Daily numbers, graphs, analysis thread 14

[BigChocFrenzy]

Welcome to thread 14 of the daily updates

Resource links:

Uk dashboard deaths, cases, hospitals, tests - 4 nations, LAs, English regions
Slides & data UK govt pressers
[[https://www.gov.uk/guidance/coronavi
rus-covid-19-information-for-the-public UK stats]] list of reports added daily by PHE & DHSC
PHE Surveillance report infections & deaths released every Thursday with sep. infographic
ONS England infection surveillance report ONS UK statistics for CV related deaths, released weekly each Tuesday
Daily ECDC report UK & EEA
Worldometer UK page
Plot FT graphs compare countries deaths, cases, raw / million pop
Covidly.com world summary & graphs
Plot COVID Graphs Our World in Data additional data

We welcome factual, data driven, and civil discussions from all contributors 📈 📉 📊 👍

The rushed Russian vaccine is one I'd personally avoid taking

• as it's a clear example of national pride and competition overcoming caution I hope it doesn't go wrong and reduce takeup of properly tested vaccines like our Oxford one

The Philippines is planning to start clinical trials for the Russian vaccine this October
If trials OK, then rollout of vaccine to their 107 million population would start in early 2021
President Duterte is reportedly to be inoculated in May 2021.

Did the lady who had a respiratory infection in February take Covid to the hospital having contracted it in their care home from an asymptomatic member of staff?

Of course, we cannot completely rule out the hypothesis that the residents might have been infected by the care home staff, however, on balance it appears less likely.

Firstly, the research seems to indicate that contrary to the intuitive explanation that temporary staff spread out infection across care homes, genetic analysis did not find any overlaps across different facilities. In other words, each of the reviewed care homes had to deal with multiple COVID introductions at approximately the same time.

Secondly, the percentage of care homes with COVID outbreaks across England appeared surprisingly consistent (43% of all facilities with London being the only exception at 100%). The number did not vary with the degree of community transmission. If the care home personnel had been the primary source of transmission, one would expect to see much lower numbers of outbreaks in the regions like South West that had a much lower COVID prevalence within general community. This didn't happen. Care homes in South West were hit equally hard.

Maybe we will find out how much control Putin has over certain leaders / govts......

Sorry I know not appropriate for this thread but I couldn't resist.

I think there is a need to accept that Germany is an anomaly because they traced, tested and counted every asymptomatic case. They had much lower infections across the country proportionally than other figures, even if their case volume looked high.

A couple of important factors in spread are population density and the propensity of the population to travel creating a huge volume of initial seed cases at the outset of the pandemic. If you travel the world you'll find baby British, French, Spanish, Indians, (and in some countries Israelis).

There's a part of me that wonders whether the Russian vaccine might actually be the Oxford vaccine...

In some ways I hope it is, much safer for the Russian people provided they've copied it right.

@Sunshinegirl82 given the hacking story a few weeks ago.... I'd say it's likely!!

Sorry to interupt, but I just wanted to say @littleowl1 your service is brilliant! My first emails came through at 8am. Thank you so much - to be able to just see clear and straight forward data which is releveant for me is what I've been looking for all along... without having to trawl numerous sites etc. I'm not a statistician by any means and this is so useful - thank you!

This pandemic has constantly amazed, me both unfortuantely with the magnitude of the ineptitude of far too many, but also with the brilliance of some 'ordinary' people. Why couldn't the government do something like this when one lone woman at her kitchen table can?!

@littleowl1 I'm sure those of us who signed up to CovidMessenger website yesterday were as delighted as me to read my email delivery/deliveries this morning, thank you!

Very clear and helpful.

One thing though, you don't spell out that it's absolute numbers, not per 100,000, so differing population sizes should be taken into account if comparing areas.

But thank you!

@Pebble21uk I absolutely agree with all you say to @littleowl1 !

I really liked it too. I especially appreciated if it was going up or down and what last week was. Very clear

Zoe Covid's estimated daily transmission rate has gone back down to 1,434.

The drop is probably not statistically significant given the small numbers but it is nice to see the estimate dropping back to the late June - early July level.

@littleowl1
got mine too. Am broadly very impressed but agree with comments of others. I honestly think a lot of people will be interested in your site and I'm very impressed. Are you manually inputting data?!

@humphries I don't think the fact that we badly undercounted deaths early on is a reason to be overcounting them now! Much better to standardise with the rest of the world. The large majority of deaths will occur within 28 days of a positive test, and it will be a consistent proportion - meaning that trends can be properly understood and death rates compared over time and between places. Unlike the previous system which was becoming increasingly inaccurate.

wintertravel you quoted one London study of 6 care homes, can you link to any study of care homes generally please.

@ancientgran Do you really need to ask if Boris lies? In case you do the answer is frequently but more often it's a distortion of the truth. You need a few empty beds to be able to take emergencies, you dont keep a lot of them in an under-resourced health service. There were, at the height of the pandemic, still ICU beds available in most places, but you achieve that by moving people out to a "covid" ward, a ward that was quite likely to have been occupied previously by those moved to nursing homes, as soon as you can. You also converted most of your theatres to ICUs, because you dont have staff available to send to Nightingale hospitals. The NHS continued to treat emergencies and in some places do elective work - "protecting the NHS" was about ensuring patients needing emergency treatment during the pandemic were able to get it and that there were still health care staff left to provide it after the pandemic.

Sending out the knowingly infected is a different issue. A lot of those who died of Covid-19 in hospital had a normal place of residence in a care home.

