Bayesian multivariate disease mapping and ecological regression with errors in covariates: Bayesian estimation of...

Area of Science:

• Biostatistics
• Epidemiology
• Public Health

Background:

• Accurate disease mapping and risk factor analysis are crucial for public health.
• Existing models may not fully account for errors in covariate data.
• Estimating disease burden and directing prevention resources requires robust methodologies.

Purpose of the Study:

• To present Bayesian multivariate disease mapping and ecological regression models that incorporate covariate errors.
• To develop a Bayesian disability-adjusted life years (DALYs) methodology for analyzing multivariate disease/injury data and ecological risk factors.
• To facilitate small-area DALYs estimation, inference, and mapping for targeted health interventions.

Main Methods:

• Developed Bayesian hierarchical formulations for multivariate disease and covariate measurement models.
• Integrated these models into a Bayesian DALYs methodology.
• Employed estimation and inference methods for small-area analysis.

Main Results:

• The methodology enables estimation of multivariate small-area disease/injury rates and risk effects.
• Facilitates evaluation of DALYs and 'preventable' DALYs.
• Identifies geographical regions for directing disease/injury prevention resources.

Conclusions:

• The presented methodology effectively integrates Bayesian disease mapping and the global burden of disease framework.
• It allows for comprehensive evaluation of disease, injury, and risk factor impacts on population health.
• Informs community health needs and priority setting for effective disease and injury prevention strategies.