Loading Notifications...

France: Covid Cases vs Deaths

This topic has been archived, and won't accept reply postings.
 ChrisLeigh19 31 Aug 2020

https://covid19.who.int/region/euro/country/fr

For me, comparing the cases vs deaths in France only confirms theories that (very unfortunately) most of the people who would have died due to Covid in western Europe due to underlying conditions, age etc most likely already have and therefore returning to 'normal life' is the best course.

I see a lot of kickback from people against this and although cases are rising (and I am obviously not advocating unnecessary contact) deaths are not. Surely more damage is being caused not returning to school, work and getting businesses back on track?

I do agree that there is far more testing going on now compared to back in March but even if true that then shows that the mortality rate of Covid is orders of magnitude less than we think?

22
 wintertree 31 Aug 2020
In reply to ChrisLeigh19:

I disagree.

The mortality rate is not fixed but depends on a bunch of factors

  1. Age and health of the infected 
  2. How overwhelmed healthcare is/isn’t.
  3. How much we’ve learnt about clinical care 

With regards 1 - there is now much better continuous testing of staff in care homes and much better protocols over care home visitors than in March.  This along with better awareness of protecting the elderly means the average age of infected is younger - given what we know about the age dependence of mortality a small shift in average age has a big effect

With regards 2 - things were desperate in France (intubated patients being shipped out of Paris to hospitals with spare capacity in specially chartered trains.)

With regards 3, a massive amount has been learnt such as when not  to use an incubator but to use CPAP and the finding that dexamethasone significantly improves survival chances (hypothesised on UKC by some other posters well in advance...)

So, we expect less infected people to be killed by covid now than in March.

Also keep in mind that fatalities lag cases by about 3 weeks.  detected infections aug 8 in France was 1719 (worldometer 7 day moving average), deaths aug 29 in France was 13 (same method).  13/1719 = 0.7% fatality rate.  The reality is higher as testing is at best 70% accurate (studies on false negative rates) and is not exhaustive.  0.7% / 70% = 1%.

If I assume there’s a tail of patients from the March/April peak overlaid with patients from the current peak in the French default deaths data, where those in the tail have taken a long time to die, the daily deaths from the new exponential phase is about 5.  With the same optimistic assumptions (all cases detected, 70% detection rate) that gives a fatality rate of (5/1719)/70% = 0.4% - this is still sufficiently high to count as a major risk for individuals if it goes population wide, and to overwhelm healthcare with all the concomitant risks.

[0.4% to 1%] is a high fatality rate and that’s a lower bound (as not all cases are detected).  There’s a lot of details I don’t know about cases and reporting methodologies in France that have importance to interpreting this.  But critically it suggests extreme caution in interpreting the numbers from France - we must remember the lag between infection, detection of infection and death when considering the early stages of an exponential rise as they lag has massive implications for analysis and interpretation.  Right now France is in the early stages of an exponential rise.  Not considering this lag was at the core of a lot of falsely reassuring posts on UKC in early/mid March that didn’t deliver on their reassurance.

I sincerely hope your take on this is proved correct, but in my considered view it is too soon to make a high quality, evidence based view on the Worldometer numbers (or similar) alone.

> Surely more damage is being caused not returning to school, work and getting businesses back on track?

Its intertwined - a lot more damage will be cause to schools and business if we end up back in lockdown.  Schools in my view are one of the highest priorities (Power, water, healthcare, sewage, food supply are higher) - above various businesses that have been open for the last month.  Something is going to have to give very soon.

Post edited at 23:27
4
In reply to ChrisLeigh19:

> For me, comparing the cases vs deaths in France only confirms theories that (very unfortunately) most of the people who would have died due to Covid in western Europe due to underlying conditions, age etc most likely already have and therefore returning to 'normal life' is the best course.

Or also a possibility: people who now (after the spring) know they are very vulnerable because of age or medical conditions, are desperately trying not to get it - so being much much more careful than many of us are. If either of my parents get covid it wouldn't be great - they know that - so although shielding for the extremely medically vulnerable has "paused", they continue to live very differently from before March.

Didn't the government deem several million people in the UK to extremely medically vulnerable? So with somewhere between 40 and 60 000 deaths, we are clearly nowhere near to a point where covid has 'killed off' all the people who are most likely to die from it.

2
 wintertree 31 Aug 2020
In reply to TobyA:

>  we are clearly nowhere near to a point where covid has 'killed off' all the people who are most likely to die from it.

Agreed. Seroprevalence data for the UK suggests about 8% of people have had it.  I get a similar number from applying reasonable fatality rates to the deaths.  So this suggests - adjusting for improved clinical care, there’s another 400,000 or so in the UK who would die if infected by a high viral load.  SAGE was estimating 85,000 dead this winter from a “reasonable worst case” which seems believable given that the average age of infected is moving ever younger in the UK showing that we’re more effective at protecting the most vulnerable as you say.  (Well, we’re effective at it now but that could change over the next few months with schools, universities and the flu season all increasing R).

Post edited at 23:22
3
 ChrisLeigh19 31 Aug 2020
In reply to wintertree:

Hi wintertree,

In regards to 1, I partially agree, although evidence that the elderly were to be at a far higher risk came out very early after Italy's initial spike of deaths which most attributed to how their family structure works, having both young kids (which went out and more actively spread/caught Covid) and more elderly members of families live in the same house. I would argue that this evidence was known long before the UK's spike.

In regards to 2, I disagree, as I believe this topic was high dramatised by the media. Taking the UK as an example, the emergency Nightingale hospital which was impressively constructed in just 9 days, only took in 54 patients in its lifetime whilst having over 4,000 beds before being shut down (and the 12 remaining patients were transferred to other hospitals). In both France and the UK and other European countries, patients being transferred from hospitals to hospitals is very common before and during Covid simply to share the load/maximise use of facilities/utilise specialists and specialist equipment. Although done with greater urgency, this at first apparent shock headline isn't such a problem when you realise neither France nor the UK reached full capacity of their healthcare infrastructure. 

