Fatores associados à mortalidade infantil em Moçambique, 1998 a 2003

Abstract

In this study a database of the Demographic and Health Survey (IDS 2003) for Mozambique was used to study demographic, socioeconomic and health care factors associated to neonatal and post-neonatal. The model of binary logistic regression was applied to analyze 9,849 cases of live births between August 1998 and July 2003. First, univariate analysis was performed, which allowed selecting variables that were eligible for multivariable models in both components (neonatal and post neonatal). Then, we built multivariate models. Seven models for neonatal and eight models for post-neonatal period were built. The results showed that being the oldest son, having small size at birth, having a mother with less than 4 years of schooling, and having a mother living in the North region were independently associated with neonatal mortality. Lower maternal age (less than 20 years) at the child's birth lost significance after inclusion of socioeconomic and health care variables in the model. For the post neonatal death, younger age of mother (less than 20) were associated with post neonatal mortality in all models. Maternal age of 35 or higher and big size at birth were found to be protective of post neonatal mortality and the result remained significant in all models. The low schooling level of the mother, small size at birth, being the oldest son, having a mother living at North or Center region of the country and domiciliary birth were statistically associated with post-neonatal mortality. In summary, the analysis showed that a more efficient implementation of health programs related to children and to give more attention to groups of mothers and newborns at increased risk for infant mortality in Mozambique are key factors in order to reach lower levels of mortality.