Analysis of recurrent event data with incomplete observation gaps

Area of Science:

• Biostatistics
• Survival Analysis
• Longitudinal Data Analysis

Background:

• Recurrent event data analysis requires methods to handle terminating events.
• Intermittent dropouts create observation gaps, complicating risk status assessment.
• Standard risk variable definitions are inadequate when observation gaps are incomplete.

Purpose of the Study:

• To propose a statistical method for analyzing recurrent event data with intermittent dropouts and incomplete observation gap information.
• To model incomplete observation gap information using an interval-censored mechanism.
• To apply the proposed method to real-world data from the Young Traffic Offenders Program.

Main Methods:

• Development of a statistical framework for recurrent event data with intermittent dropouts.
• Utilizing an interval-censored mechanism to model incomplete observation gap information.
• Application and validation of the method on conviction rates in a study of young traffic offenders.

Main Results:

• The proposed interval-censored method effectively accommodates intermittent dropouts in recurrent event data analysis.
• Valid inference on risk status is achieved despite incomplete information on observation gaps.
• The method demonstrates practical utility in analyzing conviction rates with study suspensions.

Conclusions:

• The developed method provides a robust approach for recurrent event data analysis when intermittent dropouts occur.
• Accurate risk inference is possible even with interval-censored observation gap data.
• This methodology enhances the analysis of longitudinal studies involving complex event patterns and missing data.