Daily numbers, graphs, analysis thread 6

[Barracker]

Welcome to thread 6 of the daily updates.

Resource links:
Worldometer UK page
Financial Times Daily updates and graphs
HSJ Coronavirus updates
Johns Hopkins Coronavirus Resource Centre
NHS England stats, including breakdown by Hospital Trust
Covidly.com to filter graphs using selected data filters
ONS statistics for CV related deaths outside hospitals, released weekly each Tuesday

Thank you to all contributors for their factual, data driven, and civil discussions.

https://www.newscientist.com/article/mg24632783-400-could-the-coronavirus-trigger-post-viral-fatigue-syndromes/

We may expect to see an outbreak of post-viral fatigue syndromes in some people who have had covid-19,
according to some researchers.

Viral infections have previously been linked to problems with long-term fatigue symptoms.
For example, chronic fatigue syndrome (CFS), which is also called myalgic encephalomyelitis (ME), sometimes occurs after viral infections.

What % of parents had already taken their kids out of London schools befkre closure though?
By 2 days before SW ours was down 25%.
Then you have all the families where a kid was ill with temp or cough. And then the 14 days for whole families.
Thousands were calling for school closures.

It's does seem to be the largest cities.
Then you have airports amd whether they flew to hotspots.
Public transport links.
Type of business in the city.
I did wonder about altitude.
I guess it's somewhat luck if an area had good tracing done after italy return. If thry missed hundreds it will have grown.
There were reports that loads of people didnt/wouldnt stay in for 14 days others were having friends round. Luck re whether the person got symptoms so tested if asyptomatic they were back at school.
We have to have data on whether the returning kids had caught it amd how many gave it to classmates/teachers.
Then they werent allowed to shut even to clean a school.

Interesting article with pleasing graphs showing how several countries affected their R0 thingy by the measures that they took, including Sweden, Iceland, South Korea and all the usual suspects. And some other nice graphs where you can add any countries you want.

www.nzherald.co.nz/nz/news/article.cfm?c_id=1&objectid=12326697

Can I ask a really stupid question? Sorry didnt sleep well last night so brain not fully functioning....
The graph produced by the NHS showing actual deaths by day above, which indicates the death rate peaked on the 8th/9th April.
The govt have said one of the indicators they are looking at to ease restrictions is a sustained fall in the daily death toll.
Every day the news reports how many deaths were reported that day making it sound like they happened that day, but they didnt, that's just the day they were reported.
The govt stand at that press briefing every day and read out that same figure without any explanation that that is deaths reported that day, not occurring that day.
Due to the way that is recorded it will take forever to have a sustained fall.
I doubt the govt will come out with the actual data from NHS at some point all apologetic that they didnt show this before....the public will expect to see the reported figure falling, most Oblivious to the fact that they aren't deaths occurring. So my fear is that either the govt will either use the data they have, the actual data, to bring us out of lockdown and people will be terrified or they will use what they have been reporting to save face and we'll be stuck like this for ages.
Now, dont get me wrong, I do believe that the lockdown was and is needed to manage capacity for the nhs and slow the spread of infection, and it does need to happen for as long as it needs to happen for. I just dont think the whole thing should be steered by duff data.

Sorry that was epic.
Also forgot the question. What are your opinions?

Biffle,

The graphs are pretty but seem entirely random. Firstly, the R0s are entirely subjective and don’t match any other estimates I have seen (most think the U.K. is 0.6-0.7 post lockdown). Secondly, they have declining cases and fatalities where they are estimating R0 is still 1 or greater, which contradicts the definition of R0.

Sheffield has a specialist ID department that started testing local NHS staff before lockdown. This was posited to be why it had a very high infection rate - certainly at the start. Not sure if this is still the case. BAME population is lower than other cities and main employers in the city are NHS and Universities- there aren’t huge numbers travelling into the city centre to work each day as you might find in say Leeds or Manchester. In fact by 6.00pm on weekdays the centre is quiet. So again be interested to know if the higher rates are due to the testing regimen. That said reports on the ground in the region are that severe cases needing hospitalisation are still coming to peak.

