Locating irregularly shaped clusters of infection intensity

Area of Science:

• Spatial epidemiology
• Geographic information systems (GIS)
• Infectious disease modeling

Background:

• Disease risk is often influenced by environmental factors, leading to irregular geographic patterns.
• Traditional cluster detection methods are limited to circular shapes, reducing their power for irregular spatial anomalies.
• Identifying non-circular disease clusters is crucial for understanding environmental influences on health.

Purpose of the Study:

• To introduce and evaluate a novel method, the "greedy growth scan," for detecting irregularly shaped spatial clusters.
• To assess the method's efficacy at micro-geographical scales using simulated and real-world parasitic infection data.
• To improve the identification of localized disease risk factors.

Main Methods:

• Modification of the spatial scan method to create the "greedy growth scan" algorithm.
• Application of the method to simulated data to compare performance against fixed geometry methods.
• Analysis of real-world infection intensity data for Schistosoma mansoni and hookworm in Kenya.

Main Results:

• Simulated data analysis demonstrated superior detection of irregular clusters (e.g., along rivers) compared to traditional methods.
• The greedy growth scan identified two localized areas of elevated infection intensity in the Kenyan study region.
• The method proved effective in detecting spatial anomalies at micro-geographical scales.

Conclusions:

• The "greedy growth scan" is a powerful tool for exploratory geographical analysis of disease intensity data.
• This method is particularly suitable when irregular spatial patterns are suspected, especially at fine spatial resolutions.
• The approach aids in identifying localized environmental or social factors potentially driving disease risk.