Managing school interaction networks during the COVID-19 pandemic: Agent-based modeling for evaluating possible scenarios when students go back to classrooms.

Abstract

The most unexpected and toughest phenomenon that has occurred in recent times is the global COVID-19 pandemic. One of the first measures to prevent the spread of the disease was to close educational institutions. The students were forced to start a learning process through social networks and web platforms. In some countries, a return to face-to-face classes was established. However, weeks later, some of them had to return to virtual activities due to an upswing in the COVID-19 cases. In Mexico, classes have been held virtually, with face-to-face activities only re-established in two of the 32 states. In our state, Yucatan, scholarly activities are still virtual. In this work, the dispersion of COVID-19 at different academic establishments in Yucatan was simulated. Networks of Friendship, noncordial treatment, family ties and study groups were considered. Based on these networks, we evaluated the possibility of returning to school without inducing a rebound in the COVID-19 cases in the state. Agent-based simulations were used, with each student as an agent. Interaction rules were established based on international research regarding good practices in times of COVID-19. We used seven networks from different academic institutions, ranging from primary through college level. As a result, possible contagion curves were obtained for different scenarios, which leads to a discussion about the measures that would be relevant once a return to face-to-face classes is overseen. Simulations show that isolating students and reducing the number of students in the same classroom are good strategies and substantially reduce the possible contagiousness.