Assessing Performance of the Generalized Exponential Model in the Presence of the Interval Censored Data with Covariate

Nada Alharbi,Haizum A,Wendy Ling,Jayanthi A

doi:10.17713/ajs.v51i1.1192

Nada Alharbi, Haizum A + Show 2 more

Open Access

https://doi.org/10.17713/ajs.v51i1.1192

Copy DOI

Abstract

This study aims to extend the generalized exponential model (GEM) to include covariates in the presence of interval-censored data. The maximum likelihood estimator (MLE) was obtained for the parameter of the model formulated. Afterward, a thorough simulation study was carried out to evaluate the estimator's performance based on the values of bias, standard error (SE), and root mean square error (RMSE). The result indicated that the (SE) and (RMSE) decrease with the increase in sample sizes and decrease in censoring proportions. Finally, the performance of the Wald confidence interval estimation technique for the GE model with interval-censored data covariate was assessed by a coverage probability study at several censoring proportions and different sample sizes.

Highlights

Survival time data analysis, concisely referred to as survival analysis is a collection of statistical methods for analyzing time to an event data
A severe analytical problem in analyzing survival time data arises when a portion or even all ti, i = 1, 2, ..., n, are censored data. (Klein and Moeschberger 2006) explained that the ith subject’s survival time, ti, is censored data when its exact value is unknown due to one of these four main reasons; (1) The subject lost to follow-up, (2) The subject withdraws from the study, (3) The subject does not experience the event before the study ends, and (4) The subject is not under continuous observation
The first approach in survival analysis based on interval-censored data was presented by (Turnbull 1976), who discussed the estimation of the non-parametric survival function

Summary

Introduction

Concisely referred to as survival analysis is a collection of statistical methods for analyzing time to an event data. The focus of these methods is to describe the distribution of T on a population and relationship between T and some covariates. Further studies on interval-censoring were explored by (Brunel, Comte et al 2009), where other types of smooth estimators were proposed for interval-censored data.The survival model with doubly interval-censored data and time dependent covariate in which the lifetime is the elapsed time between two related events which means that the first event and the second event are interval-censored that discussed on Kiani and Arasan (2018). The performance of the Wald confidence interval was examined by conducting coverage probability study at various cp, n, and α

Objectives

Methods

Conclusion