Complex data features, such as unmodelled censored event times and variables with time-dependent effects, are common in cancer recurrence studies and pose challenges for Bayesian survival modelling. However, current methodologies for predictive model checking and comparison often fail to adequately address these features. This paper bridges that gap by introducing new, targeted recommendations for predictive assessment and comparison of Bayesian survival models for cancer recurrence. Our recommendations cover a variety of different scenarios and models. Accompanying code together with our implementations to open source software help in replicating the results and applying our recommendations in practice.

翻译：在癌症复发研究中，复杂的数据特征（如未建模的删失事件时间和具有时变效应的变量）十分常见，并对贝叶斯生存建模提出了挑战。然而，当前用于预测模型检验与比较的方法往往未能充分处理这些特征。本文通过针对癌症复发贝叶斯生存模型提出新的、有针对性的预测评估与比较建议，弥补了这一空白。我们的建议涵盖了多种不同场景和模型。随附的代码及我们在开源软件中的实现，有助于复现结果并在实践中应用这些建议。