considering the postpartum prohibition in Aymaran culture against handling water for cooking and washing clothing. Women are said to remain entirely in

bed for at least the first eight days. Hospital-birthers may disregard this prohibition and therefore feel freer to tell an outsider what will happen when they return home.
Most importantly, because the normal hospital stay is two days postpartum and no home-birthing woman said she returned to her duties before two days after the birth, a hospital stay did not appear to delay the resumption of productive activities.
#6 Transportation: Are women with personal access to vehicles and telephones more likely to birth in the hospital? All of the interviewees used public transportation (bus or taxi) to get to the hospital for prenatal visits and/or labor and delivery attention. Only one of the home-birthers and one of the hospital-birthers said a family, friend, or coworker had a vehicle that was available to them for emergencies. Accessibility of phone services is important for calling a taxi or ambulance when labor begins in the middle of the night. Three of the home-birthers and one of the hospital-birthers had a telephone or cell phone available in their homes. Ambulance services do not seem to be frequently used. The service in La Paz uses a single emergency number for reaching all the major hospitals; but the ambulances do not go to El Alto, where most of the interviewed women lived. Furthermore, none of the hospital-birthers had used an ambulance to get to the hospital. In sum, hospital birthers did not rely on personal transportation and communication tools. Rather, they used public systems.
#7 Costs of hospital services: All of the hospitals services are free when they are ordered by a physician. Therefore, none of the interviewees paid anything for their obstetrical care.
#8 Social network buffering-. Analysis of social network buffering was not accomplished. Because social network analysis is very time-consuming, it could only have been accomplished in follow-up interviews. Logistical barriers and mistrust of outsiders made it impossible to visit with interviewees a second time. This would be a good area for further studies.
Some of the above items are quantified in Table 4.5.

Mean or N (%)
Home Hospital Combined
Vehicle available to household
Yes 1(5%) l (14%) 2(8%)
No 18(95%) 6 (86%) 24 (92%)
Telephone in home or cell phone available
Yes 3(15%) 1 (20%) 4(16%)
No 17(85%) 4(80%) 21 (84%)
Number of days of rest postpartum
none 2(14%) 4(57%) 6(29%)
eigfal days 5 (36%) 1 (14%) 6 (29%)
two weeks 2 (14%) 1 (14%) 3 (13%)
One month 5 (36%) 1 (14%) 6 (29%)
Pregnancy was intended
Yes 6 (40%) 2 (29%) 8(36%)
No 9 (60%) 5 (71%) 14(64%)
Has used IUD in the past
Yes 1 (7%) I (14%) 2 (10%)
No 13 (93%) 6 (86%) 19 (90%)
Has used calendar method
Yes 5(36%) 2(29%) 7(33%)
No 9 (64%) 5 (71%) 14(67%)
Age at first birth 20.7 22.0 21.0
Table 4.5: Tabulations of additional variables of interviewed groups
Contraception history was similar between interviewed groups. Age at first birth was slightly lower among home-birthers. See Table 4.5. One difference stood out between the two groups (see Table 4.6).
N (%)
Place of origin/history of migration Home Hospital Combined
Currently lives in rural area 2(11%) 0 (0%) 2(8%)
Migrated as adult 10 (55%) 0(0%) 10 (42%)
Migrated as child 5 (28%) 1 (17%) 6 (25%)
Bom in La Paz/El Alto 1 (6%) 5 (83%) 6 (25%)
Table 4.6: Place of origin/history of migration among interviewees
The greatest portion of home-birthers migrated to the city as adults (56%), with the youngest age of migration being 17 and the next youngest 20. All (100%) of the hospital-birthers had been bom and/or raised in La Paz/El Alto.
In sum, home- and hospital-birthing interviewees appeared similar in most respects. Among indian women, the economic costs of hospital birth, including the costs of missed work or familial obligations, did not seem to prohibit hospital birth. This finding is similar to those of other studies (Thaddeus and Maine 1994:1094-1095). Yet, because the study presented here sampled women with health insurance, we cannot assume that conclusions would be the same for uninsured women in La Paz/El Alto. Public laws dictate that health centers attend women in labor or showing signs of obstetrical complication, whether they are able to pay or not. However, women are required to show proof of residence and other official documents to

receive obstetrical care services. These and other factors may effectively limit hospital births among uninsured women.
The finding that women who birth at home were overwhelmingly migrants to the city at an adult age matches well with the results of the chart review phase, in which home-birthing women were found to be poorer and less educated than hospital birthers. They were also more likely to live in El Alto rather than La Paz and, regardless of they resided in, to live in poorly developed neighborhoods, as evidenced by the lack of sewage services. Rural-urban migrants are often poor and have come to the city hoping to improve their quality of life. In a four-volume study of migration to La Paz/El Alto, Albo, Greaves, and Sandoval report that the most common reason for leaving the countryside is the lack of economic opportunities for survival (see Albo, Greaves and Sandoval 1981:59). After a few years, most migrants do experience greater economic success than they had in the countryside, but they remain poor by the citys standards (Sandoval, Albo and Greaves 1987:122). Constrained by poverty and, initially, by limited social contacts in the city, they are forced to live in the least developed neighborhoods. Since affordable neighborhoods in La Paz have become crowded, rural migrants have become concentrated in El Alto, and El Alto is spoken of as a campesino city. Education levels in the countryside are less than in urban areas (ENDSA 1998:18), and school attendance begins at a later age (Morales and Rocabado 1988:75) This is especially true for girls, who forego schooling to assist with household duties and animal care more often than boys (Morales and Rocabado 1988; 141, 342, 177-179).
Rural-urban migration is a major social phenomenon in Bolivia. In the 1980s the percentage of La Paz residents that had migrated from rural areas was reported at 38% (Morales and Rocabado 1988:50). Migration to the city subsequently increased in response to the natural disasters (flooding and drought) of the mid-80s, the depletion and closing of mines, and the national economic crisis that followed (Sandoval, Albo and Greaves 1987:57). These migrants have neither re-created rural life in the city or adopted city life altogether. Rather, they have developed a distinct urban Aymara culture that incorporates elements of rural and urban life (Albo, Greaves and Sandoval 1983:3-37) (see background section). As I will explain in later sections, interviewed women expressed similar ideas about pregnancy and birth, regardless of where they gave birth. I conceive of the arena of reproductive health for urban Aymaras as containing multiple ideologies and priorities, possible through the meeting of rural and urban cultures. When it comes to place of birth, the choice of home or hospital birth depends not so much on which cultural beliefs a woman holds, but on which of these ideologies and priorities are weighted more heavily and how they interact with situational factors.

In the coming sections, I describe home and hospital birth. Although many writers have dichotomized home and hospital birth according to the set of ideologies behind them, on most counts, Bolivian indians do not reject or accept either set; rather, they balance them. I will also explain this balancing act and how it differs between women who grew up in the countryside and those who grew up in the city.

Ethnographic Description
Hospital Birth
The following description of labor and delivery at Hospital 18 de Mayo comes from 25 hours of observation in the labor ward and delivery rooms, on two different days. During this time, I witnessed five vaginal births and two cesareans, as well as the labors of several women. When they had free time, I was able to talk with hospital personnel, in particular the interns and residents, who perceived me as a peer. Most hospital workers assumed I was a doctor and addressed me as such, presumably because I was a white foreigner who was allowed to wander the hallways. Even after I explained I was not a doctor, I was called doctorcito (little doctor), signaling that I was considered to have some medical authority.
Hospital environment. The physical structure of Hospital 18 de Mayo is probably similar to most hospitals in the developing world in that it appears somewhat broken-down. The walls need paint, the floors are cracked, revealing at least three layers of linoleum and concrete underneath, and wards are crowded, lined with ten to twelve beds apiece. Like almost all buildings in La Paz, there is no central heating. Space heatersare found in the labor ward and the delivery rooms. Because of the above structural problems, a new hospital is being built and is planned to open sometime in 2001.
Protecting the privacy of patients is a constant problem. Doors to the labor ward and the adjoining examination room are usually left open, in spite of signs on them that read keep door shut. Women wearing ill-fitting gowns that do not close in the back are given vaginal exams while hospital personnel wander in and out.
Hospital personnel j of course, are used to this., -but patients often-complain about it to each other.
Sanitation standards are not high. Hand-washing is infrequent, and gloves, because they must be sanitized and reused, are reserved for vaginal examinations, enemas, and the delivery room. Sheets lying on stretchers that are used to transport women to the delivery room are reused, even when blood is visible on them.

