CMStatistics 2020: Start Registration
View Submission - CMStatistics
Title: Bayesian joint modelling of recurrence and survival: A conditional approach Authors:  Willem van den Boom - National University of Singapore (Singapore) [presenting]
Maria De Iorio - Yale-NUS College (Singapore)
Marta Tallarita - University College London (United Kingdom)
Abstract: Recurrent event processes describe the repetition of an event over time. A recurrent event process is often terminated or censored by another event with dependence between the termination time and recurrence process. For instance, recurrent disease events might be terminated by death, while frailty might affect both disease recurrence and survival. As such, it is important to model the recurrent event process and the termination time process jointly to better capture the dependency between them. We propose a model in which the number of gap times, i.e. the time between two consecutive recurrent events, before the terminal event occurs is a random variable of interest. Then, conditionally on the number of recurrent events before the termination event, we specify a joint distribution for the gap times and the survival time. This novel conditional approach induces dependence between the recurrence and survival process. Additional dependence between recurrence and survival is introduced by a joint distribution on their respective frailty terms. A non-parametric random effects distribution for the frailty terms accommodates population heterogeneity and allows for data-driven clustering of the subjects. Posterior inference is performed through a tailor-made Gibbs sampler strategy involving a reversible jump step and slice sampling.