|USFDC Home | USF Electronic Theses and Dissertations||| RSS|
This item is only available as the following downloads:
Breathing Easier: Ethnographic Study of Acut e Respiratory Infection in Children in Rural Ecuador by John S. Luque, M.A., M.P.H. A dissertation submitted in partial fulfillment of the requirement s for the degree of Doctor of Philosophy Department of Anthropology College of Arts and Sciences University of South Florida Major Professor: Linda Whiteford, Ph.D. Roberta Baer, Ph.D. David Himmelgreen, Ph.D. Eknath Naik, M.D., Ph.D. Graham Tobin, Ph.D. Date of Approval: December 6, 2005 Keywords: medical anthropology, health care-seeking, ethnomedicine, pneumonia, Andes Copyright 2006, John S. Luque
Dedication To Marie and Carolyn
Acknowledgments This dissertation would not have been possible without the assistance, encouragement, and positive reinforcement of all my teachers and professors who have helped me along the long road that has led to the end of my graduate school career. I would like to acknowl edge the committee for their constructive criticism and guidance throughout the proc ess: Dr. Linda Whiteford, for welcoming me to USF and providing me with the idea for the dissertation topic through my employment on the CDMHA gran t; Dr. Graham Tobin, for introducing me to the field of geography and providing valuable traini ng in statistical analysis; Dr. Robbie Baer, for her expertise in medical anthropology and for being a careful reader of my work; Dr. David Himmelgreen, for his careful guidance during the analysis of the di etary data; and Dr. Eknath Na ik, for introducing me to the field of infectious disease epidemiol ogy. The following is a short list of some of the great teachers who have had m ade an impact on me in chronological order: John Potter, Jerome Long, Ramnad Raghavan, John Chance, Bob Alvarez, Tod Swanson, Don Casimi ro Mamallacta, Neil Whitehead, Lakshminarayan Rajaram, Michael Angrosino, Robbie Baer, and Linda Whiteford. Also, I thank all of t he people in Penipe, Ec uador, and especially those who assisted and facilitated my research: Padre Jaime Alvarez, Rosa Hidalgo, Sandra Moreno, Jorge Lara, Mari a Guapulema, Juan Velastegui, Ruth Noemi Carrasco, Nelly Castillo, Yadira Villamagua, Hector Bonifaz, Guadalupe Barahona, Janet Serrano, Laura Rodrigue z, Ximena Haro, Lizeth Rodriguez, Nelly Recalde, Homero Vega, Ad ela Vimos, and Xeomara Menendez.
i TABLE OF CONTENTS LIST OF TABLES.................................................................................................iv LIST OF FI GURES...............................................................................................v ABSTRACT ..........................................................................................................vi CHAPTER ONE: INTRODUCT ION......................................................................1 Statement of Probl em.......................................................................................1 Introduction to the Research Site......................................................................2 Brief Overview of Theory and Methods.............................................................3 Research Hy potheses.......................................................................................7 Statement of Findi ngs.......................................................................................9 Structure of Di ssertatio n.................................................................................11 CHAPTER TWO: ECUADOR GEOGRAPH Y, DEMOGRAPHY, ECONOMY, AND CHILD HEALTH .........................................................................................13 Geography ......................................................................................................15 Demograp hy...................................................................................................19 Economy, Land, Right s, and D ebt..................................................................21 Child Health in Ecuad or..................................................................................29 CHAPTER THREE: RESEARCH SETTING, PRIOR RESEARCH, AND SITE SELECT ION.......................................................................................................36 Research Setting ............................................................................................36 CEBYCAMCES..............................................................................................39 Prior Rese arch................................................................................................44 Site Sele ction..................................................................................................47 CHAPTER FOUR: LITERATURE REVI EW THEORY, METHODOLOGY, AND ACUTE RESPIRATORY INFECTION................................................................49 PART I MEDICAL ANTH ROPOLOGY THEORY.............................................49 Medical Ecology and Critical Medical Anth ropology........................................50 Interchange Between Theorie s.......................................................................53 Theory in Resear ch Cont ext...........................................................................57 Political Ecology ..............................................................................................58 PART II ETHNOMEDICINE IN THE ANDES...................................................63 Studies of Andea n Ethnomedi cine..................................................................63 Biocultural Approac h.......................................................................................67 Medical Plur alism............................................................................................68 PART III METH ODOLOGY..............................................................................76 Ethnography for Epi demiologi sts....................................................................77
ii Collaborative Projects.....................................................................................82 Integration of Met hods.................................................................................... 85 PART IV EPIDEMIOL OGY OF ARI.................................................................87 Morbidity and Mort ality Tr ends........................................................................ 89 Etiology of ARI................................................................................................92 Etiology of Pneumon ia....................................................................................95 Risk Fact ors....................................................................................................96 Modifiable Risk Factor s...............................................................................97 Non-modifiable Ri sk Factor s.....................................................................106 Case Mana gement........................................................................................ 108 Meta-analysis of Intervention Tr ials on Case Management ..........................113 Antibiotic M anagement .................................................................................114 Preventio n.....................................................................................................116 Ethnographic Methods for t he Study of ARI..................................................119 Conclusi on....................................................................................................123 CHAPTER FIVE: METHODS ............................................................................126 Focus Gr oups............................................................................................... 128 Focus ........................................................................................................128 Selection Cr iteria .......................................................................................129 Participant-observation and Mother Info rmants ............................................130 Focus ........................................................................................................130 Selection Cr iteria .......................................................................................131 In-depth Interviews from Health Care Setting ...............................................131 Focus ........................................................................................................131 Selection Cr iteria .......................................................................................132 Semi-structured Q uestionnair e.....................................................................133 Sampling I ssues........................................................................................ 133 Focus ........................................................................................................135 Selection Cr iteria .......................................................................................136 Additional Re search ......................................................................................138 Data Anal ysis................................................................................................138 Limitati ons..................................................................................................... 142 CHAPTER SIX: RESULTS ...............................................................................149 PART I FOCUS GROUP DISCUSSION S......................................................151 PART II IN-DEPTH INTERVIEWS IN HEALTH CARE SETTING..................159 Training .........................................................................................................161 Access ..........................................................................................................162 Primary Health Needs ...................................................................................164 Delay in Health Care-seekin g.......................................................................164 Antibiotic Di spensing .....................................................................................167 Vaccination Co verage...................................................................................169 Traditional H ealers........................................................................................169 Costs .............................................................................................................170 PART III EPIDEMIOLOGICAL DATA FR OM MINISTRY OF HEALTH..........173
iii PART IV QUANTITATIVE DATA FROM QUESTION NAIRE.........................179 Demographi cs...............................................................................................179 Socioeconomic Status ..................................................................................182 Risk Factors fo r ARIs....................................................................................185 Childs Respirator y Problem s........................................................................189 Use of Health Care Resource s.....................................................................193 Diet of Child in the Last Twenty-four Hours ...................................................197 PART V QUALITATIVE DATA FROM STRUCTURED QUESTIONNAIRE...203 Use of Healthcare Facilitie s..........................................................................203 Attitudes towards Health care Facilit ies...................................................... 207 Barriers to Care.........................................................................................208 Hot Cough versus Cold C ough..................................................................... 209 Use of Cur anderos........................................................................................ 211 Home Rem edies........................................................................................... 214 CHAPTER SEVEN: ANALYSIS, TESTIN G THE RESEARCH HYPOTHESES, AND DISCU SSION...........................................................................................219 PART I ANAL YSIS........................................................................................219 PART II TESTING THE RESEARCH HYPOTHESES: RESULTS .................236 PART III DISC USSION..................................................................................244 CHAPTER EIGHT: CONCLUSION, RECOMMENDATIONS, AND CONTRIBUTIONS OF RESEAR CH.................................................................251 PART I CONCL USION..................................................................................251 PART II RECOMME NDATIONS ....................................................................260 PART III CONTRIBUTIONS TO APPLIE D ANTHROPOLOGY AND PUBLIC HEALTH ...........................................................................................................264 LIST OF REFE RENCES..................................................................................267 APPENDIX A: ENGLISH VERSION OF SEMI-STRUCTURED QUESTIONNAIRE .........................................................................................................................288 APPENDIX B: SPANISH VERSION OF SEMI-STRUCTURED QUESTIONNAIRE .........................................................................................................................305 APPENDIX C: FOCUS GR OUP QUESTION S.................................................322 APPENDIX D: IN-DEPTH DESCRIPT ION OF ADMINISTRATION OF QUESTIONNAIR E............................................................................................324 ABOUT THE AUTHOR .End Page
iv LIST OF TABLES Table 1: Key Child Health Indicators for Ecuadors Sier ra Provinces.................33 Table 2: Health of the Child, Accord ing to Selected Ch aracteristics...................35 Table 3: Population of Penipe County................................................................37 Table 4: Summary of Fo cus Group Discu ssions...............................................157 Table 5: Summary of Staff for H ealth Posts in Pe nipe County.........................160 Table 6: Summary of Responses from In-depth Intervie ws..............................171 Table 7: Demographics and SES of Fe male Caregiver Respondents..............180 Table 8: ARI Risk Factor s for Youngest Child..................................................186 Table 9: Youngest Childs Re spiratory Pr oblems.............................................189 Table 10: Caregiver Treatment Choi ce Depending on Symptom Cluster.........195 Table 11: Level of Severity of Symptom Cluster ...............................................196 Table 12: Daily Nutrition Intake for Ch ildren .....................................................202 Table 13: Use of Treatment Facility by Area.....................................................205 Table 14: Attitudes towards Healthcare Fa cilities .............................................208 Table 15: Hot and Cold Cough Know ledge......................................................211 Table 16: Use of Cur andero/a for Child............................................................ 213 Table 17: Commonly Used Remedi es..............................................................216 Table 18: Frequency of Nutrient Intake by Age Group .....................................231 Table 19: Frequency of Macronutrient Intake by Gender and Salary...............232 Table 20: Logistic Regression for Health Care-seeki ng Behavio r.....................238
v LIST OF FIGURES Figure 1: Map of Ec uador................................................................................... 17 Figure 2: Map of Ch imborazo Pr ovince ..............................................................18 Figure 3: Town of Peni pe...................................................................................38 Figure 4: Map of Penipe C ounty.........................................................................39 Figure 5: Entrance to CEBY CAM.......................................................................42 Figure 6: Penipe Coun ty Ambulanc e................................................................163 Figure 7: Penipe County Cases, Children < 1 yr. old........................................174 Figure 8: Penipe County Cases, Children 1 4 yr s. old...................................175 Figure 9: Penipe County ARI Morbidit y Rates for Six M onth Peri od.................176 Figure 10: Penipe County AR I Morbidity Rates................................................177 Figure 11: Map of Peni pe County To wns.........................................................181 Figure 12: Monthly Salary of Penipe Fa milies..................................................220 Figure 13: Care-seeking for Cough, Fe ver and Nasal Secretion by Monthly Salary ........................................................................................................222 Figure 14: Knowledge of ARI Symp toms..........................................................224 Figure 15: Percentage for each score on level of symptom severity................225 Figure 16: Rapid Breathing Heal th Care-seeking Behavior..............................226 Figure 17: Indrawn Chest Health Care-seeking Behavior.................................226 Figure 18: Rapid Breathing & Indrawn Chest Health Care-seeking Behavior...227 Figure 19: Vitamin A Inta ke by Age Group.......................................................233 Figure 20: Vitamin E Inta ke by Age Group.......................................................234 Figure 21: Zinc Intake by Age Group................................................................234 Figure 22: Bridge in Bilbao Pari sh....................................................................252
vi BREATHING EASIER: ETHNOGRAPHIC STUDY OF ACUT E RESPIRATORY INFECTION IN CHILDREN IN RURAL ECUADOR John S. Luque ABSTRACT This dissertation research utilizes ant hropological methods to determine the degree to which the signs and symptoms female caregivers identify as causes of acute respiratory infections (ARIs) in under-fives in rural Ecuador correspond with Western biomedical categories. By employing both a semi-structured medical history questionnaire and more open-ended ethnographic methods such as in-depth informant interviews and focus group research, the researcher identifies the factors which determine time ly health care-seeking behaviors of female caregivers in this case study. Economic factors such as the cost of medications and lost work hours were determined to be the primary financial obstacles for timely health care-seeki ng. Other barriers included limited and inconsistent hours at the health centers and transportation issues. Families of lower socioeconomic status were also more likely to have children suffering particular respiratory ailment s. Childhood respiratory ill ness was identified as an outcome of poverty, which had the pot ential to reproduce itself through the negative effect of illness on household income. However, the research determined that there was an overall lack of recognition of the biomedical signs and symptoms of serious lower respiratory infections regardless of socioeconomic status. The model of ethnomedicine supports the finding that compliance with timely health care-seek ing is limited wit hout collaboration between healthcare professionals and comm unities to work towards beneficial and achievable goals that are joined by a common purpose. By understanding local cultural beliefs towards ARIs, heal thcare professionals are in a better position to: (1) assess the accuracy or inaccuracy of ethnomedical beliefs and determine if there is a conflict in sympto m recognition and ca re-seeking behavior with the biomedical model; (2) determine cu lturally-appropriate interventions or recommendations to address the health pr oblems of the community and identify barriers; and (3) work with existing comm unity resources in order to foster effective health communication. This res earch finds that public health messages regarding ARIs be informed by ethnomedica l knowledge of hom e treatments and beliefs. Moreover, health centers need to adhere to regular hours of operation and increase staff capacity to better meet the needs of their clients.
1 CHAPTER ONE: INTRODUCTION Statement of Problem This dissertation research projec t utilizes anthropological research methods to examine mater nal health care-seeking behaviors for children under five with acute respiratory infections (A RI) in Penipe, a rural setting in Andean Ecuador. The purpose of the study is to determine the degree to which the ethnomedical signs mothers identify as causes for ARIs correspond with Western biomedical categories in order to pursue timely and appropriate medical care. Ethnomedicine is broadly defi ned as the study of how well-being and suffering are experienced bodily as well as soci ally, the mutlivocality of somatic communications, and processes of heali ng as they are contextualized and directed toward the person, household, co mmunity and state, land and cosmos (Nichter 1992:x). Biomedici ne is defined as the empirical study of medicine as it relates to all biological systems usi ng the standardized concepts, measures, and techniques that are practiced in Western societies (Browner, et al. 1988:682). The lack of immediate attention for severe ARIs based on differences of judgement concerning disease severity are barriers to timely health care-seeking. The research approaches the problem of AR Is from the perspec tive of medical pluralism, between ethnom edicine and biomedicine, in order to gauge
2 perceptions of severity of childhood ARIs in the small town of Penipe. Utilizing both available epidemiological data and ethnographic information generated by the more open-ended anthropological struct ured survey methods, the design of this project incorporates both publ ic health and applied anthropology. Introduction to the Research Site The dissertation field research visit to Ecuador was the researchers fifth time to the country over a period of f ourteen years. In 2 002, the researcher worked as a research assistant at the Un iversity of South Florida under a grant funded by the Center for Disaster Management and Humanitarian Assistance (CDMHA) titled, In the Shadow of the Volcano: Human Health and Community Resilience (Whiteford and Tobin 2002). The principal investigators, Dr. Linda Whiteford, Anthropology, and Dr. Graham Tobin, Geography, were interested in examining the health effects of ash fall and community resilience in the face of Mount Tungurahua, an active volcano which had erupted and caused forced evacuations in October 1999 (Whiteford and Tobin 2001). The research compared three distinct areas affect ed by the volcano and found different patterns related to infectious disease, respiratory illnesses, risk perceptions, and economic losses (Tobin and Whiteford 2002). By working on the CDMHA grant in various capacities between 2001 and 200 2 including field research, grant writing, and data collection and analysis, the researcher became very familiar with Penipe, the dissertation research study site, and had established some
3 contacts with in-country personnel, who woul d later be instrumental in facilitating entry to the research site. I decided that the city of Riobamba, c apital of Chimborazo Province, would be the most convenient place to live. The commute between Penipe and Riobamba was a mere forty-minute bus ride and cost forty cents. Most professionals working in Penipe also lived in Riobamba because Penipe has very few services. Riobamba is a small-siz ed city with approxim ately 125,000 people and has all the modern conv eniences (INEC 2001). The dissertation research took place between June and December of 2004, and a follow-up site visit was conducted in July 2005. Village life in Penipe followed a regul ar routine, with people working in the town from Monday until Fri day, with weekends free, exc ept in the health centers, which were open on Sundays, or market day. Many people would come to the town on Sundays from outlying rural areas to do weekly shopping or sell agricultural products, or would make the longer commute to Riobamba for its market day on Saturday. Brief Overview of Theory and Methods Estimates for worldwide child deaths due to acute respiratory infections (ARIs) are approximately two million annually, making it the leading cause of child mortality (Williams, et al. 2002). Rapid treatment of pneumonia with antibiotics is necessary for child survival in many of these cases (Mull, et al. 1994). Consequently, examining how ant ibiotics are dispensed and used for
4 respiratory infections is one component of an ethnographic study of ARIs. Identifying obstacles to health care-seeking is another vital component in any study, because when treatment is delayed, there are dire consequences for seriously ill children. While ther e have been many ethnomedical studies examining how knowledge of illness affect s health care-seeking behavior, some have argued that the links ar e tenuous (Rubel and Hass 1990:125). In the case of Penipe and childhood ARIs, many health professionals maintain the belief that mothers understanding of the biomedical model will improve illness outcomes. Therapeutic choices are a complex function of not only the caregivers beliefs but of people in her social network. Genera lly, health care-seeking behavior is not a black and white matter of choice bet ween home-treatment and a doctors attention, but a mix of strat egies for one illness episode. Expensive medical interventions are not necessary to solve the worldwide problem of childhood mortalit y from ARIs. There are many low-cost educational strategies and medical interventions such as improving vaccination coverage that can significantly improve health in rural communities in the Andes. However, before choosing which intervention will hav e the desired effect, a study of the local situation and extent of the problem needs to be ca rried out in order to tailor a program to a specific locality. This dissertation research should be considered as a descriptive case study, and the conclusion of the dissertation makes recommendations on how to improve respirat ory health for the target population. The dissertation research approaches the problem of childhood ARIs in a rural community in a developing country from a critical biocultural approach, with
5 a conception of the environment as a changing Andean landscape, imbued with cultural dimensions, and actors who negotiate relationships and health in an environment of medical pluralism (G oodman and Leatherman 1998; Greenway 1998b). This research perspective is in accord with current research in medical anthropology of the Andes and social epid emiology (Koss-Chio ino, et al. 2003; Pillai, et al. 2003). While many studies of medical pluralism in the Andes have been conducted in indigenous communiti es, the influence of indigenous ethnomedicine on the central highland provin ces of mestizo communities should not be underestimated (Crandon-Malam ud 1991; Finerman and Sackett 2003; Greenway 1998a; McKee 2003). thnomedicine is defined simply as, culturally oriented studies of illness (Rubel and Hass 1990:118). Approximately fifty percent of the population of Chimborazo Province is composed of indigenous people, and ethnomedical conceptions of health and illness, as well as diagnoses and home-based treatments, cross et hnic boundaries (Gerlach 2003). This research differs from some fe minist anthropological studies of Andean women because the research is not focused on folk illnesses that women suffer, who sometimes seek tr eatment with alternative healers, but instead examines childhood respiratory ill nesses which are either treated with home remedies or with biomedicine (Greenway 1998a; Larme 1998). The language of power and praxis that has been employed in some critical phenomenological studies of medical anthropologic al problems is not invoked in this study because the researcher em ploys a model based on the critical biocultural approach in anthropology that ex amines ARIs as the interaction of
6 political and economic forces on the experience of health and illness, not interpreting the illness as a mask for a mo re hidden, insidious malady, as in the case of Scheper-Hughes ethnomedical ex planation for infanticide in the favelas of Brazil (Leatherman 1998b; Schepe r-Hughes 1990; Scheper-Hughes and Lock 1987). In the theoretical section of the dissertation, t he synergism between political ecology and a critical biocul tural approach is proposed as a more appropriate perspective for situating acto rs in the local research context of childhood ARIs. In the process of c onducting an ethnomedical investigation of ARIs, the researcher explores the cultur al significance of an illness, and discerns the salient features of tr eatment and resolution from an ethnomedical perspective (Rubel and Hass 1990). This research elicits competing and overlapping m edical systems from mother respondents and biomedical health practitioners and gauges the variables [e.g. socioeconomic status (SES), education, num ber of children, crowding, environmental pollu tants, and nutrition] that af fect the extent to which female caregivers recognize biomedica lly defined symptoms for childhood ARIs. In addition, perceptions of health care decisions for treatment are discussed. This information is gathered through the combination of qualitative and quantitative ethnographic method techniques including focus group discussions, key informant interviews, in-depth in terviews with mother informants and healthcare practitioners, and semi -structured questionnaires. On the one hand, the research is fo cused on the biomedically-defined disease of ARI, which is distinguished from the illnesses of resp iratory infections,
7 which are culturally-defined feelings and perceptions. This is a common definitional distinction in medical anthropology (Pelto and Pelto 1990; Rubel and Hass 1990). The biomedical classification of ARIs has been defined by the World Health Organization (WHO) and is given to practicing doctors and nurses in Ecuador in the form of treatment guidelines. On the other hand, the local people use a plethora of term s to describe various resp iratory illnesses, and do not use the Spanish biomedi cal term for the disease, IRAs (ARIs). Medical anthropology studies of ARIs frequently us e methods such as free listing in order to ascertain the local terminol ogy for illnesses in order to determine whether local people have accurate recognition of various illnesses and act appropriately. This research study fo llows this tradition to discover the determinants for disease recognition in a rural population in Penipe County, Chimborazo Province, Ecuador. Research Hypotheses First Research Hypotheses 1) Female caregivers of higher so cioeconomic status with under-fives who display signs of severe respirator y distress are more likely to seek allopathic medical care. 2) Female caregivers of lower socioeconomic status with under-fives who display signs of severe respirator y distress are more likely to seek alternative medical care or home-based care.
8 Second Research Hypothesis Biomedical healthcare practitioners ma intain a perception that female caregivers delay seeking treatment for their under-fives who display biomedically defined symptoms for ARIs because of an inadequate understanding of the bi omedical model, meaning lack of education. These research hypotheses were determined a priori to performing field research. The first research hypothesis was tested using the results of the semistructured questionnaire. The hypothes is was based on the assumption that families of lower socioeconomic status would be more likely to treat many illnesses in the home or us e low-cost non-traditional healers because of the expenses of biomedical care, which incl uded transportation to health centers and the costs of medicines. Conversely, it wa s also assumed that families of higher socioeconomic status would have the monet ary resources to pay for either public or private medical care for their child ren and would have more spare time in order to seek timely m edical care for under-fives. The second research hypothesis was te sted using the information from indepth interviews with informants from the health care profession. These hypotheses guided the research agenda and were always in the researchers mind when carrying out the multiple resear ch methods. One of these questions, which is addressed in the second research hypothesis, aims to identify the constraints or barriers to timely health care-seeking. Some of the posited constraints were lack of economic capita l, limited access to health centers, and
9 lack of social support. The primary c onstraint was economic, however access issues were also identified, in the form of transportation barriers and limited hours of the medical facilities. Lack of social support was not mentioned as a barrier to health care-seeking. Statement of Findings The first hypothesis of the dissertation research stated that families of low SES would be more likely to treat respirator y infections at home or do nothing. Results from the semi-structured questi onnaires found that families with belowaverage salaries (under $150 per month) were more lik ely than those with aboveaverage salaries (over $150 per month) to treat a child with cough, fever and nasal secretion at home or do nothing ( (1) = 6.40, p < 0. 05, two-tailed). Moreover, families who lived in homes with dirt floors were more likely than those with non-dirt floors to have ch ildren with coughs and colds ( (1) = 4.22, p < 0.05, two-tailed). In addition, familie s with a low possession score were more likely than those with a high possession score to have a child with wheeze ( (1) = 7.35, p < 0.01, twotailed). Results from the logi stic regression procedures found that for the symptom cluster of cough, fever and nasal secretion health care-seeking, families with above-average salaries were five times as likely as those with below-average salaries to s eek a doctor for their childs symptoms when controlling for the other three i ndependent variables [ odds ratio (OR) = 5.97, p = .05]; moreover families with nondirt floors were 3.5 times more likely than those with dirt floors to seek a doc tor than treat at hom e for children with
10 these symptoms (OR = 3.51, p = .05). Therefore, th ere was a statistically significant association between increased incidence of respiratory problems and inhibited the ability to seek appropriat e care with poor housing, few possessions, and low income. In sum, household income and possession score were associated with differences in care-seeki ng and respiratory illness frequencies. For the second hypothesis, there was s upport from the in-depth interviews that doctors and nurses believed that their patients were unaware of the seriousness of ARIs and therefore delay ed care-seeking. Out of the seven doctors and nurses who responded to t he question on whether mothers delayed health care-seeking, five reported that mo thers delayed bringing their children to seek care because of lack of educat ion on the importance of seeking a biomedical care for certain illnesses. Although the health professionals also listed other reasons, e.g. tending animals and treating at home or with a natural healer, to explain delay in biomedical hea lth care-seeking, they stressed lack of education as a key determinant. When mothers were asked to describe what they believed to be childhood respiratory symptoms, they mentioned chest whistling, tightening of the chest, green nasal discharge, stuffy nose, cough, colds, and wheezing, all of which could be possible symptoms for ARIs as defined in the biomedical model. Some of the mothers other explanations for the cause of ARIs, such as change in climate, lack of vitamins, and ash rain from the volcano, were also plausible explanations from a biomedical standpoin t for the onset of ARIs. The main discrepancy between the biomedical and et hnomedical models was the mothers
11 use of various home remedies for an ext ended time, over the two-day limit as recommended by health professionals, to cure respiratory infections that would not resolve naturally. In these cases, the children likely had bacterial infections and should have consulted with a doctor. Therefore, there was considerable overlap between the biomedica l and ethnomedical models for the identification of ARIs; however, the mothers were inconsiste nt about the specific signs that they would recognize to prompt them to seek a medical professional, because of their overall lack of knowledge of the mo st recognizable ARI symptoms, rapid breathing and indrawn chest. This lack of knowledge was not specific to any socioeconomic group but was more visibl e in mothers who had not had children with severe respiratory infections in the past. Those few mothers who had had children with pneumonia were more cogniza nt of the seriousness of acute lower respiratory infections. Structure of Dissertation Chapter Two begins the dissertation with an introduction to the geography, demography, and recent politics and hi story of Ecuador. Chapter Three introduces the research setting, descri bes prior research, and presents the rationale for site selection. Next, in Chapter Four, in t he literature review, there is a discussion of theory in contemporary medical anthropology. This chapter presents the theoretical background of t he dissertation, which is framed by a political ecology orientation informed by a biocultural perspective. The second section of the chapter is a critical treat ment of ethnomedical studies in the Andes
12 and how the present research relates to t hese previous scholarly works. This dissertation takes a political economy approach and examines differences in socioeconomic status as a key vari able determining access to health care resources. The third section of the c hapter introduces methodologies used in the cross-fertilization of studies of medi cal anthropology and epidemiology. The argument is to undertake an ethnographic st udy before starting epidemiological research in order to identify the cultural fa ctors of illness. In the fourth section of the chapter, the researcher grounds the dissertation in a discussion of the epidemiology of ARIs in the developing worl d. This section begins with morbidity and mortality trends, etiology and modifiabl e and non-modifiable risk factors. The second half discusses case managemen t, meta-analysis of case studies, antibiotic management, prevention, and et hnographic studies of ARI. Chapter Five describes the multiple methods used to gather data for the ethnographic study. The study used a combination of focus group discussions, key informant interviews, in-depth interviews, and semi -structured questionnaires to collect ethnographic data. Chapter Six covers t he results of the instruments used in data collection. In Chapter Seven, t he researcher tests the hypothesis and provides an analysis and discussion of the results. The dissertation concludes with Chapter Eight and offers final re commendations to improve the health situation of under-fives in Penipe County.
13 CHAPTER TWO: ECUADOR GEOGRAPH Y, DEMOGRAPHY, ECONOMY, AND CHILD HEALTH Ecuador is a country that has held a fascination for anthropologists, biologists, geographers, and historians becau se of its unique history, stunning landscapes, remarkable biodiversity, ri ch cultural traditions, and diverse indigenous peoples. This chapter introduces t he reader to recent political history, population characteristics and health ca re system in Ecuador, and addresses some of the major issues that face Ecuador today. Particular historical events in Ecuador, like the 1990 Levantamiento Indigena (Indigenous Uprising), led to a pl ethora of scholarly writings about identity politics, self-determination, and ethnic nationalism amongst Ecuadorian indigenous peoples (Brown 1993; Hendri cks 1991; Martin 2003; Pallares 2002; Selverston-Scher 2001; Whitten 1996; Zamosc 1994). Other more gradual events, such as Ecuadors Agrarian Re form, that began to emerge in the late 1960s and 1970s, were the subject of numerous studies in anthropology, political science, history, and geogr aphy (Chiriboga 1988; Hirao ka and Yamamoto 1980; Korovkin 1997; Rudel 1995; Uquillas 1984; Zevallos 1989). Jungle border conflicts between Peru and Ecuador that began after World War II, erupted again in 1980 and also in 1995, prompted schol arly research in the area of peace
14 studies/conflict resolution (Herz and N ogueira 2002). While a peace agreement was signed, the tri-border area of Ecuador, Colombia, and Peru is one characterized by instability due to the mu ltinational trade in drug trafficking and counterfeiting. Moreover, because Ecuador us es the U.S. dollar, it is attractive place for cash-placement and money-launder ing operations for drug traffickers, especially along the northern border with Colombia (CIA 2005). Third World debt has gradually incr eased since the 1970s, and solutions such as debt-for-nature swaps were a ttempted in Ecuador with the help of environmental non-governmental organi zations (NGOs) and other countries because the forces of poverty were viewed as a threat to the environment (Meyer 1993; Phelp 1992; Visser and Mendoza 1993). Environmental and health crises were compounded by past illegal activities of multinational oil companies in the Amazon region of Ecuador (Sawyer 2004). Lawsuits were filed in the U.S. because of a corrupt judicial system in Ecuador, and epidemiological studies were conducted to assess cancer risks due to exposure to oil pollution in defense of indigenous peoples (Gerlach 2003; Hurtig and Sebastian 2002; Kimerling 1991). Finally, even though the 2000 dollarization of the economy curbed rising inflation, the higher price of oil has l ed to price hikes of basic necessities and consumer goods, producing economic hardshi ps for the majority While Ecuador is considered a constitutional republic, seven presidents have governed the country since 1996 (CIA 2005). The last president, Lucio Gutierrez was ousted by congress on April 20, 2005 and replaced by the then Vice-President, Alfredo Palacio, one year prior to the expiration of his term because of lack of political
15 support, corruption, and unpopular decisi on-making, among other reasons. Many of these various historical events involved various power brokers such as multinational corporations, indigenous federations, political parties, local, national, and foreign governments, the Ecuadorian military, and local and international NGOs. The preceding short list is merely a sample of the dramatic events that have occurred in recent Ecuadorian histor y. Compounding Ecuadors situation has been the ongoing internal problems of its neighbor to the north, Colombia, whose escalating drug-related violence in its major cities has led many of its citizens to emigrate to Ecuador, thus compounding Ecuadors problems of widespread poverty, lack of job opportunities, and gener ally weak government infrastructure. Geography Ecuador is a small country roughly t he size of Oregon and is divided into four major regions: the coas t, the sierra, the Oriente or eastern Amazon jungle, and the Galapagos Islands. Each area is environmentally distinct. Ecuador is divided into twenty-two provinces, which are in turn subdivided into cantons, or counties, which are also subdivided by parishes. Quito is the capital of the country, while the coastal port of Guay aquil is the largest city. Throughout Ecuadors post-colonial history, t here have been power struggles between Guayaquil, and the more conservative Andean capital of Quito. In contrast, the Oriente is sparsely populat ed; however, because of its rich natural resources,
16 namely oil, its international appeal fo r ecotourism, and its value for national security, sharing borders with Colombia and Peru, the Oriente is tied to the nations development plans and economic pr osperity. While Ecuador is a small country with a land area of 276,840 square k ilometers, it is noted for its Avenue of Volcanoes that runs through the Andes (CIA 2005). The highest peak in Ecuador is Chimborazo at 6,267 meters, to the north is Cotopaxi, the highest active volcano on the planet at 5,897 me ters, and to the south is Tungurahua, another active volcano, near the resear ch site, at 5,023 meters (CIA 2005; Whiteford and Tobin 2002).
17 Caar Manab Esmeraldas Pichincha Imbabura Carchi Sucumbios Orellana Napo Cotopaxi T ungurahua Pastaza Morona SantiagoC h i m b o r a z oB o l i v a rLos Rios Guayas Azuay El Oro LojaZ a m o r a C h i n c h i p e T ungurahua VolcanoQUITOP A C I F I C O C E A NCOLOMBIA PERU PERU 81o79o0o2o4o77o77o79o81o4o2o0o NEcuador: Provinces 0 50 50 km100 KilometersScale (approximate) CapitalLegendBase map: Defensa Civil Ecuador, undated GalpagosNot to scale Figure 1: Map of Ecuador: (Lane 2003) The research site of Penipe is loca ted in the northeastern part of the Province of Chimborazo, known as the pr ovince of high peaks, whose capital is Riobamba, nicknamed the sultan of t he Andes (d'Angelo 2001). Chimborazo covers an area of 5,637 square kilo meters, and has a population of 403,000 people (Ecuador Ministry of Tourism: 2005) The province of Chimborazo has some of the highest peaks in Ecuador including Chimborazo (6,310 m), Carihuairazo (5,020 m), and the paramos of Urbina, Cuvelln and Achupallas. In
18 the western part of the province towers the peak of Quilimas (4,919 m) and the snow-covered El Altar (5,320 m). The prov ince is traversed by two major rivers: one is the Chanchn in the southern part of the province, and the other is the Chambo River, that runs from south to north. In addition, Chimborazo has attractive lakes: Ozogoche, Atillo, and Colta. Finally, Sangay National Park is a large ecological reserve (517,765 ha.) and a UNESCO World Heritage site that extends into three provinces, borderi ng Tungurahua and Morona Santiago. The snow-capped Sangay Volcano is a destinat ion for adventurous trekkers. The province is characterized by a divers e climate, with a s ubtropical area in Pallatanga, passed by the traveler en r oute towards the coast. The average temperature of the province is 13 C (d'Angelo 2001). Cantons of Chimborazo Province Riobamba Guano Penipe Chambo Alaus Cumanda Pallatanga Guamote Colta Chunchi Riobamba Tungurahua Volcano Base Map: Ecuador On Line, undated N Legend Provincial capitalS c a l e ( a p p r o x i m a t e )3 6 k i l o m e t e r s 1 8 9 0 9 k m Figure 2: Map of Chimbor azo Province: Lane (2003)
19 Demography Ecuador has a population of 13, 363,593 (July 2005 estimate), characterized by a population growth rate of 1.24 percent (CIA 2005). In 1950, the population was only 3,202,757, and in 1974, the population was 6,521,710, half of what it is today (INEC 2001). Fifty-five perc ent of the population live in urban areas, 50 percent live in coastal areas, 45 percent live in mountainous regions, nearly 5 percent live in the Oriente, and the remaining less than one percent live in the Galapagos Islands (PAHO 2002). Most people live in the cities, with approximately three million in Guayaquil and two million in Quito (USAID 2004; Wibbelsman 2003). Interestingly, the third-largest city of Ecuadorians is New York, not Cuenca (W ibbelsman 2003). The ethnic makeup of the population is primarily mestizo (65%), of mixed Spanish and Indian descent. The remaining population is composed of indigenous peoples (25%), people of Spanish descent and others (7%) and people of African descent (3%) (Wibbelsman 2003). While Spanish is the official language of Ecuador, thirteen indigenous nationalit ies inhabit Ecuador, each with their own unique language and culture. The indigenous populati on numbers roughly 3,655,000. The Kichwa, formerly spelled as Quichua, ar e the most populous of these indigenous nationalities, with a popul ation of 3,000,000 in t he Andes and 70,000 in the Oriente (Sawyer 2004; Wibbelsman 2003). The northern provinces of Ecuador have witnessed a large influx of Colombi an refugees fleeing the political violence in their country, some 20,000 in 2004 (CIA 2005). Population trends for Chim borazo Province have paralleled those of the
20 country, with more people moving to ur ban areas. In 1950, 21 percent of Chimborazos population was rural compar ed to 26 percent in 1974, 33 percent in 1990, and 39 percent in 2001 (INEC 2001). To illustrate the more rapid growth of the cities of Quito and Guayaquil, located in Pichin cha and Guayas provinces, respectively, in 1950, Chimborazo compri sed 6.8 percent of the population of Ecuador, compared with 4.7 percent in 1974, 3.8 percent in 1990, and 3.3 percent in 2001 (INEC 2001). Therefor e, while the city of Riobamba has witnessed growth, there has been substantially more growth in the major cities of Quito, Guayaquil, and Cuenca. Of t he 403,632 people reported in 2001 for Chimborazo Province, 193,315 people live in Riobamba County, compared with 6,485 for Penipe County (INEC 2001). The city of Riobamba has grown from a population of 29,830 in 1950 to a populat ion of 124,807 in 2001. The annual growth rate of the city of Riobamba has remained relati vely stable, ranging from a high of 3.09 percent between the mi d-1970s and the mid-1980s, to a low of 2.53 percent for the decade of the 1990s (I NEC 2001). Half of the economically active inhabitants of Chimborazo Provin ce work in agriculture, with a lesser proportion working in services (15.6%), commerce (12.6%), and manufacturing (8.6%) (INEC 2001). For the entire provin ce of Chimborazo, 16.5 percent of the population is illiterate, co mpared to a country aver age of 10.8 percent, with a greater proportion of females (20. 7%) than males (11. 6%), and a higher proportion of illiteracy in rural areas (24.6%) than urban ar eas (4.9%) (INEC 2001).
21 Economy, Land, Rights, and Debt Ecuador has an export-driven economy focu sed on oil, coffee, cocoa, cut flowers, bananas, and shrimp, primarily with export partners United States (48.3%), Colombia (5.5%) and Germany (4.8%) (CIA 2005). Consequently, Ecuador has a fragile economy dependent on t he world market prices for these products. Petroleum continues to account for 40 percent of export earnings and consumes one-fourth of government spending (CIA 2005). Under the administration of Jamil Mahuad, while the October 1998 Peace Accords were signed with Peru, finally ending a centur y of conflict in 1999, the economic situation continued to worsen. Bec ause of a combination of economic and environmental factors includi ng falling oil prices, an inte rnational financial crisis, and a regional recession caused by t he weather phenomenon of El Nio, the Ecuadorian currency, the Sucre, took a dramatic downtur n, and the banking sector collapsed (Embassy_of_Ecuador 2002; Wibbelsman 2003). The 1997-98 weather pattern led to storms, floods and landslides which produced losses of $2.6 billion (Gerlach 2003). Because of the financial crisis, Ecuador was forced to default on its international debt obli gations (CIA 2005). In 2000, Ecuador dollarized its economy and Sucres were taken out of circulation. This quickly led to street protests and the ouster of President Mahuad. The subsequent administration of Gustavo Noboa sought to reach agreements with international creditors and sign agreements with inter national oil companies to increase foreign investment and to expand oil expl oration and the oil pipeline infrastructure (Embassy_of_Ecuador 2002). While Ecuadors economy continues to grow
22 because of the rise in petroleum prices, t he country continues to be vulnerable to price swings and international economic crises (CIA 2005). One way to trace recent political and economic history of Ecuador is to examine the process of agrarian reform The agrarian reform laws of Ecuador are based on a policy of integration (Davis and Wali 1993). Although the ideology of the initial agrarian refo rm was more influenced by theories of development concerning sparsely populat ed lands, the primar y beneficiaries of the reform were former hacienda tenant s (Crain 1990). The aim behind the agrarian reform was to integrate indi genous people into t he national economy and to colonize the frontier with the ov erflow populations of urban areas. The method for carrying out this process was to organize indigenous people into cooperative forms of organization with t he promise of land rewards in exchange for compliance. The Law of Agrarian Reform and Colo nization (1964), was passed in an effort to increase national food producti on in response to the United States' program for Latin America, the Alliance fo r Progress. The effect of the change was to eliminate the traditional debt -peonage system of the haciendas and make peasants legal owners of their land plots (Zamosc 1994). In order to accomplish this task, the newly created Ecuadori an Institute of Agrarian Reform and Land Settlement (IERAC) would be in charge of land redistribution. Since the agrarian reform, colonization was the most common method of land redistribution. The underlying political objectives of these agr arian reform laws included lessening demographic pressures in hi ghland urban areas and assim ilating lowland Indian
23 populations into the nationalist program (Macdonald 1981). The objective of the law was to stimulate the agricultural se ctor so that "unused" lands could be redistributed. In the 1970s, the second phase of the reform was more instrumental at transferring former hacienda lands to peasants from surrounding communities (Zamosc 1994). The Law of Coloniza tion of the Amazon Region (1977) facilitated colonization of t he four Amazon provinces of Ecuador that would act as a safety valve for the densely populat ed urban areas. The 1977 law created INCRAE (National Institute of Coloniza tion of the Ecuadorian Amazon Region), and this body was to oversee the quasi-mili tary colonization of the Oriente to strengthen the Ecuadorian pres ence along the southern border in response to the potential military thr eat from Peru (Vickers 1988). The 1994 Agrarian Development Law was an attempt to apply principles of a free market economy approach in response to pressures to inst itute neoliberal reforms by the World Bank and International Monet ary Fund. The law succee ded in eroding the land tenure stability of communal land holdi ngs and abetted corporate interests in securing private property to stimulate the raw material s export sector (Sawyer 2004). Parallel with the process of agrarian reform in the 1970s was the explosion of the Ecuadorian economy from new oil revenues. This allowed the Ecuadorian government, which at the time was a military regime, to engage in unprecedented development schemes. While the boom was primarily a boost to Ecuadors economy in the 1970s, even in 2001, oil revenues constituted almost
24 half of the countrys revenue (Gerlach 2003). While oil has driven the countrys economy, there have been disastrous environmental consequences, with frequent oil spills in the Amazon along the pipel ine, as well as pollution from oil production wastes in fragile ecological areas (Kimerling 1991; Sawyer 2004). In 1979, a radical change in government took place with the election of Jaime Roldos and Oswaldo Hurtado, the fi rst democratically elected government without the support of the oligarchy or the military. President Roldos died in a plane crash halfway into his tenure and Hurtado became his successor. Hurtado was a former professor of political science and sociology at the Catholic University of Quito, and his most infl uential work, Political Power in Ecuador was a critique of the workings of the Ecuador ian political system and glaring social inequalities (Hurtado 1980). Hurtado c oncludes his book with a critique of Ecuadors dependence on oil, concluding t hat Ecuadors development strategy leads not only to inequities in income distribution, benefiting Quito over Guayaquil, but to a larger dependence on fo reign markets. This is indeed what occurred as oil prices plummeted in the ensuing decade. At one point in the 1980s, Ecuadors foreign debt reac hed $431 billion, but the debt was continuously restructured in order to be eligible for more loans (Gerlach 2003:46). During the 1970s and 1980s, high-i nterest loans caused these debts to grow, and consequently, the debts were continuously restructured and renegotiated. One solution to the foreign debt problem involved debt-for-nature swaps (DNS) involving international financ ial institutions and environmental non-
25 governmental organizations (NGOs). T here has been a vast proliferation of NGOs in Latin America generally and in Ecuador more specifically. As the Ecuadorian governments environmental budget was small, local NGOs expanded and assumed roles that t he state agencies could not manage (Bebbington 1996; Meyer 1993). In Oct ober 1987 a DNS deal was brokered by Fundacin Natura (FN), the largest priv ate environmental NGO in Ecuador and the World Wildlife Foundation (WWF). FN received the governments permission to convert $10 million of commercial bank debt into local government bonds. WWF acquired the first $1 million debt for $3 54,000 and placed it at the disposal of FN for the maintenance of Ecuador s national parks (Phelp 1992). The government issued long-term bonds, the intere st from which funded FN activities in respect to the management of national parks. The $10 million issued in bonds was a very small percentage of the countrys total exte rnal debt. Nevertheless, while Ecuadors debt was only reduced by $1 million, the resulting conservation fund yielded revenue equal to twice the existing government budget for parks and reserves. The second stage of the swap involved $9 million in 1989 purchased for $1,068,750 (Visser and Mendoza 1993). WWF purchased $5.4 million and The Nature Conservancy (TNC) purchased $3. 6 million, including $400,000 from the Missouri Botanical Gardens for botanical re search projects in protected areas of Ecuador (FAO 1993). The interest fr om the bonds f unded educational and training programs in the Galapagos Isl ands and seven parks and reserves in the Andes and the Amazon. Some funds we re designated to change the status of
26 some parks from poorly managed paper par ks into conservation areas. Once the bonds matured, the prin cipal would become an endowment for FN activities. While this DNS transaction appeared positive there were questions about FNs commitment to environmental protecti on. Local NGOs are vulnerable to cooptation by national governments or multin ationals. FN received funding from a U.S.-owned oil company in exchange for ex ploration rights within Yasuni National Park, home to the Huaorani Indians (Phelp 1992). FN also allowed half of the park to be an open access area for mineral ex ploitation, and this led to a divide among Ecuadorian envir onmental NGOs. Another criti que of FN was that it was elitist and ill suited to work with local comm unities. FNs strengths were more in public relations to spread the environment al protection message (Meyer 1993). Indigenous peoples voices were exclude d in the DNS negotiations. Currently, Ecuadors external debt currently stands at $16.8 billion ( 2004 estimate), and President Palacio has shown resistance to implementing structural adjustment policies imposed by the World Bank and International Monetary Fund at the expense of the health and social servic es for Ecuadors citizens (CIA 2005). Future innovative means for reducing Ecuadors foreign debts are likely. Indigenous participation in powerful political blocs involved in decisionmaking in Ecuador evolved in response to t he threats of agrarian laws to their traditional lands and to the perceived ill egal activities of international oil companies in the 1960s and 1970 s. Part of the strategy to persuade the national government to hear the voices of indi genous peoples was a revalorization of Indian ethnic identity t hat culminated in the Levantamiento Indigena of 1990.
27 First, during the 1970s and 1980s, agencies of rural development stepped up their efforts to integrate marginal Indi an communities into the national political infrastructure. Second, voting rights and more educational opportunities were extended to Indians. Third, the promises of the agrar ian reform had fallen short of expectations and the economic situation of Indians had declined drastically. Before the uprising, mestizo and Indian peasants experienced the pressures of inflation, lack of credit, usurious inte rest rates, and limited assistance from the government (Zamosc 1994). F ourth, the pervasive racism in Ecuadorian national society and the pejorative connotations of the term indio produced a reaction, and Ecuadorian Indians (primarily speaking Kichwa) began to use indio to refer to themselves (Whitten 1985). A revalo rization of Indian identity under the banner of Indianism painted the Kichwa Indi ans as returning to the Inca ideal, envisioned as a philosophy of egalitar ianism and communalism. Fifth, the formation of CONAIE in 1986, the nat ional organization representing all indigenous groups in Ecuador, concentra ted on pressuring the national government to recognize Ecuador as a pl urinational state, achieved under the administration of President Rodrigo Borja, and was successful in administering the program for intercultural bilingual educ ation in all Indian parts of the country (Zamosc 1994). Working alongside CONA IE were new, powerful Indian leaders who were able to negotiate territori al autonomy for Amazonian groups and respond appropriately to pressure fr om regional delegates for a massive mobilization (Collins 2000; Zamosc 1994). The confluence of these factors coupled with the international attention of environmental NGOs led to a strategy
28 of advancing the rhetoric of Indian nat ionalities. The same leaders who had identified themselves as campesinos (peasants) in the 1970s were involved in the national uprising in 1990 demanding Indian rights to land, technical assistance, and cultural autonomy (Schroeder 1991). This is an example of how tribal identities could be created thr ough contact with the colonial state and converted into a weapon of resistanc e against dominant national sectors (Whitehead and Ferguson 1992). Because of the power vacuum created by the agrarian reform, referring to the weak ened power of the hacienda owners, Indian leaders were poised to broker co operation between rural people and development agencies (Zamosc 1994). Crain (1990) describes how in a re cent period of Ecuador ian history from 1960-1984, positive symbols of Indianness and the rhetoric of a plurinational state was used by the nati onal government in a hegemonic process to construct a national identity. While Indian leaders have partici pated in this process, counterhegemonic discourses disseminated by nascent Indian political parties such as Pachakutik-Nuevo Pais challenge the nationali st ideology. More frequently, indigenous leaders ha ve been elected to political offices in Ecuador, and formed powerful political blocs and coalitions to change the course of political history in Ecuador. Moreov er, indigenous movements have aided in developing public social sector policie s and creating councils to monitor the childrens rights and government fisca l responsibility (UNICEF 2003). The next section turns the readers a ttention to the main subject of the dissertation, that is, child health. Through the presentation of health statistics for
29 the country as a whole, and for the provin ce of the study site, the research is situated in the local context. Child Health in Ecuador Many of Ecuadors health problems are tied to poverty and marginalization (Whiteford and Tobin 2002). One recent estimate places 45 percent of Ecuadors population below t he poverty line (CIA 2005). Of this percentage, roughly half are classified as indigent, and 40 percent of the population has at least one unmet daily need regarding hous ing, food, health services, and education (PAHO 2002; Wibbelsman 2003). The recent population census estimates based on unsatisfied basic needs suggested by the Andean Community places 61 percent of the population in poverty; however, this estimate ignores ethnic and racial diffe rences, since a higher proportion of indigenous and black people live in pov erty than the general population (INEC 2001). Rural and indigenous areas are more severely affected by poverty. Nearly 70 percent of Ecuadors children cu rrently live in poverty, and 15 percent of under-fives are affected by malnutri tion (UNICEF 2003). Pneumonia, one of the diseases of poverty, is the primary cause of death in Ecuador with a rate of 27.2 per 100,000 people, followed by cerebrovascular diseases and cardiovascular diseases. Recently, there have been gains by government and society in social public investmen t, and child immunization coverage has increased from 70 percent in 1999 to almo st 95 percent in 2002 (UNICEF 2003). For children under one year old, t he country has achieved 90 percent
30 immunization coverage for diptheria, tetanus, and pertussis (PAHO 2002). Moreover, the last report ed case of polio occurred in 1990, and since 1998, there have been no reported cases of measles (PAHO 2002). There exist ethnic and racial health disparity challenges in Ecuador. Ninety percent of indigenous and black people live in poverty and experience higher mortality rates than other groups and only 39 percent complete primary school, with many forced to work at a young age (UNICEF 2003). In rural mountain and Amazon regions approximately 80 percent of indigenous children live in poverty, and along the coast, 70 perc ent of black children live in poverty (PAHO 2002). One of the r easons for these statistics is that oftentimes heads of households are forced to emigrate to urban areas in search of job opportunities. The health sector in Ecuador is a combination of public and private institutions, nonprofit and for-profit. T he researcher witnessed a speech by the Ecuadorian Public Health Minister, Dr. Teof ilo Lama, at the main public hospital in Riobamba, where he was responding to the issue of how government clinics should respond to natural disasters. In his view, it was a question of autogestin or developing from within, in order to solve the health problems caused by poverty and maternal mortalit y. The view of many Latin American government health ministries is that decentralization of health services serves to improve quality of care and responsiveness to local concerns and problems (LLoydSherlock 2000). However, what often happens is that resources are disbursed unevenly to urban areas at t he expense of more rural ones, contributing to health disparities. After the Minister spoke, the researcher heard rumblings from local
31 doctors and officials that the national government had still not done enough to assist with recovery efforts in certai n areas of Chimborazo that had suffered health and economic losses due to mudslides and volcanic ash. The Ministry of Public Health, Social Security Institut e (IESS), the Public Health Service of the Armed Forces and Police, the National Child and Family Institute, and the Ministry of Social Welf are comprise the public subsector, which serves roughly 59 percent of the population (PAHO 2002). The Ministry of Public Health serves the most people, roughly 31 percent of the population. Those who cannot afford either the public or privat e healthcare providers turn to traditional medicine. PAHO estimates that there are 13.3 physicians and 4.6 nurses for every 10,000 patients (PAHO 2002). Cla ss distinctions and preoccupations constantly come into play in daily so cial and business relationships. Having personal influence, the right contacts, and the right relatives propels the status seeker to forge alliances and improve one s social standing (Wibbelsman 2003). Access to property, quality health servic es, and loans are contingent on the ability of individuals to negotiate the so cial milieu that is Ecuadorian society. Key Child Health Indicators for Ecuadors Sierra This section presents key child heal th indicators for Ecuadors Andean region. The most recent figures from 2003 indicate an infant mortality rate of 24.0 per 1,000 live births for Ecuador ov erall; moreover, this rate is a considerable improvement from 1990 (42.8 per 1000) and 1980 (95.0 per 1000), and rates are expected to fall to 14.3 per 1000 births by 2020 (USAID 2004). In
32 under-fives, ARI and acute diarrheal diseas e accounted for 60 to 70 percent of hospitalizations, and ARI was responsible for 37 percent of deaths in infants and 32 percent of deaths in children one to four years old (PAHO 2002). Table 1 contains information on key ch ild health indicators for Ecuadors 10 Sierra provinces. The table contai ns average percentages for the country of Ecuador and the Sierra as a whole. Next, the percentages for prevalence of diarrhea, respiratory infections, vaccinat ion coverage, and chronic undernutrition for under-fives are presented. The numbers for Chimborazo Province are highlighted for comparison purposes with the averages for the Sierra and the country as a whole. Chimborazo Provin ce has lower prevalence of diarrhea and respiratory infection than the Sierra and country averages, but ranks worse for the indicators of complete vaccination scheme and chronic undernutrition. From this preliminary data collected by CEPAR (2004), it appears that despite performing worse in the areas of va ccination coverage and providing adequate nutrition, the province as a whole ranks second best for Sierra provinces in low prevalence of childhood diarrhea and thir d best in low prevalence of childhood respiratory infections.