Anyone clarify what natural immunity is? Someone used it in the context of being exposed to viral load but not being symptomatic. But surely if you are exposed to a viral load and have any type of immune response you aren’t immune? Is it more there are some people with a resistance.

Finally what do people think about the Santa Clara study showing many 50-80 times more people had been infected then they would have estimated off reported rates of infection?

Personally I’m in favour of a mass test in a UK medium city (Bradford has been mentioned and looks like it would be ideal). That ease lockdown gradually from mid May with requirements to wear masks and phases approach while adopting WHO gold standard of test and isolate symptomatic.

I think the population are starting to get restless and also medical profession are now starting to question stopping of other medical treatments and impact on death rates for non-Covid. My redeployed partner (working in a Covid ICU now) hasn’t done any 2 week wait (red flag) cancer screens for 4 weeks now. NHS has met capacity. Final years medics can hopefully start in just over a week and shore up a return to normal service if they are still needing staff in Covid areas. Only issue is the intensivists who would also do anaesthetics for ops will need to be able to return to this work.

@ShootsFruitAndLeaves
I have to ask (sorry) what's your day job?
I'm asking 1 because of your very funky graphs 😁
And 2 because you have such clear reasoning and analysis which I so often only see from others with a science and statistical background,therefore nosily curious

Livin,

Isn’t the delay 2-3 weeks rather than ‘forever’?

I think the government will consider the following before unlocking us:

1/ How is it working in our close neighbours? Ad we are behind most of Europe, we can at least have a look at what happens in Germany, France, Italy and Spain as they unlock and how they do it. We want to be (broadly) consistent in order to enable trade and travel.

2/ We need to unlock for at least 2-3 months (minimum). It would be a fiasco to unlock and then lock down again a few weeks later. So they need to do the maths, estimating the R0 post unlocking and seeing how many weeks the case load can grow until we need to lock down again (absent a vaccine). Unfortunately the cases/deaths drop a lot slower than they rise (unless we can somehow partially unlock and keep R0 below 1.25 or so).

3/ They then need to weigh 2 vs the economic cost to the country and the loss of education to a generation. And, in reality, the political risks either side of unlocking ‘too early’ (lots of avoidable death headlines) or ‘too late’ ((headlines of deliberately cratering the economy, especially for the poorest).

It is not an enviable decision!

If cases continue to go down at the current rate, my personal guess is that, at the end of this period, there will be a very partial unlocking (dentists, shoe shops for children, primary schools in deprived areas?) and then, maybe 3 weeks later, a more significant unlocking.

Thanks larry
That makes perfect sense. I think it's just lack of sleep and seeing the effect that this whole situation is having on those I work with must have stopped my brain from thinking that rationally.

The graphs are pretty but seem entirely random. Firstly, the R0s are entirely subjective and don’t match any other estimates I have seen (most think the U.K. is 0.6-0.7 post lockdown). Secondly, they have declining cases and fatalities where they are estimating R0 is still 1 or greater, which contradicts the definition of R0.

I was intrigued by Sweden... There’s a clear disconnect between the graph showing a R0 of 0.9 and its chief medical officer (or similar) stating recently that they were potentially looking at achieving herd immunity by mid-May!

Finally what do people think about the Santa Clara study showing many 50-80 times more people had been infected then they would have estimated off reported rates of infection?

I read a good article of this the other day that indicated why it could have significantly overstated infection rates. Firstly, the subjects weren’t selected at random, but were respondents to a Facebook ad, with the suspicion that this was more likely to attract people had been symptomatic. Secondly, the test used had a relatively high false positive rate that could have inflated numbers significantly.

didnt they orignally want more cases in the summer for some reason?
I imagine the scientists are communicating world wide

Derby,

A link to the study would be good!

