Bayesian Analysis of Masked Competing Risks Data Based on Proportional Subdistribution Hazards Model

Masked issues can emerge when dealing with competing risk data. Such are exemplified by the cause of a particular failure not being directly exhibited for all units to observe but only proven be subset possible causes failure. For assessing impact explanatory variables (covariates) on cumulative incidence function (CIF), process Bayesian analysis is discussed in this paper. The symmetry assumption imposed masking probabilities and independent Dirichlet priors assigned them. Markov Chain Monte Carlo (MCMC) technique utilized implement analysis. effectiveness developed model tested via numerical studies, including simulated real data sets.