UKH

Manchester Uni suggest that by 24th April 29% of the population had had Covid.

https://www.manchester.ac.uk/discover/news/over-25-of-the-uk-likely-to-have...

(includes link to paper)

PHE/ Cambridge suggest as few as 24 cases a day in London

https://www.mrc-bsu.cam.ac.uk/tackling-covid-19/nowcasting-and-forecasting-...

I'm only a simple matrix biologist and don't do epidemiology. Even with wide margins of error does this not represent some optimism that herd immunity might be starting to play a role in the epidemiology and that a second wave is less likely is we're careful?

I have not got enough understanding to know if this is nonsense or not, but as we don't know how long or effective any immunity is, I don't think we can declare it over just yet!

the manchester.ac.uk source has been suppressed, the story is carried by the mirror ... may not be reliable

Seems very high, saw this earlier in the week about Spain, suggesting that they had only got to around 5% of the population having had Covid and they were very badly hit as a country.

https://www.aa.com.tr/en/europe/study-5-of-spanish-population-has-had-coron... 

I was skeptical on the very tentative conclusion about community infection drawn by the Manchester paper - one of its authors says the same thing in the press release

“The figures are not perfect, with the numbers of severely ill patients as a proportion of the total cases being used as a market for estimates of wider infection.

“Only extensive antibody testing could give us a more accurate picture - but as that is only just becoming available, we believe this form of modelling is important in informing the best approach to unlocking the population.”

Still big questions about the strength of immunity from very mild infections as well.

UK hospital numbers are dropping quite a lot, as are deaths.  Let’s hope that continues as lockdown relaxes.

Some current data- estimates from UK test results:

“2.Number of people in England who had COVID-19

Based on tests conducted between 27 April and 10 May 2020, we estimate 148,000 people in England had COVID-19”

https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/con...

I think the Manchester modelling is likely way off given our preliminary antibody study gave 4% overall and 10% in London. Both France and Spain's antibody studies have shown 5%.

https://english.elpais.com/society/2020-05-14/antibody-study-shows-just-5-o...

I'm optimistic about things continuing to improve. I haven't had any new people with covid this week.

Interesting to see how these two studies have been reported first thing today:

Torygraph & Fail - front page headline splash

Grauniad & Auntie - not mentioned at all.

It's not up front in the Guardian but it's been there since last night:

https://www.theguardian.com/politics/2020/may/14/englands-covid-19-infectio...

I

29% of population in England+Wales is 17 million.

ONS yesterday said 148,000 infected England+Wales from 27 Apr to 10 May. 

When did the rest of the 17 million get infected? It would take 4 years at 148,000 per fortnight.

I understood that to be infections at any point in time, so perhaps 500000 over the fortnight. Still a large discrepancy though. 

https://www.bbc.co.uk/news/health-52662066m

beeb reported those figures yesterday too. Making it 0.27% of the population.

Maybe someone from the Manchester report slipped with the full stop when they pressed publish? (0.29%?)

Looking at the tracking tables on the FT website, almost every region (that has had a significant outbreak) has the same excess death profile irrespective of the strength of the lockdown imposed. Stockholm's excess death rate would appear to have peaked and is dropping despite schools / bars / restaurants being open... 

I was just about to copy in that quote before I read your post. They mention that antibody tests are just becoming available. It would seem far better to wait for these, which probably give a truer overall picture (they seem to imply that their type of  data are more readily available and might be used to give local information and perhaps local responses). I'm no expert.

Our these the Porton Down antibody studies which people have been talking about? In an older thread someone was saying that they were sampling the population every week, but I'm not clear if they have published the results.

I think those figures might suggest some discrepancy but I don't think we're comparing like with like. the other factor is those figures come from the very peak of lockdown when it had been up and running for 6 weeks my rough calculations suggest that if r was reduced to 0.7 by those measures then at the point of lockdown there were 880,500 infected at any one time. obviously this is probably total bollocks but you get an idea of how different things might be 

They were figures given by Patrick Vallance in the press conference on Monday. I assume that's where they came from.

