Parametric models for spatially correlated survival data for individuals with multiple cancers.
Parametric models for spatially correlated survival data for individuals with multiple cancers.
复制标题
患有多种癌症的个体的空间相关生存数据的参数模型。
DOI：
10.1002/sim.3141
复制
发表时间：
2008
影响因子：
2
通讯作者：
Banerjee,Sudipto
中科院分区：
文献类型：
--
作者：
Diva,Ulysses;Dey,DipakK;Banerjee,Sudipto
消除空间变异可能会增强来自生存数据的信息。此外，对来自同一患者的不同疾病（如癌症）的事件发生时间数据进行同步（联合）建模，可以提供有关这些疾病如何共同表现的有用见解。本文提出了贝叶斯分层生存模型中的比例风险（PH）和比例优势（PO）的框架内捕捉空间相关性。参数（PH为Weibull，PO为log-logistic）模型用于基线分布，同时以县癌症水平脆弱性的形式引入空间相关性。我们用来自国家癌症研究所的监测流行病学和最终结果数据库的数据对爱荷华州诊断为多种胃肠道癌症的患者进行说明。模型检查和竞争模型之间的比较，并提出了一些实现问题。我们建议对该数据集使用空间PH模型。版权所有© 2008约翰威利父子有限公司。
Incorporating spatial variation could potentially enhance information coming from survival data. In addition, simultaneous (joint) modeling of time‐to‐event data from different diseases, such as cancers, from the same patient could provide useful insights as to how these diseases behave together. This paper proposes Bayesian hierarchical survival models for capturing spatial correlations within the proportional hazards (PH) and proportional odds (PO) frameworks. Parametric (Weibull for the PH and log‐logistic for the PO) models were used for the baseline distribution while spatial correlation is introduced in the form of county–cancer‐level frailties. We illustrate with data from the Surveillance Epidemiology and End Results database of the National Cancer Institute on patients in Iowa diagnosed with multiple gastrointestinal cancers. Model checking and comparison among competing models were performed and some implementation issues were presented. We recommend the use of the spatial PH model for this data set. Copyright © 2008 John Wiley & Sons, Ltd.