How will COVID-19 persist in the future? Simulating future dynamics of COVID-19 using an agent-based network model

Despite the United States Center for Disease Control (CDC)s May 2023 expiration of the declared public health emergency pertaining to the COVID-19 pandemic (Silk 2023), approximately 3 years after the first cases of SARS-CoV-2 appeared in the United Sates, thousands of new cases persist daily. Many questions persist about the future dynamics of SARS-CoV-2s in the United States, including: will COVID continue to circulate as a seasonal disease like influenza, and will annual vaccinations be required to prevent outbreaks? In response, we present an Agent Based Networked Simulation of COVID-19 transmission to evaluate recurrent future outbreaks of the disease, accounting for contact heterogeneity and waning vaccine-derived and natural immunity. Our model is parameterized with data collected as part of the Berkeley Interpersonal Contact Survey (BICS; Feehan and Mahmud 2021) and is used to simulate time series of confirmed cases of and deaths due to SARS-CoV-2, paying special attention to seasonal forces and waning immunity (Kronfeld-Schor et al. 2021; X. Liu et al. 2021; Nichols et al. 2021). From the BICS ABM model we simulate SARS-CoV-2 dynamics over the 10-year period beginning in 2021 with waning immunity and inclusion of annual booster doses under a variety of transmission scenarios. We find that SARS-CoV-2 outbreaks are likely to occur frequently, and that distribution of booster doses during certain times of the year--notably in the late winter/early spring--may reduce the severity of a wintertime outbreak depending on the seasonal epidemiology of the pathogen.