Thesis (M.S.)--University of Michigan, 1955.

Com base em algumas variáveis sócio-antropológicas e três medidas antropométricas - altura, peso e índice de Kaup - de um Estudo realizado no Município de Santo André, elaboramos o presente trabalho. Para este trabalho excluímos crianças púberes, incluindo treze idades: O, 3, 6 e 9 meses, 1 ano, 1a e meio, 2, 3, 4, 5, 6, 7 e 8 anos, de ambos os sexos. As variáveis independentes utilizadas foram: ISSE - índice de situação sócio-econômica, construído com base na média de gasto mensal familiar per capita e segundo as 14 categorias de ocupação do pai e as de instrução. Considerou-se para tal, as interrelações das variáveis com o gasto, que foram tomadas como desvios em relação à média de gasto familiar per capita mensal da amostra e a média para cada casela, medidas como múl tiplos de T. CATANCES - construída a partir da informação sobre a nacionalidade dos ascendentes das crianças até a 3ª geração. Constitui-se de 2 categorias: \"todos ascendentes brasileiros\" e \"pelo menos um estrangeiro\". Grupo Residencial - dividido em três categorias: \"poucos\", \"médio\" e \"grande\", dependendo de quantas pessoas residissem na casa. Dado que já se conhecia a existência de uma relação positiva entre melhores condições de vida e desenvolvimento físico queríamos observar se isso se dava igualmente nos dois sexos e nas várias idades, e assim poder apreender diferenças entre esses grupos uma vez que só havia sido possível estratificar a amostra por idade e sexo. Essas interrelações ajudariam a melhor caracterizar a população estudada. Como esperado, encontramos uma associação positiva entre melhores condições sócio-econômicas, menor tamanho do grupo residencial e as médias das medidas. Porém, também verificamos a existência de relações entre as variáveis independentes: a) que a categoria de ISSE \"alto\" associava-se com a categoria \"pelo menos um estrangeiro\" da variável Catances, o inverso sendo verdadeiro quando a categoria era \"todos brasileiros\"; b) que, medido pelo tamanho médio do grupo residencial, há indicações de uma fecundidade diferencial, sendo esta maior para o grupo \"todos nacionais\" da variável Catances. Nesse sentido chamamos a atenção à grande concentração de migrantes de Minas e Nordeste nessa categoria. As médias das medidas foram testadas pelo método de ordenação de médias de DUNCAN, e também pela técnica de ajuste por mínimos quadrados para análises de variâncias. Verificamos ao analisar as tabelas das medidas, segundo os vários controles, que o índice de Kaup não se apresentava diferente de uma idade para outra, enquanto tal acontecia com altura e peso. Uma série de outros resultados somados a esse, evidenciou que o índice não media, ao menos para crianças em crescimento, aquilo que se propunha medir, ou seja, diferenças de estados nutricionais. Sugere-se então um ajuste de curva a fim de que, baseados nos dados empíricos, se encontre uma dada função para cada idade. Também constatou-se que embora as médias das medidas diferissem segundo as variáveis independentes, isso não acontecia igualmente nas mesmas idades e nos 2 sexos. Ainda, através dos resultados das análises de variância, observou-se que apesar das diferenças estatísticas significantes, as variáveis selecionadas explicavam muito pouco da variação das medidas nas várias idades e sexos. Esses resultados nos levaram a concluir que a amostra estudada, segundo as variáveis independentes deste estudo, difere entre idades em ambos os sexos. Isso levanta uma série de questões sobre quais as variáveis mais adequadas a um estudo desse gênero e sobre a utilização desses dados para construção de curvas de velocidade. Permitimo-nos sugerir que as tabelas por nós apresentadas no anexo 1, se consideradas as duas categorias de Catances, fossem consideradas como dois padrões: Leste-Nordeste e Sul, utilizando a idéia das tábuas de vida padrão. / The present dissertation was based in some of the data collected during the first semester of 1969, in the Município de Santo André, State of São Paulo, Brasil. The original data was partly constituted of several anthropometric measurements in 17 different ages of children from zero month onto twelve years. Some socio-economic information of their families and data concerning the country of birth (or the State of birth, if Brazilian) was also collected for three generations of the children\'s ancestors. The sample was based in the population of these ages living at the Município mentioned. The total sample was of 9,258 children, stratified by age and sex, so that one could estimate the degree of confidence of the actual sample. It is worthwhile to call attention to the criterium adopted to asses age, which was very strict, and not common in studies as such. Zero month children were measured on 2d or 3d day of life. From three to eighteen months, ± 7 days from the day of birth and from two to twelve years old, ± 14 days from the day of birth. For this study we choose only three of the measurements: height, weight and Kaup index and related them by age and sex, with three independent variables as follows: ISSE - index of socio-economic conditions, which was built considering the joint distribution of occupation and education of the father and the amount of the family expenses per capita, per month. CATANCES - meaning \"categories of ancestors\", divided into two categories: \"all ancestors were Brazilians\" and \"at least one was foreigner\". Residencial Group - (household). Depending of the size or the household, there were three different categories: \"small\", \"medium\" and \"large\". Our objectives were to study the relations between the socio-anthropological variables and the measurements selected. At the same time to have a better knowledge of the variability of the population, sampled by age and sex only. for our purposes we dealt only with 13 ages, from zero months to eight years old, excluding in this manner children entering puberty, after the definition of puberty adopted in the study. Amongst the main findings concerning the relationship between the socio-anthropo1ogical variables and the anthropometric ones, was the fact that many combinations of the independent variables showed differences in measurements, as expected. At the same time, we found interrelationships among the independent variables: a) that better conditions of ISSE were related to few people in the household and with the category \"at least one foreigner\" of the variable Catances; the opposite was also true in relation to the category \"all Brazilians\"; b) we found indications of different fertility between the two categories of Catances. The higher were \"all Brazilians\". This category included a high percentage of Brazilians migrants from other States. We did not have sufficient information (if this is possible anyway), to state that the differences in the means of height and weight were due to socio-anthropological conditions, since they could be due to the tendency of more endogamous marriages between the migrants, that generally got into Santo André already married. We also found that the Kaup index did not measured what it is supposed to, at least in growing children, because of the relative variability of weight and height. Thus, it was suggested that a curve should be fitted having per base our empirical data, so that it could have meaning by each age. The means were tested by DUNCAN multiple range tests, and least aquares technique of analysis of variance. By this last test we get to know that despite of significant statistic influences of the independent variables (measured by weight and height) these influences varied depending on age, sex, eatances and the anthropometric variables, showing that the sample varied differently depending on age, sex and the controls used. But however important were the selected variables, there was still 2/3 of variability to be explained by other variables. For this matter, velocity curves based in this data should be taken carefully, once one cannot say children that are 2 year old in the sample, will be one year latter, exactly as the ones that are 3 years in the sample, or that theirs velocity of growth will be the same.

Este estudo é um recorte de uma pesquisa multicêntrica, financiado pela FAPESP, intitulado Deficiência de ferro em crianças de três a doze meses: determinantes biológicos, sociais, e suas implicações para o incentivo ao Aleitamento Materno Exclusivo(AME). A temática deste recorte focalizou o crescimento, morbidade infantil e Aleitamento Materno (AM) e teve como objetivo geral analisar o crescimento, segundo os índices antropométricos e o estado de saúde, a partir da morbidade referida pela mãe, das crianças de 3 a 11 meses e 29 dias de idade, e suas relações com o tipo de AM, atendidas em dois serviços de saúde de cidades do Estado de São Paulo, no período de julho de 2005 a julho de 2006. Trata-se de um estudo transversal. Participaram do estudo 254 crianças de 3 a 11 meses e 29 dias de idade, atendidas em consulta de puericultura previamente agendada, nos serviços de saúde selecionados. Para coleta de dados, elaborou-se um formulário específico a partir do utilizado no estudo multicêntrico. As análises foram realizadas com auxílio do programa Statistical Package for Social Sciences (SPSS, versão 11,5 for Windows). O valor dos índices peso/altura, altura/idade e peso/idade, para cada uma das referências utilizadas, foi calculado em programas disponibilizados nos sites do CDC e OMS, respectivamente. Foram estabelecidas comparações entre o tipo de aleitamento materno e os índices antropométricos (peso/comprimento e comprimento/idade), segundo os referenciais CDC e OMS. Para identificação da associação ou não entre doença, uso de medicamentos antiinfecciosos e causa de internação referida pelas mães e o tipo de AM, foram utilizados os Testes Qui-Quadrado e Exato de Fisher. Os resultados revelaram uma prevalência maior de mães adolescentes na amostra, comparando-se com índices dos dois municípios. A prevalência de AM, no total da amostra, foi de 71,7% e AME de 11,8%. Tanto pelo referencial CDC como pela OMS, as crianças apresentaram maior freqüência de peso elevado para altura, consideradas obesas, dentre as crianças desmamadas, comparando-se com as que estavam em AM. A comparação entre os dois referenciais de crescimento foi realizada pelo teste de concordância estatística Kappa, mostrando maiores discordâncias nas crianças de 3 a 6 meses, principalmente no que se refere ao índice peso por comprimento. Nas crianças de 3 a 6 meses, encontramos resultados estatisticamente significantes, com valor de p=0,048, para associação entre AME e doenças respiratórias, sugerindo efeito protetor do AME para essas doenças. Os resultados encontrados neste estudo revelam achados motivadores para futuras investigações. / This study