A comparative study of prevalence-based incidence estimation techniques with application to dementia data in Germany

Area of Science:

• Epidemiology
• Biostatistics
• Mathematical Modeling

Background:

• Dementia presents a growing global health challenge, exacerbated by increasing life expectancy.
• Multistate models are essential for tracking disease progression, particularly for irreversible conditions.
• Understanding disease dynamics requires accurate estimation of incidence, prevalence, and mortality rates.

Purpose of the Study:

• To evaluate and compare different methods for estimating age-specific incidence of irreversible diseases.
• To assess the performance of an ordinary differential equation (ODE) based technique against alternative methods.
• To apply and validate these estimation techniques using German dementia data.

Main Methods:

• Utilized a simplified illness-death multistate model (healthy, diseased, dead).
• Conducted simulation studies to compare numerical approximations of an ODE model with three other estimation techniques.
• Applied the methods to cross-sectional data on dementia prevalence in Germany.

Main Results:

• The ordinary differential equation (ODE) method proved to be the most effective technique for incidence estimation.
• The ODE approach showed superior performance in both simulated scenarios and the analysis of dementia data.
• Numerical approximations of the ODE model were evaluated for their accuracy.

Conclusions:

• The ODE-based method is a predominant and high-quality technique for estimating age-specific incidence of irreversible diseases.
• This approach offers significant advantages for analyzing epidemiological data, such as dementia.
• Accurate incidence estimation is crucial for managing the public health burden of diseases like dementia.