Modelling time varying heterogeneity in recurrent infection processes: an application to serological data

Area of Science:

• Epidemiology
• Biostatistics
• Survival Analysis

Background:

• Frailty models analyze multivariate time-to-event data, crucial for infectious disease epidemiology.
• Existing models often assume lifelong immunity, but refinements address reinfections.
• Previous work quantified misspecification effects and applied time-varying frailty models.

Purpose of the Study:

• To extend existing frailty methodology to incorporate age-dependence in individual heterogeneity.
• To develop and apply age-dependent shared and correlated gamma frailty models.
• To analyze time-to-event data in infectious disease epidemiology with enhanced individual heterogeneity.

Main Methods:

• Extension of frailty methodology proposed by Abrams and Hens.
• Implementation of age-dependent shared and correlated gamma frailty models.
• Application of the methodology to two real-world data sets.

Main Results:

• The proposed age-dependent frailty models successfully account for time-varying individual heterogeneity.
• Demonstrated improved modeling of infection acquisition and association with age.
• Methodology validated through practical data applications.

Conclusions:

• Age-dependent frailty models offer a more realistic approach to modeling infectious disease dynamics.
• These models enhance the understanding of individual heterogeneity in infection risk over time.
• The developed methodology provides a valuable tool for infectious disease epidemiologists.