is part of a larger, FAPESP-funded, multicentric research study: Deficiência de ferro em crianças de três a doze meses: determinantes biológicos, sociais, e suas implicações para o incentivo ao Aleitamento Materno Exclusivo(AME) Iron deficiency in children three to twelve months old: biological and social determinants and their implications on exclusive breast feeding (EBF). This part of the study focused on child growth and morbidity, and breast feeding (BF). The study population consisted of 254 children with ages from three months to eleven months and 29 days, users of two health services in cities in the state of Sao Paulo. The main objective was to analyze the growth of those children according to anthropometric indexes and their health condition based on the morbidity reported by their mothers, and the relationships with the type of breast feeding. This is a cross-sectional study. The studied children were seen in previously scheduled puericulture appointments at the selected health services, from July 2005 to July 2006. Data collection was performed using a specific form, developed from the one used in the multicentric study. The analyses were carried out using the Statistical Package for Social Sciences (SPSS, 11.5 for Windows). The weight/height, height/age, and weight/age index values for each of the references used were calculated using the software available on the CDC and WHO websites, respectively. Comparisons were established between the type of breast feeding and the anthropometric indexes (weight/height and height/age) according to the CDC and WHO references. The Chi- Square and Exact Fisher Tests were used to identify if there was an association or not between the disease, the use of anti-infection drugs, cause of hospitalization reported by the mothers, and the type of BF. Results showed a greater prevalence of teenage mothers in the sample, compared to rates in both municipalities. The BF prevalence in the whole sample was 71.7%, and 11.8% for EBF. According to the CDC reference as well as the WHO, weaned children presented a greater frequency of high weight for their height, and were considered obese compared to those being breastfed. The comparison between both growth references was performed using Kappa statistical concordance test, revealing greater discordances in children with 3 to 6 months of age, mainly regarding weight/height indexes. For children with the age of 3 to 6 months, statistically significant results were found, with p=0.048 for the association between EBF and respiratory diseases. This suggests EBF has a protective effect for these diseases. The results found in this study reveal motivating findings for further studies.

Com base em algumas variáveis sócio-antropológicas e três medidas antropométricas - altura, peso e índice de Kaup - de um Estudo realizado no Município de Santo André, elaboramos o presente trabalho. Para este trabalho excluímos crianças púberes, incluindo treze idades: O, 3, 6 e 9 meses, 1 ano, 1a e meio, 2, 3, 4, 5, 6, 7 e 8 anos, de ambos os sexos. As variáveis independentes utilizadas foram: ISSE - índice de situação sócio-econômica, construído com base na média de gasto mensal familiar per capita e segundo as 14 categorias de ocupação do pai e as de instrução. Considerou-se para tal, as interrelações das variáveis com o gasto, que foram tomadas como desvios em relação à média de gasto familiar per capita mensal da amostra e a média para cada casela, medidas como múl tiplos de T. CATANCES - construída a partir da informação sobre a nacionalidade dos ascendentes das crianças até a 3ª geração. Constitui-se de 2 categorias: \"todos ascendentes brasileiros\" e \"pelo menos um estrangeiro\". Grupo Residencial - dividido em três categorias: \"poucos\", \"médio\" e \"grande\", dependendo de quantas pessoas residissem na casa. Dado que já se conhecia a existência de uma relação positiva entre melhores condições de vida e desenvolvimento físico queríamos observar se isso se dava igualmente nos dois sexos e nas várias idades, e assim poder apreender diferenças entre esses grupos uma vez que só havia sido possível estratificar a amostra por idade e sexo. Essas interrelações ajudariam a melhor caracterizar a população estudada. Como esperado, encontramos uma associação positiva entre melhores condições sócio-econômicas, menor tamanho do grupo residencial e as médias das medidas. Porém, também verificamos a existência de relações entre as variáveis independentes: a) que a categoria de ISSE \"alto\" associava-se com a categoria \"pelo menos um estrangeiro\" da variável Catances, o inverso sendo verdadeiro quando a categoria era \"todos brasileiros\"; b) que, medido pelo tamanho médio do grupo residencial, há indicações de uma fecundidade diferencial, sendo esta maior para o grupo \"todos nacionais\" da variável Catances. Nesse sentido chamamos a atenção à grande concentração de migrantes de Minas e Nordeste nessa categoria. As médias das medidas foram testadas pelo método de ordenação de médias de DUNCAN, e também pela técnica de ajuste por mínimos quadrados para análises de variâncias. Verificamos ao analisar as tabelas das medidas, segundo os vários controles, que o índice de Kaup não se apresentava diferente de uma idade para outra, enquanto tal acontecia com altura e peso. Uma série de outros resultados somados a esse, evidenciou que o índice não media, ao menos para crianças em crescimento, aquilo que