Read on Twitter
It's time to clarify some things about children, schools and #COVID19 
Summary: Young children seem significantly less susceptible, probably less likely to transmit. Less clear for teens. Schools mainly follow community trends, but secondary much higher risk than primary
1
Summary: Young children seem significantly less susceptible, probably less likely to transmit. Less clear for teens. Schools mainly follow community trends, but secondary much higher risk than primary
1
The best way to determine susceptibility is through household contact tracing, as it controls for *exposure* - everyone gets more or less the same
There are many of these. Results vary, which we expect because infection is complicated
That's why we need to combine results
2
There are many of these. Results vary, which we expect because infection is complicated
That's why we need to combine results
2
Here's 4 meta analyses; all find young children are much less susceptible than adults. Some that teens are too
Zhu (RR 0.6)
https://academic.oup.com/cid/advance-article/doi/10.1093/cid/ciaa1825/6024998#.X821iNvCv-E.twitter
Goldstein (RR~0.5)
http://dx.doi.org/10.1093/infdis/jiaa691
Viner (OR 0.41)
https://jamanetwork.com/journals/jamapediatrics/fullarticle/2771181
Maidwell (RR ~0.5)
https://www.medrxiv.org/content/10.1101/2020.07.29.20164590v1
3
Zhu (RR 0.6)
https://academic.oup.com/cid/advance-article/doi/10.1093/cid/ciaa1825/6024998#.X821iNvCv-E.twitter
Goldstein (RR~0.5)
http://dx.doi.org/10.1093/infdis/jiaa691
Viner (OR 0.41)
https://jamanetwork.com/journals/jamapediatrics/fullarticle/2771181
Maidwell (RR ~0.5)
https://www.medrxiv.org/content/10.1101/2020.07.29.20164590v1
3
What if cases in children were missed because they were asymptomatic?
Here are just a few which test ALL household contacts regardless of symptom, and all find lower susceptibility in children
Zhang
https://science.sciencemag.org/content/368/6498/1481
Jing
https://www.thelancet.com/journals/laninf/article/PIIS1473-3099(20)30471-0/fulltext
4
Here are just a few which test ALL household contacts regardless of symptom, and all find lower susceptibility in children
Zhang
https://science.sciencemag.org/content/368/6498/1481
Jing
https://www.thelancet.com/journals/laninf/article/PIIS1473-3099(20)30471-0/fulltext
4
Somekh
https://journals.lww.com/pidj/Fulltext/2020/08000/The_Role_of_Children_in_the_Dynamics_of_Intra.30.aspx
Rosenburg
https://academic.oup.com/cid/article/71/8/1953/5831986
Sun
https://science.sciencemag.org/content/early/2020/11/23/science.abe2424
And there's more...
5
https://journals.lww.com/pidj/Fulltext/2020/08000/The_Role_of_Children_in_the_Dynamics_of_Intra.30.aspx
Rosenburg
https://academic.oup.com/cid/article/71/8/1953/5831986
Sun
https://science.sciencemag.org/content/early/2020/11/23/science.abe2424
And there's more...
5
What if we're missing cases by PCR testing because it's difficult to NP swabs in children (even though nasal swabs are ~90% as sensitive)?
We have studies using serology. One found similar rates in children and adults
https://academic.oup.com/cid/advance-article/doi/10.1093/cid/ciaa1721/5979490
But there are more studies...
6
We have studies using serology. One found similar rates in children and adults
https://academic.oup.com/cid/advance-article/doi/10.1093/cid/ciaa1721/5979490
But there are more studies...
6
All the following tested everyone with serology, and STILL found lower rates in young children
van Der Hoek
https://www.ntvg.nl/artikelen/de-rol-van-kinderen-de-transmissie-van-sars-cov-2
Lewis
https://academic.oup.com/cid/advance-article/doi/10.1093/cid/ciaa1166/5893024
Bi
https://www.medrxiv.org/content/10.1101/2020.11.04.20225573v1
7
van Der Hoek
https://www.ntvg.nl/artikelen/de-rol-van-kinderen-de-transmissie-van-sars-cov-2
Lewis
https://academic.oup.com/cid/advance-article/doi/10.1093/cid/ciaa1166/5893024
Bi
https://www.medrxiv.org/content/10.1101/2020.11.04.20225573v1
7
On to infectiousness; this is difficult to study for 2 reasons
1) We mainly examine infectiousness of symptomatic children, who are likely more infectious. If ~50% of children are asymptomatic, this is not representative
2) Shared exposure bias - this is worth exploring...
8
1) We mainly examine infectiousness of symptomatic children, who are likely more infectious. If ~50% of children are asymptomatic, this is not representative
2) Shared exposure bias - this is worth exploring...
8
This is when 2 people became infected at the same time, but you mistakenly think one of them infected the other
It's a huge issue for children (who rarely travel alone). It completely reversed the findings of a study from South Korea when accounted for
9 https://twitter.com/apsmunro/status/1292852036720091136?s=20
It's a huge issue for children (who rarely travel alone). It completely reversed the findings of a study from South Korea when accounted for
9 https://twitter.com/apsmunro/status/1292852036720091136?s=20
The "biggest ever" contact tracing study from Southern India did not account for this bias
That means we can draw very little from the findings of infectiousness in children; did they infect each other, or all get infected at once?
