looking at the john hopkins breakdown in the US is interesting, if Trump is relying on the virus only hitting blue states/area... he maybe in for a shock.... for example caddo county louisiana
As for the role schools play This Guardian article Warrants further investigation.
So one childs evidence, excuse me for not being convinced schools should be rushed back.
Sweden has a population of around 10 million, stretched across a landmass bigger than the whole UK
I thought the population was concentrated in Stockholm and Gothenberg? About 20% in those two.
I refuse to link it, but there was a DM article yesterday on CV19 mutations, stating the California strain was the mild version, and the NY the European one.
Del - the big word in your post is REPORTED. How many cases get reported/tested in the UK. Any CFR is meaningless until we know the ACTUAL numbers infected.
There's a mild variant (strain) of CV19?
For the actual CFR of Covid to currently be 0.1% (1 death per a thousand cases), so far a quarter of the UK population would already have been infected.
Could be right. I don't know the UK figures but 4 weeks ago we were given an estimate of 65,000 folk in Scotland had been infected. That's 1.2% of the population. Infection rate appears to have slowed from its peak since then but at a peak of "doubling every 3 days" it wouldn't have taken long to get to 25%. And the %age infected is likely to be higher in England.
Could be right. I don’t know the UK figures but 4 weeks ago we were given an estimate of 65,000 folk in Scotland had been infected. That’s 1.2% of the population. Infection rate appears to have slowed from its peak since then but at a peak of “doubling every 3 days” it wouldn’t have taken long to get to 25%. And the %age infected is likely to be higher in England.
without massive levels of community testing its all speculation
it wouldn’t have taken long to get to 25%. And the %age infected is likely to be higher in England.
the dutch estimate is very low
Del – the big word in your post is REPORTED. How many cases get reported/tested in the UK. Any CFR is meaningless until we know the ACTUAL numbers infected
While I don't refute what you say, 3.4% is a lot bigger number than 0.1%. If you want to speculate and say that if even 60% of cases are never reported/tested you still get a much larger number than 0.1%
In any case, you can still work back from where they do have decent testing, and they have a good idea of how the numbers of infected relate to hospital admissions and deaths.
And as thepurist observes - we're making the comparison with regular flu again?
without massive levels of community testing its all speculation
Indeed.
The predicted numbers were that less than 10% of cases required hospitalisation. We are only testing in hospital. Add that to 50% asymptomatic cases, its pretty easy to move between 0.1 & 3.4%.
But, as above, we wont know until its all over and proper antibody resting is widespread.
The data from the Covid symptom tracker app shows an estimated peak of symptomatic cases on 1st April, so a peak in deaths a week later is consistent with that.
That symptom tracker shows the onset of people downloading and using the app. Yes I used it on my onset. Once. It's highly dependent on downloads of the app. How did they adjust for that in their projection?
I watched the video and quite like Unherd. Most of what he says is reasonable, until the speculation starts - where is the EVIDENCE that up to 50% HAVE been infected. Based on Dutch screening of blood samples, corroborated by a screen of healthcare workers, the prevalence is perhaps 3-6%. The Stanford seroprevalence suggests 1%. These numbers are an order of magnitude lower than those to even contemplate herd immunity.
Sweden has exactly the same growth rate for cases and deaths as UK and France. It is possible that their strategy is as efficacious as the UK's, but they had the benefit of considerable hindsight.
The CFR is likely 0.1-0.3% and lower in healthy subjects based on reasonable orders of magnitude calculations. About 19/20 cases are not being recorded. Social distancing will likely reduce overall deaths by up to 10x.
Del – the big word in your post is REPORTED. How many cases get reported/tested in the UK. Any CFR is meaningless until we know the ACTUAL numbers infected.
South Korea is my personal go-to for "good" numbers (massive pinch of salt for any numbers, obvs) - mainly because (a) not China and (b) doing track and trace and seemingly keeping on top of it.
Their numbers suggest around 2.5% - 3% CFR, IIRC. As of some weeks back. This might be a bit out of date, but again, order of magnitude different to that 0.1% from the speculation video.
This was them on Mar 24:
8961 confirmed cases
111 deaths
3166 recovered
90% of dead are over 60
20% asymptomatic
CFR distribution:
80+ 11.60%
70-80 6.30%
60-70 1.50%
50-60 0.40%
40-50 0.10%
30-40 0.10%
overall CFR 3% [ignoring active cases]
as of 20th Apr, 10683 confirmed cases, 237 dead, 8213 recovered. Overall CFR is 2.8% (ignoring active cases)
I would guess the Korean numbers are about as good as they get, they have very slow growth now, single digit type daily new cases and deaths.
That symptom tracker shows the onset of people downloading and using the app. Yes I used it on my onset. Once. It’s highly dependent on downloads of the app. How did they adjust for that in their projection?
Usage is quite a different pattern, but they discuss the modelling in the webinars linked from the site. There was a spike on 31 March when they introduced a reminder notificaiton, but have (or had at the time of the last webinar) seen increasing participation since then.