Course of patients movement through hospital. Upon arrival with signs of labor, a woman is taken to emergency, the exam room just off the labor ward. She changes into a hospital gown in the bathroom, emerges (always holding it together in the back), and lies on the exam table for her exam. If she exhibits any signs of labor, the doctor will ascertain how far she lives from the hospital before deciding whether she should stay or go home and return with more-active labor. If she stays, she will undergo an enema and perineal shave and be sent to the resting ward. Thus, many women who live in El Alto and other outlying-parts of the city spend hours or days in the hospital before early labor becomes regularized and active. Notably, there are just as many beds in the resting wards as there are in the postpartum wards. When labor becomes active (defined as 4 centimeters dilation and 3 or more contractions within ten minutes), the woman is moved to the labor ward. There she stays, with doctors and nurses nearby, until she is ten centimeters dilated and the birth is imminent. Next someone yells camiUal (stretcher!)into the hallway, an orderly shows up with a stretcher, the woman is wheeled into the delivery room, and she has her baby. After the birth, the baby is placed on her chest, and both mother and infantare wheeled into a postpartum ward. The usual postpartum stay is two days, or five days after a cesarean, during which time the baby sleeps in its mothers bed.
Pain management. Analgesics are seldom used, except for epidural anesthesia for cesareans. I am not sure why, but general anesthesia is used for a legmdo uterino instrumental, which is either dilation and curretage or dilation and evacuation (I was unable to find a translation). Neither do hospital personnel instruct laboring women in non-pharmaceutieal forms of pain management, such as breathing techniques, massage, body positioning, etc. (see Varney 1997:421-431). Family members are allowed into the resting wards during the afternoons but restricted from the labor ward, except for short visitsof twenty minutes or so. In general, then, the woman is on her own to deal with the pain. Breathing techniques are taught during prenatal classes at the hospital, and some women I observed seemed to be trying them. But more common sights were women grabbing the bed poles, holding their breath, and biting their sweaters. Few women cried out in pain,-but, rather, appeared stoical. Bradby (1999) writes that Quechua women of the valleys of Bolivia desire to demonstrate their strength during labor. Perhaps thisis true of mestizo and/orIndian women in the Altiplano as well.
Technologies and techniques. Technology used during labor is much more restricted than it is in the U.S.(see Davis-Floyd 1992 for an anthropological discussion of routine technologies in biomedical obstetrics). Something that has always disturbed me about hospital birth in the U.S. is that women appear somewhat tied to their beds, with intravenous tubes, electronic fetal monitor (EFM) cords, and epidural and urinary catheters if they have had an epidural. At 18 de Mayo, none of

these things are routine Unfortunately, it seems that the hospital has chosen to adopt the practice of restricting the laboring woman from food and liquids by moutb(at least during active labor) without adopting the additional practice of intravenous hydration (see Varney 1997:410-411). If women are-given an intravenous drip, it is usually only after several hours without an increase in dilation. Intravenous medications include artificial oxytocin and antibiotics, if the membranes have been ruptured for several hours. The hospital has an electronic fetal monitor (EFM), but it remains locked in a closet to which only one person has a key, and doctors complain that administrators will not buy paper for it. As mentioned above, epidurals are only done for cesareans. Thus, most women labor unconnected to any apparatus.
The state ofthe fetus is not determined by reading EFM tape, but by periodic counting of fetal heart tones with aPinardhom or therare use of a sonographic unit. The Pinard horn is a wooden or plastic instrument that is placed between the womans abdomen and the health workers ear in order to amplify fetal heart tones. In my experiences with homebirth midwives in Colorado, Iowa, and Texas, the Pinard hom was very rarely used, but it is less expensive than a fetoscope.
One practice of labor that used to be very common in the U.S. (Varney 1997 409-410) and that is standard at 15 de Mayo is that ofthe enema and perineal shave. Davis-Floyd (1992) explains how obstetrical practices can take on ritual value that supercedes their medical efficacy. At 18 .de Mayo, the enema seems to have taken on ritualistic value, since it is done when it is sometimes illogical to do so. For example, when a woman is in very early labor and is being admitted to the resting ward, she will be given an enema even though she will receive regular meals until her labor becomes active. TheJegrado (explainedabove) is another practice that is over-used. When a woman who has delivered at home presents herself at the hospital within the first day postpartum, she will undergo general anesthesia and a Jegrado. This is true even when the placenta has birthed normally and she has come because her infant is having problems, as happened to one of my informants. Performing a Jegrado in these circumstances may serve the purpose of standardizing the womans birth, giving her a baseline of clinical care so that the clinically unrecorded events of her birth become less important.
During the birth itself the woman is draped with surgical cloth, put in the lithotomy position, and instructed how to push. Both rectal pressure and fundal pressure are common. Episiotomy is routine for primiparas and very common with other patients.
The umbilical cord is cut immediately and the placenta is allowed to emerge spontaneously. The infant is dried off, measured, given an injection of vitamin K and

antibiotic eyedrops, nnd then dressed and put on the mothers chest. Mothers-are instructed to begin breastfeeding immediately, in spite of the fact that it is impossible to breastfeed in the lithotomy position, where she remains until the episiotomy is done being sutured. However, breastfeeding is facilitated at 18 de Mayo by the practice of keeping infants with their mothers throughout the postpartum stay.
Laboring and birth position. As mentioned, women birth in the lithotomy position. However, one of the residents reported that a few women have insisted on birthing in a squat, and that he complied. Perhaps an assertive patient is able to bend hospital norms.
Because women are not usually attached to an apparatus during labor, I was surprised to find that they remain in bed except to go to the bathroom. The cold atmosphere and the lightweight hospital gowns may be a factor; it is certain to be warmer under a blanket. Yet, even during midday, when temperatures were comfortable, women did not do much walking through the room, squatting at the end of their beds, or trying other positions that tend to make labor more bearable than does lying down.
Authoritative knowledge and national progress. One of the commonalties of hospitals everywhere is that doctors are considered the gatekeepers of knowledge about the birth process (Jordan [1978] 1993; Davis-Floyd and Sargent 1997). At 18 de Mayo, it does not matter to hospital personnel what a woman is thinking about the course of her labor. For example, if a woman feels her health is being threatened by the presence of open doors (a belief explained below), the doctors correct her with the knowledge that her temperature is normal. Likewise, a womans urge to push is invalid in the doctors eyes until they have determined that she is fully dilated. In the hospital context, then, authoritative knowledge is hierarchically distributed and entirely in the hands of medical staff. To underline this point, hospital personnel use juridical languagehomebirth is prohibido (prohibited), a statement that is used to scold women who have birthed at home. Hospital practices are also backed by the ideology of modernization and progress, an ideology that carries great force in Bolivia. New government programs as well as new consumer products on the market are sold to the public as modem. I have heard a hospital social worker say to a woman who had just birthed at home, Es el ano 2000. No sabes que parto en domicilio es atrasado? (Its the year 2000. Dont you know that homebirth is backwards?) This statement is more than a personal insult; it is an accusation that women who birth at home hold the country back from progress.

In sum, hospital obstetrics in Bolivia carry the sanctions of medical authority and national ideology on progress. Pharmaceutical interventions during labor are few, but the birth event itself happens much as it does in U.S. hospitals. Matemologico 18 de Mayo is short on supplies and resources, like most institutions in developing countries.
The following description of homebirth is based on information from the twenty interviews in the interviewed home-birthing group, four of the interviewed hospital-birthers, who had had previous homebirths, and two interviews with Aymara midwives.
Homebirth is distinguished from hospital birth in terms of technologies and techniques, the roles of participants, and ideas about risk. Beliefs about dangers associated with labor and the distribution of obstetrical knowledge within society are important in analysis below. I have outlined a hierarchy of risks involved in the birth process, and how Aymaras explain why or how things can go wrong. Inasmuch as ethnography is the translation of one culture into the terms of another, this hierarchy is imposed and artificial. It is constructed here for our, the outsiders, understanding, but it would never be described in the same way by a woman of the Altiplano. I speak of two categories of risk: threats, which are ever-present dangers during labor and birth, and which may cause complications, and complications, deviations from the course of a normal and safe birth that require intervention to prevent the loss of life.
Cold During the birth process, it is very important that a woman keep warm. For this reason, there is some danger involved with leaving the house in the middle of the night during labor to go to the hospital. The room or rooms she occupies are closed off to prevent air from passing through. She must abrigarse, bundle herself up with warm clothes, and drink warm mates (herbal infusions). The explanation for this belief comes from humoral medicine. Because the laboring and birthing body of the woman is in a warm state, cold is a threat. Los poros abren (the pores open), making the woman unusually vulnerable to the elements. This condition lasts until at least eight days after the birth. During this time, the woman must stay in bed and must not wash herself, although most interviewees reported washing at least their genitalia after the birth. Note that the prohibition against rising out of bed in the first days may be the best treatment for vaginal tears that have not been stitched.

