Covid-19 and schools—known unknowns

On 28 January The BMJ hosted a webinar on covid-19 and schools. An expert panel discussed the risks of community transmission to students and teachers, the clinical risks of covid-19 for children, the mental health effect of school closures, and international perspectives. Nikki Nabavi and Juliet Dobson report

The theme of the webinar was inspired by an editorial published in The BMJ called “Covid-19’s known unknowns.” The key message: “the more certain someone is about covid-19, the less you should trust them.”

Community transmission

Alasdair Munro, clinical research fellow in paediatric infectious diseases, Southampton University, began the discussion by defining the difference between “transmission” (infection passed on from one person to another) and “prevalence” (number of people infected at a given time).

“When we refer to young children, we’re normally talking about primary school age children and below. That is an important differentiator because there are some biological and epidemiological differences between young children and adolescents,” he said. This was later picked up on by Muge Cevik, clinical lecturer in infectious diseases, St Andrews, who suggested that in primary schools, children are often clustered tightly in their classrooms, whereas secondary school kids mix much more broadly, not only with other classrooms, but also have more afterschool contacts. Munro added: “Given the much larger population within secondary schools, the possibility of having very large outbreaks is bigger because there are simply more people who could potentially become infected if there were a large super spreading event.”

Susceptibility to infection

“In England, what we saw around November, December time when schools were fully open and the rest of the society was under restrictions, was an increase in cases among kids compared with adults,” says Cevik.

According to Munro, the best way to ascertain susceptibility to infection is through household contact tracing studies, as it can be assumed that everyone within a household has similar types of exposure. “Most of the studies have been well conducted and tested all contacts regardless of their symptom status, which is important as it eliminates a potential bias for children who may have a lower symptom burden.” He cited a meta-analysis in JAMA Paediatrics, which shows that the relative risk odds of a child having been infected in a household is roughly half of that of an adult. Cevik agreed that contact tracing data show that  working age adults are more likely to bring infection to the household than children. 

Clinical risks to children

Elizabeth Whittaker, senior clinical lecturer in paediatric infectious diseases and immunology, explained  that the risk of both hospitalisation and death from covid-19 is very low in children—those in hospital are mostly younger than 1 year or older teens. 

Some children have been affected by a chronic fatigue-like syndrome (fatigue driven with no evidence of end organ damage), whereas others have had paediatric inflammatory multisystem syndrome (PIMS, a post infectious phenomenon defined as “child presenting with persistent fever, inflammation, and evidence of single organ or multiorgan dysfunction, exclusion of any other microbial cause, and a history of SARS-Cov-2, or contact).” A third of these children develop severe disease and need emergency treatment, often ending up in intensive care. “These children are often negative on swab for the virus with the PCR test, but have already made antibodies, fitting with that epidemiological picture that this is a post-infectious phenomenon.” 

According to Whittaker, in the second wave, clinicians were much better at recognising PIMS early and initiating appropriate treatment, although a proportion of these children developed similar symptoms as adults with long covid.

Clinical risk to teachers

According to Rachael Wood, consultant in public health medicine, “the General Teaching Council for Scotland has provided [Public Health Scotland] with basic demographic details for almost all active teachers in Scotland, and NHS organisations have provided similar data for healthcare workers.” This allows assessment of risk to teachers  and healthcare workers separately from the general working age population. Findings have been reassuring for teachers, as there is now “strong evidence” that teachers are at a similar or lower risk of severe illness or dying from covid-19, than the general working age population, “and that finding persists when schools are open.” 

Wood reminded attendees  that similar or low risk is not no risk, and that the risk of covid for teachers reflects the background risk in the communities and populations in which they live.


Munro stated that infectiousness has been much more difficult to ascertain in children, who they can still infect others, even when asymptomatic. 

“It’s important to emphasise that [transmission] depends on various factors—it is not only one dimensional,” says Cevik. “For example, in England, we saw much more outbreaks in secondary schools in the most deprived areas. Those kids might be living in much more crowded and multigenerational households that are also at higher risk of infection outside of the household.”

Munro explained why tracing infectiousness in children can be especially complicated, as a result of their shared exposure with classmates. “For example, you find a child case A who is infected, you test their contact, B, who is also infected, and you assume that A has infected B, whereas what has actually happened was A and B were together at the time they were exposed to C (the initial infectious person). Then we don’t know if C infected A and B, if C infected A, who infected B or C infected B, who infected.”

