ABSTRACT
This chapter presents a Bayesian cure models with spatial random effects. It describes the flexible cure model with latent activation schemes and also describes the Bayesian cure model with generalized modified Weibull distribution. The chapter evaluates the mixture cure model with spatially correlated frailties for population-based cancer survival data. We apply the spatial mixture cure model to estimate the survival and cure rate for patients diagnosed with lung and bronchus cancer from 1992 to 2013 in Iowa. The chapter discusses the general implementation of Bayesian analysis using BUGS language. Bayesian cure model has been implemented using the OpenBUGS, which can be called within R using BRugs. The Bayesian framework is easy to implement and can be extended to more general situations, e.g., the Bayesian cure model with frailty and Bayesian cure model that accommodates both multiplicative and additive covariates.
