In the following chapters, I will discuss the effectiveness of radio and audiotapes as appropriate tools for health communication particularly suited to reaching deprived and isolated communities. I will refer to the striking achievements in radio and audiotape projects by drawing on the experiences of more than sixty radio and cassette projects concerned with primary health care in developing countries. I will present a detailed description of my Argentinean and South African radio experiences, focussing on how the two programmes came into existence, the different stages they have gone through, the problems and constraints encountered as well as their strengths and successes. I will describe the audiotape project I am involved in, and demonstrate how audiotapes can be used alone or in conjunction with radio for primary health education.

Dyslipidemia is increasing among U.S. children, and the prevalence is highest among children with diabetes and obesity. Recently, vitamin D deficiency has been suggested as a possible dietary risk factor for dyslipidemia. Despite the high prevalence of vitamin D deficiency amongst children, virtually no studies have evaluated the association between vitamin D and dyslipidemia among children. We evaluated the vitamin D and dyslipidemia relationship among 240 children and adolescents aged 2 through 21 years who were outpatients of a pediatric endocrinology unit at a large tertiary care facility in Western Massachusetts from April 2008 to April 2010. Eligible children were those with either obesity and/or type 1 or 2 diabetes mellitus. A total of 17.4% of children had severe (<15.0 ng/ml) vitamin D deficiency, 19.2% had moderate (15.0-19.9 ng/ml) deficiency, 36.3% were insufficient (20.0-29.9 ng/ml), and 27.1% had normal (≥30.0 ng/ml) levels. A total of 28.8% of children had high total cholesterol (TC ≥180 mg/dL), 19.6% had high triglycerides (TG; <10years: ≥110 mg/dL, ≥10years: ≥130 mg/dL), 21.3% had low high density lipoprotein (HDL <40 mg/dL), and 6.7% had high low density lipoprotein (LDL ≥130 mg/dL). Moderate vitamin D deficiency was associated with increased risk of high TC (adjusted odds ratio [OR adj] = 2.9, 95% confidence interval (CI): 1.0, 8.8) compared to children with normal vitamin D levels. Severe vitamin D deficiency was associated with an increased risk of low HDL (OR adj = 3.5, 95% CI: 1.0-12.3) and high TG (OR adj = 11.7, 95% CI: 1.9, 70.3) compared to children with normal vitamin D levels. Children with moderate vitamin D deficiency had approximately 3-fold increased risk of high TC compared to children with normal vitamin D levels. In comparison to children with normal vitamin D levels, severe vitamin D deficiency was associated with a strong and significant increased risk of low HDL and high TG; with a significant dose-response relationship. Additionally, in linear regression analyses, we found that an increase in vitamin D deficiency was associated with a significant mean increase in all four measures of dyslipidemia. Vitamin D adequacy may reduce the risk of dyslipidemia in children.

<p>NB: I had another committee member, Dr. Mark Larché; and would like to have his name included in the document.</p> <p>Thank you.</p> / <p><strong>Introduction: </strong>Increasing incidence of chronic immune diseases are mirrored by changing disease risk factors, which include maternal stress during pregnancy. To date, no studies have investigated the impact of prenatal stress challenge (PNS) on the fetal immune system. Fetal liver and bone marrow represent major sources of hematopoietic stem cells (HSC) at mid gestation, which differentiate and mature in the thymus. Disturbance of immune development may cause immune impairment in later life. Further, progesterone is recognized as a critical part of feto-maternal interaction. This study aimed to determine if PNS interferes with normal fetal immune development in mice and the impact of progesterone supplementation on stress effects. <strong>Methods: </strong>DBA/2J-mated BALB/c dams were sorted into three groups: control, PNS (gestation days (GDs) 12.5 and 14.5) and PNS plus progesterone supplementation (DHD). Fetal tissue was collected on GDs 16.5 and 18.5. Flow cytometric analysis examined frequency and phenotype of fetal immune cell populations: HSC in fetal liver and bone marrow, and different stages of T cell maturation and regulatory T (Treg) cells in the thymus. Fetal tails were collected to determine fetal sex by PCR analysis. <strong>Results: </strong>PNS induced a decrease in organ size on GD16.5, which was not seen on GD18.5 and was reversed by DHD treatment. PNS altered the percentage and absolute number of HSC within the liver and bone marrow populations, on GD16.5 and 18.5. There was a significant lag in T cell maturation as demonstrated by the altered expression of CD3 and skewed CD3-:CD3+ ratio. There was a significant decrease in Treg cells within CD3+ thymic cells in response to PNS. PNS effects in the thymus were ameliorated by DHD treatment. There was no PNS-induced sex bias. <strong>Conclusions: </strong>These results indicate that PNS compromises the developing fetal immune system, which could account for impaired immune responses in adults with chronic immune disease, and provide evidence for a therapeutic role of progesterone supplementation.</p> / Master of Science (MSc)