se propunha medir, ou seja, diferenças de estados nutricionais. Sugere-se então um ajuste de curva a fim de que, baseados nos dados empíricos, se encontre uma dada função para cada idade. Também constatou-se que embora as médias das medidas diferissem segundo as variáveis independentes, isso não acontecia igualmente nas mesmas idades e nos 2 sexos. Ainda, através dos resultados das análises de variância, observou-se que apesar das diferenças estatísticas significantes, as variáveis selecionadas explicavam muito pouco da variação das medidas nas várias idades e sexos. Esses resultados nos levaram a concluir que a amostra estudada, segundo as variáveis independentes deste estudo, difere entre idades em ambos os sexos. Isso levanta uma série de questões sobre quais as variáveis mais adequadas a um estudo desse gênero e sobre a utilização desses dados para construção de curvas de velocidade. Permitimo-nos sugerir que as tabelas por nós apresentadas no anexo 1, se consideradas as duas categorias de Catances, fossem consideradas como dois padrões: Leste-Nordeste e Sul, utilizando a idéia das tábuas de vida padrão. / The present dissertation was based in some of the data collected during the first semester of 1969, in the Município de Santo André, State of São Paulo, Brasil. The original data was partly constituted of several anthropometric measurements in 17 different ages of children from zero month onto twelve years. Some socio-economic information of their families and data concerning the country of birth (or the State of birth, if Brazilian) was also collected for three generations of the children\'s ancestors. The sample was based in the population of these ages living at the Município mentioned. The total sample was of 9,258 children, stratified by age and sex, so that one could estimate the degree of confidence of the actual sample. It is worthwhile to call attention to the criterium adopted to asses age, which was very strict, and not common in studies as such. Zero month children were measured on 2d or 3d day of life. From three to eighteen months, ± 7 days from the day of birth and from two to twelve years old, ± 14 days from the day of birth. For this study we choose only three of the measurements: height, weight and Kaup index and related them by age and sex, with three independent variables as follows: ISSE - index of socio-economic conditions, which was built considering the joint distribution of occupation and education of the father and the amount of the family expenses per capita, per month. CATANCES - meaning \"categories of ancestors\", divided into two categories: \"all ancestors were Brazilians\" and \"at least one was foreigner\". Residencial Group - (household). Depending of the size or the household, there were three different categories: \"small\", \"medium\" and \"large\". Our objectives were to study the relations between the socio-anthropological variables and the measurements selected. At the same time to have a better knowledge of the variability of the population, sampled by age and sex only. for our purposes we dealt only with 13 ages, from zero months to eight years old, excluding in this manner children entering puberty, after the definition of puberty adopted in the study. Amongst the main findings concerning the relationship between the socio-anthropo1ogical variables and the anthropometric ones, was the fact that many combinations of the independent variables showed differences in measurements, as expected. At the same time, we found interrelationships among the independent variables: a) that better conditions of ISSE were related to few people in the household and with the category \"at least one foreigner\" of the variable Catances; the opposite was also true in relation to the category \"all Brazilians\"; b) we found indications of different fertility between the two categories of Catances. The higher were \"all Brazilians\". This category included a high percentage of Brazilians migrants from other States. We did not have sufficient information (if this is possible anyway), to state that the differences in the means of height and weight were due to socio-anthropological conditions, since they could be due to the tendency of more endogamous marriages between the migrants, that generally got into Santo André already married. We also found that the Kaup index did not measured what it is supposed to, at least in growing children, because of the relative variability of weight and height. Thus, it was suggested that a curve should be fitted having per base our empirical data, so that it could have meaning by each age. The means were tested by DUNCAN multiple range tests, and least aquares technique of analysis of variance. By this last test we get to know that despite of significant statistic influences of the independent variables (measured by weight and height) these influences varied depending on age, sex, eatances and the anthropometric variables, showing that the sample varied differently depending on age, sex and the controls used. But however important were the selected variables, there was still 2/3 of variability to be explained by other variables. For this matter, velocity curves based in this data should be taken carefully, once one cannot say children that are 2 year old in the sample, will be one year latter, exactly as the ones that are 3 years in the sample, or that theirs velocity of growth will be the same.