[url= https://i.postimg.cc/7bhynC4g/Annotation-2020-04-21-114149.jp g" target="_blank">https://i.postimg.cc/7bhynC4g/Annotation-2020-04-21-114149.jp g"/> [/img][/url]
About 19/20 cases are not being recorded
I think you're right, but I'm not sure that pushes CFR down to 0.1-0.3%? Or does that low CFR discount the pre-existing medical condition cases?
It’s a magic number made up for attention. Ignore it. Move on.
[url= https://live.staticflickr.com/65535/49800667613_b19eb46fdb_h.jp g" target="_blank">https://live.staticflickr.com/65535/49800667613_b19eb46fdb_h.jp g"/> [/img][/url][url= https://flic.kr/p/2iSHrse ]Screenshot 2020-04-21 at 11.51.47[/url] by [url= https://www.flickr.com/photos/76248110@N06/ ]danthomassw13[/url], on Flickr
South Korea testing per head of population isn't much different to ours now.
I'm afraid that in seroprevalence I trust. A lot of infections in self-selected patients reporting via an app is not the best recording of prevalence. It is A recording, but anonymized seroprevalence is the only denominator of value. I have reanalysed the Californian data using a Bayesian methodology, and their prevalence is 1% (but does not rule out zero). It's definitely less than 3%.
3% seroprevalence in 60Mn population is 2Mn cases against 100K infections and 10K deaths is 0.5% CFR
Adjust appropriately, but order of magnitude is what matters here. In healthy it's 10x lower. UK CFR for reported COVID19 +ve cases cases is about 5%.
as of 20th Apr, 10683 confirmed cases, 237 dead, 8213 recovered. Overall CFR is 2.8% (ignoring active cases)
So, not 2.8% then.
I've just seen this on socail media, its behind a paywall so don't have full details:
https://www.telegraph.co.uk/news/2020/04/20/exclusivemillions-pieces-ppe-shipped-britain-europe-despite/?fbclid=IwAR39fT5kCxcj4ZMWxJfU-nvGhWYWTNu5bxSqyn3kygIgDLPnh3I4nCr31N8
I assume the reason is due to requirement to fullfill existing contracts, but this quote at the start, if correct, is concerning/unsurprising depending on cynicism level:
"On Monday night, UK firms said they had “no choice” but to keep selling the lifesaving gear abroad because their offers of help had been repeatedly ignored by the Government."
So, not 2.8% then.
if you strictly take dead/recovered.
depends on the story you want to tell.
Now ^^^ That Telegraph article is the details we should be using at Daily Press conferences to hold ministers to account. "Please find out and come back to me tomorrow, because I expect detail by then."
So, not 2.8% then.
erm, deaths divided by [recovered + deaths] gives 2.8%.
or, as I posted:
(ignoring active cases)
because their outcome is currently unknown.
Which means we don't know the CFR, and we won’t know for a long time yet. If ever.
depends on the story you want to tell.
Indeed
Apologies. Conflating numbers of those diagnosed dying vs. those dying of covid-19 associated problems among the general population. Dumb ass.
Interesting take on Germany's handling.
Especially (IMO) the case for honesty in the public communications, admitting the unknown factors and not bullshitting through it, thereby garnering more public trust and compliance for the required social distancing rules.
Which means we don’t know the CFR, and we won’t know for a long time yet
Give it another month and I'd guess most active Korean cases will be closed one way or other.
We *do* know the CFR of the currently closed cases. Obviously we don't know full infection figure for the Korean general public, but if their 'new cases' figure is low - and it is very low - you'd have to assume their reported cases number is close to the actual case number. If it was massively different then the deaths and reported cases would start piling up, and that doesn't appear to be happening.
It's a country mile different to what's happened here, spain, italy, etc. where the cabinet sat on the problem for a few weeks whilst having a Got Brexit Done pishup and let the thing slip completely out of control on the tracking and testing front.
Look at my graphic ^^^^^^^^ - S Korea testing a smaller proportion of population than the uk now.
Especially (IMO) the case for honesty in the public communications, admitting the unknown factors and not bullshitting through it, thereby garnering more public trust and compliance for the required social distancing rules.
Germany has 4x fewer cases and deaths than the UK. If you assume an initial doubling time of only 3-7 days, then three weeks is about 3-6 doubling times, and 4^(1/6) = 1.25. So a relatively small difference in attitude to illness and subsequent behavioral change, gets exponentially magnified during the epidemic phase. Germany's different outcome was not due to testing (they weren't at the start). I believe it was more likely to be behavioral. And public trust played a big part in that.
Look at my graphic ^^^^^^^^ – S Korea testing a smaller proportion of population than the uk now.
We know. They don't need to. That's what happens when you get ahead of the thing. The graphic is a bit meaningless for South Korea.
