Data and analysis thread, started 12 November

[NoGoodPunsLeft]

Previous thread here:

Data and Analysis Thread, started Oct 29 www.mumsnet.com/Talk/coronavirus/4064113-Data-and-Analysis-Thread-started-Oct-29

Regular lurker but I frequent poster, didn't want to lose the threads.

@Firefliess

It's just certain postboxes *@snowman*. I noticed a sticker on my local post box yesterday saying it was a priority post box and was wondering what that was about. So you need to look out for the stickers.

Priority post boxes are the only ones you can post your covid tests in, I believe. They are now going to be collected on Sundays - before you basically couldn't post them on a Sunday so if you requested a test on a Friday, say, you wouldn't have any conceivable way of sending it back until the next Monday. Sunday collections will hopefully really help...? Or perhaps the delay might have meant some cases were caught that would have otherwise not have developed enough to register as positive?!

Completely bonkers number in this LA today. Going to check later...!

From the dashboard "about the data" info: Cases are allocated to the person's area of residence. From 16 November 2020, PHE has updated the way it records the location of people who test positive or negative for COVID-19. It now prioritises addresses given at the point of testing over the details registered on a patient’s NHS Summary Care Record. This better reflects the distribution of cases and testing. However, it may give rise to differences in previously reported numbers of cases and rates in some areas. The change has been retrospectively applied to tests carried out from 1 September 2020, and data in the dashboard was updated to reflect this change on 16 November 2020. Due to reallocation of cases in this way, the number of cases reported by local authority may be artificially high or low on 16 November 2020.

So looks like they've finally got on top of the issue people were flagging up weeks and weeks ago about students being recorded at their GP address (often their home address not uni) rather than where they were when they got infected. It's all a bit odd today though - most places are saying zero cases, except Manchester with over 3,000!

Mystery solved - I think this will be to fix the student address issue?

Cross post- I am a slow typist!

I don’t think it will only be Manchester, all the big uni LAs will be impacted. Leeds is about 4 x usual cases today too

Bloody hell, I know we all moaned about this a while ago but I had sort of assumed there'd been a fiddle to fix it since then! Well-spotted, people.

So had this been fixed earlier many more university citys may of been on tier 3 prior to the current national lockdown?

My LA is apparently zero where recently it’s been around 20. We are a rural area with no universities or other higher or further education so that makes sense.

@TheSun The way a large randomised trial works is that you put your volunteers into two groups at random. Half are given the vaccine and half get placebo. They don't know which they've got and nor do the nurses, etc who administer the jabs. You don't need to know the details of where the people live, whether they wear masks etc, because on average the people in the vaccine group will behave the same as those in the control group (the ones who were given the placebo) You don't need to try and compare with rates of Covid in the general population either, as that's what the control group is for. It wouldn't work if your two groups were tiny, obviously, as they might just happen to differ, but with large trials there will be no significant differences overall between the behaviour of the people who were given the vaccine and those with the placebo. (A bit like if you tossed a coin 30,000 times you could be pretty sure you'd have very close to 50% heads and 50% tails)

1392 10-19 cases moved out of East of England data and into... (maybe just Manchester??)

Are Manchester students particularly rubbish at registering with a GP maybe?!

Has anyone checked the reverse - has Elmbridge, Richmond (the southern one), Merton all had large falls?

What on earth is happening in Scarborough and Hull? Anyone got any explanation for their explosion in cases post lockdown?

New cases in my area have dropped very suddenly, how did that happen?

Oh hang on just saw those comments. Um, I think we had yours.

Dint know about individual London boroughs but London generally as a region really low today so I guess ended up a net exporter of students who get covid (interesting in itself as of course it also has a fair few unis itself although not sure how they have faired for covid outbreaks)

Scarborough and Hull had very low cases pre lockdown, and I think were Tier 1. So explosion now likely due to tier 1 restrictions not being sufficient as cases started to rise - a bit like much of the South west.

Nottingham City nearly 2000 today - the university effect 🙈

but with large trials there will be no significant differences overall between the behaviour of the people who were given the vaccine and those with the placebo

Indeed, this is all fine, my concern is that there was some confounding factor in the selection of the vaccine group.

(This is not what I think is true, but this is a hypothesis that could work and be bad)

You're worried that there's not enough virus in the community that your subjects will be exposed, so you select NHS staff for the trial, vaccinate them all, see success 5 with vaccine, 95 without!

Reality is that the vaccine only works as a booster due to previous infection and you actually selected a group that had already been infected and the 5/95 was the re-infection rate in that community.

I don't think that's true (neither that selection process, nor that level of previous infection in the study group seems likely) but it feels difficult to believe that the group was representative in some way when the infection rate is so different - obviously this trial wasn't even in the UK, but the US would've suffered similar risks - testing either in an area where there was considerable existing infection or in an area where there should have been significant infection.

I'd just like some commentary on what seems odd.

We got over 2800 here...

It's been between 100-200 cases recently.

What on earth is happening in Scarborough and Hull?
Hull had very low numbers right through firstcwave and all summer, then gradually rising and were about to enter tier 2 as lockdown started.
Just watched Hull city Council leader on local news absolutely pleading with Central government to help as the virus is "ravaging the city at an astonishing and terrifying rate"

@Qasd - lots of London unis tend to pull in students from London, SE and overseas. Since September staff + student cases: UCL 514, Kings College 334 cases, Kingston 40, Royal Holloway 15. So no big numbers in big nor small unis (obviously I only bothered to look at a few, there are lots more!).

@Firefliess
ok, I get that. My interest would be to have transparency of how many of the 100 had kids in school or went to work using public transport. Or what type of work they do. Do the 2 groups (placebo/real) demographics and socioeconomic real mirror each other.
In this specific case (as in covid research generally) as much as I generally believe scientists and researchers it would be paramount to come forth with these datasets.
You might be absolutely right and this might be the best designed trial possible. But even the design description is sketchy.
And this based on the actual research paper, not some article.
So imagine few months from now when the actual issue of get vaccine or not comes up not just as a hypothetical.... these type of secretive (or don't assume anything sinister: simply mis-communicated) companies themselves will be fueling conspiracy theories. At a time when mass uptake would be needed.

I'm just thinking along the lines of .. if we would partake: kid doesn't go to school. I go shopping in n95 mask (not textile) but i do limit where I go and how frequently. I don't use public transport. If I got the vaccine and didn't get infected it would have minimal to do with the vaccine. I'd be happier to know that they didn't select people like me. Based on what they say the selection criteria is we could be part of this because we live in an area of high numbers which is enough.

@sashagabadon

Interesting to hear how many questions are asked. But why not use technology to verify where possible? Participants could agree to hand over google/apple data that has been collected of them anyway for instance....

I honestly hope that no company is cutting corners just for profit's sake. But knowing how companies work around the world I doubt it.

If the data on the vaccine trial isn't convincing with respect to efficacy then it won't get licensed, regardless of any motives the company that developed it might have. I trust the European regulator, if they are satisfied that the trial and the data it produced are sufficiently robust then I'm happy to trust that.