Este estudo é um recorte de uma pesquisa multicêntrica, financiado pela FAPESP, intitulado Deficiência de ferro em crianças de três a doze meses: determinantes biológicos, sociais, e suas implicações para o incentivo ao Aleitamento Materno Exclusivo(AME). A temática deste recorte focalizou o crescimento, morbidade infantil e Aleitamento Materno (AM) e teve como objetivo geral analisar o crescimento, segundo os índices antropométricos e o estado de saúde, a partir da morbidade referida pela mãe, das crianças de 3 a 11 meses e 29 dias de idade, e suas relações com o tipo de AM, atendidas em dois serviços de saúde de cidades do Estado de São Paulo, no período de julho de 2005 a julho de 2006. Trata-se de um estudo transversal. Participaram do estudo 254 crianças de 3 a 11 meses e 29 dias de idade, atendidas em consulta de puericultura previamente agendada, nos serviços de saúde selecionados. Para coleta de dados, elaborou-se um formulário específico a partir do utilizado no estudo multicêntrico. As análises foram realizadas com auxílio do programa Statistical Package for Social Sciences (SPSS, versão 11,5 for Windows). O valor dos índices peso/altura, altura/idade e peso/idade, para cada uma das referências utilizadas, foi calculado em programas disponibilizados nos sites do CDC e OMS, respectivamente. Foram estabelecidas comparações entre o tipo de aleitamento materno e os índices antropométricos (peso/comprimento e comprimento/idade), segundo os referenciais CDC e OMS. Para identificação da associação ou não entre doença, uso de medicamentos antiinfecciosos e causa de internação referida pelas mães e o tipo de AM, foram utilizados os Testes Qui-Quadrado e Exato de Fisher. Os resultados revelaram uma prevalência maior de mães adolescentes na amostra, comparando-se com índices dos dois municípios. A prevalência de AM, no total da amostra, foi de 71,7% e AME de 11,8%. Tanto pelo referencial CDC como pela OMS, as crianças apresentaram maior freqüência de peso elevado para altura, consideradas obesas, dentre as crianças desmamadas, comparando-se com as que estavam em AM. A comparação entre os dois referenciais de crescimento foi realizada pelo teste de concordância estatística Kappa, mostrando maiores discordâncias nas crianças de 3 a 6 meses, principalmente no que se refere ao índice peso por comprimento. Nas crianças de 3 a 6 meses, encontramos resultados estatisticamente significantes, com valor de p=0,048, para associação entre AME e doenças respiratórias, sugerindo efeito protetor do AME para essas doenças. Os resultados encontrados neste estudo revelam achados motivadores para futuras investigações. / This study is part of a larger, FAPESP-funded, multicentric research study: Deficiência de ferro em crianças de três a doze meses: determinantes biológicos, sociais, e suas implicações para o incentivo ao Aleitamento Materno Exclusivo(AME) Iron deficiency in children three to twelve months old: biological and social determinants and their implications on exclusive breast feeding (EBF). This part of the study focused on child growth and morbidity, and breast feeding (BF). The study population consisted of 254 children with ages from three months to eleven months and 29 days, users of two health services in cities in the state of Sao Paulo. The main objective was to analyze the growth of those children according to anthropometric indexes and their health condition based on the morbidity reported by their mothers, and the relationships with the type of breast feeding. This is a cross-sectional study. The studied children were seen in previously scheduled puericulture appointments at the selected health services, from July 2005 to July 2006. Data collection was performed using a specific form, developed from the one used in the multicentric study. The analyses were carried out using the Statistical Package for Social Sciences (SPSS, 11.5 for Windows). The weight/height, height/age, and weight/age index values for each of the references used were calculated using the software available on the CDC and WHO websites, respectively. Comparisons were established between the type of breast feeding and the anthropometric indexes (weight/height and height/age) according to the CDC and WHO references. The Chi- Square and Exact Fisher Tests were used to identify if there was an association or not between the disease, the use of anti-infection drugs, cause of hospitalization reported by the mothers, and the type of BF. Results showed a greater prevalence of teenage mothers in the sample, compared to rates in both municipalities. The BF prevalence in the whole sample was 71.7%, and 11.8% for EBF. According to the CDC reference as well as the WHO, weaned children presented a greater frequency of high weight for their height, and were considered obese compared to those being breastfed. The comparison between both growth references was performed using Kappa statistical concordance test, revealing greater discordances in children with 3 to 6 months of age, mainly regarding weight/height indexes. For children with the age of 3 to 6 months, statistically significant results were found, with p=0.048 for the association between EBF and respiratory diseases. This suggests EBF has a protective effect for these diseases. The results found in this study reveal motivating findings for further studies.

