Simulation-Based Analysis of COVID-19 Spread Through Classroom Transmission on a University Campus

Abstract

Airborne transmission is now believed to be the primary way that COVID-19 spreads. We study the airborne transmission risk associated with holding in-person classes on university campuses. We utilize a model for airborne transmission risk in an enclosed room that considers the air change rate for the room, mask efficiency, initial infection probability of the occupants, and also the activity level of the occupants. We introduce, and use for our evaluations, a metric $R_0^{eff}$ that represents the ratio of new infections that occur over a week due to classroom interactions to the number of infected individuals at the beginning of the week. This can be seen as a surrogate for the well-known $R_0$ reproductive number metric, but limited in scope to classroom interactions and calculated on a weekly basis. The simulations take into account the possibility of repeated in-classroom interactions between students throughout the week. We presented model predictions were generated using Fall 2019 and Fall 2020 course registration data at a large US university, allowing us to evaluate the difference in transmission risk between in-person and hybrid programs. We quantify the impact of parameters such as reduced occupancy levels and mask efficacy. Our simulations indicate that universal mask usage results in an approximately $3.6\times$ reduction in new infections through classroom interactions. Moving 90% of the classes online leads to about $18\times$ reduction in new cases. Reducing class occupancy to 20%, by having hybrid classes, results in an approximately $2.15-2.3\times$ further reduction in new infections.