SARS-CoV-2 Transmission in University Classes

We investigate transmission dynamics for SARS-CoV-2 on a real network of classes at Simon Fraser University, a medium-sized school in Western Canada. Outbreaks are simulated over the course of one semester across numerous parameter settings for a realistic compartment model, including asymptomatic and presymptomatic transmission. We investigate the control strategy of moving large classes online while small classes are allowed to meet in person. Regression trees are used to model the effect of disease parameters on simulation outputs; specifically, the total number of infections and the peak number of simultaneous cases. We find that an aggressive class size thresolding strategy is required to mitigate the risk of a large outbreak, and that transmission by symptomatic individuals is a key driver of outbreak size,