Fear. If a woman experiences fear, the baby may never emerge or it may even die. Fear is connected to the concept of verguenza, shame or embarassment. For example, several of the women who birthed at home said that they were afraid of the doctors seeing them naked. Although prenatal patients at 18 de Mayo are required to undergo an initial physical examination, including a vaginal exam, in a hospital gown, this experience is short-lived, compared to the birthing process, and is not said to incur harm on either mother or baby. In my interview sample, only one woman said she always delivered her babies in complete solitude. The reason she gave was that the presence of others at the moment of birth frightened her, and, no matter how hard she pushed, she.could not birth her baby. Neither did she accept assistance with the placenta. She explained how she discovered this: Her waters broke in the middle of the street, and a group of women formed around.her. They decided to call the nuns at the nearby health center. Upon arrival, the nuns drove her to her house and sent a midwife over.
Pero usted vera que no podia tener con,la partera.. .Podia, tal seria, no se si seria el miede, o la verguenza, y mismo no puedo decir, que es lo que ha impodido, pero no podia. La senora me tenia en la cama, me tenia, me decia, empuja, empuja, empuja, fuerzas, fuerzas... no podia. Me dolia mucha, me dolia, no podia, de repente mi hermana de afriera golpea a la puerta.. .Entonces la partera ha salido afuera... Yo ya me he aprovechado, de levantarme de la cama y ir a donde la silla, y vuelvo a ir recien tener.
But youll see that I couldnt have the baby with the midwife. ..Was it that, I dont know if it was fear, or shame, I cant say myself, what it was that stopped it; but I couldnt do it. She had me in the bed, she told me, push, push, push, hard, hard. ..I couldnt do it. It hurt a lot, I couldnt do it, all of a sudden my sister knocked on the door .. Then the midwife went outside... I took advantage of this chance to get up from the bed and go to the chair, and then I went about having my baby again.
After this birth, the midwife recommended that she birth all her babies in solitude:
La senora me ha dicho, sabes que, me ha dicho, debias que siempre acostumbres tener solita... Seguramente tienes verguenza, por eso debe ser, por las personas no somos iguales. Hay differentes clases de parto, a unas les gusta que la gente le atienden, pero algunas, no, somos bien reservadas, asi somos, no podemos.
The woman told me, you know, she told me, you should get used to having babies by yourself. ..You are ashamed, it must be that, because not all people

are the same. There are different kinds of births, some like to have people attend them, but others, no, we are very reserved, thats how we are, we cant do it like that.
A fright is also said to turn a womans breastmilk bad.
Huahua en mala position. Huahua (or wawa or guagua) is an indian word for baby and/or fetus. A malpositioned fetus is not a complication, as we would expect, but rather a threat, because it must constantly be watched for. The babys position is the most important determinant of the length of labor. It must lie directly central in the womb, what obstetricians would call occiput anterior (OA), for a quick and easy birth. Thus, even a baby lying at left occiput anterior (LOA), for example, can be considered malpositioned and the cause of prolonged labor. Massage (frotando or una sobada) to adjust the babys position is perhaps the most common feature of Aymara obstetrics. It can be performed by a midwife, a family member, or the woman herself, if she is alone during labor. External cephalic version is also practiced among the Aymara, when the baby is determined to be in a breech presentation. Yet this is a relatively rare occurrence compared to the use of massage to re-position babies that biomedical obstetrics considers normally placed. More information about this technique, and its relation to pulse-reading, is given below.
Complications. This list of complications seems very short in comparison to what biomedical obstetrics identifies as complications of labor and delivery. In a 1996 MotherCare/Bolivia survey, respondents recognized other labor and delivery problems not mentioned by my informants, including hemorrhage, infection, and eclampsia (1998:75). It is possible that the small scope of my study restricted the number of complications discussed during interviews to those most common or most severe. Alternatively, because the MotherCare survey included women of all cultural backgrounds, while I spoke only to Aymara women, the additional problems discovered may be recognized by mestizo women but not by Aymara women. In any case, all the complications discussed here are associated with pain, according to Aymaras. Although when asked pointedly, women will not deny that they experienced pain during a homebirth, it will be seen below that to claim a pain-free labor is to claim a normal labor.
Prolonged labor. Certainly most women would agree that a speedy labor is a good one. Yet, where health services are scarce, as they have been throughout Bolivias history, a long labor is not just undesirable, it is dangerous. Regardless of the problems etiology, the laboring woman may exhaust and die. Prolonged labor is most often thought to be caused by fetal malpositioning, but it may also result if the woman feels afraid.

Sobreparto. Sobreparto is an ethnomedical category for which there does not seem to be an equivalent in biomedicine. It is characterized by chills, trembling, weakness, headaches, and/or the feeling that cold has entered the nerves. It is also described as the continuation of labor-like painsthus, the name sobreparto, literally, in addition to birth. Caused by exposure to cold, especially cold water, it strikes anytime in the first month after the birth, but most frequently in the first days postpartum. To avoid it, women must refrain from handling water and keep themselves warm. Some interviewees said it would be a month before they returned to washing clothes because of this danger, a practice found in other parts of Latin America (Lefeber 1994:71). Women may either treat sobreparto at home, with warming mates, or seek clinical attention.
Placental retention. Once the umbilical cord has been cut, there is a danger of the placenta rising up into the womans chest, in search of the newborn from which it has just been detached. Thus, either the cord must not be cut before the placenta is bom, or the cut end of the cord must be attached to the womans toe or some other convenient place to prevent the placenta from rising. Women know that if the placenta is not bom, it may cause death. Consequently, nine percent of home-birthers in the medical chart review sought medical attention at Hospital 18 de Mayo in the first hours after the birth due to placental retention and/or bleeding associated with retention of fragments.
Attitudes about pregnancy and birth. Pregnancy is treated as a completely normal state that for the most part does not require a woman to stop her usual activities. When I asked interviewees what precautions a woman needed to take during pregnancy, a few replied that she should not lift heavy things. Some mentioned that she needed to stay warm but avoid being in the sun, a sensible prescription in a climate characterized by the simultaneous presence of extreme cold in the shade and fierce ultraviolet rays in the sunshine. Other precautions included in the literature are the cessation of weaving and knitting and the avoidance of stirring a spoon in a bowl, all of whi ch may cause the cord to wrap around the fetus (Cajias 1984:16-17; Huanca, Palma, and Alvarez 2000). Note that in Huanca, Palma, and Alvarezs survey of 100 women who birthed at home in La Paz/El Alto (2000), 67% of respondents said no precautions were necessary when pregnant.
Local idioms about birth suggest that it is viewed as a sickness. Women are said to enfermarse (to get sick) at the moment of birth and postpartum complications are called recaidas (relapses). Yet, in other respects, birth is treated as a completely normal event that does not require expert knowledge unless something goes awry. Huanca, Palma, and Alvarez (2000) explain:

El parto es designado con la palabra enfermedad pese a ser la culmination de un proceso nonnal. En esta dicotomia se pone de manifiesto el equilibrio precario y la convivencia de salud con enfermedad: hay enfermedades que son necesarias y hasta beneficioas y esperadas como en este caso.
Birth is designated with the word illness even though it is the culmination of a normal process. This dichotomy manifests the precarious equilibrium and the co-existence of health and sickness: there are illnesses that are beneficial and desirable, as in this case (my translation).
One interviewee explained that, Dices que estas enferma porque te sientes debil (You say you are sick because you feel weak).
Technologies of labor and birth. Homebirth in Bolivia appears to happen with the use of very few technologies. Although women do not simply squat in the bushes as romantic notions of indian birth would have it (for example, see Suzanne Arms Immaculate Deception), neither are elaborate rituals or techniques involved.
The main component of medicinas caseras, or home remedies, in Bolivia, is mate (herbal infusion). A mate is prepared by boiling the herb in water; it is then drunk. Mates are usually the first level of resort when a family member has an ailment. As in many parts of Latin America, it appears to be the women who choose and prepare such treatments for use (Browner 1989:465). During labor and delivery, mates used include azahar,flor de naranja, oregano, anis, chocolate, chamomile, and romero. Pildoraspujantes are sometimes used (Huanca, Palma, and Alvarez 2000), although none of the women I spoke with used them. I do not know what these pills contain. In any case, all the mates used in labor are referred to as pujantes (pushers), because they are meant to speed labor by pushing the baby out. They are also meant to keep women warm during labor.
Four of the women I interviewed said they had given the newborn mate following a homebirth, a practice that I have also seen in Northern Mexico and that is reported for Guatemala (Lefeber 1994:72). The reason usually given was that the womans milk had not come in yet and the baby was hungry. Colostrum is believed to be detrimental to the newborn, although babies were usually breastfed within six to twelve hours postpartum, certainly before a womans colostrum had been replaced by milk. One woman, who was not Aymara, but Quechua, said that the purpose for giving the newborn mate was to clean its stomach of the blood that had accumulated there during the pregnancy. This blood explains why the babys first stool is black. Unfortunately, I never thought to ask if sugar was mixed into mates given to

newborns. If so, the mate would provide glucose. Glucose water is often given to newborns in U.S. hospitals. Mate de romero or chamomile may also be used to treat vaginal tears, although women often implied that there was nothing that could be done for a tear at home.
The other common practice during labor is to massage the womans abdomen. In George Engelmanns 1884 survey of childbirth around the world, he wrote, There is hardly a people, ancient or modem, that do not in some way resort to massage and expression in labor, even if it be a natural and easy one (1884:180). Massage may be a form of pain management, but its expressed purpose among the Aymara is to arrange the fetus within the uterus correctly. As explained above, it is believed that the fetus must lie at the very center of the womb for the birth to be quick and easy. Pulse-reading is related to this belief, as it can indicate on which side of the uterus the fetus is lying, and therefore what direction massage strokes should take. Radial pulses on both wrists are palpated simultaneously. The side with a weaker pulse is the side in which the fetus is lying. Pulses may also be read during pregnancy to indicate when the baby will be bom and what sex it is.
I have heard that Aymaras will throw a woman in a blanket in order to treat a malpositioned fetus, a practice called manteo. However, I only heard this from mestizos, and it is not mentioned in any of the literature. One interviewee mentioned that, during her pregnancy, she liked to lie on a blanket and have her husband pull up the two comers near her feet and rock her legs back and forth. She said it helped relieve swelling in her legs and made her relaxed. I suspect that some sort of practice involving a blanket, such as this one, has been misconstrued or misunderstood by mestizos.
Laboring and birth positions. The most common position for the second stage of labor (the pushing stage) was reclining. Other positions were on the knees and squatting. This finding was the same as that found in Huanca, Palma, and Alvarezs survey of 100 home-birthing women in La Paz. During labor, women usually described themselves as moving between several positions, including lying down, sitting, and walking. Some women had strong preferences for one position:
Yo siempre de rodillas tengo mis bebes, no achada, me puedo morir, me
puedo morir, me pone de rodillas, entonces, alii que tengo mi hijo.
I always have my babies on my knees, not lying down, I could die that like, so
I get on my knees, and there I have my baby.

