Analysing the distribution of SARS-CoV-2 infections in schools: comparing model predictions with real-world observations

Abstract School closures were used to mitigate transmission in the SARS-CoV-2 pandemic. Understanding the nature of SARS-CoV-2 outbreaks in classrooms could help inform targeted, precision preventive measures and outbreak management in schools, in response to future pandemics. Infection probability distributions establish the salient features of disease dynamics, yet very few studies have attempted to model it ab initio; a key problem being systematically accounting for the high variability of the governing parameters. In this study, possibly for the first time, we analytically derive the probability density function (PDF) of SARS-CoV-2 secondary infections accounting for major sources of variability in airborne transmission like viral load, dose–response, occupancy and compare it with real-world infection distributions from reported cases across public schools in Ontario, Canada. The model output showcases a robust quantitative match with the data while demonstrating the intrinsic overdispersed nature of SARS-CoV-2 infections and their mechanistic underpinnings. The results quantify the importance of long-range transmission in triggering superspreading events, whereas short-range transmission engenders a more frequent but smaller number of secondary infections. This study provides a fundamental understanding of the overdispersed nature of school outbreaks along with a robust method to predict outbreak size in indoor environments, which could inform focused mitigation strategies.