Are schools actually the main spreader of covid?

[Newyearsamecovid]

People on here are constantly saying that schools are the biggest spreader and to start I accept the fact that obviously there will be spread in schools.

However here we are, schools have been closed 10 days and yet we have the highest confirmed cases on record. I appreciate people may have mixed over Christmas, but from what’s being said, 30 separate kids from different households being crammed in a classroom with no social distancing would surely be more spreadable than say 2/3 households round the dinner table?

Looking at the stats, infection rates for all children are consistently quite far behind the infection rates of 20-40 year olds also - who are not at school.

Given schools are one the only places open, you’d expect it to be a very different picture. You would expect schools to be the place everyone is catching it but the figures just don’t seem to point to this.

So why do we think schools are the driving factor in spreading covid?

Its been all the teachers in my sons school we had two bubbles burst due to one family in two different year groups the rest of the time the school closed year groups due to lack of teaching staff they all went back in and three days later the entire slt came into contact with a positive case and the school was shut down completely

I work in a massive secondary school. We had a flurry mid term of positives. (No more than about twenty kids in total though). Followed by about six or seven staff. We've also tested some students randomly (paid for test) who show no signs at all and who got positive results back.
Incidentally I have family members in their 90s who have tested positive and feel ok. Friends in their 40s have been bedridden for three weeks with it.

You're looking at the wrong data, OP.

Infection rate data from random sampling is the best thing to look at for children because they are more often asymptomatic or don't show the 3 adult symptoms that trigger a test (child symptoms are different but the government has not advertised this).

The ONS random sampling data shows that secondary school children are the most infected subset of the population. Infection rates in primary have also shot up, while infection rates in sixth form and uni students that were incredibly high at the start of term when uni students moved around the country dropped dramatically but have now started to rise again.

Attached is the latest infection rate data. More up to date data should have been published yesterday but has not, so we cannot see the impact of school closures on infection rates in school kids. I'm not sure if this is deliberate.

We have had many more kids test positive than staff in my primary school.
Primary infections are not far behind secondary at all.

@noblegiraffe presumably less 2-16 year olds would have had immunity by September than older age groups since they would have been less likely to pick it up at the shops etc and therefore with more mixing for a variety of reasons in September they had a faster rate of increase as fewer of them had immunity? So that wouldn’t mean it was necessarily spreading more in schools than in other places. Uni is a much less controlled environment than schools and involves the mixing of living spaces between households so I would expect the yr 12 - 24 category to have a spike then. Shame they didn’t separate that category into sixth form, uni, and young adult

The difficulty we have in containing the virus is precisely because there is no "main source" or single big thing we can do to prevent spread. Containment requires many smaller, disruptive and sustained measures. That's the trouble.

But @noblegiraffe doesn’t this coincide with the lockdown during which schools remained open, such that these age groups at the time were the most likely to be exposed given that others weren’t meant to be mixing. So it doesn’t really tell us about relative risk all other things being equal?

Covid spreads where people gather. It's that simple.

The more precautions you take against it, the less it will spread. So an unmasked dinner with 2 to 3 households in a heated home in December is risky. 20 minutes in a numbers-limited supermarket with social distancing and universal mask use, less so. (Yeah, I know, it ain't that ideal in reality!)

Then we have 25 to 30 unmasked children in a classroom for hours, variable handwashing, limited social distancing. That is very obviously a great way to spread an airborne virus.

The main spreader of Covid is a combination of inaction by the government at various stages since early March, and not requiring face coverings or shields by all without exception.

Easy to blame schools as Gavin Williamson and Boris Johnson would like to do, or people for not being responsible.

Did the government close airports in March or have the kind of quarantine that some countries did (hotels)? No.
Did the government set an example of wearing face coverings when arriving at press conferences or in the House of Commons. No.
Did the government end mass gatherings and stop 3,000 people from the Covid hotspot of Madrid visiting? No.
Did the Prime Minister sack Dominic Cummings and Robert Jenrick? No.
Did the government stop 1.2 million students going back to in-person university tuition when their own graphs showed 18-21 year olds with at the time the highest rate of infection? No.