Placenta. As discussed above, precautions must be taken to prevent the placenta from rising into the womans chest and causing death. Thus, the umbilical cord should not be cut, or should be attached to something outside the womans body, until the placenta births. The placenta is the focus of all the ritualistic elements of homebirth. It is washed, dried, buried, and/or thrown into the garbage. Three interviewees said their placentas were thrown into the garbage, which in poorly developed neighborhoods is usually a large pile in the street. They did this despite others' beliefs that the placenta must be cared for properly. One said:
Cuando le botan, dice que no tienen nada.. .Es una creencia.. .Enterrarlo en la tierra, todo tienen, nada le falta, es una casa llena.
When it gets thrown out, they say that they wont have anything... Its a belief.. .Burying it in the ground, they have everything, the house is frill.
Some women reported drying the placenta in the sun until it became like dried llama meat, then like powder. One of the midwives explained how to wash and bury a placenta:
En la casa se lo lavan, con doce aguas...Todo la sangre hay que exprimirlo, y hacerlo chiquitito, limpio de doce aguas.. .Entonces este se ponen en una bolsita, y guardo, y se nos enterrarlo para la tierra.
The placenta is washed twelve times... All the blood must be shaken off it so it is made very small and clean from the twelve rinses.. .Then it is put in a bag and buried in the ground.
Most women buried the placenta just outside the house, but it is sometimes buried inside in a comer, when the floor of the house is dirt. Other writers (Huanca, Palma, and Alvarez 2000) have reported that the placenta is buried because it is connected to the souls of the mother and the child. As one husband said during an interview,
Como es una mujer, los enterramos (Because it is a woman, we bury it). But even he, like most of the other informants, said the best explanation was that it was a custom passed down from older generations.
Expert knowledge and participation in birth. Who do Aymara women recognize as the experts on pregnancy and birth?
Physicians. Even though home-birthers in the chart review attended fewer prenatal appointments, home-birthing interviewees agreed with women who birthed in the hospital that doctors are highly qualified at prenatal care. Reasons women gave

for attending prenatal check-ups were to make sure that the fetus was developing properly and to prevent problems. However, these two responses were repeated identically at each interview, and further questioning about what prevention meant to women was usually replied to with a list of techniques performed at prenatal visits weighing the mother, measuring the belly and listening to the belly. Only two interviewees said they attend prenatal care in order to receive the governmental prenatal supplement (el subsidio prenatal). During the five to eighth months of pregnancy, a woman covered by the national health care system can receive powdered milk, butter, iodized salt, sweetener, milk substitutes, and yogurt, if she gets a doctors signature on the appropriate form attesting that she has attended prenatal care that month (see Remision Reglamento Asignaciones Familiares, 2000, Caja Nacional de Salud). I suspect that this reward program was the incentive that initially brought most home-birthing women to prenatal appointments. However, once she has monthly contact with health professionals, she may begin to feel that the routines of measuring weight and fundal height and listening to fetal heart tones are necessary for the health of the baby. Postpartum/infant care may be motivated by government supplements as well, since the program continues during the first year of an infants life, if the mother has proof that the child is receiving vaccinations.
Concerning labor and birth, the physicians special expertise is limited to performing cesareans. Home-birthing women said that other women sometimes go to the hospital to birth because they want it to be over as fast as possiblewith a cesarean. Because analgesics are not routinely used in the hospital, a cesarean is indeed the quickest way to end the painEs mas rapido, un grande ayuda (Its faster, a big help). Contrary to the Bolivian Quechua women in Bradbys study (1999), Aymaran interviewees did not seem to fear surgical intervention. Rather, they considered it necessary and even desirable at times. It was outside the scope of this study to examine perceptions of how long a normal labor can last, but home-birthing women and midwives alike said that a woman who is unsuccessful birthing at home must go to the hospital for a cesarean.
Midwives. Traditional birth attendants or midwives were hard to find in the city. Everyone agreed that they must exist, but that they didnt know one personally. Eventually I made contact with PROMUJER, a womens credit and health education association, and people there put me in touch with a midwife. I also struck some luck one day at the hospital when the mother of a woman who had just birthed at home came in with the baby and identified herself as a midwife.
Two words for midwife exist in Boliviapartera and matrona. There may, however, be additional words in Aymara that I do not know. In general, partera and matrona seemed to be used interchangeably, except that partera carried a connotation

of professionalism. Since Jordans Birth in Four Cultures (1978), there has been a tendency in anthropology to assume that there is only one type of traditional midwife, although her techniques or beliefs may vary between cultures. This midwife is a well respected bearer ofreproductive-knowledge, has-apprenticed for years to learn, her. trade, comes with a specific set of instruments to be used during labor and delivery, depends on her practice for her livelihood, andattends the majority^ of her community. Yet many cultures have more than one category of birth assistant. Some kinds of birth assistants fit the above description- others areprimarily religious^ practitioners and come to births to perform rituals; and others only take care of household duties while the mother is incapacitated (Lefeber 1994:9-10). In her review of literature on birth practices, Goldsmith concludes that midwives as we think of them are a relatively recent social development, and thaL historically, women in most tribal groups delivered with their own mothers help or without any aid at all (1990:24).
Thus, the role of the Aymaran midwife needs to be critically examined. Midwlves are notfrequently usedin Bolivia^only 7% of births in the. whole country are attended by them (ENDSA 1998; 137). Likewise, out of 56 total previous homebirths among interviewed women, 7% had a midwife present (see Table 4.7).
Birth setting Primary helper Number of previous births (% within group)
Home-birthing group Hospital-birthing group Combined groups
Home Husband 13(24) 0(0) 13 (18)
Mother-in-law IT (21) 2(10) 13(18)
Mother 6(H) atio> 8(11)
Other family member 3(6) 2(10) 5(7)
Midwife 5(9) 0(0) S
Doctortaurse auxiliary 5(9) 0(0) -.5(7)
Self 4(8) 0(0) 4(5)
Neighbor n4> r(5) 3(4)
Hospital health Hospital pasonnei - 4(8) B(65) 17(23)
Total 53(100). 20 (100) 73 (100)-.

Table 4.7: Settings and primary helpers for previous births of interviewees
Although government programs designed during the 1990s called for the training of midwives (see-Blan Decenal de Accion Para La Ninez y^La Mujer, Presidencia de La Republica 1993, p.56 and 58), and PAHO reports that trained midwives are active (2000), Lcould find no evidence.that these programs were, functioning. Interviewed midwives learned through experience at their own and others births, rather, than in training sessionsor through apprenticeship. They have two recognized pieces of expert knowledgepositioning the fetus and reading the mothers pulsealthough even positioning the fetus-through massage is not solely used by midwives. Some midwives, of course, have other skills, such as stretching the cervix manually on a woman with a matriz pequeno (small uterus). Although neither interviewed midwife provided prenatal care, other interviewees reported

having-seen a midwife duringtheir. pregnancy-when they perceivedthat the-fetus was malpositioned. Interviewed midwives did not carry instruments or materials with them to a birth, buLrather used-whatever was available in the birthing-woman! a home for making mates and cutting the umbilical cord. Compensation for a midwifes services is on a voluntary basis. Those interviewed said they usually receive between 20 and 60 Bolivianos (about US$3.50 to $10) or an in-kind gift such as fish or produce for help ata-birth. The existence-of male-midwives has. been reported in Bolivia (Morales and Rocabado 1988:224; Lefeber 1994:10), although without accompanying explanations. This would be an interesting area for further research.
Aymara midwives seem to have the same scope of practice as the Quechua midwives in Bradbys study (1999:296), Because they have other sourcesof income, attend few births, and, for the most part, practice skills that are also used by nonmidwives, midwives in Bolivia are not professionalized. Furthermore, midwives do-not participate in the national organization of traditional medicine (Sociedad Boliviana de Medicina Tradicional [SOBOMETRA]).
Common knowledge and assistance from family members. Thus, when it comes to labor and-delivery, both doctors and midwives have small arenas of expertise, according to Aymara women who birth at home. In general, they are sought out only when complications arise. The primary- complications recognized by respondents, as discussed throughout this section, are prolonged labor, placental retention, and sobreparto. Inthe- absenceof these-problems, the-home--birther sees-no need for help from outside the family. At least 32% of all births in Bolivia are aided only by a family member or friend, which is- 74%-of all homebirths (ENBSA 1998:139). Table 4.7 shows that, among the 56 previous homebirths of interviewed women, seventy percent had family members as the primary helper. Among these-births, 87% were attended by the husband, the husbands mother, or the womans own mother. Children, the-womans father, or a more distant relation like an aunt were less often the primary helpers.
In some cases, helpers were very involved, making mates, gathering warm clothes and blankets, physically supporting thewoman as she labored, and receiving-the baby and the placenta. In other cases, helpers slept through the night until woken when the woman felt the birth was imminent It should be noted, however, that receiving the baby does not necessarily mean doing perineal support or helping the baby pass through the vagina. It may simply mean picking up the baby after it is fully bom.
Bender writes that the Aymara typically seek treatment for illnesses from neither traditional practitioners nor clinical services. Rather, they treat themselves

