Seemingly Unrelated Multi-State Processes: A Bayesian Semiparametric Approach

Many applications in medical statistics and other fields can be described by transitions between multiple states (e.g. from health to disease) experienced individuals over time. In this context, multi-state models are a popular statistical technique, particular when the exact transition times not observed. The key quantities of interest rates, capturing instantaneous risk moving one state another. main contribution work is propose joint semiparametric model for several possibly related processes (Seemingly Unrelated Multi-State, SUMS, processes), assuming Markov structure dependence different captured specifying prior distribution on rates each process. case, we assume flexible distribution, which allows clustering individuals, overdispersion outliers. Moreover, employ graph describe among processes, exploiting tools Gaussian Graphical literature. It also possible include covariate effects. We use our approach disease progression mental health. Posterior inference performed through specially devised MCMC algorithm.