窗体顶端

教育部人文社科重点研究基地中央财经大学中国精算研究院学术活动

精算论坛第211期讲座

(2022年11月14日)

讲座主题：Semiparametric Copula Method for Semi-Competing Risks Data subject to Interval Censoring and Left Truncation: Application to Disability in Elderly

摘要：According to WHO, more than 46% of elderly aged 60 years and over have disabilities. It is crucial to assess the impact of covariatefactors on the time-to-disability in the elderly for early prevention and management. An important challenge for studying time-to-disability in the elderly is the high mortality because death dependently censors disability but not vice versa, leading to semi-competing risks data. One approach to analyzing the semi-competing risks data is the copula model, which can explicitly account for the dependence between disability and death. However, there is currently a gap in fitting such a model in disability data due to two major challenges. One is that the time-to-disability is subject to interval censoring due to intermittent assessments; the other is the left truncation issue when the time is on the age scale, which is more preferred than the study-entry scale in ageing studies. We develop a novel two-parameter copula-based semiparametric transformation model that handles interval censoring and left truncation in semi-competing risks data. Particularly, the two-parameter copula quantifies both upper and lower tail dependence between the marginal distributions of disability and death. The semiparametric transformation models are employed to incorporate various model assumptions in both margins. Numerical simulations show that the proposed method corrects estimation biases and provides correct coverage probabilities. We apply the proposed method to a dataset of 12,969 elders over 60 years of age from the Chinese Longitudinal Healthy Longevity Survey and assess the impact of covariatefactors (i.e., age, sex, education, marriage) on disability and death. Our work is the first research on disability that adopts a solid statistical model that appropriately handles interval censoring and left truncation in the semi-competing risks data

报告人：孙韬

中国人民大学统计学院讲师，博士毕业于匹兹堡大学生物统计系，主要研究方向为复杂生存数据模型，老年慢性病预防与管理。主持国自然青年基金项目与国家统计局重点项目，学术论文发表于Science,Biometrics,Biostatistics, Statistics in Medicine, Statistical Methods in Medical Research等期刊。

讲座时间：2022年11月14日下午14:00-15:00

报告地点：腾讯会议（会议ID：869 633 112，密码：123456。）

邀请人：韦晓

欢迎各位老师和同学积极参加！

窗体底端