Abstract:In this paper we consider a time-to-event variable $T$ that is subject to random right censoring, and we assume that the censoring time $C$ is stochastically dependent on $T$ and that there is a positive probability of not observing the event. There are various situations in practice where this happens, and appropriate models and methods need to be considered to avoid biased estimators of the survival function or incorrect conclusions in clinical trials. We consider a fully parametric model for the bivariate distribution of $(T,C)$, that takes these features into account. The model depends on a parametric copula (with unknown association parameter) and on parametric marginal distributions for $T$ and $C$. Sufficient conditions are developed under which the model is identified, and an estimation procedure is proposed. In particular, our model allows to identify and estimate the association between $T$ and $C$, even though only the smallest of these variables is observable. The finite sample performance of the estimated parameters is illustrated by means of a thorough simulation study and the analysis of breast cancer data.

What problem does this paper attempt to address?

### Problems Addressed by the Paper The paper primarily addresses the issues of model identification and parameter estimation in survival analysis when there is a cure fraction and dependent censoring. Specifically: 1. **Research Background**: - In survival analysis, random right censoring is commonly encountered, where the event time $T$ is influenced by a random censoring time $C$. - In practice, there are many situations where there is a dependency between survival time and censoring time. For example, in cancer research, patients may drop out of the study due to deteriorating health, in which case survival time and censoring time may be positively correlated; conversely, if patients drop out due to significant health improvement, they may be negatively correlated. - Traditional survival analysis assumes that all individuals will experience the event of interest, but in some cases, a portion of individuals (long-term survivors) will never experience the event, and these individuals are considered cured. 2. **Core Issues**: - When there is a cure fraction and dependent censoring, traditional survival analysis methods cannot correctly handle the data, leading to biased results. - Although there is existing literature that separately studies the issues of cure fraction and dependent censoring, there is little literature that considers both simultaneously. - The authors propose a new fully parametric mixture cure model that allows for the presence of a cure fraction and dependent censoring without assuming a known copula. 3. **Objectives**: - To study the identifiability conditions of models that can handle both cure fraction and dependent censoring without requiring strong prior assumptions. - To propose a new fully parametric mixture cure model that allows for the presence of a cure fraction and dependent censoring and to prove that the model is identifiable under more general conditions. - To provide asymptotically normal estimates of the dependence strength and cure fraction. Through the above research, the paper aims to address the issues of model identification and parameter estimation in survival analysis when both cure fraction and dependent censoring are present, thereby avoiding biased results and improving the effectiveness of the model.

Copula based dependent censoring in cure models

Copula-based Semiparametric Nonnormal Transformed Linear Model for Survival Data with Dependent Censoring

Conditional Copula Models for Right-Censored Clustered Event Time Data

Quantile regression under dependent censoring with unknown association

Flexible control function approach under different types of dependent censoring

On an extension of the promotion time cure model

Survival Estimation for Missing not at Random Censoring Indicators based on Copula Models

A copula-based boosting model for time-to-event prediction with dependent censoring

Dependent censoring with simultaneous death times based on the Generalized Marshall-Olkin model

Nonparametric estimation of conditional cure models for heavy-tailed distributions and under insufficient follow-up

Clayton copula for survival data with dependent censoring: An application to a tuberculosis treatment adherence data

Modeling excess hazard with time--to--cure as a parameter

Likelihood-Based Inference for Semi-Parametric Transformation Cure Models with Interval Censored Data

Estimation and Model Selection of Semiparametric Multivariate Survival Functions under General Censorship

Bayesian cure-rate survival model with spatially structured censoring

Copula-Based Deep Survival Models for Dependent Censoring

Deep Copula-Based Survival Analysis for Dependent Censoring with Identifiability Guarantees

Copula-based semiparametric regression method for bivariate data under general interval censoring

Generalized parametric cure models for relative survival

Time-dependent receiver operating characteristic curve estimator for correlated right-censored time-to-event data

Causal Inference for Time-to-event Data with a Cured Subpopulation.