33 Table 1: Key Child Health Indicators for Ecuadors Sierra Provinces (%) Health Indicators Ecuador Sierra Carchi Prevalence of diarrhea21.7 22.5 18.2 Prevalence of respiratory infections42.3 35.3 40.9 Complete vaccination scheme: Children 12-59 mo.67.0 68.1 87.8 Chronic undernutrition in under-fives23.2 34.5 28.7 Health Indicators ImbaburaPichinchaCotopaxi Prevalence of diarrhea 12.221.425.7 Prevalence of respiratory infections 21.036.535.1 Complete vaccination scheme: Children 12-59 mo. 64.770.456.4 Chronic undernutrition in under-fives 33.727.933.8 Health Indicators BolivarTungurahuaChimborazo Prevalence of diarrhea 28.322.7 17.6 Prevalence of respiratory infections 42.734.5 32.7 Complete vaccination scheme: Children 12-59 mo. 56.877.0 54.6 Chronic undernutrition in under-fives 39.941.0 39.8 Health Indicators CaarAzuayLoja Prevalence of diarrhea 22.529.226.7 Prevalence of respiratory infections 29.837.841.6 Complete vaccination scheme: Children 12-59 mo. 76.468.668.2 Chronic undernutrition in under-fives 39.831.328.9 Source: (CEPAR 2005) In Table 2, selected health characte ristics of child he alth were collected from the Endemain study in 2004. For breastfeeding practice, the groups with the highest percentages included rural areas (63%), Sierra (64%), women between 35-49 years old ( 58%), women lacking educat ion (68%), and indigenous women (80%). This is compared with the nat ional breastfeeding average of 49 percent. The national prevalence of diarrhea was 22 percent, with 24 percent using oral rehydration salts. Mo thers under 25 years old had the highest percentage of children with diarrhea (24%) of the thr ee age groups represented. The greatest prevalence of ARI was found in the urban areas (46%) and the coast (51%). Mothers le ss than 25 years old had more children with ARIs (44%) than the other two age groups and a lower per centage for antibiotic use (34%).
34 This suggests that young mothers are a possible risk group for having children with untreated ARIs. Less education is possibly associated with lower use of antibiotics to treat ARIs, with 26 perc ent of mothers wit h no education using antibiotics. Antibiotic usage is positi vely associated with education level. According to PAHO (2002) between 1990 and 2003, there was a national increase of 40 percent in the number of ac ute diarrheal cases and an increase of 84 percent in acute respiratory infecti on cases. Although not represented on the table, in Chimborazo, the prevalence of low weight-for-age in children between one and four years old was 40 percent, co mpared to the national average of 26 percent (PAHO 2002).
35 Table 2: Health of the Child, Accord ing to Selected Characteristics (%) of childReceivedFirstExclusively at birth (a)chesthour0 3 monthsTOTAL80.896.726.448.721.724.142.337.3 AREA Urban92.095.622.432.720.724.745.740.0 Rural67.498.131.763.223.023.538.333.3 REGION Coast86.095.820.432.220.624.551.342.5 Sierra76.197.431.764.222.824.435.730.4 Amazon73.798.534.244.722.719.020.223.9 Insular96.694.154.933.36.725.013.412.5 AGE Less than 2581.197.526.344.024.321.344.434.1 25 3482.296.726.752.718.929.842.043.3 35 4975.394.025.857.520.321.535.031.7 EDUCATION None48.298.628.268.220.540.429.326.4 Primary69.896.731.055.024.021.244.030.9 Secondary92.596.723.940.420.326.643.442.7 Postsecondary97.596.318.318.104.22.1685.949.0 GROUP Indigenous51.198.831.379.625.027.435.021.4 Mestizo84.596.726.543.721.424.142.038.3 Other83.494.921.143.321.120.052.242.1 SELECTED CHARACTERISTICSPrevalencePrevalence A ntibiotics ORT (c) BREASTFEEDING ( b ) DIARRHEA ( a ) ARI ( a ) Wei g ht (a) born between 1999 2004; (b) born in 1999; (c) oral rehydration salts therapy Source: (CEPAR 2005) The aim of this chapter has been to introduce the geography, demography, politics, history, and status of child health in Ecuador. While brief, the major recent developments in Ecuador have been placed in a context that informs the larger dissertation project. Ecuadorians are invested in their political and social institutions, and they follow curr ent events closely and critically. To study health care-seeking behavior in a sm all town in Ecuador, the researcher should not lose sight of t he larger political and soci al context wherein these decisions are made.
36 CHAPTER THREE: RESEARCH SETTING, PRIOR RESEARCH, AND SITE SELECTION Research Setting Penipe is located in the northeastern section of the Chimborazo Province, lying 22 km from Riobamba, the capital of the province. To the north lies the Tungurahua Province and to t he south are the Parish of Quimiag and Riobamba County. Penipe County lies at an av erage altitude of 2460 meters above sea level with an average temperature of 14 Celsius. The county is divided into seven parishes: Penipe (County Seat), El Altar, Bayushig, Puela, Matus, Candelaria, and Bilbao. Penipe County has two valleys, Penipe and El Altar, and two mesas, Matus and Bayushig. The c ounty is carved up by various rivers, including the Chambo, Blanco, and Puela, and has various lagoons such as the Negra, Minsas, Yaguarcocha, Ventanas Collantes, and Amarilla, all part of Sangay National Park. From national census data, the total population of the county is 6,485 (710 living in the town of Penipe, 5775 livi ng in rural areas) with 1,783 occupied dwellings (Table 3) (INEC 2001). Ther e is a population densit y of 17.5 persons per square kilometer, the lowest for a ll counties in Chimborazo Province (INEC 2001). After the October 1999 Mount Tungu rahua eruption, roughly half of the population of the county was evacuat ed, and an unknown number of these
37 returned (Whiteford and Tobin 2002). Bec ause of the negative effects of the ash on animals and crops, and the lack of economic opportunities, many young workers emigrated to other regions of Ecuador (Lane 2003; Tobin and Whiteford 2002). The breakdown on ethnicity from the same population diagnostic is 90 percent mestizo, and 10 percent indi genous. From the researchers observations, there is a discrepancy in th is data since the people of Penipe did not report any indigenous people living in the county, and there was no visible presence of indigenous people during the fieldwork. Table 3: Population of Penipe County (INEC 2001) Total Male Female Total 6485 3226 3259 Urban 710 325 385 Rural 5775 2901 2874 The primary economic activities in P enipe County are agriculture, mainly of corn and potatoes, and cattle raising, although these businesses have suffered since the volcanic eruptions. Other agr icultural products include apples, pears, claudias and capulies. The second primary economic activity is the shoe factory, which employs a large number of disabl ed individuals in the community and is funded by a non-profit Catholic charity. Penipe is a small, rural town and the seat of the county with one health sub center (S CS Penipe), a small soccer stadium, a main plaza with a church and convent, f our restaurants, and no hotels. Some travelers pass through Penipe on their way to El Altar, (at 5,319 m., it is Ecuadors fifth highest summit) and fo r views of Mount Tungurahua (5,029 m.)
38 (Rachowiecki 1992). Local food specialtie s include corn tortillas cooked on hot stones, potatoes with guinea pig, roasted pi g, and apple wine (Los_Andes 2004). Figure 3: Town of Penipe The area which is today Penipe was or iginally inhabited by a tribe called the Pinipis, which was a subgroup of t he Puruhaes, and which meant river of the snakes. The Pinipis cultivated beans, corn, potatoes, and cauliflower to provide themselves with sustenance and to trade with neighboring tribes for salt. After the Spanish conquest, most of the Pinipis fled to the jungles of the Oriente. Penipe was founded in 1563 by Don Loren zo de Cepeda and was named after the mountain Cedral of Peni pe. The patron saint of Penipe is Saint Francis of Assisi. In 1845, Penipe became part of the Guano County as a rural parish and
39 church center. Penipe County was founded in February 1984 under President Oswaldo Hurtado (COSUDE 2003). Figure 4: Map of Penipe County Source: (COSUDE 2003) CEBYCAM-CES This section introduces the Center for the Eradication of Goiter and Debilitating Diseases (CEBYCAM), which wa s created in 1983 as an initiative by the Ecclesiastical Communities in order to address severe problems of health and disease, unemployment, and educati on in Penipe (Ponce 2001). The example of CEBYCAM exemplifies a su ccessful development model for rural communities in Ecuador, and serves as an introduction to some of the problems which the people in Penipe faced in t he past and continue to face today. In
40 addition, the researcher achieved acc eptance from the community for his own research in the county through his a ssociation with CEBYC AM, so a thorough explanation of the centers activities is appropriate. A leader of the community, Father Jaim e Alvarez, arrived in Penipe in 1979 and worked together with re searchers studying the relationship between poverty and illnesses such as cretinism and goiter, beginning their work in 1981. At the time there were more than 700 cases of people with goiter, twice the number of the national average, and the c ause was attributed to lack of iodine intake, which is normally consumed in sa lt. However, goiter was not the only regional health problem, there were also many cases of deafness, language disorders, as well as mental retardation. Other problem s in Penipe included lack of potable water for half of the inhabitant s, poor sanitation, and lack of electricity for the majority. After more than a decade of work, the chief doctor on the project reported that the preventi on of the causes of goiter had been largely achieved (Ponce 2001). The initial group, which later becam e the institution of CEBYCAM, was composed of a doctor, a social worker, an educator, a physica l therapist and a psychologist. In the beginning, t he group sought to address the problems related to goiter, and other more general health problems. The research began with household surveys of families wit h members suffering from various maladies. The group sought help from other agencies in the county to assist with the work (Ponce 2001).
41 The development program worked at many levels simultaneously: social, economic, religious, and educational. T here are three basic elements of CEBYCAMs development model: 1) heal th and rehabilitation services; 2) cooperative enterprises and production associ ations; and 3) livable communities. CEBYCAM does not envision the model they have developed as one that is replicable, but as a point of reference, noting that every sit uation is different. CEBYCAM trained teachers to handle physica lly disabled and mentally retarded students. Campaigns in the communities took place to educate people on how to improve health and nutrition through heal th sanitation improvements in the physical infrastructure and starting fam ily gardens to improve nutrition. In addition, family reunions were organized fo r physically disabled individuals. Job creation was also a component of the program to help disabled persons find work in crafts production. Moreover, the proj ect aided schools to institute rehabilitation programs as well as to start initia tives aimed at improving nutrition.
42 Figure 5: Entrance to CEBYCAM (monastery in rear) With the large increase in resp iratory problems from the Mount Tungurahua eruptions beginning in late 1999, the community faced new health challenges. In the CEBYCAM health clinic a 47 percent increase in respiratory infections was observed between Oct ober 1999 and February 2000, the time of the orange alert for Mount Tungurahua (Lara, et al. 2000a) In particular, Dr. Lara observed an increase in bronchitis and pharyngitis in children under three years old due to exposure to volcanic ash. The epidemiological data collected in Penipe demonstrated significant increases in both outpatient consultation rates for ARIs after the eruption of 1999 as we ll as higher pneumonia mortality rates in infants in 2000 (Whiteford and Tobin 2002). Over 60 percent of the inhabitants of Penipe were affected with respiratory pr oblems by the ash. Moreover, the ash
43 had a negative effect on agriculture, t hus compounding the nut ritional problems that could lead to respiratory infections. Since the mid-1980s, CEBYCAM has been instrumental in aiding poor children and families in a variety of ways. One method has been to encourage national and international individuals to adopt a family in Penipe to aid with daily expenses, pay for necessary tools and small domesticated animals, and subsidize a childs education. Ar ound 400 children and 75 adul ts are supported through this program. One of the main accomplishments of CEBYCAM is the institution of the shoe factory wher e many wheelchair bound persons work alongside people without physical disab ilities. The factory is run on the cooperative model and the shoes are sold on the national market. There is also an artisan enterprise for women in the community to produce original designs made from silk, and a gift card production cooperative. All of these examples demonstrate that CEBYCAM takes community health very seriously and believes that a multi-faceted approach which in corporates microenterprise economic development, improved health services, education, religion, and social services contribute to positive change in the community. Overall, the projects which CEBYCAM funds with the financial support of the Swiss Agency for Development and Cooperation (COSUDE) and the collaboration of the municipal authorities, helps t he people of Penipe to feel a sense of solidarity in the face of severe economic hardships and health disparities. COSUDE has had a formal re lationship with CEBYCAM since 1994. Much of the work has also been in colla boration with the Interamerican Institute
44 for Agricultural Cooperation (IICA). CEBYCAM recently officially changed its name to CEBYCAM-CES, which stands for Centro de Desarollo Humano en Cultura y Economa Solidadria (Center for Human Deve lopment in Culture and Economic Solidarity) to reflect the change in emphasis of the organization from the health problem of goiter to development more generally. Prior Research In the face of the Mount Tungurahua vo lcanic eruption, Penipes residents had to organize in order to evacuate in October 1999. The Committee of Emergency Operations (COE) helped to organize 15 Centers of Population Reactivation (CEREPS). CEREPS were groups of evacuated peasants who organized shelters to wait out the ash fall until the people could return to their homes. Some decided to move to di stant towns such as Pallatanga in Chimborazo Province (43 families from B ilbao) and Chillanes in Bolivar Province (families from Pachanillay), and ca rried out the same organizational and solidarity generating activities that had helped them to survive in Penipe (La_Nacin 2003). Two new towns were created, Sucuso and Matapalos, for evacuated residents of Bilbao and other a ffected areas in Penipe County. The ash from Mount Tungurahua affected 267 families from Penipe County and 649 families from neighboring Guano County; moreover, losses in agriculture production affected over two million hecta res of land (La_Prensa 2003). The health consequences, economic upheaval and community resilience of Penipe
45 and other surrounding communities affected by the eruption of the volcano was the subject of research by two Universi ty of South Florida (USF) professors. During 2001 and 2002, the researcher worked under the supervision of Dr. Linda Whiteford, Anthropology, and Dr. Graham Tobin, Geogr aphy, from the University of South Florida (USF) on a Center for Disaster Management and Humanitarian Assistance (CDMHA) grant ex amining the health effects of volcanic ash fall and community resilience in re lation to the Mount Tungurahua disaster. The hypotheses and research questions for the present research evolved out of the work the researcher conduct ed under the CDMHA grant (Tobin and Whiteford 2002; Whiteford and Tobin 2001; Whiteford and Tobin 2002; Whiteford and Tobin 2004). More work was needed to study health care-seeking behavior and ARI incidence in rural towns. The researcher compiled a chapter on the epidemiology of infectious disease on the areas around Mount Tungurahua for the grant final report based on the data co llected by Dr. Carmen Laspina and her research team (Whiteford and Tobin 2002). The epidemiological data (19952001) supported one of the re search hypotheses of the grant that exposure to high levels of ash fall was associated wit h people experiencing higher levels of both upper and lower respiratory problems than those not living in high ash fall areas (Whiteford and Tobin 2002). Possibl e risk factors include poverty, ash fall, and access to healthcare facilities. For the same period in 1998 and 1999 (October 16 to December 31) in seven areas of Tungurahua Province and two areas of Chimborazo Province, overa ll there was an alarming increase in morbidity for upper respiratory infections (2.6 times increase), lower respiratory
46 infections (2.5), conjunctivitis (2.3), asthma (2.1), and de rmatitis (1.9) (PAHO 2000).1 While the epidemiological data suppor ted this hypothesis, individual selfreport was contradictory. In the survey results, only six percent of respondents surveyed in the town of Penipe believed t hat the volcano was a risk to respiratory health. The possibility of recall bias, or illness narrative, was high in the structured interviews; therefore, this bias needed to be explained. According to the epidemiological data, there were significant increases in outpatient consultations for ARIs dur ing 1999 and 2001 compared to earlier years. In addition, pneum onia mortality rates for infants reached high levels. Specifically, outpatient consultation ra tes for ARI nearly quadrupled in Penipe County in 1999, the year of the volcanic eruption, and two years later in 2001, a year of renewed volcanic activity, rates doubled from the previous year, especially in the under five age group. The study found that Penipe had higher outpatient consultation rates for respirat ory infections and diarrheal diseases and higher mortality rates for children under five than two neighboring counties also affected by the volcanic ash (Whiteford and Tobin 2002). This information was reported to the press, and when the authors of the study returned to Ecuador to present the results, there was not a lot of concern ex pressed on the part of the Penipe municipal authorities. The mayor of Penipe was aware of the study and mentioned the report in an in terview that Penipe was the county most affected by the volcanic ash. The mayo r reported that the county should have 13,000 people but only had 6,014 people (La_Prensa 2002a) The chief epidemiologist of 1 This data was collected from daily outpatient and emergency visits in the 2 provinces. The data from the
47 Chimborazo Province commented in a news paper interview that the governor of the province and the mayors of Penipe and Riobamba were not present at the meeting, yet the major authorities of neighboring Tungurahua province were all present to hear the research results of t he South Florida professors (Penipe sent a representative only). The chief epidemio logist went on to warn of the dangers of the silica composition of the vo lcanic ash, and expressed the need for advanced detection equipment to gauge the se verity of the risk caused by the particulates, which might include spikes in rates of pneumonia, bronchitis and tuberculosis (La_Prensa 2002c). Another official of the Chimborazo Health Department claimed that since the South Fl orida presentation of research results did not offer the local governments any mo netary or food aid in response to the crisis, the municipal author ities were not interested in attending. However, he noted that CEBYCAM, the Civ il Defense, and the Ministry of Health were all present at the meeting (La_Prensa 2002b). This case demonstrates that NGOs and public health authorities in Chimborazo Province were better partners in the dissemination of research findings than the local government of Penipe, which was more motivated by political and monetary concerns. Site Selection The main justification for selecting P enipe County as the research site was the aforementioned health di sparity regarding ARI morbid ity and mortality rates gained from the previous experience work ing on the research team. The second Ecuadorian Ministry of Health does not distinguish whether the patients were living in shelters.
48 reason was that all of the residents of the county speak Spanish, and not an indigenous language. The thir d reason was that the area is rural, and rural children are less likely to receive care for ARIs than those living in urban areas where there is greater access to healthca re facilities. In addition, Chimborazo Province ranks very low on indicators for child well-being. According to UNICEF, Chimborazo Province ranks 2.3 out of 10 in relation to rights for children and adolescents. Moreover, the province ranks 3 out of 10 for education for children, and half of the children in the province are undernourished. Fina lly, the province holds the penultimate place for the overa ll situation of children and adolescents in Ecuador (La_Prensa 2004). Through partnership with CEBYCAM-C ES, the researcher aimed to contribute to the communitys goal of im proving health through the identification of barriers to health care-seeking in childhood ARI cases. CEBYCAM-CES authorized access to its files and newspaper archives. The people of Penipe were very cooperative and helpful in the re search process. The health of the people had improved since the eruption of the volcano, and some explained that the people had become accustomed to the volc anic ash. Whatever the reason, the researcher was able to identify cont inuing health problems that needed to be addressed.
49 CHAPTER FOUR: LITERATURE REVI EW THEORY, METHODOLOGY, AND ACUTE RESPIRATORY INFECTION PART I MEDICAL ANTHROPOLOGY THEORY In medical anthropology theory, the two primary paradigms are medical ecology and critical m edical anthropology (CMA). Medical ecology is an empirical, materialist perspective which took hold in the 1960s, that focused on human adaptations in micro-populations that allowed peopl e to utilize the environment for the purposes of resource extraction. CMA, on the other hand, came later, and lobbied to shift the focu s towards broader social relations in society and the political economy in a histor ical context. The critiques of medical ecology, primarily by Singer (1989), Scheper Hughes (1990), Singer and Baer (1995), and Morsy (1996), as well as attemp ts to achieve a new synthesis of understanding local realities in global cont exts, by McElroy (1990), Armelagos (1993), Leatherman (1993, 1998), Goodman and Leatherman (1998), and Baer (1996, 1997) have led anthropologists to rethink the role of political economy in their research and examine how sociocu ltural and political-economic processes affect human biologies and how negative out comes of these processes affect the larger society (Goodman and Leatherm an 1998). This section will examine
50 these arguments and critiques and conclude with a discussion of work in political ecology, which is one example of t he new biocultural synthesis. Medical Ecology and Critical Medical Anthropology One of key tenets of m edical ecology is to explain how human populations adapt to a changing environment. Rather than focus on the biomedical causes of disease, medical ecology is interest ed in how humans adapt to particular environments through evolutionary mechani sms, that include cu ltural adaptations as well as genetic or immunological re sponses (McElroy and Townsend 1989). Like epidemiology, medical ecology recogn izes the causal assemblage of factors that cause disease. Medica l ecology defines disease as a chain of factors that stem from ecosystemic imbalances, and that both health and disease are consequences of interpenetrating subsystem s that are physical, biological and cultural (McElroy and Townsend 1989). Ther efore, the model is a holistic one, a homeostat where imbalance leads to adaptation or maladaptation, and consequently, health or disease. Like it s theoretical antecedent, cultural ecology, developed by Julian Steward and later retooled as neofunctionalism by Roy Rappaport and G. Reichel-Dolmatoff, medi cal ecology focuses on how culture enables a population to adapt to a particular geographical area. The critique of medical ecology by medical anthropologists is primarily aimed at biological anthropologists. Biological anthropologists have used an ecological model in order to underst and biocultural responses to disease (Armelagos, et al. 1992). This ecological model takes its cue from epidemiology,
51 which configures disease as the outcome of the in terplay between the host, agent, and environment. In epidemiology, the agent or pathogen is considered the primary cause of disease, to the excl usion of social and political factors. Traditional epidemiological approaches have been critic ized as too focused on individual behavioral risk factors and not atte ntive to critical contextualization (Farmer 1999). In Latin Am erica, the Centro de Estudi os y Asesora (CEAS) has developed what it calls a Ma rxist epidemiology, which examines the relationship between social class and health status (Mor gan 1998). Social epidemiologists at CEAS have critiqued traditional epidemio logical categories and methods and have examined differential morbidity and morta lity according to variables of social class and historical moments. CMA strives to combine the micro and macro to understand microparasitism (agents or proximate causes of disease) as well as macroparasitism (social relations, which are considered to be the overarching causes of disease) (Singer and Baer 1995). Nancy Schepe r-Hughes identifies the macroparasitism of class exploitation as the social cause of mortality and morbidity (Morsy 1996). George Armelagos, et al. (1992) concede that the role of social organization as well as t he ideological system has been largely neglected by medical ecologists. A rmelagos, et al. (1992) echo medical anthropologists critiques of general m edical ecology as furthering an outmoded functionalist, adaptationist model steeped in biomedical conceptions of disease and illness; however, the authors cont end that the corrective, termed the emergent biocultural perspective, ackno wledges that culture forms a part of the
52 environment and that biomedicine is simply another ethnomedicine, and consequently, social relations as well as subjective experiences should be incorporated into the new model to understand disease. In an attempt to appease CMA proponents who would toss out the notion of adaptation altogether, Armelagos, et al. (1992:42) reform adaptation by naming the concept, adaptive process and coping beha vior, and stress th at a model of adaptive process conceptually locates both proximate and ultimate causation of illness within the constraints, diminis hed options, and contradiction in goals and responses of human populations. This, in turn, emphasizes the role of politicaleconomic factors and social rela tions in the health process For Merrill Singer, an emphasis on the role of social factors is not sufficient. In Singers critique of Wiley ( 1993), he claims that bioculturalists are overly concerned with how people cope wit h environmental disruptions rather than what he calls the original causes of the disruption, i.e. social relations (Singer 1993). However, W iley counters with the argument that the crux of the debate lies with adaptation, and that as an evolutionary biocultural anthropologist, the concept of adaptati on provides the best explanatory framework for causes of health and dis ease (Wiley 1993). Singer promotes a version of CMA whose goal is not simply to understand the social causes of disease, but to correct oppr ession and exploitation in t he realm of health care through ethical research (Singer 1998).
53 Interchange Between Theories Thomas Leatherman, et al. (1993) attempt to parse out the common elements in the debate between Wileys bi oculturalism and Singers CMA. The authors claim that Singers critical bi oculturalism combined with Wileys new approaches to biocultural adaptation is fertile ground for interchange between CMA and medical ecology (Leatherman, et al. 1993). While biological anthropologists have not abandoned ecology or adaptation, attempts to address the critiques of t he critical medical anthropologists have placed a new emphasis on social, political and economic factors that affect human biology, health, and disease. For example, contrary to ear lier studies of health and stature of Andean populations which focused on biology and genetics, although some looked at socioeconomic status, more recent resear ch has shown that growth and stunting are highly associated with economic st atus and has refuted the small but healthy hypothesis (Leatherman 1998a). T hese earlier studies did not examine how differential access and control of resources could lead to varied health outcomes, an issue which is addressed in more recent studies which focus on the political economy of health. Theref ore, as a correctiv e to the adaptation model, biological anthropologists in t he 1980s began to examine the effect of social inequality on human bi ology from prehistoric ti mes to the present, which has been labeled the political econom y of human biology and critical bioculturalism, among other nam es (Goodman and Leatherman 1998). Nevertheless, one of the inherent problems of the medical ecology model is an emphasis on the pathogen, or the broa d category of insults as the disease-
54 causing agent. When an even broader caus al assemblage is considered, adding political-economic factors, the model bec omes even more difficult to ascertain the magnitude of separate causes for diseas es. Medical ecology is criticized for its inability to ascertain relative levels of factors affecting disease causation as well as the identificatio n of trends across different disease complexes (Baer 1996). Armelagos, et al. (1992) attempt to address this confusion by placing coping or adaptive behavior at both the macro and micr o level simultaneously in order to address social relations; howev er, the holistic model of adaptation is incongruent with the processual and dialectical model of CMA. As Singer (1989) cogently points ou t, the medical ecology model is teleological, as the population is prom pted to adapt to the changed environment in order to be healthy. The problem with adaptation, according to Singer (1989:230), is to answer the question of what is adapting to what. Rather than focus on how culture helps individuals or populations adapt to changes in the environment, Singer argues that medica l anthropology should be concerned with how social transformation di fferentially affects the health of individuals and populations. Because medical ecology is unable to address social transformation, Singer argues that medi cal anthropologists should consider CMA as the more defensible paradigm. Some critics have accused critical m edical anthropologists of misusing the term biomedical hegemony to refer to the process of th e proliferation of poverty and disease in the developing world and de legitimizing ethnomedicines (Csordas 1988; Estroff 1988). Another critic of CMA, states that proponents of the new
55 paradigm only shared terminology is t he hegemony of biomedicine which by itself has relatively little meaning (Pelto 1988). The problem with CMA, as Pertti Pelto sees it, is that a new paradigm should generate new research questions, but he sees CMA as a post facto theory used to interpret and understand findings. Pelto points out t hat medical anthropology is la rgely an applied field, concerned with the development of new methods and seeking answers to the question of how to make things be tter in a given situation. Paul Farmer (1999) is adept at shi fting from the medical gaze to the critical (self-critical) one. Farmers work is more akin to Scheper-Hughes poststructuralist critical reflexive medi cal anthropology than Singers political economy of medical anthropology. Howe ver, Farmer (1999: 58) clarifies his approach as ecumenical and complementar y, in contrast to Scheper-Hughes activist, revolutionary approach. Scheper-Hughes in-depth work, Death Without Weeping is a postmodern treatment of childhood death in a shantytown in northeastern Brazil. The two folk illnesses affecting infants, nervos (nerves) and doena de nervos (nervous sickness), are interpreted as Foucauldian discourses, reinforced by medical practitioners and accepted by the disenfr anchised sector, despite the belief by Scheper-Hughes that these terms mask the real, unspeakable cause of illness, namely slow, forced starvation (Schepe r-Hughes 1992). Scheper-Hughes (1990) argues for three different ways to dev elop a critically applied medical anthropology: 1) to further a program of demedicalization which recognizes that persons other than biomedical practitioner s are more adept at identifying the
56 social causes of disease; 2) to enc ourage anthropologists to collaborate with medical practitioners outside the realm of biomedicine; and 3) to initiate a revolution in how medicine is practiced in hospitals and clinics to achieve a radicalization of medica l knowledge and practice. Farmers intention is not to suppl ant complementary methodologies of related disciplines with his critical approac h; rather, he aims to add an equal dose of anthropology to biomedicine in order to lead to a more accurate understanding of disease and care for his patients. Farme r criticizes the biomedical gaze as too narrow, not as overly hegemonic. Si nger appears to be influenced by James Scotts theories of peasant resistance when he writes of the locus of the sufferers experience in the context of hegemony and resistance. Farmers critical perspective allows him to ask provocative questions such as, By what mechanisms have international changes in agriculture shaped recent outbreaks of Argentine and Bolivian hemorrhagic fever, and how do these mechanisms derive from international trade agreem ents such as GATT and NAFTA? Does privatization of health services buttress social inequalities, increasing risk for certain infections and poor outcomes among the poor of sub-Saharan Africa and Latin America? (Farmer 1999:52). Thes e are the types of questions that medical anthropologists should be asking in relation to emerging infectious diseases in the developing world. Hans Baer makes the point that CMA needs to incorporate political ecology into its political economy pers pective (Baer 1996). Baer (1996) argues that medical anthropologists should fo cus on processual, dialectical models
57 rather than homeostatic ones. Political ec ology recognizes that capitalism is the force behind the destruction of the environment and the exploitation of nonrenewable natural res ources, and only through a commitment towards changing the social relations of producti on and reproduction, is health to be achieved. Theory in Research Context The political ecology approach permits the researcher to start with the social relations imposed by the political economy and examin e the implications for the overall cultural coping mechanism s of the community in response to the volcano hazard and limited health care options. The volcano hazard has varied effects on the community depending on the inte rnal social differentiations of the community; therefore, a familys ability to seek timely medical attention is a factor of its socioeconomic status. The purpose of this research is not to gauge the environmental risk factors for ARIs, such as volcanic ash, but t he possibility of ash exacerbating the problem of respiratory infections in this community is noted. The research focuses on the response of mothers to thei r childrens respiratory infections to identify the multiple strategies for seeki ng health care. The research aim is to determine how female caregivers make heal th care decisions when their children suffer from respiratory infections depend ing on their own perceptions of illness etiology and diagnosis, and the multiple constraints on their health care-seeking behavior. This enters the realm of ethnomedicine and is the topic of the next
58 chapter. In conclusion, at the macro-leve l, actors are situated in the politicaleconomic-environmental context, but at the micro-level, et hnomedically defined conceptions of health and illness and health care-seeking decisions based on access to monetary resources and healthcare facilities comprise the mosaic of peoples everyday lives. Political Ecology Anthony Oliver-Smith is a proponent of the political ec ology (PE) approach to understanding environmental hazards and disasters (Hoffman and OliverSmith 2002; Oliver-Smith and Hoffman 1999) Oliver-Smith begins with the political economy framework and infuses it with the social science perspective of medical ecology (Oliver-Smith 1999). He b egins with the concept of vulnerability, synonymous with a communitys adaptive failure, in order to characterize a disaster, emphasizing that it is the patterns of vulnerability in a society that affect how events play out even more than the act ual catastrophic event of the disaster. Segments of society, dependi ng on the degree of social di fferentiation, are better equipped to cope with a disaster when it occurs because they are in a more privileged position in terms of power relati ons. Oliver-Smith (1999) proposes that a PE framework succeeds at explaining not only the causes and effects of disasters but at elucidating the co mplex dynamics of the social and environmental processes, cultural adapt ations, and community developments that are also implicated in a disaster thereby linking society, environment, and culture.
59 Oliver-Smiths version of political ecology represents one side of the ongoing debate between cultural constructi vists or what Vayda and Walters call the camp that has turned political ecology into politics without ecology and the scholars who have fused human ecology wi th political economy (Brosius 1999; Escobar 1999; Vayda and Walters 1999). T hat the terrain of political ecology represents a relatively new paradigmatic trend is evi denced by the definitional debates and contested definitions in re cent book chapters and journal articles (Escobar 1999; Paulson, et al. 2003; Vayda and Walters 1999; Watts 2002). Whereas Escobar and Watts represent the constructivist camp, Vayda, OliverSmith, and Stonich are more representative of the ecol ogist/materialist side in their insistence on the transformative power of the environment and their adherence to the concept of adaptation (Stonich 1999). Despite different orientations and emphases, advocates for a political ecology approach do share some common concepts. Watts (2002) defines the conceptual toolkit as: 1) marginality (t he cause-effect relationship of social relations and environmental degradation); 2) pressure of production on resources (also implicated in social relations); and 3) a plural approach (recognizing the effects of discourse on policy and practice and the politics of representation). One example of marginality examined l and degradation in the Amazon where the irrationality hypothesis of peasant land destruction by cattle ranching was refuted and replaced with an explanation that accounted for social and legal forces (Blaikie and Brookfield 1987). T he researcher explored these themes in his Masters thesis, and was largely influenced by William Durhams work to
60 understand the processes of deforesta tion in the Ecuadorian Amazon (Durham 1995; Luque 1997). Durham (1995) describes two posit ive feedback loops one labeled impoverishment and the other capital accu mulation. The capital accumulation loop begins when a consortium of entr epreneurs, companies, and small farmers work together either directly or indire ctly to create avenues for the development and deforestation of large blocks of forest in order to extract valuable natural resources. Contributin g factors include domestic and foreign demand and the appropriate government policies in place to facilitate these activities. Population displacement and land scarcity are the product of these forces and a secondary cycle of poverty results, fed by a lack of economic alternatives and dwindling household incomes. Durham infers t hat one of the possible responses to impoverishment is migration wher ein the cycle of impoverishment and environmental degradation is r eproduced in frontier regions. The previous example demonstrates t hat by stressing social relations, while maintaining a holistic framework, one can construct a model for how a disaster process may be conceptualized. In the context of Mount Tungurahua and its affected communities, the political ecology framework fits well with the research concerning vulnerability and adapta tion to adverse circumstances. In the case of Penipe, many young people left the community because of the lack of economic opportunities and the negative im pact of volcanic ash on their crops and livestock (Whiteford and Tobin 2002) However, not everyone had the economic or social resources to leave t he community, so they adapted to living in
61 a hazardous area. Vulnerability, expr essed in terms of negative health outcomes, increases as poverty increas es, and a perpetual cycle of disease and poverty ensues in the face of an ongoing hazard. Leatherman argues that vulnerability has its roots in unequal soci al relations, wherein rural producers endure illness, which in turn increases t heir vulnerability to economic exploitation because of limited househol d production and reproduc tion (Leatherman 1998b). Vulnerability therefore means exposure to risk of disease, hunger, and poor living conditions, including access to potable wa ter. The response to risk exposure is resilience, which in the case of childhood illness, takes the form of a mothers strength and perseverance to overcome these risks through coping mechanisms, which are strengthened by her educational and economic resources, as well as social support network. O ne of the limitations of t he Ecuador study is that the researcher was not able to intervie w the people who had already left the community, in order to understand why t hey had opportunities that the ones who remained did not. There are visible economic disparities in the town, and the health of the impoverished me mbers of the community, in terms of respiratory health, is worse for those who live in houses with dirt floors or are poorly insulated, and who cook with wood fi res near the areas where they sleep, because they cannot afford gas fuel. There is still much work to be done on the combination of a political ecology framework with medical anthropolog y questions. Central to a political ecology approach is a concern with power relations and how politics is used to wrest power from some to be wielded by others. An understanding of various
62 stakeholders interests and goals assists the anthropologist to ascertain which communities are considered integral to the local economy and which specific actors shape these policies. In the cont ext of medical anthropology, a political ecology approach allows t he anthropologist to consider the social causes of disease, in this case acute respiratory in fections in children. Because of lack of education, limited access to medical care, and few financial resources, studies have shown that rural, impoverished co mmunities are at a greater risk for childhood ARIs (Nichter 1994). While other factors ar e also significant, the political-economic causes of disease must be part of the causal assemblage of multiple factors.
63 PART II ETHNOMEDICINE IN THE ANDES Recent medical anthropological studies of the Andes support the thesis that medical systems act as discourses about the human body (K oss-Chioino, et al. 2003). These systems are conceptualiz ed by anthropologists dichotomously as ethnomedicine and Western biomedici ne. Ethnomedicine is composed of various folk illnesses, healing traditions, and beliefs that may or may not have a scientific basis according to Western biom edicine. Folk illnesses are those which Western biomedicine acknowledges that it possesses a lack of expertise and understanding and lacks the diagnostic cat egories to define (Browner, et al. 1988; Stein 1981). Some anthropologists identify Western biomedicine as another ethnomedicine, whose treatm ents and cures have successes and failures just as ethnomedicine, and are also products of long-term processes influenced by culture, hist ory, and politics (Rubel and Hass 1990). This chapter traces recent research on Andean ethnomedicine with an emphasis on anthropological studies carried out in Ecuador. Studies of Andean Ethnomedicine The first major collaborative work on the medical anthropology of the Andes was Health in the Andes published from the collection of papers at the American Anthropological Association meetings in 1978 dealing with Andean health, nutrition, medicine and morta lity (Bastien and Donahue 1981). The collection was divided into three secti ons, which mirrored the panel discussions:
64 Andean ethnomedicine, Andean environment, and improving Andean health. The physical environment and human adaptati on to it were dominant themes in anthropological theory of this period. Factors such as altitude and verticality were cultural determinants for biologi cal and behavioral adaptations. Another major theme was reciprocity, both in terms of labor and gender relations and ritual obligations. In the section on Andean ethnomedicine, Joseph Bastien exemplified the importance of reciprocity in his descrip tions of the people of Qollahuaya, who employed the metaphor of t he mountain to understand the health or sickness of the human body (Bastien 1978; Bastien 198 1). Bastien explained that the traditional healers function was to rest ore the bodys harmony by feeding the mountain and resolving conflicts such as land disputes. Through ritual, the healer reordered the universe through fulfil ling reciprocal relationships between people and the mountain. The secti on on ethnomedicine demonstrated how a proper understanding of folk illnesses was necessary for the delivery of health care services. Steins chapter empl oyed language which anthropologists today would not use by describing the resi dents of Vicos as ignorant, poor, and unhealthy (Stein 1981:64). Stein made this characteri zation to argue that the enemy of Western biomedicine was not erroneous ethnomedi cal understanding, but rather poverty and isol ation. While some anthr opologists today might agree with his conclusions, they would couch it in different language and not impose biomedically-based value j udgments on ethnomedical beliefs. Other authors in this section, Buechler and Urioste, de scribed different ethnomedical conditions in
65 an effort to explain how su ch beliefs shed light on the interrelationships between culture, biology, health practices, and ritual. Throughout the 1970s and 1980s, Andean anthropology shifted from ecological orientations towards approaches which incorporated hegemony, political economy, feminist theory, and post-colonial theory to analyze complex processes of cultural change and assim ilation of indigenous peoples (Miles and Leatherman 2003). In one of the most extreme examples from medical anthropology, Mary Crain describes devil possession in Andean women as an ethnomedically-appropriate response to expl oitative hegemonic relationships in society (Crain 1991). This article draws fr om previous works which view esoteric cultural phenomenon such as devil possessi ons and contracts as reactions to the introduction of the capitalist mode of production (Ong 1987; Taussig 1980). In this model, medicine is construed as a metaphor for social relations embedded in a political economy (Crandon 1986:466). Health and illness were now being explained by medical anthropologists in t he Andes as not only caused by culture and biology, but also intertwined with so cial and political relationships. Embodiment was introduced as a new theoretical concept meaning the inscription on the body of social i nequalities, and it gained currency in the anthropological literature (Miles and Leatherman 2003:7). A special issue of the journal, Social Science and Medicine titled Landscapes of Health in the Andes, re visits some of the themes and issues from Health in the Andes (Greenway 1998b). Greenway explains that medical anthropologists have moved away from adaptationist theories which emphasize
66 environmental constraints, towards interd isciplinary models which take account of the plurality of health systems in the Andes. Greenway examines the diagnosis and treatment of susto (fright sickness) in high land Peru. By analyzing susto from an emic, or insider, perspective, Greenway argues that the sufferer is playing out a deep cosmological discourse through the experience of being sick and existing outside of the normal cultur al boundaries. The Quechua worldview is a holistic one, and consequently, healing is achieved through t he restoration of cultural and cosmological balance by t he healer who reintegrates the patient to the Quechua universe who, in the experience of susto experiences a conflict of identity. Greenway concludes that ethnomedical understandi ng will facilitate improved health care delivery (Greenway 1998a). In another article focused on ethnom edicine, Larme reiterates the importance of economic reform, as stat ed in Bastien and Donahue (1981), as critical to improving Andean health (Larme 1998). Larmes article focuses on the concept of human vulnerability in A ndean ethnomedicine in order to analyze, from a feminist perspective, why wom en suffer from emotion-based illnesses more than men. She concludes that ethnomedical knowled ge in the southern Peruvian highlands which classifies wo men as vulnerable and weak has negative consequences for womens health, status and power in Andean society. Bastien has examined underutilization of health services in Bolivia and biomedical practitioners ignorance of i ndigenous traditional beliefs, which would aid in the transference of health informa tion to treat diarr heal disease (Bastien 1987). Bastien argues that the primary reason that Indian peasants resist
67 modern medicine is due to financ ial, psychological, physical, and cultural barriers (Bastien 1987:1111). Bastien demonstrates how applied anthropology can make a difference for a serious health problem in his description of how a myth was retold in a health informa tion brochure to teach peasants the importance of using rehydration salts fo r the treatment of severe forms of diarrhea with cultura lly-appropriate language. Biocultural Approach Two of the journal articles, Leathe rman (1998) and Oths (1998), employ a biocultural perspective to Andean anthr opology. Leatherman (1998) advocates for a critical biocultural model as a corrective for previous adaptationist approaches that combines biology, ecol ogy, and political economy. Biological anthropologists adopting the critical bioc ultural perspective have placed a new emphasis on social, political and economic factors that affect human biology, health, and disease. Medical ecology wa s ill-equipped to deal with resistance, conflict, and revolt as adaptive behaviors, behaviors which were omnipresent in Latin America throughout the 1970s, and continue today. The problem of specific etiology, or germ theory of biomedicine, applied to Andean peoples is highlighted in Oths wo rk in highland Peru In her discussion of the illness debilidad (weakness), Oths finds that the illness is most often experienced by women in vulnerable time s of their lives from household stresses, and is explained as resulting from a conf luence of biological and social factors (Oths 1999). She employs the concept of embodiment to demonstrate how
68 debilidad is manifested in people at the height of their status in the community, but at the same time during their least pr oductive years. In her journal article from the special issue Oths proposes a biocultural model to understand regional differences in Andean health status to demonst rate that intracultural diversity is a predictor for health differences (Oths 1998). Oths conducts a meta-analysis of health indicators to show the range of dive rsity in health status within regions. A more recent research study examines the effect of wom ens food consumption on household economic standing in the s outhern Andes and finds that lack of money is a greater risk factor for lower nutritional status than landholdings during pre-harvest season (Graham 2004). Prev ious models failed to gauge economic, cultural, and health differences within par ticular regions, thus making health interventions less effective. Medical Pluralism The third major compilation of essays on Andean ethnomedicine was the publication of Medical Pluralism in the Andes (Koss-Chioino, et al. 2003). This volume was dedicated to the life and wo rk of Libbett Crandon-Malamud, and in particular, to her contribution to t he field of medical anthropology with the publication of From t he Fat of Our Souls a study of medical pluralism in highland Bolivia (Crandon-Malamud 1991). The book is divided into four sections. The first is a collection of four essays on the contributions of Crandon-Malamud and Andean medical anthropology. The second section is composed of three case studies of medical pluralism in the Andes. The third section is an eclectic mix of
69 three essays ranging from a chapter by Ba stien on Chagas disease to an article on ethnomedicine by McKee dealing with mal aire (evil wind). The final section is subtitled, Gender, power, and health and is a further exploration of ethnomedicine and traditional healers. The edited vo lume is a significant addition to studies of medical pluralism in the Andes. In the opening chapter, Miles and Leatherman point out Crandon-Mala muds major contribution to this area of study in their assertion that s he identified medicine as a resource that was both negotiated and was a method by which different social actors expressed cultural identity and po litical and social power (Miles and Leatherman 2003:8). The authors of the Medica l Pluralism in the Andes recognize that myriad health practices and systems operate si multaneously in the Andean context. Choices of care depend on a variety of fa ctors, and different types of health practitioners, including biomedical prac titioners, are sought depending on the particular cultural, political, and social circ umstances of the actors seeking care. Crandon-Malamud demonstrated t hat in the diagnosis of susto (fright sickness), diagnosis was more of an economic decis ion because it meant that treatment would be at home rather than at a more expensive biomedical facility (CrandonMalamud 1983; Crandon-Malamud 1989). Crandon-Malamuds main argument is that medical dialogue is a means by which political and economic resources are exchanged and is thus a mechanism t hat facilitates or inhibits change (Crandon 1986:466). Broadly defined folk i llnesses such as evil eye are oftentimes treated by lay traditional health practitioners, known as curanderos
70 (healers) brujos (sorcerers) chamanes (shamans) or herbalistas (herbalists) in the Andean context, and these treatments are both more easily accessible and less expensive than biomedical care. In highland Bolivia, there are multiple medical traditions, and treatment choice involves both the negotiation of ethnic identity and exchange of po litical and economic capital across ethnic and class boundaries (Crandon 1986). Both mestizo (of mixed parentage) and indigenous people of the Andes, maintain a belief in mal aire a potentially fatal malady caused by winds from the dead upon the living (McKee 2003). The young ar e especially susceptible to this illness. McKee suggests that the ethnomedical diagnosis of mal aire in children in indigenous highland Ecuador is a method of enculturation, or internalization of world view (McKee 2003:133). McKee argues that blood and fat are symbols which stand for vitality and energy; theref ore, children who are malnourished are often diagnosed with mal aire because they are considered weak and susceptible to attacks from the evil spirits. Ot her anthropologists have noted the importance of wira, tulla, and yawar (Quichua for fat, bones, and blood) as three basic bodily principles whose well-being was seen as a sign of health for Andean peoples (Bastien 1978; Weismantel 1988). Blood and fat are, like personal property, subject to robbery, and weak bl ood is the disastrous result of such an attack. Traditional healers are employed to pur ge the victim of the malady, and oftentimes use foul-smelling herbs or tobacco to cure the patient. Of all the journal articles related to ethnomedicine in Ecuador, one by McKee in particular relates to the dissertat ion research. In this article, McKee
71 identifies three main ethnom edical categories for the di agnosis of gastrointestinal disease in children (McKee 1987). Such illnesses are classified as: 1) for the doctor; 2) not for the doctor, home treat ment; and 3) supernatural afflictions, treatment by traditional healer. All diagnos es are made after an examination of the stool. McKee concludes that the folk taxonomy of gastrointestinal diseases is accurate enough, and from the standpoint of the rural poor, most treatments will be home-based regardless. The class of di sease for the doctor, is believed to be caused by infection and is treated by antib iotics. In other cases, either dietary improvement or treatments for incurable parasitosis are recommended. In the latter case, medical attention would be in vain because of the guarantee of reinfection in poor areas with unsafe water sources. Although not represented in the three collections on Andean ethnomedicine previously discussed, t he work of Ruthbeth Finerman has contributed to the knowle dge of home-based health care and womens roles in the southern highlands of Ecuador (F inerman 1983; Finerman 1987; Finerman 1989a; Finerman 1989b; Finerman and Sack ett 2003). Finermans first published study draws from her dissertation research in Saraguro, an indigenous community in southern Ecuador and home to the former president of the Confederation of Indigenous Na tionalities of Ecuador (CON AIE), Luis Macas. In this article, Finerman ex amines diagnosis and health care-seeking behavior of Saraguro women (Finerman 1983). She finds that indigenous ethnomedical illness categories, such as the belief in hum oral imbalances (hot, cold), soul loss, mal aire, etc., do not figure in the diagnostic categories of biomedically-trained
72 physicians, and this denial threatens the wo rld view of indigenous people seeking care. Finerman concludes that because of this disconnect and lack of health access issues, people from Saraguro are more likely to seek traditional healers. She reiterates these findings in a late r article but shifts the focus to the household, which she identifies as a locu s for refuge from illness. Home-based care is considered most appropriate for some types of illnesses when leaving the home and being exposed to strangers and fore ign illnesses would be considered dangerous (Finerman 1987). Finerman adds to the knowledge on ethnomedical beliefs of Andean peoples in order to assi st health authorities adopt culturallyappropriate preventive health measures. Because of Finermans long period of re search in Saraguro, she is able to make observations regarding changes in the home-based health care system in highland Ecuador. Through the results of qualitative and quant itative studies, Finerman finds that womens gossip netwo rks are the most common sources of new knowledge on health care (Finerm an 1989b). She concludes that homebased health care is receptive to particula r elements of biomedicine, and that for home-based illness treatments, biomedica l products are combined with herbal remedies. In a recent article, Fine rman finds that the quality of womens medicinal gardens reflect a womans stat us and that judgments by peers on the quality of a womens gardens gauge that particular familys health and economic status (Finerman and Sackett 2003). Throu gh her carefully constructed research studies of home-based health care, Fine rman significantly contributes to the discourse on medical pluralism in the Andes.
73 Finerman and Sackett (2003) descri be home gardens as medicine cabinets, and claim that the majority are medicinal plants used to treat nervios (nerves), mal aire colds, colic and reproductive ailments. The authors reaffirm that Andean ethnomedicine maintains the hum oral opposition of hot and cold, so, for example, for a hot illness such as a stomach ache, cool mint tea is the appropriate remedy. Curiously, shamanism is not mentioned in this article, although the primary halluci nogenic plant in Andean shamanism is listed in garden inventories, the San Pedro cactus. The majority of shamans in Andean countries whose charge is to cure sorc ery-induced afflictions are men; however, powerful curanderas (women he alers) are also a feature in many communities, although less has been written about them (Glass-Coffin 1998; Glass-Coffin 2003; Muratorio 1998). In both the hi ghlands and the lowlands, predominantly male shamans cure sorcery-induced afflic tions in a ritual ceremony assisted through visions induced by hallucinogenic substances. These maladies are considered outside of t he purview of curanderas In the highlands, shamans or cur anderos perform all-night healing ceremonies called mesas (a table or altar with various religious icons) in order to cure their patients. One of the most famous Andean shamans in the anthropological literature is Eduardo Cal dern, whose folk-healing practices have been described by anthropologists (Joral emon and Sharon 1993; Sharon 1978; Villoldo and Jendresen 1990). Sharon (1978) compares shamanism to a form of symbolic communication, and like verbal co mmunication, it adap ts over time. The healing shamans task is to cure the victim of sorcery-induced illness caused
74 by another shaman, who was either hired or attacked the victim directly. Envy of the victim is usually the alleged mo tive behind sorcery attacks (Glass-Coffin 2003). The healing shaman, in exchange for payment, cleanses the victim with chants, cane liquor, and medi cinal plants, while in a hallucinogenically-induced trance state. Andean shamanism is cons idered syncretic, a mixture of Inca, Catholic, and Western esotericism, and many of the chants and idols used in the mesa combine multiple religious symbols (Sharon 1978). This brief review of Andean ethnomedicine discussed the major anthropological works which have shaped theor y and practice in this subfield. Researchers in medical anthropology have generally followed the trends in anthropological theory and method in their field studies and publications. Like the variegated discipline of anthropology, different researchers have followed divergent research agendas. Some like Leatherman and Oths have pursued quantitative studies based on biocultural models, while other s, like Greenway and Larme, have followed a political economy approach focused on power relations. The approach of this dissertat ion research draws on the latter, with an emphasis on power relations and an attent ion to differences in ethnomedical and biomedical understandings of illness. W hat these anthropologists all have in common is a genuine concern for improv ing Andean health. McKees insights are particular illuminating to the present research because she focuses on hometreatments for childho od illnesses, a significant finding in this research. By eliciting pertinent ethnomedical categorie s, McKee finds that diagnoses of childhood diarrhea are not only medical decis ions, but economic ones as well.
75 Rural peoples in the Andes make the most of limited medical resources available to them. Expensive medical technology is not necessary to raise the level of health care to rural peoples of the A ndes. Through understanding ethnomedical categories and explanations, the anthropol ogist comes closer to cultural understanding in the research setting, and is therefore better able to recommend educational materials and low-cost medica l interventions to improve health in struggling communities.
