Thread by @BenChu_

Is Covid really being spread significantly in and by schools? 🦠🏫🧑‍🎓

Thread 🧵

The decision by the government to relay the school return for secondary pupils by a week suggests ministers now believe so.

But what’s the evidence? 1/

The basic facts are not disputed.

The Office for National Statistics’ large-scale and random weekly survey shows that rates in the run-up to Christmas were considerably higher for school age children than adults...2/

ons.gov.uk/peoplepopulati…

Another large and random survey by researchers at Imperial College London, known as REACT, has been showing a similar picture.

Prevalence in school age children is roughly double the rest of the population...3/

spiral.imperial.ac.uk/bitstream/1004…

A modelling study released by the London School of Tropical Hygiene and Medicine (LSHTM) on 23 December takes as its assumption that children mixing in schools are indeed spreading the virus....4/ cmmid.github.io/topics/covid19…

And the LSHTM authors conclude unless the government closes primary and secondary schools and also universities the R number will not fall below 1 and the outbreak will continue to worsen.

That's may be behind the reported advice from SAGE on schools...5/ politico.eu/article/corona…

Yet how secure is the crucial modelling assumption that schools have been accelerating the spread of the virus?

As well as its national survey, the ONS also conducts a specific testing study of English schools, working with Public Health England...6/

ons.gov.uk/peoplepopulati…

This study found that the proportion of schoolchildren and teachers with coronavirus “closely mirrors” the proportion in the local community.

So schools in high prevalence areas have high prevalence and schools in low prevalence areas have low prevalence...7/

Dr Shamez Ladhani, a PHE consultant epidemiologist who ovesees the study, argues causality might run in the other direction – in other words virus being brought into schools from outside, rather than being spread by transmission between pupils...8/
uk.finance.yahoo.com/news/school-co…

Some other European countries (notably France) have been able to reduce infection growth while keeping their schools open.

Yet the evidence AGAINST school transmission is rather speculative at the moment...9/

The authors of the ONS schools study concede that the results of the first round were based on a low sample, covering just 105 schools (63 secondary and 42 primary) - and the results relate to November...10/

And the fact that during the November lockdown cases rose in schools in London while they declined in the rest of the population does support the case that schools are themselves responsible for transmission....11/

Also note that infections appeared to increase more in the second national lockdown in November compared to the first in March and April, when schools were required to close....12/

Ideally we would, of course, wait for more data.

But the danger of waiting for more conclusive evidence before closing schools is that the disease could accelerate rapidly if it is being driven by them....13/

And on the question of harm to education, if schools are kept open and the disease worsens they are likely to be closed ultimately anyway....14/

As with handling the disease more generally, the argument in favour of acting decisively and early on closing schools is that it is, in the medium term, the best way to protect the education of children....15/

Conclusion: the evidence is not clear cut and it's a tough policy call.

But vital, as always in this emergency, to bear in mind the relative risks from acting too late.

More here for @indypremium

ENDS

independent.co.uk/independentpre…

ADD:

Spoke to Mike Tildesley, a member of the Government’s advisory group on pandemic modelling.

He's also sceptical of the claim schools are significantly driving infections in the community (rather than the other way around)