Bayesian disease mapping using product partition models

Area of Science:

• Epidemiology
• Biostatistics
• Spatial Analysis

Background:

• Estimating geographic variations in disease risk is crucial for public health.
• Existing methods like the standardized mortality ratio may not fully capture complex spatial patterns.
• There is a need for advanced statistical models to accurately identify disease hotspots and coldspots.

Purpose of the Study:

• To develop and validate a model for estimating relative disease risk in defined geographic areas.
• To identify areas exhibiting unusually high or low disease risk.
• To compare the performance of the proposed model against established methods.

Main Methods:

• Utilized a product partition model (PPM) assuming relative risks are constant within specific sets of areas.
• Employed Markov chain Monte Carlo (MCMC) techniques for approximating posterior distributions of model parameters and partitions.
• Validated the model through a simulation study before applying it to real-world cancer data.

Main Results:

• The product partition model (PPM) effectively estimates relative disease risk across different geographic areas.
• Identified specific regions with statistically significant high or low disease risk.
• Demonstrated comparable or improved performance against standardized mortality ratio, empirical Bayes, spatial scan, and nonparametric Bayesian methods.

Conclusions:

• The developed product partition model (PPM) provides a robust framework for spatial disease risk estimation.
• The model enhances the ability to detect localized disease risk variations, aiding targeted public health interventions.
• This approach offers a valuable alternative to traditional methods for spatial epidemiological analysis.