Closing schools is not evidence based and harms children
BMJ 2021; 372 doi: https://doi.org/10.1136/bmj.n521 (Published 23 February 2021) Cite this as: BMJ 2021;372:n521Linked Editorial
Covid-19: Keeping schools as safe as possible
Linked Opinion
School closures have had a huge impact on children’s futures
Read our latest coverage of the coronavirus outbreak
All rapid responses
Rapid responses are electronic comments to the editor. They enable our users to debate issues raised in articles published on bmj.com. A rapid response is first posted online. If you need the URL (web address) of an individual response, simply click on the response headline and copy the URL from the browser window. A proportion of responses will, after editing, be published online and in the print journal as letters, which are indexed in PubMed. Rapid responses are not indexed in PubMed and they are not journal articles. The BMJ reserves the right to remove responses which are being wilfully misrepresented as published articles or when it is brought to our attention that a response spreads misinformation.
From March 2022, the word limit for rapid responses will be 600 words not including references and author details. We will no longer post responses that exceed this limit.
The word limit for letters selected from posted responses remains 300 words.
Dear Editor
We agree with Lewis et al. [1] that reopening schools should be the UK’s top priority. However, there is evidence that strongly contradicts two of their key claims :(i) that school closures do not reduce SARS-CoV-2 transmission, and (ii) that the overall risk to children and young people from COVID-19 is very small.
It has been clear for months that schools and children contribute to SARS-CoV-2 transmission. Many of the studies cited by Lewis et al. [1] suffer from serious limitations [2]. These include relying on symptom-based testing in children, using data from when schools were closed or attendance was low [3,4], and from settings with low community transmission or strict mitigatory measures [5]. Symptom-based testing underestimates infections and transmission. The few studies that have included asymptomatic testing in children have mostly been small and have either not involved regular surveillance [6,7], or relied on seroprevalence as an indicator of past infection [8]. This is problematic, because children may be less likely to seroconvert than adults [9].
Lewis et al. argue that adults living with children in the UK were not at increased risk of death from COVID-19. However, this was based on data from the first wave in the UK during which schools were largely closed [10]. When data from the second wave were included, living with children was associated with an increased risk of infection and hospitalisation, and an increased risk of death from COVID-19 in adults over 65 years [11]. This is despite parents of children having a much lower risk of non-COVID-19 deaths compared to the control group, suggesting this group is healthier in other ways, with residual confounding likely underestimating the relative risk of death among parents [11].
A study of over 3 million adults from Denmark showed that the risk of testing positive for SARS-CoV-2 was higher in households with young children, increasing with the number of children in the household [12]. A large US study found increased risk of infection among parents of primary and secondary school children attending school in person [13]. Transmission within schools is also reflected in the increased risk for teaching staff. The REACT-1 survey in England shows that those in teaching-associated occupations were at significantly higher risk of infection compared to other occupations, even when schools were only partially open [14]. This is consistent with the global evidence – e.g. in Sweden secondary school teachers doing in-person teaching had a two-fold higher risk of COVID-19 infection than their colleagues who were teaching remotely [15].
Having schools open without sufficient mitigations has a major impact on community transmission. Studies that have carried out careful case-control examinations of the specific impacts of thousands of interventions over many countries have shown that both primary and secondary school closures result in substantial reductions in the effective reproduction number (Rt) [16,17]. In June 2020, the UK Government’s Scientific Advisory Group for Emergencies (SAGE) warned that community transmission would rise when schools reopened for the autumn [18]. Infection positivity among children has closely tracked school openings and closures [19,20] with clear temporal associations between transmission rates and school closures and openings for mid-term holidays [19]. According to the UK’s Office for National Statistics (ONS) COVID-19 Infection Survey, 2% of 2-10 year-olds and 3% of 11-16 year-olds in England were estimated to be infected with SARS-CoV-2 at the end of term in December, higher than any other age group [19]. During the short school openings in England in March, cases increased rapidly among primary and secondary school children, declining over the Easter school holidays, with rises resuming recently.[20]
The role of schools has become more critical with the emergence of new variants. In November 2020, during the second lockdown in the UK when schools remained open, cases continued to rise in regions where the now dominant B.1.1.7 variant was prevalent [19,20]. Rt only dropped below 1 once schools were closed. Notably, between Nov-Dec, when B.1.1.7 was increasing in frequency, this variant was more dominant in school-age children suggesting schools may have contributed to early spread. Since then, several outbreaks of B.1.617.2 been reported in schools,[21] and cases among secondary school-age children are now the highest across all groups in England.[22]
With regard to the second claim, while the risk of death from COVID-19 is low in children, mortality is not the only concerning outcome. Long COVID is a relatively common outcome of infection [23,24], and can result in significant disability [25]. Data from the ONS suggest 7% of those aged 2-10 years and 8% of those aged 12-16 years have at least one persistent symptom 12 weeks after testing positive [23] and similar reports are emerging from other countries [25,26]. Given uncertainty regarding the long-term health consequences of SARS-CoV-2 infection, it is unwise to allow the virus to circulate in children.