<p>Malnutrition and Anemia in children continue to be major public health challenges in most developing countries, particularly in Africa. Malnutrition and Anemia pervade all aspects of their health, growth, cognitive and social development. They lead to irreversible and lifelong effects that prevent children from realising their full potential. This study was designed to examine the prevalence and determinants of malnutrition and anemia in children under 5 years of age in the Ghanaian population. This research is based on data from the Ghana Demographic and Health Survey (GDHS) 2008, obtained from the Ghana Statistical Service (GSS). The survey is an extensive survey conducted using a stratified, two-stage cluster sampling design. The GDHS data contains a wealth of information on health, demographic, as well as socio-economic factors but is underutilised due to the complexity of the survey data. This study therefore stands out as one of the few that use the GDHS to investigate aspects of child health in Ghana. In this study, we perform subgroup analysis by disaggregating the data by age and gender specific subgroups and then by place of residence and region. This was in order to identify sub level estimates as national estimates have a high tendency of concealing true values and deviations from general trends. Also, subgroup analysis is very significant especially for resource allocation so as to minimize the likelihood of missing the target populations. We investigated associations between the three measurements of malnutrition; stunting, underweight and wasting and anemia (assessed by haemoglobin concentration) and the various risk factors using chi-square test to examine bivariate associations and chi-square trend test to examine linear trends in association. We identified the following variables to be significantly associated with all forms of malnutrition and/or anemia: age of child, mother’s education, financial status and place of residence. Other factors that were identified to be associated with some form of malnutrition and/or anemia include duration of breastfeeding, source of drinking water, mother’s occupation and currently breastfeeding. In view of the high rate of malnutrition, approximately 36% (33.6−37.6) and the alarming prevalence of anemia, 78% (76.7 − 80.2) in children in Ghana, particularly among those less than 2 years old, and the grave consequences on their cognitive and behavioral development even in later years, there is an urgent need for effective and efficient public health interventions.</p> / Master of Science (MSc)

Background: Gestational diabetes mellitus (GDM) is a common pregnancy-related metabolic disorder associated with adverse maternal and fetal outcomes. Recent research has highlighted the potential role of the oral microbiome in the pathophysiology of various systemic conditions, including diabetes mellitus. However, the impact of GDM on the maternal oral microbiome remains relatively understudied. Understanding alterations in the oral microbiome during pregnancy complicated by GDM could provide valuable insights into the mechanistic links between systemic metabolic disorders and oral health. Objective: The objective of this scoping review was to comprehensively examine the existing literature on the relationship between GDM and maternal oral microbiome composition and diversity. The review sought to identify the microbial changes associated with GDM and explore their potential implications for maternal oral health and pregnancy outcomes. Design: Following PRISMA-ScR guidelines, a systematic search of electronic databases, including “Pubmed” and “Web of Science,” was conducted to identify relevant studies investigating the impact of GDM on the maternal oral microbiome. Eligible studies included those examining microbial diversity, abundance, and composition in pregnant women with GDM and their neonates. Data synthesis involved summarizing key findings and identifying patterns across studies. Results: A total of eight primary studies were identified and included in the scoping review. These studies employed various methodologies, including observational reports, longitudinal analyses, and cohort studies, to investigate the relationship between GDM and the maternal oral microbiome. Findings from these studies revealed distinct alterations in oral microbial profiles among neonates and pregnant women with GDM compared to healthy controls. These alterations encompassed shifts in microbial diversity, abundance, and composition, suggesting potential biomarkers or indicators of GDM status within the oral microbiome. Longitudinal analyses further elucidated dynamic changes in the oral microbiota throughout pregnancy and postpartum, underscoring the temporal nature of these associations. Conclusions: The synthesis of evidence from the scoping review highlights significant associations between GDM and alterations in the maternal oral microbiome. While these findings imply a correlation between GDM and oral microbial changes, causality cannot be directly inferred. Further research is warranted to decipher the underlying mechanisms driving these associations and to explore their potential implications for maternal oral health and pregnancy outcomes. Nonetheless, the insights gleaned from this review underscore the importance of integrating oral health assessments into routine prenatal care protocols for gestational diabetic individuals. By doing so, healthcare providers can enhance risk stratification, early detection, and management of GDM, ultimately improving maternal and neonatal health outcomes.

