Characterising COVID-19 school and childcare outbreaks in Canada in 2021: a surveillance study

Introduction

Beginning in January 2021, the Public Health Agency of Canada (PHAC) began COVID-19 outbreak surveillance in collaboration with provinces and territories (PTs).1 The responsibility for delivering health services, including public health services, lies with provincial and territorial governments; federal public health surveillance relies on PT participation. The Canadian COVID-19 Outbreak Surveillance System (CCOSS) aimed to monitor the frequency and severity (hospitalisation and deaths) of outbreak events by setting, using data contributed by eight PTs.2

Children and youth were identified as a population of interest for SARS-CoV-2 infection during development and implementation of CCOSS. There were concerns that SARS-CoV-2 may lead to increased hospitalisation for this population, as those aged younger than 5 years are susceptible to other respiratory illnesses such as influenza.3 Furthermore, children can play a role in the transmission chain of influenza, as they experience high attack rates, prolonged viral shedding and have very dense social networks characterised by close contacts.4 5

A report of the special task force led by the Chief Science Advisor of Canada on COVID-19 and Children, published in July 2020, found that many science gaps remain for fully understanding COVID-19 in children that should be urgently addressed through research and systematic data gathering. COVID-19 outbreak surveillance in school settings aims to leverage the collection of better and broader data, including prevalence studies and systematic observations in day camps, daycare facilities and schools.6

School and childcare settings are integral to communities; they provide a safe learning environment for children and employment for teachers and staff.7 Certain schools and childcare programmes provide additional services to children, including meal programmes, mental health services, and social and physical activities.7–9 In the USA, reports suggest that school closures in an attempt to slow the spread of COVID-19 during the spring of 2020, may have had a negative effect on children’s learning, as well as the emotional and mental health of children and parents.7 Further, school and childcare setting closures disproportionately impacted women working from home compared with men, as women were more likely to absorb additional housework, educational support and childcare responsibilities.10 11 Additional childcare responsibilities can also impact the workforce; a study by Bayham and Fenichel found that deaths could increase with school and childcare closures due to healthcare workforce declines.12 The risk of SARS-CoV-2 spread in schools and childcare settings should be balanced against the benefits of in-person learning, caregiving needs and workforce capacity.7 12

Early in the pandemic (2020), other countries, such as Germany, analysed their surveillance data to better understand outbreaks in schools.13 Between March and August 2020, most school outbreaks were small in Germany, with few cases per outbreak; cases were primarily among older age groups who may have been staff members or other adults linked to the school outbreak.13

This paper describes the temporal outbreak trends in school and childcare settings in Canada between January 2021 and December 2021, including comparisons with overall case incidence, differences between childcare settings and types of schools (ie, primary, secondary, etc), age distribution of outbreak-associated cases, and outbreak size and severity (hospitalisation and deaths). We also review temporal relationships between public health interventions such as school closures and vaccination and outbreak incidence.

Methods

CCOSS was implemented in January 2021 and the surveillance period for this analysis includes outbreaks declared between 3 January 2021 and 1 January 2022, as data were aggregated by epidemiological week (Sunday–Saturday).

Both case and outbreak data sources used for analyses are maintained by PHAC and housed in a Postgres database. CCOSS is an event-based surveillance system that captures outbreak-level data on administrative variables (eg, outbreak identifier, date outbreak declared, date outbreak declared over), 25 outbreak settings, number of confirmed cases and outcomes (hospitalisation, death). The national COVID-19 case surveillance system captures case-level data on demographics, clinical status and outcome, risk factors, exposures, vaccination status and variant lineage. Some jurisdictions have provided permission to have their case data linked to outbreak data to provide additional information on demographics, vaccination status and variant lineages of outbreak-associated cases.

Five PTs contributing data to CCOSS submitted outbreaks in school and childcare settings, representing 76% of the Canadian population (Ontario, Quebec, Alberta, Nova Scotia and Prince Edward Island). Of those five, three permit linkages to the case data to obtain additional information, such as age demographics.

The focus of this analysis was on outbreak trends in paediatric childcare and education settings (definitions below). Outbreaks in post-secondary, adult education and ‘other’ or unspecified school settings were excluded from analyses. Cases associated with these settings also include staff, volunteers and other individuals aged ≥18 years.