Furthermore, the mental health burden in children of “bringing the virus home” to family members, the anxiety associated with the risk of their loved ones dying or suffering prolonged illness, and the impact of losing a parent or grandparent should not be minimised. A US study estimated that over 34,000 American children aged 0-17 had lost at least one parent to COVID-19. These considerations are likely even more pronounced in those from minoritised or deprived backgrounds, worsening existing inequalities [27].
The false dichotomy of schools being either open or closed that has dominated the discourse during the pandemic has hampered the development of adequate mitigation strategies that would allow us to keep schools open more safely with more children attending in person [28], something that we all want to see. Other countries have shown that this is feasible with multi-layered mitigatory measures in place [13]. Given the role schools play in community transmission, focusing solely on keeping them open without meaningful mitigation against infection risks worsening the pandemic and forcing closures in places where community transmission is increasing. We must therefore focus on making schools safer to avoid the devastating effect of further educational disruption.
References:
1 Lewis SJ, Munro APS, Smith GD, et al. Closing schools is not evidence based and harms children. BMJ 2021;372:n521. doi:10.1136/bmj.n521
2 Hyde Z. COVID-19, children and schools: overlooked and at risk. Medical Journal of Australia 2020;213:444-446.e1. doi:https://doi.org/10.5694/mja2.50823
3 Viner RM, Russell SJ, Croker H, et al. School closure and management practices during coronavirus outbreaks including COVID-19: a rapid systematic review. The Lancet Child & Adolescent Health 2020;4:397–404. doi:10.1016/S2352-4642(20)30095-X
4 Ismail SA, Saliba V, Bernal JL, et al. SARS-CoV-2 infection and transmission in educational settings: a prospective, cross-sectional analysis of infection clusters and outbreaks in England. The Lancet Infectious Diseases 2021;21:344–53. doi:10.1016/S1473-3099(20)30882-3
5 Theuring S, Thielecke M, Loon W van, et al. SARS-CoV-2 infection and transmission in school settings during the second wave in Berlin, Germany: a cross-sectional study. medRxiv 2021;:2021.01.27.21250517. doi:10.1101/2021.01.27.21250517
6 Brandal LT, Ofitserova TS, Meijerink H, et al. Minimal transmission of SARS-CoV-2 from paediatric COVID-19 cases in primary schools, Norway, August to November 2020. Eurosurveillance 2021;26:2002011. doi:10.2807/1560-7917.ES.2020.26.1.2002011
7 Macartney K, Quinn HE, Pillsbury AJ, et al. Transmission of SARS-CoV-2 in Australian educational settings: a prospective cohort study. The Lancet Child & Adolescent Health 2020;4:807–16. doi:10.1016/S2352-4642(20)30251-0
8 Ulyte A, Radtke T, Abela IA, et al. Clustering and longitudinal change in SARS-CoV-2 seroprevalence in school-children: prospective cohort study of 55 schools in Switzerland. medRxiv 2020;:2020.12.19.20248513. doi:10.1101/2020.12.19.20248513
9. Weisberg SP, Connors TJ, Zhu Y et al. Distinct antibody responses to SARS-CoV-2 in children and adults across the COVID-19 clinical spectrum. Nat Immunol 2020; 22: 25-31. doi:10.1038/s41590-020-00826-9
10 Forbes H, Morton CE, Bacon S, et al. Association between living with children and outcomes from COVID-19: an OpenSAFELY cohort study of 12 million adults in England. medRxiv 2020;:2020.11.01.20222315. doi:10.1101/2020.11.01.20222315
11 Forbes H, Morton CE, Bacon S, et al (2021). Association between living with children and outcomes from covid-19: OpenSAFELY cohort study of 12 million adults in England. BMJ, n628. https://doi.org/10.1136/bmj.n628
12 Husby A, Corn G, Krause TG. SARS-CoV-2 infection in households with and without young children: Nationwide cohort study. medRxiv 2021;:2021.02.28.21250921. doi:10.1101/2021.02.28.21250921
