Modeling levels of overdispersion in hierarchical medical data

The presence of overdispersion in binary data is well accepted. Several methods of modeling such extravariation have been considered in the literature. However, most of these methods assume that the overdispersion has occurred at the primary level of the data. In hierarchical models such extravariation can be present for levels other than at the primary stage. In this paper we model healthcare data with a hierarchical structure while accounting for the overdispersion present at varying levels.