The first two chapters of this dissertation focus on the child health outcomes in brief, child diarrheal incidence in India, and child growth measures in Rwanda. The third chapter examines the determinants of timely access to health care in emergency departments in the United States. All three essays are different in their area of interest, data sources, and methodology.ESSAY 1 India has recorded the highest number of child diarrheal deaths at the global level. Oral Rehydration Therapy (ORT) would just provide hydration and few mineral supplements for infected children. However, their malnutrition and weakened immune system cannot be reversed. Malnutrition affects child growth, and causes stunting, and makes them susceptible to other forms of infections. The rotavirus vaccine provides a pseudo-sense of protection from non-rotaviral diarrhea. Preventing diarrhea right away from the source of the infection would be a better solution. Since most diarrheal pathogens are water borne, disinfection treatment of drinking water at the point of use could prevent diarrheal incidence of children and adults as well. Household data from the National Family and Health Survey and their estimators viz., Propensity Score Matching (PSM), and Inverse Probability Weights Regression Adjustment (IPWRA) have been employed to examine the effects of water treatment techniques in households. This chapter attempted a novel approach in studying all the popularly used water treatment techniques currently practiced in India in one study. They have been ranked for multi-value treatment effects model. Water filters with ceramic candles are more effective than other point-of-use water treatment techniques, followed by chlorination, water purifiers, and boiling. ESSAY 2 Rwanda is a sub-Sahara African country affected by genocide with a patriarchal family structure system. Higher poverty and gender imbalance were not alleviated by gender equality being on the political agenda. Despite the highest female representation in the parliament in the world, gender equality and liberty are confined to elite women. Additionally, flawed laws for women's equality made women's empowerment a paradoxical phenomenon. Women at the gross root levels, i.e., community and household, are still dependent or interdependent on men (husband/father). This has been confirmed by the findings in Chapter 3, that is, for most decisions, women are taken jointly with their partners. A minimal percentage of women are autonomous in their decisions and in their home. This study made an attempt to examine the mother’s autonomy in Rwanda and its impact on child health. Mother’s autonomy is negatively related to child’s height and weight for their age. Although the expected association between mother’s autonomy and child height/weight is positive, it would also depend on the historical and cultural context of the country of interest. Instrumental variable analysis is used to study women’s autonomy due to its complex and endogenous nature. Spousal educational difference and marriage-to-birth interval are valid instruments but weakly identified. ESSAY 3 An increasing burden on emergency services that exceeded its resources led to congestion in the emergency department (ED), with patients waiting for physicians on the examination bed and for inpatient bed transfer. This creates a blockage between access to healthcare and emergent patients. ED measures adopted to reduce ED congestion, boarding, and waiting times, such as ambulance diversion management, fast tracking of patients with low acuity, and bed coordination do not effectively control waiting time and boarding. ED crowding is a patient flow, but not necessarily a hospital resource deficiency. This is evident from the findings that even EDs with new treatment spaces still keep patients waiting for an inpatient bed, however, they reported a shorter wait time for the physician. Optimal utilization of nurses by floating them to needy units is effective in timely transfers of patients to inpatient beds compared to EDs without floating nurses.