76 PART III METHODOLOGY There has been considerable debate by academics and practitioners on the relative compatibility or incompatibi lity of the disciplines of epidemiology and anthropology, more specifically, both de scriptive and analytic epidemiology and cultural anthropology. Most of the literatur e on the subject is written by anthropologists, who are also trained in ep idemiology. Anthropology has a long tradition of crossing disci plinary boundaries, and in me dical anthropology such cross-fertilization of theor ies and methods is a product of the emergence of this particular subdiscipline. Both disciplines require methodologica l rigor, with medical anthropology employing a combination of qua litative and quantitative methods and epidemiology utilizing prim arily quantitative methods. Positive commentaries reflecting the potential colla boration between the two di sciplines highlight the complementarity of methods (Agar 1996; Dunn and Janes 1986; Hahn 1995; Inhorn 1995; Trostle and Sommerfeld 1996; True 1996). For example, Dunn and Janes (1986) note that methodological issues between the two disciplines are easily overcome because quantitat ive and qualitative methods are complementary. On the other end of the spectrum are the crit ics. One anthropologist points to the methodologic al rigor mortis of epidemiology (Nations 1986). Another anthropologist argu es that since the methodology of the two disciplines is fundamentally differen t, this limits the types of research questions that can be considered as well as the answers or explanations to these
77 questions (DiGiacomo 1999). According to DiGiacomo (1999), anthropology is both diverse and eclectic in its myriad t heoretical perspectives. Theories are frequently stated a priori to any methodological discussions; however, in epidemiology, theory is neit her a prerequisite to meth odological discussions, nor is there any required explanatio n for its absence (Krieger 1994). DiGiacomo recounts her largely negat ive experiences to integrate an anthropological perspective into an epidemiolo gical cancer research project. Her criticisms may give one reservati ons on the prospects of convincing epidemiologists of the merits of particu lar anthropological theories and methods. For example, DiGiacomo (1999) remarks that the old stat ic view of culture as a closed and stable system has persisted in ep idemiology because it provides an easier fit into causal models. Neve rtheless, other anthropologists remain optimistic. For instance, Agar (1996) views ethnography as not simply a new method that epidemiologists mi ght employ, but as a tool that could transform how epidemiology is practiced. Ethnography for Epidemiologists In order to evaluate the usefulness of ethnography fo r epidemiologists, the strengths and weaknesses of the ethnographic method mu st be assessed. One text that would appeal to the scientific-minded epidem iologist argues that ethnographic research and methods are not only practical but can be used to evaluate and test scientific hypotheses (deMunck and Sobo 1998). Ethnographic research contrasts with epidemiologica l methods in a number of ways.
78 According to deMunck and Sobo (1998), t he aim of ethnographic research is generally: 1) exploratory, 2) classificato ry, or 3) associational. In contrast, analytic epidemiology is primarily concerned wit h testing hypothetical relationships and satisfying causal criter ia, after conducting case-control, cohort, or clinical trial studies. Infectious di sease epidemiology is a more expensive undertaking than an ethnographic st udy by definition, becaus e in addition to large sample sizes, it requires laborator y support (specifically microbiology and serology) and complex statistica l analyses (Jekel, et al. 2001). Two important concepts to evaluate any scientific research project are validity and reliability. Validity refers to the accuracy of the findings, and reliability means that the methodology is sound, or that the study is replicable. Validity for ethnography refers to whether the researcher s conclusions based on ethnographic research reflect reality in the world (deMunck and Sobo 1998). Because of the ethnographers deep familiari ty with the community, observational data correlate with cultural meaning, t hus increasing validity (Weiss 1988). For ethnographers, research bias is the resu lt of too many preconceptions about a culture based on lack of familiarity or theoretical preconceptions that the researcher brings to the field and dete rmines the choice of methods (deMunck and Sobo 1998). In epidemiology, compromised validity is a result of selection bias or allocation bias. Selection bias refers to the situation where subjects are permitted to choose in which arm of the st udy (intervention or control) they wish to participate. Allocation bias occurs when study investigators fail to employ a
79 random method of allocating participants to arms of the study (Jekel, et al. 2001). Epidemiologists aim to minimize bias, whil e anthropologists attempt to explain it (Trostle and Sommerfeld 1996). In epidem iology, sample selection is based upon predetermined demographic characteri stics and disease status (Dunn and Janes 1986). Epidemiologists discuss inte rnal and external validity. External validity exists when the study findings can be extended and hold true for a larger population and not just for the study sample, whereas internal validity is found when the study results hold true only for t he study population (Jekel, et al. 2001). A study must have at least internal validi ty to be considered worthwhile. In order to establish validity, one prerequisite is the randomization of study participants, so that each individual has an equal chance of being selected. When randomization is compromised, due to se lection bias for ex ample, external validity is compromised, so the resu lts are only generalizable to the study population rather than to the general population. In the past, an ethnographic sample was infrequently based on a sampling method because ethnographers generally worked with small, delineated populations in order to est ablish a range of cultural p henomenon, not a statistical frequency distribution of such elements (d eMunck and Sobo 1998). In the field of ethnomedicine, there hav e been issues in the past of insufficient sample sizes, limiting the statistical power of ethnomedical observatio nal studies of treatment efficacy, thereby decreasing validity (A nderson 1992:9). However, this sampling strategy has been changing as more anthr opologists are trained in quantitative methods and sampling techniques. Espec ially in medical anthropology, where
80 researchers face a tremendous variation in peoples attitudes and beliefs towards traditional medicine and biom edicine, there is a need for representative, random samples to be collected (Pelto and Pelto 1990). Samples are based upon cultural features. Other older et hnographic techniques such as in-depth interviews with key informants are still esse ntial because they assist in situating the quantitative data in a cultural context. For the sake of comparison with epi demiological methods, one particular method of ethnography, the focus group, will be used as an example. The focus group method is a group interviewing technique commonly used in medical anthropological research employing fi ve to fifteen participants who are not randomly selected in order to perform explor atory research in a particular cultural domain. The purpose is to develop theor ies, devise more valid measurement instruments, and to make s ense of quantitative survey results (Schensul, et al. 1999). Focus group interviews last fr om between one hour to ninety minutes and require the facilitator to guide the group interview with previously identified discussion questions. A note-taker or tape re corder assists the facilitator so the interviews can be transcribed. Focus group questions should be open-ended in order to generate discussion. For exampl e, in this study several focus groups were conducted with mothers of young chil dren to collect information on maternal knowledge and care-seeking for childhood resp iratory infections. One example of a focus group question was, What are the different types of respiratory problems your children experience? It is the facilitators job to ensure that everyone in the focus group participates that no one dominates the discussion,
81 that key questions chosen beforehand are di scussed, and that potentially harmful misinformation is corrected. Schensul, et al. (1999) suggest a number of ways that validity can be enhanced in focus group interviewing. First, they suggest pilot testing of questions to ensure comprehension. Se cond, they encourage consulting with the group participants in order to ensure a welcoming environment. Third, they advise selecting group facilit ators who are mem bers of the target group. Fourth, they suggest choosing an appropriate locati on for the focus gr oup. Fifth, they stress ensuring clarity of the research questions and resolving ambiguity in the process, and sixth, they discuss cons ulting with participants for further interpretation before the research results are published or disseminated to ensure accuracy and validity. Epidemiologists use p -values and confidence intervals to demonstrate the statistical si gnificance of their findings, which do not necessarily translate into clinical signi ficance. On the ot her hand, ethnographers establish confidence in their findings if t hey have followed a seri es of steps, that include: 1) building a familiarity with the culture through ext ensive research and living with the people there; 2) establishing the elements involved in the topic of study through extensive ethnographic met hods; and 3) recognizing that their synthesis of the research results repres ents not the actual culture itself but a model of it (deM unck and Sobo 1998). Reliability is not important for focus group research because the results of such research are not generalizable and w ill produce different results each time, even with the same facilitator (Schensul, et al. 1999). Nevertheless, focus group
82 research alone without a combination of ot her research methods is not sufficient to establish reliability and validity. Dunn and Janes (1986) argue for the combination of methods, stating that qualit ative research is powerful in its ability to achieve high validity, whereas quantitat ive research, with its emphasis upon sampling procedures and limiting observer bias, is noted for its strength to produce reliability. Such an explanati on appears plausible at first reading; however, is validity compromised in the pur suit of reliability in epidemiological studies? Is there any way that the me thodological rigor of epidemiology can be saved from rigor mortis (Nations 1986)? Anthropologists argue that in order for statistical relationships between risk factor s and disease for example, to be valid, an ethnomedical perspective is necessary (Weiss 1988). Hahn (1995) criticizes the standardized interview instrument tool, a common epidem iological method, as not following the conventions of normal conversation and preventing the chance for clarification by research s ubjects, thus obscuring the results. Consequently, the rigorous pur suit of reliable instruments may lessen the validity of the study results (Hahn 1995). This is a valid argument for incorporating an ethnographic component into an epidemio logical study, because one primary goal of health research is to generate more useful hypotheses that connect perceptions of illness to particular health outcomes (Weiss 1988). Collaborative Projects Trostle (1986:80) provides a selective history of collaborative projects between anthropology and epidemiology in the twentieth century, a history he
83 characterizes as one of benign neglect. Trostle describes the South Africa Polela project in the 1950s that used epidemiological and social science methods to design a project to improve community health. The project identified social class and poverty as important determinant s of health. Trostle identifies key studies in the 1950s and 1960s, such as Rubels work on the epidemiology of the folk illness, susto at the University of Nort h Carolina Chapel Hill, and the projects designed to study the ecology of di seases in developing countries by the Geographic Epidemiology Unit at Johns Hopkins School of Hygiene and Public Health (Trostle 1986). A more recent example of a synergist ic effort to combine methods from both disciplines was a study of infertility in Egypt (Inhorn and Buss 1994). In this study, Inhorn and Buss used a case-cont rol epidemiological design combined with an ethnographic component. In a case -control study, two groups are selected based on either having the outco me of interest or not having the outcome of interest. The researc hers emphasized that while they were interested in identifying the significant ri sk factors for infertility, their main focus was the ethnographic contextualiz ation, i.e. the persistenc e of cultural practices that diminish fertility in the Egyptian se tting. In this case, the following potential risk factors were identified: cervical el ectrocautery, male occupational exposure to noxious agents, male waterpipe smoking, and close cousin marriage practices. To ignore the cultural reasons for the per sistence of these practices is to ignore the cultural context.
84 In a later article, Inhorn (1995) writ es that there are few examples of ethnographic-epidemiological research. She cites Janes study of risk factors for hypertension among mi grant Samoans and Zunzun eguis research on male sexual practices as a risk factor for cerv ical cancer in migrant Hispanic women in southern California (Inhorn 1995). Inhor n issues a caveat, however, because she is concerned that if medical anthr opologists become overly concerned with the creation of sociocultural risk data, th is will further the undesirable trend of the medicalization of life. The concern is that the biomedical community would identify more at-risk groups for intervention based on sociological/anthr opological studies. Another attempt at explicitly combining anthropological and epidemiological methods is the Bedouin infant feeding study (Lewando-Hundt and Forman 1993). This study was a cohort study that collected data on infant feeding patterns, the social environment, health service utilization, home health care, and anthropometrics. A cohort study differs from a case-control study in that one or more cohorts is selected based on exposure to one or more risk factors and are followed over time to determine whether the cohort develops the outcome of interest. In the data analysis phase, the researchers found that some of their hypothesized quantitat ive associations were not st atistically significant. However, because of the complem ent of ethnographic information they had collected, they were able to determine how certain variables were interrelated in a multivariate regression m odel. For example, the rese archers found that as the length of time of assistance with cooking to mothers increased, the likelihood of
85 exclusive breastfeeding rather than bottl e-feeding increased. Another study examining breastfeeding also benefited by using a co mbination of methods (Nations 1986). Nations argues that only with anthropological observations is one able to understand the complex web of c ausation in disease etiology. Such observations led Nations to hypothesize a relationship between breastfeeding and the incidence of diarrheal diseases. The anthropologists found that bottlefeeding increased the risk of diarrhea due to water contamination and lack of protective chemicals cont ained in mothers milk. Integration of Methods From this brief review of the inte rface between epidemiology and medical anthropology, one may conclude that to combine both epidemiological and anthropological methods in a study design is problematic because of the different type of research questions that are posed as well as the data collected. For this reason, the dissertation research was primarily et hnographic, with an epidemiological component built into the semi-structured questionnaire and supplemented by the colle ction of epidemiological data from the health department. Trostle (1996) argues that a preliminary qualit ative, ethnographic component preceding an epidemiological study can increase validity and generalizability as well as improve access to study populations. This call for integration of methods is echoed in anot her paper where it is argued that by failing to employ preliminary qualitativ e research studies before the analytic epidemiological study runs the risk of finding independent variables that may be
86 statistically significant but are not comp rehensible from a clin ical standpoint or from an interventionist perspective (Yach 1992). Anthropological research adds the cult ural dimension that is often missed in epidemiological studies, which focu s primarily on identification of risks and disease control. As the discipline of anthropology expands and crosses disciplinary boundaries, practitioners must adapt by becoming experts in a subspecialty where anthropology informs the practice of related disciplines. In medical anthropology, there is a need for more studies, which employ anthropological methods to complement epi demiological investigations so the cultural dimension is not subsum ed by the quantitative results.
87 PART IV EPIDEMIOLOGY OF ARI In a majority of the developing world, acute respiratory infections (ARIs) constitute the major causes of death in children under five years old. These deadly ARIs are comprised of pneumonia, br onchitis, bronchiolitis, and croup. Similar to diarrheal diseases, ARIs are o ften related to malnutrition and poverty. In fact, malnutrition is associated with about half of all child deaths; moreover, based on the results of community-based studies, a dose-response relationship exists between poor anthropometric status and increasing risk of mortality from ARIs and pneumonia in children (Rice, et al. 2000). ARI deaths are primarily caused by pneumonia, usually the bacterial type, which is potentially treatable with antibiotics. Poor acce ss to biomedical care and lack of sufficient community health workers (CHW) educated to detect ear ly signs of pneumonia contribute to the problems faced in developing countries (Kelly, et al. 2001). Annually, ARIs cause approximately 1.9 million deaths in ch ildren under five years old, making it the leading cause of death in this cohor t, with over 70 percent of these deaths occurring in Africa and Asia (Williams, et al. 2002). For comparative purposes, between one and three percent of pneumonia deaths in children under five occur in developed countries compared to 10 to 25 percent in developing countries (Benguigui, et al. 1999). Ninety-nine percent of all deaths from ARI occur in young children in developing countries, and of these, 90 percent die from pneumonia, and approximately 76 percent of these deaths occur in infants (The_Wellcome_Trust 2001). If case-fata lity ratios could be reduced to those
88 observed in the developed world, almost 98 percent of ARI deaths could be prevented (Stansfield 1987). A myriad of fa ctors aid in the explanation of this discrepancy: larger family sizes, crow ding, lack of education, poor access to health care, an elevated exposure to second-hand smoke and combustion particulate matter, nutritional deficienc ies, lower rates of breastfeeding, and exposure to a variety of environm ental pollutants (Graham 2001). ARIs are appropriately referred to by Frank Shann as the forgotten pandemic (Shann 1999). Because most types of ARI resolve naturally, there is a degree of complacency over the pr oblem. Moreover, because of the inadequacy of health services in dev eloping countries, the diversity and complexity of etiological agents, and fail ed interventions, there is a prevalent fatalistic attitude towards ARI at the time when more effective ARI control programs need to be implemented in developing countries. In the developed world, ARI managem ent and classification is based upon radiographic and microbiologic data, as we ll as physical examination and clinical history (Stansfield 1987). ARI is t he general label applied to encompass both upper (URI) and lower respiratory tract infe ctions (ALRI) whic h are distinguished by a complex etiology and display divers e symptoms. One well known URI is the common cold, which has high incidence and is of great public health concern. Pneumonia is the most severe respiratory infection which results in mortality and is linked to ALRI. In th e developing world, more t han 75 percent of ARI deaths are caused by pneumonia, primarily of t he bacterial strain (PAHO/SHA 2001). While bacterial pneumonia is a major cause of childhood mortality in the
89 developing world, it is treatable wit h antibiotics and proper case management and potentially preventable with affordable vaccines and nutritional interventions. There have been serious issues raised concerning the methods that epidemiological studies have estimated the incidence of ARI episodes (Lanata, et al. 2004). A recent estimate is 165 m illion new ALRI episodes globally for children under four years ol d, with approximately 2.1 m illion ALRI-related deaths. Between 2000 and 2003, the Child Heal th Epidemiology Reference Group (CHERG) was formed by the WHO Child and Adolescent Health and Development programme in order to study the published data on ARIs to estimate the global disease burden (W HO 2004). Out of the 2000 studies surveyed, only 28 met the criteria for accurate ALRI incidence estimates based on the merits of their research de signs for prospective community-based surveillance for new episodes of ALRI. This review of the epidemiological lit erature will discuss strategies for the prevention, treatment, and cont rol of acute respiratory infections in children in developing countries. Beginning with the morbidity and mortality trends and etiology of the disease, the review moves on to risk factors, case management, antibiotic management, prevention, and appl ication of ethnographic methods to study ARIs. Morbidity and Mortality Trends Since most types of ARI are viral in origin, the worldwide incidence of ARI is similar in both developed and developi ng countries, with higher incidence in
90 urban than rural areas. Cross-country comparisons should be approached with caution because of different research me thods, study designs, case definitions, and lab culture techniques, but generally, fo r developing countries, an estimate of 4 to 7 ARI episodes per child per y ear has been proposed (Lanata and Black 2001). Incidence rates are highest for inf ants between 6 to 11 months of age. However, for specific types of AR I, the differences between developed and developing countries are striking. T he incidence of pneumonia remains 3 to 10 times higher in developing countries (T he_Wellcome_Trust 2001). Moreover, the risk of dying from measles, pertussis, and H. influenzae type b (Hib) is greater in developing countries. The reasons for this discrepancy stem from incomplete vaccination coverage and a number of ri sk factors associated with developing countries, such as malnutrition and crowding. In terms of morbidity, ARI is the primary reason for pediatric outpatient visits and outpatient antibiotic use (The_Wellcome_Trust 2001). Rough estima tes of all outpatient pediatric consultations are between 20 and 60 percent for ARI, and between 12 and 45 percent for all hospital admissions in dev eloping countries (Stansfield 1987). While incidence rates are not di ssimilar between developed and developing countries, mortality rates are 10 to 50 times greater, indicating the greater need for ARI prevention and control in devel oping countries (Stansfield 1987). ARI is most acute in young childr en for three main reasons: 1) the anatomy of children makes them more susceptible to infection given the proximity of particular organs in the head and body; 2) childrens immune systems are not developed; and 3) there is a high risk of exposure to infection
91 (The_Wellcome_Trust 2001). Other risk fact ors such as malnutrition and lack of breastfeeding also contribute to the problem of ARI in young children. In both developed and developing countries, t he incidence of ARI is inversely proportional to age, with ch ildren under five at great est risk, and within that group, infants under two months are at highest risk (The_Wellcome_Trust 2001). After the first few months of the infant s life, the risk of death from pneumonia decreases progressively as the child ages, particularly between ages two and five (The_Wellcome_Trust 2001). Estimates of ARI morbidity and mort ality are characterized by widespread under-reporting; nevertheless, the followi ng figures estimate the burden of disease. In 2000, approximately 10,891, 000 deaths occurred in young children, and of this number around 2,126,000 were due to ALRI and 47,000 to URI (The_Wellcome_Trust 2001). Ninety-nine pe rcent of these ALRI deaths occurred in developing countries. Fu rther breaking down the 2 million deaths from ALRI, the following estimates are provided: 253,000 from neonatal pneumonia; 246,000 from pertussis; 393,000 from meas les; and 105,000 from HIV/AIDS (The_Wellcome_Trust 2001). Most of the ALRI deaths were caused by either primary or secondary pneumonia. Other than pneumonia, croup and bronchiolitis compose roughly 5 percent of ALRI deaths Deaths from secondary pneumonia are high because of the severity of t he bacterial infection, the weakened immune systems caused by co-infection with other di seases such as measles, pertussis, or HIV infection, and other contributi ng risk factors such as malnutrition (The_Wellcome_Trust 2001).
92 Etiology of ARI ARI is composed of both upper and lower respiratory infections; however, accurate data concerning the bacterial and vi ral etiologies for ARI in infants and children is incomplete because of the problems encountered in making accurate clinical and microbiologic diagnoses (S tansfield 1987). Because of lack of consolidation from t he pneumonia, young children are frequently unable to produce enough sputum, so microbiol ogic diagnoses are problematic. The etiological aspects of ARIs in clude both viral and bacterial agents, and these may occur as a single or mixed in fection. While most cases of ARI are of viral origin, only a small fraction result in severe or fatal disease (The_Wellcome_Trust 2001). A longitudinal study found that among the viruses associated with ARIs, respiratory syncyt ial virus (RSV), adenoviruses, influenza viruses, and enteroviruses were the most frequent (Portes, et al. 1998). Severe ARI often develops after a primary viral infection (e.g. RSV) is succeeded by a secondary bacterial infection (The_Wellc ome_Trust 2001). Upper respiratory infections (URI) consist of coughs, colds and croup often caused by a viral agent. Examples of URIs are the common cold, si nusitis, nasopharyngitis(coryza), otitis media, and pharyngotonsillit is, none characterized by high mortality risk (Stansfield 1987). Unlike ALRI, URI do not require antibiotics. ALRI in the International Classificati on of Diseases is defined as infections affecting the airways below the epiglottis. These infections encompass laryngitis, tracheitis, bronchitis, bronchiolitis, lung infections, as well as combinations of these, or together in co mbination with URIs, includ ing influenza. The major
93 burden of disease from ALRI involves pneumonia and bronchiolitis, according to WHO, and programmatic efforts are concent rated on case management of these diseases to reduce the global burden of dis ease from ARI in children (Lanata, et al. 2004). Laryngitis (croup), tracheobronchitis, and bronchiolitis ar e examples of viral ALRIs, and only bronchiolitis, caus ed by RSV, is characterized by high mortality risk (Stansfield 1987). RSV can cause different types of ARI, including bronchiolitis, pneumonia, cr oup, and otitis media (T he_Wellcome_Trust 2001). On the other hand, in children, bronchi olitis can be caus ed by a number of viruses, such as RSV, parainfluenza, and influenza viruses (The_Wellcome_Trust 2001). Br onchiolitis is an infect ion of the bronchioles common in infants less than 18 months. Despite the threat of viral agents, ALRI deaths are primarily caused by pneumonia, which is usually bacterial and potentially preventab le with the proper treatment of antibiotics. Respiratory bacteria may be either primary pathogens, opportunistic organisms, or constituted in the normal flora of the respiratory tract (The_Wellcome_Trust 2001). Despite difficult ies in obtaining accurate etiological data, there is a consen sus that two bacteria, Streptococcus pneumoniae and Haemophilus influenzae cause most cases of bacterial pneumonia in children from developing countries, and these occu r sporadically, not in outbreaks (Graham 2001). S. pneumoniae has been found in 30 percent of hospitalized children in 60 percent of the studies undertaken in developing countries, and H. influenzae in 11 to 50 percent of cases. These bacteria may spread to other
94 parts of the body, such as the meninges of the brain, causing bacterial meningitis. H. influenzae type b (Hib) comprise 95 per cent of the strains that cause infections (The_Wellcome_Trust 2001). In developing countries, accurate AR I morbidity and mortality data are extremely difficult to obtai n. Accurate etiologically specific morbidity and mortality data are necessary to design e ffective public health policy strategies and to provide directions for vaccine devel opment. The major bar rier to accurate data is lack of reporting. However, bec ause of the etiology of ARI, proper diagnosis is difficult; therefore, other pr oblems in data collection include failing to distinguish between prim ary and secondary infection, to determine between concurrent infections, and to classify case s correctly as either ARI, URI, or ALRI, or misdiagnosis (The_Wellcome_Tr ust 2001). Moreover, the signs and symptoms of ARI are also related to other childhood conditions such as meningitis and septicaemia, and conditions such as measles or pertussis may be recorded as the cause of death instead of pneumonia. Other problems related to the reliability of ARI morta lity data are cultural differences in the recognition and diagnosis of ARI by CHWs and mothers, and differences in methodology carried out by investigators, such as the variabl e use of hospital-based data to estimate disease burden in a community, resulting in selection bias (The_Wellcome_Trust 2001). In developing countries, many young children have reservoirs of asymptomatic nasopharyngeal S. pneumoniae and H. influenzae which may lead to opportunistic infections Therefore, reliable and a ccurate samples must
95 be taken by either blood collection or the in vasive lung aspiration technique. The blood culture technique has low sensitivit y to determine bacterial etiology of pneumonia, and the lung aspiration techniq ue, while highly sensitive, is an invasive procedure that is both ex pensive and ethically objectionable (The_Wellcome_Trust 2001). Blood culture is the most accurate and widely available technique used to determine bac terial etiology of pneumonia (Berman 1991). In any specimen of sputum, blood, or even pleural fluid, there are a myriad of possible pathogens present, making isolation of a single agent difficult (Stansfield 1987). Depending on the type of diagnostic tests used, inter-study comparisons are not possible because of th e varying sensitivity and specificity of the tests. Other problems include differential distribut ion of bacterial pathogens between communities, variati on in the prevalence of risk factors, and lack of appropriate training and laboratory facilitie s necessary to carry out reliable tests (The_Wellcome_Trust 2001). Etiology of Pneumonia Pneumonia, an inflammation of the lung, is characterized by cough and rapid breathing, which can sometime s be accompanied by fever and muscle aches (CDC 2002). From recent studies on children aged two to 59 months in developing countries, 49 percent are viral only, 34 percent are bacterial only, and 17 percent are mixed infections (The_Wel lcome_Trust 2001). The primary viral causes of pneumonia and ARI are adenovirus parainfluenza virus, and RSV, the most common viral agent to cause severe ALR I. In a clinical study in poor, urban
96 Uruguay, RSV was the main agent producing annual outbreaks (Hortal, et al. 1994). The primary bacteria l causes of pneumonia are S. pneumoniae (responsible for 30 percent of pneumonia deaths), Hib, and Staph. aureus which particularly affects infant s (The_Wellcome_Trust 2001). S. pneumoniae is responsible for high mortality in children less than two years, causes around one million deaths per year, and spreads to ot her parts of the body leading to other diseases such as acute bacterial meningitis and otitis media (The_Wellcome_Trust 2001). Even though approximately 30 percent of ARI-related mortality occurs in infants under three months of age, there is scant etio logic information available on the etiology of pneumonia for this age group (Berman 1991). Lung aspirates are expensive and invasive, and consequent ly, they are not frequently done in developing countries. Blood culture is the most common and accurate form of determining the bacterial etiology of pneumonia (Berman 1991). Young infants are especially susceptible to bacterial infection because of transmission of infection from the mother at birth and more vulnerab le respiratory tracts. Because pneumonia is primarily bacterial, t here is justification for the treatment of suspected infection with antibiotics. Risk Factors Risk factors are divided into tw o categories: modifiable and nonmodifiable. Modifiable risk factors known to increase the relative risk of ARI include vitamin A deficiency, air polluti on, parental smoking, and overcrowding.
97 Non-modifiable risk factors include the age and sex of the child, genetic factors, allergy status, congenital abnormalities, and geographical factors. The risk factor that is linked to the highest case fatalit y is malnutrition, followed by low birth weight and lack of breastfeeding. These th ree risk factors are related to the risk of severe pneumonia (T he_Wellcome_Trust 2001). Modifiable Risk Factors The relationship between ARIs, malnut rition, and poverty has been well documented since the early 1990s (Lanat a and Black 2001; Ray, et al. 2001; Shann 1999). Malnutrition reduces the bodys immunity to infection, increasing the risk of ARI. Malnutrition is linked to increased severity of ARIs, as well as increased risk of bacteraemia, pleural effu sion, and other factors that lead to pneumonia. Malnutrition also leads to malaria, and consequently, in endemic countries, it is difficult to make the proper diagnosis between pneumonia and malaria (Lanata and Black 2001; Rice, et al. 2000). Compared to a normal child, a malnourished child is 19 times more likely to acquire pneumonia (Nichter 1993). Based on large studies conducted in the Philippines, malnutrition is linked to a moderate to severe risk for deve loping ARI or pneumonia (Lanata and Black 2001). In a population-based survey in nor theastern Brazil, ARI prevalence was highest in children with either acute or chronic malnutrition, after adjusting for potential confounding variables (Cunha 2000). In a hospital-based study in the Central African Republic, re searchers found that the stat us of being a child under one year old and acute malnutrition were the two risk factors most highly
98 associated with mortality risk from ALRI (Demers, et al. 2000). Based on the available evidence and data, there is strong argument for t he cause and effect relationship between mal nutrition and ARIs. In addition, particular micronutrient deficiencies are implicated as modifiable risk factors. O ne important micronutrient is vitamin A, which is found active in the body as retinoids, and the body has cells which convert these into vitamin A when needed (Whitney and Rolfes 2005). The amount of vitamin A in the body depends on availability of vitamin A stored primarily in the liver, and of protein which acts as a carrier. The prim ary roles of vitamin A are to support the immune system, and promote visi on, growth, and cell differentiation, the latter of which helps to maintain the epithelium of the respiratory and gastrointestinal tracts (Whitney and Rolfes 2005). When a deficiency occurs, these tracts are more susceptible to bacterial infecti ons (Biswas, et al. 1994; Semba 2001). Sources of vitamin A include pre-formed vitamin A, found in butter and egg yolks, and carotenoids, found in spinach, carrots, and papayas (Semba 2001). Children with severe ARI have lower seru m vitamin A levels than those with mild forms (Dudley, et al. 1997; Kkbay, et al. 1997). Moreover, children with vitamin A deficiency suffer more ARI epi sodes per year compared to those without the deficiency (Pandey and Chakra borty 1996). Over 100 million children are burdened with vitamin A deficiency worl dwide (Whitney and Rolfes 2005). Initially, because vitamin A is lost in the urine during infection, the belief was that vitamin A supplementation would help chil dren recover from AR Is. A number of placebo controlled trials tested this hypothes is (Biswas, et al. 1994; Dudley, et al.
99 1997; Kkbay, et al. 1997; Pandey and Chakraborty 1996; Ramakrishnan and Martorell 1998; Reyes, et al. 2002). Studi es that demonstrated a decrease in childhood mortality due to vitamin A s upplementation reflected reductions in death due to acute gastroenteritis and meas les, rather than ARIs and malaria (Ramakrishnan and Martorell 1998). The hi gh mortality in children with vitamin A deficiency has also been linke d to protein deficiency, diarrhea, and respiratory tract infections (Biswas, et al. 1994). While poor vitamin A status has been linked to higher ARI and pneumonia morbidity and mortality, this does not necessarily mean that vitamin A supplementation is the answer. In fa ct, supplementation with vitamin A has not been found to be significant in the fight against lower respiratory infections or RSV infections in children (Semba 2001). RSV is the leading viral pathogen involved with ARI since approximately 30 per cent of children harboring the virus also have pneumonia (Lanata and Bl ack 2001). While vitamin A supplementation has been shown to reduce morbidity from pneumonia in acute complicated measles, there has been no co mparable observable effect with other lower respiratory infections (Semba 2004). Both WHO and UNICEF recommend regular vitamin A supplementation for ch ildren with measles in areas with high vitamin A deficiency or high rates of m easles (Whitney and Rolfes 2005). Vitamin A supplementation in children aged six months to five years old has demonstrated a lowering of childhood mo rtality from meas les-associated pneumonia from randomized trials but not with non-measles-associated pneumonia (The_Wellcome_Trust 2001). Vi tamin A supplementation has also
100 been shown to reduce the incidence of severe diarrhea (Ramakrishnan and Martorell 1998). However, in a case-cont rol study, vitamin A supplementation did not alter vitamin A deficiency status bec ause of frequent respir atory infections (Rahman, et al. 1996). Malnutrition is bot h a cause and an effect of ARI, as the two behave synergistically. Malnutrition is associated with a greater frequency of complications related to ARI, such as s epsis and meningitis, which contribute to increased mortality (Tupasi, et al. 1988). Vitamin D, selenium, and zinc deficiencie s are also risk factors for ARIs in children. In a pooled analysis of several clinical trials in different geographic locations, zinc supplementation led to a 41 percent decrease in childhood pneumonia (Lanata and Black 2001). Dietary s ources of zinc include shellfish, nuts, beans, beef, and chicken. Zinc deficie ncy can lead to increased infections as well as growth abnormalities and other disorders (S emba 2001). Zinc deficiency is usually the result of inadequ ate dietary intake, and the groups most at risk include pregnant and lactating wo men, infants, and young children; moreover, zinc deficiency interferes wit h vitamin A metabolism, so vitamin A deficiency may result (Whitney and Rolf es 2005). A possible result of poor nutrition on the part of the mother is low birth we ight babies, defined as less than 2500 grams. This outcome affects roughly 19 percent of infants in developing countries (The_Wellcome_Trust 2001). The risk of pneumonia mortality increases as birth weight decreases, and low birth weight is a factor in 25 percent of childhood pneumonia deat hs (The_Wellcome_Trust 2001). While in general
101 low birth weight does not increase the risk of ALRI, it may increase the frequency of severe ALRI (Demers, et al 2000; Dudley, et al. 1997). Breastfeeding is a behavioral practice that has obvious implications for the nutrition of infants as well as conferring pa ssive immunization. In order to reduce malnutrition among infants, breastfeedi ng should be continued for at least one year (Lanata and Black 2001). South Asia and Latin America are two regions where the combination of inadequate breastfeeding and malnutrition has led to high mortality from pneumonia (Lanata and Black 2001). Breastfeeding is a protective factor for ARI because of trans fer of immunity by breast milk, which contains antibacterial and antiviral elements such as secretory IgA and neutrophils. Breastfeeding reduces pneumonia mortality by a factor of two and reduces incident cases of otitis media and ALRI by 50 percent (The_Wellcome_Trust 2001). In a case -control study among urban Malaysian children, breastfeeding for at least one mo nth was found to be a protective factor for ARI [odds ratio (OR) = 0. 58] (Azizi, et al. 1995). Bo ttle-feeding can contribute to infections because of unsafe drin king water containing pathogens. A comprehensive approach that addresses ov erall nutrition, living conditions, and childcare practices is a more appropriate intervention strategy in developing countries. Risk factors associated with poverty are associated with ARIs. These include crowding, exposure to indoor air pollutants such as smoke from biomass fuels, low education levels, and poor housing (Ezzati and Kammen 2001a; Ezzati and Kammen 2001b; Hoque, et al. 1999; Rahman and Shahidullah 2001). A
102 recent study using a national sample of children in Bolivia suggests that socioeconomic factors are the most im portant links in the chain connecting maternal education and child nutritional stat us (Frost, et al. 2005). A prospective study carried out over a year in Calc utta, India found ARI incidence to be significantly worse for undernourished childr en in the lowest socioeconomic class (Biswas, et al. 1999). The finding that malnutrition and low socioeconomic status are tied to increased ARI case fatality rates, especially for infants, is a constant theme throughout the epidemiological lite rature (Tupasi, et al. 1988). Since respiratory infections are transmitted thr ough droplet nuclei in the air, crowding increases the probability of transmission. Case-control and longitudinal studies have demonstrated the effects of crowding on risks of transmission of respiratory infection (Azizi, et al. 1995; Lanata and Black 2001; Rahman and Rahman 1997). Overcrowding may lead to doubling of ARI episodes because unexposed, susceptible children are exposed to carrier s or those infected by ARI. Children who attend day care are at higher ri sk of both upper and lower respiratory infections. A case-control study of urban Malaysian children found that the presence of a sibling with a cough (OR = 3.76), a household comprising five or more people (OR = 1.52) and sleeping with three or more people (OR = 1.45) were all significant independent risk factor s for ARI (Azizi, et al. 1995). Increased contact with other children or adults due to sleeping arrangements or crowded quarters is a significant ri sk factor for ARI and pneumonia. In developing countries, the use of bi omass fuels, which includes wood, manure, or waste, is prev alent because heating oil, elec tricity, or other less air-
103 polluting methods of heating and cooki ng are scarce or too expensive. Consequently, because of poor ventilation or insulation, families utilizing biomass fuels are at greater risk of aggravating childhood resp iratory infections. A longitudinal study in Kenya found an ex posure/response relationship between the inhalation of particulate matter from biomass combustion and ARIs (Ezzati and Kammen 2001b). Various biomass combustion fuels produce high concentrations of particulates in the air, which increases the risk of pneumonia. Some studies have shown that indoor air pol lution increase the risk of ARI by two to five times (The_Wellcome_Trust 2001). In addition, passive smoking from parents increases childhood risk in a dose-response relationship to ARI, otitis media, bronchitis, and pneumonia. To compound the problem of indoor air pollution, in many developing countr ies, because of weak environmental legislation and enforcement, outdoor environ mental pollution from vehicles and industrial emissions is a contributing factor. In this study, the inhabitants have been exposed to volcanic ash for roughly five years (Tobin and Whit eford 2002; Whitef ord and Tobin 2001; Whiteford and Tobin 2002). Many studies of the health affects of volcanic ash do not implicate its role in the incidence of ARIs but hypothesize a relationship with chronic bronchitis and silicosis. Neverthel ess, people experiencing health threats from volcanic ash, or living in tempor ary shelters because of evacuation from a natural disaster, have a great er relative risk of developi ng respiratory infections and other infectious diseases (Noj i 1997). Results from the national epidemiological surveillance system in Nic aragua found significantly higher visits
104 to healthcare facilities for acute diarrhea and acute respiratory infections in the affected communities followin g the eruption of the Cerr o Negro Volcano in 1992 (Malilay, et al. 1996). In one of the communi ties, rates for consultations for ARIs surged from 17.4 per 1000 before the eruption to 83.6 per 1000 after the eruption. This amounted to a relative risk of 6.0 for this community, meaning that the risk was six times greater for child ren under five to seek a medical consultation for ARI after the eruption. The researchers in this study found especially high rates of ARI among in fants and attributed the finding to a combination of risk factors, including vo lcanic ash and crowding in the shelters (Malilay, et al. 1996). Two areas of epidemiological investi gations related to volcanic ash exposure are air pollution and workplace exposure to crystalline silica. Studies have shown that short-term exposure to urban air pollution is associated with higher prevalence of sympt oms related to respirat ory disorders such as bronchitis and asthma (Moor e, et al. 2002). Exposure to fine particulates in urban environments has been shown to be a ssociated with the risk of developing chronic bronchitis (Searl, et al. 2002). Silicosis, a progressive, irreversible scarring of the lungs that may lead to l ung cancer, is an outcome associated with long-term exposure to crystalline silica. In vestigations of crystalline silica in high concentrations of volcanic ash have been conducted on the eruption of the Soufriere Hills Volcano in Montserrat, ac tive since 1995 (Baxter, et al. 1999; Moore, et al. 2002; Searl, et al. 2002; W ilson, et al. 2000). In the 1991 Mount Pinatubo eruption in the Philippines and the Mount St. Helens eruption, the
105 volcanic ash contained between three to seven percent crystalline silica, which could prove to be a risk factor for silicosis and an occupational hazard for loggers or farmers if exposure were maintained through frequent eruptions (Baxter 1997; Dollberg, et al. 1986). The particular dynamics of the S oufriere Hills Volcano have produced PM10 ( 10 m in diameter) volc anic ash created through pyroclastic flows of lava dome collapses with concentrations between 10 to 24 percent on a continual basis (Baxter, et al. 1999). Explosive eruptions produce much lower percentages of crystalline silica. The second factor, concerning the health effects of volcanic ash depending on the size of the ash particles, has been described. Research into the health effects has found that fine-grained volcanic ash particles, even at low concentrations, have had adverse consequences for respiratory health in the case of the 1992 Mount Spurr eruption in Anchorage, Alaska. In contrast, the coarser particles (>15 m) produced by the 1985 86 Mount Sakurajima eruption produced only ocular problems (Moore, et al. 2002). Ash particles may cause corneal abrasions and irritations, but t hese effects are generally minor and can be prevented with protective eyewear (Baxter 1997). The third and fourth factors that de termine the health effects of volcanic ash are evident in the case of the Soufrier e Hills Volcano. The third factor relates to exposure because of climatological fa ctors, and the fourth concerns the frequency of eruptions. The long-term risk for lung disease in the Soufriere Hills Volcano case is considered high since the eruptive cycle is active, especially for people working outdoors and ch ildren playing outside. Baxter (1999) makes the
106 observation that human activity determi nes exposure levels. For example, people working in agricultural activities stir up ash while performing necessary tasks. This type of knowledge pr ompted authorities to dampen school playgrounds in Montserrat after findings t hat children exposed to the volcanic ash had a higher prevalence of wheeze sympt oms (Searl, et al. 2002). People who are involved in agricultural or other work activities that agitate ash areas are at higher risk for respiratory problems. Mo reover, those who stay indoors in wellinsulated homes are at lower risk for breathing ash than people who are otherwise exposed. For this reason, even in the absence of epidemiological data, since poor housing is a characteri stic of people in developing countries, these people are at higher risk for inha ling potentially harmful ash particles (Baxter 1997). Non-modifiable Risk Factors Non-modifiable risk factors for AR I are necessary to understand and consider, but interventions cannot be planned around them. Age and sex are risk factors for ARI. As mentioned pr eviously, children under five, especially infants, are at greatest ri sk. Moreover, boys appear to account for more of the disease burden of ARI and pneum onia than girls. However, this is most likely a result of reporting bias, since in some cu ltures, more attention is given to the health of young boys than girls, so more cases are reported (The_Wellcome_Trust 2001). Both atopic allergies and asthma lead to an increased risk of childhood ARI. Moreov er, congenital abnormalities such as
107 cystic fibrosis and heart conditions are implicated in childhood ARI (The_Wellcome_Trust 2001). Genetic abnor malities such as homozygous sickle cell disease leave individuals prone to pneum ococcal infections; therefore, such persons should receive an early vaccination against S. pneumoniae (The_Wellcome_Trust 2001). Another biol ogical risk factor not frequently reported for ALRI is birth in a cold season, found to be statistically significant in a cohort study in Chile ( Lpez Bravo, et al. 1997). Environmental and geographical factor s such as seasonal patterns have an effect on ARI incidence. RSV and bac terial epidemics are more frequent in the colder months in tem perate climates and during wet seasons in tropical climates. During cold te mperate weather, infants ma y experience chilling which increases the risk of ARI, and during wet we ather, the protective barrier of the respiratory system becomes less effect ive, likewise predisposing to ARI (The_Wellcome_Trust 2001). A case-contro l study in rural Bangladesh studying environmental risk factors found a signifi cant association between sources of drinking water and deaths due to ARI afte r controlling for confounding variables (Hoque, et al. 1999). For this research because the study site is near the equator, the environmental conditions for ARI are present year-round. Finally, quality and access to health services varies between geographical regions depending on geographic location, terrain, population density, and whether the region is urban or rural. This review of risk factors for ARI ill ustrates why ARI incidence is so much greater in developing rather than developed countries. Modern care, in the form
108 of antibiotic case management and prevent ive vaccines, can accelerate the process of lessening the burden of diseas e. However, other factors are also important. One study by Caldwell, which examined the correlation between infant mortality and a number of risk fact ors in 99 developing countries, found the highest correlation with female literacy, and concluded that t he most dramatic effect on reducing infant mortality would be found by increasing access and affordability of basic health services to regions which had previously been wanting (Douglas 1990). Female literacy was also found to be a significant factor in correct attitudes and practices towards ARI in a study in India (Khan, et al. 1995). A review by Douglas made general recommendations for interventions in two parts of the tripartite epidemio logical equation, namely the host and environment, which included better nutrition and hygiene, access to medical care combined with maternal education, and increased female autonomy (Douglas 1990). Great gains can be made to lessen the burden of ARI and pneumonia mortality by increasing access and effi ciency of health services, together with mothers educated to detect the signs and symptoms of ARI in order to act appropriately. Case Management In the late 1980s, the World Heal th Organization (WHO) National ARI Control Programme was launched in many developing countries. The strategy used early detection, case managem ent, and health education of CHWs and mothers to reduce childhood pneumonia and limit the misuse of antibiotics.
109 WHO guidelines recommend that CHWs cla ssify ARI by signs and symptoms in order to: 1) treat with antibiotics; or 2) treat as an outpatien t; or 3) refer for hospital admission (Stansfield 1987). An ARI control program follows four basic steps: 1) recognition by the primary car egiver that the child needs treatment; 2) clinical assessment by a physician or trained CHW; 3) ARI treatment with oral antibiotics for outpatient non-severe ca ses and with parenteral antibiotics for inpatient severe cases, admission to heal thcare facility and supportive therapy for severe ARI; and 4) follow-up (The_Wel lcome_Trust 2001). Pneumonia, if not severe, can be treated on an outpatient basis with proper antibiotics, and URIs without ear infections or streptococca l throat infections can be managed on an outpatient basis without antibiotics (B erman 1991). The results of the WHO strategy have been largely successful, but in the most remote, rural regions, effective interventions remain problematic. WHO recognizes the clinical symptoms of early pneumonia, such as rapid breathing and an indrawn chest, to be the standard upon which to gauge the incidence of critical ARI episodes, and to be the best predictor for antibiotic therapy. ALRI diagnosis is based on presenc e of a cough, as well as one of the following symptoms: rapid breathing (tachy pnea) (>60/min. in infants less than two months, >50/min. in infants 2-11 m onths, and >40/min. in children older than one year), crepitations, wheezing, stridor and chest indrawing (Lanata and Black 2001). Chest indrawing, inability to drink, nasal flaring, and cyanosis are signs of more severe ALRI (Lanata and Black 2001). In infants less than two months of age, fever or low body temperature (seps is and meningitis) are additional signs
110 used to identify illness (Berman 1991) The presence of a cough, chest indrawing, and an increased respiratory rate is 70 percent sensitive and specific in diagnosing pneumonia (Lanata and Black 2001). In one study of 395 children in the Central African Republic who we re hospitalized according to the WHO clinical definition of pneumonia, chest i ndrawing was found to be the independent risk factor most strongly associated with mo rtality in a multivariate model (OR = 22.99; 95% CI 3.81 935.20) (Demers, et al. 2000). For all types of ARI, supportive tr eatment includes fluids, continued food intake, antipyretics, maintaining moder ate room temperatur e, and clearing of nasal and ear cavities (Stansfield 1987). Because in developing countries, half of pneumonia cases in children are bacterial treatment with antibiotics, such as parenteral penicillin because of its affo rdability and effectiveness, is suggested for both moderate and severe ARI (Be rman 1991). Depending on whether the recommended treatment is of an outpatient or inpatient variety, different combinations and types of antibiotics ar e used. A later section on antibiotic management discusses this topic in more depth. Another strategy of the WHO ARI program is to improve maternal education. In order to address educat ion in terms of childhood infectious diseases, it is necessary to reach a cu ltural understanding in order to improve childcare practices, which includes care -seeking, maternal education, and child spacing. Some of the material and econom ic barriers that prevent timely careseeking include lack of transportation, i nability to pay for drugs or medical attention, scarcity of needed antibiotic treatments, and lack of social support
111 (Gove and Pelto 1994; McNee, et al. 1995). Poor access to medical services combined with misunderstanding of the severity of pneum onia symptoms can lead mothers to delay treatment or apply re medies that are harmful, either one of which could lead to increased mortality ri sk for the child (Rashid, et al. 2001; Senz de Tejada 1997; Teka and Dagnew 1995). The work of anthropologists and medi cal scientists sensitive to the importance of cultural understanding has led to new approaches towards ARI control programs in developing countries in conjunction with WHO (Gove and Pelto 1994; Hudelson, et al. 1995; Hudelso n 1994; Kresno, et al. 1994; Mull and Mull 1994; Mull, et al. 1994; Nichter 1994; Nichter and Nichter 1994; Pelto 1996; Rashid, et al. 2001; Stewart, et al. 1994; WHO 1993). A qualitative study in rural Bangladesh found that it was important not only to educ ate mothers to detect the early signs of pneumonia, but other family members as well (Rashid, et al. 2001). Care-seeking is also an economic decision and is not always made by the mother of the child. In this same study the researchers discovered that it was important to differentiate hot illnesses fr om cold ones. A child may suffer from diarrhea, a hot illness, and a cough, a c old illness, at the same time and would consequently, be deprived of both hot and cold foods, thus exacerbating the childs nutrition deficiencies and infect ions (Rashid, et al. 2001). Education is also necessary to di scourage potentially harmful treatment practices, such as the use of kerosene for body rubs of the child observed in poor, rural Bolivia, which can cause bur ning or damage to the lungs (Hudelson, et al. 1995). In a study among women tr aders in urban Ghana, mothers were
112 found to use castor oil or enemas of gi nger or pepper as preventive measures for ARIs (Denno, et al. 1994). Another study of mothers in a rural district of Ghana found that because of lack of finances and poor attitudes towards ARI by CHWs, inadequate treatment was prov ided for mild and severe forms of ARI, which was compounded by withholding of food during disease episodes (Amofah, et al. 1995). While perceptions of severe forms of ARI were high, community members found difficulty in di stinguishing severe signs for children less than two months old (Amofah, et al. 1998). Si milar findings were reported in an urban Ghanaian population, where poor mater nal understanding of the signs and symptoms of ARI were common. For inst ance, mothers interviewed reported that they would delay treatment if they observed chest retrac tion, rapid breathing, and cough and fever, all which could be symptom s of severe respiratory problems (Demers, et al. 2000). Community-based interventions whic h include education on detecting the signs of pneumonia and case management instruction for CHWs have proven effective in preventing severe ARI-rela ted diseases and deaths (Fagbule, et al. 1994; Pandey, et al. 1989). In one study, a practical recommendation relating to the detection of rapid breathing by CHWs and thus better case recognition was increasing access to watches or timers in order to facilitate this task (Kambarami, et al. 1996). In rural In dia, a community-based intervention trial with a casecontrol design using mass education about childhood pneumonia and casemanagement of pneumonia with antibiotic use as interventions, the case-fatality rate among 612 cases was 0.8 percent comp ared to 13.5 percent in the control
113 area (Bang, et al. 1990). After one year pneumonia-specific childhood mortality, infant mortality, and mortalit y for children under five were all significantly lower in the intervention group. In another case in rural northern Pakistan, a communitybased case management program was im plemented, and the result was a reduction in total and ALRI-related mortalit y (Khan, et al. 1990). An example of a successful application of the WHO National ARI Control Programme occurred in Pakistan where the use of antibiotics was cu t in half, as was the case fatality rate of children admitted to healthcare fac ilities with signs and symptoms of ARI (Qazi, et al. 1996). Another pilot pr oject in rural Nepal demonstrated the effectiveness of active health service outreach programs focusing on education of the community and CHWs in order to reduce ARI-related childhood mortality (Pandey, et al. 1989). In a later study in t he same region, a carry-over effect was observed where prevention of pneumonia deaths thr ough proper antibiotic use indirectly affected mortality from meas les and diarrhea (Pandey, et al. 1991). Ethnographic data on local cult ural understandings of ARI as well as practices related to care and treatment are useful in the design and implementation of ARI control programs (Awedoba 1996; Stewart, et al. 1994). Meta-analysis of Intervention Trials on Case Management Sazawal and Black (1992) perform ed a meta-analysis of six published intervention trials on ARI case managemen t. The researchers controlled for the effects of immunization or diarrhea treatment, and found a pooled relative risk estimate that amounted to a 35 percent reduction in overall ALRI-specific
114 mortality. From their ca lculations on pooled rate differences, the researchers reported a 20 percent reducti on in infant mortality and a 35 percent reduction in mortality among children aged one to four (Sazawal and Black 1992). A major finding was that in areas with an infant mortality rate of at least 90/1000, a pneumonia case management approach had a si gnificant effect on lowering both infant and under-five mortality rates (Sazawal and Black 1992). This metaanalysis provides strong evidence that a pneumonia case management approach should be considered for any major intervention concerning child survival and ARI control. The WHO ARI Control Programme has been shown to be capable of reducing ARI mortality by 25 to 30 percent (Sazawal and Black 1992). However, because of increased antibiotic resistanc e among pneumococcal isolates, lack of antibiotics in some rural areas, and inappropriate health care-seeking behavior, the current emphasis has shifted to vaccination for pneumococcal diseases (Lucero, et al. 2003). Nevertheless, pneumonia case management remains an integral component of any ARI control program. Antibiotic Management A final issue related to education deals with the proper use of antibiotics to treat ARIs. According to the Integrat ed Management of Childhoo d Illness (IMCI) guidelines, antibiotics are recommended for pneumonia, acute otitis media, and mastoiditis, but not for ch ronic otitis media and any fo rm of pharyngitis besides the bacterial form (The_Wellcome_Trust 2001) IMCI is a fairly recent WHO-
115 UNICEF initiative whose aim is to im prove childhood illness prevention through nutrition, immunization, maternal health education, and vitamin A supplementation (Nicoll 2000). Antibioti cs prescribed to patients for ARI depend on whether the patient is receiving inpati ent or outpatient ther apy. For outpatient antibiotic therapy, patients receive cotrim oxazole, amoxicillin ampicillin, or procaine penicillin (Berman 1991; The_Wellcome_Trust 2001). WHO recommends treatment for five days. Fo r inpatient antibiotic therapy, especially when the mortality risk is high, patients receive broad-spectrum antibiotics. WHO recommends chloramphenicol for severe pneumonia cases and for children suffering malnutrition (Berman 1991). For infants less than two months of age with pneumonia, the recommendation is for treatment with benzyl penicillin plus gentamicin, and for other children, treatment with ben zyl penicillin is recommended (Berman 1991). Additio nal treatments include oxygen, bronchodilators, steam humidification, fluids, and other supportive care. There are primarily three main probl ems regarding antibiotic use: 1) use for upper-respiratory problems such as coughs or colds; 2) adherence problems to the full prescribed regim en; and 3) using leftover me dicines for treatment of future episodes (Gove and Pelto 1994; Sim on, et al. 1996). This is an area where proper case management trai ning for CHWs, as well as home management, is crucial. Misuse of antibio tics may lead to bacterial resistance to anti-infective drugs. This is a problem for ARI case management, which relies on these drugs. In addition, S. pneumoniae and H. influenzae are characterized by showing the highest resistance to common antibiotics, such as penicillin, co-
116 trimoxazole, chloroamphenicol, and eryt hromycin. Drug resistance occurs rapidly and renders previously used dr ugs ineffective. Nevertheless, antimicrobial resistance varies great ly among regions, independent of economic status, because some population groups may have higher nasopharyngeal carriage of bacteria that demonstrate resist ance to antibiotics, or inappropriate use of antibiotics allows the bodys nasopharyngeal bacteria to develop resistance (The_Wellcome_Trust 2001). One recent study of private pharmacy staff in Hanoi, Vietnam demonstrated that inappropriate dispensing of antibiotics for mild cases of ARI was a common practice (Chuc, et al. 2001). Another study based on 29 healthcare facilities in Rio de Janeir o, Brazil found that antibiotics were prescribed incorrectly in 8.9 percent of ARI cases (Cunha 2002). A study of physicians in Havana, Cuba provided evid ence that a refresher training program may rapidly reduce inappropriate prescr ibing of antibiotics, but that public education efforts alone are not sufficient (G onzlez Ochoa, et al. 1996). Better training, supervision and organization of se rvices are all necessary for improving antibiotic case management (Iqbal, et al. 1997). Prevention Immunization is the primary preventive measure of ARI control programs. Successful immunization progr ams decrease the incidence of ARI; therefore, the burden on health services is lessened and inappropriate use of antibiotics is curtailed. Immunization programs in developing countries target measles,
117 pertussis, and diptheria in order to prevent transmission of ARI (The_Wellcome_Trust 2001). In developi ng countries, young children are at increased risk for ARIs because vaccination coverage is limited, and clean water and sanitation is frequently not available. In contrast to developed countries, where the goal is to eliminate these child hood diseases, in developing countries, the goal is to control morbidity and mo rtality rates to acceptable levels. H. influenzae type b (Hib) conjugate vaccines were introduced into vaccination programs in the 1990s and hav e caused significant drops in the incidence of Hib diseases in developed countries (The_Wellcome_Trust 2001). While Hib is only responsible for a sma ll amount of pneumonia, it is 21 percent protective against broadly defined pneumon ia, thus making the case for the Hib vaccine as another protective measure (Lanata and Black 2001). In addition, the Hib vaccine prevents bacterial mening itis and significantly reduces the nasopharyngeal carriage of Hib (The_Wellcom e_Trust 2001). According to the Centers for Disease Control and Prevent ion (CDC), conjugate vaccines for Hib and S. pneumoniae (licensed in the U.S. in 2000) provide the greatest hope of reducing pneumonia morbidity and mortality in the developing world, yet such hopes remain unrealized (CDC 2002). Use of the Hib conjugate vaccine combined with the conjugated S. pneumoniae vaccine is currently too expensive to implement in many of these count ries (Lanata and Black 2001). Today, only some countries in Latin America, Sout h Africa, and The Gambia are using the newly available Hib vaccine. Some of t he other barriers to it s use besides cost include: absence of data on t he burden of Hib disease; la ck of proper etiological
118 diagnoses of pneumonia for surveill ance purposes; and introducing another vaccine into the standard chil dhood immunization schedule (The_Wellcome_Trust 2001). The older pneumococcal polysaccharide vaccine has proven effective in preventing pneumococcal disease in ch ildren older than two years of age and adults, but the challenge has been to develop a vaccine effective for young children and infants (The_Wellcome_Trust 2001). The pneumoc occal conjugate vaccines only potentially prevent particu lar serotypes; however, research has found that ten serotypes comprise appr oximately 80 percent of pneumococcal disease in the world, so the future devel opment of a single vaccine of nine to 11 serotypes is a plausible solution (The_W ellcome_Trust 2001). Such a candidate has apparently been found in the new pneumococcal vaccine called Prevnar, approved by the U.S. Food and Drug Administration in February 2000. Prevnar was found to be 90 percent effective agai nst all strains of the pneumococcus bacteria in a Kaiser Permanente study of 38,000 infants in northern California (childrensvaccine.org 2004). Despite the development of the pneumococcal conjugate vaccine, its potentia l utility in the developi ng world remains unknown because of issues relating to affordability, supply, and route of immunization in early infancy (The_Wellcome_Trust 2001). The applicat ion of ethnographic methods to the study of ac ute respiratory infections is the subject of the remainder of this chapter.