OBJECTIVE/ METHOD The study examined implementation challenges faced by primary health care workers within the Ekurhuleni Metropolitan Municipality in Gauteng South Africa. Data collection was based on semi-structured interviews carried out on a purposive sample (n=19) of frontline clinicians working within the district as primary health care practitioners. RESULTS Participants confirmed that work within the primary health care service disproportionately focussed on curative and rehabilitative functions of their roles with little prioritisation of preventive and promotive interventions. Primary identified reasons included, institutional culture that prioritised short-term curative approaches. Clinicians also cited a range of other organisational barriers, such as – poor strategic planning, and a lack of understanding of health promotion and illness prevention. CONCLUSIONS Although the challenges that exist in implementing primary health care are clearly understood, clinicians perceive the solutions for these as being within the control of policy makers and those with power within the organisation. / Health Studies / M.A. (Public Health)

La vie des femmes du continent africain et de leurs enfants continue d’être mise en danger lors de chaque accouchement car les risques de décès maternels et infantiles sont encore très élevés. Il est estimé chaque année à environ le quart du million le nombre de décès maternel et de près de quatre millions celui des enfants de moins de cinq ans. La comparaison de la situation sanitaire avec d’autres contextes permet de mieux cerner l’ampleur du problème : en Afrique sub-Saharienne, le risque de décès lié à la grossesse est de l’ordre de 1 pour 31, alors qu’il n’est que de 1 pour 4300 dans les pays industrialisés. Cette situation est évitable et, le plus souvent, résulte de la sous ou non-utilisation des services de santé maternelle, du manque de structures adéquates de soins ou de personnel de santé qualifié. Notre thèse cherche à comprendre la manière dont les inégalités de genre au sein du ménage et dans la communauté renforcent les inégalités quant à l’utilisation des services de santé maternelle, ainsi qu’aux relations empiriques qui lient les différents recours aux soins. Concrètement, elle vise à 1) proposer une mesure des normes de genre favorables à la violence contre les femmes et à analyser son influence sur leur prise de décision au sein du ménage, 2) analyser simultanément l’influence de ces normes et de l’autonomie des femmes sur le recours aux soins prénatals et à l’accouchement assisté et finalement, 3) cerner l’influence des soins prénatals sur le recours à l’accouchement assisté. Chacun de ces objectifs se heurte à un problème méthodologique substantiel, soit de mesure ou de biais de sélection, auxquels l’approche par modèles d’équations structurelles que nous avons adoptée permet de remédier. Les résultats de nos analyses, présentés sous forme d’articles scientifiques, s’appuient sur les données issues des Enquêtes Démographiques et de Santé (EDS) du Ghana, du Kenya, de l’Ouganda et de la Tanzanie et concernent les femmes vivant en milieu rural. Notre premier article propose une mesure des normes de genre et, plus exactement, celles liées à la violence contre les femmes en recourant à l’approche des variables latentes. Les cinq questions des EDS relatives à l’attitude des femmes sur la légitimation de la violence ont permis de saisir cette mesure au niveau contextuel. Les résultats suggèrent d’une part que cette mesure a de bons critères de validité puisque l’Alpha de Cronbach varie de 0.85 pour le Kenya à 0.94 pour le Ghana; les chi-deux sont non significatifs partout; le RMSEA est en dessous de 0.05; le CFI supérieur à 0.96 et les saturations sont pour la plupart supérieures à 0.7 dans tous les pays. D’autre part, à l’aide du modèle d’équations structurelles multiniveaux, nous avons trouvé qu’au-delà de leur propre attitude envers la violence contre les femmes, celles qui vivent dans un milieu où les normes de genres sont plus favorables à la violence ont plus de chances d’être de faible autonomie ou sans autonomie (comparativement à forte autonomie) dans l’ensemble des pays étudiés. Le second article documente l’influence des inégalités de genre, cernées au niveau contextuel par les normes favorables à la violence contre les femmes et au niveau individuel par l’autonomie de prise de décision au sein du ménage, sur la survenue des soins prénatals au cours du premier trimestre et sur les recours à au moins 4 consultations prénatales et à l’accouchement assisté. En utilisant également les modèles d’équations structurelles multiniveaux sur les mêmes données du premier article, nous constatons que chacune de ces variables dépendantes est fortement influencée par la grappe dans laquelle la femme vit. En d’autres mots, son lieu de résidence détermine le comportement de santé maternelle que l’on adopte. De même, en contrôlant