Generalized mean residual life models for survival data with missing censoring indicators

Area of Science:

• Biostatistics
• Survival Analysis
• Statistical Modeling

Background:

• The mean residual life (MRL) function is a valuable tool for analyzing time-to-event data.
• Standard survival analysis methods often struggle with missing censoring indicators.
• Developing robust models for such data is crucial in medical research.

Purpose of the Study:

• To develop and evaluate generalized mean residual life (MRL) models.
• To address right-censored survival data where censoring indicators are missing at random (MAR).
• To provide reliable statistical inference for these complex scenarios.

Main Methods:

• Augmented inverse probability weighted estimating equations are proposed for parameter estimation.
• Non-missingness and uncensored observation probabilities are estimated using parametric or nonparametric kernel smoothing.
• Asymptotic properties of the developed estimators are theoretically established.

Main Results:

• The proposed augmented inverse probability weighted methods provide consistent parameter estimation.
• Extensive simulation studies demonstrate the finite sample performance of the estimators.
• The methods are applied to real-world brain cancer data, showing practical utility.

Conclusions:

• Generalized MRL models offer a powerful alternative to hazard function-based models.
• The developed statistical methods effectively handle survival data with MAR censoring indicators.
• The approach is validated through simulations and a practical application in medical research.