<p>How households and individuals respond to adverse and unanticipated shocks is an important concern for both economists and policy makers. This is especially true in developing countries where poverty, weak infrastructure, and a lack of social safety nets often exacerbate the effects of adverse shocks on household welfare. My research addresses these issues in the context of three economic outcomes and behaviors - early life health and the accumulation of human capital, willingness to take on financial risk, and behavior in the labor market. The results of this research project both adds to our understanding of how life experiences shape individuals' well-being and behavior and how policy can help individuals achieve long-term improvements in the lives following adverse events.</p><p>My research focuses on households and individuals affected by a large-scale natural disaster, the 2004 Indian Ocean tsunami. I utilize data from the Study of the Tsunami Aftermath and Recovery (STAR), a unique longitudinal survey of individuals and households living in coastal communities in Aceh and North Sumatra, Indonesia, at the time of the tsunami. The STAR surveys were conducted annually for five years after the disaster and include a wide range of demographic, economic, and health measures.</p><p>In the first chapter, Child Height after a Natural Disaster, co-authored with Elizabeth Frankenberg, Duncan Thomas, and Jed Friedman, we investigate the immediate and long-run impacts on child health of in utero exposure to stress induced by the tsunami. We investigate whether in utero exposure to stress, as measured by tsunami-induced maternal posttraumatic stress, affected the growth of children born in the aftermath of the tsunami in the critical first five years of their lives. Although previous studies suggest that in utero exposure to stress is related to a number of adverse birth outcomes such as prematurity and lower birth weight, there is little evidence of the impact on linear growth, a strong correlate of later life income. We find evidence that children exposed to high levels of stress beginning in the second trimester experienced reduced growth in the first two years of their lives. We also find evidence that growth reductions largely disappear by age five. This suggests that significant catch-up growth is possible, particularly in the context of pronounced post-disaster reconstruction and economic rehabilitation.</p><p>In the second chapter, The Impact of a Natural Disaster on Observed Risk Aversion, I investigate the short and long-term impacts of the 2004 Indian Ocean tsunami on attitudes toward risk. Attitudes toward risk are important determinants of economic, demographic, and health-related behaviors, but how these attitudes evolve after an event like a natural disaster remains unclear because past research has been confounded by issues of selective exposure, mortality, and migration. My study is the first to directly address these problems by utilizing exogenous variation in exposure to a disruptive event in a sample of individuals that is representative of the population as it existed at the time of the event. In addition, intensive efforts were made to track migrants in the sample population, which is important for this study because migration is common following events like natural disasters and is likely related to attitudes toward risk. I find that physical exposure to the tsunami (e.g., seeing or hearing the tsunami or being caught up in the tsunami) causes significant short-term decreases in observed aversion to risk, especially for the poor, but few longer-term differences. This finding has important implications for the design of effective post-disaster assistance policies. In particular, it implies that post-disaster assistance programs should include aid that is consistent with the observed risk attitudes of the survivors such as job training and capital to start-up businesses.</p><p>In the last chapter, Labor Market Outcomes following the 2004 Indian Ocean Tsunami, I investigate how labor market outcomes changed in coastal communities in Aceh and North Sumatra following the tsunami and the post-disaster recovery efforts. Although restoring the livelihoods of survivors of adverse events is critical for their long-term recovery, there is little evidence from developing countries of how labor market outcomes change after such events. Using the STAR data, I find a significant and persistent increase in paid employment for younger women in urban communities. The increase occurred in communities that were heavily damaged by the tsunami and those that were not, suggesting that the impacts of the disaster on livelihoods are likely long-lasting and extend beyond the communities that were directly stuck by the disaster.</p> / Dissertation

Child growth reference charts have been used since the 1960s to assess children´s growth - enabling comparison of different population groups and the implementation of nutritional surveillance. In 2006, an important critical juncture occurred in the history of anthropometry and nutritional assessment, when the WHO released new growth charts for international comparison after promoting, since 1975, the use of the charts developed by the Centers for Disease Control (CDC) and US National Center for Health Statistics (NCHS). According to the WHO, these charts indicate how children should grow for the best health outcome in contrast to the NCHS/CDC charts that indicated how the average child grows. This shift from a descriptive to a prescriptive -and rather normative - approach allowed the WHO to state that all children in the world have the potential to grow and develop to within the same range of height and weight, thus implying that all children should develop in specific standardised ways, regardless of ethnicity, socioeconomic status and type of feeding. By 2011, approximately 125 countries had adopted the WHO charts for individual growth monitoring as well as the means of producing statistics for under- and over-nutrition, which would be used to assess and monitor a population's health status. This shift between charts has direct implications for how children's growth is measured and how malnutrition is assessed. The adoption of the WHO charts has immediate consequences for the calculation of underweight, overweight, stunting, and wasting prevalence. In this sense, the adoption of the new charts considerably changes the estimates to predict nutrition-related emergencies, the assessment of appropriate weaning practices, and the screening and monitoring of populations at risk or with growth deficiencies or excesses. In my doctoral research, I use Colombia as a case study to unpack how a standard developed by an international organisation is negotiated, adopted and constantly transformed once it is scaled down to a specific country. Using the theoretical approach to standards by authors such as Star, Bowker, Timmermans, Berg, and Epstein, in this dissertation I show how, far from being 'stable' and 'value-free' (as the World Bank would describe them), growth charts are political tools of measurement, charged with specific values regarding children's bodies. Given that Colombia had previously used the NCHS charts, this research explores how the WHO charts have been adopted within individual growth monitoring programmes in Colombia. I also describe how the change in charts has destabilised the production of under and over-nutrition indicators by national bodies, such as the Ministry of Health and the Instituto Nacional de Salud. My data includes twenty-eight interviews with policy makers, experts and civil servants who actively participated in the process of adopting and adapting the standards in Colombia at the national level; seventeen interviews with nurses and doctors; observation of 158 anthropometric assessments of children under five years old within six health facilities in the Caribbean region that were implementing a growth monitoring programme. By exploring how the WHO charts are interpreted and used in practice, this research contributes to the study of standards and standardisation as a field of study in its own right.