119 Ethnographic Methods for the Study of ARI A comprehensive collection of studies of health-related research methods from anthropology and epidemiology, as appl ied to the study of ARIs in young children, was published in a special edition of Medical Anthropology (Nichter 1994). These ethnographic studies of ARI looked at program needs such as: mothers recognition of signs and symptoms of pneumonia, ARI home management practices, and health care-seeking behavior. Current research recognizes the clinical symptoms of early pneumonia to be the standard upon which to gauge the incidence of critical ARI episodes. Further anthropological studies which use highly focused methodological tools to answer specific research hypotheses are needed to examine this worldwide problem. The most effect ive method to elucidate cultural understanding of ethnomedical terms and practices is through ethnographic interviews, which might include focus groups or key informant interviewing (Hudelson, et al. 1995; McNee, et al. 1995; Nichter 1994; Oyejide and Oke 1995; Rashid, et al. 2001; Stewart, et al. 1994). This methodological turn represents an improvement from earlier Knowledge, Attitudes, and Practices (KAP) surveys that were more limited in the type and scope of information collected. For example, in Bangladesh and Pakist an, an indrawn chest was seen as a health concern, but rapid breathing was not. In contrast, in Indonesia, the opposite occurred, with rapid breathing view ed as a cause for alarm, but not an indrawn chest. In other places like India and the Philippines, neither sign was regarded as a health concern (Nichter 1994). The focused ethnographic study
120 (FES) was developed at WHO in or der to understand peoples cultural explanatory models of ARIs in young ch ildren and to model health care-seeking behavior (WHO 1993). Study results of the FES methodology illustrate the merits of this ethnographic approach to studying childhood diseases, and the value of social/cultural factors for t he functioning of successful health care program interventions (Pelto and Pelto 1996). The FES method relies on the triangulat ion of qualitative methods and small sample quantitative techniques. T he focus of FES is to answer National ARI Control Program questions that incl ude the following: Do mothers recognize fast breathing and chest indrawing? If so, w hat terms do they use to describe it? What are the sequence and timing of care-seeking for a child with ARI? How do they vary with the perceived severity of illness and age of a child? (Gove and Pelto 1994:411). The ethnographer emplo ys collection of illness episodes, freelisting exercises, viewing of illness epi sode videos, and structured sorting tasks. The basic ethnographic methods used for th is type of study are described in two texts by Susan Weller and Russ Bernard (Bernard 1994; Weller 1998). The varied sorts of data collected through the di fferent techniques allows for crosscomparisons of information concer ning illness behavior and terminology. For an infectious disease like ARI, Nicht er (1994) makes the case for why medical anthropological methods are necessa ry to tell the whole story. Before the FES protocol, most of the emphasis on studying ARI related to identifying symptoms, classifying illne ss, and determining causes. However, other factors need to be understood and explained, such as access to resources and health
121 care, treatment expectations, understandings of prognosis, cultural variation in illness determination, and poli cy implications of findings In both epidemiology and anthropology, budget concerns limit the success of policy interventions proposed by the research findings. In the final analysis, what distinguishes medical anthropology from epidemiology re garding peoples health behavior is its ability to model decision making proce sses in a particular cultural context. Future improvements to decrease morb idity and mortality from childhood ARIs require a worldwide commitment and inter-institutional collaboration. Agencies and institutions such as the World Health Organization, the International Union against Tuberculo sis and Lung Disease, UNICEF, research institutions and universities, national government agencies involved in foreign aid, and professional lung health associations must coll aborate to improve health care in developing countries to addre ss the problem of ARI (Miller 1999). For example, the Integrated M anagement of Childhood Illne ss (IMCI) (developed by WHO and UNICEF) works on improving case management skills of CHWs, strengthening national health systems fo r effective management, and improving family and community practice in the detection of signs and symptoms of ARIs (Kalter, et al. 2003; Nicoll 200 0). This initiative is a switch from vertical child health programs of the past towards an integrated approach to the management of childhood illness. WHO is advocati ng IMCI and seeking additional partner agencies to advance this strategy (Miller 1999). The WHO Programme for the Control of Acute Respiratory Infections has recognized that ethnographic findings relat ed to respiratory illnesses are relevant
122 both in terms of preventi on and treatment of ARIs. Through the development of the focused ethnographic study, data we re collected concerning cultural practices towards care-seeking behavior. Since one of the tasks of cultural understandings of treatments for ARIs involves asking people how they use antibiotics or other over-t he-counter medicines, it is the hope that through proper education on the use of such medicines, fu rther resistant strains of bacteria can be avoided, and education materials devel oped will be practically relevant and locally specific towards ARI prevention and control. Capacity building and health sector re form are two processes currently being made by governments of many developi ng countries in order to strengthen their health care systems, provide better curative and preventive services, and offering more comprehensive and a ccurate health surveillance systems (The_Wellcome_Trust 2001). Reducing indoor air pollution through the introduction of cleaner indoor energy sources is one area where government interventions may be targeted. While the last twenty years has s een remarkable achievements in the control of ARIs, including better surve illance, new vaccines, and refinement of case management guidelines, there are st ill gaps in collaborative research and the mobilization of resources that need to be realized in order to effect significant changes in ARI morbidity and mortality in the developing world (Miller 1999). Moreover, there are both methodological and quality issues in epidemiological studies of ARIs in developing countries that need attention in order to improve
123 future research and allow meta-analysis in order to more accurately predict the global burden of the disease (Lanata, et al. 2004). This review has highlighted the inte rstices between malnutrition, poverty, and the transmission of ARIs. The ke y factors implicated are crowding, malnutrition, environmental degradation, poor housing conditions, and lack of access to health care and maternal educat ion. Primary avenues for intervention include micronutrient supplementati on, increased childhood vaccination, improving sanitation and water purificat ion systems, better education of CHWs, widening health care access, and improving caregivers ability to recognize signs and symptoms of ARIs. While rapid and reli able identification of ARI etiology is still elusive, controlled field trials hav e demonstrated that the combination of maternal education and proper case management for CHWs has had a significant impact upon ARI-related morta lity among under-fives. Epidemiological studies have contributed to the identificati on of successful intervention strategies for addressing this worldwide problem. Conclusion This review of the literature began with medical anthropology theory and narrowed the discussion to Andean ethnomedi cine in a context of medical pluralism. Next, the argument was made for the addition of an initial ethnographic component to epidemiologi cal investigations. Finally, the epidemiology of ARI was elaborated upon and examples of the use of ethnographic methods for the study of the disease were emphasized.
124 From this discussion, there are thr ee main points concerning ARIs in the developing world that call our attention. First, childhood ARI is a preventable disease that requires a worldwide commi tment of resources and people to make a difference. Strong immunization progr ams providing affordable vaccines will play a major role in the future to reduce the disease burden, especially by improving vaccination coverage of the pneum ococcal conjugate and Hib vaccine. Second, one of major challenges fo r treatment will be growing antibiotic resistance to different strains of bacter ial infections because of improper case management of respiratory illnesses gener ally. Improved case management through proper training of medical staf f and CHWs is another key component of ARI programs. Second, malnutrition has be en identified as a major risk factor for the disease; however, vitamin A supplement ation has not proven effective as it has for other childhood diseases such as measles, although the use of micronutrient supplementation has shown promise. There is no magic pill that cures severe infection, thus making the case for early detection and treatment as crucial strategies for intervention. Th ird, the ethnographic literature has shown that training peer educators or CHWs to teach caregivers to recognize the signs of ARI has met with success, strengthening the case for addressing the problem of childhood mortality fr om ARIs with low-cost educational strategies. Ethnographic research has also determined that barriers to health care-seeking need to be measured and understood in order to measure overa ll risk. While epidemiological methods allow the re searcher to gauge independent risk
125 exposures, it misses the ethnographic component of cultural and political context and ethnomedical categor ies of illness.
126 CHAPTER FIVE: METHODS This dissertation project collects qual itative and quantitative data by three main methods: focus group sessions; in-depth interviews with female caregivers and informants from the health care setti ng; and a semi-structured questionnaire. The methodology borrows from the focu sed ethnographic study (FES), but does not follow the FES protocol precisely, whic h is a rapid survey technique. Further explanation of the methods used is the focus of this chapter. CEBYCAM-CES served as the base for the initial research. The researcher presented a letter of introduction to Padre Jaime Alvarez, the director of CEBYCAM-CES, who approved the study. This formal relationship had been made possible as a result of the work of Dr. Linda Whitef ord and Dr. Graham Tobin on the USF CDMHA Grant (Whitefo rd and Tobin 2002). A copy of the letter was also submitted to Mayor Juan Salazar in Penipe town hall. Sandra Moreno, a CEBYCAM-CES sta ff member, is in charge of the adoption of families from a distance program. This progr am is funded by Italian donors. The program provides assistance to over 500 ch ildren. Part of t he requirement for a family to receive aid from this pr ogram (about $28 every two months for one child) is to attend monthly meetings. T he families receiving aid are divided into 10 separate groups of 20 to 30 people. At these meetings, the pediatrician from
127 CEBYCAM-CES would give talks about first aid, family planning, and other health topics. Each meeting lasted two hours, with the first hour spent on the catechism led by a nun from the convent. All of the meetings took place in a classroom on a separate piece of property from the CEBYCAM-CES headquarters on the opposite side of the main plaz a. The people attending t hese meetings were from various socioeconomic backgrounds, but most were on the lower-end of the scale and worked as farmers. At the end of the two hour meeting, the researcher would ask the women who had children un der five to stay for a focus group discussion, which typically last ed between a half hour and an hour. For the first two months, the researc her rented a room fr om a family near the bus terminal in Riobamba; however this arrangement became untenable, and the researcher moved to a one-bedroom apartment in the center of town. This new arrangement worked out better bec ause of proximity to restaurants, internet cafes, and a gymnasium. Moreov er, the researcher found part-time work teaching English to professionals in the evenings for five weeks in the same building where he was living as a diversion fr om the daily routine. By working for a local professional, the researcher was invited to social events such as soccer matches which provided a window into the local culture. He used this same approach in Penipe, where he attended local league soccer matches on Sundays. There was a very active lo cal soccer league in Penipe County with around 32 teams. The researcher traveled via bus to Penipe in the morning, worked all day conducting interviews and questionnaires, a nd returned late in the afternoon,
128 making valuable contacts and conducting informal interviews during the daily morning commute of 45 minutes with t eachers and healthcare workers. These conversations produced field notes that enriched the overall picture of working professionals opinions of the people of Penipe. Focus Groups Focus Through the focus group discussions, lo cal ARI terminology was collected through techniques of fr ee listing (Weller and Rom ney 1988). For this task, respondents were asked to list the terms th ey used for coughs and colds as well as all of the home treatments they used. Other topics included: what was their opinion of these home treatments; how l ong did mothers delay treatment before seeing a doctor or lay health practitioner ; what sort of prescriptions had they used; what were the most severe ch ildhood illnesses; had they had heard of pneumonia or the symptoms of severe respiratory distress such as rapid breathing; did they think t hat they had all the information they needed to treat their children; and whose task was it to pr ovide them with this information. By collecting all of this information, t he researcher was able to gather local terminology, identify new research ques tions, refine research hypotheses, and further develop the structured questionnair e. The sessions were tape-recorded and field notes were taken. The tapes were later transcribed.
129 Selection Criteria There were a total of six focus groups conducted with female caregivers about barriers to seeking a doctor, home treatments for common childhood illnesses, and local illness terms. There was an average of ten participants in the groups. For five of the groups, from CEBYCAM-CES, female participants were asked to stay after the meeting if they were in care of a child un der five years old. Each of the five groups had different participants because each of the ten adopted families groups had a name assigned by CEBYCAM-CES. The five groups chosen for the focus groups were a convenience sample of the ten groups already established by CEBYCAM-CES. Initially, the researcher attended two adopted families group meetings by lect uring on first aid and family planning when the doctor was absent, without carrying out the focus group afterwards. After five focus groups the focus groups were not producing any new information, so the remaini ng five adopted families groups were not interviewed. In one of these meetings, Sandra Moreno assisted with the tape recorder, but the researcher conducted the other four fo cus groups alone. Comparison of the groups did not reveal significant diffe rences, despite the presence of this authority figure in one of the focus groups. The sixth and final focus group was held with a womens group in Subcentro de Salud (SCS) Penipe (Health Subcenter), which held monthly meetings to discuss various issues. The researcher hypothesized that this group might produce different data than the prev ious five focus groups. However, the discussions and responses to the focus group questions paralleled the other
130 focus groups, even though the participants were not affiliated with CEBYCAMCES. This particular focus group was conducted after their regular meeting and informed consent was obtained. By chanc e, all the women in this focus group were mothers or grandmothers, so t he focus group composition was different than the others because not all participants were female caregivers of children under five years old; however, all we re able to respond to the focus group questions related to childrens health care. Participant-observation and Mother Informants Focus Concurrent with the focus group meet ings, the researcher spent time taking field notes in the CEBYCAM-CES Heal th Clinic, as well as SCS Penipe. The researcher observed how long people we re waiting for attention, how many people were in the waiting room, what we re the characteristics of the waiting room and surrounding area, what informa tion was posted on the walls, what procedures were followed between che ck-in and departure, and what was the general atmosphere. In CEBYCAM-CES, t he health staff assisted in recruiting mother informants to che ck the information collected from the focus groups. Eight in-depth interviews were conduct ed with individual mother caregivers on the same themes as the focus group discussions, while in CEBYCAM-CES. Mother informants were asked to list the terms they used to describe respiratory problems, what home remedies they used, how long they waited to seek care, how much the care cost, and what were the most common child hood illnesses.
131 Participants were also asked why on that particular day they had come to the clinic and were asked to describe their childs symptoms. Selection Criteria Because this part of the research was conducted in the beginning of the field study, all informants were select ed in CEBYCAM-CES where the researcher had the most rapport wit h the health clinic staff. Only mothers accompanied by their children under five years old were ask ed to participate in the interview. The mother informants were predominately se lected by the clinic staff while the mothers were waiting in the waiting r oom to participate in the study. The interviews were brief, typically taking ar ound ten minutes, while the mother would be waiting to see the nurse or doctor in the waiting room. In formed consent was obtained. In-depth Interviews from Health Care Setting Focus In addition to focus groups and in-depth interviews with mother caregivers, the researcher interviewed informants in the health care setting These in-depth informant interviews with members of the health profe ssions allowed the researcher to gauge the most pressing health problems and resource needs in Penipe County. These individuals in cluded doctors, nurses, nurse aides, pharmacists and other health sector employ ees, including traditional healers. They were asked to describe their educatio nal preparation for t heir job, their job
132 responsibilities, what skills or traini ng they needed to perform their job better, what were the most pressing health needs and problems of the community, and what resources were lacking in the health care facilities. The semi-structured interview also included the following questions (when interviewing a doctor or a nurse): 1) who provides care to child ren with ARI; 2) who is in charge of prescribing antibiotics; 3) do the health providers identify a delay in mothers seeking care for their children with AR I; 4) do the health practitioners use local terminology when dealing with mothers who bring their children in for care; 5) what do local healthcare practitioners perceive to be the mothers knowledge regarding the signs and symptoms of ARI; 6) what were the regular clinic hours and how many people worked there; 7) what were the costs to the patients; and 8) were there too many or too few patients. For all of the interviews and focu s groups, both written and computer versions of field notes were maintained. Usually, more information would later be added in the computer version of the fiel d notes. This stage of the research was conducted from June until July of 2004. Selection Criteria The informants from the health care setting were selected based on the personnel that were available to be inte rviewed. All health staff agreed to be interviewed. There were two cases w here interns in SCS Penipe were not interviewed because of their limited familia rity with the study site. Because the interview was semi-structured, new fo llow-up questions and discussions were
133 also common. A total of 25 different people in the health care setting were interviewed in four different areas of t he county. Two of these were curanderos, or lay natural health practitioners. In addition, two people were interviewed who worked for the Municipality of Penipe; one, an agronomist in charge of a joint development plan with CEBYCAM-C ES, and the other, an engineer knowledgeable of the towns potable water system. Semi-structured Questionnaire Sampling Issues The questionnaire phase of the resear ch began with a mapping exercise. With the assistance of a map provided by the town hall planning department, all 147 households in the town of Penipe and the nearby hamlet of Penicucho Bajo were identified. All residences we re surveyed and the 31 households where children under five years old were living wit h female caregivers were identified. The initial plan was to take a random sample of these 147 households for the survey. Unfortunately, because of the lo w number of households that satisfied these criteria, all 31 households had to be interviewed. The same procedure was conducted in the two other areas su rveyed (Puela and Matus), locating 30 households in each area which satisfied the inclusion criteria, for a total of 91 interviews. In each of the two areas, the inhabited households were surveyed fully for eligibility in the survey. A ll households in the other two areas were surveyed near the hamlets of Palictahua, Puel a, El Altar, Pachanillay, Matus, and Calshy. Maps were created for each community, noting the number of the
134 household where the questionn aire was administered, because the twenty-fourhour dietary recall section of the questionnaire required a return to the household after eight days in order to perform a sec ond dietary recall. While the sample is a convenience sample and not randomly selected, the sample does offer a large cross-section of households and exhausts the number of target households for the areas represented in the sample. Gi ven the financial and time constraints of the project, there was no ot her feasible alternative than the sampling method ultimately chosen to carry out the research. The twenty-four-hour recall was used pr imarily because of it ease of use and its incorporation into the semi-struc tured questionnaire. The other option was to conduct a food frequency questionn aire which would require the researcher to compile a list of all the ty pes of foods consumed as well as a list of traditional dishes for respondents to keep track of their childrens diets with a food diary. This approach wa s unfeasible because of la ck of financial incentives and the fact that many caregivers in the study were illiterate. Because of time constraints, the twenty-fou r-hour recall was the preferre d method to characterize the habitual food intake of children thr ough the collection of two twenty-four-hour recalls. The disadvantages of the twentyfour-hour recall are that it depends on the respondents memory of foods consum ed and ability to remember accurately the sizes of food portions, and it may collect inaccurate information relating to the ingredients and cooking me thods used to prepare t he foods (Gibson 1993).
135 Focus With all of this information from the previously-described research techniques, an in-depth semi-structured questionnaire was finalized to administer in three different areas. The ques tionnaires and interviews employed a vocabulary appropriate to t he study site. The language of the semi-structured survey instrument was slightly modified wi th the assistance of one of the doctors from CEBYCAM-CES to ensure that subjec ts would understand the terminology, and the questionnaire was pilot tested wit h one key informant from the focus groups to ensure comprehensibility and valid ity of responses. For example, on the questionnaire, the Spanish word for clinic was changed to doctor, because in Ecuador, health clinics are private and expensive, and the intent of the question was to ask the mother respondent under what circumstances she would seek a doctor. In addition, question s on home treatments were supplemented with knowledge gained from the focus group discussion s. These discussions aided the researcher to use local idioms and expressions in the wording of the questionnaire. Because of time and financ ial constraints, a more thorough pilot study was not possible before the adminis tration of the questionnaires. The semi-structured questionnaire was administered in Penipe County with female caregivers to obtain more data on health care-seeking behavior and childhood respiratory problem s. The administration of the questionnaires was completed in two months from August to October, 2004. The questionnaire is a detailed health record of respiratory illness in the youngest child, the female caregivers use of healthcare facilities, a detailed twenty-four-hour dietary recall
136 of the child, and an inventory of medicati ons used in the home. There are also questions about socioeconomic status, employment, education, ethnomedical terms, and symptom recognition for respir atory illnesses. The questionnaire produced 81 categorical and continuous variables, answers to 11 open-ended questions, and a twenty-four-hour dietary record. Selection Criteria The original inclusion criteria for t he questionnaire stated that the female caregiver be the primary caregiver of a child under five, had lived in the community for at least 3 y ears, had the mental capacity to answer questions for the 45 minute questionnaire, and was a Spanish speaker.2 In the town of Penipe, 31 female caregivers satisfied the inclusion criteria. Two female caregivers did not satisfy the inclusion criteria.3 The goal was to match the number of interviews in Penipe to two ot her areas of Penipe Co unty. Of the two other areas, one area was located in t he designated Area of Risk for the Tungurahua Volcano, and an adjacent ar ea was not, even though this region also was exposed to volcanic ash. T he second area spanned from the Parish of Puela to Ganzhi (in El Altar Parish), including the communities of Pungal, Palictahua, and Pachanillay. Exactly 30 fe male caregivers matched the inclusion criteria in this second area: one female caregiver chose not to participate in the 2 The early assumption that there were Qu ichua speakers living among the study population turned out to be false so this criterion was unnecessary 3 The researcher changed the inclusion criteria with USF IRB to increase the age limit from 44 to 65 because it was not uncommon to find grandmothers or older mothers as the primary caregivers.
137 study, and one did not satisfy the inclusi on criteria. The third area included mainly the Parish of Matus, with a few female caregivers interviewed in nearby Calshy and upper El Altar Pa rish. Thirty female car egivers met the inclusion criteria in the third area. One female caregiver did not meet the inclusion criteria for living at least three years in the co mmunity. Overall, 91 interviews were completed with female caregivers about t he health of their yo ungest child under five in Penipe County. Upon contacting potential research subjec ts in their homes, the researcher identified himself, explained the inclusion criteria and that all responses were confidential, and the risks and benefits of the study as outlined in the verbal informed consent agreement (see Appendix A). Participants were compensated with either a baseball cap or an eye shadow compact at the conclusion of the interview. There were no early withdr awals. On the cover sheet of the questionnaire, the subject identificati on number was recorded (no names were used), as well as the date and the place of the interview. Eight days after the initial interview, the same study partic ipant was contacted to conduct a follow-up twenty-four-hour dietary recall. The sa me subject identification number was employed on the dietary recall sheet for ease in collation. This second recall was not possible for 16 of 27 interviews in Ma tus because of the threat of aggressive dogs. Therefore the dietary informat ion for the Matus respondents had lower validity than the other tw o groups of Penipe and Puela, which combined had 43 out of 47 records with two dietary recalls collected.
138 At the end of each day, t he total number of comp leted questionnaires was calculated and the map was upda ted. In addition, the co mpleted questionnaires were checked for accuracy. A calendar served to schedule the second dietary recall visit. According to convention, it was important not to perform the second dietary recall on the same day of t he week as the initial questionnaire. Additional Research After completing the questionnaires, fu rther ethnographic research was conducted in the two health centers in the town of Penipe. The researcher also carried out newspaper archival research and collected epidem iological data from the Ministry of Health in Riobamba. T he information collected from the Ministry of Health included population projections for the counties of Penipe and Guano, as well as general morbidity data from 2002 to 2004. The morbidity data were problematic because for 2002 and 2004 the data covered the months from January to June, but for 2003 the data covered January through August, making meaningful comparison difficult. Mor eover, the population estimates were unreliable, thus computing morbidity ra tes would be problematic. Finally, impromptu interviews with heal thcare workers and other persons of interest were sometimes possible during the morning commute to Penipe. Data Analysis Qualitative data analysis was iterative and depended on careful interpretation of observational and intervie w data. Transcriptions from the focus
139 groups were entered into Microsoft Word and grouped into thematic categories for analysis. Simple frequencies of re sponses were tabulated for analytic purposes. Similarly, in-depth intervie ws with healthcare workers and mother informants were transcribed from handwritten field notes into Microsoft Word and analyzed for content. Quantit ative data from the semi-structured questionnaires were entered into a Microsoft Access database, and imported into SPSS Release 13.0 software for statistical analysis (SPSS 2005). Logistic regression procedures were conducted with SAS Rele ase 8.2 statistical software (SAS Institute 2001). Dietary information from the twenty-four-hour food recall was entered using First DataBank software Nutr itionist Pro Version 1.2.207 (First DataBank 2002). After the data were entered, the dietary information was exported to Microsoft Excel and import ed into SPSS for statistical analysis. For the quantitative data from the semi-structured questionnaire the key index variable created from other vari ables is referred to as SES score or possession score. There were 9 vari ables which when combined created a possession index. This created a ten-point (0-9) scale. Those who scored 5 or above ranked high on the SES score and thos e who scored below 5 were ranked as low on the index. It was not possibl e to calculate a reliability coefficient because after the questionnaires were comp leted, neighbors were not asked to identify households of greater economic m eans to verify the results on the scale; however, these scales are very similar ac ross rural regions in the world (Weller 1998; Weller, et al. 1997). Based on t he Weller, et al. (1997) study in Guatemala, the SES items included: sala ry above $150 per month, more than
140 three rooms in the house, a non-dirt floor, wall of blocks or bricks as opposed to wood, land, television, bicycle, and appl iance ownership, and ownership of more than three agricultural animals.4 One check on the valid ity of the possession score was to crosstabulate ownership status of land for agriculture with possession score and salary above or below t he mean. The results found that of the 10 families that rented land, signify ing low socioeconomic status, only one family earned above $150 per month. Sim ilarly, seven families ranked as low on the possession score index and three ranked hi gh. In contrast, of the 18 families that tilled land that they owned, five ranked low on the possession score index, and 13 families ranked high on the inde x, suggesting an association between land ownership and high socioeconomic status. Two demographic continuous variables were recoded into categorical variables for analytical purposes: age of respondent and monthly salary. Four key qualitative variables were coded and conv erted into categorical variables for analysis: hot and cold cough knowledge, us e of curanderos, barriers to seeking a doctor, and attitudes towards healthcare facilities. The initial stage of data analysis entailed calculating frequencies and descriptive statistics for the quantitat ive variables. In the next step, crosstabulations were performed bet ween self-reported childhood illness categories and demographic variables to perfo rm chi-square tests of association. The same procedure was carried out to co rrelate hypothetical treatment seeking strategies for specific childhood ill ness symptoms with de mographic data. 4 Weller et al. (1997) used an 8 item scale. Television ownership was added as a possession variable in this study.
141 Responses were compared between the two groups of questions listing illness symptoms, i.e. treatment c hoice and level of severity, to determine whether there was a consistency of responses based on t he chi-square test. In other words, do respondents who answer that cough, phlegm and nasal secretion is a case for the doctor, also rank the symptom as s erious to emergency? In addition, a score was generated to quantify ARI kno wledge based on correct responses to the question on perceived level of severity for the symptoms of rapid breathing and indrawn chest as well as the combi nation of the two (s ee Appendix A). An answer was scored as 0 or 1, for inco rrect or correct, respectively, for a maximum score of 3. The same pr ocedure was used to generate an overall score for all nine symptom clusters to generate a knowledge score of respiratory illness. After determining where the biva riate statistical associations were located, a binomial logist ic regression procedure was used, employing the forward stepwise method, to gauge the predi ctive effect of demographic and SES independent variables on the dependent variable of treatment choice for each of the nine symptom clusters (Kinnear and Gray 2004:396). The nutritional data were tested for normality and descriptive statistics were calculated to determine mean, m edian, range, and standard deviation. These resultant macroand micronutri ent values were compared with RDA nutrition goals according to age group, and nonparametric Kruskal-Wallis tests, similar to one-way (between subjects) AN OVA, were conducted to compare the means of these macroand micronutrient values between the three age groups to determine if there were significant differences between the means. Next,
142 Mann-Whitney tests, a nonparametric version of the independent samples t test, were carried out to compare the means of the macroand micronutrient data stratified by gender and income to determine if there were significant differences between the means. The combination of quantitative and qual itative methods produced a rich body of data that could be analyzed to test the research hypotheses of the study. In addition to the primary data that we re collected and analyzed, secondary data from the Chimborazo Minist ry of Health was collect ed and analyzed to calculate outpatient consultation rates for ARIs in under-fives in Penipe County in order to compare with previously collected data by Whiteford and Tobin (2002). The triangulation of methods increased the validity of the findings. Limitations This final section delineates some of the limitations of the research methods employed. In five of the fo cus groups, participants were chosen based on an opportunistic sample of the partici pants in the health educational sessions held by CEBYCAM-CES. Therefore, the participants who stayed for the focus group discussions may have been biased towards CEBYCAM-CES, since they were receiving monetary support from them for their children. This may have produced bias in responses related to acce ss to health resources. However, the participants in the focus groups understood t hat the researcher did not work for CEBYCAM-CES, and were very frank in their responses. The presence of a CEBYCAM-CES staff worker in one of t he focus groups and of a nurse in the
143 final focus group at SCS Penipe may have inhibited responses because of the presence of an authority figur e. However, upon analysis of the focus group data, the responses of these two groups did not contrast with the other four focus groups. Another limitation was logistical. Si nce the researcher did not have an assistant for the focus groups, in order to obtain a quality recording of the participants, he had to circulat e the room with the tape reco rder. In the first focus group, the tape recorder was placed in the c enter of the circle of chairs, resulting in a poor quality recording. Therefor e, even though written notes were taken during and after the focus gr oup, since a tape recorder was not used, extensive note-taking during the focus group was unfeas ible. However, by transcribing the focus group session on the same day, the researcher was able to complement note-taking with the literal tr anslation of the focus groups. Since the in-depth interviews with t he mother informants were frequently short, the amount of information gleaned from these interviews was not as extensive as the focus groups. However, since some of these interviews were conducted in the week before the focus groups, some questions emerged that had not occurred to the researcher in t he initial formulation of the focus group questions, and were henceforth incorporat ed to the list of questions. Another limitation was that all mother informants were from one heal th clinic. More varied responses may have been collected if t he mother informant s had been expanded to other health clinics in the county. Finally, because the interviews were
144 conducted in the public waiting room space, the range of responses may have been limited. The in-depth interviews proceeded without incident. The primary limitation was the time constraint placed on the healthcare workers to take the time necessary to complete the interview. However, this was not problematic because in general, the healthcare workers were interviewed during slow times or times of their choosing. Because the interviews were conducted in offices or areas away from other peopl e, the healthcare workers were very forthcoming with responses to the researchers questions. The specific challenges of the semi-structured questionnaire are presented in Appendix D. More gener ally, the questionnaire was primarily composed of yes/no responses, so t here was the potential drawback that respondents might fall into a rhythm of answe ring all yes or all no. There was one section of the questionnaire that present ed a moderate level of difficulty to respondents, who had probably never answe red a questionnaire before. The question was, of the following list of sympt oms, which type of provider you would seek first and why. Many answered for exam ple, first I treat at the home, and if the illness did not improve, I would seek a doctor. This was problematic because the researcher had to choos e one answer for each symptom. The second set of questions asked the respondents to rate the seve rity of the same symptoms on a Likert scale. By comparing the responses to these two sets of questions, the aim was to determine whet her for the symptoms the respondents rated as serious or emergency in the second set of questions, they chose
145 doctor in the first set of questions and likewise for symptoms rated as moderate or not serious in the sec ond set, chose home remedy or nothing in the first set of questions. Responses correlated for these two related sets of questions for the symptom clusters of c oughs, fevers, and phl egm but not for the symptom clusters of ARIs, which should have been classified as serious to emergency and treatment choice of doctor. This may be explained by the generalized lack of recognition of the se riousness of ARI symptomology, which is discussed later in the results and analysis c hapters. A complete pilot study of the questionnaire may have led to the developm ent of a different format for asking disease severity and help seeking questions Therefore, a gener al limitation of the questionnaire format rather than an open-ended interview was the difficulty for respondents to answer according to t he pre-determined list of responses as delineated by the researcher. A related limitation to this question was the choice of illness categories. The researcher was interested in t he two primary symptoms of ARI, chest retraction and tachypnoea, or rapid breathi ng. The list of symptoms should also have included dyspnoea, or labored breathing, anorexia, lethargy and restlessness, in combination with other symptoms, in order to more fully elicit the ethnomedical model (Denno, et al. 1994). In Dennos study, the question of treatment choice was asked directly a fter the severity of the symptom was established. In this study the treatment choice was a sked first according to the list of symptoms, and the level of severi ty second. Whether the order of the questions biased responses to the level of severity is unknown; however,
146 according to the results of the data analysi s, the responses to the first group of questions largely parallel t hose of the second group. Another limitation that the researcher had to negotiate was the case where the respondent would answer the question with her husband present. As much as possible, the researcher attempt ed to limit the input of the husband by maintaining eye contact with the female respondent and emphasi zing that there were no right or wrong answers to the ques tions. If the researcher sensed that the woman might only be answering what her husband had said, the respondent would be asked again if that was what she thought. Although there were only a few cases of husbands influencing the res ponses, the researcher sought to limit the effect of their presence as much as possible. Usually, in these cases the husband would wander off half way thr ough the questionnaire session, which typically lasted between 30 to 45 minutes. A related issue to spousal relationships was the inability of the ques tionnaire to capture who was in charge of the household finances. The ques tionnaire asked for overall household income but did not differentiate the sala ry of the husband fr om the wife nor who controlled the money and made financial deci sions related to medical care. In one case, a mother did not know how much her husband earned, which suggests that he controlled the money; however her husband was a well-paid working professional, an agronomist, and he reported his income to the researcher in order to complete the questionnaire. Finally, relating to sampling issues the 91 surveys captured approximately 28 percent of the children under five years old in Penipe County. According to
147 data from the Ministry of Health fo r Penipe County in 2004, there were 152 children under one year old and 169 chil dren between the ages of one and four, for a total of 321 children. Some of the children from the most rural areas of the county were missed as well as children from the parishes of Bayushig and Candelaria (which were not designated as areas of risk from Mount Tungurahua), which combined, accounted fo r 53 children. These parishes were not one of the three areas selected for the questionnaires because of their distance from the volcano and logistic al factors such as regular bus transportation, in the case of Candelaria. The methods used in the research were a combination of qualitative and quantitative data collection techniques. The information generated from these data collection instruments is complement ary and paints a more complete picture than would one method used exclusively. The initial plan was to informally interview more mother informants, but t hey were yielding the same information as the focus group meetings, and the focus group meetings were more successful in generating lists of terms and vocabularly. The in-depth interviews of the healthcare worker informants was an ongoing project throughout the fieldwork, and many of the respondent s were genuinely interested in the research study. The data generated was both qualitativ e and quantitative and aided the researcher in interpreting the other resear ch results, such as the epidemiological data from the Ministry of Health. Results from the quantitative (semi-structured questionnaires) and qualitative (in-depth in formant interviews and focus group
148 discussions) methods provide insights into maternal knowledge of home treatment, health care-seeking behavio r, and ethnomedica l and biomedical models for treating childhood respiratory infections.
149 CHAPTER SIX: RESULTS As detailed in the methods chapter, t he researcher collected qualitative and quantitative data by three main met hods: focus group sessions; in-depth interviews with mothers and informants fr om the health care setting; and a semistructured questionnaire admin istered to 91 female caregivers in Penipe County. Part I is a discussion of the focus gr oup sessions. The six focus groups are organized according to 13 different ma in discussion topics carried out with mother respondents. The in-depth interv iews with female caregivers were carried out concurrent with the focus group discussions and centered around the same themes. Part II summarizes the main themes from the in-depth interviews with the informants from the health care se tting, who were comprised of doctors, nurses, and other healthcare professionals Some of the major themes that emerged relate to primary health needs, peo ples opinions of health service, and health service utilization. Part III is a discussion of the epidemiological information gathered from the Chimborazo Minist ry of Health related to outpatient consultation rates for AR I, pneumonia, and malnutriti on. In Part IV, the descriptive statistics from the semi-s tructured questionnaire are presented in tabular form with the accom panying explanations. In addi tion, the results of the dietary analysis are presented in this secti on. Finally, in Part V, the qualitative
150 data from the semi-structured questionnaire is presented, which consisted of use and attitudes towards healthcare facilit ies, barriers to care, ethnomedical knowledge of coughs, use of cu randeros, and herbal remedies.
151 PART I FOCUS GROUP DISCUSSIONS The six focus group discussions carried out with mothers of children under five were broken down into 13 different t hematic categories. The focus groups were all transcribed and responses were grouped in relation to these thematic categories. Particular themes emerged from the anal ysis of the focus group discussions, and the results are presented in this section. The focus groups lasted from 30 to 45 minutes depending on the number of people in the groups. The first category dealt with childhood diseases. Childhood diseases were described as respiratory problems, health problems from the volcanic ash, stomach problems, and other health probl ems. Volcanic ash, coughs, and colds were all mentioned once at each focu s group as major health problems for children. Diarrhea was listed as a major health problem in th ree out of the six focus groups. Intestinal infections, phlegm, eye and skin problems, varicella, fever and cold weather were each mentione d in two of the six focus groups. Other health problems mentioned once were allergic rhinitis, asthma, lack of hygiene, respiratory problems, unsafe water, and general stomach disorders. Complaints about health problems caused by the volcano ran the gamut. The health effects described were respiratory, skin, and eye problems. Other issues concerned water and food contamination from the ash. One mother commented, lo que necesita para que formar los nios, una ceniza bien finita, problema del agua tambin, cosas del comer, se enferman los nios [what you need for children to grow, the very fine ash, probl ems with the water, f ood, the children get
152 sick]. Finally, focus group participants co mplained of their childrens stomach problems which included parasitosis, diarr hea, and other intestinal infections. Other health concerns included lack of vitamins, poor hygiene, and protection against chicken pox (varicella zoster). The second discussion topic focused on household remedies. For common colds and coughs, lemonades and teas with various herbs were used. Most frequently mentioned we re the herbal remedies, poleo and borraja which were used to cure coughs.5 Also frequently mentioned were chamomile, garlic and lemon as remedies. Clearly, t he section on the semi-structured questionnaire dealing with hom e treatments yielded more valuable information for this category. The third topic asked focus group members to talk about hot coughs versus cold coughs. Respondents were asked if they knew the difference between the two types, and what remedies were used to treat these illnesses. More information on this topic wa s generated in the semi-structured questionnaires. A general belief regarding the difference between the two types of coughs is illustrated by the following quote: Tos fra, cuando tiempo es muy fro, to s caliente cuando hay mucho calor por los soles [Cold cough, when the climate is cold, hot cough when it is very hot because of the sun]. 5 Borraja pertains to the family Boraginaceae and is of European and North African origin. It has attractive blue flowers with a white center and is cultivated widely. It is used as a sudorific and to calm the chest. It is also reported to help the liver when eaten in a salad. The herb poleo was not able to be identified.
153 One person said that for a hot cough, herbal remedies such as borraja and poleo were to be used, and for a cold cough one should breathe the vapor of eucalyptus leaves. Another woman commented: Tos caliente, ya ha salido al fro, tos fra, no se ha abrigado, sin chompa, poner mentol en pecho y espalda con un peridico caliente [For a hot cough, the cold has gone out, for a cold cough, you have not covered yourself up with a jacket, so put menthol on your chest and back with a hot newspaper]. On the fourth topic, t he researcher asked people to discuss what signs or symptoms they recognized in their childr en where they would seek a doctors help. Six caregivers in five of the six groups agreed that one would wait for two days and treat a cough or cold with home remedies. Then, if the illness did not resolve, they would take the child to see the doctor. One person said it was difficult to cure a child of cough and fever with home remedies. Another remarked that if the fever did not subside, they would take the child to see the doctor the next day. One mother comm ented, est con bronquitis, dice, el doctor conoce dando las medicinas [(the ch ild) has bronchitis, the doctor knows to give medicines]. The fifth topic addressed the terms t he doctors used to describe illnesses and whether they used the same termi nology as the mothers. One person commented that the doctors know about hot coughs, and children have more
154 rapid recoveries after visi ting the health clinic. Three people said that they agreed that the doctors used the same terminology, although one person complained that a doctor advised against using a particular herbal remedy that was apparently good for coughs. Both the nurses in CEBYCAM and in SCS Penipe have worked for many years in Peni pe, so they use the same language as the local people to describe illness terms. The sixth topic asked participants to lis t the obstacles to seeing a doctor. The most common obstacle to seeking care for their children was lack of money. One person mentioned having to borrow money from a neigh bor in a dire situation, suggesting the importance of social support mechanisms in the community. This respondent remarked, dinero en el campo, no hay una facilidad de dinero diario, cosecha cinco meses adelante, all tenemos dinero extra [money in the rural area, it is not easy to have money daily, the harvest is five months away, then we have extra money]. The concern about money was directed at the issue of having to pay for medications. In accord with the result s of the semi-structured questionnaires, money was the primary concern related to health care. Another obstacle was limited bus transportation for people who live in rural areas outside of Penipe, but this was only a problem in the evenings. Another obstacle wa s the limited health clinic hours, so if a child fell ill on Fri day, the family would have to wait until Sunday for medical treatment. The seventh topic asked participant s if they had seen a curandero, or natural healer. The researcher used t he term curandero, which was the common
155 term to describe these alternative healers, while some used the term fregaduras (bone-setters), who were distinguished fr om curanderos because they were not as well trained or prepared. A few people said they had never seen a curandero. Another admitted using curanderos, but added that people used curanderos more for problems related to childbirth t han for childhood illnesses. A couple of people commented that curanderos cure folk illnesses such as mal aire (bad air) and espanto (fright). When a curandero cure s an individual, it is called a cleansing, and he or she employs eggs, gui nea pig and specific herbs such as Santa Maria in practice. More information on this topic was gathered in the semistructured questionnaire and in the indepth interviews with two curanderos. The eighth topic for discussion asked participants to comment on whether they felt they had the information they needed to make informed decisions on childhood illnesses and how to act on them. The consensus of the women in the focus groups was that they did not have all the information they needed to diagnose childhood illnesses on their own. They wanted the doctors from CEBYCAM to give talks at the Development Center or in their communities so they could become better educated. The ninth topic asked participants whos e job it was to provide them with this health information. One person suggested that they needed more information about which herbal remedies were efficacious. People thought it was the job of SCS Penipe to give them ta lks on these topics, but conceded that CEBYCAM had done a good job at providing t hem with some information. For example, the pediatrician from CEBYCAM had given talks on first aid and family
156 planning to the adopted families groups. One participant remarked, personas ms preparados que nosotros porque no tenemos la capacidad para saber, capacidades, charlas en la medicina, rem edias caceres, a veces hace mal, no se sabe [people more knowledgable than us because we do not have the training to know, classes, talks in medicine, hom e remedies, sometimes do harm, there is no way to know]. The tenth topic addressed vaccinati on coverage. People reported that their children were vaccinated, although so me reported that there was no vaccine available against varicella. The elevent h topic asked participants to list their knowledge of pneumonia. Most of the women recogniz ed the term pneumonia but did not know what it was. More commonly used was the term pulmona or sickness of the lungs. The women did not recognize an indrawn chest or rapid breathing as symptoms of respiratory infe ctions. One person commented that in the case of pneumonia, you would have to br ing your child to the hospital or risk death. Another person said their child had pneumonia, and the doctor said to keep the child inside and covered up, w hereupon the child was supposedly cured with cough syrups. The twelfth topic asked responde nts about prescription medicines, especially antibiotics. Sometimes t he women would only buy half of the prescription if they did not have enough money. People most frequently mentioned Tempra, which was given for free by the government health clinics to treat fevers. This was confirmed by t he semi-structured questionnaire section on household medici ne inventory.
157 Finally, focus group participants were asked to make suggestions to improve the current situation. People mentioned that they needed to earn more money for basic food items such as rice and pasta. Moreover, one person suggested that they needed more co llaboration with the people from the neighboring province of Tungurahua, who had suffered similar economic losses because of the destructive effects of the volcano on agriculture and animal husbandry. One person suggested one area she saw needing improvement, alimentacin, en el campo, aliment os bsicos, desnutricin en los nios [nourishment, in the rural areas, basic foods, undernutrition in the children]. Table 4: Summary of Focus Group Discussions Thematic Category of Focus GroupMost Frequent Responses Most Common Childhood Health Problemsvolcanic ash, coughs, colds, diarrhea Home Treatmentspoleo, borraja, lemon juice Time to Wait before Seeking a Doctortreat at home for 2 days, then seek doctor Doctors and Nurses Use Language you Understanduse same language we understand Major Obstacles to Seeking a Doctorlack of money and restricted bus hours Health Informationwe lack information on treating children Responsibility for Providing Health Informationdoctors, SCS Penipe, and CEBYCAM Awareness of Pneumonia referred to a pulmona Suggestions for Improvementsmore money for food, food assistance These focus group discussions were helpf ul in the final formulation of the semi-structured questionnaire. Moreover for the ethnomedical question on hot and cold coughs in the semi-structured ques tionnaire, the researcher was able to further probe on this question because of the experience with the focus groups. Finally, the focus group discussions provided another perspective that was useful in the analysis of the in-depth inte rviews with the healthcare staff.
158 The primary findings of the focus gr oups were that mothers in Penipe viewed money as the primary obstacle to seeking medical care for their children because they needed money to pay for the bus and for medications. The most common childhood illnesses were listed as coughs and colds, and skin, eye, and throat problems caused by the aggravating effects of the volc anic ash. There was a consensus among the mothers that they lacked the necessary health information to make inform ed health care decisions when their children suffered from respiratory infections and they believed that it was the responsibility of CEBYCAM and the government health c enters to provide them with this education and training. There was a general lack of knowledge of the signs and symptoms of pneumonia and other severe re spiratory infections. Moreover, when mothers perceived that their childr en were suffering from severe coughs and colds, some would use antibiotics i nappropriately by not following the full course of the prescription or using lefto ver medications. There was a delay in seeking a doctors attention because mothers would attempt to treat their children at home with herbal remedies, alth ough they would not seek the help of curanderos, because they were not viewed as effective in healing these types of childhood respiratory illnesses. Finally, the mothers believed that the doctors and nurses used terminology that they coul d understand, a positive finding of the focus groups, since good health communica tion is one necessary component for a healthy community.