13 Lessler J, Grabowski MK, Grantz KH, et al. Household COVID-19 risk and in-person schooling. Science 2021.
14 Riley S, Wang H, Eales O, Haw D, Walters CE, Ainslie KEC et al. REACT-1 round 9 final report: Continued but slowing decline of prevalence of SARS-CoV-2 during national lockdown in England in February 2021
15 Vlachos J, Hertegard E, H BS. The effects of school closures on SARS-CoV-2 among parents and teachers. Proc Natl Acad Sci U S A 2021; 118(9).
16 Li Y, Campbell H, Kulkarni D, et al. The temporal association of introducing and lifting non-pharmaceutical interventions with the time-varying reproduction number (R) of SARS-CoV-2: a modelling study across 131 countries. The Lancet Infectious Diseases 2021;21:193–202. doi:10.1016/S1473-3099(20)30785-4
17. Haug N, Geyrhofer L, Londei A, Dervic E, Desvars-Larrive A, Loreto V et al. Ranking the effectiveness of worldwide COVID-19 government interventions. Nature Human Behaviour 4, 1303-1312. 2020.
18. SPI-M-O Expert narratives for potential Autumn and Winter Events, 29th July 2020.
19. Scientific_Advisory_Group_for_Emergencies. Children's Task and Finish Group: update to 4th Nov 2020 paper on children, schools and transmission. UK; 2020.
20. Office_for_National_Statistics. Coronavirus (COVID-19) Infection Survey 2020
21. Public Health England. SARS-CoV-2 variants of concern and variants under investigation in England. Technical briefing 14. 3 June2021
22. Public Health England. Weekly national Influenza and COVID-19 surveillance report Week 22 report (up to week 21 data). 3 June 2021
23 Office for National Statistics. Prevalence of ongoing symptoms following coronavirus (COVID-19) infection in the UK: 1 April 2021.
24 Buonsenso D, Munblit D, Rose CD, et al. Preliminary Evidence on Long COVID in children. medRxiv 2021;:2021.01.23.21250375. doi:10.1101/2021.01.23.21250375
25 Over 200 Kids in Sweden Diagnosed with Long-term Covid-19 Symptoms with Some Being Bed-ridden. https://in.news.yahoo.com/over-200-kids-sweden-diagnosed-130000439.html (accessed 6 Mar 2021).
26 Sante GD, Buonsenso D, Rose CD et al. Immune profile of children with post-acute sequelae of SARS-CoV-2 infection (Long Covid). MedRxiv (preprint). 2021
27 Kidman R, Margolis R, Smith-Greenaway E, Verdery AM. Estimates and Projections of COVID-19 and Parental Death in the US. JAMA Pediatrics. 2021 doi:10.1001/jamapediatrics.2021.0161
28 Jeffreys, B. More children in England missing school over Covid-19. BBC News. 27 October 2020.
Thanking you,
Yours sincerely
Deepti Gurdasani, Queen Mary University of London, UK
Lonni Besancon, Monash University, Australia
Nisreen A Alwan, University of Southampton, UK
Sarah Dean Rasmussen, University of Cambridge, UK
Trish Greenhalgh, University of Oxford, UK
Luke Johnson, University of Southampton, UK
Martin McKee, London School of Hygiene and Tropical Medicine, and Independent SAGE, UK
Susan Michie, University College London and Independent SAGE, UK
Kim Prather, University of California, San Diego, US
Stephen Reicher, University of St. Andrews and Independent SAGE, UK
Paul Roderick, University of Southampton, UK
Hisham ZIauddeen, University of Cambridge, UK
Zoë Hyde, University of Western Australia, Australia
Competing interests: No competing interests
Dear Editor,
(In what follows I am using the original reference numbers)
I have serious concerns regarding the editorial “Closing schools is not evidence based and harms children” by Lewis at al. Neither the paper’s title nor its contents accurately portrait state-of-the-art information regarding the impact of school closures on transmission. Such lack of accuracy is contrary to the responsible role that the scientific community should aim for in communicating scientific topics.