Background: Intimate partner violence (IPV) is a pervasive, worldwide public health problem and one of the most common violations of human rights. The aim of this thesis were twofold: (i) to study the process and factors related to ending of IPV of Nicaraguan women and (ii) to examine to what extent IPV exposure is associated with the child linear growth and women’s contraceptive use after pregnancy. Methods: Data were collected from a panel study which followed 398 women who were inquired about their IPV exposure during pregnancy and at follow-up a median of 43 months after delivery. Three hundred seventy five of their children were available for anthropometric assessment. Thirteen in-depth interviews were conducted with women exposed to physical/sexual IPV during pregnancy but not at follow-up. For analysis both quantitative and qualitative methodologies were used. Results: Women experienced four patterns of abuse: never abused, ending abuse, continued abuse, and new abuse. Of the women who experienced any IPV before or during pregnancy, 59% (95% CI 52-65%) reported no abuse at follow-up (135/229).  Women exposed to a continued abuse pattern and those exposed to any IPV, emotional or physical IPV at follow-up had higher odds of reversible contraceptive use. Further, exposure to any IPV and controlling behavior by a partner during pregnancy impaired the index child linear growth. Girls whose mothers had low social resources during pregnancy were the most affected. Women felt that being inquired about IPV while pregnant contributed to process of ending the abuse. Ending IPV was experienced as a process with three phases: “I came to a turning point,” “I changed,” and the “Relationship ended or changed.” Successful strategies to ending abuse mainly involved utilizing informal networks. Ending IPV did not always mean ending the relationship. IPV awareness, severity of the abuse, and economic independence were individual factors associated with ending of abuse. At the relationship level, diminishing or no exposure to controlling behavior by their partner was a key element. At the community level, a supportive and less tolerant to IPV environment as well as exposure to IPV inquiry during pregnancy facilitated the process of ending abuse. Conclusion: The study found that IPV exposure is associated with the children’s linear growth and women’s reversible contraceptive use. In addition, it is clear that gender norms regarding IPV are not static and that they play an important role in facilitating the process ending the abuse by increasing abused women’s access to emotional and material support. Our results emphasize the relevance of improving public services response to IPV.

Nutritional insults and conditions in fetal life and infancy may influence later growth and body composition as well as the development of chronic diseases in adult life. We studied the effects of maternal food and micronutrient supplementation and exclusive breast-feeding counseling on offspring growth 0-54 months and body composition at 54 months of age. We also validated and developed equations for a leg-to-leg bioimpedance analyzer in order to assess body composition of Bangladeshi children aged 4-10 years. In the MINIMat trial in Matlab, Bangladesh, pregnant women were randomized to Early (around 9 weeks) or a Usual invitation (around 20 weeks) to food supplementation and to one of three daily micronutrient supplementations with capsules of either 30 mg Fe and 400 µg folic acid, or 60 mg Fe and 400 µg folic acid, or multiple micronutrient supplements (MMS) (15 micronutrients including 30 mg Fe and 400 µg folic acid). They were also randomized to exclusive breastfeeding counseling (EBC) or to usual health messages (UHM). Growth of their children was measured from birth to 54 months, when body composition also was assessed. There were no differences in background characteristics across the different intervention groups. There was no differential effect by prenatal interventions on birth weight or length. Early invitation to food supplementation reduced stunting from early infancy up to 54 months for boys (average difference 6.5 percent units, 95% CI=1.7 to 11.3, p=0.01), but not for girls (average difference 2.4 percent units, 95% CI=-2.2 to 7.0, p=0.31). MMS resulted in more stunting than standard Fe60F (average difference 4.8 percent units, 95% CI=0.8 to 8.9, p=0.02). Breast-feeding counseling prolonged the duration of exclusive breastfeeding (difference 35.0 days, 95% CI 30.6-39.5, p&lt;0.001). Neither the pregnancy interventions nor the breast-feeding counseling influenced body composition at 54 months. Early food supplementation in pregnancy reduced the occurrence of stunting in boys 0-54 months, while prenatal MMS increased the proportion of stunting. Early food and multiple micronutrient supplementation or exclusive breastfeeding intervention provided to rural Bangladeshi women during pregnancy did not affect offspring body composition at 54 months of age. The effects on postnatal growth suggest programming effects in early fetal life.