In regards to 3, I can't comment much as I know somewhat little about how clinical care has improved.

I don't quite agree with your math however-you claim 0.4% to 1% to be the lower bound, yet you say in the same sentence that not all cases are detected. As we both agree that realistically there are/have been far more cases than we have tested for, surely that would decrease the mortality rate? Using your example, 13 divided by (a number bigger than 1719) would result in a smaller percentage, therefore decreasing the mortality rate, and therefore making ~0.7% the upper bound?

In response to the lag time, I believe the only lag time we will see is vastly reduced numbers of deaths due to flu this winter. People forget that (again very unfortunately) the average life expectancy of people put into care homes in the UK is roughly between 3-6 months. People put into care homes in the UK are almost always one way or another very sick, and again unfortunately when flu season comes around a large amount of them die. I strongly believe that Covid has just brought flu season half a year early, and that the total death tally this year will be within 20% of any other year.

7
 wintertree 31 Aug 2020
In reply to ChrisLeigh19:

> In regards to 2, I disagree, as I believe this topic was high dramatised by the media. Taking the UK as an example, the emergency Nightingale hospital which was impressively constructed in just 9 days, only took in 54 patients in its lifetime whilst having over 4,000 beds before being shut down (and the 12 remaining patients were transferred to other hospitals). In both France and the UK and other European countries, patients being transferred from hospitals to hospitals is very common before and during Covid simply to share the load/maximise use of facilities/utilise specialists and specialist equipment.

I disagree.  The entire NHS was reconfigured almost totally - it wasn’t just nightingale hospitals.  Almost every routine operation and cancer treatment was cancelled and we’ll likely have a lot of excess deaths due to cancer (in turn due to covid).  The number of ITU beds was something like doubled by cancelling everything else and reconfiguring. This wasn’t quite enough hence starting to use the field hospitals. The nightingale spaces were in part insurance as we were so close to the wire and I think in part forwards looking to winter 20/21.

> Although done with greater urgency, this at first apparent shock headline isn't such a problem when you realise neither France nor the UK reached full capacity of their healthcare infrastructure. 

I can’t speak for France, but the UK exceeded the usual capacity by cancelling most other healthcare.

> In regards to 3, I can't comment much as I know somewhat little about how clinical care has improved

Data is out there to read.

> I don't quite agree with your math however-you claim 0.4% to 1% to be the lower bound, yet you say in the same sentence that not all cases are detected.

I didn’t put that across very well, I’m trying not to go on indefinitely.  Testing is not exhaustive (not all people are contact traced,) and testing is not 100% effective (time since infection has a relation to false negative rate).  In reality I think the false negative rate outweighs the contribution of non-complete contract tracing *now that contract tracing capacity exceeds daily infections, unlike in march*.  It’s all spitballing but independent comparisons with seroprevalence land in the same ballpark as these assumptions of mine.

> As we both agree that realistically there are/have been far more cases than we have tested for, surely that would decrease the mortality rate? Using your example, 13 divided by (a number bigger than 1719) would result in a smaller percentage, therefore decreasing the mortality rate, and therefore making ~0.7% the upper bound?

You are - accidentally I hope - ignoring that we were massively under testing for the first peak and are now estimated to be catching at least 50% of cases with test/trace.  You also conveniently ignore the false negative rate to the point I start to think you’re pushing an agenda over considering the range of unknowns.  In my view the under-testing is more than compensated for by the false negative rate (70% is the best case for this; time averaged its much worse)

> In response to the lag time, I believe the only lag time we will see is vastly reduced numbers of deaths due to flu this winter

We know deaths lag infections by several weeks with COVID.  That matters, and that is critical to analysing the numbers.

> I strongly believe that Covid has just brought flu season half a year early, and that the total death tally this year will be within 20% of any other year.

I hope you are right.  I remain to be convinced. I certainly couldn’t in good conscience argue for an unwinding of controls based on your belief.  I suggest we park this discussion for 3 weeks and then revisit the French data. 

Unless you can come back with an evidence base for the fraction of cases detected in France being much less than 50% then you’re just spitballing - as am I.  When spitballing in a global pandemic I prefer to err on the side of caution and using what imperfect evidence I have.  If you want to throw caution to the wind, please do advocate for that - but directly not through unevidenced arguments.

Post edited at 00:18
1
 elsewhere 01 Sep 2020
In reply to ChrisLeigh19:

Very wrong. How familiar are you with care homes? This study tallies with my experience of visiting. If you can't get the basic facts right why should your analysis be believed?

"The median length of stay was 19.6 months for all admissions. Median length of stay for people admitted to nursing beds was 11.9 months and for residential beds it was 26.8 months. In the PSSRU study, average length of stay was predicted at 29.7 months following admission."

https://eprints.lse.ac.uk/33895/1/dp2769.pdf

Post edited at 00:18
1
 wintertree 01 Sep 2020
In reply to ChrisLeigh19:

Are you posting from a sock puppet account?

Forgive my paranoid nature, but the flags include:

Prior to this your account only has posts on one previous thread.

Your previous post was on a thread shared with another account I shall call ‘x’

In common with ‘x’, you have the following attributes:

  • You start your reply to me with “Hi wintertree” - something no one else has done, ever, in thousands of messages
  • Your forename starts with a ‘C’; their username is presumably the capitalised first letter of their first name then their familial name with its first letter capitalised.  If my assumption is correct, their first name starts with a ‘C’ as does yours.
  • You both list 7a and E3 as your on sight grades
  • Both profiles have or link to photos of similar looking males with similar looking beards.  The photo on the younger profile looks a lot like a scan of an old fashioned film photograph.
  • The last update to each account was within 8 days of each other.

That’s a lot of coincidences.  Perhaps I’m wrong and have become a paranoid raving lunatic - but I’ve seen enough messages from suspect accounts advocating for a more “relaxed” approach to covid  in the last 6 months that I have no problem putting my paranoia front and centre.  
 

