Bayesian structured additive regression modeling of epidemic data: application to cholera

Area of Science:

• Spatial epidemiology
• Public health
• Infectious disease modeling

Background:

• Limited use of spatial data and nonlinear effects in traditional epidemiological studies.
• Need for advanced statistical models to capture complex risk factor relationships.
• Growing interest in spatial epidemiology for identifying infection risk factors.

Purpose of the Study:

• To develop and apply a Bayesian Structured Additive Regression (BSAR) model for analyzing cholera epidemic data.
• To investigate the spatial distribution and risk factors associated with cholera outbreaks.
• To incorporate nonlinear effects of risk factors and spatial dependencies in the analysis.

Main Methods:

• Utilized a Bayesian Structured Additive Regression (BSAR) model.
• Employed Markov Chain Monte Carlo (MCMC) simulations for model estimation and inference.
• Modeled proximity and density of refuse dumps, slum settlement presence, population density, and spatial effects.

Main Results:

• Cholera risk is significantly associated with slum settlements and high population density.
• Proximity to refuse dumps and their density influence cholera risk, with higher risk near more dumps.
• Distinct spatial variations in cholera infection risk were observed across communities.

Conclusions:

• Bayesian semi-parametric regression models are effective for analyzing public health data, particularly for infectious diseases.
• Findings provide crucial information for health planners and policymakers to control and prevent cholera epidemics.
• The study underscores the importance of considering spatial factors and nonlinear relationships in epidemiological research.