with medicinas easeras like mates, saving that they know just asmueh about theuse of herbal medicines as-professional healers (Bender-1983:13). In normal labor-and birth, then, too, Aymara families are typically self-sufficient. Goldsmith believes this has been the dominant pattern throughout the worlds In a large part of. the worlds except in difficult cases, a birth required no more specialized help than the members of a womans own family could provide (1990:25). Knowledge about birth is horizontally distributedeach woman who has gone through a birth knows as much about it as the next one.
Balancing Ideologies and Situational Factors
Clearly, there are important differences between birth in the hospital and birth at home for urban Aymara women. Yet, womens evaluations of each birth setting were similar regardless of- where-ihey. gave birth themselves. For example, women who birthed in the hospital were just as likely to say that one is better cared for at home as the women who birthed at home. There was- also agreement that the hospital was cold and drafty, but that being attended there guarantees that labor will not last too long, becausea cesarean canke-done We. have already concludedthat differences-in financial and logistical accessibility to hospital services cannot fully explain the choice of birth setting. If women, who birthu at home and women who birth-in the hospital have similar ideas about the advantages and disadvantages of each birth setting, what then determines their choice?
Two themes frequently arose within interview conversationsan inability to reach the hospital and personal assessments of danger during labor. When asked about their choice of birth setting, most women who birthed at home first spoke of the difficulties of getting to the hospital in the middle of the night, or of precipitous labors that caught them unawares and left no time to get to the hospital. These are acceptable rationale, as every Bolivian, knows the inefficiencies of the-countrys.. transportation system. Yet, as conversations continued, women moved away from this stance and began to refer directly or indirectly to understandings of normalcy and risk during labor and birth. Women who birthed in the hospital did not speak of logistical barriers, since obviously, they had overcome them. Instead, they referred to another structural factorfinancial access to health care. They simply said they birth in the hospital because they have health insurance. Four hospital-birthing women.hadhad previous homebirths and said they began birthing in the hospital once their husbands
As mentioned earlier, women are usually household managers of health and make treatment selection (Browner 1989:465). Likewise, wherrhusbands were present during interviews, theimarratives of their wives births demonstrated that the help they gave during labor and delivery was under the direction of their wives or mothers.

had found employment that provided National -Health Fund coverage. Yet, eventually, they too came to discussions of obstetrical risk.
I propose that the way a woman conceives of risk interacts with situational factors to produce choiceofbirth, setting- These situational factors include an ongoing evaluation of risk during the course of labor and an evaluation of the logistical difficulties involved in getting to.the hospital.. Here I will summarize-the ideologies of risk at play in the arena of obstetrical health for urban Aymaras, the perception of risk within the unique circumstances of a birth event, and-the evaluation of barriers associated with transportation and communication systems in La Paz/El Alto.
Ideologies of risk. We can, identify two ideologies of risk, one associatedwith hospital birth and the other associated with homebirth. By dichotomizing two ideologies, I do not mean, to suggest that they are entirely, in opposition. Note, for example, that it is possible to believe that complications are avoided through both monitoring with biomedical techniques and wardingoff cold and fear. Nor do I think that all hospital workers always agree with the first two statements and always disagree with the second two, or that Aymara women always adhere to the traditional ideology of obstetrical risk. I only mean to present a framework for thinking about different conceptions, of risk that operate within urban Aymara- culture Theterm. biomedical is used because hospital birth is a biomedical event. The term traditional is used because homebirth preceded the arrival of hospital facilities in Bolivian society.
The biomedical ideology of obstetrical risk:
1) Expert knowledge is needed to determine that labor and birth are happening normally,
2) Complications are avoided by monitoring labor and delivery with biomedical techniques.
The traditional ideology of obstetrical risk:
1) Common knowledge is sufficient to determine that labor is happening normally.
2) Complications are prevented by warding off a set of threats, including cold, fear, and a fetus that is not optimally positioned.
These statements summarize the important ideological differences between the ways birth are done in the hospital and at home and are based in the descriptions of home and hospital birth provided above. Fuller explanations of these ideas are found in those sections.

Situational perception- of risk. Before andduring labor, women must-assess the dangers to which they are being exposed, the course of their labors, and the overall degree of risk involved. In interviews, women often defended-their choices to birth at home by saying that they felt no pain. Statements that labor was normal were usually coupled with- statements that it was relatively painless':Estaba trabajando normal; no he sufiido mucho (I was laboring normally; I wasnt suffering much).
Yet this does not mean that they believed that any pain was a sign that something was going wrong with their labors. When asked directly, they did not deny that labor is at least very uncomfortable and usually very painful. Therefore, I suggest that pain is used as a shorthand in conversation for the type of suffering associated with problematic births.
In anthropological studies of choice of health service, the perceived severity or gravity of an illness is usually a primary consideration (Young. 1980;. Thaddeus and Maine 1994). In labor, risk assessment is an ongoing process as women consider possible obstetrical outcomes. Obviously, what factors they consider- depends on their particular conception of risk. Women who planned on birthing in the hospital from the beginning never suggested that labor is always too dangerous for anywojnan to birth at home. Rather, their viewed each situation as unique.
Logistical difficulties Few residents of La Paz/El Alto have personal autos. Most people use the bus system, which is organized by an extensive set of routes through the city. During the daytime, the main roads of La Paz and El Alto are filled with buses, and one never has to wait more than a few minutes to catch one on the right ro ute. To get to places off the main roads, one may have to change buses two or three times and do some walking. Taxis are also available but, in usual circumstances, are used only by those-of higfi economic standing. After about ten oclock at nightr both buses and taxis become scarce, even on the main roads, and impossible to find a few hours later.
Thirty-four percent of residents in Bolivias urban areas have a telephone (ENDS A 1998). The rest of them use public phones provided by the national Cotel system and the international Entel system, which are found in phone booths on the street. Usually these booths require that the user previously purchased a phone, card or token. In addition, many small shops make their private phone lines available for a fee. To understand the accessibility of these services, I wandered through several neighborhoods in El Alto where most interviewees lived. I found that both public phones and shops with phones are very scarce. Phone booths are found only on the largest plazas and intersections, which are concentrated near the center of the city. In the outlying neighborhoods, I was lucky to find a single shop with a phone. Shops

close for the night. Of course, residents of El Alto are familiar with their neighborhoods and may know of resources that I did not discover.
All the major hospitals in La Paz are connected by a city-wide ambulance service. Women are given the emergency phone number at prenatal visits. But, for some reason, the ambulances do not go to El Alto.
I had originally expected to find that indian women who birthed at home would be more economically disadvantaged than hospital-birthing, women. Therefore, they would live in more marginalized neighborhoods and have less access to resources needed for transport to the hospital. This distinction may apply between home-birthers and hospital-birthers when all Pacena women are considered. Yet, among urban indian women, interview-data showed that hospital birthers were just as likely to live in poorly developed neighborhoods of El Alto and were not more likely to have phones and/or autos. By income-and education, they fell into the same socioeconomic stratum. Yet interviewed hospital-birthers were able to get themselves to the hospital, sometimes in the middle of the night, and always with public transportation.
Certainly structural barriers related to weak transportation and communication systems constrain options for health care, especially for conditions like labor, the onset of which cannot be predicted. Yet the fact that they can be overcome demonstrates that, at least in this urban, area^-they are not the final boundaries on choice of birth setting. Rather, they are interpreted and evaluated by women and their families. Interviewed women assessed the difficulty of obtaining transportation according to the time of day or night in which they labored and the distance to a telephone. These considerations affected their decisions on where to birth.
Case Studies
The strength of ideas and beliefs a woman has about birth is balanced with situational factors, so that the context in which women make birth choices is both ideological and material. For example, the woman who starts labor at home but begins to experience problems must consider how severe those problems are and how much effort she is willing to exert to get to the hospital, a consideration that is regulated by her ideas of risk factors during labor. Will she put herself more at risk by stepping out into a cold, windy night? Will attempting to straighten the baby resolve the problem, or is a doctor required to decide whether or not she needs a. cesarean? Even the woman who plans to birth in the hospital must decide at what point after labor begins she should go to the hospital. Will it be too dangerous to wait until