This is just what I was thinking but hadn't got round to doing the maths. Surely the cumulative total must be reasonably high by now on these (the 148,000) figures? Maybe not 19 million but certainly many millions.

Yeah I'd make the numbers nearer 5 million but by that point I've made so many assumptions that the number is meaning less I may as well say that its between 148000 and 60 million.

At 24 cases per day it would take London just over a thousand years to reach 29% infection and another thousand or few thousand years to reach herd immunity.

That's not a good plan.

24 per day is not even enough keep up with live births in London.

Something is wrong with one or both or 29% of population and 24 cases per day in London.

Even to get 148k they surveyed 10000 people. So they will have made loads of assumptions within that already. 

Yes that was kinda of my point if i had one, the only semi reliable number is the number of deaths, but its hard to use this for anything other than spotting trends as we don't actually know what the mortality rate is all this really tells us is that approximately 3 weeks ago R was >1 or <1 it doesn't tell us why or what the actual number is because that depends on how the number of deaths compares to the number of infections. the more surveys done the better idea we'll have and I assume it's a "representative" sample being used but there's a huge amounts of unknowns and we know it didn't include hospitals or care homes. basically i'd use any of these numbers with extreme caution and while i welcome their publication I'd caution against reading too much into it unless you have a background in stats . 

I think the point is that London may be at herd immunity already, hence why there are so few new cases... 

Real case for optimism looking at the Cambridge uni data, the fact they are doing the modeling for public health England means it's should be fairly reliable.  Main take is 12 percent of English population look to have been infected with 20 percent for london which now has an R value of 0.4, so a degree of herd immunity is obviously helping slow down transmission in London, with community transmission down to a low level there so probably not such a crazy idea encouraging more people back to work there as long as they are avoiding public transport.  I am not qualified to state what effect easing the lockdown will have on R, but the halving times in the model suggest daily new infections should be below the 5000 a day level to allow some primary school kids to return by 1st of June , that's if things don't get dramatically worse of next two weeks, but can't see how people going to gardening centres and sun bathing or traveling for excerise is going to push R above 1 in meantime, but I am sure the government and its scientific advisors factored all this in with its decision to ease restrictions earlier in the week.

Obviously still a long way to go, and getting the country and economy opened up more by July still looks extremely challenging, but if some kids are able to go back to school from June that can only be a positive for the country.

Unsurprisingly countries with a similarly significant outbreak look similar.

Sweden doesn't look so good when compared to a dissimilar country like Greece.

https://ig.ft.com/coronavirus-chart/?areas=grc&areas=swe&areasRegio...

I'm sure it's as reliable as it can be but there has to be a large amount of uncertainty in this, uncertainty doesn't mean bollocks just that you can only be so accurate with the information available 

This would be great but it's pretty hard to go to work in london and avoid public transport for most people ( I gather I'm not daft enough to live in london). Also I think there's a tendency for the government to take an optimistic view of the data given to them and combined with a lack of background in these things this could lead to questionable decision making. conversely I think there's a tendency to take a conservative (small c) view of the same data in the devolved regions and that's before you factor in the differences in the initial infection rate there and political biases.  

Well, the ONS suggest 0.29% of the UK have it so your first link appears to be rather over egging it

Haha yeah I am not daft enough to live there either, but from the start of the pandemic London has had a lot of bad press, packed tubes, over crowded parks, and cyclist riding in groups up box hill, but despite all the negativity and being worse hit part of the country, things dont look so bad now in London, even if the data is far from accurate, all the evidence I come across still suggest London has lowest level of R in country.

I have an optimistic view of life, one thing I know the country can't continue as it now for the next couple of years, so any optimism is welcome, and I don't view the current easing of restrictions with gloom and doom. 

Living in Scotland I don't think our devolved government has handled the situation much better than UK government, but they do get a easier time of it in press etc.

What's that a number of though I think most people suspect that more than 174000 people in the uk have had it. using the ONS number of Covid deaths of 33841 that'd give a mortality rate of nearly 20% I don't think anyone has suggested this to be the case.

