Fitting semiparametric cure models
Fitting semiparametric cure models
复制标题
DOI：
10.1016/s0167-9473(02)00184-6
复制
发表时间：
2003-01-28
影响因子：
1.8
通讯作者：
Peng, YW
中科院分区：
文献类型：
--
作者：
Peng, YW
在一些癌症临床研究中经常会遇到具有相当大治愈率的生存数据，最近半参数比例风险治愈模型已经被研究来分析这样的数据。然而模型的估计方法需要专门的C程序。该模型的限制性假设也限制了其在更广泛环境中的应用。本文提出了一种新的治愈模型计算方法。该方法结合了逻辑回归和考克斯比例风险模型的计算方法，易于在许多统计软件包中实现。我们还展示了这种方法如何允许一些有用的扩展模型放松限制性假设。提供了淋巴瘤患者生存时间的说明性实例。(C)2002 Elsevier Science B.V.保留所有权利。
Survival data with a sizable cure fraction are commonly encountered in some cancer clinical researches and the semiparametric proportional hazards cure model has been recently investigated to analyze such data. However the estimation method of the model requires a special C program. The restrictive assumptions of the model also limit its application in broader settings. We present a new computational method for the cure model in this article. The method combines the computational methods for logistic regression and the Cox proportional hazards models and is easy to implement in many statistical packages. We also show how this method allows a number of useful extensions of the model to relax the restrictive assumptions. An illustrative example with survival times of lymphoma patients is provided. (C) 2002 Elsevier Science B.V. All rights reserved.