Daily numbers, graphs, analysis thread 24

[BigChocFrenzy]

Welcome to thread 24 of the daily updates

Resource links

UK:
Uk dashboard deaths, cases, hospitals, tests - 4 nations, English regions & LAs
UK govt pressers Slides & data
R estimates UK & English regions
Imperial UK weekly LAs, cases / 100k, table, map, hotspots
School statistics Attendance
ICNRC Intensive Care National Audit & Research reports
NHS t&t England & UK testing Weekly stats
Datasets for ONS surveillance reports
ONS Roundup deaths, infections & economic reports
Modelling real number of UK infections February to date

England:
NHS England Hospital activity
NHS England Daily deaths
MSAO Map of English cases
Cases Tracker England Local Government
ONS England infection surveillance report each Friday
ONS MSAO Map English deaths
PHE Surveillance reports & LA Local Watchlist Maps by LSOA
PHE surveillance reports Covid, flu, respiratory diseases
CovidMessenger live update by council district in England

Scotland, Wales, NI:
Scot gov Daily data
Scotland TravellingTabby LAs, care homes, hospitals, tests, t&t
PH Wales LAs, tests, ONS deaths
NI Dashboard

Miscell:
Zoe Uk data
ECDC rolling 14-day incidence EEA & UK
Worldometer UK page
Our World in Data GB test positivity etc, DIY country graphs
FT DIY graphs compare deaths, cases, raw / million pop
Alama Personal COVID risk assessment
Local Mobility Reports for countries
UK Highstreet Tracker for cities & large towns Footfall, spend index, workers, visitors, economic recovery

⏭ Our STUDIES Corner ⏮

We welcome factual, data driven and analytical contributions
Please try to keep discussion focused on these
📈 📉 📊 👍

D'you know I was just thinking 'I actually need to calculate probabilities here', this is far too simplistic (although technically if a school was perfectly reflective of community spread....).

I'm leaving @Nellodee to it as I suspect she's a far better mathematician than me...

[quote ceeveebee]From gov website - full details of restrictions and tiers
www.gov.uk/government/news/prime-minister-announces-new-local-covid-alert-levels[/quote]
Thanks

The spread is
2 x year 7
1 x year 9
2 x year 10
5 x year 11
1 x year 12
5 x staff.
Fairly evenly distributed.

@MarshaBradyo For a school of 1500, you would have a 41% chance of having zero cases, with that community transmission, and a 37% chance of having 1 case.

So, not at all unusual to have zero cases.

I just have a nice statistical calculator!

@Nellodee

The spread is 2 x year 7 1 x year 9 2 x year 10 5 x year 11 1 x year 12 5 x staff. Fairly evenly distributed.

@MarshaBradyo For a school of 1500, you would have a 41% chance of having zero cases, with that community transmission, and a 37% chance of having 1 case.

So, not at all unusual to have zero cases.

Very interesting. But are you saying our school is not unusual because it fits with what you’d expect from community transmission, but yours is unusual because it doesn’t?

I think the problem is that schools do not simply reflect community spread. The bigger the school, the more combinations of connections you can have.

It's my theory that the amount of connection combinations you have is just as important as age.

Primary schools, you have 30 people who can connect with each other. Under 14, you have a middling group of people who can connect with each other, because they don't have option groups. 15-16, you have many more connections.

This would be very simple to prove - simply compare the amount of cases in 17-18 year olds in sixth form colleges, as opposed to the amount of cases in 17-18 year olds in Further Education colleges, within the same local area. If its much higher in Further Education colleges, then it's the network effect that is the problem, not the age.

Well, I suppose every school fits with community levels until it gets its first case? It's what happens then that tells us how well it spreads in schools.

@Nellodee

Well, I suppose every school fits with community levels until it gets its first case? It's what happens then that tells us how well it spreads in schools.

Your spread looks evenly dispersed over the years, so not close contacts? I’m surprised you have so many cases and we don’t actually.

Am listening to the briefing. There is no mention in any of this about reintroducing shielding and protections for the vulnerable. Are we suppose dot just assume tough restrictions protect them, even if they are continuing, for example, attending their workplaces?