Post edited at 00:40
2
 ChrisLeigh19 01 Sep 2020
In reply to wintertree:

Haha I was starting to write a response to your previous message but I thought I should reply to this first.

> Are you posting from a sock puppet account?

No, the account 'x' I assume you are referring to is one of my climbing partners.

> Prior to this your account only has posts on one previous thread.

I think after this thread I'll keep it that way!

> You start your reply to me with “Hi Wintertree” - something no one else has done, ever, in thousands of messages

Sorry for being polite...

> Your forename starts with a ‘C’; their username is presumably the capitalised first letter of their first name then their familial name with its first letter capitalised.  If my assumption is correct, their first name starts with a ‘C’ as does yours.

We share the same first name actually

> You both list 7a and E3 as your on sight grades

Having been climbing partners we climb very similar grades

> Both profiles have or link to photos of similar looking males with similar looking beards.  The photo on the younger profile looks a lot like a scan of an old fashioned film photograph.

I think you'd find we look very different in person

> The last update to each account was within 8 days of each other.

We have both been climbing a lot recently due to me being between school/uni.

> That’s a lot of coincidences.  Perhaps I’m wrong and have become a paranoid raving lunatic - but I’ve seen enough messages from suspect accounts advocating for a more “relaxed” approach to covid  in the last 6 months that I have no problem putting my paranoia front and centre.  

Again being parters/friends we probably share very similar views. This has given me a good chuckle tho!

Edit: I have made my logbook public if you still don't believe me!

Post edited at 00:57
6
 wintertree 01 Sep 2020
In reply to ChrisLeigh19:

That’s a remarkable number of coincidences by any measure - if you hadn’t both addressed me in the same way - that I’d never seen before - within 24 hours of each other doing so - I wouldn’t have flagged it.  Same first name, same onsight grades, same (otherwise unique) form of address to me (not used to other posters), same sort of views and it transpires climbing partners.  No wonder I got suspicious... 

You are right to flag up ambiguity on my treatment of false negatives rate - - you get different biases if you apply it to detected cases, determination of cause of death or both.  The permutations shift the upper bound on mortality through the range of 0.5% to 1%. How ever one ascribes it, the French figures - when accounting for lag in infection > detection > death do not support the view from your first post; we need to look again in 2-3 weeks time.  Either way, case are more confined to younger people right now so the present fatality rate tells us nothing about what’ll happen if it gets back in to older people in significant numbers.  Your theory that the vulnerable are already dead holds little water as infection rates are so low in the elderly (due to better protection measures) that it’s not being tested to any statistical significance.

Post edited at 01:40
3
In reply to TobyA:

> people who now (after the spring) know they are very vulnerable because of age or medical conditions, are desperately trying not to get it - so being much much more careful than many of us are.

And, conversely, the people who know they are not vulnerable (or think they aren't) are now taking very little care,  and are thus getting infected, but not dying.

My parents are also continuing to shield. More so than previously, as the world around them opens up, and this increasing number of people around them take less care to keep distance. My dad, who commented how eerie it was to see his local town almost empty, is now reluctant to go there at all, except for essential medical reasons, because there are now more people there than ever (it's a coastal resort town), and they're not being careful.

 wbo2 01 Sep 2020
In reply to ChrisLeigh19:

I would be very wary of taking observations from summer conditions and assuming they will apply to the winter, when the NHS is stressed at the best of times, and use that as the basis for a return to normality. 

You may find you like this new normal less than the old.. but life changes with time , usually for the better but sometimes for the worst..

1
 marsbar 01 Sep 2020
In reply to ChrisLeigh19:

I have a friend who hasn't died.  However several months after having the virus she is still not well.  I understand this isn't uncommon. 

In reply to ChrisLeigh19:

The Nightingale hospitals were never going to work.  Taking an exhibition hall and sticking beds in it does not make a functioning ICU or produce more doctors and specialist nurses.

It is also completely wrong to focus only on deaths as if not dying was the same as fully recovering.  

9
In reply to tom_in_edinburgh:

> The Nightingale hospitals were never going to work.  Taking an exhibition hall and sticking beds in it does not make a functioning ICU or produce more doctors and specialist nurses.

I think the plan was for them to have less serious patients on oxygen etc.. but not those who would need icu care, ventilators etc.  

It was a reasonable plan and better than the alternative of leaving them to die on the streets, if covid had turned out to be more severe. 

1
 wintertree 01 Sep 2020
In reply to tom_in_edinburgh:

> The Nightingale hospitals were never going to work.  Taking an exhibition hall and sticking beds in it does not make a functioning ICU or produce more doctors and specialist nurses.

 In my view it's better - more human and more pragmatic - to try and to risk failure than to just let the bodies pile up.  Thank the gods we didn't get there - and it was I think a close shave.  The nightingale hospitals would have been a lot better than not trying.

> It is also completely wrong to focus only on deaths as if not dying was the same as fully recovering

It is completely wrong, but surprisingly common.

It would be a rare disease that kills ~0.5% of those who catch it and leaves the other 99.5% with no long term consequences.  For those still being ventilated or put on ECMO who survive, that's a horrific experience for the body in itself.  You don't just get up and walk out of multiple weeks of sedation and ventilation.

For all this positive thinking that we've killed all those who were going to die in the coming flu seasons, a negative take would be that perhaps 250,000 people have had their respiratory health significantly worsened by the covid that they have shaken off, and go in to this coming flu season weaker than they otherwise would.  

The reduction in deaths from improved clinical care does not logically translate in to a reduction in admissions to hospital either, although it may translate into a reduced time in the hospital.  So the improved clinical care may improve outcomes but it does much less to prevent hospital overloading.

 neilh 01 Sep 2020
In reply to tom_in_edinburgh:

They were out together by a Major in the army who had learnt what to doas he had direct experience of Ebola and had helped control the outbreaks in Africa .I would suggest his experience is more helpful than any of ours!