I think that number is either number of people currently infected or number of positive tests. neither are the same as percentage of the population to have had covid-19.

I really hope you're right and i agree that we can't continue under lockdown for the next few years but that doesn't mean things won't be grim, I think london probably looks better now precisely because of those things at the beginning R seems to be reduced by immunity probably not below 1 in a normal situation with no lockdown though, who knows though.

It feels like the government have had a very easy ride in the english press I dunno what it's been like north of border. the tails of people who have been working from home being told to go back to offices by their bosses gives me huge cause for concern though.    

The Manchester paper is coming in for some heavy criticism, and one of the authors is blocking anyone on Twitter who points out flaws. See:

https://twitter.com/CT_Bergstrom/status/1261058066495074304

The only ways to get herd immunity is for lots of people to be previously infected (say 4.5 million in 3 months averaging 50,000 per day) and since then the London infection rate has declined to 24 per day.

This requires me to believe infections have plummeted by a factor of two thousand (or more) since the peak without a huge drop in hospital admissions and deaths.

There are time lags between infections, hospital admissions (10 days?) & deaths (23 days?) but nothing that could hide a two thousand fold drop in infections.

I think alot of people are miss understanding the ONS

"Our latest estimates indicate that at any given time during the two weeks from 27 April to 10 May 2020, an average of 148,000 people in England had the coronavirus (COVID-19) (95% confidence interval: 94,000 to 222,000). This equates to 0.27% (95% confidence interval: 0.17% to 0.41%) of the population in England. This estimate is based on tests performed on 10,705 people in 5,276 households."

So i guess it means on any given day during that period you could reasonably assume that 148000 people where  infected.

Not exactly a large scale survey or anything, but I've now gone 3 shifts without seeing a patient who looked like Covid. Not had a run like that since February.

I'm not seeing much reason for optimism in Derbyshire yet. These graphs are from a daily spreadsheet I've been keeping:

https://bit.ly/2WD1THH

Some pieces of data:

The recent antibody tests the government did suggested that ~10% of a random sample in London tested positive, so about 1m people depending on what you count as London. Its not known how many people with minor/asymptomatic infections do not get antibodies.

The deaths in the week to 5th May (in London) were 84% down on a month before

Whichever way you look at it the cases/deaths in London are dropping off far quicker than in other parts of the country, despite being under the same lockdown rules, with I am guessing no less/worse compliance with those rules than elsewhere in the country. A plausible explanation would be that the virus is simply running out of suitable hosts and this is driving down the R factor... I am not in any way an expert so cannot state that is the case, but what other plausible explanations are there for this discrepancy between London and the rest of the country? 

There was someone on the news, an epidemiologist, recently that suggested our figure was around 4%.

Not sure if this is true or not.

I think it is possible but that doesn't mean it's the case up till the end of the initial tight lock down it's claimed that cases were halving every 3.5 days that means that it'd take 5 and a half weeks for cases to drop 2000 fold at that rate. the exponential effect working in reverse. 

It's complicated though as you probably don't need to have herd immunity to reduce R  significantly, we don't have herd immunity to seasonal flu otherwise it'd die out but we reckon that it has an R of 1.4 at the moment this rate of transmission combined with it's mortality rate mean that we can manage it. I am way out of my competancy on this so happy to be corrected. 

I wonder how many people who regularly commute on the tube have had it?  

Who is catching it now? 

this tells us how we move forward so we should be testing as much as possible and this information should be shared with the public. I feel like a school teacher but the goverment need to show their working. 

There's true as in best estimate on information available and then there's accurate. There's a huge amount of unknowns at the moment. I can't see how any figure can be accurate to any sensible tolerance.   

Up here in Scotland all the blame seems to be aimed at UK government with Scottish government seemingly getting of lightly despite various mistakes made in past and currently still being made now.  I guess half of England voted for the Tory government and half of Scotland voted for SNP, so supporters won't want to let obvious failings change there core beliefs, in Scotland we have a similar number of deaths as Sweden and half the population, kinda says it all.