• Primary school: schools with grades kindergarten–grade 8 (including middle schools).

• Secondary school: schools with grades 9–12.

• K-12: schools with grades kindergarten–grade 12.

• Childcare: home daycare, school daycare, external (privately and publicly funded) daycare, day/overnight camps.

The national definition for COVID-19 outbreaks was applied to the dataset, and outbreaks with fewer than two cases or missing case counts were excluded. Duplicate outbreaks were removed using the unique outbreak identifier.

Two or more confirmed cases of COVID-19 epidemiologically linked to a specific setting and/or location. Excluding households since household cases may not be declared or managed as an outbreak if the risk of transmission is contained. This definition also excludes cases that are geographically clustered (eg, in a region, city or town) but not epidemiologically linked, and cases attributed to community transmission.14

One jurisdiction changed their outbreak definition in October 2021 to only include outbreaks with a minimum of 10 cases.

Data were cleaned and analysed using R statistical software (V.4.0.4).15 Epidemiological curves were generated comparing case incidence with outbreak incidence. Trends in outbreak setting detail and cases’ age distribution over time were examined to assess for any correlation with vaccination implementation in different age groups. Descriptive statistics on outbreak size and proportionate representation by outbreak setting detail were computed to describe the outbreaks.

Results

School outbreaks and community transmission

School and childcare centre outbreak incidence trends align temporally with trends observed in COVID-19 case data (figure 1). In 2021, 367 914 cases were reported among those 0–19 years of age (24.36% of all cases). Prior to the introduction of the Omicron variant in November 2021, community case incidence rose and reached its peak prior to an increase and peak in outbreak incidence in school and childcare settings (figure 1).

Request permission

Figure 1

Weekly COVID-19 case incidence, including outbreak-associated and non-outbreak-associated cases (A), and COVID-19 outbreak incidence in school and childcare settings (B) from January to December 2021.

During the Omicron wave (beginning 5 December 2021), we also observed a rise in cases prior to outbreaks; however, it appears the increase in outbreaks rose faster than cases in fall 2021 (see online supplemental figure 1).

School outbreaks and vaccination

Between January and June 2021, the number of outbreaks was consistently distributed across school and childcare setting and age groups (figure 2). On 5 May 2021, Health Canada approved the first COVID-19 vaccine for individuals aged 12–17 years.19 During the start of the 2021–2022 school year (September 2021), there were fewer outbreaks in secondary schools and fewer cases among those aged 12–17 and 18+ years (ie, vaccine-eligible age groups) compared with January–June of 2021. In December 2021, with the emergence of Omicron, a variant with vaccine escape,20 there was an increase in the number of cases aged 12+ years; however, they continued to make up a small proportion of cases compared with younger age groups not eligible for full vaccination. Since the start of the 2021–2022 school year, primary schools and childcare settings account for the highest proportion of outbreaks (figure 2). Children aged 5–11 years have been disproportionately represented in school and childcare outbreaks (figure 2). Beginning in September 2021, there was consistent increase in outbreaks in primary schools and cases among those aged 5–11 years until school closures for the holiday season in mid-December 2021. Overall, we observed a decrease in the number of outbreaks in childcare settings in fall 2021 compared with the January–June 2021 period.

Request permission

Figure 2

Weekly outbreak incidence in school settings (A) and weekly school outbreak-associated cases by age group (B) from January to December 2021.

Size of school outbreaks

Most outbreaks in school and childcare settings reported fewer than 10 cases (figure 3); less than 4% of outbreaks had 25 or more cases. Overall, the mean case count per outbreak was six individuals and the median was four individuals. Both mean and median outbreak size varied over time and increased slightly during the Omicron wave, but this increase was not significant during the study period, which only included the emergence of Omicron (figure 4). The size of the largest outbreak reported in each setting was 236 cases in a primary school, 144 cases in a K-12 school, 132 cases in a secondary school and 69 cases in a childcare setting. The overall size of an outbreak is dependent on the size of the population at risk (eg, childcare settings tend to have smaller numbers of individuals compared with school settings). Childcare settings had the smallest proportion of outbreaks with more than 10 cases, while K-12 settings had the highest proportion of outbreaks with more than 10 cases (>25%) (figure 3). The largest outbreaks across all childcare and school settings were associated with variants of concern (VOCs) (Alpha, Delta and Omicron); however, most cases reported to CCOSS across all settings were associated with VOCs during the study period in 2021.

