Bayesian cluster detection via adjacency modelling.

Area of Science:

• Spatial epidemiology
• Biostatistics
• Geographic Information Systems (GIS)

Background:

• Disease mapping estimates spatial patterns in disease risk.
• Current Bayesian methods struggle to delineate contiguous high-risk clusters.
• Identifying the geographical extent of disease clusters is crucial for targeted interventions.

Purpose of the Study:

• To develop and validate a novel two-stage approach for disease mapping.
• To accurately identify and delineate spatially contiguous high-risk disease clusters.
• To improve the estimation of spatial patterns in disease risk.

Main Methods:

• A two-stage approach combining clustering and Bayesian modeling.
• Stage 1: Spatially adjusted hierarchical agglomerative clustering.
• Stage 2: Poisson log-linear model to select the optimal cluster structure.

Main Results:

• The proposed method successfully identifies high-risk clusters.
• Applied to Chronic Obstructive Pulmonary Disease (COPD) in England, several high-risk clusters were detected.
• The methodology provides a robust way to define the geographical extent of disease clusters.

Conclusions:

• The novel two-stage method enhances disease mapping capabilities.
• It accurately identifies spatially contiguous high-risk clusters, overcoming limitations of existing techniques.
• This approach offers improved tools for spatial epidemiological analysis and public health planning.