Yeah despite the immunity levels in London business as normal would most probably result in R becoming greater than 1 but not quite at levels seen pre lockdown 

I think the data presented by government on monday is now nearly 6 weeks old, hence my thinking that the numbers suggested in on going Cambridge modeling is more likely to be current estimation,  I am guessing the Cambridge scientist did the maths in there modeling which got us from 4 percent out side London six week ago to the 12 percent for England as a whole now, and 10 percent in London back then to 20 percent now.

There's a lot of towing the line going on to at the start of lockdown there was lots of grim stories of young people dying now these have all but disappeared. 

A lot of what you observe i think is the inherent bias of media in england and scotland. the UK government has apparently been slated in other countries. the Scottish government must have felt to a certain extent tied to the rest of the UK at least at the start of it all so perhaps it's not totally unfair of them to shift some of the blame. 

With the benefit of hindsight we can see where we could have been if we locked down when spain did but that's only knowing that we'd lockdown anyhow. I'd argue that that much was obvious at the time but the goverment may disagree.  

I'm not sure what you mean by "were infected" (had the virus or got the virus?). Presumably the number who got the virus would be the number with the virus divided by the average length of the infection. So maybe 148000/14 or roughly 10000.

I t wasn't from the government, he was a professor from somewhere or other, (sorry I know that's vague), but I'm not sure how you can assume he was using 6 week old data, in favour of you Cambridge model, it seems a bit wishful thinking, bu who knows.

For what it's worth our town seems to think it's all over and has gone back to normal, there's a massive queue in the drive through for KFC and more cars on the road since I've seem before Christmas!

If we don't get a monster upsurge in new cases in a couple of weeks it'll be proof of something.

We have evidence it's pretty effective immunity, convalescent plasma is working.

How long? We think it's not mutating fast and think a vaccine will work.

Do you not remember Boris saying we could take it on the chin? That's not hindsight, seeing where we could have been, then him shaking hands with every Tom, Dick and Harriet.

There's only one protocol which has worked anywhere, and we knew about it in February, lockdown, to control numbers when the numbers are low enough, testing and, track and trace, then start opening things up. We're currently opening up before the numbers have come down enough, and we have NO tracking and tracing in operation.

Were testing 100,000 which in reality means somewhere between 70,000 and 80,000, but it looks like many of them don't really count.

I can't see how ALL of this wasn't obvious at the time.

How do you work that out? At most we have 5 months of data, and the WHO still are saying there's no proof of immunity, so how do you come to the conclusion it's pretty effective?

Where's the vaccine though? We've only just got a "possible" effective anti-body test!

I sometimes wonder if your need for things to get back to normal colours the facts.

That's not true. You can keep repeating things but that doesn't make it true.

Convalescent plasma has recovered patients, its not been fully confirmed by controlled trials because they've not had times but its now a treatment. My wife is an MD at a local academic hospital, they'v used it there.

How the f*ck did I say things were going back to normal?

How did a say a vaccine WILL work, they THINK it will, we've about 20 at least going through trials and a long history of effective vaccines.

Dr Jeremy Rossman, Honorary Senior Lecturer in Virology at the University of Kent, said:

“The WHO have recently warned that there is no evidence that people who have been infected with SARS-CoV-2 will be immune to a subsequent infection. This posting was then retracted to avoid confusion as it seems to imply that people don’t become immune to subsequent infections. The message is that we just don’t know enough at present to say if people that recover will be immune or not. We know that the virus causes a robust immune activation, we know that survivors do generate antibodies to the virus but we don’t yet know if generates an immune memory that protects against re-infection, we don’t know what percentage of people will be protected and we don’t know how long this protection would last.”

https://www.sciencemediacentre.org/expert-reaction-to-who-clarification-of-...

Sorry I am dyslexic so extremely difficult to express thoughts in writing rather than verbally, and for it to make grammatical sense.

But from reading the data, partipants in the study were given a PCR test, from these results they concluded on any given day of the study you could assume that on average 148000 people in england was currently  infected by the virus.  So today I imagine a bit less then 148000 people are currently infected, and this daily number should continue to decrease as long as r is below 1.