morning, when her husband will not have to walk an hour to find a phone for calling a taxi?
Below I present three case studies of urban Aymara women and their decision-making processes about birth setting. Like nearly, all the women I interviewed, they had had at least one birth at home, and had also had a hospital birth, or in some other way had used biomedical obstetrical services during the birth process.
Flora. Flora grew up in a rural area of the Altiplano and has lived part-time in El Alta for five-years, since she-was 28. Sheis now 32 and has just had her. fourth child. Her husband is a rural schoolteacher who is given a room to live in at his school. He comes to EL Alto occasionally on weekends. When. Flora lived.there with
him, she tended sheep and a few cows..She decided to get. a house in El Alto in order
to begin earning cash to supplement the subsistence, her livestock provided She began selling fruits at an in El Alto. Flora had dark patches on her cheeks (mowc/jas), like. many women, of the Altiplano. who are exposed to intense sunlight, and wore a traditional Aymara outfit, except she had replaced the womens hat with a knit stocking cap. Flora spoke Aymara as a first-language.
During her last pregnancy-, she attended seven prenatal visits, beginning at the 18th week of her pregnancy She says that prenatal visits are important because the doctors can tell if the baby is developing normally:
Ya que ellos- sabery han estudlado desarrollomiento envientre de una mujer. Entonces, yo creo que estan entera capacitados para saber como esta el nino al dentro del estomago.
They know, they have studied development inside the womb of a woman. So,
I believe that they are whplly qualified to know how the child is inside the stomach.
Her previous births all happened in her rural home. She emphasizes that she never had any assistance at these births, but that she gave birth como ovejita (like a little sheep). Her last labor began at night, and the baby had been bom by the next morning. Her husband was in the countryside, and her eldest daughter of 12 years was watching the younger children while she labored. She says she thought about coming to the hospital. However:

No tenia ni dolor, nada, EL doctor esta diciendo cualquier dolor tienes, pues a venir. Queria venir, no tenia ni dolor, normal, caminaba, caminaba, no tenia ni dolor.. .El doctor me ha dicho, cualquier dolor, vas a llegar a la emergencia.
I wasnt having any pain. The doctor says, whatever pain you have, then come. I wanted ter come, I didnt have any pain, it was normal, I was walking and walking, I didnt have any pain.. .The doctor told me, come to emergency with the first pain.
After the babys birth, she felt normal and was tranquila (relaxed) watching TV.
But the placenta did not birth, for several hours, and so her daughter, flagged, down, a taxi to take her to the nearest clinic. The placenta was manually removed at the clinic, and then she was sent to Hospital 18 de Mayo, since she is covered by the national health service through her husbands job. Her hemoglobin was found to be extremely low, and an ultrasound determined-that there were, still placental fragments .in her uterus. In the next two days, she underwent two legrados (explained above). By the time I met her on her third day in the hospital, she was angry and. impatient to be discharged. She said that she felt fine and wanted to go home. Furthermore, she was tired of having so many people trying to help her.
Floras explanation of her decision-making process both validated the authority of her doctors and demonstrated her desire for autonomy in her health care. She wished to show me that she complied with the doctors orders to come to the hospital at the first pain by explaining that she never felt any pain, a point that underlined the normalcy of her labor. She also believed that physicians are best qualified to judge the health of her unborn baby, and she sought, their attention when her placenta failed to emerge. Yet, once it had been removed, she felt the hospital had served it purpose and that she was capable of returning home and taking care of herself. Before her five day stay was over, I frequently saw Flora peeking around walls in the hallways of the hospital, as if she were going to try to escape, not an unheard of event at 18 de Mayo.
Juana. Juana was a 37 year old woman who was bom and grew up in the neighborhoods that border the highway between La Paz and El Alto. Shefirst said-she was a homemaker, and then added that she also worked at home as a tejedora. lt took her. three days to make a manta,, which she could, sell for about US$4, Occasionally she supplemented that income with sewing jobs that other women brought to her. Her husband had been a police officer for eleven-years and at that time worked, as a night guard at a school. She spoke Aymara, but said she preferred to speak Spanish with her five children. She. wore. atraditional./?o//era, a derby hat, m& chalita shoes,, and her hair was in the traditional style of two braids attached at their ends.

Her first child was bom, in the home-of her mother- andfatherdm-law, where she and her husband were living at the time. Both of these in-laws helped at her birth, as did her husband. Her second pregnancy ended in- a miscarriage. When labor began with her third pregnancy, her husband and his parents were busy with the harvest in a rural area. Becauseshe was-alone, her neighbors, both men and women, helped her by making mates, massaging her abdomen, and sending for help at the nearest health clinic. The health clinic sent a nurse to the house, something they do for emergency situations, but by the time the nurse arrived, the baby had been bom. The nurse was able to receive the placenta, and then she left. During-Juanas third labor and birth (her fourth pregnancy), her husband and his parents helped once again.
About this time, her husband became a police officer, which meant he and his family became insured by the national health service. Thus, with Juanas fifth pregnancy, she wasableto attend prenatal care and deliver her baby at Hospital IS de Mayo. She did not have a phone or a car, but when labor started in the middle of the night, her husband went to a public phone and calleda taxi. Again with her most recent pregnancy and the birth of her fifth child, she was attended at 18 de Mayo. Her labor had once again started during the night, but this time, rather than-using a, taxi, she waited until morning came and took a bus to the hospital.
When I asked-her why she- decided to start birthing her babies in the hospital after having three homebirths, she said:
Antes era joven y me sentia mas fuerte. No me sentia mis partos como ahora. Ahora me siento cansancio... Con m,i primero hijo, me sentia ninguno hasta el momento del parto.
Before, I was young and I felt stronger. I didnt feel my births like now. Now I feel tired... With my first child, I felt nothing until the moment of birth.
She also talked about the fear of death, although she put it in third person, as if speaking about other women who choose to birth in the hospital. [Ellas] estande miedo, de que van a morir, estan bien controlada. Muchas veces otras mueren, no ve? ([They] are afraid^that they are going to die, they see the doctor a lot. Lots of times, other women die, dont you see?) When asked why some women would choose to birth at home, she said:
[En el hospital] se hace caminar. Despues el parto, va a levantarte, lavarte, caminarte, y en la casa, creo que se cuidan tres dias, en la cama. Despues sc lavan, se abrigan bien en tu casa.. .Los que tengan en la casa dicen, en el

hospital, se meten frio, y le afecta. Camison no mas ponen aqul En lacasa, es mas cuidadosa^ hay agua caliente... Que te cuidas mejor en la casa, te abrigas bien.
In the hospital, they make you walk. After the birth, they are going to get you up, wash you, make you walk, and in the house, I think they take care of you for three days, in bed. After you wash, you bundle up well at home...Those that have it at home say, in the hospital, you. catch cold, and that affects yoiL-You only wear a nightgown here. At home, the care is better, theres hot water... You take better care at home, you dress warm.
When I asked if she worried about these things when she was in labor in the hospital, she replied, A veces,.tambieivpero que.vapos a hacer? (Sometimes, this too, but what are we going to do?
After the interview, I asked Juana if she had any questions for me. She asked, Seria bueno-usar siempre anticonceptivos? (Would it be good to always use contraceptives?) and went onto explain that she was conformada (satisfied) with five children. In fact,, she had been satisfied with three children, but the calendar method did not work consistently and she did not trust the IUD. Her sister-indaw had used the IUD, and it had caused her to go looking, for other men besides, her husband, which eventually destroyed her home. Finally, the woman got cancer and diedse lo abro, y todo cancer (They opened her,.and it was all cancer). Yet, Juana wanted, to find a birth control method she could trust, and so she sought my opinion about whether or not other contraceptives were as dangerous as the IUD.
Juanas comment about the possibility of death showed that she felt safer birthing in the hospital. But she more directly used Aymara notions-about what constitutes normal labor and birth to justify her decision to birth in the hospital, saying-that homebirth isTor women who have normal, painless labors. In her case, older age made her births more-risky. Sheagreed with most interviewed women that one is better cared for at home, and that, in the hospital, a woman is not protected well enough from the cold. Yet her comment that nothing could be done about it implied that this risk was less-important than the risks she would expose herself to if she birthed at home. Juana did not, however, trust all the advice givenat the hospital. She strongly believed that the IUD coulddestroy ones home and health and was, by extension, mistrustful of other contraceptives.
Antonia. Antonia migrated.:to-LaPaz froma rural area when she was eleven-years old. At the time of our interview, she was 26 years old, working at an egg distribution center, and had just delivered her third baby. She wore sweatpants and a

