Comparing different spatio-temporal modeling methods in dengue fever data analysis in Colombia during 2012-2015

Area of Science:

• Epidemiology
• Geographic Information Systems (GIS)
• Environmental Science

Background:

• Dengue fever is a significant public health concern, particularly in tropical and subtropical regions.
• Understanding the spatio-temporal dynamics of dengue transmission is crucial for effective disease control.
• Environmental factors play a critical role in the geographic distribution and incidence of dengue fever.

Purpose of the Study:

• To compare various spatio-temporal conditional autoregressive models for dengue fever analysis in Colombia.
• To investigate the influence of environmental variables on dengue fever outbreaks using Bayesian modeling.
• To develop a predictive model for identifying regions at high risk of dengue fever.

Main Methods:

• Application of Poisson-lognormal and binomial models with Bayesian spatio-temporal methods.
• Incorporation of an innovative data transformation technique for enhanced analysis.
• Utilizing a dengue fever dataset from Colombia, including variables such as population density, elevation, land surface temperature, precipitation, and vegetation.

Main Results:

• The selected spatio-temporal mixed effects model effectively captures data variations.
• Population density, elevation, and land surface temperatures (daytime and night) are significant predictors of dengue fever outbreaks.
• Precipitation and vegetation variables were found to be not significant in the developed model.
• Generated dengue fever probability maps highlight a geographic risk distribution aligned with the elevation gradient.

Conclusions:

• The study successfully identified key environmental determinants for dengue fever risk in Colombia.
• The findings provide valuable insights for targeted public health interventions and resource allocation.
• The developed modeling approach offers a robust framework for future dengue fever research and surveillance.