Cevik added  “When schools are open, there’s evidence to suggest in school transmission, and that these are stochastic events . . . There are still many unknowns, especially the impact of schools on community prevalence, hospitalisations, and deaths . . .How and when to open close schools is not solely a scientific decision, because I think schools provide more than education.”

Impact of school closures

Elisabeth Gilpin, the headteacher of St Mary Redcliffe School, one of the largest secondary schools in Bristol, started the session by pointing out that “schools are not closed.” As well as real time remote learning, “children are on site” because they “are vulnerable, perhaps because of their special needs or their home circumstances.” 

Gilpin highlighted the emotional trauma that covid-19 is having on young people, saying that “our country is going through a pandemic and so there are so many more bereavements . . . parents are more anxious, perhaps because of potential job losses, and children will be soaking up all this trauma.”

According to Sunil Bhopal, NIHR academic clinical lecturer in paediatrics at Newcastle University: 

“We’ve been causing great harm in many ways to children through our pandemic response,” he said, “is this harm justified by the benefits brought about towards pandemic control?” “Schools matter not just for education and learning, but for everything else. Children need children. They need to spend time with them for socialisation and interaction,” Bhopal said. “When one in four young people are saying that they feel unable to cope with life, I worry.”

Bruce Adamson, children and young people’s commissioner for Scotland, agreed that  “this has been the biggest human rights crisis that we faced in a very long time.” 

Haroon Chowdry, Director of Evidence for Children’s Commissioner England, talked about how worried children are: .” “Children talk about how they feel like they’ve been blamed for covid . . . they feel like they’re being blamed for spreading it.” 

International perspectives

Ibukun C Akinboyo, professor of paediatrics at Duke University Hospital, North Carolina, US, discussed the data driven approach that she and her colleagues have taken to measure and inform their response to covid-19. “We saw very few, not zero, very few secondary cases.” Where they saw cases, she said, “we adjusted, and we haven’t seen secondary spread. So it is possible to have in-person education safely during a pandemic. It’s certainly not the panacea for all of our problems, but we should have ongoing discussion.”

Margrethe Greve-Isdahl from the Institute of Public Health in Norway, discussed the Norwegian strategy. The country’s  schools closed in March 2020 and gradually started reopening by the end of April—a lot earlier than many other countries. 

A “traffic light” system helped. From May to November, the yellow level was sustained. But when cases rose from the end of October measures were tightened, but schools remained.

“But we have also had a lot of communication relating to staying home when sick and implementing hygiene measures. And we are one of the few countries that have been able to avoid masking in schools, either for teachers or pupils at any age,” she said.

Armand Fontanet, Pasteur Institute, explained that France’s president and minister of education are very much pro school opening. Nevertheless, schools were closed in France from March until May. 

In May, schools resumed, but with “quite tough restrictions. We put masks on all kids from the age of 6and above, and things went OK. . . . a  few outbreaks here and there, but nothing really major,” Fontanet said. Over autumn, France went into lockdown, but kept schools open.

However, now the new variants of covid-19 and the experience from other parts of Europe is causing concern. Fontanet: “ I really cannot tell you what decision will be made . . . schools  will be probably the last thing that the president will agree to close.”


This webinar was part of The BMJ‘s series of covid-19 known, unknowns webinars. Find out more and register for future events here. 

Nikki Nabavi, editorial scholar, The BMJ

Juliet Dobson, editor, The BMJ

Competing interests: none declared.


Panel and Agenda

Community prevalence/ transmission/ dissemination 

Chair: George Davey Smith, Bristol University

Alasdair Munro, University of Southampton

Muge Cevik, University of St Andrews

Clinical risks (severe disease/long covid) 

Chair: Allyson Pollock, Newcastle University

Risks to children: Elizabeth Whittaker, Imperial College, London

Risks to teachers: Rachael Wood, Public Health Scotland and David McAllister, University of Glasgow

Impacts of school closures 

Chair: Caroline Relton, University of Bristol

Teachers’ perspective: Elisabeth Gilpin, St Mary Redcliffe School, Bristol,

Consequences for children: Sunil Bhopal, Newcastle University

Rights of the child:

Bruce Adamson, Children and Young People’s Commissioner for Scotland

Haroon Chowdry, Director of Evidence for Children’s Commissioner England

International perspectives 

Chair: Fiona Godlee, The BMJ

USA: Ibukun C. Akinboyo, Duke University Hospital

Norway: Margrethe Greve-Isdahl, Institute of Public Health, Norway

France: Arnaud Fontanet, Pasteur Institute

North America/ Europe: Jennifer Couzin-Frankel, Science Magazine


Chair: Phil Hammond, Gp, Journalist and Broadcaster