Empirical and fully Bayesian methods for identifying the spatial distribution of the adolescent pregnancy rates in a brazilian state

Abstract

Spatial statistical methods are often employed to describe the spatial distribution of the occurrence of an event of interest in a geographical space. In this paper we compare the global and local empirical Bayesian methods and the fully Bayesian method for identifying the spatial distribution of the adolescent pregnancy rates in Minas Gerais, the largest State in the Southeastern of Brazil. The fully Bayesian method consider a conditional autoregressive (CAR) distribution to the spatial components of the model. The methods used in this study suggest that the percentages of live births to adolescent mother (LBAM) are not randomly distributed across the State of Minas Gerais. The percentages of LBAM tend to be higher in the North than in the South of the State. This suggest an important association between the occurrence of teenage pregnancy and socio-economic indicators, given that the North is the poorest region of the State while the South region concentrates the municipalities with the highest rates of development.