The list goes on.

Here is a great Twitter thread gathering a lot of evidence around spread in schools:

mobile.twitter.com/apsmunro/status/1337426764189020160

Bear in mind that this predates the new strain B117 and it is too early to have clear evidence on the effect of this.

However, it shows that younger children really do seem to play a smaller role than other in transmission than other groups and that in general that schools don’t appear to be drivers of transmission, but obviously transmission does occur there.

@ATieLikeRichardGere

But *@noblegiraffe* doesn’t this coincide with the lockdown during which schools remained open, such that these age groups at the time were the most likely to be exposed given that others weren’t meant to be mixing. So it doesn’t really tell us about relative risk all other things being equal?

Look at the graph. Secondary kids are massively more infected than the adult population and have been since mid September.

The dip in infection rates in secondary kids coinciding with half term shows that being in school is a big area of transmission for them.

We know from PHE data that schools are the most likely place where someone testing positive had a positive close contact.

We know that covid spreads best indoors in poorly ventilated cramped environments with no masks. This makes classrooms the perfect conditions for transmission.

Kids are spreading covid in school and taking it home to their families.

I think, very understandably, in this dire situation we are all searching for silver bullets, myself included. For some, the silver bullet of choice has become to close schools. Unsurprisingly, the reality is more complicated and difficult to untangle. Now that the new strain has a naturally increase R number we may have to close schools from the point of view that we have very few tools left in the arsenal to control transmission - many are already fully in use. That is, until vaccination can begin to make a real dent. However, that shouldn’t be confused with the notion that schools have been key drivers of transmission.

@noblegiraffe I agree that it looks like keeping older children home in particular could be a useful tool, especially because it is of lower cost than some other measures, such as keeping younger children home. But we can’t draw too many conclusions from a single graph.

But we can’t draw too many conclusions from a single graph.

What conclusions are you trying to avoid drawing?

I mean, we know secondary school kids are indoors mixing freely daily in ideal conditions for covid transmission and are also the most infected subset of the population.

Are we supposed to pretend that the two are unconnected?

If you lived in our area it looks pretty obvious.

We were low, until just before half term, when we started rising slowly. There had been maybe half a dozen cases in school (large secondary) and similar in the other secondaries.

Then the week before half term over 50% of ds' year was isolating due to three cases, and the other secondary schools had similar situations.

Two weeks later they went back. Then by the end of November of the three secondary schools, one was totally closed and the others had at least 50% off, due to numbers of cases and staff absences.
You can see the point on the heat map, where the 10-14 age category goes into darker colours.

But then here's the interesting point:
That age category eases as there's fewer in schools to pass it on... but then 2 weeks later you can see the categories 45-60 ages shoot up-the parent's age.
Two of the schools get back for last couple of weeks in December before Christmas, and after a week you begin to see that rising slowly again... but the primaries, who now haven't generally been off, are shooting up.

If you look at the local heat map with local information about school closures it really does follow a pattern.
Schools open, starts to increase. Schools closures, decreasing, then rise in parents' ages. Then reopen and back the increases start again.

Fascinating from a statistics prospective. Unfortunately rather scary in the real world.

@starrynight19

We have had many more kids test positive than staff in my primary school. Primary infections are not far behind secondary at all.

Whereas we've had barely any cases at all, in infants or juniors, amongst staff or pupils.

We had a Y4 and a Y6 case plus an infant teacher isolating because of contact with somebody who tested positive. And that's in schools with 120 pupils per year group.

In think to draw better conclusions we would need to drill down to more local data as well.

So Richard you think that despite school kids mixing freely in the ideal conditions for covid transmission and after schools went back the infection rate shooting up in that group will they're the most infected subset of the population, with a dip in infection rates correlating with half term when schools were closed, more info is needed?

Bear in mind that this predates the new strain B117 and it is too early to have clear evidence on the effect of this.

So therefore it’s out of date no ?