1
 cp123 01 Sep 2020
In reply to wintertree:

Hi wintertree,

I think you will find I am the OG CXXXXX and ChrisLeigh is an imposter....

Cheers

Chris

1
In reply to ChrisLeigh19:

> For me, comparing the cases vs deaths in France only confirms theories that (very unfortunately) most of the people who would have died due to Covid in western Europe due to underlying conditions, age etc most likely already have and therefore returning to 'normal life' is the best course.

A recent study suggests that in this country less than 6% had antibodies at the end of June, and that's  typical for Western Europe.

https://www.imperial.nhs.uk/about-us/news/largest-home-antibody-testing-publishes-results

Given low infections rates that will not have gone up much.

So "most of the people who would have died due to Covid in western Europe due to underlying conditions, age etc most likely already have" is clearly untrue.

I do see the current mortality rate as good news but it is dangerous to be spreading unfounded theories.

This week the schools go back..... that's a pretty big change and it is sensible to see what the result of that is first. There's a lot of posts like yours online urging us to get back to 'normal life' all in one go...... if things go tits up again (like march), that will do far more harm to the economy than us carrying on like this for a few more months.

Post edited at 10:26
In reply to cp123:

You two climbing together must be fun.

"Good effort Chris!"

"Cheers Chris. Tough pitch that one. I'm building belay, don't take me off yet."

"No worries Chris."

"OK, I'm safe Chris!"

"Excellent Chris. I'm tied in but need to put my shoes on."

"Alright Chris. You are on belay though, so climb when ready Chris."

"Cheers Chris. Climbing."

"Nice one. By the way Chris, don't you think the whole Covid thing is over blown?"

"Hang on a sec Chris, just getting your first nut out... Good! Got it! Yes, absolutely Chris. Possibly part of the New World Order of fear-led mind control*." 

"Yes... oh, it's good finger lock out on the left by the way Chris... you know what we should do, we should go on UKC and argue about it because its an obvious way of changing government policy."

"Good idea Chris! ...Ohhh, that is a good finger lock isn't it?"

And a big ;-) just in case that wasn't obvious from tone.

*See Mr Woods's contributions on the other thread.

3
 DancingOnRock 01 Sep 2020
In reply to wintertree:

Latest figures suggest 0.3-0.7% IFR. 
 

Seroprevelance is looking for Antibodies and there seems to be evidence they’re only detectable between 3 and 14 weeks after infection and then become undetectable. 
 

Many people seemed to be fighting off the infection using T-cells and by having immunity from previous Coronavirus type infections. 
 

If a high proportion of people were already resistant to the disease, it would explain how it could have been here since early February in the community and there were relatively few deaths. Especially in London where I just can’t understand why the disease didn’t appear to infect the whole city inside 6-7 weeks before lockdown. 
 

I think France and Spain are to be watched carefully to see whether something is going on and maybe the 3 weeks to death is actually a lot longer and the virus is going into remission after infection in older people. Although I suspect the modelling should have shown that very clearly if it was happening. I don’t know how sensitive PCR testing would be. Doesn’t AIDS and Herpes do something similar? 

Post edited at 11:37
2
In reply to neilh:

> They were out together by a Major in the army who had learnt what to doas he had direct experience of Ebola and had helped control the outbreaks in Africa .I would suggest his experience is more helpful than any of ours!

Well if they were put together by the army they must be good.   The army is an organisation that does what it is told even when the orders are stupid.   

The reason the Nightingale hospitals were not used is that the hospitals were only allowed to send patients there if they also seconded staff and they thought they'd be better off keeping their staff in the proper ICU than sending them to a shed full of dodgy and unfamiliar equipment.

It turned out that limited numbers of ventilators was not the issue because ventilators were not that effective and even if they had them there wouldn't have been the staff to use them.   The whole idea that flight attendants could be trained in a couple of weeks to look after multiple ICU patients in a shed was bullsh*t.   The same kind of propaganda bullsh*t as trying to design ventilators from scratch rather than increasing production of tested designs.   

The Nightingale hospitals would have been places where large numbers of people went to die.   Which is what we would have needed if we hadn't locked down and where we are headed again if we don't keep R < 1.

4
In reply to summo:

> I think the plan was for them to have less serious patients on oxygen etc.. but not those who would need icu care, ventilators etc.  

Starting to rewrite history there.  The media were full of how many ventilators they were going to buy for the Nightingale hospitals.

Using them for less serious patients might have been more sensible but then you have the issue of how in hell do you look after several thousand sick and infectious people who are not sedated and need to eat and sh*t in a big shed.

> It was a reasonable plan and better than the alternative of leaving them to die on the streets, if covid had turned out to be more severe. 

Yes, it was a reasonable plan but only because they f*cked up so badly they had no time left to do anything else.

Post edited at 11:49
1
 wintertree 01 Sep 2020
In reply to DancingOnRock:

Latest figures suggest 0.3-0.7% IFR. 

Yup; the OPs theory however is that most of the dead who gave us that rate represented the majority of highly susceptible individuals and so any future IFR will be much lower.   https://www.youtube.com/watch?v=XBZUz4C6kqk&

> Many people seemed to be fighting off the infection using T-cells and by having immunity from previous Coronavirus type infections. 

Yes; the T-cell immunity is the wildcard in terms of what we don't yet know at a population level; some suggestion that another potentially non-coronavirus also confers T-cel immunity.   I don't know how many vaccines specifically targeting T-cells are in the pipeline; I'm hoping for some informed comment from a couple of posters though...  As ever I ponder the link between viral load at infection and strength of immunity (both as protection against that load, and as a result of the infection based on the strength of infection) - so "immunity" is not a binary thing.

> I think France and Spain are to be watched carefully to see whether something is going on

Agreed.  Data from the early part of an exponential phase before deaths start ramping up is not a good basis for making predictions. I'm having deja-vu back to March/April...  

