Random right censoring is a common phenomenon in survival analysis. In such cases, the estimation of the marginal distribution of the survival time can be challenging. We often assume that the survival time T and the censoring time C are independent of each other to identify the distribution. However, in practice, T can be stochastically dependent on C. In cancer-survival or other medical studies, and individual may leave the study for reasons that are related to the treatment given or the health condition of the patient. The individual may even die of another disease sharing common risk factors. We review the copula-based approaches that can handle dependent censoring. In particular, we consider the model proposed by Czado and Van Keilegom (2021). Unlike the models proposed earlier, this approach does not assume that the associated copula function is known. We extend this model to include the effects of covariates, study the small sample performance through extensive simulations and illustrate the model through an application on Pancreas cancer data.      «

Random right censoring is a common phenomenon in survival analysis. In such cases, the estimation of the marginal distribution of the survival time can be challenging. We often assume that the survival time T and the censoring time C are independent of each other to identify the distribution. However, in practice, T can be stochastically dependent on C. In cancer-survival or other medical studies, and individual may leave the study for reasons that are related to the treatment given or the healt...     »