We have high infection rate 10-14 days after schools closed as it takes up to 10 days for someone to catch it after exposure. That’s why isolation period is 10 days. I tested positive 8 days after exposure to a covidvpositive child.

Being devil's advocate here, it isn't that straightforward to interpret causality. For example, you could have a lot of children being infected at the same time by a single teacher, but then relatively few of those children actually passing on the infection. It would look like the highest infection rate was amongst the children and therefore that they were the driver of infection and as such "the problem", but they might not actually be the drivers.

Right now, it could be that chance has caused the more infectious new strain to become seeded in schools because at a certain time when this strain emerged, they were open while everything else was locked down. Thus the infection rate among pupils where the new more infectious strain is more seeded is higher than among other age categories who are primarily sharing a less infectious strain. This could again skew our interpretation of what is happening. Over time obviously, the more infectious strain one will bleed from schools into other populations and it could therefore be said that schools are causing this, but it wouldn't necessarily lead to the same policy conclusions.

To be clear, I think that given then unknowns we should proceed with caution and consider school closures among other measures, but obviously these are also really high cost so it's not something you would want to do.

I'm just saying in general we have to also be really cautious about interpreting things because what can seem really obvious might not even be real at all. At the same time, we don't always have the luxury of time and answering the unknowns - I recognise that too.

There's so many great viral epidemiologists and similar to follow on Twitter. I recommend Muge Cevik and Adam Kucharski as some of these.

Thus the infection rate among pupils where the new more infectious strain is more seeded is higher than among other age categories who are primarily sharing a less infectious strain.

The infection rate in secondary pupils was way higher than other age categories way before this new strain became a thing.

For example, you could have a lot of children being infected at the same time by a single teacher, but then relatively few of those children actually passing on the infection.

Which would be bizarre. We know how covid spreads but we are supposed to ignore what we know about covid spreading when it comes to kids sat shoulder to shoulder not wearing masks in poorly ventilated indoor spaces?

There's so many great viral epidemiologists

Do they think that covid suddenly behaves differently in a classroom? If so, then I won't be giving them a follow.

They’re not blanket testing.
Most people are asymptomatic, therefore I would imagine most school kids have had it or will spread it. During Xmas it will gave been more adults mixing and they get symptoms and test.

My daughters school closed. All teachers were tested, one was symptomatic. They all but one tested positive.
Out of my daughters class, about three children got symptoms. Seven decided not to test as per government guidelines as they did not have symptoms. The rest of us decided to anyway. Out of those of us that tested, one child tested negative.

My daughter tested positive but was asymptomatic, we then got tested as a family and were all positive. I was the only one to develop symptoms, even then very mild.

As a family we ended up isolating for a month until everyone was clear. However if we’d followed the government guidelines my daughter would have been the only one to isolate for fourteen days following her contact. The rest of us would not have had to as she had neither symptoms or would have tested positive.

Now can you imagine how many families are going about their business regardless. From just my daughter’s class there were seven families that were doing just that and may or may not have been carriers.

@noblegiraffe It wouldn't be bizarre if children were less likely to pass on infection - prior to the new strain at least, this is what the evidence points to. There are quite a few different plausible explanations for this, to do with children's immune systems, prior coronavirus exposures, their likelihood to be asymptomatic etc... This isn't the same as saying that they don't spread the virus, just that they may be less likely to.

One thing to point out is that evidence suggests true asymptomatic infections (not pre-symptomatic) do spread the virus but they are less likely to do so.

The epidemiologists would agree that classrooms can be problematic environment's for spreading covid for sure, but perhaps not that they should be considered the key driving factors. There are a lot of other things to point fingers at as well.

We don't need to be dogmatic about this. I think everyone is frustrated and worried that not enough is being done.

On the plus side, we now have more stable PPE supplies, better covid treatments and two approved effective vaccines so there are reasons to be positive despite the ugly numbers - which are definitely ugly.

Sorry not sure why I've got a rogue apostrophe in environments.

I just think there is a lot of panic and misinformation and bias and emotion and we should take a step back. I know the government are rubbish but nonetheless, scientists with a much better grasp of all this than we do are working on it.