Reopening schools may cause COVID spike

With the UK making plans to reopen schools on 8th March, scientists have tried to predict the consequences...
23 February 2021

Interview with 

Deepti Gurdasani, Queen Mary University of London


A classroom.


How much of a pandemic risk is reopening schools? With the UK making plans to do so on 8th March, two new studies have tried to answer this question. Both are preprints, so haven’t yet been peer-reviewed; one is from the London School of Hygiene and Tropical Medicine, the other is from the University of Warwick; and they reach almost opposite conclusions. Phil Sansom asked epidemiologist Deepti Gurdasani - who’s independent of both studies - to take a look, starting with the one from the London School of Hygiene and Tropical Medicine…

Deepti - What the study showed was that if schools were opened, we would very likely, in most scenarios, see the pandemic go from shrinking to starting growing again, and cases start growing exponentially like we've seen before.

Phil - What assumptions have they made going into this model, to make this prediction?

Deepti - The assumptions made are about how likely a child is to get infection when exposed compared to an adult. That's what we refer to as susceptibility. And it assumes a susceptibility of between 30% - which means a child is much less likely than an adult to get infection when exposed - compared to equal susceptibility, which is 100%. Irrespective of the assumptions made, what we find in almost all scenarios is that R would rise above one. The study also states that the results may depend on how dominant the new variant is in different parts of England.

Phil - Right, so what are we to make of this? Is opening schools going to put us back in outbreak city?

Deepti - I think it's very likely. I mean, it's not just evidence from this model. If we look at the real world evidence from England: if we go back to November, when we had a national lockdown and schools were open, areas where the new variant was dominant - which is I guess, comparable data to now, because the variant is dominant across the UK - R rates in those places were higher than one, despite a national lockdown. There's no reason to think that if we open schools in exactly the same way as they were open at that point in time, with a new variant circulating, the results would be any different, unless we make sure that we open them in a different way,

Phil - This sort of contrasts to the other preprint that we have to look at, which is from Warwick. And they seem to say something different. What are they saying?

Deepti - They're looking at the rates of absenteeism - when children are absent from school because of COVID-19. And it's looking at the rates over autumn and winter last year, and trying to correlate that rate with when community transmission rises.

Phil - So they're using the real figures of how many kids were absent around October/November, and they say, "we conclude that there is not significant evidence to suggest that schools are playing a significant role in driving spread in the community." Where are they getting that from?

Deepti - They didn't find evidence that, for example, absenteeism rates rose before there were rises in infection in the community, which was interpreted as, "no strong evidence to suggest that schools were contributing to community transmission".

Phil - What are we to make of this then? Especially because the other preprint, they said if you open schools we're going to get a bunch of new infections.

Deepti - Yes. I think the problem is that studies that look at absenteeism as a proxy for infection are problematic. Most children who develop infection are actually asymptomatic. They don't develop symptoms. And that means that they're very unlikely to become absent. We actually have very good studies now in the UK that don't suffer from these flaws. They are on random surveys of the population, which means that they include children who are asymptomatic as well as children who are symptomatic.

Phil - What data is this? Where are we getting this?

Deepti - The Office for National Statistics household survey data clearly show that children contribute to transmission. Primary school children were two times as likely as adults, and secondary school children was seven times more likely than adults, to be the first person in a household infected. Once infected, both those groups were two times more likely to transmit to household members compared to adults. And that shows that children have played a very important role in transmission.

Phil - Kids are really struggling. Is there any way to do this right?

Deepti - Yes, absolutely. I mean, the impact of school closures on children is huge. Our current policy in schools is actually completely out of line with the evidence on aerosol transmission. The CDC actually recommends using masks in secondary schools, but also encouraged in primary schools, particularly in high transmission areas. This is something that's being done in many parts of the world. In France and Spain, for example, all children six years and above wear masks in schools. The second area that hasn't received much attention in the UK is ventilation. While there have been guidelines to, for example, open windows and doors to improve ventilation settings - which definitely does help - there's not been any formal attention to this like there has been other countries like Germany. We know, for example, that filtration devices and air purifiers can really reduce the amount of aerosol in a classroom; and monitoring devices like carbon dioxide monitors can help us assess whether we're reaching the level of ventilation that we really need to keep people safe in those environments. The other issue we have in schools, particularly in England, is the class sizes are actually one of the largest ones in Europe. In many schools, bubble sizes reach to the hundreds. But to reduce this, of course we need more resources; so we need more staff, larger spaces... I'm really worried that them reopening without those safety measures in place will lead to a rise in R above 1, which means we may actually have to go into another lockdown and close schools again, which I think would be devastating.


Add a comment