sweater both times Lsaw her, and her hair was cut fairly short. She said she knew-Aymara but spoke Spanish at home. Her husband worked as a mechanic.
Her first child had died when it was a few months old. She said the baby fell off the bed, and then developed .a fever, vomiting, and diarrhea. When she took him to the hospital, it was too late to save him.
All three of her births had happened at home with the help of her husband and her mother. Her husband massaged-her abdomen and made mate de oregano during her labor and mate de chocolate after the birth. The placentas were, dried for several days and then buried ma comer of the outdoor patio. At first, Antonia said that her husband had called the ambulance service with his cell phone, but the ambulance never came. When I asked if the same thing had happened during her two previous births, Antonia replied that she was confident about her abilities to birth at home. She said, Casi no sufromucho mispartos... Yo secomo es mi cuerpa (I dont suffer much during my births. .I know how my bodyis). She also said it was mostly fear of the doctors that kept her fronr delivering in the hospital, even though she had been insured since before her first birth. In addition to her fear of doctors, she was concerned about open windows and doors in the hospital:
Para mi, que es en mi casa mejor, porqueen el hospital, bien esta abierto las puertas por el eruza deviento, todo eso. Hay una le sopla... Asi te mete la enfermedad.
For me, that its better in my house, because in the hospital, the doors are wide open and the wind crosses, all of that. It blows on you.. .Thats how you get sick.
Yet, after each of her births, she has gone to the hospital in the first hours postpartum, even though everything had happened normally. Each time she and her baby have been admitted for a regular postpartum stay of two days. Each time they have given her a legrado (explained above), in spite of the feet that the placentas delivered normally. Antonio found this reassuring, since the doctors could examine her and her baby to make sure everything was okay.
Antonias pattern of admitting herself to the hospital only after the birth was complete is another example of how Aymara women optimize their options for obstetrical care. By combining care provided at home during labor and birth with postpartum care at 15 de Mayo, Antonia-experienced the more personal attention available with homebirth as well as the medically sanctioned confirmation that she and her baby were healthy. She was confident in her ability and knowledge about

birthing-at home, but, at the-same-time, she wanted to do all shexouldto ensure that she and her baby were healthy. Eor her, this-entailed a doctors attention. Her fear of doctors and the risk involved with the cold atmosphere in the hospital were put aside after the birth so that she could get this attention. 1

This research found that women who birth at home are likely to be of rural origin, and, correspondingly, to be indian, to have lower education levels, and-to live in the newest, most underdeveloped neighborhoods of the city, In their birth accounts, my informants balanced ideological and situational factors related to conceptions of risk, ongoing evaluation of levekof risk, and difficulties encountered in getting to the hospital from another part of the city. Hospital-birthing women were motivated to overcome the inefficiencies of the citys transportation and communication systems because, for them, the risks involved if they were to have a homebirth were greater than the risks involved with hospital birth. This model may not be comprehensive. Other factors that I did not explore thoroughly during interviews may weigh in as well, for example, treatment by hospital staff.
I began with the hypothesis that women who birth at home in La Paz/El Alto do so because they are less able to incur the economic costs of hospital birth, including the costs of missed work or familial obligations. I conclude that economic disadvantage is not a direct determinant of birth setting for these urban women. It is an extrinsic but related factor. Women who migrate to the city from the countryside tend to weight more heavily traditional Aymara understandings of obstetrical risk and the role of obstetrical experts. They also tend to be poorer than their urban counterparts.
Their poor economic standing, higher gravidity,, fewer prenatal visits, and lower levels of education do not seem to result in less healthy pregnancies and births, as judged by hemoglobin, blood pressures, weight gain during prenatal care, birthweight, and length of gestation. As analyses from the larger project on high altitude and pregnancy demonstrate, Aymaran women less often experience diseases of pregnancy such as pre-eclampsia, pre-term labor, and IUGR, compared to mestizo women living at the same altitude. Thus, Aymaran ancestry may provide some genetic protection against complications associated with high altitude (Armaza et al 2000). One hypothesis is that indian women and mestizo women appear similar in outcomes because genetic protection may offset the effects of poverty during pregnancy for indian women, while many mestizo women may experience less severe

poverty but suffer greater effects of an oxygen poor environment. However, what happens immediately after the birth and during the first years of the childs life may differ in important ways for home-birthing women. As mentioned, they may experience greater postpartum complications, and they are more likely to suffer the death of children.
The balance between the factors that affect choice of birth setting differs between women who grew up in the countryside and women who grew up in the city. It is not that women of urban origin have experienced greater losses related to birthing. To the contrary, it was the home-birthing interviewees who had experienced all previous cases of stillbirth, as I reported above. Perhaps women who grew up in La Paz/El Alto have been reached by more public health messages about the dangers of homebirth. Perhaps lower levels of education in rural areas have limited womens exposure to biomedical concepts. Or perhaps they have observed that women who use hospital services do not die in childbirth. Hospital 18 de Mayos rate of one maternal death out of around 5000 live births produces a rate of200 deaths per 100,000 live births, less than half the national estimate of480. It would be impossible for women living in rural areas to notice that hospitals are successful at preventing maternal deaths, because communities there are often hours away from the nearest health center. But explaining just what it isin the personal histories of women of rural and urban origins that causes them to prioritize risk concerns differently is outside the scope of this study.
Further research is needed to address this question. This research should incorporate social and familial network analyses. One of the greatest weaknesses of the study presented here is that it treats the woman as the solitary decision-maker about birth setting. Women gave hints that these choices were affected and constrained by the desires of other family members, and other family members were always involved with bringing women to the hospital. In addition, the role of non-familial social networks needs to be considered. McClains study (1987) of social network and choice of birth setting in the United States found that women who birthed at home were much more likely to have friends that had birthed at home. Aymaras are accustomed to passing and gaining information orally (see Morales and Roeabado 1988:243). Clusters of Aymaran women in heated discussion are common sights in La Pazs and El Altos plazas. I imagine that they discuss, among other things, the pros and com of reproductive health care options. Clinic waiting rooms are also sites where information and opinions are shared between women. Asone informant said, Nosotros hablamos juntos en los controles (We talk together at the doctor visits). How do these discussions affect womens ideas about obstetrical risk and the need for expert knowledge about birth? Unfortunately, there was not enough time in interviews to look into these questions.

My conclusions are very similar to those made by Sargent in her study of home and hospital birth among the Bariba of Benin. Bariba women agreed with each other on indigenous beliefs about childbirth, whether they lived in the city or in the country. In the city, new agendas emerged,.such as status aspirations, that, induced women to birth in the hospital. Yet women continued to refer to traditional beliefs as well for making their decisions of where to deliver. Sargent writes, I conclude that medical choice represents the individuals effort to accommodate several agendas and reflects the articulation of ideological and material considerations (1989:204).
An important difference between Bariba and Aymara women in these studies is that Bariba women are sometimes motivated to birth in the hospital by status aspirations. Despite the ideological connection between national progress and hospital services in Bolivia, and literature reporting that rural-to-urban Aymaras often seek to align themselves with modem and civilized aspects of Bolivian society (Albo, Greaves, and Sandoval 1981), I could find no evidence that status aspirations affected womens decisions of birth setting. For example, no woman said she birthed in the hospital because she was not a poor campesina. Perhaps the difference between the situation of the Aymara and that of the Bariba is that the Bariba must pay for their hospital care. Because delivery and birth care are free for all Bolivian women, using the hospitals services is not a sign of high economic status.
In sum, this study concludes that the choice of birth setting is a multi-factorial process resulting from the meeting of rural and urban culture. The choice is not entirely structural, related to access to health services, nor pragmatic, based in the efficacy of health care services, nor entirely based in indigenous and/or biomedical beliefs about health. Ideological and material factors interact to produce choice. It is also important to stress that decision-making is a process and thus it is affected by changing circumstances. To improve maternal health care in Bolivia, this process must be well understood.

Albo, Xavier, Tomas Greaves, and Godofredo Sandoval
1981 Chukiyawu: La Cara Aymara de La Paz. I. El Paso a la Ciudad. Cuademo de Investigacion #20, CIPCA.
Albo, Xavier, Tomas Greaves, and Godofredo Sandoval
1981 Chukiyawu: La Cara Aymara de La Paz. IH. Cabalgando Entre Dos Mundos. Cuademo de Investigacion #24, CEPCA. Wayar and Soux Impresores.
Armaza, F., E. Vargas, M. Villena,.and L.G. Moore
2000 Contributions of population ancestry and pregnancy complications to altitude-associated intrauterine growth restriction (IUGR). Paper presented at the FASEB Summer Conference: Perinatal Regulation of the Cardiovascular System. Cooper Mountain, June 4-9, 2000.
Bradby, Barbara
1999 Will I Return or Not?: Migrant Women in Bolivia Negotiate Hospital Birth. Womens Studies International Fomm 22(3):287-301.
Browner, C.H.
1989 Women, Household and Flealth in Latin America. Social Science and Medicine 28(5):461-473.
Browner, Carole H., and Carolyn F. Sargent
1996 Anthropology and Studies of Human Reproduction. IN Medical
Anthropology: Contemporary Theory and Method. Rev. ed. Carolyn F. Sargent and Thomas M. Johnson, eds. Pp.219-234. Westport, CT: Praeger.
Gajias, Lupe
1984 El Embarazo y Las Tradiciones en Bolivia. 1st ed. La Paz, Bolivia: Antonio Paredes-Candia.

