Robust estimation for panel count data with informative observation times and censoring times

Area of Science:

• Biostatistics
• Longitudinal Data Analysis
• Survival Analysis

Background:

• Panel count data analysis is crucial for recurrent events in longitudinal studies.
• Existing methods often rely on restrictive assumptions like independent censoring and Poisson processes.
• These assumptions can be questionable in real-world follow-up studies.

Purpose of the Study:

• To develop new joint models for semiparametric regression of panel count data.
• To address the limitations of existing methods by incorporating informative observation and censoring times.
• To provide a more robust framework for analyzing recurrent event rates.

Main Methods:

• Proposed novel joint models for panel count data.
• Developed semiparametric regression techniques.
• Established asymptotic normality of the proposed estimators.

Main Results:

• The proposed joint models effectively handle informative observation and censoring times.
• Asymptotic normality of the estimators was theoretically established.
• Simulation studies and a real data example demonstrated the advantages of the new method.

Conclusions:

• The developed joint models offer a more flexible and accurate approach to analyzing panel count data.
• This method overcomes the restrictive assumptions of prior techniques.
• The findings have significant implications for the analysis of recurrent events in longitudinal research.