A heterogeneity measure for cluster identification with application to disease mapping

Area of Science:

• Epidemiology
• Spatial Statistics
• Public Health

Background:

• Disease mapping is crucial for epidemiology and public health.
• Identifying spatial clusters of elevated disease rates is a key challenge.

Purpose of the Study:

• To develop a new approach for identifying spatial disease clusters.
• To estimate relative disease risk associated with risk factors and detect clusters simultaneously.

Main Methods:

• Proposed a heterogeneity measure for comparing clusters and their complements.
• Developed a quasi-likelihood procedure for parameter estimation and cluster identification.
• The method identifies clusters of arbitrary shapes, accounting for risk factors and spatial correlation.

Main Results:

• The methodology establishes asymptotic properties.
• Simulation studies demonstrate sound finite-sample properties.
• Applied the method to map and cluster enterovirus 71 infections in Taiwan.

Conclusions:

• The novel approach enhances spatial clustering by identifying arbitrary-shaped clusters.
• This method improves disease risk assessment and public health monitoring.
• The illustrated application demonstrates the practical utility in infectious disease surveillance.