 DancingOnRock 01 Sep 2020
In reply to wintertree:

>Yup; the OPs theory however is that most of the dead who gave us that rate represented the majority of highly susceptible individuals and so any future IFR will be much lower.  

 

The IFR has been dropping all the time though. The more testing we do, the lower it is dropping. Even at 1% you would expect  1% of the population to die of old age Or serious illness in one year. The issue was always that these deaths would happen in 2-3 weeks instead of spread across the whole year. We seem to have spread them out (apart from one really horrendous week), across a couple of months. 
 

Cases in the U.K. appear to have been rising, but not exponentially, and deaths are still falling. 

Post edited at 12:12
In reply to tom_in_edinburgh:

> Using them for less serious patients might have been more sensible but then you have the issue of how in hell do you look after several thousand sick and infectious people who are not sedated and need to eat and sh*t in a big shed.

Anyone with experience of event management, or military field operations/ hospitals could easily organise food, water, toilets, showers etc.. it's their day job. 

The UK has many folk who ran hospitals with Ebola in Africa, I think they'll cope with something less serious like covid. 

> Yes, it was a reasonable plan but only because they f*cked up so badly they had no time left to do anything else.

Didn't Saint Nicola also do the same? 

 DancingOnRock 01 Sep 2020
In reply to summo:

It’s the 21st century. Besides, putting old people in field hospitals, many of them with co-morbidities, will kill them pretty quickly. 

Post edited at 12:23
In reply to DancingOnRock:

> It’s the 21st century. Besides, putting old people in field hospitals, many of them with co-morbidities, will kill them pretty quickl

What's the option, you have 50,000 beds sitting vacant waiting for another pandemic, when it comes along all the equipment is obsolete anyway? 

Post edited at 12:26
 DancingOnRock 01 Sep 2020
In reply to summo:

Keep them in the Care Homes. 

In reply to DancingOnRock:

> Keep them in the Care Homes. 

Fine for old folk but the temporary hospitals were built for anyone, not just folk already in care. 

 Roadrunner6 01 Sep 2020
In reply to ChrisLeigh19:

In the US about 94% had underlying conditions, but that can be 'obesity' which affects about 35% of Americans, so you are looking at a good 40-50% of the population with some sort of at risk condition.

I think we can return to a new normal, playing sports, schools, but close bars, no mass indoor gatherings without significant distancing or masks.

 wintertree 01 Sep 2020
In reply to DancingOnRock:

> Cases in the U.K. appear to have been rising, but not exponentially, and deaths are still falling. 

Agreed - PHE data and VictomOfMathematic's much better presentation thereof make it clear that the average age of infection is falling.  So there are two possible things going on:

  1. Better isolation and protection of the elderly
  2. All those who were going to die, have died.

I think it's 1 going off various estimates at the IFR and the independent seroprevalence data.  The OP suggests 2 with no evidence.   One way or another we'll find out eventually.  I suspect that the better isolation and protection of the elderly ins't so sustainable if we go in to winter with an ever-rising baseline infection level in the younger folks.

Post edited at 13:07
 cp123 01 Sep 2020
In reply to TobyA:

Cheers Toby I enjoyed that - its even worse as another person who Chris and I climb with is called Chris as well - leading to email exchanges like this:

Hi Chris,

Yes I'm free, lets invite Chris.

Cheers,

Chris

 wintertree 01 Sep 2020
In reply to Roadrunner6:

> I think we can return to a new normal, playing sports, schools, but close bars, no mass indoor gatherings without significant distancing or masks.

How are the megachurches dealing with that?  They can take half a million dollars in the cash collection on a Sunday service, this must be hitting them hard in the wallet...

 Dr.S at work 01 Sep 2020
In reply to tom_in_edinburgh:

re: type of patient for nightingales - IIRC the london unit was going to have 500 ventilator equiped beds and 2,500 which did not have vent capacity - so space for less serious patients who might need escalating to ventilators, or for step down care after ventilation.

re:equipment - certainly seems like the required equipment mix has changed a fair bit as understanding of the pathogenesis has developed, however whilsts the pathway selected may have not been the best it certainly seemed sensible in the early part of the year to have the equipment available if possible.

You are quite correct that staffing would have been a big issue, and far less effective than a normal ICU, but there was quite a lot of effort (by indiviusal NHS trusts) to procure staff like vets and vet nurses who could be rapidly trained up as respiratory care assistants to supplement medical staff.

In reply to summo:

> Anyone with experience of event management, or military field operations/ hospitals could easily organise food, water, toilets, showers etc.. it's their day job. 

Bollocks.  You are talking about a huge open shed with thousands of people in it many of whom are in extreme pain or dying.   All of whom are infectious with an air borne virus and cared for by untrained staff that need to wear PPE.    They would have been a nightmare of screaming patients, stink of human waste, loud noises from equipment, terrified and totally overwhelmed staff and huge numbers of deaths.   

> The UK has many folk who ran hospitals with Ebola in Africa, I think they'll cope with something less serious like covid. 

What was the death rate in those Ebola hospitals?

> Didn't Saint Nicola also do the same? 

What else was she supposed to do?  Like I said before this was f*cked up to the point there were no other options by not reacting fast enough.

 DancingOnRock 01 Sep 2020
In reply to wintertree:

Have just seen some evidence that wearing of masks increases the proportion of asymptomatic cases. From 40% to 95%. ie masks reduce the severity of the disease. Maybe that’s what’s happening. 

 DancingOnRock 01 Sep 2020
In reply to summo:

But ‘ordinary people’ are in the vast minority of people not requiring ventilators. 

Post edited at 13:50
In reply to cp123:

I did a trip to Lofoten once where I met up with the, at the time, 'other Toby' on UKC to climb for a week. My two Finnish mates who I drove up with were both called Toni. There was plenty of comic potential there - although I was never super impressed that whilst the two Tonis quickly became Big Toni and Little Toni, it seemed the suggestion from everyone else is that two English Tobys should be Thin Toby and Fat Toby. It's not like I was actually fat, I was just a lot less skinny than Thin Toby! 

