The novel coronavirus SARS-CoV-2 emerged in 2019 and subsequently spread throughout the world, causing over 600 million cases and 6 million deaths as of September 7th, 2022. Superspreading events (SSEs), defined here as public or social events that result in multiple infections over a short time span, have contributed to SARS-CoV-2 spread. In this work, we compare the dynamics of SSE-dominated SARS-CoV-2 outbreaks, defined here as outbreaks with relatively higher SSE rates, to the dynamics of non-SSE-dominated SARS-CoV-2 outbreaks. To accomplish this, we derive a continuous-time Markov chain (CTMC) SARS-CoV-2 model from an ordinary differential equation (ODE) SARS-CoV-2 model and incorporate SSEs using an events-based framework. We simulate our model under multiple scenarios using Gillespie's direct algorithm. The first scenario excludes hospitalization and quarantine; the second scenario includes hospitalization, quarantine, premature hospital discharge, and quarantine violation; and the third scenario includes hospitalization and quarantine but excludes premature hospital discharge and quarantine violation. We also vary quarantine violation rates. Results indicate that, with either no control or imperfect control, SSE-dominated outbreaks are more variable but less severe than non-SSE-dominated outbreaks, though the most severe SSE-dominated outbreaks are more severe than the most severe non-SSE-dominated outbreaks. We measure severity by the time it takes for 50 active infections to be achieved; more severe outbreaks do so more quickly. SSE-dominated outbreaks are also more sensitive to control measures, with premature hospital discharge and quarantine violation substantially reducing control measure effectiveness.
Keywords: Continuous-time Markov chain; Gillespie’s direct algorithm; Human behavior; SARS-CoV-2; Superspreading events.
Copyright © 2022. Published by Elsevier Ltd.