Nonparametric estimation in a Markov "illness-death" process from interval censored observations with missing...

Area of Science:

• Biostatistics
• Clinical Trial Methodology
• Survival Analysis

Background:

• Clinical trials often involve intermittent patient assessments.
• This leads to interval-censored data and missing status for intermediate transitions.
• Accurate estimation of nonfatal event-free survival time is crucial.

Purpose of the Study:

• To develop nonparametric maximum likelihood (ML) estimation for an illness-death model.
• To address interval-censored observations and missing status of intermediate transitions.
• To provide new methodology for analyzing incomplete clinical trial data.

Main Methods:

• Utilized an "illness-death" model framework.
• Developed nonparametric maximum likelihood (ML) estimation.
• Proposed a self-consistent algorithm for ML estimators.

Main Results:

• Successfully developed ML estimators for the illness-death model with incomplete data.
• Demonstrated that the ML estimators are self-consistent.
• The new methodology was applied to a cancer clinical trial dataset.

Conclusions:

• The developed methodology offers a robust approach for analyzing incomplete data in clinical trials.
• This method improves the estimation of nonfatal event-free survival time.
• The findings provide a valuable alternative to existing conventions for data analysis.