Design of panel studies for disease progression with multiple stages

Area of Science:

• Biostatistics
• Epidemiology
• Longitudinal Studies

Background:

• Panel studies collect data from individuals at periodic follow-up visits.
• Understanding disease progression requires analyzing data collected over time.
• Transition intensities are key metrics for modeling disease stage changes.

Purpose of the Study:

• To investigate critical design factors in panel studies for disease staging.
• To analyze the impact of time intervals between follow-up visits on transition intensity estimation.
• To explore sample size considerations and statistical power for analyzing disease progression.

Main Methods:

• Simulation studies to evaluate the effect of varying time intervals.
• Analysis of transition intensity estimators' precision.
• Power calculations for statistical tests on intensity ratios.
• Incorporation of frailty models to account for population heterogeneity.

Main Results:

• The choice of time interval significantly impacts the precision of transition intensity estimators.
• Shorter intervals generally increase precision but may affect feasibility.
• Statistical power for detecting differences in transition intensities is sensitive to interval length.
• Frailty models can effectively capture subject-specific disease progression rates.

Conclusions:

• Optimizing the time interval between follow-up visits is crucial for accurate disease staging analysis in panel studies.
• Careful consideration of sample size and interval length is necessary to ensure adequate statistical power.
• Frailty models provide a robust framework for analyzing heterogeneous disease progression in panel data.