No i agree we should not overcount now, bring it in line now with other countries since it is important to know what is going on, but the discrepancy between ons, excess deaths is still very big so they could have left the actual count where it was and give us the new daily figures from now on. We now have a 25,000 discrepancy which looks like we are not counting accurately ( much like spain). I have more or less given up with uk data ( though daily stuff is a bit better for me now)but like this thread since it makes it a bit more clear.

@littleowl1

Thank you for the website. I'm viewing 3 totally different areas: Mine, my parents and where dd is at uni. It's going to be very helpful, and I must persuade my dad to sign up as he's been plotting statistics.

I'm not totally surprised the government has managed to manipulate the numbers downwards; I think the only surprise is they didn't manage to drop it further to put us under Italy. But I am annoyed by it, because we're clearly under reporting by excess deaths, and one of the worst countries for it, so bringing us into line with others sounds somewhat hollow.
I would also like to point out that they've altered the more recent figures so that almost all the deaths over 50 since the start of July are now miraculously under 50. Before this over half were over 50. This has amazingly brought our current 7 day average down to being comparable with the other European countries.
Now this I find suspicious, because you would expect a few blips-why would the days of over 50 be effected more? It looks like careful manipulation to me.

New Zealand is going to give one example of whether a student transmits infection

"Thirty-six cases are now active in the country, including those in managed quarantine facilities. One of the new cases announced on Thursday was at the Mount Albert grammar school in Auckland, and was confirmed to be a relative of someone in the initial outbreak. The infected student attended class on Monday, and local health authorities said they were contacting and isolating around 100 close contacts they had while at school."

Thanks for the figures @littleowl1, it looks good. Is there a possible future option to have the areas concatenated in one scrolling email, with intro that says "jump to X cases"?

I'm really pleased to see Covid-Zoe going down as well as ONS for estimates; if we can believe it then ~2/3 of cases are being picked up, which would have a huge impact with new cases dropping fast in a few weeks. I really hope it is true.

The difficulty with the stats reporting is that each category of reporting needs to be consistent. If the deaths over 28 days hadn't been removed from that category then it would have made comparisons difficult. The 28 day cut off data is flawed due to the undercounting in the early months but including deaths that occurred after 28 days doesn't solve that issue it just means the data is inconsistent.

We are lucky in some ways because we do have access to different categories of data. The ONS "Covid on death cert" data and excess deaths data are far more accurate in terms of overall total number of deaths. The 28 day cut off data will never be accurate for overall numbers, that ship has sailed.

At least now the daily date being released is more meaningful in terms of trends and real time comparisons to other countries.

I don't think the 28 day cut off has been the cause of any undercounting during the peak. The original count only included those with a positive test result and these would all have been within 28 days for deaths during March and most of April. You can see on the chart that they published that it's made no perceptible difference to the numbers during the peak. The big change has been to the recent weeks - due to people dying (mostly not of Covid) 1-5 months after a positive test.

@Firefliess

I think the undercounting is mainly due to the lack of testing in March and April. As that category only counts deaths following a positive test when we we're doing almost no testing the numbers would have been artificially low hence the difference between that figure and total excess deaths or deaths with Covid on the death certificate.

I still think it was right to make the change, it just means we need to look at other sources for a more accurate assessment of total deaths which, to be fair, Valance and Whitty have been saying from day 1.

alreadytaken

The genome study has only been done for London.

The distribution of care home outbreaks is provided in the PHE report:

assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/903085/COVID19_Care_Homes_22_July.pdf

The areas of low COVID prevalence (e.g. South West) started seeing outbreaks as early as high risk areas (e.g. London). If the transmission had been primarily driven by personnel, the picture would have been driven - regions with high community transmission would have been leading the way. On the cumulative basis, while South West care homes still appeared "safer", the differential to other areas does not fully reflect the comparative difference in COVID cases/community transmission.

I was wrong when I quoted the 100% number for London. I referenced it from memory but, in fact, I was thinking about "high risk boroughs", e.g. Camden, Hackney, Kensington and Chelsea, etc. The overall hit rate for Greater London care homes was in fact 50% (vs 48 for East of England, another area with lower community spread).

Also, just to clarify:

• It is entirely possible that care home personnel were secondary vectors in the COVID spread;
• Of course, UK is not on its own in failing to deal with care home transmissions. It has been a problem across most of the European countries (with Germany again doing much better than anyone else)

However... these caveats do not change the fact that discharging patients into care home without testing was most likely a tragic mistake.

I don't quite understand those numbers and have multiple questions:

1. If the hypothesis is that patients returning from hospital brought covid-19, how can there be so many new outbreaks in May, June, July etc when it was so long past the release to get hospitals ready for covid patients?
2. Care homes manage flu and norovirus epidemics every winter, so why weren't there isolation capabilities, at least in large care homes?
3. Outbreaks are 2 people, this isn't useful because it doesn't show how many care homes had significant transmission within the home. We need a @boys3 decile graph.
4. What percentage of cases found were and remained asymptomatic? Were any of these outbreaks purely asymptomatic and only found through the general testing?

@wintertravel1980 - you're making an assumption implicitly that the hospitals all spread covid to their patients and the only other source could be staff. I'm not sure that is true if comparing hospital outbreaks with care home outbreaks. I know of one case where one person in a care home and his friend caught covid from his brother (Italy connection) visiting in March. Many people travel long distances to see family in care homes. There will surely be other cases like that, care homes were not in a bubble.