Controlling multiple COVID-19 epidemic waves: an insight from a multi-scale model linking the behavior change dynamics to the disease transmission dynamics

Abstract

COVID-19 epidemics exhibited multiple waves regionally and globally since 2020. It is important to understand the insight and underlying mechanisms of the multiple waves of COVID-19 epidemics in order to design more efficient non-pharmaceutical interventions (NPIs) and vaccination strategies to prevent future waves. We propose a multi-scale model by linking the behavior change dynamics to the disease transmission dynamics to investigate the effect of behavior dynamics on COVID-19 epidemics using the game theory. The proposed multi-scale model was calibrated and key parameters related to disease transmission dynamics and behavioral dynamics with/without vaccination were estimated based on COVID-19 epidemic data and vaccination data. Our modeling results demonstrate that the feedback loop between behavior changes and COVID-19 transmission dynamics plays an essential role in inducing multiple epidemic waves. We find that the long period of high-prevalence or persistent deterioration of COVID-19 epidemics could drive almost all population to change their behaviors and maintain the altered behaviors, however, the effect of behavior changes faded out gradually along the progress of epidemics. This suggests that it is essential not only to have persistent, but also effective behavior changes in order to avoid subsequent epidemic waves. In addition, our model also suggests the importance to maintain the effective altered behaviors during the initial stage of vaccination, and to counteract relaxation of NPIs, it requires quick and massive vaccination to avoid future epidemic waves.