Request permission

Figure 3

Proportion of COVID-19 outbreak size by setting (A) and number of outbreak-associated cases by setting (B) from January to December 2021.

Request permission

Figure 4

Mean and median outbreak size in school and childcare settings over time between January and December 2021.

Discussion and policy implications

Although there were a notable number of outbreaks in school and childcare settings of COVID-19 in Canada during 2021, the analyses presented here illustrate that these outbreaks were small, with lower severity (hospitalisation and deaths) compared with other outbreak settings and followed increases in community transmission.2 Severity (proportion hospitalised and deceased) was lower in schools compared with industrial/agricultural settings, congregate living settings, long-term care and other community settings.2 Median outbreak size was also smaller in schools and childcare settings compared with congregate living, long-term care, acute care and correctional facilities.2 Although majority of outbreaks occurred in primary schools, it should be noted that there are 3.8 times more primary schools in Canada than secondary schools.21

These findings align with studies conducted globally, whereby outbreaks in school settings documented limited transmission. Other studies have associated low transmission with the implementation of public health measures22–24; further, when schools have reopened, they have not been associated with increases in community transmission.25 26 Even when cases have increased in communities following school openings, it is challenging to disentangle whether increases in school outbreak incidence may have led to an increase in community incidence, as workplace mobility patterns also play a role in community incidence.27 In Ontario, workplace mobility patterns showed a sharp increase in workplace mobility following the reopening of schools, likely caused by parents and school staff being able to return to work—which may also increase community transmission.27 In-person learning, as well as working outside the home, may both contribute to an increased risk of transmission into the home. A study in New York City found that the secondary attack rate across public schools was 0.7% from October 2021 to December 2021.28 Most studies have observed that children were most likely to contract COVID-19 from family members.9 25 26 29 While transmission potential is high among children, systematic reviews have found less than 10% of index cases in a household were children.30 The Lancet COVID-19 Commission has stated that, globally, schools are able to stay open without significantly increasing the risk to students and adults, and highlighted the harms associated with school closures, stating that ‘in-person learning should be prioritised’.31

During the observation period of this study, vaccines available were effective in reducing infection against the circulating variants (eg, Alpha, Beta, Gamma, Delta),32 and Canada had a high vaccination rate among adults.33 Teachers were considered essential workers and prioritised for vaccination; however, timelines for eligibility varied by PT.34 Studies on school outbreaks have surmised that even when outbreaks occur in schools, onward transmission may be controlled if other household members were vaccinated.35 Declines in outbreaks were observed among the vaccine-eligible populations following the implementation of vaccine mandates in fall 2021. While the overall number of outbreaks increased, this disproportionately impacted primary schools and children aged 5–11 years (restricted eligibility for vaccination). Although we were unable to assess the vaccine status of individuals associated with these outbreaks, a modelling study found that vaccination significantly reduced the risk and size of school outbreaks36 and a systematic review found that vaccination was effective at protecting individuals against infection for variants prior to Omicron.32 As vaccine effectiveness against infection has waned and transmission has increased with the Omicron variants,32 these protective features may be less relevant, and the risk of school outbreaks and community case incidence could increase. However, vaccination continued to be highly protective against hospitalisation and death from pre-Delta, Delta and Omicron variants.37 38

Conclusions

The results presented here are the first to examine COVID-19 outbreaks in school and childcare settings over time in Canada. Based on the five jurisdictions that submitted school outbreaks to CCOSS in 2021, most (42%) occurred in primary schools. Although there were many school outbreaks in 2021, they were small, with the majority reporting fewer than 10 cases per outbreak. Additional research and analyses to assess transmission dynamics, and calculate attack rates, differences in regional community transmission and public health interventions would further inform these results. Children remain a population of interest for SARS-CoV-2; however, severity (hospitalisation and deaths) in paediatric populations has remained low throughout the pandemic.