pour les autres variables explicatives, nos résultats montrent que les femmes qui vivent dans un milieu où les normes de genre liées à la violence contre les femmes sont élevées ont, en moyenne, une plus grande chance de ne pas accoucher auprès d’un personnel qualifié au Ghana et en Ouganda, de ne pas débuter leurs soins prénatals dans le premier trimestre dans les mêmes pays, et de ne pas recourir à au moins quatre consultations prénatales en Tanzanie. Par contre, cette variable contextuelle n’influence pas significativement le recours aux soins de santé maternelle au Kenya. Enfin, les résultats montrent que les normes de genre favorables à la violence contre les femmes sont plus déterminantes pour comprendre le recours aux soins de santé maternelle dans les pays étudiés que l’autonomie de prise de décision de la femme. Dans le cadre du troisième et dernier article empirique de la thèse, nous nous sommes intéressés à l’importance des soins prénatals dans le processus de recours à l’accouchement assisté et à la place du contenu des soins reçus avant l’accouchement dans cette relation. Cet article met en exergue l’existence de biais d’endogénéité au Kenya et en Tanzanie, où sans sa prise en compte, l’effet des soins prénatals sur le recours à l’accouchement auprès d’un personnel qualifié serait fortement biaisé. De plus, il ressort qu’à l’exception du Ghana et dans une moindre mesure de la Tanzanie, cet effet est totalement médiatisé par le contenu des soins prénatals que les femmes reçoivent. L’article met ainsi en relief le rôle des prestataires de soins qui pour atteindre plus efficacement les populations doivent agir en tant que leaders au sein de leur communauté. / In Africa, the lives of women and their offspring continue to be threatened at every birth because of the underuse of maternal health care. It is estimated that every year about a quarter of a million the number of maternal deaths and nearly four million children die before age five. Comparing the health patterns in sub-Saharan Africa with other contexts helps to better understand the reality of sub-Saharan Africa, where the risk of death in pregnancy is of the order of 1 to 31, while it is only 1 per 4300 in industrialized countries. This situation is preventable and most often is the result of under or non-use of maternal health services, lack of adequate health care facilities or skilled health personnel. This thesis seeks to understand how gender inequality within the household and in the community reinforces inequalities in the use of maternal health services, as well as the empirical relationship linking the various types of maternal health care uptakes. Specifically, it aims to 1) provide a measure of gender norms favorable to violence against women and to analyze its influence on women’s decision-making authority within the household, 2) simultaneously analyze the impact of these gender norms along with the women’s decision-making authority on the use of antenatal care and delivery assistance, and finally, 3) determine the influence of antenatal care on the use of assisted delivery. Each of these objectives faces substantial methodological issues, either measurement or selection bias, which the structural equation modeling approach we adopted overcomes. The results of our analysis, in the form of scientific articles, are based on the data from Demographic and Health Surveys (DHS) of Ghana, Kenya, Uganda and Tanzania and are related to women living in rural areas. Our first article provides a measure of gender norms and, more specifically, those related to violence against women by using the approach of latent variables. The five questions of the DHS on women's attitudes about the legitimacy of violence allowed measuring these gender norms at contextual level. The results suggest that this measure has good validity criteria as the Cronbach's alpha ranged from 0.85 to 0.94 for Kenya to Ghana, the chi-square is insignificant everywhere, the RMSEA is below 0.05 and the CFI above 0.96 and saturation are mostly higher than 0.7 in all countries. On the other hand, using multilevel structural equation modeling, we found that beyond their own attitude toward violence against women, women who live in an environment where gender norms are more favorable to violence are more likely to be of low or no decision-making autonomy (compared to high decision-making autonomy) in all the countries studied. The second Article documents the influence of the gender inequality identified at the contextual level by the gender norms favorable to violence against women and at the individual level by the women’s autonomy of decision-making within the household, on the occurrence of antenatal care during the first quarter, and the use of at least four antenatal care and assisted delivery with skilled professional. Also using multilevel structural equation modeling on the same data of article 1, we find that each of these