Simulating the effect of school closure during COVID-19 outbreaks in Ontario, Canada

Abstract

Background: The province of Ontario, Canada, has instituted indefinite school closures (SC) as well as other social distancing measures to mitigate the impact of the novel coronavirus disease 2019 (COVID-19) pandemic. We sought to evaluate the effect of SC on reducing attack rate and the need for critical care during COVID-19 outbreaks, while considering scenarios with concurrent implementation of self-isolation (SI) of symptomatic cases.

Methods: We developed an age-structured agent-based simulation model and parameterized it with the demographics of Ontario stratified by age and the latest estimates of COVID-19 epidemiologic characteristics. Disease transmission was simulated within and between different age groups by considering inter- and intra-group contact patterns. The effect of SC of varying durations on the overall attack rate, magnitude and peak time of the outbreak, and requirement for intensive care unit (ICU) admission in the population was estimated. Secondly, the effect of concurrent community-based voluntary SI of symptomatic COVID-19 cases was assessed.

Results: SC reduced attack rates in the range of 7.2-12.7% when the duration of SC increased from 3 to 16 weeks, when contacts among school children were restricted by 60-80%, and in the absence of SI by mildly symptomatic persons. Depending on the scenario, the overall reduction in ICU admissions attributed to SC throughout the outbreak ranged from 3.3 to 6.7%. When SI of mildly symptomatic persons was included and practiced by 20%, the reduction of attack rate and ICU admissions exceeded 6.3% and 9.1% (on average), respectively, in the corresponding scenarios.

Conclusion: Our results indicate that SC may have limited impact on reducing the burden of COVID-19 without measures to interrupt the chain of transmission during both pre-symptomatic and symptomatic stages. While highlighting the importance of SI, our findings indicate the need for better understanding of the epidemiologic characteristics of emerging diseases on the effectiveness of social distancing measures.

Keywords: COVID-19; Pandemic; School closure; Self-isolation; Simulation; Social distancing.

Fig. 1

Projected daily incidence of COVID-19 cases per 10,000 populations during outbreaks in Ontario for different proportions of SI, concurrent with SC of varying durations. Top panels (a, b, c, d) and bottom panels (e, f, g, h) illustrate 60% and 80% reduction of daily contacts among school children aged 5 to 19 years. Color curves correspond to 3 weeks (black), 6 weeks (brown), 12 weeks (blue), and 16 weeks (cyan) duration of SC. Y-axis represents the daily incidence of infection and X-axis represents time in 30-day increments

Fig. 2

Projected attack rates of COVID-19 per 10,000 population during the outbreak in Ontario for different proportions of SI by symptomatic persons with mild illness, concurrent with SC of varying durations. Color plots correspond to 3 weeks (black), 6 weeks (brown), 12 weeks (blue), and 16 weeks (cyan) duration of SC

Fig. 3

Projected total ICU admissions of COVID-19 patients per 10,000 population during outbreak in Ontario for different proportion of SI by symptomatic persons with mild, concurrent with SC of varying durations. Color bars correspond to 3 weeks (black), 6 weeks (brown), 12 weeks (blue), and 16 weeks (cyan) durations of SC