Two options with track and trace, I would propose;
a) get on with it right good proper like, at the start, end up not needing that many tests and not having that many cases or bodies
b) f off on holiday, only to find when you get back its already everywhere, and you now need to test, er, well, everyone, because the crematoria are already 24/7
Which one is South Korea?
Isn't German population density less than the uk too?
Mr M - I don't disagree, but they still aren't testing the whole population to discover a true CFR.
Isn’t German population density less than the uk too?
Yes, but in the early stages, this is probably not important since transmission is localized to a relatively dense area. The transmission unit is probably a city (London) or at most a region (Lombardy).
Look at my graphic ^^^^^^^^ – S Korea testing a smaller proportion of population than the uk now.
Best to look at testing vs. deaths to get an idea of how thoroughly testing has been implemented.
Quick calculation, Korea has tested 2400 people per CV death. UK has tested 30 people per CV death.
Therefore Korea has tested 80x more extensively than the UK. (In relation to the size of their epidemic)
We are only really testing people who have severe illness and end up in hospital.
Korea were testing everyone with symptoms however mild and also in some cases all their close contacts.
Korea were testing everyone with symptoms however mild and also in some cases all their close contacts.
They probably weren't. But the case acquisition rate was likely higher than the approximate 5% in the UK. The previous SARS outbreak was most likely the cause for this.
Quick calculation, Korea has tested 2400 people per CV death. UK has tested 30 people per CV death.
That says what I was trying to say, but far more succinctly (and better).
They probably weren’t. But the case acquisition rate was likely higher than the approximate 5% in the UK. The previous SARS outbreak was most likely the cause for this.
They definitely were.... And they were also using smartphone data to make sure they got all the contacts
https://www.theguardian.com/world/2020/mar/18/covid-19-south-koreans-keep-calm-and-carry-on-testing
extract from article
"By retracing how someone came to be infected, South Korean authorities have been able to give the public details about new infections and identify infection clusters early on.
After an employee at the call centre tested positive, authorities quickly set up a tent to test everyone who worked or lived in the Koreana Building, as well those had visited the premises."
Sorry can you explain what case acquisition rate is?
Al Jazeera news report on UK death figures, and the unreported care homes numbers. Note, contains some speculation.
Summary, ONS reports 3833 extra covid deaths, as of April 10. Official government number was at 9288 on that day. That's 40% extra. Apply 40% to today's government figure, 16509, and UK could actually be at 23113 deaths.
Sorry can you explain what case acquisition rate is?
It's the fraction of all true infected subjects who are detected and counted as a case (and in this case tested COVID19 +ve). In the UK I believe it is about 5% - hence 19/20 people who have had the infection may be missed. Trying to identify cases when there are mild infections which present as common (or no) symptoms, means that the case acquisition rate is largely unknown, even in S Korea. You have to meet a case definition that may NOT include all people who have the infection, especially early on when little is known.
A seroprevalence survey tells you what you missed.
As an example - suppose in S Korea you tested EVERYONE with a dry cough or temperature, you'd miss me as I only had muscle pain and eventual shortness of breath.
Summary, ONS reports 3833 extra covid deaths, as of April 10. Official government number was at 9288 on that day. That’s 40% extra. Apply 40% to today’s government figure, 16509, and UK could actually be at 23113 deaths.
Pretty sure we wont be the only country where this is the case (iirc Spain werent including care homes in their tally) but it is still a grim number
meanwhile at select commmittee today turns out it was a political decision not to join EU procurement scheme after all...
https://twitter.com/nick_gutteridge/status/1252613227655106561
we wont be the only country where this is the case
WHO standard reporting of deaths is for hospitals only. So that is every country. The additional deaths may be reported differently from country-to-country. But that is why they are standardised.
>Sorry can you explain what case acquisition rate is?
It’s the fraction of all true infected subjects who are detected and counted as a case (and in this case tested COVID19 +ve). In the UK I believe it is about 5%
Ah, ok. That is the million dollar question that nobody really knows the answer to.
Worth mentioning that the WHO are saying some people don't produce antibodies, so serology data might not be accurate.
WHO standard reporting of deaths is for hospitals only. So that is every country.
Ok, that's useful to know.
Thanks TiRed.
WHO are saying some people don’t produce antibodies
Absence of evidence is not evidence of absence. We do have several million years of evolution that says the innate and adaptive immune systems in humans produce antibodies that are protective for reinfection (I'm hopeful), but it might be possible that those with the most severe infections might in fact have failed to mount a neutralizing response in lung tissue.
Therapeutic antibodies (screened from humans who have recovered) have proven effective in influenza and ebola infections. So there is nothing wrong with the humans as producers of antibodies per se. Whether they make enough is a different matter.
https://twitter.com/rhonddabryant/status/1252612029195726851?s=21
No surprise there, but the dissembling and blame shifting has long since worked… and the part time PM and his band of yes men and women continue to rise in popularity for their handling of the response to this health emergency. Tell the lie now, bury the truth later. Expect lots of this, and not just as regards PPE provision failings.
