Covid-19 and children: what does the science tell us, and what does this mean as the lockdown is eased?

Key points
As this is a longish post – perhaps a 10-minute read – here are the main takeaways:

• Children are much less likely to become seriously ill from Covid-19 than adults, and appear less likely to become infected.
• Unlike with influenza, it appears that children are not more likely than adults to spread the disease, and may be significantly less likely.
• There are good grounds for thinking that outdoor environments present a low risk of infection compared to indoor ones, especially where the time spent in close proximity to other people is short.
• Pandemic control measures are likely to lead to significant collateral damage to children, with the most vulnerable and disadvantaged children worst affected.
• Government, local authorities and other public agencies should take a balanced approach to supporting children through the pandemic. They should:
  • Encourage schools and child care centres to take learning activities outdoors, prioritising play and breaks, and maximize outdoor play time, as they reopen.
  • Open all remaining closed parks, review the closure of playgrounds, and take a supportive approach to the oversight of children’s play and socialising in public space.
  • Address the circumstances of disadvantaged children as a matter of urgency.
  • Prioritise children’s active travel to school, to help reduce peak hours congestion.
  • Closely monitor emerging evidence, especially from countries that have relevant experience of relaxing measures.
  • Encourage the public to engage with and understand the evidence base, and keep them informed as it grows.

Introduction
The government’s plans for relaxing the lockdown, including greater freedom to spend time outside, and the possible re-opening of schools, have unsurprisingly generated huge debate. At the same time, evidence is growing on how Covid-19 affects children, and of children’s role in the spread of the disease. This post shares my take on that evidence base and its implications.

The post starts with a summary of the clinical and epidemiological evidence base. It then looks at the collateral damage to children of the pandemic control measures. It closes with implications for policy and practice, with a particular focus on children’s play and mobility.

My discussion of the science does not claim to give a systematic or comprehensive review of what is a complex picture. I am not a clinician or epidemiologist. However, I have benefited from the expertise of Dr Alasdair Munro, a clinical research fellow in paediatric infectious diseases.

I draw heavily on Dr Munro’s 20 April 2020 tweetorial and on this website, which Dr Munro contributes to. As these links show, there is lively debate about many aspects of the science of Covid-19.

How does Covid-19 affect children, and how does this compare to the disease in adults?
Since the outbreak began in China, it has been clear that children with Covid-19 have much milder symptoms than adults – especially children under 10. While serious complications are not unheard of, and should not be ignored, they are much rarer than for older age groups.

How likely are children to become infected?
Children appear less likely to become infected than adults. A review of international studies by the Dutch National Institute for Public Health and the Environment states “the percentage of children among the confirmed COVID-19 patients is small, varying from 1% in young children up to 6% in older children.”

Contact tracing studies in Japan and China give a similar picture, with one Chinese study finding that adults aged 15-64 were three times as likely to be infected by others in their household than children aged 0-14 were, while adults aged over 65 were at even greater risk.

In Iceland, where a large-scale testing has been carried out, a population screening programme did not reveal a single infected child under the age of 10 (out of a total of 100 confirmed cases).

Taken together, the lower infection rates and typically milder symptoms in children mean that the disease is affecting the child population far less severely than for older age groups. The UK Office for National Statistics mortality dataset – probably the most reliable source of information on fatalities – shows that March 2020 saw just one Covid-19-related child fatality (in fact aged between 15 and 19). The chart below shows the ONS figures for different age groups.

How likely are children to spread the disease?
Studies of household clusters of infection show that children are the ‘index case’ (i.e. the likely original source) in only around 5-10% of cases. The Dutch review cited above found no records of spread from children under the age of 19, in a sample involving 55 cases, and stated that children “play a much smaller role in the spread than adults and the elderly.”

Similarly in Sweden (where schools and childcare have not been closed) the National Public Health Agency states that none of the outbreaks it has seen so far have been linked to schools or children. An Australian contact tracing study following up 18 cases from 15 schools in New South Wales found very low rates of infection in school children and low rates of spread, with no cases of adults becoming infected from children (although absolute numbers are low).

This is in marked contrast to other viral diseases such as influenza, where studies suggest children are the source in around half of cases.

A different picture is given by one German study (not yet peer -reviewed, but widely-discussed in the media) which concluded that children may be as infectious as adults. However, Dr Munro’s team argue that this study and its conclusions are seriously flawed. It assumes that infectiousness is just a matter of viral load as shown by a swab test (when other factors are plausibly important, most obviously whether the person is coughing). Moreover, the sampling methodology is not clear, the sample of children is very small (and also unclear) and perhaps most seriously, one of their own analyses does show a statistically significant difference between children and adults.

Data from Iceland, where alongside extensive testing schools and childcare have largely remained open, adds weight to the view that children are less likely to spread the disease than adults. Experts there state that none of the 1800 cases identified up to April 2020 were infected by a child under 10 years of age.

As this discussion shows, it is not easy to study how likely children are to spread the disease compared to adults. This is partly down to the low numbers of infected (and especially of symptomatic) children. What is more, in many countries, schools and childcare settings were closed early on in the outbreak. This may have led to children appearing less often as index cases. It will be critical to study this closely as countries move out of lockdown and reopen schools and childcare settings.

What do we know about how infection spreads indoors compared to out of doors?
The evidence base for transmission routes is patchy, partly because of the challenges in carrying out direct research. The emerging picture is that outdoor, airborne transmission is rare compared to transmission via close personal contact, especially in indoor settings. One Chinese study found that 80% of outbreaks of multiple cases were associated with the home environment, while a third were linked with transport (note that many outbreaks were linked to more than one venue). Out of 318 outbreaks in total, only one was linked to an outdoor environment.