Based on reference 12 (a preprint) Lewis et al. assert that “Although school closures reduce the number of contacts children have, and may decrease transmission, a study of 12 million adults in the UK found no difference in the risk of death from covid-19 in households with or without children. Only 3% of people aged over 65 live with children”. This formulation is misleading in the absence of further clarification: Covid-19 mortality is related to virus incidence in the population as a whole, not to virus incidence in household with children (or children and elderly). Elderly people living in nursing homes do not die from covid-19 because they co-live with children. They become infected because there are viruses circulating at the nursing home, and there are viruses circulating at nursing homes because there are viruses circulating in communities. The likelihood of the virus "entering" a nursing home (or a household) increases with its incidence in the community(a). Thus, the formulation and content chosen by the authors fails to place emphasis where it belongs: the question of schools – and therefore of children – is whether or not they contribute to transmission, and thus to infection level, in the community.
The evidence reviewed for children transmission in the editorial is not extensive, but generally confirms previous findings: transmission by younger children is less frequent (but not zero); infection dynamics in younger age cohorts broadly follow the curve for the rest of society. The latter is important because it suggests there is indeed permeability between age cohorts regarding infection incidence, despite lower levels of infection and transmission in younger children.
But one thing is to document transmission, another is to estimate the effect of school closures on community incidence. To address the potential role of school closures in controlling infection levels, the authors consider two meta studies. Study 25 (https://www.nature.com/articles/s41562-020-01009-0) finds an effect of school closures. However, Lewis et al. dismiss the conclusions of study 25 as “confounded”. Instead, they present study 26 (another preprint) as a bias-controlled analysis.
The authors of meta-study 26 (doi.org/10.1101/2021.01.02.21249146), define a set of criteria that reduces their initial pool of 3318 papers to 10. Seven of these papers actually support an effect for school closures in transmission. However, only 2 of the 10 are considered at a low risk of bias – precisely two of the three that do not show an effect for schools. It is worth noticing that both of these at lower risk of bias studies relate to first wave data (as sampling efforts then were likely to have excluded more asymptomatic cases).
Hsiang et al. (2020; doi.org/10.1038/s41586-020-2404-8) analyse policies adopted in six countries using econometric techniques. School closure is an isolated policy in only three countries (Italy, France, and the USA). Also, the order in which the policy is introduced varies between countries. School closure effect is lowest or null in the USA and largest in Italy. On their conclusions, the authors focus not on single policies but on their combined effect, showing that large-scale anti-contagion policies have slowed the COVID-19 pandemic in all countries studied.
Courtemanche et al. (2020; doi.org/10.1377/hlthaff.2020.00608) use a model to estimate the impact of different social-distancing measures in the USA. These authors conclude that neither school closures, nor bans on large social gatherings influenced the growth rate of confirmed covid-19 cases. That no effect is found for bans on large social gatherings is odd, given what we know of the role of cluster transmission, and may indicate a general lack of model sensitivity. Most importantly, however, Courtemanche et al. do not conclude that there is no evidence for an impact of school closures. Rather, they state that confidence intervals allow for moderate effects of school closures (and also of bans on large social gatherings).
In fact, authors of the preprint meta-study 26, similarly do not conclude that evidence for an impact of school closures is lacking. Instead, they conclude that evidence for effectiveness is varied: “With such varied evidence on effectiveness, and the harmful effects, policymakers should take a measured approach before implementing school closures.” After rejecting all existing data as biased except two studies from May 2020, their conclusion remains in line with other, more inclusive reviews – including meta study 25, which Lewis et al. rejected as “confounded”.
