ORCID

Abstract

Time-to-event, bivariate, semi-competing risk data occur when a terminal event can censor a non-terminal event, but not vice versa. There are potential correlations between these endpoints as they are measured on the same individual. However, traditional methods to estimate the correlations cannot be used directly due to the censoring of time-to-event endpoints. We develop methods using a copula-based approach to study the dependence structures between the two survival endpoints. We use a variety of copulas to estimate the correlation between endpoints and to acknowledge different dependence structures. The estimated association parameter in the copula function is transformed into Spearman's rank correlation coefficient. We conduct a simulation study to evaluate the estimation from the proposed models along with the effects of misspecification of the copula functions and survival distributions. The proposed methods are applied to two real-life data sets.

DOI

10.1002/bimj.202000226

Publication Date

2021-05-01

Publication Title

Biometrical Journal: journal of mathematical methods in biosciences

ISSN

0323-3847

Embargo Period

2021-10-16

Organisational Unit

Peninsula Medical School

Share

COinS