[Lecture] Mengjiao Peng: Analysis of complex survival data subject to semi-competing risks
Time: 14:00-15:00pm, Oct. 14th, 2020, Wednesday
Venue: RoomA1716, Science Building, North Zhongshan Road Campus
Spreker: Mengjiao Peng, Assistant Professor, Academy of Statistics and Interdisciplinary Science, East China Normal University
Abstract:
In this talk, I will introduce my past and recent work in the analysis of complex survival data subjectto semi-competing risks. Semi-competing risks data is a type of survival data including multiplesurvival endpoints, which is quite common in medical research and clinical trials for assessingtreatment efficacy and disease progression. In the work to analyze clustered endpoints with semi[1]competing risks via a flexible semiparametric modeling framework,we develop a nonparametricmaximum likelihood estimation procedure via a Monte Carlo EM algorithm, and establish desirableasymptotic properties of the resulting estimators. In the study of multiple time-to-event endpointssubject to two sets of semi-competing risks, we propose a novel statistical approach that jointlymodels such data via a pair of copulas to account for multiple dependence structures, and develop apseudo-likelihoodestimationprocedure.Wealsoproposetwomodel-freefeaturescreeningmethods to select the important features for semi-competing risks data, which are often collectedwith ultrahigh-dimensional gene features in modern cancer studies. Extensive simulation studieshave been carried out and resultsshow promising finite sample performance of all proposedmethods. Analysis of real data sets from various medical studies demonstrate the practical utility ofour proposed methods.
