Children in schools as “super-spreaders” of Sars-Cov-2: myth or truth?
Having established during the first COVID-19 outbreak that protecting school-aged children from Sars-Cov-2 infection was largely unnecessary, the rationale for restricting in-presence school attendance lay entirely in its potential epidemiological benefits in protecting the adults in their households and the educational staff in their schools. Keeping students at home directly reduces in-school and public transportation social contacts for all of pupils, teachers and school staff, thus limiting the mixing of households and horizontal viral transmission, as well as limiting the interactions between children, who have been repeatedly labelled as potential “super-spreaders” of infectious diseases on account of their increased crowding compared to adults. It also holds secondary pandemic mitigation potential signalling the severity of the public health emergency and keeping caregivers at home, albeit harder to quantify.
The questions now arise: in the case of the COVID-19 pandemic and based on what is currently known, is the idea of children in the school-setting acting as “super-spreaders” plausible? Is it reasonable to suspect that the second wave of the Sars-Cov-2 virus spread in the WHO European region has been driven also by school re-opening to in-presence learning for the 2020-2021 academic year? Two key unknowns should be investigated to understand Sars-CoV-2 infectiousness in paediatric age. First, what is the susceptibility of children and adolescents to the infection, and are there differences in attack rates, length of the infection and viral shedding between the two age groups? Second, depending on their susceptibility to the infection and their opportunity for onward transmission, how likely are children and adolescents to become the index case of COVID-19 clusters, respectively?
With regards to the first key unknown, preliminary evidence collected during the first COVID-19 outbreak elucidated that adults appeared to be at a higher chance of becoming infected with Sars-CoV-2 on exposure compared to younger age groups. This starting point for interpreting the infectiousness of Sars-CoV-2 in children comes from a large metanalysis published at the end of September 2020 [13] that included 32 studies. Evolving evidence strengthens these preliminary ones: children and adolescents are not spared from the infection, but they are a repeatedly underrepresented population of COVID-19 cases. This is logically true for the very few seroprevalence studies conducted on symptomatic cohorts [14-15], susceptible to an inclusion bias due to the high prevalence of subclinical infection in the younger populations. More significantly, this is also true for the few large seroprevalence studies on random cohorts selected independently of symptoms, in countries where children were continuously exposed to the infection because of continued in-presence learning or in which the exposure probabilities were comparable across generations [16-19]. Unfortunately, this epidemiological fact is not established biologically: when it comes to viral load during the infection and duration of viral shedding, especially in symptomatic cases, there seem to be fewer differences between adults and children of all ages [20]. Less has been confirmed for asymptomatic cases: the full demonstration of a significantly shortened or lessened infectiousness of asymptomatic cases would indisputably strengthen all positions favouring school opening, giving a reproducible explanation for the observed lower susceptibility to Sars-CoV-2 of the younger demographic groups.
The consistency of the epidemiological evidence alone encourages the claim that children should not be a primary target population of COVID-19 pandemic mitigation strategies [21] and has been used as sufficient evidence to disclaim the role of children as super-spreaders [22], but it does not disregard the risk that children are susceptible to the infection and may hence transmit it both in their classroom and within their household. As a very recently published modelling study [23] discloses, when it comes to the dangers of seeding COVID-19 cases across a large community through extensively undetected school clusters, even an extremely low number of child index cases might eventually amplify and result in a sizeable community spread with increased overall incidence and mortality. Confirming this concern are two large early studies coming from Asia [24,25]. A number of other studies strongly contradict this evidence, 57 of which were included in an international metanalysis on household Sars-CoV-2 transmission clusters during the first pandemic wave [26], finding that children had lower secondary attack rates than adults and that they infrequently represented the index case. This data was partly re-confirmed for the second wave in a large UK study [27].