The year 11s are definitely a school related cluster. I'd lay my house on it being transmitted at least partially at school because it's jumped from a very Instagrammy bunch of girls to a very nerdy group of fortnite loving boys in the same class. I will eat my scientific calculator if they mixed outside school (sorry to dip into anecdata).

You shouldn't be surprised that we have more cases than you - we have higher community rates and a bigger school. And 6 days ago, we had no cases here either.

Your spread looks evenly dispersed over the years, so not close contacts? I’m surprised you have so many cases and we don’t actually

This is really not surprising, cases are not evenly distributed, cases are obviously heavily clustered, so the probably of if a school has one case, then it's more likely to have others even if there was no school transmission.

Cases are not randomly distributed among the community.

The echo have published this - This makes it clear that the lest section is where all the negotiations are

@sirfredfredgeorge

Your spread looks evenly dispersed over the years, so not close contacts? I’m surprised you have so many cases and we don’t actually

This is really not surprising, cases are not evenly distributed, cases are obviously heavily clustered, so the probably of if a school has one case, then it's more likely to have others even if there was no school transmission.

Cases are not randomly distributed among the community.

Why might one school have higher cases? Even if no school transmission

Not saying I disagree just interested in why this might be so

@Nellodee

The year 11s are definitely a school related cluster. I'd lay my house on it being transmitted at least partially at school because it's jumped from a very Instagrammy bunch of girls to a very nerdy group of fortnite loving boys in the same class. I will eat my scientific calculator if they mixed outside school (sorry to dip into anecdata).

You shouldn't be surprised that we have more cases than you - we have higher community rates and a bigger school. And 6 days ago, we had no cases here either.

Oh right you’re at 184 thought you said 80 for some reason

Yes it’s higher so you’ll see that reflected

Too much going on here

Overall I don’t think it’s surprising that schools reflect community levels

And since it’s in a variety of years not one year I’d say it doesn’t particularly show in school spread

@Nellodee
could you calculate for a school of 1600? It's in Richmond were the rate per100k is 108.

How many cases would we expect based on this?
Actual number of confirmed cases is 4, but I know of 3 kids who were home for 2wks and couldn't get a test.

Just had an evasive answer from Chris Whitty on shielding.

Did Chris Whitty just say that tier three isn't enough?

[quote ceeveebee]From gov website - full details of restrictions and tiers
www.gov.uk/government/news/prime-minister-announces-new-local-covid-alert-levels[/quote]
That list states areas rather than council?

My in-laws are in a council near those areas but different council which had been under restrictions for several weeks. So clear as mud whether they are in the high tier or not....

Only time will tell when the plug your postcode in thing is launched

Yes regulus. It doesn't feel like 'clarity' to me, all this.

Why might one school have higher cases? Even if no school transmission

Because cases are not evenly distributed throughout the community, a particular cluster of people will have a higher chance - let's imagine a virus where the only 2 places it can be transmitted is the pub and the home. Three of the twenty pubs in the town have cases, will the children of those pubgoers be equally likely to go to the 10 schools, or is it actually likely that the same way that their parents go to particular pubs, they'll also go to particular schools.

If cases were randomly distributed, it would be impossible for "track and trace" to achieve anything, but viruses are never randomly distributed as they rely on contacts.

There seems to be very little difference between the tiers. I would have expected there to be more difference.
Can't see it making much difference, but hopefully I'll be proved wrong.

"People must not meet in groups larger than 6, indoors or outdoors"

and

"People must not meet in a group of more than 6 outside, including in a garden or other space"

I don't understand the difference?

@sirfredgeorge is absolutely right that cases are not evenly distributed, but are in clusters. I would say once you have 1 case, your chances of having another increase, and with every additional case, they increase more.

This said, for @TheSunIsStillShining working on a very basic model that DOES assume independence, to work out the most probable amount of cases, you just take your probability (0.00108) and multiply it by the amount of people in your school. That would give you 1.7 cases. Of course, you can't get 1.7 cases. The probability of getting exactly 1 case would be 31% and the probability of getting exactly 2 cases 27%.

The probability of having 4 or more cases is 10%, so its quite a high number, but not in any way unusual. p(X>3), B(1600, 0.00108)

The probability of having 5 or more would be 3%, and 6 or more would be 1% only. If you get to 7 cases, then you are definitely seeing spread at a much higher rate than the rest of your area.