dependent variables is strongly influenced by the cluster in which the woman lives. In other words, her place of residence determines the maternal health behavior she adopts. Similarly, controlling for other variables, our results show that women who live in an environment where gender norms related to violence against women are higher, have on average, a greater chance of not giving birth with a qualified staff in Ghana and Uganda, of not starting their antenatal care in the first trimester in the same countries, and of not reaching at least four antenatal care visits in Tanzania. By cons, this contextual variable does not significantly affect the use of maternal health care in Kenya. Finally, the results show that norms in favor of gender violence against women are more critical to understand the use of maternal health care in the countries studied than the women’s decision-making autonomy. Chapter VII is the last empirical paper of the thesis and examines the effects of prenatal care on the use of skilled birth attendance and also investigates the role of the content of prenatal care in this relationship. Previous studies on the topic have failed to control for possible endogeneity biases, limiting the validity of their conclusions. Our findings point to the existence of important biases in Kenya and Tanzania, where the estimated effect of prenatal care on the use of skilled birth attendance is highly biased downwards when endogeneity is not taken into account in the statistical model. Furthermore, with the exception of Ghana and, to a lesser extent, of Tanzania, our findings indicate that the beneficial effects of prenatal health care are completely mediated by the content of services that women receive. This study emphasizes the role of healthcare providers who, to effectively reach populations, must also act as leaders in their community.

La vie des femmes du continent africain et de leurs enfants continue d’être mise en danger lors de chaque accouchement car les risques de décès maternels et infantiles sont encore très élevés. Il est estimé chaque année à environ le quart du million le nombre de décès maternel et de près de quatre millions celui des enfants de moins de cinq ans. La comparaison de la situation sanitaire avec d’autres contextes permet de mieux cerner l’ampleur du problème : en Afrique sub-Saharienne, le risque de décès lié à la grossesse est de l’ordre de 1 pour 31, alors qu’il n’est que de 1 pour 4300 dans les pays industrialisés. Cette situation est évitable et, le plus souvent, résulte de la sous ou non-utilisation des services de santé maternelle, du manque de structures adéquates de soins ou de personnel de santé qualifié. Notre thèse cherche à comprendre la manière dont les inégalités de genre au sein du ménage et dans la communauté renforcent les inégalités quant à l’utilisation des services de santé maternelle, ainsi qu’aux relations empiriques qui lient les différents recours aux soins. Concrètement, elle vise à 1) proposer une mesure des normes de genre favorables à la violence contre les femmes et à analyser son influence sur leur prise de décision au sein du ménage, 2) analyser simultanément l’influence de ces normes et de l’autonomie des femmes sur le recours aux soins prénatals et à l’accouchement assisté et finalement, 3) cerner l’influence des soins prénatals sur le recours à l’accouchement assisté. Chacun de ces objectifs se heurte à un problème méthodologique substantiel, soit de mesure ou de biais de sélection, auxquels l’approche par modèles d’équations structurelles que nous avons adoptée permet de remédier. Les résultats de nos analyses, présentés sous forme d’articles scientifiques, s’appuient sur les données issues des Enquêtes Démographiques et de Santé (EDS) du Ghana, du Kenya, de l’Ouganda et de la Tanzanie et concernent les femmes vivant en milieu rural. Notre premier article propose une mesure des normes de genre et, plus exactement, celles liées à la violence contre les femmes en recourant à l’approche des variables latentes. Les cinq questions des EDS relatives à l’attitude des femmes sur la légitimation de la violence ont permis de saisir cette mesure au niveau contextuel. Les résultats suggèrent d’une part que cette mesure a de bons critères de validité puisque l’Alpha de Cronbach varie de 0.85 pour le Kenya à 0.94 pour le Ghana; les chi-deux sont non significatifs partout; le RMSEA est en dessous de 0.05; le CFI supérieur à 0.96 et les saturations sont pour la plupart supérieures à 0.7 dans tous les pays. D’autre part, à l’aide du modèle d’équations structurelles multiniveaux, nous avons trouvé qu’au-delà de leur propre attitude envers la violence contre les femmes, celles qui vivent dans un milieu où les normes de genres sont plus favorables à la violence ont plus de chances d’être de faible autonomie ou sans autonomie (comparativement à forte autonomie) dans l’ensemble des pays étudiés. Le second article documente