Application of a General Family of Bivariate Distributions in Modelling Dependent Competing Risks Data with Associated Model Selection

TL;DR

This paper introduces a flexible bivariate distribution family for modeling dependent competing risks data, including ties, and provides an easy-to-implement inferential framework validated through simulations and real data analysis.

Contribution

It proposes a new general bivariate distribution family for dependent competing risks data with ties and develops comprehensive inferential methods for model fitting and selection.

Findings

Inferential methods yield reasonable estimates in simulations.

Model effectively captures dependence in competing risks data.

Real data analysis demonstrates practical applicability.

Abstract

In this article, a general family of bivariate distributions is used to model competing risks data with dependent factors. The general structure of competing risks data considered here includes ties. A comprehensive inferential framework for the proposed model is presented: maximum likelihood estimation, confidence interval construction, and model selection within the bivariate family of distributions for a given dependent competing risks data. The inferential methods are very convenient to implement. Through detailed simulations, the inferential methods are observed to provide quite reasonable results. Analysis of a real data from the Diabetic Retinopathy Study is carried out with the help of the proposed model as an illustrative example.