Reconstructing a COVID-19 outbreak within a religious group using social network analysis simulation

Kim, Namje; Kang, Su Jin; Tak, Sangwoo

Objective: We reconstructed the outbreak of the COVID-19 to understand how a large cluster at a church setting progressed before being detected and estimate the potential effectiveness of complying with mask-wearing guidelines recommended by the government. Methods: A mathematical model with social network analysis approach (SNA) was used to simulate a COVID-19 outbreak. A discrete-time, stochastic simulation model was used to construct COVID-19 spread within the SJ Church. A counterfactual experiment using a calibrated baseline model was conducted to examine the potential benefit of complying with mask-wearing policy. Results: Simulations with a mask-wearing percentage of 60.5% (range 55-65%) at the time of the outbreak yielded results most consistent with the confirmed case data. If the church had followed government mask-wearing guidelines, the outbreak might have been one-tenth the size. The counterfactual experiment with 95% mask-wearing estimated an average of 45.6 (95% CI: 43.4-47.9) cases with a standard deviation of 20.1. Discussion: SNA is an effective tool for monitoring and controlling outbreaks of COVID-19 and other infectious diseases. Although our results are based on simulations and thus subject to limitations, the precautionary implication of social distancing and mask-wearing is still relevant. Because person-to-person contact/interaction is unavoidable in social and economic life, it may be wise to consider developing precise measures and guidelines for particular organizations or places that are susceptible to cluster outbreaks.