Shared parameter models for the joint analysis of longitudinal data and event times.

Area of Science:

• Biostatistics
• Longitudinal Data Analysis
• Survival Analysis

Background:

• Longitudinal studies frequently collect joint data on event times and serial outcomes.
• Analyzing these datasets requires methods that can account for non-ignorable dropout and associations between trends and event times.

Purpose of the Study:

• To present shared parameter models for jointly analyzing repeated measurements and event time data.
• To estimate and compare serial trends over time while adjusting for informative censoring due to patient dropout.

Main Methods:

• Utilized parametric and semi-parametric survival models for event times.
• Employed generalized linear or non-linear mixed-effects models for repeated measurements.
• Developed estimation methods based on a generalized non-linear mixed-effects model compatible with existing software.

Main Results:

• The proposed shared parameter models offer flexible analysis of event time distributions and longitudinal response variable relationships.
• Demonstrated the model's utility in a multi-center study on renal disease progression, diet, and blood pressure control.

Conclusions:

• Shared parameter models provide a robust framework for joint analysis of longitudinal and event time data.
• This approach facilitates accurate estimation of trends and understanding of factors influencing event times in clinical studies.