In England, 1% of nurseries reported a COVID-19 outbreak to PHE over a three-month period encompassing a rapid spread of the Alpha variant, which is associated with increased transmissibility,27 more severe disease and higher case fatality.28 Children attending nurseries reporting an outbreak had very low infection rates compared to staff, irrespective of whether the index case was a child or staff. Consistent with community trends in < 5 year-olds, incidence rates were highest in infants and then declined with age. In contrast, infection rates were up to 10-fold higher in staff, especially when the index case was another staff member. Nationally, although the Alpha variant was first identified in September 2020, the rapid spread due to this variant during December 2021 was responsible for a third national lockdown in England, which included restricted primary and secondary school attendance for children of keyworkers and vulnerable children in January 2021.26 Nationally, cases in < 5 year-olds remained low throughout the period and followed trends in older age-groups, increasing during December 2020 and then declining during January and February 2021. While the number of nursery outbreaks were similar in November 2020 and January 2021, we found some evidence of larger outbreak sizes and higher incidence rates in January 2021, when the Alpha variant predominated. Outbreaks reported in nurseries declined from mid-February to the end of April 2021.
Between 31 August and 18 October 2020, when all education settings were fully open in England, only 0.3% (87/32,852) nurseries reported a COVID-19 outbreak compared to 2% (450/18,943) primary schools and 10% (519/5,409) secondary schools. This compares with 1% (324/32,852), 4% (684/18,943) and 12% (646/5,409), respectively, between 02 November 2020 and 31 January 2021, which included a period when primary and secondary schools had restricted attendance for children of keyworkers and vulnerable children in January 2021. When outbreaks did occur in nurseries, however, incidence rates in staff (24.26%) and children (2.34%) during November 2020, when there was little circulation of the Alpha variant, were significantly higher than primary school teaching staff (9.81%), secondary school teaching staff (3.97%), secondary school students (1.20%) or primary school students (0.84%) during September and October 2020, although these estimates were generated during a period of different community prevalence (Fig. 3a).24 Others have also reported extensive transmission in individual nursery outbreaks, with high SARS-CoV-2 positivity rates and seeding of infection into households.30 Although children are more likely than adults to be asymptomatic, serosurveys where antibody testing is used to capture both symptomatic and asymptomatic infections, have reported similar or lower seropositivity in children compared to staff members,31,32 adding to the growing evidence that children – especially infants and toddlers – are not the main drivers of infection in the household, educational settings or the wider community.33 In France, only 3.7% of 327 children and 6.8% of 197 staff in 22 day centres were seropositive for SARS-CoV-2 antibodies in June 2020, compared to 5.0% in non-clinical hospital staff controls.31 Additionally, almost half the seropositive children (45%) had been in contact with an adult household member with confirmed COVID-19,31 similar to our seroprevalence study in primary schools where most children were infected at home.32
Among children attending nurseries reporting a COVID-19 outbreak, the highest incidence was in infants and then declined with age, which was also observed in the in the national surveillance data.3 A likely explanation is that it is not possible to maintain social distancing or strict infection control between the younger infants and staff or parents because they require frequent and prolonged close contact for feeding and self-care. This was also frequently reported by nurseries participating in our investigations, along with difficulties in maintaining physical distancing between staff both within and across bubbles, which was compounded by sharing of staff rooms and bathrooms by staff across different bubbles in many nurseries. These factors likely contributed to the high incidence rates among staff, especially when the index case was also a staff member.
Not all outbreaks would have been reported to PHE, as smaller, less complex outbreaks would have been managed by the settings themselves, with support from the National Schools Advice line and other partners such as Local Authorities. Additionally, HPTs in some regions with high COVID-19 incidence only recorded larger outbreaks with at least 5 cases, complex outbreaks or those requiring public health action. We also relied on the settings to report positive cases among staff and children and, since the outbreaks were not investigated with mass testing for SARS-Cov-2 infection, it is possible that the index case may have been an asymptomatic child or adult who remained undetected. We are also unable to comment on the extent of asymptomatic spread in the affected settings for the same reason. In a recent Polish nursery outbreak, for example, wider RT-PCR testing of nursery staff, children and family members found that most of those infected with SARS-CoV-2 were asymptomatic.30 Another limitation is that the lack of viral genome sequencing data precludes assessment of transmission since it is not possible to determine whether the outbreaks were due to widespread transmission of a single strain or multiple introductions of different virus strains into nurseries.15 A higher community SARS-CoV-2 infection rate, for example, would result in more opportunities for virus introduction into nurseries and, therefore, higher estimates of incidence rates. Additionally, the lack of genome sequencing data meant that we were only able to assess the potential impact of Alpha indirectly by comparing outbreak size and incidence rates over time. We were unable to assess the nurseries who did not complete the survey. As such, this may have biased the results as those with time and capacity to complete the questionnaire may have had some systematic differences, such as staffing numbers, which could have affected the results. Finally, we were unable to objectively assess or compare adherence to physical distancing and other infection control measures adopted by the nurseries.