Name of the Project
Mortality of children under five years and their risk factors associated in selected countries in Latin America
Lead researcher
Pamela Cordova, PhD
Soraya Román, PhD
Contact: pcordova@upb.edu, sorayaroman@upb.ed
Summary
The high mortality rates of children under five in Latin America continue to be a major public health problem today, despite the fact that most deaths can be prevented by known and relatively low cost technologies. The evidence suggests that both the individual characteristics (including the sex of the child, birth order, birth interval, maternal education, family planning as well as the use of contraceptives), of the household (such as household size, access to basic services and index of household wealth) as well as community or contextual (the region, place of residence, community infrastructure-water, electricity, roads-, deliveries in hospitals at the community level, maternal education in the community and the level of poverty) have an influence on health outcomes. The main objective of the present study proposal is to establish the determinants of the mortality of children under five in selected countries in Latin America with emphasis on the role of the community context. Using two years of the National Demographic and Health Survey (ENDSA) of six countries for which the surveys are recent and available, we will estimate the change in relative risk of death, as well as the main contributions to the change in mortality at over time in relation to individual, household and contextual characteristics. Taking into account the compatibility of microdata for a multi-country analysis, the study will focus on: Bolivia (2003-2008), Colombia (2005-2010), Dominican Republic (2007-2013), Haiti (2006-2012), Honduras (2006-2012) and Peru (2007-2012). Using these surveys, the study proposes, (1) to examine the effects of the characteristics at the individual, household and community level (context) on infant and childhood mortality and (2) determine the extent to which these characteristics and especially the context ( as a proxy for possible effects of public policy) they influence the variations in mortality in the sample of six Latin American countries. Three levels of analysis are proposed: univariate, bivariate and multivariate-multilevel. At the univariate level, a descriptive analysis of the characteristics identified at the individual, household and contextual level will be carried out in order to establish the distribution of the individuals under study by the selected variables. At the bivariate level, a cross-tabulation will be performed and a Pearson's chi-square test will be performed to examine the relationship between the outcome variables of infant and childhood mortality and the independent variables selected at each level of characteristics. The multivariate analysis will be carried out using the Cox Proportional Risk Model and the multilevel Cox proportional model to examine the effects of individual, household and community characteristics on the mortality of children under five years of age and determine to what extent these characteristics explain the variations in mortality in the selected countries and thus approximate results for Latin America. This study suggests the need to look beyond the influence of individual and household factors in the treatment of variations in the survival of children under five in Latin America and seeks to focus on contextual factors as key determinants of it. which are somehow strongly explained by public policy