Department of Biostatistics, St. Jude Children's Research Hospital, Memphis, TN 38105, U.S.A..
Stat Med. 2011 Nov 10;30(25):3010-23. doi: 10.1002/sim.4306. Epub 2011 Jul 22.
Recurrent event data occur in many clinical and observational studies, and in these situations, there may exist a terminal event such as death that is related to the recurrent event of interest. In addition, sometimes more than one type of recurrent events may occur, that is, one may encounter multivariate recurrent event data with some dependent terminal event. For the analysis of such data, one must take into account the dependence among different types of recurrent events and that between the recurrent events and the terminal event. In this paper, we extend a method for univariate recurrent and terminal events and propose a joint modeling approach for regression analysis of the data and establish the finite and asymptotic properties of the resulting estimates of unknown parameters. The method is applied to a set of bivariate recurrent event data arising from a long-term follow-up study of childhood cancer survivors.
在许多临床和观察性研究中都会出现复发事件数据，在这些情况下，可能存在与感兴趣的复发事件相关的终末事件，例如死亡。此外，有时可能会出现不止一种类型的复发事件，即可能会遇到具有某种相关终末事件的多变量复发事件数据。对于此类数据的分析，必须考虑不同类型的复发事件之间以及复发事件与终末事件之间的相关性。在本文中，我们扩展了一种用于单变量复发和终末事件的方法，并提出了一种联合建模方法，用于对数据进行回归分析，并建立了未知参数的估计值的有限和渐近性质。该方法应用于一组源于儿童癌症幸存者长期随访研究的双变量复发事件数据。
