A joint overdispersed marginalized random-effects model for analyzing two or more longitudinal ordinal responses

Area of Science:

• Biostatistics
• Rheumatology
• Medical Statistics

Background:

• Disease severity requires objective measurement for accurate assessment.
• The Medsger scale is a validated tool for evaluating systemic sclerosis severity across organ systems.
• Longitudinal data analysis is crucial for tracking disease progression.

Purpose of the Study:

• To develop and validate a statistical model for analyzing correlated longitudinal ordinal data.
• To assess systemic sclerosis severity using the general and skin system components of the Medsger scale.
• To account for overdispersion and temporal correlations in patient data.

Main Methods:

• Utilized an overdispersed marginalized random-effects model for correlated ordinal responses.
• Employed a random-effects approach to handle intra-subject correlations and temporal dependencies.
• Investigated statistical properties of estimators via extensive simulations.

Main Results:

• The proposed joint model effectively handles overdispersion and correlated longitudinal ordinal data.
• Demonstrated the marginal interpretability of certain model parameters.
• The methodology proved reliable in analyzing systemic sclerosis patient data.

Conclusions:

• The developed statistical model provides a robust framework for assessing disease severity, specifically for systemic sclerosis.
• This approach enhances the reliability of longitudinal disease assessment by accounting for complex data structures.
• The findings support the use of advanced statistical methods in clinical research for better patient management.