l’influence des inégalités de genre, cernées au niveau contextuel par les normes favorables à la violence contre les femmes et au niveau individuel par l’autonomie de prise de décision au sein du ménage, sur la survenue des soins prénatals au cours du premier trimestre et sur les recours à au moins 4 consultations prénatales et à l’accouchement assisté. En utilisant également les modèles d’équations structurelles multiniveaux sur les mêmes données du premier article, nous constatons que chacune de ces variables dépendantes est fortement influencée par la grappe dans laquelle la femme vit. En d’autres mots, son lieu de résidence détermine le comportement de santé maternelle que l’on adopte. De même, en contrôlant pour les autres variables explicatives, nos résultats montrent que les femmes qui vivent dans un milieu où les normes de genre liées à la violence contre les femmes sont élevées ont, en moyenne, une plus grande chance de ne pas accoucher auprès d’un personnel qualifié au Ghana et en Ouganda, de ne pas débuter leurs soins prénatals dans le premier trimestre dans les mêmes pays, et de ne pas recourir à au moins quatre consultations prénatales en Tanzanie. Par contre, cette variable contextuelle n’influence pas significativement le recours aux soins de santé maternelle au Kenya. Enfin, les résultats montrent que les normes de genre favorables à la violence contre les femmes sont plus déterminantes pour comprendre le recours aux soins de santé maternelle dans les pays étudiés que l’autonomie de prise de décision de la femme. Dans le cadre du troisième et dernier article empirique de la thèse, nous nous sommes intéressés à l’importance des soins prénatals dans le processus de recours à l’accouchement assisté et à la place du contenu des soins reçus avant l’accouchement dans cette relation. Cet article met en exergue l’existence de biais d’endogénéité au Kenya et en Tanzanie, où sans sa prise en compte, l’effet des soins prénatals sur le recours à l’accouchement auprès d’un personnel qualifié serait fortement biaisé. De plus, il ressort qu’à l’exception du Ghana et dans une moindre mesure de la Tanzanie, cet effet est totalement médiatisé par le contenu des soins prénatals que les femmes reçoivent. L’article met ainsi en relief le rôle des prestataires de soins qui pour atteindre plus efficacement les populations doivent agir en tant que leaders au sein de leur communauté. / In Africa, the lives of women and their offspring continue to be threatened at every birth because of the underuse of maternal health care. It is estimated that every year about a quarter of a million the number of maternal deaths and nearly four million children die before age five. Comparing the health patterns in sub-Saharan Africa with other contexts helps to better understand the reality of sub-Saharan Africa, where the risk of death in pregnancy is of the order of 1 to 31, while it is only 1 per 4300 in industrialized countries. This situation is preventable and most often is the result of under or non-use of maternal health services, lack of adequate health care facilities or skilled health personnel. This thesis seeks to understand how gender inequality within the household and in the community reinforces inequalities in the use of maternal health services, as well as the empirical relationship linking the various types of maternal health care uptakes. Specifically, it aims to 1) provide a measure of gender norms favorable to violence against women and to analyze its influence on women’s decision-making authority within the household, 2) simultaneously analyze the impact of these gender norms along with the women’s decision-making authority on the use of antenatal care and delivery assistance, and finally, 3) determine the influence of antenatal care on the use of assisted delivery. Each of these objectives faces substantial methodological issues, either measurement or selection bias, which the structural equation modeling approach we adopted overcomes. The results of our analysis, in the form of scientific articles, are based on the data from Demographic and Health Surveys (DHS) of Ghana, Kenya, Uganda and Tanzania and are related to women living in rural areas. Our first article provides a measure of gender norms and, more specifically, those related to violence against women by using the approach of latent variables. The five questions of the DHS on women's attitudes about the legitimacy of violence allowed measuring these gender norms at contextual level. The results suggest that this measure has good validity criteria as the Cronbach's alpha ranged from 0.85 to 0.94 for Kenya to Ghana, the chi-square is insignificant everywhere, the RMSEA is below 0.05 and the CFI above 0.96 and saturation are mostly higher than 0.7 in all countries. On the other hand, using multilevel structural equation modeling, we found that beyond their own attitude toward violence against women, women who live in