Disease Mapping, Hierarchical Models For

Abstract

Hierarchical models for disease mapping. Hierarchical models provide a convenient framework to combine information across multiple observational units. Such models are particularly convenient for small area estimation of disease rates to improve statistical precision of estimates associated with small local sample sizes. We review the basic framework of model‐based approaches for combining information across small administrative units through the use of random effects. Varying definitions of the random effects distribution allows one to combine information from all locations, neighboring locations, or a combination of both.