Chakraborty, R., S.A. Barton, R.E. Ferrell, and W.J. Schull
1989 Ethnicity determination by names among the Aymara of Chile and Bolivia. Human Biology 61(2): 159-177.
Chrisman, Noel J.
1977 The Health-Seeking Process: An Approach to the Natural History of Illness. Culture, Medicine and Psychiatry 1 (4):351 -377.
Colson, Anthony C.
1971 The Differential Use of Medical Resources in Developing Countries. Journal of Health and Social Behavior 12:226-237.
Crandon-Malamud, Libbet
1991 From the Fat of Our Souls: Social Change* Political Process, and Medical Pluralism in Bolivia. Berkeley: University of California Press.
Davis-Floyd, Robbie E.
1992 . Birth as an American Rite of Passage. Berkeley: University of
California Press.
Davis-Floyd, Robbie E., and Carolyn F. Sargent
1997 Childbirth and Authoritative Knowledge: Cross-Cultural Perspectives. Berkeley: University of California Press.
Daviss, Betty-Anne
1997 Heeding Warnings from the Canary, the Whale, and the Inuit: A Framework for Analyzing Competing Types of Knowledge About Childbirth. IN Childbirth and Authoritative Knowledge: Cross-Cultural Perspectives. Robbie E. Davis-Floyd and Carolyn F. Sargent, eds. Pp.441-473. Berkeley: University of California Press.
Duranti, Alessandro
1997 Theories of culture. IN Linguistic Anthropology. Pp. 23-50. Cambridge: Cambridge University Press.
Encuesta National de Demografia y Salud (ENDSA) 1998
1998 Instituto Nacional de Estadistica, Ministerio de Hacienda, Republica de Bolivia, Demographic and Health Surveys Program. Calverton, MD: Macro International, Inc.

Englemann, George
[1882] 1977 Labor among Primitive Peoples, New York: AMS Press. Foster, George M.
1976 Disease Etiologies in Non-Western Medical Systems. American Anthropologist 78:773-782.
Goforth, Lynnel
1988. Household Structure and Birth Attendant Choice in a Yucatec Maya Community. Ph.D. dissertation, University of California, Los Angeles.
Goldsmith, Judith
1990 Childbirth Wisdom -Prom tire Worlda Oldest Societies. Brookline, MA: East West Health Books.
Hahn, Robert A.
1987 Divisions of Labor: Obstetrician, Woman, and Society in Williams Obstetrics, 1903-1985. Medical Anthropology Quarterly l(3):256-282.
Huanca, Sonia, Miriam Palma, and Melva Alvarez
2000 Factores Relevantes Que Inciden en La Opcion deun Parto
Domiciliario en LaMatemidad 18 de Mayo, C.NJS, La Paz, Bolivia. Unpublished document.
Huntj Linda M., Brigitte Jordan, and Susan Irwin
1988 Views of Whats Wrong: Diagnosis and Patients Concepts of Illness. Social Science and Medicine 28(8):945-956.
Institute Nacional de Estadistica de Bolivia
2000 Indices Demografias. Electronic document, http: //www-
International Planned Parenthood Federation
1999, Country Profiles: Bolivia. Electronic document.
Janzen, John M.
1978 The Quest for Therapy in Lower Zaire. Berkeley: University of California Press.
1987 Therapy Management: Concept, Reality, Process. Medical Anthropology Quarterly 1:68-84.

Jordan, Brigitte
[1978] 1993 Birth in Four Cultures: A Crosscultural Investigation of
Childbirth iirYucatan, Holiand^Sweden; and the United: States. 4th ed. Prospect Heights, IL: Waveland Press, Inc.
1997 Authoritative Knowledge and Its Construction. IN Childbirth and Authoritative Knowledge: Cross-Cultural Perspectives. Robbie E. Davis-Floyd and Carolyn F. Sargent, eds. Pp.55-79. Berkeley: University of California Press.
Kaufert, Patricia A. and John ONeil
1993 Analysis of a Dialogue on Risks in Childbirth: Clinicians,
Epidemiologists, and Inuit Women. IN Knowledge, Power, and Practice: The Anthropology of Medicine and Everyday Life. Shirley Lindenbaum and Margaret Lock, eds. Pp. 32-54. Berkeley: University of California Press.
Kitzinger, Sheila
1997 Authoritative Touch in Childbirth: A Cross-Cultural Approach. IN
Childbirth and Authoritative Knowledge: Cross-Cultural Perspectives. Robbie E. Davis-Floyd and Carolyn F. Sargent,.eds. Pp.209-232. Berkeley: University of California Press.
Kleinman, Arthur
1978 Concepts and a Model for the Comparison of Medical Systems as Cultural Systems. Social Science and Medicine 12:85-93.
1988 The Illness Narratives: Suffering, Healing, and the Human Condition. New York: Free Press.
Koenig, Barbara
1988 The Technological Imperative in Medical Practice: The Social Creation of a Routine Treatment. IN Biomedicine Examined. Margaret Lock and Deborah Gordon, eds. Pp; 465-496. Dordrecht: Kluwer Academic Publishers.
Kroeger, Axel
1982 Anthropological and Socio-Medical Health Care Research in
Developing Countries. Social Science and Medicine 17(3): 147-161.

Lefeber, Yvonne
1994 Midwives Without Training: Practices and Beliefs of Traditional Birth Attendants in Africa, Asia and Latin America. Assen, The Netherlands: Van Gorcum.
Lewis, Oscar
1958 Five Families: Mexican Case Studies in the Culture ofPoverty. New York: Mentor.
Markens, S., Carole H. Browner, and Nancy Press
1999 Because of the risks: How U. S. Pregnant Women Account for
Refusing Prepatal Screening. Social Science and Medicine 49(3):359-369.
Martin, Emily
1987 The Woman in the Body. Boston: Beacon Press.
McClain, Carol Shepherd
1987 Some social network differences between women choosing home and hospital birth. Human Organization 2:146-152.
Mclver Weatherford, Jack
1987 Cocaine and the economic deterioration of Bolivia. IN Conformity and Conflict: Readings in Cultural Anthropology. James P. Spradley and David W. McCurdy, eds. Pp. 412-423. Boston: Little, Brown and Company.
Morales A., Rolando, and Fernando Rocabado
1988 Los Grupos Vulnerables en las Economias en Desarrollo: El Caso Boliviano. La Paz, Bolivia Ministerio de Planeamiento y Coordination.
1998 Estudio de liaea-de base en salad matema y perinataL La Paz, Bolivia: Stampa Grafica Digital.
Nations, Marilyn K., and L.A. Rebhun
1996 Development and Disease: Lessons from Brazil.. IN At the Edge of Development. Richard L. Guerrant, et al., eds. Durham, NC: Carolina Academic Press.

Navarro, Vincent
1974 The underdevelopment of health or the health of underdevelopment. Politics and Society 4(2):267-293.
Pan American Health Organization (PAHO)
2000 Health Analysis: Country Health Profiles: Electronic document.
Plan Decenal de Action para LaNinez y La Mujer
1993 Ministerio de Planeamiento y Coordination. Presidentia de la Republica de Bolivia.
Rapp, Rayna
1988 The Power of Positive. Diagnosis: Medical and Maternal Discourses on Amniocentesis. IN Childbirth in America: Anthropological Perspectives. Karen L. Michaelson, ed. Pp. 103-116. South Hadley MA: Bergin and Garvey Publishers.
1993 Accounting for Amniocentesis. IN Knowledge, Power, and Practice: The Anthropology of Medicine and Everyday Life. Shirley Lindenbaum and Margaret Lock, eds. Pp, 55-76. Berkeley: University of California Press.
Rothman, Barbara Katz
1988 The Decision to Have or Not to Have Amniocentisis [sic] for Prenatal Diagnosis. IN Childbirth in America: Anthropological Perspectives. Karen L. Micliaelson, ed. Pp. 90-102. South Hadley MA: Bergin and Garvey Publishers.
Sandoval, Godofredo, Xavier Albo,. and Tomas Greaves
1987 Chukiyawu: La Cara Aymara de La Paz. IV. Nuevos Lazos Con El Campo. Cuademo de Investigation #29, CIPCA. La Paz, Bolivia: CIPCA.
Sargent, Carolyn Fishel
1982 The Cultural Context of TherapeuticChoice: Obstetrical Care
Decisions Among the Bariba of Benin. Boston: D. Reidel Publishing Company.
1988 Maternity,. Medicine, and Power: Reproductive Decisions in Urban Benin. Berkeley: University of California Press.

Scheper-Hughes, Naiicy
1992 Death Without Weeping: The Violence of Everyday Life in Brazil. Berkeley: University of California Press.
Schwartz, Lola Romanucci
1969 The Hierarchy of Resort in Curative Practices: The Admiralty Islands, Melanesia. Journal of Health and Social Behavior 10:201-209.
Schuler, Sidney Ruth, Maria Eugenia Cheque, and Susanna Ranee
1994 Misinformation, mistrust, and mistreatment: family planning among Bolivian market women. Studies in Family Planning 25(4):211-222.
Varney, Helen
1997 Varneys Midwifery. 3rd ed. Sudbury, MA: Jones and Bartlett Publishers.
Winkelstein, Warren
1992 Determinants of Worldwide Health. American Journal of Public Health 82:931-932.
Worsley, Peter
1981 Non-Western Medical Systems. Annual Reviews in Anthropology 11:315-348.
Young, James
1980 A Model of Illness Treatment Decisions in a Tarascan Town. American Ethnologist 7(1): 106-131.

Full Text