an environment where gender norms are more favorable to violence are more likely to be of low or no decision-making autonomy (compared to high decision-making autonomy) in all the countries studied. The second Article documents the influence of the gender inequality identified at the contextual level by the gender norms favorable to violence against women and at the individual level by the women’s autonomy of decision-making within the household, on the occurrence of antenatal care during the first quarter, and the use of at least four antenatal care and assisted delivery with skilled professional. Also using multilevel structural equation modeling on the same data of article 1, we find that each of these dependent variables is strongly influenced by the cluster in which the woman lives. In other words, her place of residence determines the maternal health behavior she adopts. Similarly, controlling for other variables, our results show that women who live in an environment where gender norms related to violence against women are higher, have on average, a greater chance of not giving birth with a qualified staff in Ghana and Uganda, of not starting their antenatal care in the first trimester in the same countries, and of not reaching at least four antenatal care visits in Tanzania. By cons, this contextual variable does not significantly affect the use of maternal health care in Kenya. Finally, the results show that norms in favor of gender violence against women are more critical to understand the use of maternal health care in the countries studied than the women’s decision-making autonomy. Chapter VII is the last empirical paper of the thesis and examines the effects of prenatal care on the use of skilled birth attendance and also investigates the role of the content of prenatal care in this relationship. Previous studies on the topic have failed to control for possible endogeneity biases, limiting the validity of their conclusions. Our findings point to the existence of important biases in Kenya and Tanzania, where the estimated effect of prenatal care on the use of skilled birth attendance is highly biased downwards when endogeneity is not taken into account in the statistical model. Furthermore, with the exception of Ghana and, to a lesser extent, of Tanzania, our findings indicate that the beneficial effects of prenatal health care are completely mediated by the content of services that women receive. This study emphasizes the role of healthcare providers who, to effectively reach populations, must also act as leaders in their community.

Dans cette thèse, nous décrivons les résultats d’un projet de recherche visant à mesurer et évaluer la qualité des soins obstétricaux des hôpitaux de référence au Mali et au Sénégal. Dans ces pays, la mortalité maternelle hospitalière est élevée et est liée en partie à la pratique médicale inadéquate. Cette recherche a été réalisée dans le cadre de l’étude QUARITE, un essai randomisé en grappe évaluant l’efficacité du programme GESTA International visant à réduire la mortalité maternelle hospitalière. GESTA a été mis en œuvre entre 2008 et 2010 et consistait en la formation des professionnels de santé et en la revue des cas de décès maternels. En parallèle de QUARITE, les programmes de prévention de la transmission du VIH de la mère à l’enfant (PTME) ont été mis à l’échelle à travers les pays. Ces derniers ayant également la capacité d’augmenter la qualité des soins obstétricaux, nous avons donc évalué les effets des deux programmes (GESTA et PTME) sur la qualité des soins. Dans un premier temps, à l’aide d’une recension des écrits nous avons évalué la capacité d’un audit clinique basé sur des critères à mesurer la qualité des soins obstétricaux. Cet audit vérifiait si l’offre des soins avait respecté les critères cliniques définissant la meilleure prise en charge selon l’évidence scientifique et l’avis des experts. Nous avons démontré que cet outil est largement utilisé dans les pays à faibles et moyens revenus, malgré le peu d’évidence sur sa validité (article 1). Dans un deuxième temps, nous avons développé un audit clinique basé sur des critères qui s’applique au contexte ouest-africain et qui a été approuvé par des experts-obstétriciens nationaux et internationaux. À partir des dossiers obstétricaux, les actes médicaux posés pendant le travail et l’accouchement ont été évalués à l‘aide de cet instrument. La qualité des soins a été estimée sous forme de pourcentage de critères atteints. Appliqué dans différents contextes et par différents auditeurs, nous avons démontré que notre instrument est fiable et valide (article 3). Néanmoins, l’expérience de l’audit nous a amenés à nous questionner sur le mauvais remplissage des dossiers médicaux et ses conséquences sur la qualité des soins (article 2). Dans un troisième temps, l’outil a été appliqué à large échelle pour évaluer les effets de l’intervention GESTA (article 4). Nous