Passive immunization seems to work (based on plasma transfers to particularly sick patients). Some medical groups at my university have therefore started to isolate B cells that generate neutralizing ABs from the blood of recovered patients, clone out the antibody genes, and produce humanized monoclonal AB lines to mass produce these ABs.

AFAIU the problem seems more that CVs seem to suppress the AB immune response and possibly memory cell formation as part ot the natural infection process. This would not be the case when the exact same antigens were presented by a live vaccine vector or DNA/RNA immunization.

Unlike for most other infections, vaccination may therefore give the immune system a better chance to mount a longer lasting immune response than a regular infection.

CB

Proof that the current changes in restrictions haven't raised R above 1 nothing more nothing less, I'm not sure that KFC drive through was ever thought to be a big contributor to transmission. I think that either people should have gone back to work in big numbers or  we should have eased restriction on exercise not both. The difficulty is that you always could go to work just people didn't (possibly sensible) there's still no one going to pubs clubs etc and I hope not going to parties at friends houses, outside of london i expect public transport is still pretty much empty, that's got to have an impact on the transmission rate.

He was from the government as I watched him speak live on tv at the daily press conference and he was a professor being the chief scientific advisor for UK, he stated that these results was from a limited amount of antibody testing done so far, and due to the nature of the testing they were already then 5 weeks old and probably now 6 weeks old, as the Cambridge university team is working with the government I assume the data they now give  has taken account of the previous antibody test results allowing them to make current assumption of percentage of UK previously infected as seems to best data anyone has to work on just now.

The roads are busy up here in Scotland as well, no idea where everyone is traveling to and leaa good for us cyclist but shouldn't really change the fact the virus is largely supressed and R is under 1.

The scientific papers I have read suggest infection of other cornaviruses usually give at least 12 months immunity for those previously infected, so I think there is a degree of hope the it will be similar for covid 19.

Spoke too soon ☹️

5.5 weeks is much longer than the time lags so that would now be very obvious in hospital admissions and deaths halving every 3.5 days.

That halving every 3.5 days hasn't happened in admissions & deaths so I don't think infections are declining that fast.

My most optimistic calculation is R=0.8 over 4 days using official stats from https://www.england.nhs.uk/statistics/statistical-work-areas/covid-19-daily...
and my calculations of R added in red to the spreadsheet at https://gofile.io/d/gZTX0n  

this was in London not across the country, further confused by Covid running riot through the care homes not all fun and games but the two things can both be true, I'm not saying for certain that it is the case just that it's possible and within the bounds of uncertainty.  

Numbers such as hospital admissions (only 5 day lag behind infection) for London are not halving every 3.5 days.

Can you find any data indicating infections, admissions, deaths or any other indicator halving in 3.5 days?

I am assuming the data must be flawed in someway as Cambridge uni has temporarily taken the modelling data down from there website,  I am sure it still stands that London are having a low level of daily newInew but maybe not the 24 they stated on there website.

only basing that statement on this article https://www.independent.co.uk/news/uk/home-news/coronavirus-london-lowest-r... 

Can't access your file without downloading additional software, not going to do that I'm afraid happy to accept there's differing information out there.

Sorry for the repeat edit but surely there's more than a 5 day lag as it takes that long to show symptoms, not the most important bit but its just another complicating factor. 

I think there are reasons for optimism.  The numbers have fallen a lot during lockdown and we have seen several other countries manage to suppress it so there is definitely a path out of lockdown into something closer to normality.

But the estimates for the number of people who have had it already are all over the place.  The testing isn't working properly yet - they aren't even counting the subcontracted test results into the total diagnoses yet.   There are all kinds of conflicting indications suggesting errors or things we don't understand yet.

Also, things are getting better in the medium term.  The anti body testing looks like it will be available soon.  They are finally doing random sampling although not nearly enough.  They will presumably eventually get test and trace and a working app but it will take a few weeks.   Things like anti-virus surface coatings are being developed which could make for a safer environment.  The doctors are getting better treatment protocols.

