Bayesian Joint Modeling for Longitudinal Magnitude Data With Informative Dropout: An Application to Critical Care...

Area of Science:

• Biostatistics
• Biomedical Data Analysis

Background:

• Biomedical studies often analyze data magnitudes where signs are irrelevant.
• Repeated measures studies require random effects for individual heterogeneity and precise parameter estimation.
• Existing regression methods lack specific approaches for magnitude outcomes with random effects.

Purpose of the Study:

• To introduce Bayesian regression modeling for magnitude data with random effects.
• To extend these models to handle informative dropout using joint modeling.
• To assess the performance and applicability of the proposed methods.

Main Methods:

• Developed Bayesian regression models for magnitude data incorporating random effects.
• Extended models with a joint modeling strategy to address informative dropout.
• Validated methods through two numerical simulation studies.

Main Results:

• The proposed method demonstrated good estimation accuracy for magnitude data.
• Joint models effectively mitigated bias caused by missing data.
• The models were applied to analyze diaphragm thickness changes in ICU patients.

Conclusions:

• Bayesian regression with random effects provides a robust approach for magnitude data analysis.
• Joint modeling successfully addresses informative dropout in repeated measures.
• The developed methods offer valuable tools for biomedical research, including the analysis of patient data.