Evolution of spatial disease clusters via a Bayesian space-time variability modelling.

Area of Science:

• Epidemiology
• Spatial Statistics
• Bayesian Modeling

Background:

• Understanding the temporal dynamics of disease clusters is crucial for public health interventions.
• Existing models may not fully capture complex spatial and temporal variations in disease patterns.

Purpose of the Study:

• To develop and demonstrate a Bayesian space-time random-effects model for analyzing cluster dynamics.
• To classify spatial trends (hot-spots, cold-spots) and local time trends (increasing, decreasing, stable).

Main Methods:

• Utilized exceedance posterior probabilities from a space-time random-effects model.
• Employed a multivariate Markov Random Field for spatial and temporal trend modeling.
• Estimated parameters within a fully Bayesian framework.

Main Results:

• Developed a 3x3 table classifying time trends within spatial clusters.
• Applied the methodology to intestinal parasite infections in Ghana, revealing spatial clustering evolution.
• Demonstrated the model's applicability to various tropical diseases.

Conclusions:

• The proposed Bayesian methodology effectively analyzes spatial-temporal disease cluster dynamics.
• The approach provides a robust framework for public health surveillance and intervention planning.
• The model is adaptable for diverse infectious diseases with varying space-time characteristics.