Assessing heterogeneity and correlation of paired failure times with the bivariate frailty model

We consider bivariate survival times for heterogeneous populations, where h eterogeneity induces deviations in an individual's risk of an event as well as associations between survival times. The heterogeneity is characterized by a bivariate frailty model. We measure the heterogeneity effects through deviations associated with hazard functions and an association function de fined through the conditional hazard functions: the cross-ratio function pr oposed by Oakes. We show how the deviation and association measures are det ermined by the frailty distribution. A Gibbs sampling method is developed f or Bayesian inferences on regression coefficients, frailty parameters and t he heterogeneity measures. The method is applied to a mental health care da ta set. Copyright (C) 1999 John Wiley & Sons, Ltd.