In reply to tom_in_edinburgh:

> Bollocks.  You are talking about a huge open shed with thousands of people in it many of whom are in extreme pain or dying.   All of whom are infectious with an air borne virus and cared for by untrained staff that need to wear PPE.    They would have been a nightmare of screaming patients, stink of human waste, loud noises from equipment, terrified and totally overwhelmed staff and huge numbers of deaths.   

I think you've watched too many movies . 

> What was the death rate in those Ebola hospitals?

It's irrelevantthe staff with experience of managing infectious diseases that we are talking about. 

> What else was she supposed to do?  Like I said before this was f*cked up to the point there were no other options by not reacting fast enough.

As health, social services etc etc are ALL devolved, if super sturgeon had a better plan she could have implemented it. 

Post edited at 13:54
 cp123 01 Sep 2020
In reply to tom_in_edinburgh:

> Bollocks.  You are talking about a huge open shed with thousands of people in it many of whom are in extreme pain or dying.   All of whom are infectious with an air borne virus and cared for by untrained staff that need to wear PPE.    They would have been a nightmare of screaming patients, stink of human waste, loud noises from equipment, terrified and totally overwhelmed staff and huge numbers of deaths.   

I think to paint the nightingale hospital at max capacity (which except for the wasted resources, including staff time, material and cost we are all glad didn't occur) as a sort of Gone With The Wind battle scene type environment shows an overactive imagination. It also does a disservice to the various organisations which pulled together at short notice, including military and NHS planners and logistical staff.  These people, as their day job, either do contingency worst case planning (military) or have specialist knowledge in organising healthcare (NHS).

1
In reply to DancingOnRock:

> But ‘ordinary people’ are in the vast minority of people not requiring ventilators. 

Isn't that only because iit wasn't so serious and the temporary hospitals weren't needed? 

 DancingOnRock 01 Sep 2020
In reply to summo:

The vast proportion of sick people were the elderly. If you only put the young in the Nightingale hospitals It would have been because the main hospitals were full. 
 

Also we have to remember that when they were building the Nightingale hospitals, they had no idea whether the lockdown was slowing the rate of infection or how many were already infected. They were purely designed as emergency overflow wards. If we had to use them, it would have been because lockdown had failed.

I doubt the general public had fully grasped that fact. 

 DancingOnRock 01 Sep 2020
In reply to cp123:

Tom likes to be quite colourful. 

In reply to DancingOnRock:

Perhaps building them enforced the potential seriousness of the virus, which increased the effect of the lock down and thus reduced the chance they'd even be needed?

The most effect measures are those which everyone thinks with hindsight was overkill. 

I suspect they'd shove whoever they needed to in them, did they really have such an age specific target audience? 

Post edited at 14:58
 Roadrunner6 01 Sep 2020
In reply to wintertree:

Not very well, and being conservative they wont wear masks. God protects.

There was a recent outbreak at a church near us.

#prolife.. unless it doesn't fit your political narrative.

There are fit healthy people getting struck down. There seems to be a genetic factor, possibly a race issue, which is causing some to get sicker. Some of my wife's colleagues have been hit hard and on ventilators yet fairly young and healthy. But as MD's I suspect they got a massive infectious dose.

Since the early days though the amount of hospital infections has plummeted as they now assume everyone is infected. It was one early heart attack victim who infected many of those who responded to the resus call at her hospital.

I was just in a car inspection station and despite mask regulations nobody wore a mask, not even those at the front desk. 

In reply to summo:

> I think you've watched too many movies . 

I think you've not read enough about people's experiences in hospital with Covid.  That didn not come from a movie it came from someone describing what it was like to be in a ward with Covid and the person next to them screaming.   Imagine 4,000 people in an open shed instead of three or four in a small room in a modern hospital.

> It's irrelevantthe staff with experience of managing infectious diseases that we are talking about. 

Whether the patients survived is highly relevant.  'African Ebola Hospital' is not the gold standard for patient care.

> As health, social services etc etc are ALL devolved, if super sturgeon had a better plan she could have implemented it. 

The better plan would have been not to get into the situation in the first place.  The way to do that would have been to react and close the f*cking border when it took off in Italy rather than let people go on skiing holidays and come back straight into the community with no isolation.  And when it kicked off in London have a lockdown of London and keep it under control in the rest of the country. 

None of this is rocket science, other countries were watching their borders and doing local lockdowns, the Tories were obsessed with Brexit and listening to that moron Cummings and his 'herd immunity' and did nothing.

Post edited at 15:54
In reply to cp123:

> I think to paint the nightingale hospital at max capacity (which except for the wasted resources, including staff time, material and cost we are all glad didn't occur) as a sort of Gone With The Wind battle scene type environment shows an overactive imagination.

No it's a realistic projection of what it would have been like if we hadn't locked down, we'd have had hundreds of thousands of cases and thousands of dying people in these conference centre sheds.  We actually got about 60,000 people dying and a lot more getting seriously ill.  The predictions for letting it run throught the community were 250k deaths if I remember right.   

It would have been pretty f*cking horrific.  What actually happened was horrific for the people who worked in hospitals, the rest of us were isolated from it,  multiply that by ten in terms of numbers of patients and imagine it happening in a shed instead of a hospital and most of the staff untrained.

Post edited at 16:02
1
 tom r 01 Sep 2020
In reply to wintertree:

"Also keep in mind that fatalities lag cases by about 3 weeks.  detected infections aug 8 in France was 1719 (worldometer 7 day moving average), deaths aug 29 in France was 13 (same method).  13/1719 = 0.7% fatality rate.  The reality is higher as testing is at best 70% accurate (studies on false negative rates) and is not exhaustive.  0.7% / 70% = 1%.

