In some applications, clustered survival data are arranged spatio-temporally such as geographical regions over multiple time periods. Incorporating spatio-temporal variation in these data not only can improve the accuracy and efficiency of parameter estimation, but it also investigates spatial pattern of survivorship over the study period for identifying high-risk areas. Competing risks in survival data concern a situation where there is more than one cause of failure, but only the occurrence of the first one is observable. In this paper, we considered several Bayesian hierarchical survival models in the setting of competing risks for the spatio-temporally clustered HIV/AIDS data. An intrinsic conditional autoregressive (ICAR) distribution and a multivariate intrinsic conditional autoregressive (MICAR) distribution were employed to model random effect terms. The comparison between competing models was performed using the deviance information criterion and log pseudo-marginal likelihood. We illustrated the gains of final model through simulation study and application to the HIV/AIDS data.

在某些应用中，聚类的生存数据是按时空排列的，例如多个时间段上的地理区域。在这些数据中纳入时空变化不仅可以提高参数估计的准确性和效率，而且还可以研究研究期内生存空间的格局，以识别高风险区域。生存数据中存在竞争风险的情况涉及失败原因不止一个的情况，但是只有第一个失败的原因是可以观察到的。在本文中，我们在时空聚类的HIV / AIDS数据的竞争风险设置中考虑了几种贝叶斯分层生存模型。内在条件自回归（ICAR）分布和多元内在条件自回归（MICAR）分布用于对随机效应项进行建模。使用偏差信息标准和对数伪边际对数进行竞争模型之间的比较。我们通过模拟研究并将其应用于HIV / AIDS数据说明了最终模型的收益。