The problem with these theories of much higher R0s and far lower CFRs is that herd immunity would be reached much more quickly and we would see cases plateauing and then rapidly falling of their own accord.

It is probable that we are overestimating the CFR and underestimating the number of mild or asymptomatic cases, but not by that magnitude.

First of all, I want to thank everyone who contributed to these 6 threads so far and for sharing some very information and analysis. I’ve been following regularly and am grateful you are all taking the time to do so.

I read this article last night but it appears earlier versions were released so I apologise if it is old news to the regular contributors here but I couldn’t see a reference to this particular information (I did miss a number of posts at times and simply don’t have the time to go over everything again)
It is a useful visual reference even if in some cases a city is listed as opposed to a country depending on how number of deaths are recorded in particular countries.

www.nytimes.com/interactive/2020/04/21/world/coronavirus-missing-deaths.html?fbclid=IwAR2dpfCYYf8P5uXMI-jIIadzwp1sZOLmvsSbyv14cupTnFLA_p1f9KFbZAQ

The Covid app suggests a prevelance of 3-4% possibly as high as 10% in London. I think you can add maybe 30% for assymptomatic cases. So top end 5% of uk population, 12.5% in London/ West Middlands. I think anything above 20% is wishful thinking.

I’m also exercised by what seems to be a loose definition of immunity.
Is there a way of verbally differentiating between “exposed to virus but not infected because the virus is unable to infect that individual” and “exposed to virus, infected but individual does not get unwell/develop symptoms because of effective immune or alternative bodily system response”?
Because both of these seem to fall into the category “immune” but are actually quite different, though appear (not on a cellular level) the same.
I may be misunderstanding/splitting hairs here, I’m aware.

The Pasteur Institute’s central estimate is that 12.5% of Paris (Ile de France) was infected with a range up to 20% infected.

The estimated death rate is 0.5% but this is highly skewed by age. The estimated death rate is 1 in 100,000 for those under 20 (0.001%), rising to 0.05% for those 40-49, 2.2% for 70-79 year olds and rising sharply to over 8% for the 80 year olds. This is purely death rate across all the population whether they have underlying conditions or not.

If a third of the population have underlying conditions, but 0.5% die of the disease, the vast vast majority of those with underlying conditions will live, particularly if they are under 70 or 80.

hal-pasteur.archives-ouvertes.fr/pasteur-02548181/document

Analysis vs politics:

twitter.com/SeshNadathur/status/1252699110206996481

It's a good example of tricks being played with numbers.

didnt they orignally want more cases in the summer for some reason?

I would imagine this plus flu season will be horrendous. Personally, not that I want to get it, As an asthmatic for whom winter is rubbish and viruses are even more rubbish I keep thinking I'd rather have it during the summer than in the winter.

The best chance of relaxing, so that the economy can resume at say 90% for 3 months is to schedule this for the summer months
which means moving the school vacation & terms

so, e.g. resume school around 1 June and plan a 4-6 weeks winter / Christmas holiday

Coronavirus pandemic death toll in UK twice as high as official figure | Free to read

FT estimate has been updated to reflect latest mortality trends

https://amp.ft.com/content/67e6a4ee-3d05-43bc-ba03-e239799fa6ab?

The coronavirus pandemic has already caused as many as 41,000 deaths in the UK, according to a Financial Times analysis of the latest data from the Office for National Statistics.
.......
The FT extrapolation, based on figures from the ONS that were also published on Tuesday,
includes deaths that occurred outside hospitals updated to reflect recent mortality trends
.........
The analysis also supports emerging evidence that the peak of deaths in the UK occurred on April 8
with the mortality rate gradually trending lower since,
despite the 823 hospital deaths announced on Tuesday, which were sharply up on the 449 in the previous 24 hours.

FT:

@borntobequiet

I would describe the first as innate immunity and everything else as acquired immunity (but I am not an expert in this field so it is just lay person shorthand)

Those terms may have very specific scientific meanings