If I assume there’s a tail of patients from the March/April peak overlaid with patients from the current peak in the French default deaths data, where those in the tail have taken a long time to die, the daily deaths from the new exponential phase is about 5.  With the same optimistic assumptions (all cases detected, 70% detection rate) that gives a fatality rate of (5/1719)/70% = 0.4% - this is still sufficiently high to count as a major risk for individuals if it goes population wide, and to overwhelm healthcare with all the concomitant risks."

I agree with pretty much all of your post but don't quite get this math. You are saying that instead of 13 deaths there will be missed deaths (assuming they are counting deaths by testing dead people for coronavirus) so 13 / 70 * 100 = 18.57 deaths divided by number of cases 3 weeks prior to death. Surely the 70% detection rate would be reflected in false negatives of the tests on alive people so 1719/70 *100 = 2455 cases so still 0.7?

Also I bet only a small fraction of the asymptomatic cases are getting detected.

Post edited at 16:03
 Blunderbuss 01 Sep 2020
In reply to tom_in_edinburgh:

Was wee Jimmy Crankie pushing the UK government for an earlier lockdown?

4
 wintertree 01 Sep 2020
In reply to tom r:

> I agree with pretty much all of your post but don't quite get this math. 

It's not all maths - partly cargo cult stuff gazing at graphs.  To explain in more steps...  I took the worldometer 7-day average for deaths and ignored the last couple of weeks.  It is then about 10 deaths a day.  I then assume - as was the case in the UK before the reporting rules were changed - that most of these deaths are patients who were infected more than a month ago but took a long time to die.  I then project that ahead as a baseline of deaths unrelated to the current spike - so I treat the ~15 deaths/day currently seen as the sum of this baseline ~10 deaths/day from the March/April spike and ~5 deaths/day from the new spike.  So now I'm assuming only 1/3rd of the deaths relate to the new spike to get a lowball estimate of the fatality rate.  I'm pretty sure I goofed in the next step of the maths and should have done (5 deaths/day) / ((1719 detected cases per day) / (70% false negative rate)) to get the estimate which gives a 0.2% fatality rate.  I put the 70% in the wrong place before - that's what you get for doing shoddy estimates in a UKC edit box and not with pen and paper!  Although this detection rate swings both ways - there are about 20k excess deaths in the UK of people that's either from an epidemic of being driven over by busses (an often proposed interpretation) or from people not being tested before they died.  

So that's my generous assumption on the fatality rate.  My less generous one assumes all these deaths are within a month of infection and so ascribable to the new exponential rise in cases, and is 3x higher.

A fatality rate in the range 0.2% to 0.6% fits with the data showing the average age of infections has fallen.  It in no way proves the OP's feel good claim that all those who contributed to the initial, higher fatality rate have been killed and so this winter is going to be all peachy.  Me - as well as not agreeing with the OP, I worry about those who lived and had their health clobbered by covid going in to the winter.

> Also I bet only a small fraction of the asymptomatic cases are getting detected.

Perhaps.  Going off cases/deaths and the independent ONS Pilot Infectivity survey in the UK I reckon test and trace is identifying about half of the infected people (and asking more infected by false-negative to self isolate as part of the process); apply the same ratio to the UK or France (this really is playing fast and loose with everything now...) and we get a fatality rate of 0.1% to 0.3% - still giving between 60k and 180k deaths in each country if things get out of hand in the winter.  But these low numbers are with things largely confined to younger people, a likely unsustainable proposition if cases rise significantly. 

Now, as it stand I'm optimistic that things won't get out of hand - unless a lot of people start acting as if the risk is passed.  

Everything I've done is extreme navel gazing on limited data and without as much understanding of the French reporting as I have of the UK, and with some sloppy porting of assumptions from the UK to France.  Still, it's a lot more than the OP has done with their wishy washy attempt to draw an over-arching conclusion from data in the early stages of an exponential rise where deaths significantly lag infections.  

Post edited at 16:27
In reply to Blunderbuss:

> Was wee Jimmy Crankie pushing the UK government for an earlier lockdown?

That's not snp policy to remain in office. They have to make it sound like they could do better if independent, only never actually be responsible, otherwise folk will hold them to account when they can't blame Westminster for health related matters for example, even though health is already devolved. I suspect the average snp voter has no idea how much is already devolved and they should be looking to holyrood for accountability not Westminster. 

5
In reply to wintertree:

I'm exaggerating here but the data is basically irrelevant to your discussion above, what you've got here is a conflict of attitude.

There's not enough reliable data to support reliable conclusions. This seems to be agreed.

On the one side, can't tell what's going to happen but if we get it wrong it's going to be hellish, so let's play it safe.

On the other side, can't tell what's going to happen but there's nothing at the moment to show that it's going to be hellish, so let's return to normal.

Post edited at 18:37
 wintertree 01 Sep 2020
In reply to Michael Hood:

You could be right.  I have said openly we should wait a couple of weeks before we can meaningfully interpret the French data on this new spike.

> On the other side, can't tell what's going to happen but there's nothing at the moment to show that it's going to be hellish, so let's return to normal.

The problem is the view from this side is logically flawed - just as it was back in March.  In the early stages of an exponential growth phase such as France now has, absence of evidence is not evidence of absence.  We just don’t know.

Further, I and others would - and have - argue that there is sufficient data to cast this other side’s premise into serious doubt.

I have been consistent in my views that when we don’t have enough data in the early stages, erring on the side of caution is the better approach.  The other side tries to make the case that it will be fine based on wolly references to data rather than putting their view as you do.  You frame a much more honest debate. We didn’t err on the side of caution in March/April and we saw where they ended up.  With France we’re back in an early stage and the “we’ve all had it” argument is no more compelling to me now than when some posters tried to sell that crock’o’shite back in March.  They were wrong then and I’ve seen no compelling evidence they are right now, and plenty of evidence to question it (seroprevalence).

