Matching theory and evidence on Covid-19 using a stochastic network SIR model

Area of Science:

• Epidemiology
• Mathematical modeling
• Infectious disease dynamics

Background:

• The COVID-19 pandemic highlighted challenges in accurately tracking infection rates.
• Under-reporting of cases complicates epidemic analysis and intervention effectiveness.

Purpose of the Study:

• To develop an individual-based stochastic network SIR model for COVID-19 analysis.
• To estimate transmission rates and quantify the extent of under-reported COVID-19 cases.
• To analyze the impact of interventions like social distancing and vaccination.

Main Methods:

• Developed an individual-based stochastic network SIR model.
• Derived moment conditions for infected and active cases in single and multigroup models.
• Proposed a joint estimation method for transmission rates and under-reporting magnitude.

Main Results:

• Empirical analysis on six European countries validated the model's simulated outcomes after accounting for under-reporting.
• Estimated actual COVID-19 cases were 4-10 times higher than reported in October 2020, decreasing to 2-3 times by April 2021.
• Counterfactual analyses simulated the effects of social distancing and vaccination on epidemic trajectories in the UK and Germany.

Conclusions:

• Accurate estimation of COVID-19 transmission requires addressing significant under-reporting of cases.
• The developed model provides a robust framework for analyzing epidemic dynamics and evaluating intervention strategies.
• Findings underscore the importance of timely and effective public health interventions in controlling infectious disease outbreaks.