比例风险模型下病例队列设计中两种加权估计方法及其应用

INFERENCE AND APPLICATION OF CASE-COHORT DESIGN UNDER THE PROPORTIONAL HAZARDS MODEL

对于大型队列研究或观察型研究,基于生存数据的病例队列设计是一种能有效节约成本和提高效率的抽样机制.这种抽样设计仅对一个随机抽取的子队列以及子队列之外所有经历了感兴趣事件的病例个体进行关键协变量的测量,具有显著的成本效益.本文研究如何应用比例风险模型拟合病例队列研究数据.探讨逆概率加权和与时间相关加权这两种基于加权估计方程的统计推断方法和其渐近性质等理论结果.通过一系列的统计模拟研究展示了病例队列设计的优良性以及相较于传统简单随机抽样设计的高效性.进一步,应用这两种推断方法分析了两个实际数据,展示了其在实际中的应用价值和前景.

A case-cohort design is a cost-effective sampling scheme in large cohort studies.The key idea of such a design is to assemble the measurements of expensive covariates only on a subset of the entire cohort and all the subjects outside the subcohort that experience the event of interest.In this paper,we study the inference methods for case-cohort data under the Cox model.We consider two weighted estimating equation approaches,the inverse-probability and time-varying weighted methods.The asymptotic theories are established.A series of simulation studies are conducted to assess the finite-sample performance of the proposed methods and exhibit the superiority and efficiency of the case-cohort design.Some real data examples are analyzed to illustrate the application of the proposed methods.