Alternative Strategies for the Estimation of a Disease's Basic Reproduction Number: A Model-Agnostic Study

Area of Science:

• Epidemiology
• Mathematical Modeling
• Infectious Disease Dynamics

Background:

• The basic reproduction number (R0) is crucial for understanding disease spread.
• Accurate estimation of R0 is vital for effective public health interventions.
• Various modeling approaches exist, each with theoretical and practical considerations.

Purpose of the Study:

• To conduct a model-agnostic evaluation of four distinct methods for estimating the basic reproduction number.
• To compare the theoretical underpinnings and simulation-based performance of these R0 estimation models.
• To apply the evaluated models to a real-world case study of the COVID-19 outbreak in Colombia.

Main Methods:

• Theoretical review and comparison of four R0 estimation models: a Bayesian method (Thompson et al.), a state-space method, and two Poisson-based simulation approaches.
• Impartial testing of each model using eight distinct simulation scenarios designed to meet model assumptions.
• Application of the models to analyze the COVID-19 outbreak evolution in Colombia.

Main Results:

• The study provides a comparative analysis of the strengths and weaknesses of each model from both theoretical and practical perspectives.
• Simulation results offer insights into the performance and reliability of different R0 estimation techniques under controlled conditions.
• The Colombian COVID-19 case study demonstrates the practical utility and challenges of applying these models in real-world epidemiological scenarios.

Conclusions:

• The evaluation highlights the trade-offs associated with different basic reproduction number estimation models.
• Understanding model assumptions and limitations is critical for accurate epidemiological assessments.
• This comparative framework aids in selecting appropriate R0 estimation methods for disease surveillance and control.