This widely-shared blog post by a comparative epidemiologist, while not peer reviewed, discusses documented case studies of multiple spreading events. It argues that the time spent in proximity to infectious people is a critical factor, and that the environments most likely to lead to outbreaks are “enclosed, with poor air circulation and high density of people.”

How are pandemic control measures affecting children?
It is of course hugely reassuring that Covid-19 rarely makes children ill. By contrast, pandemic control measures are likely to be hitting them hard.

Children are under what is in effect house arrest. Many face extreme social isolation and domestic stress. They are anxious for their own health, worried about sick friends and relatives, and grieving for loved ones who have died before their time.

An editorial in the British Medical Journal warns of harm to their education and learning, an increase in domestic violence and abuse, mental health problems, and lower levels of play and physical activity. The most vulnerable children are likely to be particularly badly affected, including those in poor housing and homeless circumstances, and children with special educational needs and disabilities.

These warnings are echoed in a paper from leading British child psychology academics and experts, who have shared their concerns with Ministers. It noted that as of early May, every child in the UK had spent 6 weeks without seeing any of their friends face-to-face. Drawing on a rapid evidence review, it warned of the impact of social isolation on children, and called for opportunities for play with friends to be a priority.

“The worst things about the lockdown are no friends, not having siblings so I’m lonely and having to hear my dog bark all day. I’m getting on good with my parents I guess, but sometimes we argue.” 11-year-old child, quoted by the Children’s Commissioner for England.

Implications for education and childcare
The scientific evidence offers a great deal of reassurance as far as the infection risks to children are concerned, and especially younger children. As for teachers and childcare workers, the picture is less clear. However, they are arguably at greater risk of infection from their colleagues than from the children in their care, and may be well advised to be more careful about what happens in the staffroom than the playground.

The evidence also strengthens the case for taking learning outdoors. This move is recommended by the World Bank. It is also encouraged by the Danish National Board of Health [pdf link to Danish language publication – see p.8], and is being explored seriously in Scotland. National Dutch public health guidance also states that “going to school and playing outside is possible.”

In the UK and elsewhere, schools rarely give children’s play and socialising serious thought. However, play and breaktimes are hugely important to children themselves, even in normal times. And play has an important role in helping children to keep physically and mentally well, and to make sense of, and bounce back from, uncertain and challenging situations. It will be more important than ever to allow them to rebuild their friendship and peer support networks.

Almost nothing about the future working of schools and childcare will be easy. A detailed discussion is beyond the scope of this post. However, the graphic below (sourced from twitter here, and showing an extract of the pandemic arrangements for one school in Quebec, with key sections highlighted by me) is surely an example of how *not* to respond.

Implications for children’s play and mobility
Much of the above discussion is also relevant to children’s lives outside of school and childcare. The evidence base strengthens the case for children to be given more opportunities for outdoor play, and for official measures that support this, while continuing to manage the risk of spread. One key issue is the continued closure of public playgrounds due to concern about transmission from equipment surfaces. Evidence from countries that are reportedly reopening their playgrounds may shed light on this complex topic.

While the UK response has emphasised the value of exercise, one striking and disturbing feature has been the complete lack of consideration of what exercise looks and feels like for children. Put simply, for many children – and especially for younger children – ‘exercise’ *equals* ‘play’. It is essential that this is recognised in future guidance and policy.

Once lockdown measures recede in earnest and children return to school, the road network is set to become more congested. The morning peak in particular could see huge congestion, with public transport capacity likely to remain limited for months to come. Measures to encourage more children to walk, cycle and scoot to school would make a substantial contribution to reducing the demand for car trips, and should be a transport planning priority.

Conclusions and reflections
When the pandemic began, there was huge uncertainty about its impact on children. Coupled with assumptions about children’s possible vulnerability and role in disease spread, this led to the early adoption of measures such as school and childcare closures. In the light of the evidence reviewed here, it is right that government is re-examining its position on school closures. It now urgently needs to re-examine the position and guidance on childcare, on children’s play – in and outside of school – and on their mobility.

Our understanding of the virus and its spread is growing fast. So is our knowledge of the impact of infection control measures, including their potential side-effects. The different approaches in different countries, when combined with good data, will help to improve the evidence base and inform responses.

The pandemic will raise many, many questions that we will rarely if ever be able to be answer with complete certainty. We need to do the best we can, using the best available evidence.

What is more, the public should be encouraged to make sense of the evidence base, and to be kept informed as it grows. Official guidance will never be able to cover all the situations that we will face (as the criticism of the UK government’s recent changes in its core messaging shows).  Guidance and responses risk being misunderstood, distrusted, undermined and even ignored, if people are not actively helped  to understand the science behind them.

My expertise on children, risk and public policy is largely drawn from decades of involvement in play safety. A key insight from this experience is that emotive reactions and a demand for absolute safety are unhelpful, if not positively harmful. Progress depends on adopting a more measured, balanced approach.

The Covid-19 pandemic presents threats and challenges that are orders of magnitude more serious than found on a children’s playground. However, responding to it also needs balance, and constructive, reasoned, respectful discussion and debate. I would welcome comments and reactions offered in this spirit.

Acknowledgements
I am extremely grateful to Dr Alasdair Munro for his input into this post, and for responding so enthusiastically and generously to my initial approach on twitter. I am also grateful for input and feedback from Prof Mariana Brussoni: longstanding collaborator, kindred spirit and global champion of children’s right to play. The views given here are mine, not theirs.