159 PART II IN-DEPTH INTERVIEWS IN HEALTH CARE SETTING In Penipe County there are tw o government health posts or subcentros de salud (SCS), one in Penipe proper (SCS Penipe), and one in the town of Matus (SCS Matus). These health centers are distinguished from the ancillary health outposts in the county because they hav e at least one permanent staff person ( de planta ), either a physician or nurse, assi gned by the Ministry of Health of Chimborazo Province and their facilities are of higher quality. Because Penipe is the major town in Penipe County, the two healthcare facilities in town, CEBYCAM (the only private clinic in the county) and SCS Penipe (the major public health center in the county) were selected fo r the healthcare staff interviews. SCS Penipe has a permanently assigned docto r, nurse, and nurses assistant, along with a dentist completing her year of rural service and a permanent dentists assistant. Moreover, SCS Penipe has the health inspector and another doctor who works once a week. The private health clinic CEBYCAM in the town of Penipe operates with a pediatrician, gynec ologist, dentist, laboratory technician, pharmacist, nurse, nurses assistant, and p sychologist. Its facilities are superior to the governments Ministry of Public H ealth centers, and it is the only health facility with a laboratory for blood, urine and fecal tests. SCS Matus in the town of Matus has a permanently assigned nurse and a doctor completing her year of rural service. There are health outposts or puestos de salud (PS) in El Altar, Bilbao, Bayushig, Candelaria, Pali ctahua, Shamanga, and Nabuzo. Only Palictahua and Bayushig are regularly staff ed. Only PS Palictahua, in the tiny
160 hamlet of Palictahua, has a permanently assigned nurses assistant and a doctor completing her year of rural service. PS Bayushig has a nurses assistant and a doctor completing her year of rural servic e who visits intermittently. Because the only other healthcare facilities in Penipe County with a doctor were PS Palictahua and SCS Matus, the researcher also interviewed the health staff at these health posts. By covering all four health centers (CEBY CAM, SCS Penipe, SCS Matus, and PS Palictahua) all health workers in the county were contacted and interviewed. The unstaffed health outpost s rely on visits from the doctors or nurses from the public health centers or outposts approximately every two weeks. These health outposts have very limited hours and weak infrastructure. The purpose of these in-depth interviews was to identify barriers to access which could explain delays in health care-s eeking for mothers of young children. Table 5: Summary of Staff for Health Posts in Penipe County Name of Health PostDoctorsNursesOther ProfessionalsOther Staff CEBYCAM2244 SCS Penipe1132 SCS Matus1100 PS Palictahua1100 PS Bayushig0100 In the in-depth interviews with info rmants from the health care setting there were eight doctors, six nurses and eleven healthcare workers from other fields for a total of 25 people. Of the doc tors, two were male, and the other six female (including a dentist and a psychol ogist). All the nurses were female, although only one was a nurse professional all the others were classified as
161 nurse assistants. The el even healthcare workers were a varied group, and included a physical therapist, two cur anderos, a nurse/social worker, a health inspector, a pharmacist, a cashier, a laboratory technician, a program coordinator, and an agronomist. During the ti me of the fieldwor k, all essential healthcare personnel in Penipe County were interviewed, with the exception of one dentist in CEBYCAM, who was not able to be interviewed. The interviews revolved around fourteen major categories of questions. The following discussion summarizes the findings of t he three categories of respondents, and conclusions will be drawn from the comparison and summation of their responses. Training For the public sector facilities under the Ministry of Public Health, which included SCS Penipe, PS Palictahua, and SCS Matus, the trend was to have doctors who were carrying out their year of rural service, a r equirement after they completed their seven years of training befor e they start their career. The typical salary of these doctors was approximat ely $350 per month. In contrast, at CEBYCAM, a private clinic, ther e was an experienced psychologist, gynecologist, pediatrician, dentist and laborat ory technician. Towards the end of the fieldwork session, a permanent doc tor had been assigned to SCS Penipe. Most of the nurses interviewed were ac tually nurse assistants. The one nurse professional in the county spent five years studying nursing and an additional two
162 years for the masters degree. The nur se assistants generally had a high school education and had received additional technical training. Access Two of the doctors in the rural clinics of PS Palictahua and SCS Matus would travel to rural areas every 15 da ys such as Bilbao, Nabuzo and Candelaria in order to attend patients t here. In Bilbao, there wa s regular medical attention until the Mount Tungurahua erupt ion of 1999, which has effectively cut off this region from the rest of the county, due to the destruction of bridges. In emergencies, residents of Bilbao travel to Baos for medical attention. The rural doctors reported elderly patients living in the countryside who could not afford medicines, so in these cases the doctors would bend the rules and dispense medicines and care without cost. Because of the issues of the research related to barriers to care, doctors were asked what they believed to be people s opinions of the health clinics. One of the doctors from SCS Penipe belie ved that there should be information campaigns to educate people of the services t hey offer. In the case of one of the rural clinics, the doctor commented that the people are content with the services, but sometimes they do not want to pay One nurse from SCS Penipe explained the need for a laboratory in or der to perform urine and fecal testing, and a vehicle to transport patients in emergencies. The request for an ambulance had been presented to the head of the Ministry of So cial Welfare at a public event in town in summer 2004, and a year later, the prom ise was fulfilled. The nurses at the
163 rural clinics also commented on the lack of necessary equipment, such as more sterile equipment, stethoscopes, as well as supplies of medications and vitamins. Moreover, the nurses in the rural clin ics pointed to staff needs such as an obstetrician, dentist, and a permanent doctor position. According to the doctor, SCS Matus was lacking in certain areas. Some of the prim ary identified needs included a dentist and equipment for chil dbirth. In addition, the doctor recommended community education courses on hygiene and boiling water. Finally, the doctor recommended dispensing free medications to the elderly, who were often unable to pay, and this sentim ent was echoed by other doctors. Therefore the consensus among healthcare professionals was that the primary barrier to care was education, foll owed by services and transportation. Figure 6: Penipe County Ambulance
164 Primary Health Needs There was a consensus among docto rs and nurses that respiratory problems, parasitosis, and diarrhea were the primary health problems affecting children. Some mentioned poverty as t he underlying cause of health problems in the community, leading to undernutrition. The nurses focused on the health risks caused by the volcanic ash, particularl y respiratory problems. Another nurse commented that agriculture and cattle -raising had been devastated by the volcanic eruption, thus exacerbating poverty. The health inspector gave a long list of health problems in the community. He identified the unsafe water as a ma jor problem causing parasitosis in 80 percent of the population. Secondly, he listed the active volcano as a major hazard, causing people to lose their crops and to now be dependent upon raising pigs. He also listed the limited hours of the health clinics, the open dumping of garbage in ravines, and poverty as majo r health problems in the community. Delay in Health Care-seeking Another major focus of the in-depth interviews was to identify what the health professionals believed to be t he reason that mothers delayed careseeking for their children when they becam e sick with respiratory ailments. One nurse commented that people are accustom ed to treating childhood respiratory problems at home by purchasing medicati ons in the pharmacy. If the problems did not resolve at home, mothers woul d come to the clinic and were given Tempra for fevers or anti biotics like ampicillin.
165 A very interesting question which elic ited different answers from mothers and doctors was whether the mothers delay seeking the doctors attention for their childs health problems. In one of the rural clinics, the doctor commented that the people did not come down to the clinic readily and were more likely to use curanderos or fregaduras (bone-setters). It was t he doctors responsibility to convince the people of the importance of seeking a doctor. Another doctor observed that the mothers w ould use home remedies, such as herbal teas, for as long as four days before bringing a child to the doctor. In SCS Penipe and one of the rural clinics, the doctors did not per ceive a delay in mothers bringing their children to the health centers. One doctor suggested that the people knew very little about respiratory infections and had become accustomed to the ash since it had been falling for around five years. The doctor added t hat poor nutrition, which weakened the bodys defenses, and the cold climate were contributing factor s. Another doctor commented that people would use terms such as chest whistling and green discharge from the nose to describe t he symptoms of ARIs. Another doctor commented that the public was quite sophi sticated in the treatment of ARIs, because after the 1999 Mount Tungurahua er uption, volcanic ash was a health threat, and there were numerous health ca mpaigns. Therefor e, the people know that they could buy amoxicillin for re spiratory problems and cotromixidazol for intestinal problems. However, the docto r said the public did not follow the proper regimen for these prescriptions.
166 Nurses were asked their opinion of whether mothers delayed in seeking medical attention for their sick children. There was some consensus that the nurses believed the people were more concerned with tending their crops and caring for their animals than with seek ing immediate care. Another nurse commented that the largest obstacle was tr ansportation, a sentiment that was not echoed by the people themselves, who list ed money as the greatest barrier. Related to the costs of the medica tions, one nurse commented that at the government clinic, the cost of medicine equaled a sma ll donation, and the patient would only have to pay one dollar. Finally related to barriers, another nurse commented that the hours of the clinic were restri ctive, and the people would seek care more readily if their hours were extended. Another health worker believed the people lacked education and would usually attempt to treat illness at home, but that they do not know if t he treatments they used caused more harm than good. One reason cited for this treat ment choice was that people felt they could save money on medications by using herbal remedies instead. In sum, the healthcare professionals believed t hat the people of Pe nipe County delayed treatment because of lack of education of the seriousness of respiratory infections, preferred to treat these illnesses at hom e because they thought it would be cheaper, and would use a comb ination of herbal remedies and prescription medicines that in the end could do more harm than good.
167 Antibiotic Dispensing A difficult topic to get accurate information on was the frequency of antibiotic dispensing. At SCS Penipe, amoxicillin and am picillin were prescribed according to the AIEPI chart the doctor showed me.6 For example, first a nebulizer was used to clear out most of the phlegm, and then either amoycillin or cephalexin would be prescribed for a sev en day regimen. At one of the rural clinics, the doctor said amoxicillin was prescribed for fever. At CEBYCAM, they used a nebulizer which delivered a pain re liever, cortisone, and saline solution. They would also prescribe antibiotics. Some doctors commented that if the case was severe, such as pneumonia, the pati ent would be referred to a hospital in Riobamba. Nurses were permitted to prescribe antibiotics. One nurse described a particular scenario. She said, If he come s with a fever, I take vital signs, and then put a cold cloth on his head. I gi ve him medications. We prescribe amoxicillin or give Tempra if he has a lot of cough, lung secretions, or rhinorrhea. Another nurse answered, I can give amoxicillin, dicolofenar, paracetomol, and ampicillin. I cannot give medicine for arterial pressure, only a doctor can give medicine for attacks. One nurse explained that ARIs and bronchitis were very frequent in child ren, and they would frequently give amoxicillin and Tempra, or another antibiotic, cephalexin, for more serious cases. Another nurse observed that the volcano had led to ma ny cases of adenoiditis, but the people had learned to protect thei r children by keeping them inside, 6 AIEPI stands for Atencion, Integral, Enfermedades Prevalentes en la Infancia [Attention, Integral, Sicknesses, Prevalent in Infancy] and was developed in conjunction with WHO.
168 sweeping their homes, and using breathing ma sks or handkerchiefs to cover their mouths. When asked to describe how t hey detected ARI in children, one nurse responded that fever, headache, inflamed adenoids, cough, and nasal secretion were the prevalent symptoms. One health worker mentioned giving penic illin for adults and children, and nebulizations for women and children. A nother health worker gave nebulizations with medications such as salbutamol and ventolin to assist breathing. The community health nurse, who was a P enipe resident, worked for the Ecuadorian equivalent of the Departm ent of Children and Families and had her office separate from the health centers in Penipe in one of the old municipal offices, had medications such as penicillin to gi ve to people with respiratory problems, and she said they did not need a prescription if there was no doctor available. The nurse said the mothers were not very timely in bringing their children to seek medical care. She said the mothers w ould sometimes wait until their children were in a very grave condition before s eeking care, and she would refer them to SCS Penipe or the Multicli nic Hospital in Riobamba, sometimes helping with the bus fare. Another health worker was dismayed about how some mothers would leave their young children at home alone while they went to work in the fields. The community health nurse talked about respiratory illness treatments she prescribed such as Berodual (Boehni ger Ingelheim) and Flixotide (Glaxo Wellcome), as well as herbs like salbareal poleo peach flowers, violets, and blackberry leaves.7 7 Berodual is a bronchodilater inhaler used to treat asthma and COPD. Flixotide is a corticosteroid used to treat asthma by reducing swelling and irritation in the wall of air passages
169 Vaccination Coverage Finally, the researcher asked doc tors about vaccination coverage and medication costs. The vaccines were giv en in the government health clinics, not CEBYCAM. There is a new vaccination called pentavalente that is being used, but its implementation has encountered so me resistance according to Penipe doctors because occasionally there are si de effects like a rash. Pentavalente protects against diphtheria, tetanus, hepatitis B, and haemophilus influenzae type b. This vaccine will soon be introduc ed in the U.S. under the brand name Pentacel, and it has been given in C anada for about a decade. It is given in three doses to children less than one year old. Other va ccinations included bacillus Calmette-Gurin (BCG), against tuberculosis, measles vaccine, triple vaccine DPT (against pertussis, diphtheria and tetanus), varicella zoster vaccine, and the polio vaccine. As far as costs, at CEBYCAM the costs of the laboratory and pharmacy are split in half, an d the consultation is free. For SCS Penipe, the cost for consultation was fi fty cents and the medications that they had, which were not extensive, were di spensed either without cost or for a nominal fee. Traditional Healers One elderly curandera interviewed expl ained that she attended patients on Tuesdays and Thursdays. The curander a explained that she cured patients in lungs, thereby alleviating respiratory distress The herbal remedies are discussed in a later section in this chapter.
170 using cuy (guinea pig), cane liquor, tobacco, and herbs such as Santa Maria and marco to cure patients of mal humor and mal aire (bad bile and bad air, both folk illnesses). Another cur andero explained that he cur ed children by putting an egg in the glass, examining the egg, cleaning the back of child with perfume, and then blowing on them with cigarettes. The typical day for this curandero was to attend patients on Sunday after the market The curandera identified two types of coughs: tos de fro and tos de calor (cold and hot cough). For cold coughs she recommended poleo and borraja and for hot coughs, teas of malva (an herb used to treat bronchitis), thereby conf irming the informati on from the semistructured questionnaires. Costs Finally on the question of costs, pat ients would pay fifty cents for a consultation at SCS Penipe; however the major costs were for the medications, especially at CEBYCAM, w here there was a pharmacy. The CEBYCAM cashier said the most expensive medications we re hormone medications and vitamins, and a routine clinical ex am with laboratory tests would cost around twelve dollars. The pharmacist listed regimens of antibiotic medications or other medications that people would purchase for childhood infections. A 30 tablet, 500 mg prescription of amoxicillin woul d cost $5.40 compared to a 24 capsule regimen of ampicillin at $2.88. Another antibiotic wa s cephalexin at $3.00 for a 30 tablet regimen of 250 mg each. A peni cillin treatment for childhood bronchitis cost $2.17. A child with an intestinal in fection would take trimetoprin, costing
171 $3.60 for the 30 pill regimen. The mo st common cough syrup was Stoptos at $1.46 for a 60 ml bottle, but t here were others in the three to four dollar range. While these costs do not appear high to people in the United States, for people making less than $150 per month, a prescrip tion costing $5 to $10 is a significant slice of their income. Table 6: Summary of Responses from In-depth Interviews Thematic Category of In-depth InterviewMost Frequent Responses Primary Health Needsrespiratory problems, parasites, diarrhea Suggested Improvementslab equipment, ch ildbirth facilities, more staff Public Opinion of Health Servicegenerally positive Number of Patientsmanageable number of patients Need for More Trainingamenable to idea of further training Delay in Patient Health Care-seekingtoo much emphasis on home treatment Use of Local Terminologyappropriate use of local terms Cost of Service consultation costs fifty cents at health posts, compared to higher costs of medicines While the opinions of the healthcare workers were varied, some common themes emerged. There was a general co nsensus that the government clinics needed to expand their staff, build infras tructure, extend their hours, and reach out to the community to better inform them of the medical services offered. Moreover, there was a concern that t here was a delay in care-seeking by mothers, who needed to be better educated to recognize the signs of serious illnesses in their children instead of trying to treat at home with herbal remedies because they were cheaper than prescription medicines. Finally, the healthcare workers shared a belief that the public they served generally respected their
172 professionalism and expertise in offering the services th at they were capable of providing.
173 PART III EPIDEMIOLOGICAL DATA FROM MINISTRY OF HEALTH Extensive epidemiological data were co llected by the research team under the CDMHA grant (Whiteford and Tobin 2002). One objective of further epidemiological data collection was to co llect information from 2002 to 2004 to complement the data collected earlier. However, because of unreliable population data, it was not po ssible to calculate ARI morbidity rates to compare with earlier years; moreover, the morbidit y data collected represented six months for the years 2002 and 2004 and eight mont hs for 2003, thus making comparison across the years difficult. In this study, secondary morbidity data for ARI, pneumonia and malnutrition were collected and analyzed from records at the Chimborazo Ministry of Public Health. The following figures show the number of cases reported in Penipe County by SC S Penipe, CEBYCAM, SCS Matus, and PS Palictahua (No data were provided by CEBYCAM for 2002). There were 63 total pneumonia cases over a three-year period for under-fives; however, there were no cases in 2002. The pneumonia incidence rate for 2003 in the less than one-year-old age group was 649 per 10,000 and for 2004, 216 per 10,000 population. In the one to four-year-old age group, the pneumonia incidence rate for 2003 was 173 per 10,000 and for 2004, 95 per 10,000 population. The highest rate recorded in the prior research by Whiteford and Tobin (2002) was in 1999 where the rate in the less than one-year-old age group was 250 per 10,000 and 53 per 10,000 in the one to four-year-old age group (Whiteford and Tobin 2002). The rates in 2003 appear unusually high; however, the small population
174 sample size numbers for the age groups cause even small increases in pneumonia consultations to produce compar atively higher rates. The data for cases are presented for two age groups less than one-year-olds and children between one and four years old. Penipe County Cases Children < 1 yr. old (Jan. June, 2002 2004)76 161 65 0 24 8 6 4 18 0 20 40 60 80 100 120 140 160 180 20022003*2004Number of cases ARI Pneumonia Malnutrition Figure 7: Penipe County Ca ses, Children < 1 yr. old 2003 covers January thru August, Source: (Ministry of Health Chimborazo 2004)
175 Penipe County Cases Children 1 4 yrs. old (Jan. June, 2002 2004)152 323 142 0 20 11 35 15 32 0 50 100 150 200 250 300 350 20022003*2004Number of cases ARI Pneumonia Malnutrition Figure 8: Penipe County Case s, Children 1 4 yrs. old 2003 covers January thru August, Source: (Ministry of Health Chimborazo 2004) Population estimates from the Ministry of Health in Chimborazo for under one-year-olds are 372 (2002), 370 (2003), an d 370 (2004). For children between one and four the population estimates for Penipe County are 1152 (2002), 1153 (2003), and 1153 (2004). However, there are another set of figures for 2004 which count 152 children under one year old and 169 for children between one and four. From the researc hers observations, this latter figure is more accurate, considering the researcher canvassed half of the county asking for households with under-fives and only encountered less than one hundred households satisfying the criteria. The pneumonia incidence rate s demonstrate that infants had higher risk for pneumonia than the young child age group. Moreover, using the population overestimates produces some alarming ARI consultation rates. To calculate the rates, the number of cases is divided by the population subgroup
176 and multiplied by 10,000. For the under one-year-old group the estimated ARI consultation rates are as follows (F igure 9): 2,043 per 10,000 (2002); 4,351 per 10,000 (2003); and 1,757 per 10,000 (2004). Si milarly alarming statistics for one to four year olds are presented for the six-month period: 1,319 per 10,000 (2002); 2,801 per 10,000 (2003); and 1,232 per 10, 000 (2004). Both the pneumonia and ARI incidence rates for the under one-yea r-olds are higher than the older age group, suggesting higher vulnerability to re spiratory infections among infants in this region. Penipe County ARI Morbidity Rates for Six Month Period (Jan. June, 2002 2004)2043 4351 1757 1319 2801 1232 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 20022003*2004Rate per 10,000 people Children < 1 yr. old Children 1 4 yrs. old Figure 9: Penipe County ARI Morbid ity Rates for Six Month Period 2003 covers January thru August, Source: (Ministry of Health Chimborazo 2004) The trend of the above graph shows a doubling of ARI morbidity rates from 2002 to 2003, and then lower rates again in 2004. In the prior research by Whiteford and Tobin (2002), outpatient c onsultation rates for ARI more than
177 quadrupled in Penipe County in 1999 for under-fives, the year of the volcanic eruption (see Figure 10). There is no evi dence that this occurred again in 2003, but this is one possible explanation for t he spike in outpatient consultation rates for ARI, before returning to pre-1999 rates in 2004. Du ring the field period in 2004, volcanic activity was relatively low, with occasional emissions of volcanic ash. A separate report carried out by Dr. Jorge Lara of CEBYCAM found that bronchitis and pharyngitis incidence ra tes among under three-year-olds rose dramatically after the vo lcanic eruption of Mount Tungurahua in October, 1999 (Lara, et al. 2000a). The rates of common respiratory infections rose from an average of 16 to 18 percent to 42 percent within the year. Penipe County ARI Morbidity Rates (1995-2001)0 1000 2000 3000 4000 5000 6000 1995199619971998199920002001Rate per 10,000 people Children < 1 yr. old Children 1 4 yrs. old Figure 10: Penipe County ARI Morbidity Rates (1995-2001) Source: Adapted from Whiteford and Tobin (2002) The attitude of the nurses in Penipe County was that the people had grown accustomed to the vo lcanic ash in the air, and consequently, rates of
178 outpatient consultations for respiratory infe ctions had stabilized. According to the literature, there is no evidence that s uggests that people can develop resistance to volcanic ash in the way descri bed by the nurses, and this should be considered a folk belief. Moreover, t here was no evidence that the health workers themselves protected themselv es against the ash by wearing masks or scarves, suggesting that they may have thought that they themselves had developed a resistance to the volcanic as h. The epidemiological data suggest that respiratory infections will continue to be an acute problem for children in the zone.
179 PART IV QUANTITATIVE DATA FROM QUESTIONNAIRE This section introduces the descriptive statistics results from the semistructured interview instru ment. The information incl uded in this section are frequencies and percentages of the vari ables discussed. There are 108 quantitative variables, and twelve ques tions which ask purely qualitative information. The quantitative data discussi on is broken down into the following sections: demographics, socioeconomic stat us, risk factors for ARIs, childs respiratory problems, use of health res ources, and diet of youngest child in the last twenty-four hours. The qualitat ive data discussed include use of local healthcare facilities and resources, at titudes towards healthcare options, knowledge of home remedies, use of curanderos, and folk illnesses. Demographics The total number of female respondent s was 91. There were three groups of respondents distinguished by geographic area. The first 31 respondents were from the town of Penipe proper. The second 30 respondents were from an area the researcher designated as Puela, but ac tually included the nearby hamlets of Palictahua, Pungal, and Ganzhi. The final 30 respondents were from an area the researcher designated as Matus, which also included a few respondents from neighboring Calshy and El Altar. T he age of respondents range d from 18 to 57, with an average age of 30. Ei ghty-four respondents (92.3% ) were mothers of the children whose health history was questioned. The others were relatives such as
180 grandmothers or aunts. Twenty-eight respondents (30.8%) were between the ages of 31 to 44. Six respondents were over 45 years old. Roughly half of respondents (53.8%) were married, the rest were either in a unin libre (cohabitating) or single. Only one of the 16 single fema le-headed households had a monthly salary over $150 per mont h and only two out of 16 ranked above average on the possession score index (5 items or above on a scale of 9). Table 7: Demographics and SES of Fema le Caregiver Respondents (n=91) CategoryNumber Female caregivers 18 to 3057 (62.6%) Female caregivers 31 to 4428 (30.8%) Female caregivers 45 or older 6 (6.6%) Female caregivers (mothers)84 (92.3%) Female-headed households16 (17.6%) Primary education only65 (71.4%) Secondary level education16 (17.6%) College education10 (11.0%) Primary occupation at home76 (83.5%) Work in agriculture77 (84.6%) Salary above $150/month28 (30.8%) Mean monthly salary ($)$126.71 Median monthly salary ($)$100.00 Receive outside help33 (36.3%) CEBYCAM source of help28 (30.8%) SES score above average53 (58.2%) Home ownership41 (45.1%) Land ownership39 (42.9%) Five or more dependents13 (14.3%)
181 Figure 11: Map of Penipe County Towns Source: (COSUDE 2003)
182 The average level of education of respondents was primary school with some secondary, or approximately 8.78 y ears of schooling. There was a positive correlation between years of education with monthly salary (r = 0.54, p < .001, two-tailed). Sixty-five respondents (71.4%) only had prim ary schooling. Seventy-six respondents (83.5%) clai med homemaker as their principal occupation. Seventy-seven respondents ( 84.6%) also worked in agriculture, at least in some capacity. There was an open-ended question that asked for the primary breadwinners occ upation. Most women responded that their husbands either worked in agriculture or construc tion jobs, such as bricklaying. Other common occupations for husbands or boyfr iends were driving trucks or buses and chainsaw work. Ten respondents (10. 9%) answered that their husbands had professional jobs such as teaching or wo rking for the government or a company. Respondents were required to live in the area for at least three years to participate in the questionnaire. The av erage time in town for respondents was approximately 10 to 11 years. Socioeconomic Status The primary hypothesis of the study involves soci oeconomic status (SES). Therefore a number of questions were posed to gauge SES. The first and most obvious question involved household income The monthly salary of families ranged from $0 to $500 a month. The aver age salary in the sample was $126.71 per month, with a median sa lary of $100. A baseline of $150 per month was set as the average salary for the area based on population census data for
183 Chimborazo Province (INEC 2001). Si nce roughly 85 percent of respondents worked in agriculture in some capacity, the average salary for Penipe County residents would be lower if the few respond ents with high salaries from the town of Penipe were removed. The proportion of people working in agriculture in this sample is greater than the proportion of agriculturalists for Chimborazo Province (49%), emphasizing that P enipe has primarily a rural, agriculturally-based economy (INEC 2001). As mentioned, other forms of income such as agriculture could increase household income. Fifty percent of respondents worked in agriculture seasonally, and forty-seven percent of respondents owned more than three agricultural animals. In addition, many respondents answered that they received outside help. Of the thirty -three respondents who rece ived such help, most received help from CEBYCAM, which c onsisted of monetary aid (around $28 every two months for t heir childrens welfare). A series of questions were asked in order to create a possession or SES score. These questions asked whether the respondent owned any kitchen appliances, a bicycle, a television, more than three agricultural animals, lived in a home with walls of blocks or bricks rat her than wood or plastic, had a non-dirt floor, and had a home with more than three rooms. Together with the question on monthly household salary (which was conv erted into the dichotomous variable above or below mean salary), a SES sco re was generated which ranged from 0 to 9. Thirty-eight respondents (41.8% ) rated low on the SES score (below 5), and fifty-three respondents (58.2%) rated hi gh on the scale (5 or above). The
184 mean SES score was 4.7 (SD = 1.94; mi nimum = 1; maximum = 8). The SES score moved ten people who had salaries below $150 per month into the high end of the SES score, so there were differ ences between the two measures. It was determined that all of the variables on the SES score receive equal weight, because there were no criteria to dete rmine whether some of the possession items were more important than others. There was a positive correlation between SES score and monthly salary (r = 0.54, p < .001, two-tailed), indicating that as monthly salary rose, socioeconomic status rose proportionately according to the scale. The validity of the SES score as an accurate measure of the respondents socioeconomic status was limited because it did not take into account either education or occupational status; however, because of complex issues involved in weighting these vari ables, a simpler model was adopted for analytical purposes (Weller 1998). Forty-one respondents (45.1%) ow ned their home, and thirty-five answered (38.5%) that it was owned by a relative. Respondents had an average of three rooms in their homes. Fifty-two respondents (57.1%) did not own land compared to thirty-nine (42.9%) who did. Respondents were asked to report how many dependents lived in the household. The responses ranged from one to eight dependents, with an average number of three dependents (mean = 3.24 dependents).
185 Risk Factors for ARIs A number of questions asked respondent s about the environment in which their child or children lived. These ques tions identified the age and gender of the child, how many people lived in the household and shared t he childs bedroom, whether the person who shared smoked ci garettes, if the child shared a bed, what were the methods used for heating and cooking, and whether the family owned household pets. This section continues with the frequencies for each response. The average age of a child in the study was between 17 and 18 months old. There were more boys (49) than girls (42) in the study. There were questions to determine if crowding was a ri sk factor for the health of the child. Therefore, respondents were asked about the number of persons in the household and the characterist ics of the childs home environment. The average number of people in a household was fi ve, with a minimum of three and maximum of twelve people. Only nine percent of households surveyed had children under five with his or her own r oom. The average child shared his or her room with two or more people. This number is inflated by the very young children surveyed, who generally shared their bed and room with their parents. Sixty-eight of the children (81.9%) who shared their r ooms, also shared their bed with others.8 8 The female caregivers of children who had his or her own room were not asked the question of whether the person with whom their child shared a room smoked cigarettes or whether the child had his or her own bed and fall into the category not applicable, hence eight cases are excluded from the percentages for those two questions.
186 Table 8: ARI Risk Factors for Youngest Child (n=91) CategoryNumber Child less than 6 months old 8 (8.8%) Child 6 17 months old23 (25.3%) Child 18 29 months old26 (28.6%) Child 30 months 5 years old34 (37.4%) Gender of child (male)49 (53.8%) Gender of child (female)42 (46.2%) Over 5 people in household31 (34.0%) Child shared room83 (91.2%) Child shared bed76 (81.9%) Child shares room with smoker18 (21.7%) Smoker in family36 (39.6%) Boils drinking water57 (62.6%) Pet owners75 (82.4%) Pets enter home41 (54.7%) Heat home with wood47 (51.6%) Cook with wood only19 (20.9%) Child sleeps in kitchen 7 (7.7%) Child vaccinated90 (98.9%) There were questions about smoking, wh ich is another risk factor for ARI in children. Of those children who shared a room with another person, twentytwo percent of these children shared t heir room with a smoker. The other question on smoking asked more generally w hether or not there was a smoker in the household. Thirty-six people (39.6%) responded that there was a smoker in the household. When asked how much the person in the household smoked (usually a husband or boyfriend), the answer was generally only when he goes out and drinks, or very rarely. It was not possible to gauge the smoking frequency of the family member in question, who was always someone other than the respondent. There were three questions about water usage. The first question asked whether the respondent received water fr om a well, fresh water source, or a
187 municipal source. Virtually everyone responded that their water was piped in from a municipal source (three peopl e responded well). Next, respondents were asked if they boiled their water to drink and if so, for how many minutes. Fifty-seven respondents (62.6%) answered that they boiled their water, and thirty-four respondents (37.4%) did no t. Of the fifty-seven respondents who boiled water, one was the mi nimum number of minutes that water was boiled, and the maximum was 30 minutes (60 minute outlier removed from analysis). Seven and a half minutes was the aver age number of minutes that respondents boiled water to drink and cook; however, some people seemed to be aware that ten minutes was the standard. Pets are another risk factor for respir atory infections because of the dust and dander that they track into the home. Respondents were asked if they had cats and dogs, and whether the pets were allowed to enter the house. Frequently, respondents would answer that the cats could enter, but not the dogs; however, a limitation of the questionnaire is that it did not distinguish which type of pet was allowed to enter. Dogs were more common pets than cats, and many people had more than one dog. The dogs were trained to be hostile to strangers. Sometimes the dogs were tied, but more ofte n, dogs were allowed to roam freely. The majority of respondents owned cats, dogs or both (82.4%). Of the respondents who had pets, forty-one (54.7% ) allowed their pets to enter the home while thirty-seven (45.3%) did not. Another risk factor for ARI is exposure to cooking smoke in the home (Mishra and Retherford 1997). Theref ore, there were two questions about
188 cooking method and home heating. Fi rst, respondents answered how they heated their home. Twenty respondents (22%) responded that they did not heat their home. The rest answered that they primarily heated their home with wood (51.6%) or gas (25.3%). While heat ing the home may have been a foreign concept to some, everyone understood the cooking question. Many respondents answered that they cooked with both w ood and gas (34.1%), but more relied on gas exclusively (45.1%). A smalle r proportion depended on cooking with wood only, because they could not afford gas (20.9%). A common feature for households who cooked with wood was to have a separate kitchen area away from the living quarters bec ause of the smoke problem. A related question to these two questions on fuel consumpti on was whether the child slept in the kitchen, where exposure to smoke is a problem. Seven respondents (7.7%) answered that their child slept in t he kitchen, usually because the house was very small. Finally, another risk factor for ARIs is vaccination status. Only one respondent (1.1%) in the sample responded that her child had not been vaccinated.
189 Childs Respiratory Problems The bulk of the questionnaire asked about the childs respiratory health history. Most of these questions were in a yes/no format, and there were a number of follow-up questions that were contingent upon an affirmative response to the first question. Half of respondent s said that their children normally had cough with colds. Twent y-nine (31.9%) respondent s answered that their youngest child normally had cough apart from colds. Table 9: Youngest Childs Respiratory Pr oblems (n=91, unle ss otherwise noted) CategoryNumber Cough with cold46 (50.5%) Cough without cold29 (31.9%) Frequent cough (n=52)19 (36.5%) Eliminates phlegm58 (63.7%) Frequent phlegm (n=55)24 (43.6%) Attacks of phlegm24 (26.4%) Wheeze47 (51.6%) Wheeze with cold (n=48)37 (77.1%) Wheeze apart from colds (n=48)18 (37.5%) Wheeze days and nights (n=48)21 (43.8%) Wheeze attacks cause rapid breathing9 (9.9%) Wheeze attacks with exercise3 (3.3%) Severe chest illness28 (30.8%) Hospitalized for chest illness18 (19.8%) Other chest illness before age 210 (11.0%) Measles or Rubeola3 (3.3%) Varicella12 (13.2%) Sinusitis1 (1.1%) Bronchitis19 (20.9%) Pneumonia10 (11.0%) Adenoiditis14 (15.4%) Bronchopneumonia6 (6.6%) Asthma4 (4.4%) Allergies11 (12.1%)
190 If the respondent answered affirmativel y to either of these questions on coughs and colds, the respondent was then asked whether her child had a cough more than four times a week for at least three months out of the year. For those who answered this question, nineteen respondents (36.5%) had children with a frequent cough and thirty-three respondents ( 63.5%) did not. Of those who had a persistent cough, the average child had had a cough for approximately two years. Fifty-eight children (63.7%) elimi nated phlegm or were congested when they had a cold. Of those who elimi nated phlegm, twenty-f our cases (43.6%) were congested, and eliminated phlegm fo ur times a week for at least three months a year.9 Of those who eliminated ph legm, they had done so for an average of one and a half years. Twenty-four children (26.4%) had atta cks of cough, congestion or phlegm that lasted at least a week each year Of those who had had attacks, the average number of years living with this health problem was around a year and nine months (mean = 1.84 year s). In addition, five colds per year was the average number for those w ho suffered these attacks. Forty-seven respondents (51.6%) had a child whose chest had wheezed sometime, normally with colds (77.1%). In contrast, wheezi ng apart from colds was less common (37.5%). For those children who had wheeze symptoms, when asked if they wheezed most days and nights, the answers were split, fortyfour percent of respondents answered yes and fifty-six percent answered no. 9 Three cases not applicable because the child was less than a year old and therefore too young for the question to be relevant.
191 For the respondents who were asked the number of years their child had had wheeze symptoms, the average was aroun d one year and eight months (mean = 1.73 years). A small percentage of respondents (9.9%) answered that their child had wheeze attacks which had caused rapid breathing. These respondents were asked five follow-up questions. The firs t asked if the child had had two or more wheeze episodes, and only one respondent answered in the affirmative. The second question asked whether the child took any medicine for the wheeze attacks when they occurred. Seven respondents (77.8%) responded in the affirmative. The third question asked for the age of the child when the wheezing symptoms began. The average age was one and a half years old for the beginning of wheeze symptoms. The f ourth question asked whether there was normal breathing between the attacks. Mo st respondents (55.6%) answered that there was normal breathing. Finally, if there was a doctors diagnosis, the respondent was asked to describe the resu lts of the consultation. The most common responses to this question were either pneumonia or bronchitis as the doctors diagnosis. The last question on wheezing asked the respondent if the child had had wheeze attacks while exercisi ng, a sign of asthma. Only three respondents (3.3%) answered in the affirmative. The next section of the questionnaire asked about chest illnesses. Twenty-eight respondents (30.8%) had a child with a chest illness that lasted three days and had impeded normal activities in the last three years. Within the group of severe chest illness sufferers eighteen children (64.3%) had more
192 congestion than normal with these illnesse s, and twenty children (71.4%) had between one and seven chest illnesses per year. Moreover fifteen children (53.6%) suffered chest illnesses t hat lasted seven days or longer. Eighteen children (19.8%) in the st udy were hospitalized between one and three or more times for chest illnesses. Ten respondents (11%) answered that their child had suffered another grave ches t illness before reaching two years old. The most common childhood respiratory illn esses were bronchitis (19), followed by adenoiditis (14), chicken pox (12) and pneumonia (10). Bronchopneumonia (6), asthma (4), sinusit is (1), and measles/rubeola (3) were less frequent illnesses reported by respondents. There were three follow-up questi ons for the four respondents who answered yes to the asthma question. With such a small sample of respondents, the following stat istics are reported for informational purposes only. The first question asked at what age as thma symptoms start ed. The average age for the four respondents was around two y ears old. Of the four respondents with asthmatic children, three out of f our no longer had asthma. For the three respondents who reported that their child s asthma had resolved, the question was posed at what age this had occurred. For the first child, asthma symptoms began at around nine months and resolved at one year old. The second child began asthma symptoms at one and a half and resolved at age two. The third child began to have asthma at four year s old and resolved at four and a half, and the fourth child began to have symptoms at a year and seven months and never became asymptomatic. For the thr ee children whose asthma symptoms
193 resolved, the average time was five mont hs. When asked if the child continued to take medicine for t heir asthma now, one of the respondents who reported that asthma symptoms had resolved continued to give asthma medicine to her child. The next section of the questionnai re asked respondents to report on allergies in their children. The first ques tion asked whether the child was allergic to food or medicine, or anything else. The most common response to medicine allergy was penicillin. Eleven respondents (12.1%) answered in the affirmative that their child had allergies of some form. All respondents were asked the follow-up questions regardless of their re sponse to the first question. Seven respondents reported that their child was allergic to pollen or dust, and one reported an allergy to chemicals, in this case, particles from the active volcano. Five respondents reported that their child received inject ions for their allergies. Use of Health Care Resources The most difficult yet interesting part of the questionnaire dealt with use of health care resources and perception of se verity of illness symptoms. It was necessary to take time to explain this part of t he questionnaire to respondents because initially, the questions had the potent ial to cause confusion. There were a series of nine separate illness symptoms listed. The final symptoms on the list, indrawn chest and rapid breathing, are the classic symptoms of ARI. The objective of these questions was to det ermine the female caregivers careseeking behavior depending on the severity of the symptoms.
194 The first set of questions asked res pondents to answer where they would treat their child first: at home, at the doctors office or do nothing. For the symptom cough and nasal secretion, 80 per cent of respondents would go to a doctor first. For cough and fever, 75 per cent would see a doctor first. For fever and nasal secretion, cough, fe ver, and nasal secretion, and cough and phlegm 76 percent would see a doctor first. When hearing the word fever, many respondents deemed this term serious and suggested that seeing a doctor was appropriate. For the symptoms, cough, phlegm, a nd nasal secretion, the result was 84 percent seeking a doctor first. For all of these symptoms, relatively few respondents answered nothing (maximum = 3.3%). The results of the final three symptom combinations were telling. For rapid breathing, 84 percent of respondents answered that they would seek a doctor first, and 11 percent would do nothing. When asked what action they would take in the case of indrawn chest, 77 percent of res pondents would seek a doctor first, and 26 percent would do nothing. Finally, for the comb ination of symptoms rapid breathing and indrawn chest, 78 percent would see a doctor first, and 19 percent would do nothing. Therefore, for these final three symptoms, female caregivers were more likely to seek a doctor first than the aver age for the previous symptoms, but they were also more likely to take no action at all.
195 Table 10: Caregiver Treat ment Choice Depending on Symptom Cluster (n=91) Category of Symptom ClusterHome remedyDoctorNothing Cough with nasal secretion17 (18.7%)73 (80.2%)1 (1.1%) Cough and fever22 (24.2%)68 (74.7%)1 (1.1%) Fever and nasal secretion19 (20.9%)69 (75.8%)3 (3.3%) Cough, fever and nasal secretion21 (23.1%)69 (75.8%)1 (1.1%) Cough and phlegm20 (22.0%)69 (75.8%)2 (2.2%) Cough, phlegm and nasal secretion14 (15.4%)76 (83.5%)1 (1.1%) Rapid breathing5 (5.5%)76 (83.5%)10 (11.0%) Indrawn chest6 (6.6%)61 (67.0%)24 (26.4%) Rapid breathing and indrawn chest3 (3.3%)71 (78.0%)17 (18.7%) The next section of the questionnaire asked respondents to rank the level of severity for the same list of symptoms as previously described. In the analysis section, the responses between the two groups of questions are compared to determine whether there was a consistency of responses. In other words, do respondents who answer that indrawn chest is a case for the doctor also rank the symptom as serious to emergency? For the illness symptoms cough and nasal secretion, 71 percent ranked it as not serious to moderate, and 29 percent ranked it as serious to eme rgency. For the symptoms of cough and fever, 69 percent ranked the level of severi ty as not serious to moderate, and 31 percent classified it as serious to emergency. For both fever and nasal secretion, and cough, fever and nasal secretion, 65 percent ranked the symptoms as not serious to moder ate and 35 percent ranked them as serious to emergency. Cough and phlegm were ranked as not serious to moderate by 70 percent of respondents an d as serious to emergency by 30 percent of respondents. The symptoms of cough, phlegm and nasal secretion
196 were considered not serious to moder ate by 65 percent of respondents, and as serious to emergency by 35 percent of respondents. Table 11: Level of Severity of Symptom Cluster (n=91) Cate g or y of S y mptom Clusternot seriousmoderateseriousemer g enc y Cough with nasal secretion28 (30.8%)37 (40.7%)18 (19.8%)8 (8.8%) Cough and fever22 (24.2%)41 (45.1%)19 (20.9%)9 (9.9%) Fever and nasal secretion27 (29.7%)32 (35.2%)25 (27.5%)7 (7.7%) Cough, fever and nasal secretion25 (27.5%)34 (37.4%)25 (27.5%)7 (7.7%) Cough and phlegm27 (29.7%)37 (40.7%)20 (22.0%)7 (7.7%) Cough, phlegm and nasal secretion21 (23.1%)38 (41.8%)24 (26.4%)8 (8.8%) Rapid breathing25 (27.5%)31 (34.1%)24 (26.4%)11 (12.1%) Indrawn chest43 (47.3%)18 (19.8%)15 (16.5%)15 (16.5%) Rapid breathing and indrawn chest39 (42.9%)19 (20.9%)14 (15.4%)19 (20.9%) The last three cluster of illness sy mptoms, of most importance to the study, were ranked in the following fash ion: rapid breathing was deemed not serious to moderate by 62 percent of respondents, and as serious to emergency by 38 percent of women su rveyed; indrawn chest was ranked as not serious to moderate by 67 perc ent of respondents and as serious to emergency by 33 percent; and finally, rapid breathing and indrawn chest were ranked as not serious to moderate by 64 percent of respondents, and as serious to emergency by 36 percent of respondents. In contrast to the other illness symptoms, for the la st three questions on symptom s, the answers in the not serious to moderate categories were weighted more towards the not serious response category. After these series of questions, respondents were asked to look at a photo of a child with an indrawn chest, taken from an educational CD-ROM on ARI (The_Wellcome_T rust 2001). Only seven percent
197 of respondents identified the photo correctly as either a respiratory ailment or needing a doctors immediate attention. The remaining 93 percent responded that they did not know what the problem was or thought the sign was a result of a fall and that the child had possibly suffe red a broken rib. The descriptive statistics results suggest that there was poor recognition of the severe signs of ARIs. There was greater recognition of the more common signs of colds and flus. Diet of Child in the Last Twenty-four Hours Undernutrition affected roughly a third of the children in Penipe County in October 2000, a year afte r the eruption of Mount T ungurahua, partly due to the disruption of agricultural production in t he sector (Lara, et al. 2000b). An unpublished study carried out by CEBYCAM in 2001 f ound that poor nutrition affected 43 percent of childr en under 14, and if only ch ildren less than six years old were considered, the figure of under nourished children rose to 64 percent. From a survey of 362 school children in Penipe County, daily consumption of foods included mainly sugars (empty calo ries), tubers (mostly potatoes), and oils (such as lard and cooking oils). The children also consumed a high quantity of corn and peas. Bananas were the mo st commonly eaten fruits. The study also found that milk and eggs were cons umed daily; however, daily consumption of meat and fish was rare because t hese and other products could only be obtained at the weekly market. The main diet of people of Penipe is based on corn and pasta, while the ingestion of protei ns is scarce. According to the chief
198 doctor at CEBYCAM, there also exists a cultural barrier that adults should eat better than children (Lara, et al. 2000b). The nutrition of people in the zone has been negatively affected by the volcanic activity of Tungurahua, which has destroyed and damaged crops, and diminished incomes. One of the last secti ons of the semi-structured questionnaire asked respondents to report their childs dietary in take in the last twenty-four hours. This section yielded quantit ative data in the form of nutrient and micronutrient frequencies in order to calcul ate the childs daily nutriti on intake. The researcher recorded the time and place the food or beverage was consumed, the type of food or beverage, the brand if applicable, and the quant ity, which was measured using a typical bowl and cup. The bo wl measured 21 oz./650 ml, and the cup measured 12 oz./375 ml. The researcher would ask for example the quantity of tea the child consumed, and if the responde nt answered half a cup, after being shown the cup that the researcher carried with him, the quantity would be registered as six ounces for data entry pur poses. Efforts were made to make a second visit to each family to collect a se cond dietary recall afte r eight days. If, for example, the child were sick on the fi rst visit and not eating properly, a more exact dietary record would be collected the following week. A typical breakfast for a child was an oatmeal beverage with milk and sugar, juice, coffee or tea, bread, and fruit. For lunch, the child usually ate the same meal as the adults, albeit smaller por tions, such as a potato or yucca soup appetizer before the main course, chicken or beef with rice, beans and vegetables. Lunch was the main course of the day. For dinner, the child
199 generally ate soup, as well as another helping of oatmeal, and either soda or tea. Between meals, some children occasionally ate ice cream, chochos and mote (beans and pealed white corn, respective ly), fruit, oatmeal, or an occasional soda. The childrens diets were high in proteins from the milk, beans and chicken, and in carbohydrates fr om the potatoes and plantains. The diet of the family consisted ma inly of rice, potatoes, pasta, beans, chicken, and vegetables. Eighteen differ ent varieties of soup were recorded, consisting of different combinations of vegetables (corn and green peas), rice, legumes, grains, beef, chicken, fish, pot atoes, and pasta. One soup particular to the Andes is called locro a potato soup with eggs and cheese, and some recipes include turnips. Another popular soup is known as sancocho made from plantains, yucca, beef, and peas. A popular food for children is known as colada or porridge, and consists of grains mixed with sugar and milk. Different varieties of coladas were made from bean, oatmeal, cornmeal, barley, and quinua. Popular main courses included beef or chicken stew, potato omelette, chicken with rice, and fried fish, chicken or pork. In times of celebration or festivals, popular dishes include potatoes and peanut sauce with roast cuy (guinea pig), fritada (roast pig), and llapingachos (fried potato patties with eggs, cheese, and sausage). In this study, there are four group s of children: under six months (17 children); six to 17 months (14 childr en); 18 to 29 months (26 children), and 30 months to under five years old (34 childre n). No dietary records were collected for the children in the infant baby ca tegory, who were almost exclusively
200 breastfed. That left a total of 74 children with dietary information. After entering the daily diet for the 74 children for the twenty-four hour period, Nutritionist Pro software allows the user to generate a diet ary analysis for each indi vidual client. The dietary analysis includes a br eakdown of macronutrients, amino acids, vitamins, and minerals. The macr onutrient data extract ed from the dietary analysis included carbohydrates, kilocalories fats, and proteins The collection of macronutrient data is standard for nut ritional analysis to gauge the energy and general nutritional value of foods ingested. The micronutrient data extracted, based on the ARI literature, were for vitami n A and E. Vitamin A forms a family of fat-soluble compounds that help to regulate the immune system, as well as maintain the healthy lining of the respirat ory, urinary, and intestinal tracts. Vitamin A is a more likely micronutrient deficiency than others because it is not widely distributed across different foods. Vitamin E regulates the immune system and acts as an antioxidant, protecting cells from free radicals. The minerals data, also collected based on the ARI literature, were extracted for zinc and selenium. Zinc is an essential mineral for the body and supports a healthy immune system. Zinc deficiency can lead to observable signs such as growth retardation, diarrhea, and loss of appetite. Selenium is a trace mineral with antioxidant properties, aiding immune functions, and is r equired in small quantities to ensure good health. The nutrition data were analyzed using the statistical software, SPSS for Windows, Release 13.0 (SPSS 2005). Of the 74 children with dietary data (the other 17 children in the sample were excl usively breastfed re ferred to as Group
201 1), there were 14 children in the 6 to 17 month age category (Group 2); 26 children in the 18 to 29 month age cat egory (Group 3); and 34 children in the 30 month to 5 years old age category (Group 4). Of these 74 children, 50 (67.5%) children had two days of dietary data colle cted. For these 50 cases, the average of the two days of nutritional data was used for statistical analyses. The distribution of the data was tested for no rmality, and descriptive statistics were generated. Because the macronutrient and micronutrient data were not normally distributed, nonparametric statistical test s were employed, and because of the small sample size, outliers were not excl uded. The mean valu es of the nutrients were compared with recommended daily allowa nces (RDA) for the three different age groups.10 For the purposes of dietary in take comparisons, the RDA dietary goal was used instead of the 1990 WHO Study Group proposed population nutrient goals which are expressed as ranges, rather than exact values, and focus on macronutrients (Truswell 1999). After these values were compared with the nutrition goals, nonparametric tests we re performed to determine if there were differences in nutrition intake when stratified by age groups, gender, and SES. The results of these comparisons an d statistical tests are presented in the next chapter under data analysis. The followi ng table is a summary of the results of the nutritional intake fo r the children in the study from all three age groups compared to the RDA values for a three-y ear-old child in the United States. 10 The RDA values were taken from an inside cover table in Whitney and Rolfes (2005) and are based on average daily intakes of energy and nutrients which would be considered sufficient for health in the U.S. The main problems in applying RDAs to Ecuador are the scarcity of food composition data and the range of nutritional status (undernutrition in the poor and overnutrition in high SES groups) (Truswell 1999).
202 Table 12: Daily Nutrition Intake for Ch ildren (6 Months to < 5 Years Old) n = 74 (day 1) kcal/daycarbsfatproteinvit. Avit. Ezincselenium Mean 1133.44176.9731.0542.09597.453.255.5741.14 Std. Deviation 415.5365.6519.2519.79501.502.342.8019.35 n = 50 (day 2) Mean 1074.67168.2629.0539.41520.373.105.0837.15 Std. Deviation 435.2765.2117.2517.80401.141.752.6916.08 n = 74 (day 1 & 2) Mean 1100.87172.5632.4340.71593.843.295.3540.18 Std. Deviation 363.6258.2925.4115.26421.832.241.9915.43 RDA Goal (Children 1-3) 1300.00130.0030.0016.00300.006.003.0020.00 = g/day; = g/day, [vitamin A unit is termed retinol equivalents (RE) = 1 g of alltrans retinol or 6 g beta-carotene or 12 g of other vitamin A precursor caro tenes]; = mg/day [RDA values based on National Academies of Sciences (2004)] According to the table, the average childr en in the study were meeting their daily nutritional requirements for the macroand micronutrients selected for analysis, with the possible exception of vitamin E, which was lower than the RDA goal. The data is presented for the first day of dietary data (n =74), the second day of dietary data (n=50), and the two days of dietary data co mbined (n=74). A second dietary recall was carried out with 67.5 per cent of the respondents of the semistructured questionnaire.
203 PART V QUALITATIVE DATA FR OM STRUCTURED QUESTIONNAIRE In addition to the primarily close-ended questions, there were sections of the questionnaire that were open-ended and generated a vast amount of qualitative data. These data fell into five main categories: use of local healthcare facilities and resources, knowledge of hot and cold coughs, use of curanderos, and knowledge and use of home remedies. The qualitative data were rich and some categories were specific enough to be coded and converted into categorical data. This sect ion will begin by explaining the variation in use of local healthcare facilities and barriers to doctor access. Use of Healthcare Facilities Because there were three major site areas where the questionnaires were administered, there was considerable variati on in the use of healthcare facilities. The three government health outposts us ed by questionnaire respondents were the main center in Penipe (SCS Peni pe), a secondary health center in Matus (SCS Matus), and an ancillary post in Palictahua (PS Palictahua). There was also the private health clinic at CEBYCA M, an arm of the San Francisco Catholic Church of Penipe. At C EBYCAM, there were around tw enty to thirty patients a day, half of whom were children. At SCS Penipe, the average was higher, at thirty to forty patients daily, a manageable number for the clinic. The other two rural health posts attended fewer pati ents because of their limited hours and smaller area population. The highest number of patients seeking attention in
204 Penipe occurred on Sunday, market day. Waiting times were around 15 minutes, and patients at all clinics received betw een fifteen minutes to a half an hour of attention. CEBYCAM had a psychologist, gynecologist, pediatrician, dentist, and laboratory technician. All four doctors were well trained and experienced. There was also a professional nurse, a pharmaci st, and a nurse assistant. SCS Penipe had one doctor and one dentist carrying out thei r year of rural service without the advantage of a laboratory. There was also a health inspector, two nurse assistants, and a dentist assistant. W hen the SCS Penipe doctor completed her year of rural service, a permanent doctor was assigned to the clinic, and she had previously worked there a year befo re. SCS Matus and PS Palictahua each had a female doctor performing their year of rural service, and each had a nurse assistant. The nurse assistant at SCS Matus was studying to become a professional nurse. The aim of the question was to gauge mo thers opinions of the healthcare facilities in order to identify potential barriers of access. The major barrier identified was money for medicines and tr ansportation [52.8% of all respondents (medicine 31.9%; transportation 20.9%)]. However, while money for medicines was a complaint equally distributed across the three surveyed areas, money for transportation was primarily a complaint of Matus area residents (2:1 compared with Puela and 12:1 compared with Penipe). Because the SCS Matus hours were not as frequent or regular as SCS Penipe, Matus area residents would have to travel to Penipe for healthcare or
205 Riobamba for a pharmacy (46.7% of Matus residents reported using SCS Penipe; 36.7% using CEBYCAM; and 46. 7% using a pharmacy). Puela area respondents reported a differ ent proportion (51.7% repor ted using SCS Penipe; 56.7% using CEBYCAM; and 63.3% us ing a pharmacy). Penipe respondents used the health centers in their town mo re than the other two areas which used PS Palictahua for the Puela area resident s and SCS Matus for the Matus area residents. In Penipe, 87.1 percent reported using both SCS Penipe and CEBYCAM, and 41.9 percent visited pharmaci es. The proportion of those using private doctors in Riobamba was the follo wing: in Penipe two respondents (6.5% of Penipe respondents); in Puela one respondent (3.3% of Puela respondents); and in Matus two respondents (6.7% of Matu s respondents). Overall, only five percent of respondents report ed using a private doctor, so according to the data, most people with young childr en use either the local go vernment health centers or CEBYCAM, depending on whether they live in Penipe. Table 13: Use of Treat ment Facility by Area Category of Treatment FacilityPenipePuelaMatus CEBYCAM87.1%56.7%36.7% SCS Penipe87.1%51.7%46.7% SCS Matus0.0%0.0%100.0% PS Palictahua0.0%100.0%0.0% Pharmacies41.9%63.3%46.7% Private Doctors6.5%3.3%6.7% One of the barriers identified in Penipe and Puela was the sometimes limited and erratic hours of t he government clinics (14.3% of all respondents, 6 in Penipe, 6 in Puela, and 1 in Matus). Si nce all but one of the health staff of the government health posts lived in Riobamba, if for some reason they did not work
206 on a particular day, the post remained clos ed; moreover, sometimes the hours of operation were both limit ed and confusing, such as regular hours on Monday, Tuesday, and Thursday, from 8 am to 1 pm.. For the Penipe area, people used both CEBYCAM and SCS Penipe. Ev eryone would have to go to the government health posts for childhood vacci nations, which were not offered at CEBYCAM. In Puela, Palictahua and Pach anillay, there was more difficulty in accessing health services because of the restrictive hours of the local health outpost. There was planning to establish a puesto de salud (health outpost) both in El Altar and Puela to resolve this problem.11 In the hamlet of Palictahua, there was a small health outpost (PS Palict ahua) that people in the area used primarily, or if the outpost was closed, the mothers woul d bring their children to SCS Penipe. In Matus, they had their own government health center (SCS Matus), so people would use this resource, however the hours were often erratic. In addition to the health centers, people had the option of going to pharmacies in Riobamba, or to use a curandero or her balist for folk illnesses. For each resource, the respondent answered the la st time she had used the resource for her child and for what purpose. The conclusion of the Development Plan for Penipe st ates that the problem of quality and r egularity of service in the go vernment health clinics was more of a problem than the absence of services, thus corroborating the field observations (COSUDE 2003). The r eport quotes an unknown source from Matus that the health services are defici ent in the case of emergencies, since 11 The buildings were in place, they were just waiting for the equipment.