The title and content formulation in Lewis et al. are misleading in that they do not reflect current data on the efficacy of school closures on transmission – they also do not reflect the conclusions of the studies referenced. Moreover, the authors' choice of studies relies on preprints perhaps more than would be advisable to comment properly. The broad consensus so far is not that evidence is lacking for an effect of schools closures, it is that given its harmful impacts, school closures should be considered only when community transmission is high.
We are living through difficult times. Dramatic titles and pruned evidence do not help us inform policymakers, journalists, and our communities. They only add to the emotional burden of the pandemic. As scientific community I believe we can do better.
Competing interests: No competing interests
Dear Editor
The Editorial “Closing schools in not evidence based and harms children” is written with a view of the world through a Paediatric lens (not a criticism, but an observation). While this is appropriate, would an article written by advocates for the elderly come to the same conclusions? Writing off the elderly as “socioeconomically privileged people” as a group is unfair and frankly ageist. Through another lens the elderly are the generation who have worked their whole lives and improved society through their efforts and now deserve a safe and healthy retirement. The elderly had their schooling disrupted for six years by a far worse event (World War 2), and yet they survived and thrived to bring us up to the adults we are now. Children are tougher and more resilient than some like to make out.
The authors offer one view of the purpose of schools. Others might argue (quietly) that schools are for education. How much help is taking a child out of an unsafe environment (home) for eight hours a day, five days a week, and then returning them to that environment for the rest of the week? Much bullying, shaming and harm does not occur in schools but occurs outside of schools and (online), from peers not parents.
Only 3% of people over 65 may live with children, but there is a preponderance of multi-generational living in ethnic minorities who are most at risk from COVID-19. Additionally although only 3% live with older people, how many rely on older generations for out of school care whilst parents work?
The role of children in community transmission is not clear, but very concerning. The current Papatoetoe High School outbreak in Auckland New Zealand has the potential to spread the virus into a large at risk community with potentially devastating results. Fortunately the response appears to have mitigated this risk at the present time. Children, I would argue have most to lose from COVID-1: their parents.
The call for a systematic evaluation comes too late. Parachutes have never been formally and systemically evaluated, should we do that (Hopefully we aren’t in the control group)? Optimistically the pandemic appears to be on the wane. What is clear is that prior to vaccines, halfhearted and partial attempts at lock down with early opening up only lead to temporary suppression of the pandemic and recurrence on easing. This has led to the appalling death toll and pressure on health services in many parts of the world. Looking for special interest groups to exempt from a true international, national and community wide response could be part of the problem. We are all in this together.
Competing interests: No competing interests
Dear Editor,
This leading article is very timely, just as schools are about to return. But given the poor evidence for closing schools, and the known harms to children, it does beg the question why the government is introducing yet more mitigations. At a time when case numbers are now back to levels seen last September and with the added knowledge that everyone in the top 4 risk categories has been offered a vaccine which we are assured is highly effective, surely this is not only unnecessary but potentially harmful.
Firstly, the use of masks in classrooms will undoubtedly be detrimental to learning, particularly for any children with hearing impairments or any special educational needs. When the WHO first suggested masks in schools last August[1], the guidelines specifically stated that this should be accompanied by monitoring not only of any effect in reducing Sars-CoV-2 transmission but also of any harms to either mental or physical health, but this has not been done. There have been numerous publications highlighting potential risks, ranging from paradoxically increasing transmission[2], through to inhalation of microfibres[3]. A study of over 25,000 children in Germany[4] reported headache (53%), difficulty concentrating (50%), less happiness (49%), reluctance to go to school/kindergarten (44%), malaise (42%) impaired learning (38%) and drowsiness or fatigue (37%).
The second proposed mitigation is that of regular testing for secondary school pupils. Public Health England reports that lateral flow tests have around 99.7% specificity[5] which would be excellent for a test in a clinical situation. However, if testing 5 million secondary school pupils twice a week, those 10 million tests would be expected to generate 30,000 false positives. These children would presumably all be sent home from school, with their 30 classmates, leading to almost a million children incorrectly out of school each week.
Concerns have already been highlighted about children’s mental health[6], yet the combination of testing and then requiring masks in class even when tests are negative, will simply compound anxiety about Covid-19 in a population for whom the true risk is so small[7].
