Measuring differences between phenomenological growth models applied to epidemiology.

Area of Science:

• Epidemiology
• Mathematical Modeling
• Computational Biology

Background:

• Phenomenological growth models (PGMs) are crucial for understanding epidemic trajectories and forecasting.
• These models, often expressed as ordinary differential equations (ODEs), require few parameters but vary in their data-fitting capabilities.
• The COVID-19 pandemic highlights the need for effective epidemic modeling.

Purpose of the Study:

• To systematically study differences in dynamics between various PGMs.
• To explain why certain PGMs outperform others in fitting epidemic data.
• To introduce a novel metric for quantifying these dynamic differences.

Main Methods:

• Defined an empirical directed distance (EDD) to measure dynamic differences between PGMs.
• The EDD quantifies how well one PGM fits data generated by another.
• Applied EDD calculations to synthetic data and real-world epidemic data (influenza, Ebola, COVID-19).

Main Results:

• Demonstrated that PGMs exhibit distinct dynamic behaviors, even with similar parameter counts.
• The EDD metric effectively quantifies these dynamic differences.
• The study provides a framework for comparing and selecting appropriate PGMs for specific outbreaks.

Conclusions:

• The EDD offers a valuable tool for assessing PGM performance and improving epidemic forecasting.
• Understanding dynamic differences between models is key to selecting the best fit for real-world epidemic data.
• This research contributes to more robust epidemiological modeling strategies.