207 there are no services later in the afternoon or in the evening. This complaint was echoed by one respondent in Matus. Subcentro de Salud Matus, un mes, esta ba enferma de la garganta, le dio fiebre. No atiende todo el da, hasta medioda no ms. [SCS Matus, one month ago, she was sick in her throat, and had a fe ver. They do not attend all day, until noon only]. Attitudes towards Healthcare Facilities The attitude of the female caregi vers towards CEBYCAM (95%), SCS Penipe (82%) and the other government health cent ers (81%) was largely positive. Of those who received serv ices at CEBYCAM, there was a higher proportion of positive responses for the facility than those who went to SCS Penipe because the doctors were more experienced and the treatment more comprehensive, since they had laboratory facilities as well as a psychologist. The doctors at the government clinics were typically do ctors performing their one year rural service requirements whereas the doctors at CEBYCAM were hired as permanent staff. However, the respondent s also complained that the medicines were expensive in the CEBYCAM pha rmacy (the only pharmacy in Penipe County). The negative responses towa rds pharmacies in Riobamba also referred to the high cost of medicines. Those who went to CEBYCAM generally did not go to SCS Penipe, except for vaccinations. When giving reasons for visiting healthcare facilities for thei r children, stomach parasites figured
208 prominently as the primary reason, as well as general control measures for young children. Table 14: Attitudes towards Healthcare Facilities CEBYCAMSCS Penipe Other* Pharmacy negative 1 (1.1%)5 (5.5%)3 (3.3%)9 (9.9%) positive 52 (57.1%)46 (50.5%)50 (54.9%)24 (26.4%) neutral 2 (2.2%)5 (5.5%)9 (9.9%)13 (14.3%) not applicable 36 (39.6%)34 (37.4%)29 (31.9%)45 (49.5%) private doctor 5(5.5%); PS Palictahua 32(35 .2%); SCS Matus 25(27.5%); n/a 29(31.9%) There were only a few complaints of SCS Penipe. One described a problem with a dentist who supposedly carried out unnecessary tests and was not well trained. There was also one complaint against a nurse who, according to the respondent, did not know how to treat people, but the same person said she liked SCS Penipe because the medicine s were free. Another complained that they used to give medicines for free, but now they charged a dollar, and fifty cents for the consultation. In one case, a respondent co mplained that the doctor at PS Palictahua gave her a medicine for her childs parasites that worsened the childs condition. In yet another case, a mother reported t hat the pediatrician referred her to a hospital Riobamba because her child had pneumonia. Barriers to Care Women talked about barriers to care. The number one reason cited as an obstacle to seeing a doctor was lack of money for transportation (20.9%) and to purchase medicines (31.9%). A few peopl e also mentioned transportation issues
209 in the evenings and the restri ctive hours of the clinics as barriers (14.3%). For example in Penipe, one complaint was t hat in SCS Penipe, they would attend people coming in from rural areas on S unday, the market day, so people who lived in the town of Penipe only re ceived attention from Monday through Thursday, since the clinic was closed on Friday and Saturday. Another related issue was the complaint that the health centers were too fa r from their homes (6.6%). Two people (2.2%) mentioned t heir long work hours as barriers to seeking care, and two people (2.2%) mentio ned that it was a barrier when the clinic employees were on strike. During the fieldw ork session, there was a national nurse assistants strike that lasted an entire month. Fifteen people (16.5%), primarily from Penipe and Matus, did not identify any barriers to seeking a doctors care. Hot Cough versus Cold Cough From the focus group discussions, a category emerged termed hot coughs and cold coughs. These are et hnomedical terms to describe a type of cold that children experience. There was some consensus on the definition of the difference between the two types based on the fifty-eight respondents (63.7%) who had knowledge of the differ ence. First, cold coughs were considered more severe than hot coughs; moreover, the causes and cures of the two coughs were different. Cold coughs were caused by the cold weather and the wind, and the remedies were to keep warm and drink particular herbal remedies. Hot coughs were caused from spending too much time in the sun, and
210 the remedies included bathing in temper ate water and drinking cold beverages. Two herbs, tilo and borraja were frequently ment ioned as proper herbal treatments for cold coughs.12 The following three quotes illustrate some of the individual variations in treatment strategy. Tos fra, resfriada, es de calor ataca lo s pulmones. Flota mentol chino, funda de gallina, pone en la espalda. Tos calient e, aguitas frescas, poniendo con limn [Cold cough, get a cold, is when the heat attacks the lungs. With Chinese menthol, a bag of chicken br oth, you put on the back (of the child). For hot cough, cool liquids, with lemon]. Tos fra, peligroso resfriada, jarabes, cons igue borraja, tilo. Tos caliente, tos, lo ms problema, mora comuna, cocina bi en, pone borraja, agua caliente, tos fra, ajo mochacado con leche [Cold cough, dangerous sickness, infusions, get borraja, tilo Hot cough, cough is the problem, mora is common, cook well, put borraja hot water, for cold c ough, ground garlic with milk]. Es de fro, le da limonada con trago (solo tosiendo), es de calor, baar con trago manteca de cacao en la barriga, bastant e calor sufre mucho, queda con aguas frescas [For the cold cough, give lemonade with whisky ( only if coughing), if it is a hot cough, bathe with whisky, cocoa butter on the stomach, if very hot and you are suffering, give cool liquids]. 12 Tilo pertains to the family Tiliaceae in the gen us Tilia. Its flowers and stems are used as sedatives. It is used as a sudorific for coughs, colds and fevers, and as an expectorant.
211 Cold cough was viewed as more se rious in the ethnomedical model and the aim was to heat the body. The opposit e was true for hot cough, although in one case a person prescribed a hot herbal remedy for hot cough. Also interesting is that all cures use herbal remedies or baths, not drugs from the pharmacies. The following table summarizes the number of respondents who had knowledge of hot coughs and cold coughs Fifty-eight respondents (63.7%) knew what the terms meant and the diffe rences between them while 33 (36.3%) did not. In general, those who were kno wledgeable of hot and cold coughs also had a more extensive knowledge of medicinal plants and herbs. Table 15: Hot and Cold Cough Knowledge frequenc y percent y es 5863.7% no 3336.3% Total 91100.0% Use of Curanderos As with knowledge of hot and co ld coughs, some people were more knowledgeable about seeing a curandero. Curanderos are natural healers who heal the body and mind through magical means by working in the spirit world. These natural healers were both male and female, although the gender of the healer reported in the ques tionnaire was not quantifi ed. Some people had taken a child to see a curandero, but were unf amiliar with what was treated or why. Also included in this class of healers are fregaduras (bone-setters). Twenty-one
212 respondents (23.1%) had taken their child to see a curandero. They usually talked about a limpieza or cleansing, which cost around ten dollars. To cleanse someone of illness, the curandero would pass a cuy (guinea pig) and medicinal herbs over the body.13 Some of the folk illnesse s that curanderos would cure were mal aire (bad air), mal humor (bad fluid), arco iris (rainbow), and espanto (fright). An illness such as mal aire which is generally believed to be transmitted through the air from the dead to the living, is a long-st anding traditional belief in Andean cultures (McKee 2003:131). All respondents reported a positive experience with the curanderos, who had all reportedly cured their children. The following quote illustrates an illness episode of bad air for a child. Vive acacito, mal aire les da al nio, dos meses, para pequeo, limpia con unas plantas bendecitas, bendicin, si es bueno, limpia nios comen mejor, no cuesta nada, lleva plantitas [He lives close by, my child got bad air, two months old, he was cleansed by blessed plants, received a benediction, yes it is good, he cleans children, they eat better, it is free, you just have to br ing the plants (that they use for the healing)]. While mothers did not seek the hel p of the curandero in the case of respiratory infections, apparently this was not the case with diarrhea. The following quote illustrates this point. 13 More information is presented on curanderos in the key informant interview section in the beginning of the chapter.
213 Curandera, seis meses, tena mucha dia rrea, se compuso. Si es bueno, es de Penipe [Woman healer, my child was six m onths old, she had a lot of diarrhea, she got better. Yes the healer is good, she is from Penipe]. There is a generational difference in the belief of curanderismo. The younger generation is more skeptical r egarding the use of curanderos, illustrated by the following statement: Una vez, estaba espantada. Limpieza, no creo en estos cosas, ms para mi mama [One, time, he receiv ed the fright. After a cleansing, cured, I dont believe in these things, I did it because my mother does]. The small sample shows that curanderismo is alive and well in Penipe County. Twenty-one respondents (23.1%) had used curanderos to treat their children. Considering that curanderos ar e more often employed to cure folk illnesses in adults than children in th is area, the proportion would have been higher had the respondent been asked if t hey had seen a curandero for anyone in their family, rather than only for their child. Table 16: Use of Curandero/a for Child frequenc y percent y es 5863.7% no 3336.3% Total 91100.0%
214 Home Remedies All respondents were asked to give an inventory of the medications they keep in the home, whether herbal, over-thecounter, or prescrip tion medicines. Oftentimes, for herbal remedies, the fa milies maintained herb gardens around their home, and they would pick the her bs as needed. Some people responded that they did not have thes e herbs, but they could acqu ire them from a neighbor. Most herbal remedies were used daily in teas to aid digestion and ease cough symptoms. For the medications, the re spondent would often go to where they stored the medicine and show the product so the researcher could accurately write down the name. Many of the medications were obtained in the government health centers without cost. For each medicine or herb, t he question was posed as to the use, cost, and last time used. When asking informants for a medici ne inventory, the answers were extremely diverse. Some people mentione d hot cloths as a remedy as well as particular herbs to cure folk illnesses. Of the 129 different herbs and medicines mentioned, 60 were either over-the-c ounter or prescription medicines and 69 were herbal remedies. The tw o most common herbs used were manzanilla [chamomile, 48 (52.7%) respondents], or egano [26 (28.5%) respondents] and toronjil [19 (20.9%) respondents], all used to aid the stomach in digestion.14 The most common cold remedy was Tempra, a cough and fever suppressant 14 Chamomile is considered good for the digestion, by combating gases in the intestinal tract. It is taken in the form of tea. The plant has red l eaves and yellow flowers. Oregano, also known as wild marjoram, has small leaves and pink flowers. The leaves and flowers are used as tonics, condiments, and perfumes. The plant is also used as a mouthwash to fight mouth and throat infections. Toronjil, like manzanilla or chamomile, is also used in teas. It is a herbacious plant and has pink flowers and a smell similar to lemon. It is used to treat gout and urinary tract infections.
215 available in either tablet or syrup form [31 (34.1%) respondents]. Tempra was so common because the health centers gave this remedy away for free in the tablet form (the syrup version was available for around a dollar). The next most common remedies were tilo [17 (18.7%)] and borraja [17 (18.7%)], herbs used to cure cold coughs, as previously mentioned. Other herbs that were very common were llantn [16 (17.6%)], malva [14 (15.4%)], menta (mint) [14 (15.4%)], cedrn [10 (11.0%)], mora (raspberry) [10 (11.0%)], poleo [9 (9.9%)], and eucalipto (eucalyptus) [8 (8.8%)].15 These herbs were used in teas to cure coughs, colds, colics, and other stomach ailments. Antibi otics were present in a few peoples homes, such as amoxicillin [3 (3.3%)] and ampicillin [3 (3.3%)]. Two respondents (2.2%) reported having Ventolin asthma inhalers in their home for their child. Finally, a handful of respondents report ed having medicines to cure folk illnesses. These remedies included ruda (unknown root), guantu blanco (morning glory), and huevos (eggs).16 The eggs were not eaten, but passed over the body of the sick person by a curandero in order to have the contents of the egg examined for diagnostic purposes. 15 Llantn pertains to the genus Plantago, of the family Plantaginaceae. It is a herb that is used as a tea against chest problems and chronic dysentery. Malva is an herb of the family Malvaceae. Its infusions are used against bronch itis. Menta, or Mint, is an herb used as a stimulant and anti-spasmodic, as well as a condime nt and flavoring. Cedrn is an arboreal plant with aromatic fruits and leaves. Infusions of the plant are used to treat stomach and chest ailments such as asthma. Mora, or blackberry, is used to make drinks and teas. It is used as a purgative. The leaves of the eucalyptus tree ar e used to treat respiratory problems for its decongestive properties. The herb poleo was not able to be identified. 16 According to a pamphlet on natural medicines, ruda is a well-known plant whose infusion is used to wash the vagina to initiate menstrual flow.
216 Table 17: Commonly Used Remedies Name of RemedyNumber chamomile48 (52.7%) Tempra31 (34.1%) oregano26 (28.5%) toronjil 19 (20.9%) tilo 17 (18.7%) borraja 17 (18.7%) llantn 16 (17.6%) malva 14 (15.4%) mint14 (15.4%) cedrn 10 (11.0%) raspberry10 (11.0%) poleo 9 (9.9%) eucalyptus8 (8.8%) amoxicillin3 (3.3%) ampicillin3 (3.3%) The respondents were also asked the pr ice of the remedy. All the herbs were acquired without cost. The most ex pensive medications were the vitamins, which cost around ten dollars. Ventolin the asthma drug, cost around eight dollars. The next on the list were t he cough syrups, which cost between three and six dollars. Antibiotics were not expens ive, ranging from one to two dollars. People would generally go to Riobamba to purchase these remedies and find the pharmacy with the best price. The combination of qualitative a nd quantitative research methods produced a rich, multi-layered matrix of data. The strategy of beginning the research with the focus group discu ssions aided the researcher in the development of a culturally-sensitive semi -structured questionnaire which elicited emic information from respondents. By co llecting the local terminology in these focus group discussions, questions in t he semi-structured questionnaire were
217 tailored to elicit emic categories, such as the differences between hot and cold coughs and local terminology for common childhood illnesses. Meticulous care was taken to accurately record responses that emerged from all the ethnographic tools used in the dissertation research. The focus group sessions were a pr oductive qualitative data gathering exercise. The focus groups established a cultural knowledge domain regarding types of respiratory illnesses, health care-seeking strategies, home remedies, and barriers to seeking care. The major bar rier identified to seeking medical help was lack of financial resources, and this was corroborated by the semi-structured questionnaires. Moreover, an ethnomedical model of tr eatment for respiratory illnesses emerged from the focus groups which included assessment, treatments with medicinal herbs, and decisions on whether or not to seek medical help. The mothers in the focus groups also acknowledged that they needed more education and training in order to make in formed treatment decisions relating to their childrens respiratory infections. In the in-depth interviews, the healthcare workers reiterated this concern and bel ieved that if the people were better educated on symptom recognition, they woul d not delay seeking care for their ill children. Nevertheless, the healthcare workers identified education as the primary barrier for the mot hers to seek medical care, whereas the mothers cited lack of money and limited hours of the clinics as the primary barriers. In the next chapter, statistical test s from data gathered from the semistructured questionnaires dem onstrate the positive asso ciation of low SES with frequency of respiratory infections, and the negative association between low
218 SES and timely health care-seeking behavior. Moreover, the results of the semistructured questionnaire found a generalized lack of rec ognition of the primary signs of ARIs. Through t he explanation of the res earch analysis, research hypothesis results, and discussion, a more in-depth understanding of the significance of the research results are presented in the following chapter.
219 CHAPTER SEVEN: ANALYSIS, TESTIN G THE RESEARCH HYPOTHESES, AND DISCUSSION PART I ANALYSIS In order to justify perfo rming statistical tests on the study data taken as a whole, it was necessary to establish that the three groups of geographicallydispersed respondents (Penipe, Puela, and Matus) were similar for certain key demographic variables. Because of the sma ll sample size for the three different areas of 30 respondents each, the validity of the result s from the data analysis would be strengthened by the larger sample size of 91 respondents. A KruskalWallis test, a nonparametric equivalent of the one-way ANOVA, was performed to determine if there were any signific ant differences between the means for the respondents age, salary, and years of educat ion for the three areas. The test determined that the differences for two of the variables were not significant (age: (2) = 4.24, p > 0.05, two-tailed; education: (2) = 2.73, p > 0.05, two-tailed). The mean ranks for age and education indicated that the average age was highest in Puela, and the av erage number of years of education was highest in Penipe. However, the only statistically significant result was for mean salary between the three locations. This was to be expected since most of the
220 professional occupations we re located in Penipe, and a few high salaries inflated the mean (salary: (2) = 10.23, p < 0.05, two-tailed) The mean rank for salary in Penipe was significantly higher than Matus, which was in turn higher than Puela. Nevertheless, statistical test s were performed on the whole data set, rather than comparing thes e three groups separately, because a larger sample size increases the validity of statistical re sults. The average salary for the region was $127 per month, with a median of $100 per month, and as the following histogram demonstrates, the salary number s were not evenly distributed (p < .0001). Monthly salary (U.S. dollars)500.0 450.0 400.0 350.0 300.0 250.0 200.0 150.0 100.0 50.0 0.0Monthly Salary of Penipe FamiliesFrequency30 20 10 0 Std. Dev = 110.10 Mean = 126.7 N = 91.00 2 2 3 2 14 10 24 20 12 Figure 12: Monthly Salary of Penipe Families After the descriptive statistics for the semi-structured questionnaire were tabulated, crosstabulations of variabl es of interest were performed with
221 appropriate chi-square statisti cs calculated. Next, a binom ial logistic regression procedure using the forward stepwise method was employed to determine if socioeconomic status (SES) variables c ould improve the predictive accuracy of reported care-seeking behaviors, as de scribed in the following section on research hypotheses results. In the semi-structured questionnaire the most frequently mentioned barrier to seeking care was lack of money. Agricultural work contributed to this problem because of the long hours involv ed in this activity and limited hours of the clinics. Low possession score was highly associated with not owning land ( (1) = 15.90, p < 0.001, tw o-tailed). Therefore, t hese families survived by working on rented land, land owned by ki n, or land owned by an employer. Therefore, because of high work demands fr om agriculture, the first hypothesis of the dissertation research was that fa milies of low SES, as measured by possession score, would be more likely to treat respiratory infections at home or do nothing. While the overall possession sc ore did not turn out to be the most significant variable in the results, t he individual compon ents of the possession score indicated some trends in favor of the hypothesis. Families with below average salaries ($150 per m onth and less) were more likely to treat a child with cough, fever and nasal secretion at home or do nothing ( (1) = 6.40, p < 0.05, two-tailed). When families with salaries of over $150 per month were compared with families earning $75 per month or less, the difference was even more significant with mothers more likely to treat a child at hom e when presenting with cough, fever and nasal secretion ( (1) = 6.96, p < 0.01, two-tailed). Since
222 salary, as the key measure for SES, was a significant factor in the bivariate tests for treatment seeking, the other eight SES indicators were used in the initial binomial logistic model. In addition, so me of the other SES indicators were associated with negative resp iratory illness outcomes. Care-seeking for Cough, Fever and Nasal Secretion by Monthly Salary 0% 20% 40% 60% 80% 100% home remedies or nothing Doctor under $150/month over $150/month Figure 13: Care-seeking for Cough, Fe ver and Nasal Secretion by Monthly Salary For example, families with dirt floors were more likely to have children with coughs and colds ( (1) = 4.22, p < 0.05, two-tail ed). Families with less than three agricultural animals were more likely to have their youngest child hospitalized for a chest illness ( (1) = 5.64, p < 0.01, two-tailed). Finally, families with a low possession score were more likely to have a child with a wheeze ( (1) = 7.35, p < 0.01, tw o-tailed). Therefore, low SES, as measured by salary, housing quality, and possession score, was associated with a medical history of respiratory problems in ch ildren and a lesser inclination to seek appropriate medical care for particular symptoms. Even though single female-
223 headed households had lower incomes t han married or open union households, there was little measurable effect on AR I symptom recognition or treatment choice. Contrary to expectation, fo r the symptom of indrawn chest, single mothers were more likely to seek a doctor than married or open union women ( (1) = 3.68, p = .05, two-tailed). In sum, household income and possession score exhibit some differences in care-seeking and respiratory illness frequencies. The female caregivers education level and the childs gender had little effect on care-seeking behavior for childhood symptom clusters in the bivariate statistical analysis. For the specific symptoms of ARIs, t here was an overall lack of knowledge of rapid breathing and indraw n chest, or the significanc e of the combination of the two. Only seven percent of respondent s correctly identified a photo of a child with an indrawn chest as a respiratory ailment or in need of a doctors attention. The lack of symptom recognition is illust rated in the following table and figure which show that more than half of t he respondents (54.9%) did not rank the symptoms of ARI either as serious or emergency, while 29 percent ranked the symptoms of rapid breathi ng and indrawn chest correctly as serious to emergency.
224 Knowledge of ARI Symptoms11% 6% 54% 29% none little some a lot Figure 14: Knowledge of ARI Symptoms Based on the answers to questions relati ng to the severity of nine clusters of symptoms, ranging from cough and feve r to indrawn chest and rapid breathing (refer to Table 11), a score was generated to tabulate the number of incorrect responses versus correct responses, scored as 0 or 1, respectively (compare with Denno, et al. 1994:299). For the pur poses of creating a range of scores, each question received the same weight ing, even though the researcher was most interested in symptom recognit ion and appropriate tr eatment for acute lower respiratory infections. The mean score for all respondents was 5.2 out of a possible score of 9. Figure 15 illustrate s the range in scores for the outcome of this computation.
225 Percentage for each score on level of symptom severity(9 is a perfect score) 0% 5% 10% 15% 20% 25% 30% 0123456789 Figure 15: Percentage for each scor e on level of symptom severity There was a generalized lack of rec ognition of appropriate biomedical treatment choice for the last three sympt om clusters related to severe ARIs. Nineteen percent would do nothing if t heir child presented with both indrawn chest and rapid breathing, and 43 percent th ought these signs were not serious. Therefore, there was a la ck of connection between perception and care-seeking. The following pie charts (Figures 16 18) illustrate this discrepancy.
226 Rapid Breathing Health Care-seeking Behavior 11% 6% 83% nothing household remedies doctor Figure 16: Rapid Breathing Health Care-seeking Behavior Indrawn Chest Health Care-seeking Behavior26% 7% 67% nothing household remedies doctor Figure 17: Indrawn Chest H ealth Care-seeking Behavior
227 Rapid Breathing and Indrawn Chest Health Care-seeking Behavior19% 3% 78% nothing household remedies doctor Figure 18: Rapid Breathi ng & Indrawn Chest Health Care-seeking Behavior According to the figures, rapid breath ing was a more recognizable sign of respiratory infection than an indrawn c hest (83% seeking a doctor for rapid breathing compared to 67% for indrawn chest). Oftentimes, when shown the picture of the child with an indrawn chest, the respondent thought the sign was an injury caused by a fall. For both symptoms, the number of respondents who would do nothing outnumbered those who would treat the illness with household remedies. These data are sharply c ontrasted with the number of respondents who would do nothing regar ding the other illness symptoms mentioned in the same part of the questionnaire, such as cough, fever, or nasal congestion. In the case of all other illness symptoms related to coughs and colds, very few respondents answered that they would do nothing if their child presented with these other illness symptoms. Another interesting finding was th e lack of connection between peoples perception of severity of rapid breat hing and indrawn chest symptoms, and their approach to seeking care for these sym ptoms. For the symptom of rapid
228 breathing only, the chi-square statistic wa s not significant between perception of severity and treatment choice. However, for indrawn chest there was a lack of connection between percept ion and treatment ( (1) = 10.68, p < 0.01, twotailed). Furthermore, for the aggregate of symptoms indrawn chest and rapid breathing, there was also a lack of c onnection between perceived severity and treatment choice ( (1) = 7.65, p < 0.01, two-tail ed). This suggests that the symptom of an indrawn chest was not well understood, and that while the symptom may have been perceived to be serious, fewer respondents were inclined to seek a doctors attention t han for the symptom of rapid breathing. Crowding is another risk factor for ARIs. The degree of crowding in household was determined using the co mbined data between the number of rooms in the house and the number of dependents. One-room households, tworoom households with four or more dependent s, three-room households with five or more dependents and households with si x or more dependents were coded as crowded because there would be a least two or more dependents living in a room. This resulted in 24 crowded hous eholds and 67 regular households. However, when the chi-squar e statistic was calculated comparing the variables for history of childhood illnesses and respiratory problems with crowded households, the results were not signifi cant. When the variable for crowded households was compared with possession sco re, the chi-square statistic was borderline significant ( (1) = 3.68, p = .05, two-tail ed). Almost half (47.3%) of the households that were classified as not crowded ranked on the high end of the possession index. In additi on, more than twice as many households with lower
229 than average incomes had households with five or more dependents than households with higher than average incomes, suggesting that lower income households tended to have mo re dependents, thus exac erbating their negative economic situation. Caregivers leve l of education crosstabulated with the variable for crowded households produced a significant chi-square statistic ( (1) = 4.13, p < 0.05, twotailed). This result suggests that caregivers with secondary education or more were associated with households that were not crowded when compared with caregivers wit h primary education only. Therefore, in this research crowding was associated wit h lower SES and less education, but not with respiratory symptoms or history of childhood illness. Moreover, the same assumption for cr owding as a risk factor applied to exposure to smokers and indoor smoke from cooking in the household. When the variable for smoker in the household was compared with child frequently has cough or cold, child hospitalized fo r chest illness, child had bronchitis, and child had pneumonia, the results were not significant. This may partly be explained by the respondents assertions that their husbands, brothers, fathers, or boyfriends who smoked, usually smok ed elsewhere or outdoors. Finally, the chi-square test was performed with cook ing and the four illness variables described above. Cooking with wood as op posed to gas or coal, likewise did not make a difference in outcomes for ch ildren to suffer from frequent cough with colds, bronchitis, pneumonia, or history of hospitalization for a chest illness. However, there was a significant chisquare statistic between homes with nondirt floors and homes that cooked with wood only ( (1) = 15.18, p < 0.01, two-
230 tailed). This result suggests that homes with non-dirt floors were less likely to cook with wood exclusively than homes with dirt floors. Therefore, homes with dirt floors, in addition to having the problem of poor insulation from outdoor environmental pollutants such as volcanic as h, were more likely to add to this indoor air pollution with smoke from wood fires. Poor nutrition is another significant modi fiable ARI risk factor. In order to gauge the effect of nutrition on child heal th, nonparametric tests were performed on the macronutrient and micronutrient dat a to determine if there were any differences when stratified by age, gender and income. A Kruskal-Wallis chisquare test, a nonparametric equivalent of one-way (between subjects) ANOVA, was performed to compare the three age gr oups (Groups 2, 3 and 4) for which dietary data were collected [under 6 mont hs (Group 1, exclusively breastfed); 6 17 months (Group 2); 18 29 months (Group 3); and 30 months to under five years old (Group 4)] (Kinnear and Gray 2004). The Kruskal-Wallis chi-square test was not found to be significant for any of the macronutrients or micronutrients, except for the mineral sele nium and vitamin E, meaning that there was only a minor difference in mean diet ary nutrition intake between the three age groups. For the mineral selenium, t he chi-square statistic was significant, with a higher intake among Group 4 than Group 3, which in turn was higher than Group 2 ( (2) = 9.17, p < .05, two-tailed). In addition for vitamin E, the chisquare statistic was significant, with t he highest intake among Group 4, and the lowest intake among Group 3 ( (2) = 7.58, p < .05, two-tailed).
231 Table 18: Frequency of Nutrient Intake by Age Group Age GroupnMeanStd. Dev.Mean Rank kcal6 17 mo.141129.46268.0140.36 kcal18 29 mo.261046.51310.8234.54 kcal30 mo. 5 yrs.341130.66432.9138.59 carbs6 17 mo.14174.0449.0237.93 carbs18 29 mo.26167.1351.5036.69 carbs30 mo. 5 yrs.34176.1067.2637.94 protein6 17 mo.1440.839.7739.43 protein18 29 mo.2639.1916.4334.12 protein30 mo. 5 yrs.3441.8316.4439.29 fat6 17 mo.1432.7311.3343.07 fat18 29 mo.2626.6410.4833.69 fat30 mo. 5 yrs.3436.7335.3438.12 vitamin A6 17 mo.14488.86347.2228.71 vitamin A18 29 mo.26624.63528.5937.00 vitamin A30 mo. 5 yrs.34613.52358.6641.50 vitamin E6 17 mo.142.860.9334.50 vitamin E18 29 mo.262.581.0929.65 vitamin E30 mo. 5 yrs.344.022.9744.74 X = 7.58; P = .02 zinc6 17 mo.145.321.6138.14 zinc18 29 mo.265.222.3234.62 zinc30 mo. 5 yrs.345.461.8939.44 selenium6 17 mo.1431.2811.8525.00 selenium18 29 mo.2637.2413.0534.62 selenium30 mo. 5 yrs.3446.0816.3144.85 X = 9.17; P = .01 Kruskal-Wallis test on ranked data A Mann-Whitney U test, a nonparametric equivalent of the independent samples t test, was performed to compare the means of the nutrient data stratified on gender and income (Kinnear and Gray 2004). Although the mean ranks were higher for all macronutrients in boys than girls, the Mann-Whitney U test failed to show statistical significanc e. Likewise, the mean ranks were higher for all macronutrients in families with in comes above $150 per month except for carbohydrates, yet the Mann-Whitney U test failed to show significance. While the mean nutritional intakes demonstrated some trends of higher nutrition among
232 boys and families of higher household incomes, the results were not statistically significant. Table 19: Frequency of Macronutrient Intake by Gender and Salary nMean (g)Std. Dev.Mean Rank Boys: kcal371117.94366.2738.35 Girls: kcal371083.80365.1936.65 Boys: carbs37173.9058.0938.65 Girls: carbs37171.2259.2736.35 Boys: protein3742.7216.8339.89 Girls: protein3738.7113.4535.11 Boys: fat3735.4531.5540.65 Girls: fat3729.4117.1734.35 > Avg. Salary: kcal241115.80334.3239.63 < Avg. Salary: kcal501093.70379.9536.48 > Avg. Salary: carbs24168.7551.4036.58 < Avg. Salary: carbs50174.3961.7437.94 > Avg. Salary: protein2443.6417.2840.71 < Avg. Salary: protein5039.3114.1735.96 > Avg. Salary: fat2431.5913.9741.00 < Avg. Salary: fat5032.8329.4935.82 Compared with the nutrition goals for t he appropriate age ranges in the United States, the children in the study ranked above the recommended daily allowances (RDA) for macroand micr onutrients in the following age groups: carbohydrates, proteins, selenium, and zinc in all three groups; kilocalories in groups 2 and 3; fats in groups 2 and 4; vitamin A in groups 3 and 4. The following age groups performed below the RDA: group 4 for kilocalories; group 3 for fats; group 2 for vitamin A; and all th ree age groups for vitamin E. The major trend of the data indi cated vitamin E deficiency. However, dietary vitamin E deficiency occurs very rarely Deficiency of vitamin E only occurs in two classes of people: 1) low birth weight infants w hose low vitamin E levels have been linked
233 with their medical problem s; and 2) people who do not absorb fat normally such as sufferers of cystic fibrosis (Nationa l_Research_Council 1989: 99). Therefore, the apparent low intake of vitamin E compared to the RDA goal is likely insignificant in this study and can be expl ained by insufficient sample size or incomplete dietary records. Ther e were no reported cases of childhood blindness nor measles-complicated pneumonia, which would be a sign of possible vitamin A deficiency.17 Figures 19 through 21 illustrate where the different age groups rated for vitamin A, E, and zinc for the two-day combined recall compared to RDAs. Vitamin A Intake by Age Group (RE/day)0 100 200 300 400 500 600 700 6-17 months18-29 months> 30 months mean nutrition goal Figure 19: Vitamin A Intake by Age Group 17 There were 10 cases of pneumonia and 3 cases of meas les in the medical histories of the 91 children in the semi-structured questionnaires, but none of the children suffered from both diseases.
234 Vitamin E Intake by Age Group (mg/day)0 1 2 3 4 5 6 7 8 6-17 months18-29 months> 30 months mean nutrition goal Figure 20: Vitamin E Intake by Age Group Zinc Intake by Age Group (g/day)0 1 2 3 4 5 6 6-17 months18-29 months> 30 monhs mean nutrition goal Figure 21: Zinc Intake by Age Group The dietary data has limitations because the researcher was unable to determine the cooking method used in t he preparation of t he meals, so the dietary recall was not able to capture wh ich type of fats and oils were used in
235 cooking nor which condiments were used in seasoning of the f ood. Therefore, the dietary analysis should be considered as incomplete and suggestive only, and cannot be used to accurately correlate with frequency of self-reported childhood respiratory illness outcomes. Moreover, if nutritional deficiencies are indeed present in the childr en of Penipe, a thorough nutr itional study would have to be undertaken to measure food intake more precisely and include anthropometric data. T he exercise was instructive to assist the researcher to understand the basic diet of the childr en in the study and to determine the quantity and frequency of consumption of part icular foods such as meat, dairy products, and vegetables to estimate whether they were receiving enough calories, carbohydrates, fats, protein, vi tamins, and minerals in their daily diet. On the basis of the prelim inary dietary data collected, this study concludes that the childrens daily dietary requirements were being met. The analysis of the quantitat ive data results was useful to show statistical associations for the primary variables of interest. While the chi-square test of association has limitations, some infe rences can be made. In the following section on research hypothesis results, the researcher em ploys the binomial logistic regression procedure to better understand the relationship between SES and treatment choice.
236 PART II TESTING THE RESEAR CH HYPOTHESES: RESULTS The aim of the binomial logistic regr ession analysis was to determine the socioeconomic factors which were associ ated with seeking a doctor rather than employing household remedies for a seri es of symptom clusters. The three categories of health care-seeking behavio r were recoded into two categories: doctor and household remedies or nothi ng in order to create dichotomous variables for chi-square tests. Thes e variables were used as the dependent categorical variables, and the nine SES indicators were used as independent variables based on the results of the biva riate statistical tests. According to convention in binomial logistic regre ssion, the number of independent variables cannot exceed 10 percent of the sample size Using the treatment choice for the nine clusters of respiratory illness symptoms as the dependent variable, nine logistic regression procedures were carried out using the forward stepwise likelihood ratio method. After running t he logistic regression procedures, five SES independent variables were remo ved from the analysis during the procedure because of lack of ef fect on the outcome variable.18 It was important to leave all of these SES variables in t he initial model because it was not clear which variables would be associated with help-seeking behavior, although from the bivariate tests, salary had emer ged as a significant variable. This left four remaining independent variables in the models which included: salary above the mean, non-dirt floor, owning a television, and owning 18 The variables removed were appliance ownership, more than 3 rooms in the home, wall of blocks or bricks, land ownership, and owning more than 3 agricultural animals.
237 a bicycle. For the symptoms of interest (rapid breathing and indrawn chest), high SES on the independent variables of owning a bicycle and a television were associated with treating the child at home for rapid breathing and indrawn chest, respectively [odds ratio (OR) = .14, p < 05; OR = .34, p < .05] This finding was the opposite for the other sym ptom clusters, which foun d an association between high SES and treatment choice of a docto r, yet the odds ratios are not high enough to discount the other more signifi cant findings. The result may be explained by an overall lack of recognition for these se vere ARI symptoms, thus confounding the results (Mull, et al. 1994). For fever and nasal secretion health care-seeking, families with a non-dirt floor were two times as likely to seek medical care for their child (OR = 2.07, p < .01). For c ough, fever and nasal secretion health care-seeking, families wi th an above-average salary were five times as likely to seek a doctor for thei r childs symptoms when controlling for the other three independent variables (OR = 5. 97, p = .05); moreover families with non-dirt floors were 3.5 times more likely to seek a doctor than treat at home for children with these symptoms (OR = 3. 51, p = .05). Finally, for cough and phlegm health care-seeking, families with a non-dirt floor were six times more likely to seek medical care (OR = 5.91, p < .05), and families who owned a television were 3.6 times more likely to seek a doctor for these symptoms (OR = 3.62, p < .05). The resu lts of the binomia l logistic regression procedures demonstrated an association between owni ng particular material possessions and earning an above-average salary with a pa ttern of seeking help at a medical facility rather than treating at home to address childhood illness signs. In
238 Table 20: Logistic Regression fo r Health Care-seeking Behavior V ariableOdds Ratio95% CI s.e.P-value salary2.380.59-9.430.86500.70330.2187 non-dirt floor3.950.98-15.951.37240.71270.0541 television0.420.12-1.47-0.86570.63950.1758 bicycle0.380.11-1.29-0.97970.63340.1219 salary1.180.36-3.850.16330.60490.7482 non-dirt floor3.080.86-11.041.12450.65170.0844 television0.900.31-2.62-0.11000.54800.8409 bicycle0.920.28-3.06-0.08070.61170.8950 salary1.640.42-6.350.49510.69030.4732 non-dirt floor7.912.02-31.012.06830.69700.0030 television1.180.37-3.750.16900.58790.7738 bicycle0.230.07-0.80-1.45920.63020.0206 salary4.970.94-26.191.60280.84820.0588 non-dirt floor3.510.97-12.731.25510.65760.0563 television1.610.54-4.800.47830.55610.3897 bicycle0.530.14-1.96-0.63470.66810.3421 salary0.670.17-2.66-0.39730.70230.5715 non-dirt floor5.911.42-24.571.77640.72710.0146 television3.621.03-12.791.28730.64350.0454 bicycle0.120.03-0.45-2.09830.66340.0016 salary1.680.36-7.880.51950.78840.5100 non-dirt floor2.170.52-9.010.77410.72640.2866 television2.240.64-7.850.80460.64050.2090 bicycle0.340.09-1.31-1.07600.68490.1162 salary1.280.31-5.210.24640.71610.7308 non-dirt floor2.050.42-10.120.71790.81450.3781 television1.020.28-3.780.02050.66770.9756 bicycle0.140.04-0.51-1.94730.64910.0027 salary0.890.31-2.55-0.11270.53580.8333 non-dirt floor3.240.84-12.491.17480.68890.0881 television0.340.11-1.00-1.08660.55450.0501 bicycle0.540.19-1.60-0.60980.54960.2672 salary1.020.32-3.250.02160.59020.9709 non-dirt floor1.370.31-6.010.31490.75420.6763 television0.550.71-1.74-0.60760.59130.3042 bicycle0.600.19-1.92-0.50570.59110.3922 Rapid Breathing and Indrawn Chest Cough and Phlegm Cough, Phlegm, and Nasal Secretion Rapid Breathing Indrawn Chest Cough and Nasal Secretion Cough and Fever Fever and Nasal Secretion Cough, Fever and Nasal Secretion
239 particular, having a non-dirt floor impr oved predictive accuracy of the models, except in the case of the symptoms of rapid breathing and i ndrawn chest, which displayed inconsistent treatment choice s irrespective of SES because of overall lack of recognition of t hese biomedical symptoms. First Research Hypotheses 1) Female caregivers of higher so cioeconomic status with under-fives who display signs of severe respirator y distress are more likely to seek allopathic medical care. 2) Female caregivers of lower socioeconomic status with under-fives who display signs of severe respirator y distress are more likely to seek alternative medical care or home-based care. For the first hypothesis, from the semi -structured questionnaire results, no significant difference in health care-seeking behavior based on salary level or possessions for the two primary signs of ARI, rapid breathing and indrawn chest, was found, possibly explained by an overa ll lack of recognition of the biomedical symptoms of severe ARIs by the subject population. However, the results of the binomial logistic regression tied ownership of particula r material possessions and above-average salaries with seeking a doctors help for other respiratory symptoms. One gauge of a caregivers use of home remedies is recognition of the ethnomedical category of hot versus cold cough. A crosstabulation between television ownership and hot/cold c ough knowledge was performed, and the
240 resulting chi-square statisti c was highly significant ( (1) = 8.12, p < 0.01, twotailed). This result suggests that caregi vers without televisions, i.e. of low SES, were associated with possessing hot/col d cough knowledge, s uggesting the use of home herbal remedies, the preferred tr eatment for this illness. Therefore, there is some evidence that caregivers of low SES were likely to employ home or herbal remedies. The results of the binomial logistic regression confirmed the results of the preliminary chi-square tests, which found an association between certain SES variables and treatment choice. Mor eover, since families of low SES were associated with children who had a medical history of frequent coughs and colds, wheeze, and hospitalization for a chest illnes s, their strategy of attempting to treat childhood ailments at home first was a cost-saving mechanism to avert the potentially high costs of medical care, which could place a further strain on already limited household economic resources in the case of a hospitalization, for example. While not a definitive result, from the research results of the focus group sessions and the in-depth interviews with healthcare professionals, there is a higher tendency for people of low SES to tr eat their children at home initially rather than seek help at a healthcare fac ility, where they may incur expenses for medicines and travel, and lose ti me from agricultural work. Second Research Hypothesis Biomedical healthcare practitioners ma intain a perception that female caregivers delay seeking treatment for their under-fives who display
241 biomedically defined symptoms for ARIs because of an inadequate understanding of the bi omedical model, meaning lack of education. For the second hypothesis, there was support that doctors believe that their patients are unaware of the serious ness of ARIs due to lack of health education and therefore delay care-seeking. This conclusion is based on the qualitative data from the in-depth intervie ws with the healthcare workers and the focus groups with mother respondents. Of the seven doctors and nurses who responded to the question on maternal health care-seeking delay, five emphasized the need for more health educ ation on the import ance of seeking biomedical care for certain serious condi tions and following prescription regimens properly. The doctors from CEBYCAM had undertaken monthly health education classes with the groups of families that CEBYCAM was assisting with money for education costs courtesy of Italian donors for the children of thes e poor families. As described in the section under focu s groups, the doctors were educating these people on topics of first aid and fam ily planning, so there was the belief on the part of the doctors that they needed to receive this health information. The healthcare workers expressed the need fo r more health education on respiratory infections for their clients during the in terviews, yet this was not part of the curriculum of the health education lectur es by CEBYCAM. For example, one doctor observed that the mothers would use home remedies, such as herbal teas, for as long as four days before bringi ng a child to the doctor. However, the health professionals also listed other reas ons to explain the delay in health care-
242 seeking, e.g. tending animals and treating at home or with a curandero Another doctor mentioned that mothers know about ARIs, but do not follow their prescriptions properly. The healthcare professionals recommendation was to wait no longer than two days to seek tr eatment if the child s condition did not improve. The nurses and doctors biomedical m odel followed AIEPI guidelines for the treatment and symptom recognition of ARIs.19 In the pamphlet produced by the Ministry of Public Health to warn of ARIs, several risk factors are mentioned, including poor nutrition, poor vaccination coverage, lack of breastfeeding, as well as smoke and lung irritants, quick changes in outside temper ature, and exposure to dirt or dust. The primary symptom s mentioned in the pamphlet are cough, wheeze, fever, pain in the throat and ear, and nasal obstruction. The pamphlet offers some simple home remedies that can be applied to alleviate some symptoms, such as tapping the back to relieve phlegm, placing cold compresses on the forehead and stomach to lower the ch ilds temperature, and offering more liquids. The pamphlet warns of the signs of pneumonia such as difficulty breathing, indrawn chest, a pallid hue to the skin, inability to eat or drink, and lack of response to stimulation, and suggests t hat if these signs ar e present to bring the child to the doctor immediately. When mothers were asked to describe what they believed to be ARI symptoms, they mentioned c hest whistling, tightening of the chest, green nasal discharge, stuffy nose, cough, colds, and wheezing, all of which could be 19 AIEPI stands for Atencion, Integral, Enfermedades Prevalentes en la Infancia [Attention, Integral, Sicknesses, Prevalent in Infancy] and was developed in conjunction with WHO.
243 possible symptoms for ARIs as defined in the biomedical model. Some of the mothers other explanations for the caus e of ARIs, such as change in climate, lack of vitamins, and ash rain from the volcano, were all also plausible explanations from a biomedical standpoin t for the onset of ARIs. The main discrepancy between the biomedical and et hnomedical models was the mothers use of various home remedies for an ext ended time, over the two-day limit as recommended by healthcare professionals From the information collected regarding the ethnomedical ca tegory of hot/cold cough s, the use of cold compresses was only employed for hot coughs, which were considered less severe than cold coughs. Therefore, there was c onsiderable overlap between the biomedical and ethnomedical models for the i dentification of ARIs, but some inconsistencies between the type and length of applicati on of home remedies employed. Moreover, the results of the chi-square te sts confirm the doctors assertions of lack of recognition and treatment delay, especially for indrawn chest and the aggregate of symptoms of indrawn chest and rapid breathing, which indicated a lack of congruence between perception and treatment, meani ng that while caregivers perceive the symptoms as serious they choose to treat the illness at home initially.
244 PART III DISCUSSION The initial research began with the focus group discussions and participant observation in the health posts in Penipe. The focus group discussions were useful in generating new information, increasing the list of terms used to describe illnesses and tr eatments, and establishing rapport with the research subjects. The discove ry for example of the ethnomedical classification system of hot coughs versus cold coughs, generated new areas of inquiry in the semi-structured questi onnaire, and additionally, led to fruitful discussions concerning herbal remedies The information generated by the focus group discussions was an excellent st arting point to narrow the field of inquiry for the development of the semi-structured ques tionnaire. Without the experience of talking to the mother respondents in a semi-informal manner, the semi-structured questionnaire exercise would not have been very productive, nor accurate. Such an experience lends cr edence to the earlier assertion that epidemiological or medical studies in formed by an ethnographic component lead to more meaningful re search results (Dunn and Janes 1986; Weiss 1988). By spending time in the wait ing rooms of both CEBYCAM and SCS Penipe, the researcher was able to take notes on the surroundings, such as all the posters and paintings ador ning the walls, to gauge t he relative experience of waiting. At SCS Penipe, the expe rience of waiting was improved through painting the interior with vibrant co lors, adding plants, making various health information brochures and pamphlets ava ilable, offering water and tea, and
245 having a television with educational videos to entertain the waiting public. All of these improvements were made during the c ourse of the research. In contrast, at CEBYCAM, the same row of chairs and lack of stimulation for the waiting public continued throughout the research period. While there were some educational posters and pictures, the client s did not appear to take an interest. Both at SCS Matus and PS Palictahua, the waiting hall was very barren and uninviting. These venues are prime lo cations for displaying health promotion messages, but they remain underutilized. In the in-depth interviews with t he healthcare professionals, doctors, nurses and other healthcare workers m entioned diarrhea, stomach parasites, and respiratory illnesses as the primar y health problems in the region with poverty as the underlying cause. When compared to the focus group discussions with mother respondents, t here was agreement on this point; however, the mothers had suggestions for ho w to improve their situation through reinstituting government food programs. As for barriers to seeking timely treatment, some nurses comment ed that the hours of the clinics were restrictive, and the people would seek care more readily if the hours were extended. This sentiment was echoed in the focus gr oup discussions with mother respondents; however, the mothers primar y identified barrier for seeking medical care was the cost of medications. While the healthca re workers perceived the primary barriers to care as economic, in the sense that mo thers did not want to leave their fields and treated most respiratory a ilments in their children at home, in the mothers opinion, the expenses of visiting the health posts were prohibitive because of the
246 costs of bus fares and medicines. T hese identified barriers are echoed in ethnographic studies of ARIs in other devel oping countries in Latin America, the Middle East, and Asia (Gove and Pelto 1994; Hudelson 1994: Mull and Mull 1994; Nichter 1994; Stewart, et al. 1994; H udelson, et al. 1995; McNee, et al. 1995). There was a general consensus among the healthcare workers that the government clinics needed to expand their staff, add infrastructure, expand their hours of operation, and better reach out to the community to in form them of the medical services offered. For example, throughout the whole county, there were inadequate facilities for women to bear child ren. In the same vein, there was a consensus among mother informants from the focus groups that the government clinics needed to extend thei r hours, reach out to the public to educate them on vital health issues, and expand the services that they offered. Nevertheless, the mothers maintained a very positive view towards the health clinics, especially CEBYCAM, for the monetary support t hey offered to poor families. The epidemiological data collected from the Chimborazo Ministry of Health suggested that after 2001, there cont inued to be a variation in outpatient consultation rates for ARIs in under-fives in Penipe County. This finding suggests a rise in risk factors for the at-risk population, causing a doubling in rates from 2002 to 2003; moreover, the re sult contradicted nurses assumptions that the people of Penipe had become accu stomed to inhaling volcanic ash, and outpatient rates for respiratory infections had stabilized. While the rates did fall again to average levels in 2004, in 2003 rates were comparable to the year
247 following the first volcanic eruption in late 1999 (Whiteford and Tobin 2002). The collection of epidemiologica l data was useful to compare the number of cases collected between the various health centers in Penipe County. According to the statistician at the Ministry of Heal th, CEBYCAM would often underreport its outpatient consultations, or miss repor ting deadlines; however, the researcher was unable to corroborate this information. Nevertheless, the possibility remains that the number of cases reported for Penipe is an underestimation of the true number of ARI outpatient consultations. The combination of qualitative a nd quantitative data collection and analysis techniques enabled the researc her to gain a greater understanding of childhood ARIs in Penipe County than would the utilization of one method alone. Because of the inconclusive results of some of the quantit ative statistical procedures, results from t he focus group discussions and in-depth interviews aided in the corroboration of research hypotheses and assumptions. For example, while one cannot conclusively asse rt from the statisti cal tests that low SES is the key variable determining the choice of home-based rather than medical care for severe ARIs, because of an overall lack of knowledge irrespective of SES, there is evidence from the qualitative findings in both the semi-structured questionnaires and focu s group discussions that lack of economic capital was the pr imary reason for delaying appropriate health careseeking in the case of childhood ARIs. Mo reover, from the qualit ative findings, it was apparent that caregivers who had ex perience with children with pneumonia were more likely to identify the signs of rapid breathing and indrawn chest as
248 serious and deserving of a doctors vi sit on the semi-structured questionnaire compared to those who did not. This resu lt parallels findings of a case-control study on ARI symptom recognition in Paki stan (Mull, et al. 1994). Moreover, consistent with the findings of Mull, et al. (1994), this research found that chest indrawing was not as well recognized as rapid breathing, perhaps since chest indrawing is a late stage of pneumonia and can be a relatively uncommon clinical symptom. Since this is an ethnographic study, qualitative rather than quantitative methods are given more emphasis. The qualitative findings suggest that other factors besides simple symptom rec ognition are important determinants for seeking medical care. For example, time and money are two factors which could determine whether a caregiver is able to br ing a child to the doctor. One of the criticisms of the focused ethnographic st udy (FES) technique for studying ARIs was that it placed too much emphasis on the cognitive dimension of illness recognition without attending to the soci al context (Nichter and Nichter 1994). Moreover, since FES is a rapid study tec hnique, it is not possible to conduct a more in-depth study which can identify t he structural factors which impede timely health care-seeking behavior. Gove and Pelto (1994) have defended FES that its purpose is not to investigate these la rger structural constraints such as poverty but to describe and discover the ethnomedical illness categories that most closely correlate with biomedical categories in order to assist ARI programs. This research concurs with Gove and Pelto (1994) that the major purpose
249 of the ethnographic project is to foster collaboration between anthropologists and healthcare practitioners in order to formulate more effective health communication, which is possible during cl inician visits. Nurses and doctors need to ensure that the correct regimen of prescribed an tibiotics is employed to treat ARIs in order to limit inappropriate use of these medications which may lead to antibiotic resistance. Specific recommendations for health promotion strategies are discussed later in the concluding chapter. Where this research has gone beyond a FES approach is to identify the modifiable risk factors t hat could contribute to higher respiratory illness frequencies and delay in timely biomedical tr eatment seeking. For example, this research found that poor housing was an important factor for both increased respiratory illness problems in children and lack of timely health care-seeking. Because homes with dirt floors were associated with using wood fuel for cooking, the result was greater indoor air pollu tion, and consequently, children would be more likely to suffer from respiratory in fections. Therefore, a public health intervention would be to not only ensure that houses have well insulated roofs, but also cement floors rather than dirt floors, which ar e easier to keep clean from dirt and ash. Another modifiable risk fa ctor includes improving nutrition, by incorporating more fruits and vegetables into the diet. Finally, the ethnographic fi ndings of this research suggest that the research subjects are aware of the larger structural factors that limit their choices related to health care. In the focus group discussions, mothers pointed to the stress of poverty and limited food supply as negatively affecting their familys
250 health. In the mothers view, it was the responsibility of the government to provide them with food and economic a ssistance because their communities were the victims of an ongoi ng natural hazard. They continued to suffer the health consequences of ash fall, and it was their view that the government was responsible for improving their situation.