This thing is definitely controllable but the UK government is determined to rush it and take a huge risks which could result in another peak of deaths and another long lockdown to get it back in control.  They could have taken another two or three weeks and used the time to get the test and trace systems staffed up and debugged and get a firmer grip on the numbers.  It's like the ventilator fiasco, what you get when people with no f*cking clue about project management are put in charge and they get obsessed with stupid and unrealistic deadlines.  What they don't get is that things take longer if you bungle them by rushing.

https://www.bbc.co.uk/news/health-52677194 this is deeply concerning if it is creeping towards 1.

Also seems that lots of claims being made about london are entirely bollocks so lets hope they weren't the basis for government policy? 

I'm sorry but to suggest Box Hill is in London requires Ryanair levels of geographical imagination!

It's south of leeds isn't it? all the same to me

I take last statement back as now appears to be back on the Cambridge website

Excel or open office is you have them.

You are right, my mistake. 5 days for symptoms, 10 days for hospital admission (averages).

https://www.mrc-bsu.cam.ac.uk/now-casting/ shows the following:

• halving time 7.2 days for London
• daily infections 24 for London
• R=0.40 for London and has been below 0.5 since lockdown

Does it really matter where it is?, whole point was a long zoom press photographer making out a bunch of cyclist where all cycling close together in a group when in reality a lot of London based cyclist who were in the photo came back to say they were spaced out over a considerable distance.   

It’s being reported that everything i was basing my points on is basically bollocks so lets not worry bout that. 
 

Possibly the same for me!

Perhaps grounds for a little extra optimism in the good weather. I wonder if the approaching summer indeed  has a useful effect in decreasing Covid-19 (in the northern hemisphere). In UK we've had a pretty warm dry period; worse in the north but then Scotland and northern England may be slightly worse regarding both weather and the R level. Also Vitamin D levels up with increasing sunlight. New Zealand and Australia  entered the pandemic in summer and both have handled it far better than us, possibly with a little meteorological help?

I cling to the hope that the UK government gamble won't push R too high. Surely they are not completely delusional and must base their decisions on at least some sound scientific advice. If the disease levels do start to rise at least all the UK will have learnt something, though maybe they knew it already. 

Probably best spring in living memory up here in West of Scotland and despite a little rain this week and this coming weekend things till look great, I think vitamin D has a big part to play in immune system response, when most Scottish people are known to be deficient after a long winter, got to the question the current restrictions on sunbathing or reading a book in park etc, especially considering a large proportion of the population live in tenement flats with little or no garden space.

The government obviously messed up the initial response, but I do think current thinking is based on the science, and I still optimistic things will be looking good come the start of June.

That fine but I think the operative word is "hope". I hope that too, or even longer would be great, but there's no evidence for any such length of time yet.

It's only been two days FFS! the transmission takes four days before you get symptoms, if you're nor asymptomatic. How do you know KFC wasn't a big contributor, there's no wide-spread testing so in theory at least KFC might have been the worst spreader of the virus.

I haven't been able to go to work for six weeks, so it's simply not true that , "you could always go to work"  the whole factory was shutdown, that 1,000 people who couldn't do what you suggest they could have.

Yes drive through could be a huge problem but it’s not likely to be for it to be likely there would have to be a plausible mechanism for transmission. 
 

My point was that the law or even guidelines never at any point said that factories had to close. That was my entire point. There are arguments that it should have done but I wasn’t arguing that they should open up just that they always could if they wanted too. 
 

I’m not really sure what your point is.

Reminds me of a point made to me in hospital. I had an insitu melanoma removed and was told to use at least factor 30 suncream and take vit D supplements. If people have fair skin after winter then nowadays many will use suncream, and thus presumably not form their own vit  D (the old especially are often carefully shielded from the sun and are liable to deficiency).

I don't know if it's been posted but the Manchester study has been fairly robustly challenged by people who understand this far better than I do 

https://twitter.com/CT_Bergstrom/status/1261058066495074304?s=19