A double-semiparametric approach for extending mixture cure models with interval-censored data

Area of Science:

• Biostatistics
• Survival Analysis
• Epidemiology

Background:

• Mixture cure models (MCMs) analyze failure time data with a cured subgroup.
• Traditional MCMs often assume linear effects or parametric forms for cure probability, which can be limiting.
• Unobserved cure status in right-censored data poses challenges for standard models.

Purpose of the Study:

• To develop a more general mixture cure model for interval-censored data.
• To incorporate nonparametric covariate effects into the incidence component of MCMs.
• To allow semiparametric frameworks for both latency and incidence components for greater flexibility.

Main Methods:

• Developed a spline-based sieve maximum likelihood estimator.
• The estimator handles both model parameters and unknown functions.
• Established desirable asymptotic properties for the proposed estimator.

Main Results:

• The proposed method allows for flexible, nonparametric modeling of covariate effects on cure probability.
• Demonstrated desirable asymptotic properties of the spline-based sieve maximum likelihood estimator.
• Validated the method's performance through simulation and a real-world cardiac allograft vasculopathy study.

Conclusions:

• The new mixture cure model offers enhanced flexibility for interval-censored survival data.
• The method effectively captures complex relationships between risk factors and cure status.
• Provides a valuable tool for analyzing diseases with a cured fraction, improving understanding of disease progression and risk.