251 CHAPTER EIGHT: CONCLUSION, RECOMMENDATIONS, AND CONTRIBUTIONS OF RESEARCH PART I CONCLUSION In the final analysis, the people of Peni pe are marginalized by the larger society of Ecuador. Due to the eruption of Mount Tungurahua, 12 bridges were destroyed, connecting Bilbao Parish wit h the rest of Penipe County, and connecting Riobamba with the popular tourist to wn of Baos. The closing of this road has had negative economic consequences for Penipe County. Only this year has survey work finally begun, five years later, to reconstruct these bridges and reestablish the road infrastructure. Many county residents believe that since Penipe has a small population of merely 6,485, making up a mere 1.6 percent of the population of Chimborazo Province ( 403,362), their voices are not heard by national government authorities whose job it is to repair the infrastructure damaged by natural disasters, since the votes of the people of Penipe could be considered inconsequential. The re searcher witnessed one event where the former President Lucio Gutierrez was scheduled to make a stop in his helicopter in the county during the congressional el ections, but cancelled his appearance and sent a local party candidate in his place. One of the major issues in these
252 elections was the reconstruction of the r oads and bridges in the county. A bridge is an important symbol for connecting people and places. When bridges are precarious, there is a feel ing of insecurity and dread wh en crossing, a feeling the researcher experienced on one trip from Bilbao Parish. Figure 22: Bridge in Bilbao Parish From the critical biocultural pers pective, overall negative health outcomes for people from Penipe can be linked to stru ctural factors relating to Penipes peripheral place in the regional and nat ional economy and to its slow recovery from the 1999 disaster caused by Mount Tungurahua. In the course of fieldwork, the researcher continued to hear comp laints from residents who still had not had their roofs replaced. Another common complaint was the lack of employment and business opportunities, which relates not only to local but to national
253 economic crises. Those with the necessary means to leave Penipe, either in the form of economic resources or social support, usually left town to find opportunities elsewhere, while those unable to leave remained. Abandoned houses in Penipe remained unoccupied five years later, only to be revisited during the annual festivities of the Patron Saint Francis Nevertheless, with the assistance of Italian funding through CE BYCAM, new enterprises were showing signs of life, such as a milk-processing plant and a canned meat factory. Despite the national governments inadequate response to Penipes needs, international donors, foreign governments, and NGOs t ook the initiative and collaborated with local institutions in order to further Penipes development goals, which had been sidetracked following the 1999 er uption of Mount Tungurahua. This research has contributed to the knowledge of the ethnomedicine of ARIs and health care-seeking behaviors in a marginalized, mestizo population in Andean Ecuador. In particular, th e ethnomedical research added to the knowledge of beliefs concerning hot and cold coughs, an ethnomedical classification system which crosses ethnic boundaries, between mestizo and indigenous people (McKee 1987). Acco rding to McKee (1987:1147), a Hispanically-imported belie f in a humoral theory of disease has become intertwined in the Ecuadorian Andes wit h indigenous conceptions of hotness and coldness and has developed a folk taxonomy for common childhood illnesses like diarrhea. Of the respondents surveyed in the questionnaire, 64 percent had knowledge of the ethnomedical classification system of hot/cold coughs and were familiar with the accepted remedies to employ for each. Female caregivers
254 attended to both hot and cold coughs with home-based remedies, even though they considered cold coughs to be potentia lly serious illnesses. The caregiver respondents reported 69 different types of herbal remedies to treat common illnesses at home, and if t hey did not have these herbs growing in their own gardens, they usually had a neighbor who di d. The caregivers reported that many of the herbs they used were effect ive in treating various types of coughs and colds and saved them the time and expense of seeking a doctors help based on their responses to the questionnaire s. Moreover, since the eruption of Mount Tungurahua, the general populace be came more aware of particular pharmaceutical drugs according to area ph ysicians, and this was reflected in the questionnaires which found 60 different types of over-the-counter and prescription medicines in peoples homes. One nurse in Matus noted that after the eruption of Mount T ungurahua, there were numerous talks given to community members by the Ministry of Public Health, the Civil Defense force, and the Red Cross on the dangers of volcanic ash to health. Some members of the community understood the importance of using antibiotics for some types of infections and had also bought into the conc ept of vitamins to address potential nutritional deficiencies in t heir children. This greater awareness of the uses of particular medications was largely due to the collaborative efforts of the public and private health centers in Penipe Count y, as well as other governmental and non-governmental entities, in response to the hazard of volcanic ash from Mount Tungurahua. Like other Andean ethnom edical studies by Bastien (1987), Finerman
255 (1987), Crandon-Malmud (1983, 1989), and Mc Kee (1987, 2003), this research found a combination of strat egies in the treatment of childhood illnesses. Given particular hypothetical groups of illness symptoms, mothers we re reluctant to choose between treating a child at home and seeking a medical doctor because health care for each illness symptom was c ontingent on a multitude of factors. Some of these factors included whether t he child had a low or high fever (not specified in the questionnaire), which hom e treatments were available, and the day and time of the week when the illness occurred. There were constraining factors placed upon the caregivers when deciding types of treatment which included costs of treatment s or medicines, limited hour s of the health centers, and work demands. Asking close-ended que stions on treatment choice was challenging; however, the re sults of this line of ques tioning combined with the qualitative data produced ri ch ethnographic data. This research modeled the se mi-structured questionnaire on an epidemiologically-based medical hist ory survey with an added section to generate open-ended responses in order to g enerate cultural categories. The traditional modifiable risk fact ors found in epidem iological studies of ARIs were not found to be significant for self-report ed negative respiratory health outcomes, namely undernutrition, crowding, exposu re to air pollutants, and maternal education (Lanata 2004). Mor eover, non-modifiable risk factors such as child gender and age were not associated wit h self-reported negat ive respiratory health outcomes. However, other risk fa ctors were found to be significant for negative child respiratory health such as poor housing conditions and low
256 socioeconomic status. This finding para llels other ARI studies in Latin America and Asia (Biswas, et al. 1999; Frost, et al 2005). Because of study limitations, other important risk factors for ARIs could not be determined, such as breastfeeding rates, low birth weigh t, and exposure to volcanic ash. In the final analysis, the qualitat ive sections of the questionnaires generated more useful explanations fo r mothers health care-seeking behavior and identification of ethnomedical signs than the close-ended sections of the questionnaire aimed at elicit ing biomedical risk factors for ARIs. The standard epidemiological risk factors for ARIs in the literature whic h include crowding, number of children, gender, education, nutrition, and ex posure to environmental pollutants such as smoke were not f ound to be significant risk factors for respiratory infections in this study (Graham 2001). Of course, these health outcomes were self-reported, rather than clinically observed symptoms; nevertheless, the data did not support this finding. For the past 25 years, there has been a strong tradition in social epidemi ology of using self-reported health as valid indicators (Berkman and Breslow 1983; Berkman and Syme 1979). From this research, the two identified risk fact ors for both respiratory infections and delayed health care-seeking were low SE S and lack of health education. Health education is necessary to teach mothers how to identify the signs of serious respiratory infections and to properly use antibiotics. The study concurs with the conclusion of Douglas (1990), who found t hat increased access to medical care combined with maternal education to impr ove health literacy should result in improved child health outcomes.
257 While low SES and lack of health education contribute to the likelihood of childhood respiratory infections, the illness should be viewed in the larger context of political-economic factors, which im pinge upon the health of the population. Because the experience of health or illness depends on peoples relative position within the social fabric, a mothers decisi on to seek medical care for her child is not a solitary one, but one that affects the whole family and is contingent upon a number of interlocking factors, such as the childs gender, time during the harvest, and access to money. While the final decision to seek medical care resides with the mother in this case, she is not alone in the process that leads to such decision-making. Treatment choice is a reflection of not only the mothers beliefs and predilections, but of family and friends also (Rubel and Hass 1990). She has to take into account the time she might lose from agricultural work, and how much money she will have to spend on transportation and medications in order to treat her child. While some doctors and nurses perceived that mothers delayed treatment for their children, the mothers themse lves asserted that they would seek medical attention promptly and regularly as a preventive measure. This response bias was difficult to measure; however, the discrepancy between reported and actual behavior was corroborat ed by repeated health professionals testimonials of delayed health care-seeking for ARIs as well as health statistics for the region, which indicate a high in cidence of ARIs, indicating that the healthcare workers had many cases of childhood ARIs on which to base their negative opinions. This descriptive case study was modes t in its goals of identifying possible
258 risk factors for delayed health care-seek ing and negative child respiratory health outcomes, and studying ethnomedical systems of respirat ory infections in rural Ecuador. Like any other part in the world, coughs and colds are frequent maladies in children, and people have adapted to treat such common ailments at home because it is both more cost-effective and less time-consuming than consulting a doctor or nurse. The pr oblem arises when mothers mistakenly identify serious lower respiratory infecti ons as regular coughs and colds that can be treated at home. The pr imary finding of this stud y, which is widespread across the literature on ARIs, is that mo thers lack the knowle dge to identify the two primary signs of ARIs, rapid breathing and indrawn ches t (Mull, et al., 1994). This result parallels the findings of Nic hter (1994) for India and the Philippines. Since antibiotics to treat lower respirator y infections cost between two to three dollars, serious complications from ARIs are preventable and unnecessary. However, mothers who had had their ch ildren become sick with pneumonia were familiar with the signs. The question then becomes, why did these mothers not share this knowledge with others about th is very serious disease? Because childhood deaths from the disease fortunately were rare, perhaps, the seriousness of the disease had not been conveyed to others. The answer relates to peoples social networks and lines of communication. One way to improve peoples access to health info rmation had been initiated by CEBYCAM with health education classes in the comm unity given by the doctors, but there was no curriculum for respiratory infections The subject matter focused on first aid and family planning and was limited to families receiving assistance from
259 CEBYCAM. While the research identified poor recognition of severe ARI signs as a health risk for young children, when asked, some caregivers identified structural inequalities and ecological factors as the primary causes for poor health outcomes. These factors included lack of employment and educational opportunities, rising consumer prices, weak in frastructure of irrigation and roads, corrupt politicians, unsafe drinking wa ter, and frequent ash fall which damaged their homes, crops, and livestock. This dissertation began with Ecuadori an history, explored the culture of ethnomedicine, and ends with po litics. As the dist inguished anthropologist William Roseberry wrote, even at thei r most esoteric, anthropological ideas about culture involve a series of (often uns tated) ideas about history, capitalism, the state, political action (Roseberry 1989:231). While the residents of Penipe are not involved in organized protests against injustice, their political action takes another form. By aligning themselv es with CEBYCAM and forming producer cooperatives in order to fu rther their development goals, they have kept one foot forward, in spite of the uneven hand that they have been dealt by political entities and geological forces.
260 PART II RECOMMENDATIONS Both the quantitative and qualitative resu lts demonstrated that the people of Penipe County need more health informa tion and education in order to make timely health care decisions when their children are sufferin g from respiratory infections. The Ministry of Public Health has prepared a colorful brochure illustrating ARI symptoms and appropriate prevention measures. However, the introduction of this pamphlet in t he government health centers has not been accompanied by any organized health ca mpaign or public health education strategy. General recommendations for preventing ARIs involve public health education efforts which focus on bette r nutrition, hygiene, immunization coverage, and proper use of medications, es pecially antibiotics, to control this deadly disease. The specific recommendations of this study indicate that public health messages regarding ARIs be info rmed by ethnomedical knowledge of hot and cold coughs as well as appropriate hom e remedies for treating mild upper respiratory infections. In addition, such education efforts need to include targeted messages to help caregivers recognize the severe signs of lower respiratory infections a nd use appropriate home remedies such as cold compresses and increasing the ingestion of liquids. Further, borrowing from social marketing principles, messages should be targeted to the audience of mothers of low SES, identified in the study as the population at risk. Moreover, health centers need to adhere more closely to regular hours of operation, extend evening hours, and increase staff capacit y. Such extended hours would also
261 provide more opportunity for health education in the health centers. The people of Penipe have become highly resilient to the negative effects of ash from the active volcano, Mount Tungurahua, yet despite this adaptation, the vulnerable members of the population are still ex periencing negative health outcomes due to undernutrition, poverty and exposure to environmental contaminants such as dust and smoke. Through a focused, concerted health campaign, great strides against the negative effects of ARIs can be achieved. A health campaign should foster community pa rticipation. There is already a group of 10 to 15 women who meet monthly at SCS Penipe to discuss womens health issues and empowerment. If the wo mens group were to address the ARI problem in partnership wit h local doctors and nurses, and create buzz about the topic, this could be a first step in planni ng a health promotion strategy that would involve peer education, educational sessions led by doctors, nurses, and community leaders, radio advertisements (there is a local radio station in Penipe), and distribution of pamplets and flyers. In conjunction with a health prom otion campaign, the ethnographic literature has shown that tr aining peer educators or CHWs to teach caregivers to recognize the signs of ARI and teac h proper case management has met with success, strengthening the case for addr essing the problem of childhood ARIs with low-cost educational strategies (P andey, et al. 1991; Kelly, et al. 2001). Presently, there is only one community heal th worker in Penipe, the part-time nurse from the Department of Children and Families, an d she works primarily in the urban area of the town of Penipe only. Training a group of CHWs in the
262 community to disseminate ARI educationa l information would help to decrease cases of untreated severe pneumonia. Proper treatment and care of childhood ARIs requires the following: (1) early detection of rapid or labored breat hing and indrawn chest at the household level; (2) timely maternal health care -seeking behaviors for children displaying these signs; (3) access to health care facilities with proper case management practices in place; (4) appropriate use of ant ibiotics; and (5) cultural sensitivity to ethnomedical illness categorie s and culturally appropriate health communication (Herman, et al. 1994). In this re search, the main problem was not underutilization of biomedical health se rvices because of a preference for traditional health practitioners; rather, the primary barri er was recognition of the seriousness of lower respiratory infecti ons. Consequently, a health promotion program employing a social marketi ng approach should target mothers, especially younger mothers, to identify and appreciate the identifiable symptoms of rapid or labored breathi ng and indrawn chest as signs of lower respiratory infections that require consultation with a healthcare professional (Kotler, et al. 2002). Targeted efforts to the harder to reach rural population in the hills, beyond the main towns and hamlets of Penipe, should be a priority to improve the overall public health of the inhabitants. The targeting of high-risk segments using a social marketing strategy exam ines perceived benefits, barriers, and costs in order to design a culturally appropriate health prom otion program in order to change behavior, known as the product, in social marketing terms. The methodology of this study was designed using the concept of
263 triangulation, which comb ined quantitative and qualitat ive data gathered through a combination of research methods. Given budget constraints, this was determined to be a valid method of determining the research outcome. The research concludes that caregivers s hould be educated on the recognition of the signs of lower respiratory infections in young children. Based on the study results, the following summary recommendati ons are offered: (1) initiate a public health campaign sensitive to ethnomedica l conceptions of health and illness to warn of the signs of serious lower resp iratory infections and to seek a doctor immediately as well as to recognize le ss serious upper respiratory infections which can be treated at home; (2) lobby provincial health authorities to adhere to regular hours of operation, expand clinic hours and increase staff; (3) establish policies for disbursement of antibiotics and ensure that patient s follow prescribed regimens; and (4) negotiate with bus com panies to offer reduced fares when women are bringing their sick children to healthcare facilities.
264 PART III CONTRIBUTIONS TO APPLIE D ANTHROPOLOGY AND PUBLIC HEALTH The combination of applied anthropology and public health in this research produced findings with relevance to both disciplines. For applied anthropology, the combination of ethnographic research and quantitative survey techniques produced data that could be used for the purpose of a health development project, using culturally-sensitive educat ion materials by peer health educators to improve maternal recognition of severe respiratory signs. In contrast with a public health approach, which focuses on disease control and prevention, an applied anthropology participat ory research project em phasizes the importance of culture in applying research findings to address community health issues. By examining the problem of ARIs fr om the multiple perspectives of the caregiver, epidemiologist, doctor, nur se, social worker, and curandero, the researcher was able to develop a more in-depth understanding of the factors involved in maternal health care treatm ent decision-making. This research suggests that low socioeconomic stat us has negative respiratory health outcomes for children, as well as reduces the likelihood of timely health careseeking behavior. This finding resonates with research in social epidemiology, which examines the effect of social and economic factors on illness outcomes. Moreover, the results of this study add further evidence to support medical anthropologys contribution to understanding the variables influencing health care-seeking behavior in a specific cultural context.
265 The religious and civic leaders of P enipe are active collaborators with government authorities and in ternational development organizations to improve capacity in agricultural enterprises and heal th infrastructure. Research studies such as this one may assist local NGOs in Penipe to seek funding for better healthcare facilities. Recently, with Ita lian funding, CEBYCAM has constructed a respiratory health therapy center which will be headed by a regional pulmonologist and will attract patients from all over the province of Chimborazo. Given the poor respiratory health of area residents, primarily among children and the elderly, there will be a plac e where professional medi cal staff can treat their illnesses. Future studies of the respir atory health of children should be pursued in the clinical setting to measure sympto m recognition by caregivers, as well as frequency and severity of respiratory illne ss episodes as recorded by physicians and nurses. The contribution of applied anthropology to this equation is to educate healthcare professionals to recognize ethnomedical signs and symptoms and use the language and idioms appropriate to the people they serve. This research study is a combi nation of qualitativ e and quantitative research with the aim of achieving result s of both high validity and reliability. The two parts of the research, the qualitat ive, pursued through in-depth interviews and focus groups, and the quantitative, gained through semi-structured questionnaires and epidemiological record s, informed and complemented each other at different stages of the research endeavor. T he researchers combined
266 training in anthropology and epidemiolo gy permitted this strategy of interdisciplinary research with positive resu lts. In this research, the qualitative methods provided the cultural dimensi on that contributed to the quantitative research. Likewise, the quantitative meas ures indicated patterns of association that informed the tentative qualtitative research findings. The goal of this research has been to contribute to the anthropology of childhood respiratory infections . While not established as a subfield, the work by Hudelson (1994), Nichter (1994), Simon, et al. (1997), Stewar t, et al. (1994), and Tobin and Whiteford (2002) has brought the social and cultural factors of this illness to the attention of fellow academ ics and policy makers. Through further interdisciplinary collaboration, general health education intervention efforts may be undertaken with the support of anthropologica l research to ensure culturally relevant health promotion messages. Finally, anthropological investigations which incorporate epidemiological findings have the potential to contribute to anthropological theory. This research has drawn from the work of other anthropologists to illustrate the asso ciation between childhood respiratory illnesses and broader social and cultural fo rces resulting from income inequality, political marginality, and natural disasters.
267 LIST OF REFERENCES Agar, M. 1996 Recasting the "Ethno" in "E pidemiology". Medical Anthropology 16:391-403. Amofah, G.K., I.T. Essegebey, S. A. Opoku, and J. Oduru 1995 Intended Response of Caretakers of Pre-school Children to Acute Respiratory Infection in a Rural Dist rict in Ghana. East African Medical Journal 72(10):619-622. 1998 Perceptions of Caretakers of Pre-school Children on Acute Respiratory Infection (ARI) in a Ru ral District in Ghana. West African Journal of Medicine 17(2):64-69. Anderson, R. 1992 The Efficacy of Ethnomedicine: Research Methods in Trouble. In Anthropological Approaches to the Study of Ethnomedici ne. M. Nichter, ed. Paris: Gordon and Breach. Armelagos, G.L., T. Leatherman, M. Ryan, and L. Sibley 1992 Biocultural Synthesis in Medical Anthropology. Medical Anthropology 14:35-52. Awedoba, A.K. 1996 Some Ghanaian Mothers Re call of ARI Episode. African Anthropology 3(1):65-90. Azizi, B.H., H.I. Zulkifli, and M. Kasim 1995 Protective and Risk Factors for Acute Respiratory Infections in Hospitalized Urban Malaysian Childr en: A Case Control Study. The Southeast Asian Journal of Tropical Medicine and Public Health 26(2):280-285. Baer, H.A. 1996 Bringing Political Ecology into Critical Medical Anthropology: A Challenge to Biocultural Approaches Medical Anthropology 17:129-141. Bang, A.T., R.A. Bang, O. Tale, P. Sont akke, J. Solanki, R. Wargantiwar, and P. Kelsarkar 1990 Reduction in Pneumonia Morta lity and Total Childhood Mortality by Means of Community-based Intervention Trial in Gadchiroli, India. The Lancet 336:201-206. Bastien, J. 1978 Mountain of the Condor: Met aphor and Ritual in the Andean Ayllu: West.
268 1981 Metaphorical relations between sickness, society, and land in a Qollahuaya ritual. In Health in the Andes. J. Donahue, ed. Pp. 19-37. Washington D.C.: American An thropological Association. 1987 Cross-cultural Communicati on between Doctors and Peasants in Bolivia. Social Science and Medicine 24(12):1109-1118. Bastien, J., and J. Donahue 1981 Health in the Andes: American Anthropological Association Special Publication No. 12. Baxter, P.J., C. Bonnadonna, R. Dupree, V.L. Hards, S. C. Kohn, M.D. Murphy, A Nichols, R.A. Nicholson, G. Norton, A. S earl, R.S.J. Sparks, and B.P. Vickers 1999 Cristobalite in Volcanic As h of the Soufriere Hills Volcano, Montserrat, British West Indies. Science 283:1142-1145. Baxter, P.J. 1997 Volcanoes. In The Public Health Consequences of Disaster. E.K. Noji, ed. Pp. 179-204. New York: Oxford University Press. Bebbington, A. 1996 Organizations and Intensificati ons: Campesino Federations, Rural Livelihoods and Agricultural Technol ogy in the Andes and Amazonia. World Development 24(7):1161-1177. Benguigui, Y., F. Lopez, G. Schumunis, and J. Yunea 1999 Respiratory Infection in Children. Washington, DC: PAHO. Berkman, L.F., and L. Breslow 1983 Health and Ways of Living: The Alameda County Study. New York: Oxford University Press. Berkman, L.F., and S.L. Syme 1979 Social Networks, Host Resi stance and Mortality: A Nine-year Follow-up of Alameda County Residents. American Journal of Epidemiology 109:186-204. Berman, S. 1991 Acute Respiratory Infections. Infectious Disease Clinics of North America 5(2):319-336. Bernard, H.R. 1994 Research Methods in Anthropology. London: Sage. Biswas, A., R. Biswas, B. Manna, and K. Dutta 1999 Risk Factors of Acut e Respiratory Infections in Underfives of Urban Slum Community. Indian Journal of Public Health 43(2):73-75. Biswas, R., A.B. Biswas, B. Manna, S.K. Bhattacharya, R. Dey and S. Sarkar 1994 Effect of Vitamin A Suppl ementation on Diarrhoea and Acute Respiratory Tract Infection in Childr en. A Double Blind Placebo Controlled Trial in a Calcutta Slum Community European Journal of Epidemiology 10(1):57-61. Blaikie, P., and H. Brookfield 1987 Land Degradation and Society. London: Methuen.
269 Brosius, P. 1999 Comments on After Nature by Arturo Escobar. Current Anthropology 40(1):16-17. Brown, M.F. 1993 Facing the State, Facing the World: Amazonia's Native Leaders and the New Politics of Ident ity. L'Homme 92(4):307-326. Browner, C. H., B.R. Ortiz de Montellano, and A.J. Rubel 1988 A Methodology for Cross-cultural Ethnomedical Research. Current Anthropology 29(5):681-702. CDC 2002 Pneumonia among Childr en in Developing Countries. www.cdc.gov/ncidod/dbmd/diseaseinfo/ Centers for Disease Control and Prevention (accessed April 13, 2003). CEPAR 2005 Endemain 2004 Ecuador: Enc uesta Demogrfica y de Salud Materna e Infantil, Informe Preliminar. childrensvaccine.org 2004 Pneumococcus Quick Facts. www.childrensvaccine.org (accessed January 30, 2005). Chiriboga, M. 1988 La Reforma Agraria Ecuatoriana y los Cambios en la Distribucin de la Propiedad Rural Agricola, 1974-1985. In Transformaciones Agrarias en el Ecuador. P. Gonard, ed. Quito: CEDIG. Chuc, N. T., M. Larsson, T. Falkenberg, N.T. Do, N.T. Binh, and G. Tomson 2001 Management of Childhood Acute Re spiratory Infections at Private Pharmacies in Vietnam. The Annal s of Pharmacotherapy 35(10):12831288. CIA 2005 CIA The World Factbook. Central Intelligence Agency, www.cia.gov (accessed September 30, 2005). Collins, J.N. 2000 Unity, Dignity, Autonomy: A Pr ofile of CONAIE President Antonio Vargas. NACLA Report on t he Americas 33(5):42-3. COSUDE 2003 Penipe: Plan de Desarrollo 2003-2013. Riobamba: Municipio del Cantn Penipe, Agencia Suiza para el Desarrollo y la Cooperacin (COSUDE), Centro de Desarrollo Humano en Cultura y Economa Solidadria (CEBYCAM-CES). Crain, M. 1990 The Social Construction of Na tional Identity in Highland Ecuador. Anthropological Quarterly 63(1):43-59. 1991 Poetics and Politics in the Ecuadorean Andes: Women's Narratives of Death and Devil Possession. Am erican Ethnologist 18(1):67-89. Crandon-Malamud, L. 1983 Why Susto? Ethnology 22:161.
270 1986 Medical Dialogue and the Political Economy of Medical Pluralism: A Case from Rural Highland Bolivia. American Ethnolog ist 13(3):463-476. 1989 Changing Times and Changi ng Symptoms: The Effects of Modernization on Mestizo Medicine in Rural Bolivia (The Case of Two Mestizo Sisters). Medical Anthropology 10(255-264). 1991 From the Fat of Our Souls: So cial Change, Political Process, and Medical Pluralism in Bolivia. Berkeley : University of California Press. Csordas, T.J. 1988 The Conceptual Status of Hegemony and Critique of Medical Anthropology. Medical Anthropol ogy Quarterly 2(4):416-421. Cunha, A. 2000 Relationship Between Acute Resp iratory Infection and Malnutrition in Children under 5 Years of Age. Acta Paediatrica 89(5):608-609. 2002 Manejo de Infeces Respiratr ias Agudas em Crianas: Avaliao em Unidades de Sade do Rio de Janeiro. [Management of Acute Respiratory Infections in Children: Evaluation in Rio de Janeiro Health Care Facilities]. Cadernos de Saude Publica 18(1):55-61. d'Angelo, G., ed. 2001 Enciclopedia del Ecuador. Ba rcelona: Oceano Grupo Editorial. Davis, S., and A. Wali 1993 Indigenous Territories and Tropi cal Forest Management in Latin America. Washington D.C.: World Bank. Demers, A.M., P. Morency, F. Mberyo-yaah S. Jaffar, C. Blais, P. Somse, G. Bobossi, and J. Pepin 2000 Risk Factors for Mortality am ong Children Hospitalized Because of Acute Respiratory Infections in Banj ui, Central African Republic. Pediatrics Infectious Diseases Journal 19:424-432. deMunck, V.C., and E.J. Sobo 1998 Using Methods in the Field. Walnut Creek: Alta Mira Press. Denno, D. M., A. Bentsi-Enchi ll, C.N. Mock, and J. Adelson 1994 Maternal Knowledge, Attit ude and Practices regarding Childhood Acute Respiratory Infections in Kumasi, Ghana. Annals of Tropical Paediatrics 14(4):293-301. DiGiacomo, S. M. 1999 Can There Be a "Cultural Epidemiology". Medical Anthropology Quarterly 13(4):436-457. Dollberg, D. D., M. L. Bo lyard, and D. L. Smith 1986 Evaluation of Physical Health Effects Due to Volcanic Hazards: Crystalline Silica in Mount St. Helens Volcanic Ash. American Journal of Public Health 76(3 (Suppl.)):53-58.
271 Douglas, R. M. 1990 The Impact of Socio Economic Fa ctors on Respiratory Infections in Children. Bulletin of the International Union against Tuberculosis and Lung Disease 65(4):8-11. Dudley, L., G. Hussey, J. Huskissen, and G. Kessow 1997 Vitamin A Status, Other Ri sk Factors and Acute Respiratory Infection Morbidity in Children. Sout h African Medical Journal 87(1):67-70. Dunn, F.L., and C.R. Janes 1986 Introduction: Medical An thropology and Epidemiology. In Anthropology and Epidemiology: Interdi sciplinary Approaches to the Study of Health and Disease. S. Gifford, ed. Pp. 3-34. Boston: D. Reidel Publishing Company. Durham, W. 1995 Political Ecology and Environmen tal Destruction in Latin America. In The Social Causes of Destruction in Latin America. M. Painter, ed. Pp. 249-264. Ann Arbor: Universi ty of Michigan Press. Embassy_of_Ecuador 2002 Ecuador: History Contempor ary. www.ecuador.org (accessed June 4, 2003). Escobar, A. 1999 After Nature: Steps to an Antie ssentialist Political Ecology. Current Anthropology 40(1):1-16. Estroff, S.E. 1988 Whose Hegemony? A Critical Commentary on Critical Medical Anthropology. Medical Anthropol ogy Quarterly 2(4):421-426. Ezzati, M., and D. Kammen 2001a Indoor Air Pollution from Biomass Combustion and Acute Respiratory Infections in Kenya: An Exposure-response Study. Lancet 358:619-624. 2001b Quantifying the E ffects of Exposure to I ndoor Air Pollution from Biomass Combustion on Acute Respirat ory Infections in Developing Countries. Environmental Heal th Perspectives 109(5):481-488. Fagbule, D., D. B. Parakoyi, and R. Spiegel 1994 Acute Respiratory Infections in Nigerian Children: Prospective Cohort Study of Incidence and Case Management. Journal of Tropical Pediatrics 40(5):279-284. FAO 1993 Debt-for-Nature Swaps to Prom ote Natural Resource Conservation. Rome: Food and Agriculture Organization of the UN. Farmer, P. 1999 Infections and Inequalities: The Modern Plagues. London: University of California Press.
272 Finerman, R. 1983 Experience and Expectation; Conflict and Change in Traditional Family Health Care Among the Quichua of Saraguro. Social Science and Medicine 17(17):1291-1298. 1987 Inside Out: Women's World View and Family Health in an Ecuadorian Indian Community. Social Science and Medicine 25(10):11571162. 1989a The Forgotten Healers: Women as Family Healers in an Andean Indian Community. In Women as Healers: Cro ss-cultural Perspectives. C.S. McClain, ed. Pp. 24-41. New Bruns wick: Rutgers University Press. 1989b Tracing Home-based Health Care Change in an Andean Indian Community. Medical Anthr opology Quarterly 3:162-174. Finerman, R., and R. Sackett 2003 Using Home Gardens to Decipher Health and Healing in the Andes. Medical Anthropology Q uarterly 17(4):459-482. First DataBank, Inc. 2002 Nutritionist Pro Vers ion 1.2.207. San Bruno, CA. Frost, M.B., R. Forste, and D.W. Haas 2005 Maternal Education and Child Nu tritional Status in Bolivia: Finding the Links. Social Scie nce & Medicine 60:395-407. Gerlach, A. 2003 Indians, Oil, and Politics: A Rec ent History of Ecuador. Wilmington, DE: Scholarly Resources. Gibson, R.S. 1993 Nutritional Assessment: A Labor atory Manual. New York: Oxford University Press. Glass-Coffin, B. 1998 The Gift of Life: Female Spir ituality and Healing in Northern Peru. Albuquerque: University of New Mexico Press. 2003 Anthropology and Shamanism: Bottom-line Considerations in Image and Practice. In Medical Pluralism in the Andes. C. Greenway, ed. Pp. 234-249. New York: Routledge. Gonzlez Ochoa, E., L. Armas Prez, J.R. Bravo Gonzlez, J. Cabrales Escobar, R. Rosales Corrales, and G. Abreu Surez 1996 Prescription of Antibiotics for Mild Acute Respiratory Infections in Children. Bulletin of the Pan Americ an Health Organization 30(2):106-117. Goodman, A.H., and T. L. Leatherman, eds. 1998 Building a New Biocultura l Synthesis: Political-economic Perspectives on Human Biology. Ann Arbor : University of Michigan Press.
273 Gove, S., and G.H. Pelto 1994 Focused Ethnographic Studies in the WHO Programme for the Control of Acute Respiratory Infect ions. Medical Anthropology 15(4):409424. Graham, M.A. 2004 No Somos Iguales ": The Effect of Household Economic Standing on Women's Energy Intake in the Andes. Social Science & Medicine 58:2291-2300. Graham, N.M.H. 2001 The Epidemiology of Respiratory Infections. In Infectious Disease Epidemiology. N.M.H. Graham, ed. Pp. 439-476. Gaithersburg, MD: Aspen Publishers. Greenway, C. 1998a Hungry Earth and Vengeful Star s: Soul Loss and Identity in the Peruvian Andes. Social Sci ence and Medicine 47(8):993-1004. 1998b Introduction: Landscapes of Heal th in the Andes. Social Science and Medicine 47(8):989-992. Hahn, R.A. 1995 Sickness and Healing: An Anthropological Perspective. New Haven: Yale University Press. Hendricks, J. 1991 Symbolic Counterhegemony among the Ecuadorian Shuar. In Nation-States and Indians In Latin Am erica. J. Sherzer, ed. Pp. 53-71. Austin: University of Texas Press. Herman, E., R.E. Black, S. Wahba, and N. Khallaf 1994 Developing Strategies to Encourage Appropriate Care-seeking for Children with Acute Respiratory In fections: An Example from Egypt. International Journal of Health Planning and M anagement 9:235-243. Herz, M., and J. Pontes Nogueira 2002 Ecuador vs. Peru: Peacemaking amid Rivalry. Boulder, CO: Lynne Rienner Publishers. Hiraoka, M., and S. Yamamoto 1980 Agricultural Development in the Upper Amazon of Ecuador. Geographical Review 70(4):423-445. Hoffman, S.M., and A. Oliver-Smith, eds. 2002 Catastrophe & Culture: The An thropology of Disaster. Santa Fe: School of American Research Press. Hoque, B. A., J. Chakraborty, J.T. Chow dhury, U.K. Chowdhury, M. Ali, S. el Arifeen, R.B. Sack 1999 Effects of Environmental Factor s on Child Survival in Bangladesh: A Case Control Study. Public Health 113(2):57-64. Hortal, M., M. Contera, C. Mogdasy, and J.C. Russi 1994 Acute Respiratory Infections in Children from a Deprived Urban Population from Uruguay. Revista do Instituto de Medicina Tropical de Sao Paulo 36(1):51-57.
274 Hudelson, P., T. Huanca, D. Charaly, and V. Cirpa 1995 Ethnographic Studies of ARI in Bolivia and their Use by the National ARI Programme. Social Sci ence and Medicine 41(12):1677-1683. Hudelson, P.M. 1994 The Management of Acute Respir atory Infections in Honduras: A Field Test of the Focused Et hnographic Study (FES). Medical Anthropology 15(4):435-446. Hurtado, O. 1980 Political Power in Ecuador. N.D. Mills, transl. Albuquerque: University of New Mexico Press. Hurtig, A., and M. San Sebastian 2002 Geographical Differences in Ca ncer Incidence in the Amazon Basin of Ecuador in Relation to Residence N ear Oil Fields. International Journal of Epidemiology 31:1021-1027. INEC 2001 VI Censo De Poblacin y V de Vivienda. Quito: In stituto Nacional de Estadstica y Censos. Inhorn, M.C. 1995 Medical Anthropology and Ep idemiology: Divergences or Convergences? Social Scienc e and Medicine 40(3):285-290. Inhorn, M.C., and K.A. Buss 1994 Ethnography, Epidemiology and Infertility in Egypt. Social Science and Medicine 39(5):671-686. Iqbal, I., S. Pervez, and S. Baig 1997 Management of Children with Acute Respiratory Infections (ARI) by General Practitioners in Multan An Ob servational Study. The Journal of the Pakistan Medical Association 47(1):24-28. Jekel, J.F., D.L. Katz, and J.G. Elmore 2001 Epidemiology, Biostatistics, and Preventive Medicine. New York: W.B. Saunders Company. Joralemon, D., and D. Sharon 1993 Sorcery and Shamanism: Curander os and Clients in Northern Peru. Salt Lake City: University of Utah Press. Kalter, H.D., R. Salgado, L. H. Moulton, P. Nieto, A. Contreras, M.L. Egas, and R.E. Black 2003 Factors Constraining Adherence to Referral Advice for Severely Ill Children Managed by the Integrated Management of Ch ildhood Illness Approach in Imbabura Province, Ecuador Acta Paediatrica 92:103-110. Kambarami, R.A., S. Rusa kaniko, and L.A. Mahomva 1996 Ability of Caregivers to Recogn ise Signs of Pneumonia in Coughing Children Aged below Five Years. The Central African Journal of Medicine 42(10):291-294.
275 Kelly, J.M., B. Osamba, R.M. Garg, M.J. Hamel, J.J. Lewis, S.Y. Rowe, A.K. Rowe, and M.S. Deming 2001 Community Health Worker Performance in the Management of Multiple Childhood Illness: Siay a District, Kenya, 1997-2001. The American Journal of Pub lic Health 91(10):1617-1624. Khan, A.Z., R. Tickoo, T. Arif, and M. Zaheer 1995 Mothers' Attitudes to Children's C hest Infections in India. Journal of the Royal Society of Health 115(5):314-317. Khan, A.J., J.A. Khan, M. Akbar, and D.G. Addiss 1990 Acute Respiratory Infections in Children: A Case Management Intervention in Abbottabad District, Paki stan. Bulletin of the Pan American Health Organization 68(5):577-585. Kimerling, J. 1991 Amazon Crude. Washington D. C.: Natural Resources Defense Council. Kinnear, P.R., and C.D. Gray 2004 SPSS 12 Made Simple. New York: Psychology Press. Korovkin, T. 1997 Indigenous Peasant Struggles and the Capitalist Modernization of Agriculture: Chimborazo, 1964-1991. Latin American Perspectives 24:2549. Koss-Chioino, J.D., T. Leat herman, and C. Greenway, eds. 2003 Medical Pluralism in t he Andes. New York: Routledge. Kotler, P., N. Roberto, and N. Lee 2002 Social Marketing: Improvi ng the Quality of Life. Thousand Oaks, CA: Sage. Kresno, Sudarti, G.G. Harrison, B. Sutrisna, and A. Reingold 1994 Acute Respiratory Illnesses in Children under Five Years in Indramayu, West Java, Indonesia : A Rapid Ethnographic Assessment. Medical Anthropol ogy 15(4):425-434. Krieger, N. 1994 Epidemiology and the Web of Causation: Has Anyone Seen the Spider? Social Science and Medicine 39(7):887-903. Kkbay, H., C. Yakinci, F.Z. Kkbay, and M. Turgut 1997 Serum Vitamin A and Beta-c arotene Levels in Children with Recurrent Acute Respiratory Infecti ons and Diarrhoea in Malaysia. Journal of Tropical Pediat rics 43(6):337-340. La_Nacin 2003 Cuatro Aos de la Evac uacin de Penipe. October 19. In La Nacin. Pp. 8. Quito. La_Prensa 2002a Entrevista Alcalde de Penipe Ju an Salazar Lopez. September 25. In La Prensa. Pp. 12. Riobamba. 2002b Entrevista Jorge Donoso, Director de Salud, Provincia de Chimborazo. September 23. In La Prensa. Riobamba.
276 2002c Se Incrementan Enfermedades Respiratorias en Chimborazo: Entrevista Adela Vimos. September 24. In La Prensa. Pp. 12. Riobamba. 2003 La Ceniza de Tungurahua Af ecta a 916 familias. June 16. In La Prensa. Pp. 1. Riobamba. 2004 Los Nios Exigen un Plan de Cambio y Desarollo. October 4. In La Prensa. Pp. 8. Riobamba. Lanata, C.F., and R.E. Black 2001 Acute Lower-Respiratory Infections. In Nutrition and Health in Developing Countries. M.W. Bloem ed. Pp. 131-162. Totowa, NJ: Humana Press Inc. Lanata, C.F., Igor Rudan, C. Boschi-Pinto, L. Tomaskovic, T. Cherian, M. Weber, and H. Campbell 2004 Methodological and Quality Iss ues in Epidemiological Studies of Acute Respiratory Infections in Children in Developing Countries. International Journal of Epidemiology 33:1362-1372. Lane, L. 2003 Hazard Vulnerability in Socioeconomic Context: An Example from Ecuador. Master's Thesis, University of South Florida. Lara, J., P. Jaramillo, and X. Haro 2000a Incremento de las Infecciones Respiratorias Agudas por Ceniza Volcnica en Nios (Penipe). Penipe: Pueblo de la Solidaridad 18:14. 2000b La Parasitosis y la Desnutrici on en el Canton Penipe. Penipe: Pueblo de la So lidaridad 19:17-18. Larme, A.C. 1998 Environment, Vulnerability, and Gender in Andean Ethnomedicine. Social Science and M edicine 47(8):1005-1015. Leatherman, T.L. 1998a Changing Biocultural Perspective on Health in the Andes. Social Science and Medicine 47(8):1031-1041. 1998b Illness, Social Relation s, and Household Production and Reproduction in the Andes. In Building a New Biocultural Synthesis: Political-economic Perspectives on Hu man Biology. T.L. Leatherman, ed. Ann Arbor: University of Michigan Press. Leatherman, T.L., A.H. Goodm an, and R. Brooke Thomas 1993 On Seeking Common Gr ound between Medical Ecology and Critical Medical Anthropology. Medica l Anthropology Quarterly 7:202-207. Lewando-Hundt, G.A ., and M.R. Forman 1993 Interfacing Anthropology and Epidemiology: The Bedouin Arab Infant Feeding Study. Social Sc ience and Medicine 36(7):957-964.
277 LLoyd-Sherlock, P., ed. 2000 Healthcare Reform and Poverty in Latin America. London: Institute of Latin American Studies. Lpez Bravo, I.M., H. S eplveda, and I. Valds 1997 Acute Respiratory Illnesses in the First 18 months of Life. Pan American Journal of Pub lic Health 1(1):9-17. Los_Andes 2004 La Tierra de la Solidaridad Festeja 437 aos de Fundacin. In Los Andes. Pp. 3. Riobamba. Lucero, M.G., G. Williams, H. No hynek, P.H. Makela, and I. Riley 2003 Pneumococcal Conj ugate Vaccines for Pr eventing Pneumococcal Disease in Children under Two Years of Age (Protocol for a Cochrane Review). The Cochrane Library. Luque, J. 1997 Power in an Ecuadorian Indi genous Foundation: Conservation Planning in Grassroots Development. Master's Thesis, Arizona State University. Macdonald, T. 1981 Indigenous Response to an Ex panding Frontier: Jungle Quichua Economic Conversion to Cattle Ranching. In Cultural Transformation and Ethnicity in Modern Ecuador. N. E. Wh itten, ed. Pp. 356-386. Chicago: University of Illinois Press. Malilay, J., A. Henderson, M. McGeehin, and W.D. Flanders 1996 Public Health Surveillance afte r a Volcanic Eruption: Lessons from Cerro Negro, Nicaragua 1992. Bull etin of PAHO 30(3):218-226. Martin, P. 2003 Globalization of Contentious Politics: The Amazonian Indigenous Rights Movement. New York: Routledge. McElroy, A., and P.K. Townsend 1989 Medical Anthropology in Ecological Perspective. London: Westview. McKee, L.A. 1987 Ethnomedical Treatm ent of Children's Diarr heal Illnesses in the Highlands of Ecuador. Social Sci ence and Medicine 25(10):1147-1155. 2003 Ethnomedicine and Enculturat ion in the Andes of Ecuador. In Medical Pluralism in the Andes. J.D. Koss-Chioino, T. Leatherman, and C. Greenway, eds. Pp. 131-147. New York: Routledge. McNee, Andrew, N. Khan, S. Dawson, J. Gunsalam, V.L. Tallo, L. Manderson, I. Riley 1995 Responding to Cough: Boholano Illness Classification and Resort to Care in Response to Childhood ARI. Social Science and Medicine 40(9):1279-1289. Meyer, C. 1993 Environmental NGOs in Ec uador: An Economic Analysis of Institutional Change. Journal of Developing Areas 27(2):192-210.
278 Miles, A., and T. Leatherman 2003 Perspectives on Medical Anthropology in the Andes. In Medical Pluralism in the Andes. J.D. KossChioino, T. Leatherman, and C. Greenway, eds. Pp. 3-15. New York: Routledge. Miller, D. 1999 The Future Role of Internat ional Agencies in Control of Acute Respiratory Tract Infections. Clinical Infectious Diseases 28(2):230-233. Mishra, V., and R. Retherford 1997 Cooking Smoke Increases the Ri sk of Acute Respiratory Infection in Children. Honolulu, H I: International Institute for Population Sciences : East-West Center, Pr ogram on Population. Moore, K.R., H. Duffel, A. Nicholl, and A. Searl 2002 Monitoring of Airborne Partic ulate Matter during the Eruption of Soufriere Hills Volcano, Montserrat. In The Eruption of Soufriere Hills Volcano, Montserrat, from 1995 to 1999. B.P. Kokelaar, ed. Pp. 557-566. London: Geological Society of London. Morgan, L.M. 1998 Latin American Social Medi cine and the Politics of Theory. In Building a New Biocultural Synthesis : Political-economic Perspectives on Human Biology. T.L. Leatherman, ed. Pp. 407-424. Ann Arbor: University of Michigan Press. Morsy, S.A. 1996 Political Economy in Medical Anthropology. In Medical Anthropology: Contemporary Theory and Method. M.J. Thomas, ed. Pp. 21-40. Westport, CT: Praeger Publishers. Mull, D.S., and J.D. Mull 1994 Insights from Community-based Research on Child Pneumonia in Pakistan. Medical Anth ropology 15(4):335-352. Mull, D.S., J.D. Mull, and M.Z. Malik Kundi 1994 Mothers' Perceptions of Severe Pneumonia in their Own Children: A Controlled Study in Pakistan. Social Science & Medicine 38(7):973-987. Muratorio, B. 1998 Indigenous Women's Identitie s and the Politics of Cultural Reproduction in the Ecuadorian Am azon. American Anthropologist 100(2):409-420. National Research Council 1989 Recommended Dietary Allowanc es. Washington, D.C.: National Academy Press. National Academies of Sciences 2004 Recommended Daily Allowanc es. Washington, D.C.: National Academy Press. Nations, M.K. 1986 Epidemiological Research on In fectious Disease: Quantitative Rigor or Rigormortis? Insights from Ethnomedicine. In Anthropology and Epidemiology. S.M. Gifford, ed. Pp. 97123. Boston: D. Re idel Publishing Company.
279 Nichter, M. 1992 Anthropological Approaches to the Study of Ethnomedicine. Amsterdam: Gordon and Breach. 1993 Social Science Lessons fr om Diarrhea Research and their Application to ARI. Hum an Organization 52(1):53-67. 1994 Acute Respiratory Infection. Introduction. Medical Anthropology 15(4):319-334. Nichter, M., and M. Nichter 1994 Acute Respiratory Illness: P opular Health Culture and Mother's Knowledge in the Philippines. Medical Anthropology 15(4):353-375. Nicoll, A. 2000 Integrated Management of Child hood Illness in Resource-Poor Countries: An Initiative from the Worl d Health Organization. Transactions of the Royal Society of Tropical Medicine and Hygiene 94(1):9-11. Noji, E.K., ed. 1997 The Public Health Consequences of Disasters. Oxford: Oxford University Press. Oliver-Smith, A. 1999 "What is a Disaster?" Anthropolog ical Perspectives on a Persistent Question. In The Angry Earth: Disaster in Anthropological Perspective. A.H. Oliver-Smith, Susanna M., ed. Pp. 18-34. New York: Routledge. Oliver-Smith, A., and S.M. Hoffman 1999 The Angry Earth: Disaster in Anthropological Perspective. New York: Routledge. Ong, A. 1987 Spirits of Resistance and Capita list Discipline: Factory Women in Malaysia. Albany: State University of New York Press. Oths, K.S. 1998 Assessing Variation in Health Status in the Andes: A Biocultural Model. Social Science and Medicine 47(8):1017-1030. 1999 Debilidad: A Biocultural A ssessment of an Embodied Andean Illness. Medical Anthropol ogy Quarterly 13(3):286-315. Oyejide, C.O., and E.A. Oke 1995 An Ethnographic Study of Acute Respiratory Infections in Four Local Government Areas of Nigeria African Journal of Medicine and Medical Sciences 24(1):85-91. PAHO 2000 Erupciones Volcanicas y Prote ccion de la Salud. Pp. 67. Quito, Ecuador: PanAmerican Health Organization. 2002 Ecuador: Data updated for 2001. www.paho.org Pan American Health Organization (accessed September 30, 2005).
280 PAHO/SHA 2001 Charts on the Morbidity Situat ion on Children under Five in the Region of the Americas, Vol. 2003: Pan American Health Organization. Special Program for Health Anal ysis (PAHO/SHA) and Program on Communicable Diseases PAHO/HCP/HCT). Pallares, A. 2002 From Peasant Struggles to Indian Resistance: The Ecuadorian Andes in the Late Twentieth Centur y. Norman, OK: University of Oklahoma Press. Pandey, A., and A. Chakraborty 1996 Undernutrition, Vitamin A Deficiency and ARI Morbidity in Underfives. Indian Journal of Public Health 40(1):13-16. Pandey, M.R., P.R. Sharma, B.B. Gubhaj u, G.M. Shakya, R.P. Neupane, A. Gautam, and I.B. Shrestha 1989 Impact of a Pilot Acute Resp iratory Infection (ARI) Control Programme in a Rural Community of t he Hill Region of Nepal. Annals of Tropical Paediatrics 9:212-220. Pandey, M.R., N.M.P. Daulai re, E.S. Starbuck, R.M. Houston, and K. McPherson 1991 Reduction in Total Under-five Mortality in Western Nepal through Community-based Antimicrobial Treat ment of Pneumonia. The Lancet 338:993-997. Paulson, S., L.L. Gezon, and M. Watts 2003 Locating the Political in Polit ical Ecology: An Introduction. Human Organization 62(3):205-217. Pelto, G. 1996 Control of Acute Respiratory Infections. Medical Anthropology 18:28-31. Pelto, P.J. 1988 A Note on Critical Medical Anthropology. Medical Anthropology Quarterly 2(4):435-437. Pelto, P.J., and G.H. Pelto 1990 Field Methods in Medical Anthropology. In Medical Anthropology: A Handbook of Theory and Method. C.F. Sargent, ed. Pp. 269-297. Westport, CT: Greenwood Press. 1996 Research Designs in Medical Anthropology. In Handbook of Medical Anthropology: Contemporary T heory and Method. T. Johnson, ed. Pp. 293-394. Westport, CT: Greenwood Press. Phelp, J. 1992 Crisis and the Debt-for-Nature Strategy: A Discussion and Critique. Melbourne: Monash. Pillai, R.K., S.V. Williams, H.A. Glick, D. Polsky, J.A. Berlin, and R.A. Lowe 2003 Factors Affecting Decisions to Seek Treatment for Sick Children in Kerala, India. Social Sci ence & Medicine 57(5):783-790.
281 Ponce, J. 2001 Cultura y Economia Solidaria: Siete Lecciones de Solidaridad. Riobamba: Editoria l Pedagogica Freire. Portes, S.A., E.E. da Silva, M.M. Siqueira, A.M. De F ilippis, M.M. Krawczuk, and J. Nascimento 1998 Enteroviruses Isolated from Patients with Acute Respiratory Infections during Seven Years in Rio de Janeiro (1985-1991). Revista do Instituto de Medicina Tropica l de Sao Paulo 40(6):337-342. Qazi, S.A., G.N. Rehman, and M. Khan 1996 Standard Management of Acute Respiratory Infections in a Children's Hospital in Pakistan: Impact on Antibiotic Use and Case Fatality. Bulletin of the World Health Organization 74(5):501-507. Rachowiecki, R. 1992 Ecuador and the Galapagos Isla nds. Hawthorne, Australia: Lonely Planet Publications. Rahman, M.M., D. Mahalanabi s, J.O. Alvarez, M.A. Wahed, M.A. Islam, D. Habte, and M. Khaled 1996 Acute Respiratory Infections Prevent Improvement of Vitamin A Status in Young Infants Supplemented with Vitamin A. The Journal of Nutrition 126(3):628-633. Rahman, M.M., and A. Rahman 1997 Prevalence of Acute Respirat ory Tract Infection and its Risk Factors in Under Five Children. B angladesh Medical Research Council Bulletin 23(2):47-50. Rahman, M.M., and M. Shahidullah 2001 Risk Factors for Acute Respir atory Infections among the Slum Infants of Dhaka City. Bangladesh Medical Research Council Bulletin 27(2):55-62. Ramakrishnan, U., and R. Martorell 1998 The Role of Vitamin A in R educing Child Mortality and Morbidity and Improving Growth. Salud Publ ica de Mexico 40(2):189-198. Rashid, S.F., A. Hadi, K. Afsana, and S.A. Begum 2001 Acute Respiratory Infections in Rural Bangladesh: Cultural Understandings, Practices and the Ro le of Mothers and Community Health Volunteers. Tropica l Medicine & Internationa l Health 6(4):249-255. Ray, S. K., A. Haldar, B. Bi swas, R. Misra, and S. Kumar 2001 Epidemiology of Undernutriti on. Indian Journal of Pediatrics 68(11):1025-1030. Reyes, H., S. Villalpando, R. Prez-Cuev as, L. Rodriguez, M. Prez-Cuevas, I. Montalvo, and H. Guiscafre 2002 Frequency and Determinants of Vita min A Deficiency in Children under 5 years of Age with Pneumonia. Archives of Medical Research 33(2):180-185.
