Modelling local patterns of child mortality risk: a Bayesian Spatio-temporal analysis

Abstract

Background: Globally, child mortality rate has remained high over the years, but the figure can be reduced through proper implementation of spatially-targeted public health policies. Due to its alarming rate in comparison to North American standards, child mortality is particularly a health concern in Mexico. Despite this fact, there remains a dearth of studies that address its spatio-temporal identification in the country. The aims of this study are i) to model the evolution of child mortality risk at the municipality level in Greater Mexico City, (ii) to identify municipalities with high, medium, and low risk over time, and (iii) using municipality trends, to ascertain potential high-risk municipalities. Methods: In order to control for the space-time patterns of data, the study performs a Bayesian spatio-temporal analysis. This methodology permits the modelling of the geographical variation of child mortality risk across municipalities, within the studied time span. Results: The analysis shows that most of the high-risk municipalities were in the east, along with a few in the north and west areas of Greater Mexico City. In some of them, it is possible to distinguish an increasing trend in child mortality risk. The outcomes highlight municipalities currently presenting a medium risk but liable to become high risk, given their trend, after the studied period. Finally, the likelihood of child mortality risk illustrates an overall decreasing tendency throughout the 7-year studied period. Conclusions: The identification of high-risk municipalities and risk trends may provide a useful input for policymakers seeking to reduce the incidence of child mortality. The results provide evidence that supports the use of geographical targeting in policy interventions.