10 https://twitter.com/apsmunro/status/1311616478844780544?s=20
That means we can draw very little from the findings of infectiousness in children; did they infect each other, or all get infected at once?
10 https://twitter.com/apsmunro/status/1311616478844780544?s=20
Note: because of overdispersion, if you find multiple close contacts infected it may even be MORE likely they all became infected at once than infected each other, because they were likely infected by a super spreader
What else do we know about infectiousness?
11
What else do we know about infectiousness?
11
The gold standard in contact tracing is using genomics, which can rule out transmission links if viral genomes are different
Iceland have done this well, and found young children to be significantly less susceptible AND infectious than adults
https://www.nationalgeographic.com/science/2020/12/we-now-know-how-much-children-spread-coronavirus/
12
Iceland have done this well, and found young children to be significantly less susceptible AND infectious than adults
https://www.nationalgeographic.com/science/2020/12/we-now-know-how-much-children-spread-coronavirus/
12
What about schools?
Prior evidence has shown in low community prevalence, schools are safe
Also that when community prevalence is high, secondary school is much higher risk for transmission
This has not changed
But are schools driving infection in the second wave?
13
Prior evidence has shown in low community prevalence, schools are safe
Also that when community prevalence is high, secondary school is much higher risk for transmission
This has not changed
But are schools driving infection in the second wave?
13
When we say "driving" infection, we usually mean the group with the highest prevalence, or that accelerates transmission first, driving up rates for the others
This is the case for young children and schools for other respiratory viruses, but has not been for COVID-19
14
This is the case for young children and schools for other respiratory viruses, but has not been for COVID-19
14
All across Europe, *young adults* have clearly driven the second wave of infection, which started before schools even opened
https://www.medrxiv.org/content/10.1101/2020.11.11.20230177v1.full.pdf
Teens have followed up behind, and slightly older adults after that
But then we've done something strange...
15
https://www.medrxiv.org/content/10.1101/2020.11.11.20230177v1.full.pdf
Teens have followed up behind, and slightly older adults after that
But then we've done something strange...
15
Many places have then closed down almost all other areas of society for mixing *Except schools*
If you stop mixing everywhere else, then you will get a disproportionate number of cases in schools, as in England; note this is mainly SECONDARY schools
16
If you stop mixing everywhere else, then you will get a disproportionate number of cases in schools, as in England; note this is mainly SECONDARY schools
16
Despite substantially higher rates of contacts and mixing than adults, young children have had infection rates the same or even lower than older adults throughout the second wave via *UNBIASED random population testing*
This is convincing evidence of reduced susceptibility
17
This is convincing evidence of reduced susceptibility
17
What about modelling studies showing school closure is effective?
We would expect it to reduce infection rates, because it would prevent 10million people in the UK from mixing; but how much is unclear
There are many biases difficult to account for
18
We would expect it to reduce infection rates, because it would prevent 10million people in the UK from mixing; but how much is unclear
There are many biases difficult to account for
18
e.g. simultaneous implementation of NPIs, and as school closures happen earlier than other NPIs it exaggerates their effect
Some modelling studies have even shown school opened reduced transmission rates
https://www.cream-migration.org/publ_uploads/CDP_22_20.pdf
Take all these studies with a handful of salt
19
Some modelling studies have even shown school opened reduced transmission rates
https://www.cream-migration.org/publ_uploads/CDP_22_20.pdf
Take all these studies with a handful of salt
19
Further evidence that schools aren't "driving" transmission, is that many countries in Europe have brought cases down whilst keeping schools open
Schools certainly contribute to transmission, but do not drive it
https://www.ft.com/content/7e1ad517-762b-4212-a6cb-79579e619409#
20
Schools certainly contribute to transmission, but do not drive it
https://www.ft.com/content/7e1ad517-762b-4212-a6cb-79579e619409#
20
What does this all mean?
-Schools mainly reflect community transmission
-If you close everything except schools, you'll obviously see a disproportionate number of cases in schools
-Secondary schools are much higher risk than primary schools
What should we do about this?
21
-Schools mainly reflect community transmission
-If you close everything except schools, you'll obviously see a disproportionate number of cases in schools
-Secondary schools are much higher risk than primary schools
What should we do about this?
21
We must recognise the important of in person schooling; prolonged closures will deny children future life chances that cannot be reversed
They should be last to close, first to open
How can we facilitate this?
22 https://twitter.com/apsmunro/status/1336341769043238912?s=20
They should be last to close, first to open
How can we facilitate this?
22 https://twitter.com/apsmunro/status/1336341769043238912?s=20
We must do several things:
-Keep community transmission low to keep cases out of schools
-Provide resources to enable safe behaviour, especially for young adults
-Improve school infection mitigation, especially in secondary (ventilation, masks, testing etc)
END
23
-Keep community transmission low to keep cases out of schools
-Provide resources to enable safe behaviour, especially for young adults
-Improve school infection mitigation, especially in secondary (ventilation, masks, testing etc)
END
23
BONUS TWEET
For an overview of all the evidence regarding children and COVID-19, we have been reviewing it all since the beginning here on @DFTBubbles https://dontforgetthebubbles.com/evidence-summary-paediatric-covid-19-literature/
For an overview of all the evidence regarding children and COVID-19, we have been reviewing it all since the beginning here on @DFTBubbles https://dontforgetthebubbles.com/evidence-summary-paediatric-covid-19-literature/
