Issues in the mapping of two diseases

Area of Science:

• Epidemiology
• Biostatistics
• Geographical Information Systems (GIS)

Background:

• Growing interest in geographical modeling for multiple diseases.
• Need for standardized methods and comparative analysis of univariate vs. bivariate disease mapping.
• Identifying geographical patterns in disease risk is crucial for public health.

Purpose of the Study:

• To explore issues in geographical modeling of two or more diseases.
• To propose a proportional mortality approach for identifying geographical risk similarities/dissimilarities.
• To investigate the benefits of bivariate modeling for improving risk estimates by 'borrowing strength' between diseases.

Main Methods:

• Comparison of univariate and bivariate disease mapping models.
• Application of a proportional mortality approach.
• Utilizing lung and bladder cancer incidence data (1996-2000) from Washington state for illustration.

Main Results:

• The proportional mortality approach provides insights into geographical areas with similar or dissimilar disease risk distributions.
• Bivariate modeling can enhance the precision of risk estimates by sharing information across diseases.
• Demonstration of various modeling strategies using real-world cancer data.

Conclusions:

• Geographical disease modeling for multiple diseases offers valuable insights into spatial epidemiology.
• The proposed proportional mortality method is effective for detecting geographical risk patterns.
• Bivariate modeling is a powerful tool for improving disease risk estimation in public health surveillance.