282 Rice, A.L., L. Sacco, A. Hyder, and R.E. Black 2000 Malnutrition as an Underly ing Cause of Childhood Deaths Associated with Infectious Diseases in Developing Countries. Bulletin of the World Health Organization 78(10):1207-1221. Roseberry, W. 1989 Anthropologies and Histories: Essays in Culture, History, and Political Economy. London: Rut gers University Press. Rubel, A.J., and M.R. Hass 1990 Ethnomedicine. In Medical Anthropology: A Handbook of Theory and Method. C.F. Sargent, ed. Pp. 115-131. New York: Greenwood Press. Rudel, T.K. 1995 When do Property Rights Matter ? Open Access, Informal Social Controls, and Deforestation in the Ecuadorian Amazon. Human Organization 54(2):187-194. Senz de Tejada, S. 1997 Manejo de las Infecciones Resp iratorias Agudas (IRA) en una Comunidad Kaqchiquel de Guatemal a.[Management of Acute Respiratory Infections in a Kaqchiquel Community in Guatemala]. Revista Panamericana de Salud P ublica 1(4):259-265. SAS Institute, Inc. 2001 SAS Release 8.2. Cary, NC. Sawyer, S. 2004 Crude Chronicles: Indigenous Politics, Multinational Oil, and Neoliberalism in Ecuador. Durham, NC: Duke University Press. Sazawal, S., and R.E. Black 1992 Meta-analysis of Interventi on Trials on Case-management of Pneumonia in Community Se ttings. The Lancet 340:528-533. Schensul, J.J., M.D. LeCompte, B. K. Nastasi, and S.P. Borgatti 1999 Enhanced Ethnographic Met hods: Audiovisual Techniques, Focused Group Interviews, and Elicit ation Techniques. London: Altamira Press. Scheper-Hughes, N. 1990 Three Propositions for a Critic ally Applied Medical Anthropology. Social Science and M edicine 30(2):189-197. 1992 Death Without Weeping. Berkeley : University of California Press. Scheper-Hughes, N., and M. Lock 1987 The Mindful Body: A Prolegomen on to Future Work in Medical Anthropology. Medical Anthropology Quarterly 1(1):6-41. Schroeder, B. 1991 Indians in the Halls of Academ e: Rural Andean Peoples Confront Social Science. Peasant Studies 18(2):97-116. Searl, A., A. Nicholl, and P.J. Baxter 2002 Assessment of the Exposure of Islanders to Ash from the Soufriere Hills Volcano, Montserrat, Brit ish West Indies. Occupational and Environmental Medi cine 59(8):523-531.
283 Selverston-Scher, M. 2001 Ethnopolitics in Ec uador: Indigenous Rights and the Strengthening of Democracy. Miami: No rth-South Center Press. Semba, R.D. 2001 Nutrition and Infectious Diseases. In Infectious Disease Epidemiology. N.M.H. Graham, ed. Pp. 283-297. Gaithersburg: Aspen. 2004 Vitamin A. In Diet and Human Immune Function. D. Hughes, L.G. Darlington, and A. Bendich, eds Pp. 105-132. Totowa, NJ: Humana Press. Shann, F. 1999 Introduction: Acute Respirator y Tract Infections-The Forgotten Pandemic. Clinical Infecti ous Diseases 28(2):188-191. Sharon, D. 1978 Wizard of the Four Winds: A Shaman's Story. London: The Free Press. Simon, A., M. Janabi, and G. Kalmayem 1996 Caretakers' Management of Child hood Acute Respiratory Infections and the Use of Antibiotics, Bohol, the Philippines. Human Organization 55(1):76-83. Singer, M. 1993 A Rejoinder to Wiley's Critique of Critical Medical Anthropology. Medical Anthropology Quarterly 7:185-191. 1998 The Development of Critical M edical Anthropology: Implications for Biological Anthropology. In Building a New Biocultural Synthesis: Politicaleconomic Perspectives on Human Biol ogy. T.L. Leatherman, ed. Pp. 93123. Ann Arbor: University of Michigan Press. Singer, M., and H. Baer 1995 Critical Medical Anthropology. Amityville, NY: Baywood Publishing Company. SPSS, Inc. 2005 SPSS, Release 13.0. Chicago, IL. Stansfield, S.K. 1987 Acute Respiratory Infections in the Developing World: Strategies for Prevention, Treatment and C ontrol. Pediatrics Infe ctious Diseases Journal 6(7):622-629. Stein, W.W. 1981 The Folk Illness: Entity or Nonentity? An Essay on Vicos Disease Ideology. In Health in the Andes. J. D onahue, ed. Pp. 50-67. Washington D.C.: American Anthropological Association. Stewart, M.K., B. Parker, J. Chakraborty, and H. Begum 1994 Acute Respiratory Infecti ons (ARI) in Rural Bangladesh: Perceptions and Practices. Medi cal Anthropology 15(4):377-394.
284 Stonich, S.C. 1999 Comments on After Nature by Arturo Escobar. Current Anthropology 40(1):23-24. Taussig, M. 1980 The Devil and Commodity Fetishis m in South America. Chapel Hill: University of North Carolina Press. Teka, T., and M. Dagnew 1995 Health Behaviour of Rural Mot hers to Acute Respiratory Infections in Children in Gondar, Ethiopia. East African Medical Journal 72(10):623625. The_Wellcome_Trust 2001 Acute Respiratory Infection. In Topics in International Health. New York: CABI Publishing. Tobin, G.A., and L.M. Whiteford 2002 Community Resilience and Volcano Hazard: The Eruption of Tungurahua and Evacuation of the Fal das in Ecuador. Disasters 26(1):2848. Trostle, J. 1986 Anthropology and Epidemiology in the Twentieth Century: A Selective History of Collaborative Proj ects and Theoretical Affinities, 1920 to 1970. In Anthropology and Epidemiology. S.M. Gifford, ed. Pp. 59-94. Boston: D. Reidel Publishing Company. Trostle, J.A., and J. Sommerfeld 1996 Medical Anthropology and Ep idemiology. Annual Review of Anthropology 25:253-274. True, W.R. 1996 Epidemiology and Medical Anthropology. In Handbook of Medical Anthropology: Contemporary Theory and Method. T.M. Johnson, ed. Pp. 325-346. Westport, CT: Greenwood Press. Truswell, A.S. 1999 Dietary Goals and Guideli nes: National and International Perspectives. In Modern Nutrition in Health and Disease. M. Shils, J.A. Olson, M. Shike, and A.C. Ross, eds. Pp. 1727-1741. Baltimore, MD: Williams & Wilkins. Tupasi, T.E., N.V. Mangubat, M.A.S. Sunico, D.M. Magdangal, E.E. Navarro, Z.A. Leonor, S. Lupisan, F. Medalla, and M.G. Lucero 1988 Determinants of Morbidity and Mo rtality Due to Acute Respiratory Infections: Implications for Interventi on. The Journal of Infectious Diseases 157(4):615-623. UNICEF 2003 At a Glance: Ecuador. www.unicef.org (accessed October 30, 2003). Uquillas, J. 1984 Colonization and Spontaneous Settlement in the Ecuadoran Amazon. In Frontier Expansion in Am azonia. C.H. Wood, ed.
285 USAID 2004 USAID Country Health Statis tical Report. www.usaid.gov United States Agency of International De velopment (accessed September 30, 2005). Vayda, A., and B. Walters 1999 Against Political Ecology. Human Ecology 27:167-179. Vickers, W. 1988 Process and Problems of Land Demarcation for a Native Amazonian Community in Ecuador. Law and Anthropology 3:203-245. Villoldo, A., and E. Jendresen 1990 The Four Winds: A Shaman's Odyssey into the Amazon. San Francisco: Harper & Row. Visser, D., and G. Mendoza 1993 Debt-for-nature Swaps in Latin America. Journal of Forestry 92(6):13-16. Watts, M. 2002 Political Ecology. In A Companion to Economic Geography. E. Sheppard, ed. Oxfo rd: Blackwell. Weismantel, M.J. 1988 Food, Gender and Poverty in th e Ecuadorian Andes. Philadelphia: University of Pennsylvania. Weiss, M.G. 1988 Cultural Models of Diarrheal Illness: Conceptual Framework and Review. Social Science & Medicine 27(1):5-16. Weller, S. 1998 Structured Interviewing and Questionnaire Construction. In Handbook of Methods in Cultural An thropology. H.R. Bernard, ed. Pp. 365-410. London: Alta Mira Press. Weller, S.C., and A. Kimball Romney 1988 Systematic Data Collect ion. London: Sage Publications. Weller, S.C., T.R. Ruebush, and R.E. Klein 1997 Predicting Treat ment-Seeking Behavior in Guatemala: A Comparison of the Health Services Research and Decision-Theoretic Approaches. Medical Anthropol ogy Quarterly 11(2):224-245. Whiteford, L.M., and G.A. Tobin 2001 Por los Nios: Health and Illness among Volcano Evacuees. Annual Meeting of the Society for A pplied Anthropology, Merida, Mexico, 2001. 2002 In the Shadow of the Volc ano: Human Health and Community Resilience following Forced Evacuation and Periodic Ash Fall Technical Report. Tampa: University of So uth Florida Center for Disaster Management and Humanitarian Assistance.
286 2004 Saving Lives, Destroying Li velihoods: Emergency Evacuation and Resettlement Policies. In Unhealthy Health Policies: A Critical Anthropological Examination. M. Spri nger, ed. Pp. 189-202. Walnut Creek, CA: AltaMira Press. Whitehead, N., and B. Ferguson 1992 The Violent Edge of Empire. In War in the Tribal Zone: Expanding States and Indigenous Warfare. B. Ferguson, ed. Santa Fe: School of American Research Press. Whitney, E., and S. Rolfes, eds. 2005 Understanding Nutr ition. Belmont, CA: Thomson Wadsworth. Whitten, N.E. 1985 Sicuanga Runa. The Other Si de of Development in Amazonian Ecuador. Urbana: University of Illinois Press. 1996 The Ecuadorian Levanta miento Indigena of 1990 and the Epitomizing Symbol of 1992: Reflec tions on Nationalism, Ethnic-Bloc Formation, and Racialist Ideologies. In History, Power, and Identity: Ethnogenesis in the Americas. J.D. Hill, ed. Iowa City: University of Iowa Press. WHO 1993 Focused Ethnographic Study of Acute Respiratory Infections. Geneva, Switzerland. 2004 Child Health Epidemiology Reference Group. www.who.int/childadolescent-health/New_Publicat ions/Overview/ari_db.htm. Wibbelsman, M. 2003 Appendix: General Information on Ecuador. In Millenial Ecuador: Critical Essays on Cultural Transfo rmations and Social Dynamics. N.E. Whitten, ed. Iowa City: University of Iowa Press. Wiley, A.S. 1993 Evolution, Adaptation, and t he Role of Biocultural Medical Anthropology. Medical Anthropology Quarterly 7:192-201. Williams, B.G., E. Gouws, C. Bo schi-Pinto, J. Bryce, and C.Dye 2002 Estimates of World-wide Distri bution of Child Deaths from Acute Respiratory Infections Lancet 2(1):25-32. Wilson, M.R., V. Stone, R.T. Cullen, A. Searl, R.L. Maynard, and K. Donaldson 2000 In Vitro Toxicology of Resp irable Montserrat Volcanic Ash. Occupational and Environmental Medicine 57(11):727-733. Yach, D. 1992 The Use and Value of Qualitative Methods in Health Research in Developing Countries. Social Sc ience and Medici ne 35(4):603-612. Zamosc, L. 1994 Agrarian Protest and the Indi an Movement in the Ecuadorian Highlands. Latin American Res earch Review 29(3):37-68.
287 Zevallos, J.V. 1989 Agrarian Reform and Structur al Change: Ecuador Since 1964. In Searching for Agrarian Reform in Latin America. W.C. Thiesenhusen, ed. Boulder, CO: Westview Press.
288 APPENDIX A: ENGLISH VERSION OF SEMI-STRUCTURED QUESTIONNAIRE
289 ENGLISH guide for Spanish version INTRODUCTION Hello, my name is Juan Luque. We are doing a survey in this community with permission from the Univer sity of South Florida to learn about childrens health. We are looking fo r the participation of people who have: 1) lived in the community for at least 3 years; 2) who are between the ages of 18 and 65 years of age; 3) who is a mother or caretaker of a child less than 5 years of age; and 4) speaks Spanish. If you or someone in your household meet s these requirements, we would very much appreciate any help you could provide us. I will be asking about the health of your child and their experiences with re spiratory problems, and your efforts to treat your child. All information you prov ide will be kept confidential. Your participation is voluntary. There are no foreseeable risks if you agree to participate. If you participate in this study, we have small beauty compacts in exchange for your participation. If you c hoose not to participate, there are no consequences, and if you decide that you w ant to end the interview at any time, you may do so and still receive the compens ation. The survey will take about 30 minutes to complete. If you have any questions about the research, I will be in the community for six months in CEBYCAM and will be happy to answer any questions you might have. My phone number in Riobamba is 2968-087. This study will interview approximately 260 participants. DATE: _____________________ SUBJECT ID NO. ____________
290 1) How long have you lived continuously in this community? 1 = between 3 and 5 years 2 = 6 to 10 years 3 = 11 to 15 years 3 = more than 15 years 2) How old are you? _______ 3) What is your marriage status? 1 = Single 2 = Married 3 = Living Together 4 = Divorced 5 = Widow 4) How many people including your self live in your home? _______ 5) What relationship are the people liv ing with you and what are their ages? a) Family Friend b) Family Friend c) Family Friend d) Family Friend e) Family Friend 6) What is the age of the younges t child living in the home? 1 = less than 6 months 2 = 6-17 months 3 = 18-29 months 4 = 30 months< 5 years 7) What is the sex of your child? 1 = Male 2 = Female 8) How many people shar e his/her bedroom? 1 = own room 2 = 1 person 3 = 2 persons 4 = 3 or more persons
291 IF OTHER THAN OWN ROOM: 8A) Does this person or any of those people smoke cigarettes? 1 = Yes 2 = No 8B) Does this child have his/her own bed? 1 = Yes 2 = No, shared with 1 3 = No, shared with 2 4 = No, shared with 3 + 9) How is your home heated? 1 = Coal 2 = Wood 3 = Gas 4 = Not heated 10) What fuel is used most for cooking in your home? 1 = Coal 2 = Wood 3 = Gas 4 = Wood and Gas 5 = Fuel, oil, kerosene 11) Does your child sleep where you cook? 1 = Yes 2 = No 12) Do you have a cat, dog, or bird living in your home? 0 = No (skip to 13) 1 = Cat 2 = Dog 3 = Cat and Dog 13) Do the pets enter the home? 1 = Yes 2 = No
292 These questions pertain mainly to y our youngest childs chest. Please answer yes or no if possibl e. If a question does not a ppear to be applicable to your child, answer that it does not apply. COUGH 14) Does he/she usually have a cough with colds? 1 = Yes 2 = No 15) Does he/she usually have a cough apart from colds? 1 = Yes 2 = No IF ANSWER IS YES TO 14 OR 15: 15A) Does he/she cough on most days (4 days or more days per week) for as much as 3 months of the year? 1 = Yes 2 = No 8 = Does not apply 15B) For how many years has he/she had this cough? _________________ number of years 8 = Does not apply CONGESTION AND/OR PHLEGM 16) Does the child usually seem conges ted in the chest or bring up phlegm with colds? 1 = Yes 2 = No IF ANSWER IS YES TO 16: 16A) Does this child seem congested or bring up phlegm, sputum, or mucus from his/her chest on most days (4 or more days per week) for as much as 3 months a year? 1 = Yes 2 = No 8 = Does not apply
293 16B) For how many years has he/s he seemed congested or raised phlegm, sputum, or mucus from his/her chest? _________________number of years 8 = Does not apply 17) Does this child get attacks of (inc reased) cough, chest congestion or phlegm lasting for 1 week or more each year? 1 = Yes 2 = No IF ANSWER IS YES TO 17: 17A) For how many years? ______ Number of years 8 = Does not apply 17B) On average, how many chest colds per year does he/she get? ______ Avg number per year 8 = Does not apply WHEEZING 18) Does this childs chest ever sound wheezy or whistling: 1 = Yes 2 = No ( Skip to #18 ) 18A) When he/she has a cold? 1 = Yes 2 = No 18B) Occasionally ap art from colds? 1 = Yes 2 = No 18C) Most days or nights? 1 = Yes 2 = No IF ANSWER IS YES TO 18B OR 18C: 18D) For how many years has wheezing or whistling in the chest been present? ______ # of years 8 = Does not apply
294 19) Has this child ever had an attack of wheezing that has caused him/her to be short of breath? 1 = Yes 2 = No IF ANSWER IS YES TO 19: 19A) Has he/she had 2 or more such episodes? 1 = Yes 2 = No 19B) Has he/she ever required medici ne or treatment fo r the attack(s)? 1 = Yes 2 = No 19C) How old was this child when he/she had his/her first such attack? ______ Number of years 8 = Does not apply 19D) Is or was his/her breathing completely normal between attacks? 1 = Yes 2 = No 8 = Does not apply 20) Does this child ever get attacks of wheezing after he/she has been playing hard or exercising? 1 = Yes 2 = No CHEST ILLNESSES 21) During the past 3 years has this ch ild had any chest illness that has kept him/her from his/her usual activi ties for as much as 3 days? 1 = Yes 2 = No IF ANSWER IS YES TO 21: 21A) Did he/she bring up more phlegm or seem more congested than usual with any of these illnesses? 1 = Yes 2 = No 8 = Does not apply
295 21B) How many illnesses like this has he/she had in the past 3 years? 1 = < 1 illness per year 2 = 1 illness per year 3 = 2-5 illnesses per year 4 = > 5 illnesses per year 8 = Does not apply 21C) How many of these illnesses have lasted for as long as 7 days? ______ Number of illnesses 8 = Does not apply 22) Was he/she ever been hospitalized for a severe chest illness or chest cold before the age of 2 years? 1 = Yes, only once 2 = Yes, 2 times 3 = Yes, 3 or more times 4 = No 23) Did this child have any other severe chest illness or chest cold before the age of 2 years? 1 = Yes 2 = No OTHER ILLNESSES 24) Has the child had any of the followin g illnesses, and if yes, at what age? 24A) Measles or Rubeola 1 = Yes 2 = No ______ At Age 24B) Sinusitis 1 = Yes 2 = No ______ At Age 24C) Bronchitis 1 = Yes 2 = No ______ At Age
296 24D) Pneumonia 1 = Yes 2 = No ______ At Age 24E) Amygdalitis 1 = Yes 2 = No ______ At Age 24F) Broncopneumonia 1 = Yes 2 = No ______ At Age 25) Has a doctor ever said t hat this child has asthma? 1 = Yes 2 = No IF ANSWER IS YES TO 25: 25A) At what age did his/her asthma begin? ______ Age in years 25B) Does he/she still have asthma? 1 = Yes 2 = No 25C) Does he/she currently take m edicine or treatment for asthma? 1 = Yes 2 = No IF ANSWER IS NO TO 25 B: 25D) At what age did his/her asthma stop? ______ Age in years ALLERGIES 26) Has a doctor ever said that this ch ild had an allergic reaction to food or medicine? 1 = Yes, food only 2 = Yes, medicine only 3 = Yes, food & medicine 4 = No
297 27) Has a doctor ever said that this chil d had an allergic reaction to pollen or dust? 1 = Yes 2 = No 28) Has a doctor ever said that this child had an allergic skin reaction to detergents or other chemicals? 1 = Yes 2 = No 29) Did this child ever receive allergy shots? 1 = Yes 2 = No FAMILY HISTORY We would like to obtain some informa tion about the parents or guardian living with this child. 30) Does anyone in the household smoke cigarettes? 1 = Yes 2 = No If yes, specify ________________________ _____________________ ___________________ 31) Please indicate whether the female adult is: 1 = Natural mother 2 = Stepmother 3 = Other 32) What is the highest grade of school completed? 1 = Primary School 2 = Secondary School 3 = Superior _________ total years 33) What is her present job? __________________ ________________________ ______________________ 34) How long have you had this job? ________________________ _____________________ ___________________
298 USE OF HEALTH RESOURCES Please choose which types of provider you seeks for the following childhood illnesses and why: 35) Cough and runny nose 1 = Household remedy 2 = Health clinic 3 = Nothing Why? ________________________ _____________________ ___________________ 36) Cough and fever 1 = Household remedy 2 = Health clinic 3 = Nothing Why? ________________________ _____________________ ___________________ 37) Fever and runny nose 1 = Household remedy 2 = Health clinic 3 = Nothing Why? ________________________ _____________________ ___________________ 38) Cough, fever, and runny nose 1 = Household remedy 2 = Health clinic 3 = Nothing Why? ________________________ _____________________ ___________________ 39) Cough and sputum 1 = Household remedy 2 = Health clinic 3 = Nothing Why? ________________________ _____________________ ___________________ 40) Cough, sputum and runny nose 1 = Household remedy 2 = Health clinic 3 = Nothing Why? ________________________ _____________________ ___________________ 41) Rapid breathing 1 = Household remedy 2 = Health clinic 3 = Nothing Why? ________________________ _____________________ ___________________ 42) Indrawn chest 1 = Household remedy 2 = Health clinic 3 = Nothing Why? ________________________ _____________________ ___________________
299 43) Rapid breathing and indrawn chest 1 = Household remedy 2 = Health clinic 3 = Nothing Why? ________________________ _____________________ ___________________ Please rate the severity of the followin g symptoms: 44) Cough and runny nose 1 = not serious 2 = moderate 3 = serious 4 = emergency 45) Cough and fever 1 = not serious 2 = moderate 3 = serious 4 = emergency 46) Fever and runny nose 1 = not serious 2 = moderate 3 = serious 4 = emergency 47) Cough, fever, and runny nose 1 = not serious 2 = moderate 3 = serious 4 = emergency 48) Cough and sputum 1 = not serious 2 = moderate 3 = serious 4 = emergency 49) Cough, sputum, runny nose 1 = not serious 2 = moderate 3 = serious 4 = emergency 50) Rapid breathing 1 = not serious 2 = moderate 3 = serious 4 = emergency 51) Indrawn chest 1 = not serious 2 = moderate 3 = serious 4 = emergency 52) Rapid breathing, indrawn chest 1 = not serious 2 = moderate 3 = serious 4 = emergency Description of Chest Indrawing 53) Please view this photograph of a child and describe the symptoms you observe ________________________ _____________________ ___________________ ________________________ _____________________ ___________________ Please tell me the last time you used each of the followi ng health resources for your child and for what purpose, and what your opinion is of the resource
300 54) CEBYCAM Health Clinic ________________________ _____________________ ___________________ ________________________ _____________________ ___________________ 55) Health Subcenter ________________________ _____________________ ___________________ ________________________ _____________________ ___________________ 56) Pharmacist ________________________ _____________________ ___________________ ________________________ _____________________ ___________________ 57) Traditional Healer ________________________ _____________________ ___________________ ________________________ _____________________ ___________________ 58) What is the difference between a cold cough and a hot cough and what are the treatments? ________________________ _____________________ ___________________ ________________________ _____________________ ___________________ 59) Are the children under 5 in your house vaccinated? 1 = Yes 2 = No
301 INVENTORY OF HOUSEHOLD REMEDIES 60) Record all remedies in household, what is each good for? Where do they come from? What does it co st? When was it last used? Remedy Use Origin Cost Last Used
302 61) DIET OF CHILDREN UNDER FIVE (24 Hour Recall) Place Eaten Time Description of Food or Drink Brand Name Amount
303 62) Are there certain foods that you withhold from your child when he/she is sick? ________________________________________________________________________ HOUSING We would like to obtain some in formation about the type of housing. 63) What type of tenant are you? 1 = Own 2 = Rent 3 = Family 4 = Other _______ 64) How many rooms are in the house? 1 = Two 2 = Three 3 = Four 4 = Five or more 64) Where do you get your water to cook and drink? 1 = Well 2 = Faucet 3 = Bottled water 4 = Rain 5 = Other ________ 65) Do you boil the water to drink? 1 = Yes 2 = No SOCIOECONOMIC STATUS 66) Answer yes if you have any of these things: 1 = monthly income _______ 2 = any appliance 3 = > three rooms in home 4 = non-dirt floor 5 = Brick or block walls 6 = land ownership 7 = television 8 = bicycle 9 = > 3 livestock animals
304 67) What is the main income contri bution for the family and from whom? __________________ ________________________ __________________ 68) How many family mem bers are economically dependent? __________________ ________________________ __________________ 69) Do you receive monetary aid from friends or family outside of Penipe? 1 = Yes ________________ 2 = No If the respondent works in agriculture ask: 70) How often do you work in agriculture? 1 = permanently 2 = occasionally 3 = seasonally 71) The land you work on is: 1 = owned 2 = rented 3 = belongs to relative 4 = belongs to employer Thank you for your time today. A ll the information you provided is confidential. This information will be us ed to gain a better understanding of the respiratory sicknesses that are affecting the children in your community. The findings of this stud y will be related to the appropriate local health authorities.
305 APPENDIX B: SPANISH VERSION OF SEMI-STRUCTURED QUESTIONNAIRE
306 INTRODUCCION Hola, me llamo Juan Luque. Estamos realizando una encuesta en este barrio para un estudio que realiza de la Universidad del Sur de la Florida, a fin de obtener datos sobr e la salud de los nios. Buscamos la participacin de personas quienes: 1) hayan vivido en esta comunidad por 3 aos al menos; 2) tengan entre 18 y 65 aos o ms aos de edad; 3) sean madres de nios menores de cinco aos o alguien que los cuida; 4) que hablen espaol. Si usted o alguien en la casa cumple c on estos requisitos, les agradeceramos mucho cualquier ayuda que puedan prestar nos. Preguntamos sobre la salud de su nio o nia y las experiencias con problemas respiratorios agudos, y sus mtodos para sanarles. Todas sus re spuestas sern confidenciales. Su participacin es voluntario. No hay riesgos si est de acuerdo en participar. Si usted participa en este estudio, t enemos bolsitas de cosmticos como intercambio por su participacin. Si elige no participar, no habr ningn problema, y si elige terminar la entrevista en cualquie r momento, puede hacerlo y ya recibe su compensacin. La en cuesta no ser muy larga, dura aproximadamente 30 minutos. Si tiene cualquier pregunta sobre la investigacin, voy a estar en esta comunidad en el CEBYCAM por seis meses y puedo dar respuestas a sus preguntas. Puede contac tarme tambin en Riobamba. Mi nmero de telfono es 2968-08 7. Este estudio va a entrevistar aproximadamente a 260 participantes. FECHA: ________ _____________ SUBJECT ID NO. _____________
307 1) Hace cunto tiempo que vive si n interrupcin en este barrio? 1 = de 3 a 5 aos 2 = de 6 a 10 aos 3 = de 11 a 15 aos 4 = ms de 15 aos 2) Qu edad tiene usted? _______ aos 3) Cul es su estado civil? 1 = Soltero/a 2 = Casado/a 3 = Unin Libre 4 = Divorciado/a 5 = Viudo/a 4) Cuntas personas incluyendo Ud. viven en su casa? _______ personas 5) Qu parentesco o relacin tienen la s personas que viven con usted y sus edades? a) b) c) d) e) f) g) 6) Cuntos aos tienen el nio/a menor en su casa? 1 = menos de 6 meses 2 = 6-17 meses 3 = 18-29 meses 4 = 30 meses a 5 aos 7) Cul es el gnero de su nio/a menor? 1 = Hombre 2 = Mujer
308 8) Cuntas personas comparten el dormitorio de su nio/a? 1 = solo l/ella (skip to #9) 2 = 1 persona 3 = 2 personas 4 = 3 o ms personas SI NO TIENE SU PROPIO DORMITORIO: 8A) Los que comparten fuman cigarillos? 1 = S 2 = No 8B) Su nio/a tiene su propia cama? 1 = S 2 = No, comparte con 1 3 = No, comparte con 2 4 = No, comparte con 3 o + 9) Qu utiliza para abrigar su vivienda? 1 = Carbn 2 = Lea 3 = Gas 4 = No abrigamos 10) Qu utiliza para cocinar los alimentos? 1 = Carbn 2 = Lea 3 = Gas 4 = Lea y gas 5 = Gasolina, keroseno 11) Su nio/a duerme en lo mismo cuarto donde cocina? 1 = S 2 = No 12) Tiene mascotas, como gatos, pe rros, o gallinas en su casa? 0 = No (skip to #14) 1 = Gato 2 = Perro 3 = Gatos y perros 4 = Ms de un animal 13) Las mascotas entran a la casa? 1 = S 2 = No
309 Estas preguntas se refieren al pecho de su nio/a menor. Por favor, de respuestas con S o No si es posible. Si una pregunta no aplica a su nio/a, simplemente da respuesta que no aplica. TOS 14) Normalmente, su nio/a menor tiene tos con gripe ? 1 = S___ veces por ao 2 = No 15) Normalmente su nio/a menor tiene tos y no tiene las gripes? 1 = S___ veces por ao 2 = No SI LA RESPUESTA ES S A 14 O 15: 15A) Su nio/a est con tos ms de 4 veces por semana por lo menos 3 meses al ao? 1 = S 2 = No 8 = No aplica 15B) En total, cuntos aos su nio/a ha estado con esta tos? _________________num ero de aos 8 = No aplica CONGESTION Y/O FLEMA 16) Cundo el nio/a est con gripe, elimina la flema, gargajo, o est congestionado el pecho? 1 = S 2 = No (skip to #17) SI LA RESPUESTA ES S A 16: 16A) Si su nio/a est congestionado, el imina gargajo o moco de su pecho 4 veces por semana por lo menos 3 meses al ao? 1 = S 2 = No 8 = No aplica 16B) En total, cuntos aos su ni o/a estaba congestionado o elimina gargajo o moco de su pecho? _________________numer o de aos 8 = No aplica
310 17) Su nio/a tiene ataques de tos, congestin o flema que dura entre una semana o ms cada ao? 1 = S 2 = No (skip to #18) SI LA RESPUESTA ES S A 17: 17A) En total, cuntos aos? ______ numero de aos 8 = No aplica 17B) Generalmente, cuntas gripes o resfriados por ao tiene su nio/a? ______ promedio por ao 8 = No aplica RONCAS O SUENAS 18) El pecho de su nio/a alguna vez ronca o suena? 1 = S 2 = No ( skip to #19 ) 18A) Cundo el/ella est con gripe? 1 = S 2 = No 18B) Ocasionalmente aparte de las gripes? 1 = S 2 = No 18C) La mayora de los das y noches? 1 = S 2 = No SI LA RESPUESTA ES S A 18B O 18C: 18D) En total, cuntos aos el pecho de su nio/a ronca o suena? ______ # aos 8 = No aplica 19) Alguna vez, su nio/a ha teni do ataques de suena el pecho que caus respiracin rapida? 1 = S 2 = No (skip to #20)
311 SI LA RESPUESTA ES S A 19: 19A) Ha tenido 2 o ms de estos episodios por ao? 1 = S 2 = No 19B) Ha recibido medicamientos o tratamientos por estos ataques? 1 = S 2 = No 19C) Cuntos aos tena su nio/a cuando ocurri su primer ataque? ______ numero de aos 8 = No aplica 19D) En el tiempo entre los ataques, la respiracin de su nio/a fue normal? 1 = S 2 = No 8 = No aplica 19E) Si le llev al mdico cul fue el diagnostico? __________________ __________________ _________ 20) Cundo su nio/a est haciendo ejercicio o jugando, alguna vez tiene ataques de suena el pecho? 1 = S 2 = No ENFERMEDADES DEL PECHO 21) Durante los ultimos 3 aos su nio/a ha estado con cualquiera enfermedad de pecho que dur ms que 3 das e impedi las actividades normales? 1 = S 2 = No (skip to #22) SI LA RESPUESTA ES S A 21: 21A) Elimin ms gargajo o flema o pareca ms congestionado, que fuera normal con cualquiera de estas enfermedades? 1 = S 2 = No 8 = No aplica 21B) Cuntas enfermedades como esto s ha tenido en los ultimos 3 aos? 1 = < 1 enfermedad por ao 2 = 1 enfermedad por ao 3 = 2-5 enfermedad por ao 4 = > 5 enfermedad por ao 8 = No aplica
312 21C) Cuntas enfermedades como es tos dur por 7 das o ms? ______ # de enfermedades 8 = No aplica 22) Fue hospitalizado por una enferm edad grave del pecho o gripe antes de los 2 aos? 1 = S, una vez 2 = S, 2 veces 3 = S, >= 3 veces 4 = No 23) Su nio/a tena cualquier otra enfermedad grave del pecho o gripe antes de los 2 aos? 1 = S 2 = No OTRAS ENFERMEDADES 24) Su nio/a tena algun as de las siguientes enfermedades, y s, a qu edad? 24A) Sarampin o Rubeola? 1 = S 2 = No ______ edad 24B) Sinusitis? 1 = S 2 = No ______ edad 24C) Bronquitis? 1 = S 2 = No ______ edad 24D) Neumonia? 1 = S 2 = No ______ edad 24E) Amigdalitis? 1 = S 2 = No ______ edad
313 24F) Bronconeumonia? 1 = S 2 = No ______ edad 25) Alguna vez, un medico dijo que su nio/a tena asma? 1 = S 2 = No (skip to #26) SI LA RESPUESTA ES S A 25: 25A) A qu edad empez su asma? ______ edad 25B) Su nio/a tiene asma ahora? 1 = S 2 = No 25C) Su nio/a toma medicina o tr atamientos para su asma ahora? 1 = S 2 = No SI LA RESPUESTA ES NO A 25 B: 25D) A qu edad se san del asma? ______ edad ALERGIAS 26) Alguna vez, un mdico dijo que su nio/a tiene alergias a la comida o medicina? 1 = S, comida solamente 2 = S, medicamientos solamente 3 = S, comida y medicamientos 4 = S, pero desconoce la causa 5 = No 27) Alguna vez, un mdico dijo que su nio/a tiene alergias al polen o polvo? 1 = S 2 = No 28) Alguna vez, un mdico dijo que su nio/a tiene alergias a detergentes o qumicos? 1 = S 2 = No
314 29) Alguna vez, su nio/a recibi inyecciones para las alergias? 1 = S 2 = No HISTORIA DE LA FAMILIA Nos gustara obtener alguna informacin sobre los padres que viven con el nio/a. 30) Hay alguin en la casa que fuma cigarillos? 1 = S 2 = No si S, especifica ________________________ _____________________ ___________________ 31) La mujer que cuida el nio/a es: 1 = Madre propia 2 = Madrastra 3 = Otra _____________ 32) Cul es su nivel de educacin completada? 1 = Primaria 2 = Secundaria 3 = Superior _________ aos en total 33) Cul es su principal ocupacin? ________________________ _____________________ ___________________ 34) Cunto tiempo ha tenido este trabajo? ________________________ _____________________ ___________________
315 USO DE RECURSOS DE SALUD Cuando su hijo/a o familiar se enfe rma de los pulmones, a donde acude primero: 35) Tos y secrecin nasal 1 = Remedios caceros 2 = Mdico 3 = Nada Porqu? ________________________ _____________________ ___________________ 36) Tos y fiebre 1 = Remedios caceros 2 = Mdico 3 = Nada Porqu? ________________________ _____________________ ___________________ 37) Fiebre y secrecin nasal 1 = Remedios caceros 2 = Mdico 3 = Nada Porqu? ________________________ _____________________ ___________________ 38) Tos, fiebre y secrecin nasal 1 = Remedios caceros 2 = Mdico 3 = Nada Porqu? ________________________ _____________________ ___________________ 39) Tos y gargajo/flema 1 = Remedios caceros 2 = Mdico 3 = Nada Porqu? ________________________ _____________________ ___________________ 40) Tos, gargajo/flema y secrecin nasal 1 = Remedios caceros 2 = Mdico 3 = Nada Porqu? ________________________ _____________________ ___________________ 41) Respiracin rpida 1 = Remedios caceros 2 = Mdico 3 = Nada Porqu? ________________________ _____________________ ___________________ 42) Hundimiento de las costillas 1 = Remedios caceros 2 = Mdico 3 = Nada Porqu? ________________________ _____________________ ___________________
316 43) Respiracin rpida y hundimiento de las costillas 1 = Remedios caceros 2 = Mdico 3 = Nada Porqu? ________________________ _____________________ ___________________ Por favor, digame el nivel de severidad de los siguientes signos: 44) Tos y secrecin nasal 1 = no muy serio 2 = moderado 3 = serio 4 =emergencia 45) Tos y fiebre 1 = no muy serio 2 = moderado 3 = serio 4 =emergencia 46) Fiebre y secrecin nasal 1 = no muy serio 2 = moderado 3 = serio 4 =emergencia 47) Tos, fiebre y secrecin nasal 1 = no muy serio 2 = moderado 3 = serio 4 =emergencia 48) Tos y gargajo/flema 1 = no muy serio 2 = moderado 3 = serio 4 =emergencia 49) Tos, flema y secrecin nasal 1 = no muy serio 2 = moderado 3 = serio 4 =emergencia 50) Respiracin rpida 1 = no muy serio 2 = moderado 3 = serio 4 =emergencia 51) Hundimiento de las costillas 1 = no muy serio 2 = moderado 3 = serio 4 =emergencia 52) Respiracin rpida y 1 = no muy serio 2 = moderado 3 = serio 4 =emergencia hundimiento de las costillas Descripcin de Hundimiento de las Costillas 53) Por favor, vea este foto y explcame los signos que observe: __________________ __________________ ____________________________
317 Por favor, cuntame cuando fue la ultima vez que usted us los siguientes recursos de salud para su nio/a y la razn, y tambin cul es su opinin sobre el recurso 54) CEBYCAM ________________________ _____________________ ___________________ ________________________ _____________________ ___________________ ________________________ _____________________ ___________________ 55) Subcentro de Salud ________________________ _____________________ ___________________ ________________________ _____________________ ___________________ ________________________ _____________________ ___________________ 56) Farmacia ________________________ _____________________ ___________________ ________________________ _____________________ ___________________ ________________________ _____________________ ___________________ 57) Curandero/a ________________________ _____________________ ___________________ ________________________ _____________________ ___________________ 58) Cul es la diferencia entre to s fria y tos caliente y cuales son los tratamientos? ________________________ _____________________ ___________________ ________________________ _____________________ ___________________ 59) Los nios menos de cinco aos son vacunados? 1 = S 2 = No
318 INVENTARIO DE REMEDIOS CACEROS 60) Escriba todos los remedios que ti ene en la casa. Cul es el uso? De donde provienen o adquiren? Cunto cues tan? Cundo los us por ltima vez? Remedio Uso Origen Precio Cundo
319 61) DIETA DE NIOS MENOS DE CINCO AOS (24 HORAS) Lugar Hora Tipo de comida y bebida Marca Cantidad
320 62) Qu comida no le da al nio/a cuando est enfermo? ________________________________________________________________________ 63) Cules son los obstculos para ver un mdico cuando su nio/a est enfermo? ________________________________________________________________________ VIVIENDA Nos gustara obtener alguna informacin ac erca de su tipo de vivienda.. 64) Su casa o tierra es ? 1 = Propia 2 = Arrendada 3 = Familiar 4 = Otro _________ 65) Cuntos cuartos existen en esta casa? 1 = Uno 2 = Dos 3 = Tres 4 = cuatro o ms 66) De donde provienen el agua para beber y cocinar? 1 = Pozo 2 = Entubada 3 = Embotellada 4 = Lluvia 5 = Otro ________ 67) Se hierve el agua para beber? 1 = S ________________ 2 = No
321 CONDICION SOCIOECONOMICA 68) Diga S si tiene cualquie ra de las siguientes cosas: 1 = salario mensual _______ 2 = cualquier herramienta 3 = > 3 cuartos en la casa 4 = piso no hecho de tierra 5 = pared de bloques/ladrillo 6 = dueo de tierras 7 = televisin 8 = bicicleta 9 = > 3 animales criados 69) Cul es la principal fuente de su stento para la familia y de quin viene? ________________________ _____________________ ___________________ 70) Cuntos miembros de la familia son dependientes? __________________ _________________ 71) Recibe dinero de amigos o fa miliares que viven fuera de Penipe? 1 = S ________________ 2 = No Si el participante trabaja en agricultura, se pregunta: 72) Con que frecuencia trabaja us ted en la agricultura? 1 = permanentemente 2 = ocasionalmente 3 = por temporada 73) La tierra donde trabaja es: 1 = propia 2 = arrendada (alquiler) 3 = de un familiar 4 = empleado 5 = al partir Muchas gracias por su tiempo. Esta informacin va a ayudar en la comprensin de las IRAs aqu.
322 APPENDIX C: FOCUS GROUP QUESTIONS
323 Focus Group Questions 1) What are the most serious childhood illnesses around here? 2) What are the household remedies? 3) Do you use curanderos? 4) Do you know the difference between a hot cough and a cold cough? 5) What are the types of treatm ents you use to treat your child? 6) How long do you wait to see the doctor when your child is sick? 7) What are some of the obstacles to seeking care? 8) Do doctors use the same language as you? 9) What are some of t he reasons that you do not take your child to the doctor? 10) What kind of prescriptions does the doctor bring you? 11) Have you heard of pneumonia? 12) Have you heard of an indrawn chest as a sign of illness? 13) Have you heard of rapid br eathing as a sign of illness? 14) Do you have all the informati on you need about childhood illness? 15) Have you had talks about this ki nd of information from knowledgeable people? 16) Whose job is it to give you these talks?
324 APPENDIX D: IN-DEPTH DESCRIPT ION OF ADMINISTRATION OF QUESTIONNAIRE
325 In-depth Description of Admi nistration of Questionnaire I will not explain each one of the 73 questions with the contingent subquestions for the whole questionnaire, but I will explain particular questions that involved some interpretation on my part t hat could contain research bias. In general, the first five pages of the questi onnaire were easiest because they involve mostly yes/no answers. Beginni ng on page five, there are two sections on the use of health resources. These se ctions were the most difficult for the respondent to understand and answer conf idently. The open-ended questions on page six and seven were not difficult for respondents to answer. The last two pages, which included the dietary recall and questions on housing and socioeconomic status were also well understood. Beginning on page one, question number three asks for marriage status. There was some confusion about marr ied versus living together. Most respondents answered living together but when talking about their partner, would use the term, esposo, or husband. There were also cases where the woman would answer single, but still talk ed about the father of their child or children. In these cases I assumed they were never married. I did not probe on this question and wrote down what the re spondent told me. Questions four and five were related because they asked to list the persons in the household and the total number in the household (includi ng the respondent), and occasionally I would have to correct question number four depending on the answers given in number five. Question number six was self -explanatory. If the child had just turned five years old within the month, I would still incl ude them in the study.
326 Otherwise the female caregiver would not qualify for inclusion in the study. There were a few instances where I had to stop the interview at this point because the respondent did not underst and my verbal informed consent presentation and thought that they would qualify fo r the study if they had any young children. Question nine was pot entially confusing, because it appeared that the subject population did not have t he idea of heating their home, but they answered regardless. This question was closely related to question 10 on cooking. Many respondents however diffe rentiated cooking, where they would use gas if they could afford it, and heat ing, which was frequently related to wood fires. The rest of the questions in this section were understood. The next section of the question naire deals with the childs chest. Question 14 was understood, but there was some difficult y in explaining question 15 which asked if the child had coughs apart from colds, since the two were very closely related. Question 15A generally required me to repeat the question at least once because of the length of the question. The same problem applied to question 16A. Under the section on W heezing, if applicable I asked question 18B. I did not have the same difficulty in asking whether the child wheezed apart from colds as I did on the similarl y-phrased question 15. Question 18C asks whether the childs chest wheezed most days or nights, but the respondent invariable responded nights. Under t he section on Chest Illnesses, I did not experience any trouble or confusion in eliciting responses. The next section dealt with childh ood diseases. I had to add a category right from the beginning asking about vari cella, which was a relatively common
327 childhood illness. It is somewhat like m easles, so initially, when I asked about measles, respondents would answer for example, no, but he had varicella. Therefore, I added the ques tion on the questionnaire. Occasionally the respondent would not know what disease their child had had, but once I gave them the list once again, it would jog thei r memories. Respondents with children with allergies were fairly rare, so I did not have any problems with this section. The subsequent section deals with family history. Of all the questions where there was a possibility the responden t was not telling t he truth, question 30 had the highest probability. Frequently, the respondent would answer that their husband or boyfriend smoked, but the person usually answered for example, only outside the house, or on weekends when he is drinking. It is difficult to gauge the accuracy of res ponses to this question and whether the child was exposed to cigarette smoke. Question 32 did not always produce honest answers either, but there was a c heck on this question. First I asked what was the highest grade completed. Ne xt I asked how many years of school completed. For example, if a respondent claimed that they finished secondary school but only attended 5 of the 6 requir ed years necessary to graduate, I would have to change the response that they gav e me from secondary to primary as the highest level of education attained. In the data analysis, I created a new category of primary with some secondary education. The first two question sets under H ealth Resources were the most problematic to explain to respondents. From my general observations, the less well-educated set of my respondents did not understand these question sets very
328 well and tended to give the same respons e to each question. My question was, of the following list of symptoms, which ty pe of provider would you seek first and why. Many answered for exam ple, first I treat at t he home, and if the illness did not improve, I would seek a doctor. The second set of questions asked the respondents to rate the severity of the same symptoms on a Likert scale. By comparing the responses to these two set of questions, it was my aim to determine whether for the symptoms t he respondents rated as serious or emergency in the second set of questi ons, they chose doctor as the first recourse, and likewise for sy mptoms rated as moderate or not serious, chose home remedy or nothing in the first set of questions. Question 53 provided the respondents with a picture of a baby with an indrawn chest (The Wellcome Trus 2001). Most respondents had not seen this symptom before, and if they had seen something like it, they described it as the result of a fall and broken ribs. The next section was a set of open-ended questions on the health centers the respondents had used and visited. For the town of Penipe the respondents generally commented on CEBYCAM and the Heal th Center of Penipe. For the Puela group, the respondents ha d experiences with the Health Center of Penipe and the Subcenter of Health Palictagua. For the Ma tus group, the respondents generally had experiences with the Health C enter of Matus. Question 58 probed the ethnomedical kno wledge of informants on the difference between hot coughs and cold coughs. If the informant was not familiar with the terms, I wrote on the questionnaire, does not know.
329 Under the section, Hous ehold Remedies, I asked the respondent to list all household remedies currently in the home. This generally gener ated a long list of herbal remedies. Occasionally, t he respondent would also have medicines obtained in the pharmacy. I asked res pondents to tell me the use of each medicine, its origin, price, and when it was last used. Question 61 was a 24-hour dietary reca ll. I showed respondents pictures of vegetables and typical meals to help jog their memories. I used a typical bowl and a cup so that the respondents could te ll me portion size. I used the same form to ask them the diet of the child eight days late r. Often I would have to probe to determine the exact contents of the soups and other dishes that they were consuming or to clarify what the def inition of the food item consisted. I asked respondents to tell me where the f ood was eaten, the hour of the day, they type of food or beverage, t he brand name if applicable, and the quantity. After the diet question, I asked two open-ended ques tions. The first question asked if there were any foods not given to sick children. The second question I asked was for the respondent to list the obstacl es to seeing a doctor when their child was sick. These questi ons were well understood. Under the section on Housing, the most confusing question was question 65, on number of rooms in the house. Sometimes people did not know if they should include kitchens or bathrooms as room s. I told them just to count the number of other rooms. For question 66, people invari ably answered that they got their water from the faucet, so this question did not elicit any new information. There was probably recall bias in question 67 because people had trouble
330 remembering how long they usually boil ed water. However, this question was useful because some people replied that they did not boil water at all. The final section of the questionnaire deals with socioeconomic status. These questions were necessary to place people on a socioeconomic scale depending on whether they answered yes or no to a list of questions about possessions and income. For the monthly income, I simply asked what their combined total monthly income was (i f they had a partner), and then the determination was made later whether their income was above or below average. This section also asked respondent s to answer who was the principal breadwinner of the family, how many of the family members were dependents, and whether they received money or assi stance from anyone outside their town. The final section was only asked to respondents who worked in agriculture. There was some confusi on about asking whet her the respondent worked in agricultural permanently, occasi onally, or seasonally. To further gauge economic status for agricultural worker s, the final question asked whether the respondent owned, rented, borrowed, or s hared the agricultural land she or her partner worked on. Recoding of variables At the end of each day, I would keep track of how many questionnaires I had completed and make sure my map was accurate. I also looked over the questionnaires to make sure I did not make any mistakes. I kept a calendar so I knew when I would have to return to do the second dietary recall. It was
331 important not to do the sec ond dietary recall on the same day as the week as the initial questionnaire. Once I completed all the questionnaires, I designed a database in Microsoft Access in order to enter the data. I had to code all the variables for the quantitative data. For two questions, I checked more than one box, so I had to create a new response. For example, in question number 10, which asks what you use to cook food, many respondent s answered both wood and gas, so I created a new response that was not on t he questionnaire, wood and gas. The same applied to pets (question 12) wher e I created a new response, cats and dogs. Question 33 was open-ended, asking for principal occupation, but I modified the responses for ease in data ent ry into three categories, housewife, agriculture and other. The same applied to question 71 which asks whether the respondent received outside aid. If they answered yes, then I coded responses into aid from either CEBYCA M or family. These categories arose from the responses to the questionnaire s and were created because the breadth of responses that was expected when the questions were written did not manifest itself in the data. After the data h ad been successfully entered into MS Access and checked for accuracy, I exported the database into the statistical software program SPSS Release 13.0 for data analysis.
ABOUT THE AUTHOR John Juan Luque attended Wesleyan Univer sity and graduated with a B.A. in Cultural Anthropology and ear ned the distinction of receiving the Mellon Mays Undergraduate Fellowship. His undergraduate thesis was titled The Black Jaguar: Shamanic Practice and Mythical Imagery. He continued to pursue graduate studies in cultur al anthropology, completing a Masters degree at Arizona State University. His M.A. thes is was titled, Power in an Ecuadorian Quichua Foundation: Conservation Planning in Grassroots Development. He furthered his graduate studies in envi ronmental anthropology and completed another Masters degree at University of Wisconsin, Madison. Juan transferred to University of South Florida and fini shed a Masters in Public Health in Epidemiology while pursui ng the doctorate in Applied Anthropology. He has presented his research at numerous prof essional meetings and conferences both nationally and internat ionally. Juan currently resides in Tampa, Florida with his wife Marie and enjoys playing soccer and tennis.
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001787165
007 cr mnu|||uuuuu
008 070507s2006 flu sbm 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001420
Luque, John S.
Breathing easier :
b ethnographic study of acute respiratory infection in children in rural Ecuador
h [electronic resource] /
by John S. Luque.
[Tampa, Fla] :
University of South Florida,
ABSTRACT: This dissertation research utilizes anthropological methods to determine the degree to which the signs and symptoms female caregivers identify as causes of acute respiratory infections (ARIs) in under-fives in rural Ecuador correspond with Western biomedical categories. By employing both a semi-structured medical history questionnaire and more open-ended ethnographic methods such as in-depth informant interviews and focus group research, the researcher identifies the factors which determine timely health care-seeking behaviors of female caregivers in this case study. Economic factors such as the cost of medications and lost work hours were determined to be the primary financial obstacles for timely health care-seeking. Other barriers included limited and inconsistent hours at the health centers and transportation issues. Families of lower socioeconomic status were also more likely to have children suffering particular respiratory ailments. Childhood respiratory illness ^was identified as an outcome of poverty, which had the potential to reproduce itself through the negative effect of illness on household income. However, the research determined that there was an overall lack of recognition of the biomedical signs and symptoms of serious lower respiratory infections regardless of socioeconomic status. The model of ethnomedicine supports the finding that compliance with timely health care-seeking is limited without collaboration between healthcare professionals and communities to work towards beneficial and achievable goals that are joined by a common purpose. By understanding local cultural beliefs towards ARIs, healthcare professionals are in a better position to: (1) assess the accuracy or inaccuracy of ethnomedical beliefs and determine if there is a conflict in symptom recognition and care-seeking behavior with the biomedical model; (2) determine culturally-appropriate interventions or recommendations to address the health problems of the commu nity and identify barriers; and (3) work with existing community resources in order to foster effective health communication. This research finds that public health messages regarding ARIs be informed by ethnomedical knowledge of home treatments and beliefs. Moreover, health centers need to adhere to regular hours of operation and increase staff capacity to better meet the needs of their clients.
Dissertation (Ph.D.)--University of South Florida, 2006.
Includes bibliographical references.
Text (Electronic dissertation) in PDF format.
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
Title from PDF of title page.
Document formatted into pages; contains 331 pages.
Adviser: Linda Whiteford, Ph.D.
x Applied Anthropology
t USF Electronic Theses and Dissertations.