avons mené une révision de plus de 800 dossiers obstétricaux dans 32 hôpitaux de référence (16 bénéficiaires de l’intervention et 16 non-bénéficiaires). Grâce à cet audit clinique, nous avons démontré que le programme GESTA contribue à l’amélioration de la qualité des soins, spécifiquement l’examen clinique lors de l’admission et le suivi après l’accouchement. Dernièrement, nous avons utilisé cet instrument afin d’évaluer les effets des programmes de PTME sur la qualité des soins obstétricaux (article 5). Notre travail a documenté que seulement certaines composantes du programme de PTME améliorent la qualité des soins telles que la formation des professionnels et les services complémentaires en nutrition. En conclusion, cette recherche a identifié plusieurs pistes d’intervention pour améliorer la qualité des soins obstétricaux en Afrique de l’Ouest. / In this thesis, we describe the results of a research project that aimed to measure and evaluate quality of care in referral hospitals in Mali and Senegal. In these countries, hospital maternal mortality is high and linked, in part, to inadequate medical practice. This research was conducted as part of the QUARITE cluster randomized trial that assessed whether the program, ALARM International, could reduce facility maternal mortality. ALARM was implemented from 2008 to 2010 and consisted of the training of local health professionals and the use of maternal death reviews. At the same time as QUARITE was ongoing, programs for the prevention of maternal to child transmission of HIV (PMTCT) were scaled- up; these can also improve obstetrical quality of care. Thus, we evaluated the effects of both programs (ALARM and PMTCT) on quality of care. We began with a systematic review of the literature to evaluate the capacity of a criterion-based clinical audit to measure the quality of obstetrical care (article 1). This type of audit verifies if the care provided meets criteria indicative of best clinical practices, according to the literature and expert opinion. Our review demonstrates that this tool has been used in a variety of low- and middle-income settings, but the way it has previously been employed leaves doubts as to its validity (article 1). We thus developed a criterion based clinical audit specific to the West African context and approved by national and international expert obstetricians. Using patient medical records, with this instrument we evaluated obstetrical care provided during labour and delivery. Quality of care was calculated based on the percentage of care criteria met. Applied to different sites and by different auditors, our instrument demonstrated concordant results and provided a valid image of the quality of obstetrical care provided at hospitals in the region (article 3). Nonetheless, the audit experience raised concerns about the implications of poor medical recordkeeping and archiving on quality of care (article 2). We used the criterion-based clinical audit to review over 800 medical records at 32 QUARITE hospitals (16 intervention and 16 control hospitals) in order to evaluate the effects of the ALARM intervention. We demonstrated that the ALARM program contributes to better obstetrical quality of care, especially during the first clinical examination and postpartum monitoring of women treated at intervention hospitals (article 4). Finally, we used this instrument to evaluate the effects of PMTCT programs on obstetrical quality of care (article 5). Our work demonstrated that certain components of a PMTCT program, specifically training of healthcare professionals and supplementary nutritional services, are associated with better obstetrical care. In all, this research identified several mechanisms that can be targeted by quality improvement interventions in West Africa.

OBJECTIVE/ METHOD The study examined implementation challenges faced by primary health care workers within the Ekurhuleni Metropolitan Municipality in Gauteng South Africa. Data collection was based on semi-structured interviews carried out on a purposive sample (n=19) of frontline clinicians working within the district as primary health care practitioners. RESULTS Participants confirmed that work within the primary health care service disproportionately focussed on curative and rehabilitative functions of their roles with little prioritisation of preventive and promotive interventions. Primary identified reasons included, institutional culture that prioritised short-term curative approaches. Clinicians also cited a range of other organisational barriers, such as – poor strategic planning, and a lack of understanding of health promotion and illness prevention. CONCLUSIONS Although the challenges that exist in implementing primary health care are clearly understood, clinicians perceive the solutions for these as being within the control of policy makers and those with power within the organisation. / Health Studies / M.A. (Public Health)