[1] Advice on the use of masks for children in the community in the context of COVID-19. 21/08/2021 https://www.who.int/publications/i/item/WHO-2019-nCoV-IPC_Masks-Children...
[2] Proposed mechanisms by which masks increase risk of Covid-19. Huber C. 08/01/2021 https://pdmj.org/papers/masks_false_safety_and_real_dangers_part4/
[3] Masks_false_safety_and_real_dangers_Part_1_Friable_mask_particulate_and_lung_vulnerability? Borovoy B, Huber C, Makeeta Q. https://pdmj.org/papers/masks_false_safety_and_real_dangers_part1/
[4] Corona children studies "Co-Ki": First results of a Germany-wide registry on mouth and nose covering (mask) in children. Schwartz S, Jenetzky E, Krafft H, Maurer T, Martin D. 05/01/2021
Corona children studies "Co-Ki": First results of a Germany-wide registry on mouth and nose covering (mask) in children | Research Square
[5] Preliminary report from the Joint PHE Porton Down & University of Oxford SARS-CoV-2 test development and validation cell: Rapid evaluation of Lateral Flow Viral Antigen detection devices (LFDs) for mass community testing: 08/11/2020. UK evaluation_PHE Porton Down University of Oxford_final.pdf
[6] Teenagers' mental health under severe pressure as pandemic continues - new research. 31/01/2021. Teenagers' mental health under severe pressure as pandemic continues - news research | Mental Health Foundation
[7] Use of “normal” risk to improve understanding of dangers of covid-19. Spiegelhalter D. 09/09/2020. Use of “normal” risk to improve understanding of dangers of covid-19 | The BMJ
Competing interests: This response is written in a personal capacity. RAJK is also a member of UsforThem in a voluntary capacity.
Re: Pandemic policy must address SARS-CoV-2 transmission in schools
Dear Editor
In their response Gurdasani et al [1] misrepresent our editorial and published studies in a manner that exaggerates risks from covid transmission in children.
First, we state that “school closures reduce the number of contacts children have and may decrease transmission”. We do not state, as they claim, “that school closures do not reduce SARS-CoV-2 transmission”. We argue that the overall contribution of children to community transmission is unclear and must be weighed against the detrimental impact of school closures. A recent randomized-controlled trial comparing the effectiveness of isolating contacts of cases for 10 days or allowing children to continue to attend school with daily lateral flow tests found that onward transmission of the virus among contacts within schools was low (around 1.6%)[2]. A large, well conducted, detailed study of transmission in Spanish schools, including testing of all exposed children within bubbles, found very low rates of transmission with a reproduction rate of 0.4 overall (ranging from 0.2 in preschool to 0.6 in secondary schools)[3].
Second, Gurdasani et al also misrepresent the data on serology in school transmission studies saying that children, “may be less likely to seroconvert than adults”. The study they cite demonstrates the opposite: rates of seroconversion to Spike IgG, the almost universally used assay in serosurveillance studies, including schools, was equal or higher to that found in adults [4].
Third, the risk to children and young people from covid is very small, with more than half of all infections being asymptomatic. Gurdasani et al’s claim that long covid is a common outcome and may result in significant disability relies on data from the ONS survey [5]. However that study did not include an appropriate control group; the controls had not been unwell, self-isolated, or otherwise tested for SARS-CoV-2, and represent an extraordinarily healthy comparator group. Moreover, as Bhopal and Absoud highlight, ONS figures on long covid symptoms and prevalence in children is in line with or lower than the usual population prevalence of these symptoms - an unexpected finding given the high level of community and family stress [6]. Well-designed studies vary from finding no difference in persistent symptoms between those testing positive for covid antibodies and controls in Switzerland [7]: to estimates of 2 – 4% of children with more persistent symptoms in Australia and UK [8,9,10]. It is worth noting the higher prevalence of post viral syndromes in other conditions, such as infectious mononucleosis, estimated to affect >10% of patients after 6 months [11,12,13]. Recent studies confirm an exceptionally low risk to children from acute infection, with 25 deaths - including only 6 among previously healthy children - from Covid19 during the first year of the pandemic in England [14].
