Assessing Predictive Discrimination Performance of Biomarkers in the Presence of Treatment-induced Dependent Censoring

Authors

Cuihong Zhang
Jing Ning
Steven H Belle
Robert H Squires
Jianwen Cai
Ruosha Li

Publication Date

11-1-2022

Journal

Journal of the Royal Statistical Society Series C: Applied Statistics

Abstract

In medical studies, some therapeutic decisions could lead to dependent censoring for the survival outcome of interest. This is exemplified by a study of paediatric acute liver failure, where death was subject to dependent censoring due to liver transplantation. Existing methods for assessing the predictive performance of biomarkers often pose the independent censoring assumption and are thus not applicable. In this work, we propose to tackle the dependence between the failure event and dependent censoring event using auxiliary information in multiple longitudinal risk factors. We propose estimators of sensitivity, specificity and area under curve, to discern the predictive power of biomarkers for the failure event by removing the disturbance of dependent censoring. Point estimation and inferential procedures were developed by adopting the joint modelling framework. The proposed methods performed satisfactorily in extensive simulation studies. We applied them to examine the predictive value of various biomarkers and risk scores for mortality in the motivating example.

Keywords

area under curve; dependent censoring; joint modelling; net quantities; paediatric acute liver failure; predictive discrimination