Competing risks survival data under middle censoring-An application to COVID-19 pandemic

Area of Science:

• Biostatistics
• Survival Analysis
• Epidemiology

Background:

• The COVID-19 pandemic highlighted the need for robust survival analysis methods, particularly under middle censoring.
• Competing risks are common in medical studies, necessitating specialized statistical approaches.

Purpose of the Study:

• To develop and evaluate statistical methods for survival data analysis under a middle censoring scheme.
• To apply quantile function modeling within a competing risks framework, suitable for pandemic scenarios.

Main Methods:

• Utilizing cause-specific quantile inference based on the cumulative incidence function.
• Employing a cause-specific proportional hazards model with a Weibull baseline distribution.
• Estimating parameters and cause-specific quantile functions using both classical and Bayesian methods.

Main Results:

• The proposed methods provide reliable estimates for survival data under middle censoring.
• Both classical and Bayesian approaches demonstrate effectiveness in parameter and quantile function estimation.
• Monte Carlo simulations confirm the relative performance of the developed estimators.

Conclusions:

• The middle censoring scheme and quantile function modeling offer a powerful approach for analyzing competing risks survival data.
• The methods are applicable to real-world data, as illustrated by a case study.
• The findings have implications for statistical modeling in epidemiological research, especially during health crises.