Poisson-Gamma Mixture Spatially Varying Coefficient Modeling of Small-Area Intestinal Parasites Infection

Area of Science:

• Epidemiology
• Spatial Statistics
• Public Health

Background:

• Understanding the spatial and temporal patterns of intestinal parasite infections is crucial for effective public health strategies.
• Demographic factors and environmental conditions significantly influence disease transmission and prevalence.

Purpose of the Study:

• To evaluate the spatially varying effects of demographic factors on intestinal parasite morbidities in Ghana.
• To identify key risk factors, including unsafe water and sanitation, driving spatial variations in infection rates.

Main Methods:

• A hierarchical Bayesian spatially varying coefficient model was employed.
• Poisson and Poisson-gamma models were developed using morbidity data from the District Health Information Management Systems.
• Integrated Nested Laplace Approximations (INLA) were used for model fitting, incorporating covariates like unsafe drinking water, toilet facilities, and waste disposal.

Main Results:

• The overall risk of intestinal parasite infection was estimated at 10.9 per 100 people, with considerable district-specific spatial variations.
• Unsafe drinking water, unsafe toilets, and unsafe liquid waste disposal showed substantial, spatially varying multiplicative effects on infection risk.
• Structured residual spatial variation dominated, indicating spatially continuous unaccounted-for risk factors.

Conclusions:

• Spatial distribution of posterior risk and exceedance probability maps are vital tools for monitoring and controlling intestinal parasites.
• Targeted interventions addressing water and sanitation are necessary to mitigate spatially varying risks of intestinal parasite infections.