A Spatial and Bayesian Approach to Infant Mortality in Puerto Rico: A Study of Risk, Clustering and Determinants

Alexis R. Santos-Lozada, University of Texas at San Antonio

The purpose of this research is to employ spatial statistics to study infant mortality in Puerto Rico. A spatial approach is deemed appropriate for this case as research done on spatial dynamics in Puerto Rico tend to point to space playing an important role in most outcomes. Spatial statistics will be used to study, risk, clustering of the outcome and the covariates. At the same time, this paper includes a Bayesian approach for the study of Infant Mortality in the island and will also apply multilevel models to test for existence of multilevel influences in the outcome. Numerous clusters were identified in the island, with these areas exhibiting higher indicators of risk. County level factors influence infant mortality in a consistent way when considered using multiple regression model approaches.