Estimating the Impact of School Classroom Sizes on the Probability of Severe Acute Respiratory Syndrome Coronavirus-2 Infectivity or Exposure

ABSTRACT

Background. Despite schools reopening across the United States in communities with low and high Coronavirus disease 2019 (COVID-19) prevalence, data remain scarce about the effect of classroom size on the transmission of severe acute respiratory syndrome coronavirus-2 (SARS-COV-2) within schools. This study estimates the effect of classroom size on the risk of COVID-19 infection in a closed classroom cohort for varying age groups locally in Durham, North Carolina. Total number of Coronavirus Disease 2019 (COVID-19) infections over a 28-day follow-up period for varying classroom reproduction number (R0) and varying classroom cohort sizes of 15 students, 30 students and 100 students in Durham County, North Carolina. Methods. Using publicly available population and COVID-19 case count data from Durham County, we calculated a weekly average number of new confirmed COVID-19 cases per week between May 3, 2020 and August 22, 2020 according to age categories < 5 years, 5-9 years, 10-14 years, and 15-19 years. We collated average classroom cohort sizes and enrollment data for each age group by grade level of education for the first month of the 2019-2020 academic school year. Then, using a SEIR compartmental model, we calculated the number of susceptible (S), exposed (E), infectious (I) and recovered (R) students in a cohort size of 15, 30 and 100 students, modelling for classroom reproduction number (R0) of 0.5, 1.5 and 2.5 within a closed classroom cohort over a 14-day and 28-day follow-up period using age group-specific COVID-19 prevalence rates. Results. The SEIR model estimated that the increase in cohort size resulted in up to 5 new COVID-19 infections per 10,000 students whereas the classroom R0 had a stronger effect, with up to 88 new infections per 10,000 students in a closed classroom cohort over time. When comparing different follow-up periods in a closed cohort with R0 of 0.5, we estimated 12 more infected students per 10,000 students over 28 days as compared to 14 days irrespective of cohort size. With a R0 of 2.5, there were 49 more infected students per 10,000 students over 28 days as compared to 14 days. Conclusion. Classroom R0 had a stronger impact in reducing school-based COVID-19 transmission events as compared to cohort size. Additionally, earlier isolation of newly infected students in a closed cohort resulted in fewer new COVID-19 infections within that group. Mitigation strategies should target promoting safe practices within the school setting including early quarantine of newly identified contacts and minimizing COVID-19 community prevalence.