Fourth, Gurdasani et al claim that “Having schools open without sufficient mitigations has a major impact on community transmission”. However the ecological studies they cite do not isolate the effect of in-school transmissions, from movement in the community [15,16]. Contrary to Gurdasani et al’s claims, schools opening on the 8th March 2021 in England were not followed by a surge in cases, despite blanket testing with lateral flow tests of secondary school children. It is a fact that prevalence of infection continued to decline following full reopening [17,18]. By the end of July 2021, primary school aged children had an estimated population prevalence less than half that of young adults, and even less than adults age 25 – 35 [19], over 90% of whom had received one dose of vaccine and nearly 40% being double vaccinated [19]. As in all previous waves, and as documented in other countries, young adults are the principle cohort driving community infection [20].
Fifth, Gurdasani et al state that schools were “largely closed” during the first wave in the UK and misleadingly imply that evidence of transmission in households with children in that period is irrelevant. However schools closed in the UK when infection levels were high during the first wave [21] indicating there was substantial transmission of infection in the community prior to closure, at a time when there were no mitigations in schools or isolation of contacts of cases. They then misrepresent the data in the OpenSAFELY study [22] . They state that that “living with children was associated with an increased risk of infection and hospitalization, and an increased risk of death from covid in adults over 65 years. This is despite parents living with children having much lower risk of non-COVID deaths compared to the control group, suggesting this group is healthier in other ways”. What Gurdasani et al should have reported is that non-covid mortality was lower for adults 65 years and younger living with children compared to the control group (those not living with children) and deaths from covid-19 were also lower to a similar degree. However, among adults over 65 years, in those living with children, non-covid deaths were not lower. Conflating two separate pieces of information in this way misleads the reader in a characteristic fashion, and yet again demonstrates that a long list of signatories/”co-authors” is no guarantee of accuracy.
The “essential prerequisite of all good public health measures is that careful estimates should be made of their advantages and disadvantages, for both the individual and the community, and that they should be implemented only when there is a significant balance of advantage” [23]. Mitigation measures, including school closures, must be evidence based and subject to rigorous scientific evaluation to establish the costs, benefits and harms to children. Rather than acknowledging uncertainties in the scientific evidence and need for evaluation, Gurdasani advocate policies which are not grounded in evidence [24]
The negative impacts of school closures are substantial and disproportionately affect the most disadvantaged children, seen recently in the widening inequalities in educational outcomes between independent and state school educated children. The pandemic and pandemic response casts a long shadow over their lifecourse. Non-evidence based school closures is one driver of this.
1. Gurdasani D, Besancon L, Alwan N. Pandemic policy must address SARS-CoV-2 transmission in schools. BMJ 2021;372: https://www.bmj.com/content/372/bmj.n521/rr-2
2. Young BC, Eyre DW, Kendrick S et al. A cluster randomised trial of the impact of a policy of daily testing for contacts of COVID-19 cases on attendance and COVID-19 transmission in English secondary schools and colleges. medRxiv 2021.07.23.21260992; doi: https://doi.org/10.1101/2021.07.23.21260992
3. Sergio A, Alvarez-Lacalle, E, Català, M et al. Age-dependency of the Propagation Rate of Coronavirus Disease 2019 Inside School Bubble Groups in Catalonia, Spain, The Pediatric Infectious Disease Journal 2021; July 27
doi: 10.1097/INF.0000000000003279
4. Weisberg SP, Connors TJ, Zhu Y et al. Distinct antibody responses to SARS-CoV-2 in children and adults across the COVID-19 clinical spectrum. Nat Immunol 2020; 22: 25-31
5. Office for National Statistics. Prevalence of ongoing symptoms following coronavirus (COVID-19) infection in the UK: 1 April 2021.
6. Bhopal S. and Absoud M. Vaccinating children to prevent long covid ? More caution is needed in interpreting current epidemiological data. BMJ 2021;372:n520. https://www.bmj.com/content/372/bmj.n520.
7. Radtke T, Ulyte A, Puhan MA, Kriemler S. Long-term Symptoms After SARS-CoV-2 Infection in Children and Adolescents. JAMA. Published online July 15, 2021. doi:10.1001/jama.2021.11880
8. Say D, Crawford N, McNab S, Wurzel D, Steer A, Tosif S. Post-acute COVID-19 outcomes in children with mild and asymptomatic disease. Lancet Child Adolesc Health. 2021 Jun;5(6):e22-e23. doi: 10.1016/S2352-4642(21)00124-3.
