Likelihood Inference for Unified Transformation Cure Model with Interval Censored Data

Area of Science:

• Biostatistics
• Survival Analysis
• Statistical Modeling

Background:

• Cure rate models are essential for analyzing data where a portion of the population may never experience the event of interest.
• Existing models often struggle with interval-censored data, which is common in longitudinal studies.
• The Box-Cox transformation cure rate model (BCT) offers a flexible framework for survival data.

Purpose of the Study:

• To extend the unified Box-Cox transformation cure rate models (BCT) to effectively handle interval-censored data.
• To develop robust statistical inference methods for these extended models.
• To evaluate the performance of the proposed methods using simulations and a real-world dataset.

Main Methods:

• Development of likelihood inference using the Expectation-Maximization (EM) algorithm for the BCT cure model.
• Estimation of the BCT transformation parameter using simultaneous maximization and profile likelihood within the EM framework.
• Monte Carlo simulations to assess bias, root mean square error, and confidence interval coverage probability.

Main Results:

• The proposed EM algorithm demonstrates effective estimation for BCT cure models with interval-censored data.
• Simulation results indicate good performance in terms of bias, RMSE, and coverage probability.
• The EM algorithm's efficacy is comparable or superior to direct maximization methods.

Conclusions:

• The extended BCT cure rate model provides a powerful tool for analyzing interval-censored survival data.
• The EM algorithm offers a reliable and efficient method for parameter estimation in these models.
• The approach is validated through a practical application in a smoking cessation study.