9. Molteni E, Sudre CH, Canas LS, et al. Illness duration and symptom profile in symptomatic UK school-aged children tested for SARS-CoV-2 [published online ahead of print, 2021 Aug 3]. Lancet Child Adolesc Health. 2021;S2352-4642(21)00198-X. doi:10.1016/S2352-4642(21)00198-X
10. Miller F, Nguyen V, Navaratnam AMD, Shrotri M, Kovar J, Hayward AC, Fragaszy E, Aldridge RW, Hardelid P. Prevalence of persistent symptoms in children during the COVID-19 pandemic: evidence from a household cohort study in England and Wales. medRxiv 2021.05.28.21257602; doi: https://doi.org/10.1101/2021.05.28.21257602
11. Hickie I, Davenport T, Wakefield D, Vollmer-Conna U, Cameron B, Vernon S D et al. Post-infective and chronic fatigue syndromes precipitated by viral and non-viral pathogens: prospective cohort study BMJ 2006; 333 :575 doi:10.1136/bmj.38933.585764.AE
12. White PD, Thomas JM, Amess J, et al. Incidence, risk and prognosis of acute and chronic fatigue syndromes and psychiatric disorders after glandular fever. Br J Psychiatry. 1998;173:475-481. doi:10.1192/bjp.173.6.475.
13. Buchwald DS, Rea TD, Katon WJ, Russo JE, Ashley RL. Acute infectious mononucleosis: characteristics of patients who report failure to recover. Am J Med. 2000;109(7):531-537. doi:10.1016/s0002-9343(00)00560-x
14. Smith C, Odd D, Harwood R et al. Deaths in Children and Young People in England following SARS-CoV-2 infection during the first pandemic year: a national study using linked mandatory child death reporting data, 07 July 2021, PREPRINT (Version 1) available at Research Square [https://doi.org/10.21203/rs.3.rs-689684/v1]
15. Li Y, Campbell H, Kulkarni D, et al. The temporal association of introducing and lifting non-pharmaceutical interventions with the time-varying reproduction number (R) of SARS-CoV-2: a modelling study across 131 countries. The Lancet Infectious Diseases 2021;21:193–202. doi:10.1016/S1473-3099(20)30785-4
16. Haug N, Geyrhofer L, Londei A, Dervic E, Desvars-Larrive A, Loreto V et al. Ranking the effectiveness of worldwide COVID-19 government interventions. Nature Human Behaviour 4, 1303-1312. 2020.
17. Office for National Statitistics. Coronavirus (COVID-19) Infection Survey, UK: 23 April 2021. https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/...
18. Office for National Statitistics. Coronavirus (COVID-19) Infection Survey, UK: 23 July 2021. https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/....
19. Office for National Statitistics. Coronavirus (COVID-19) latest insights: Antibodies- 21 July 2021. https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/...
20. Monod M, Blenkinsop A, Xi X et al. Age groups that sustain resurging COVID-19 epidemics in the United States. Science 2021;371 (6536), eabe8372.DOI: 10.1126/science.abe8372
21. Gov.uk. Coronavirus (COVID-19) in the UK. https://coronavirus.data.gov.uk/
22. Forbes H, Morton CE, Bacon S, et al (2021). Association between living with children and outcomes from covid-19: OpenSAFELY cohort study of 12 million adults in England. BMJ, n628. https://doi.org/10.1136/bmj.n628
23. Smith CEG. Prospects for the Control of Infectious Disease. Proceedings of the Royal Society of Medicine. 1970;63(11P2):1181-1190.
24. Davey Smith G, Blastland M and Munafo M. Covid-19’s known unknowns. 2020;371:m3979
Competing interests: SL has campaigned for schools to be reopened during the pandemic. AM was the lead fellow in commercial studies of vaccines and antibiotics funded by Merck Sharpe and Dohme, GlaxoSmithKline, Johnson & Johnson, Janssen, AstraZeneca, Novavax, and Valneva.