A while ago I made a cartoon to illustrate this sequence of arguments from the first time around...  A flawed argument based on a premise that I believe is far from being justified.

Let’s hope the next version or this cartoon shows me being wrong.

Post edited at 18:51

In reply to wintertree:

Oh I agree with you, especially the "absence of evidence is not evidence of absence" bit.

I keep on wondering why the number of deaths hasn't gone up yet, seeing the number of new cases and the stupidity everywhere of people disregarding the simple steps to keep the spread down and seeing the increase in deaths in some other countries. Hopefully the virus is less deadly now (for whatever reason) than it was earlier in the year, but I fear it's going to hit hard here again, just a little later.

Maybe all the deaths are happening 29 days after being tested - must have another look at excess deaths.

 Blunderbuss 01 Sep 2020
In reply to anyone:

Spain is averaging only about 30 deaths a day despite cases starting to rocket about 5 weeks ago...

 wintertree 01 Sep 2020
In reply to Michael Hood:

> I keep on wondering why the number of deaths hasn't gone up yet [...] Hopefully the virus is less deadly now (for whatever reason) than it was earlier in the year

As Roadrunner said on another thread recently, perhaps the increased awareness and use of masks and hand washing is reducing viral loads to the point that many people who get it, get a milder dose and so the fatality rate is down.  If that's a part of it, it's incredibly sad as it means that following disease control knowledge that's 400 years old would have been more effective than "waiting for the right time to lock down"  

> Maybe all the deaths are happening 29 days after being tested - must have another look at excess deaths.

An interesting thought; the dexamethasone trial showed a significant decrease in mortality; I have't seen any data on what it does to how long it takes the dead to die; if it slows the progress of the disease but fails to stop it for some, it could conceivably be stretching out hospital stays of those not saved by it.   It does my head in trying to keep all the different interpretations and possibilities clear, and the regular changes to reporting measures and methods doesn't help. 

 wintertree 01 Sep 2020
In reply to Blunderbuss:

> Spain is averaging only about 30 deaths a day despite cases starting to rocket about 5 weeks ago...

This gives a fatality rate of 0.7% using the 7-day moving average for deaths (30, Aug 31) and detected cases 3 weeks previously (4486, Aug 10).  Again, it's probably lower by a factor 2 or so as a result of incomplete test/trace and false negatives in the test results (although false negative may bias the death count in a compensating way); it's the same sort of ballpark as my estimates for the French case and in no way justifies the OP's claim that  "the mortality rate of Covid is orders of magnitude less than we think".   It's pretty much in line with estimates of the fatality rate from the first spike in cases.  Unlike the French data there is no possibility of a drawn-out tail of deaths from the first spike confusing the analysis - they had bottomed out at zero before this rise.

To show their claim, the OP needs to produce evidence that detected cases in France and Spain are orders of magnitude less than 100%; they haven't even produced evidence it's one order of magnitude less.  My limited understanding is that test/trace is running at a similar level in the UK, France and Spain, and that it's far too effective to justify even one order of magnitude change in the true fatality rate.

What I think is misleading to many people is that the size of the first and second spikes in detected cases are similar for many countries on the worldometer plots (and national dashboards) but the deaths are much lower this time round; that's little to do with changes in the true fatality rate and massively to do with the fact we were detecting almost no cases in March/April and were almost exclusively detecting seriously ill cases.   To illustrate this I've taken the worldometer daily cases plot for the UK and attached it.  I've then put a modified version in where I've shown my crude estimate at real daily infection numbers in as well - this makes it clear how small the new growth phase currently is compared to the first go around.  The disparity is due to orders of magnitude more testing with time.  However there is no such differential scaling in death counts - which is why they seem so low compared to the last time around.  My estimate of peak daily infection comes from two methods - one is using reported true fatalities rates and death figures and the other is in the seroprevalance data ramping up and estimating how many people a day would need to be infected to make the ramp.  When I did this I got about 190,000 infections/day by both super crude estimates.  

I think a very similar analysis applies to both the French and Spanish Worldometer plots - it is an incredibly misleading pair of graphs to look at if you don't put all the pieces together.  I can see how the someone could reach the OP's conclusion by just looking at the Worldometer Spain or France pages and not considering everything else that's known, but I'm afraid it just doesn't wash.

Post edited at 20:37

 DancingOnRock 01 Sep 2020
In reply to wintertree:

>it could conceivably be stretching out hospital stays of those not saved by it. 

 

The numbers in ITU are down as well. 
 

The data on mask wearing reducing viral load is compelling. 95% were asymptomatic in one outbreak. 

 NathanP 01 Sep 2020
In reply to summo:

> That's not snp policy to remain in office. They have to make it sound like they could do better if independent, only never actually be responsible, otherwise folk will hold them to account when they can't blame Westminster for health related matters for example, even though health is already devolved. I suspect the average snp voter has no idea how much is already devolved and they should be looking to holyrood for accountability not Westminster. 

You should understand by now, it is quite simple: If it is a Bad Thing then that is the fault of the Tory Westminster Government. If it is a Good Thing then that is thanks to the SNP Government but it could have been a much Better Thing if only Scotland could be free from English Tyranny. I hope that has helped.

2
In reply to NathanP:

I thought this thread was about France?

Post edited at 20:47
 neilh 01 Sep 2020
In reply to tom_in_edinburgh:

I will stick with the major on this one as clearly he had far more experience of these situations than you and I 

In reply to DancingOnRock:

>  Cases in the U.K. appear to have been rising, but not exponentially, and deaths are still falling. 

They are rising more than exponentially, it's just that the doubling rate is very low compared to what it was back in the spring. 7 day average of daily percentage rise in cases bottomed out at 0.19% around 10th July, it's now up to 0.4%.  Back in April it was 13%. 13% means doubling in 6 days, 0.4% means doubling in 172 days.

Early July was the lowest point in the stats, since then the raw infections per day, percentage infections per day, and  number of live cases have all steadily increased. Testing has increased, but that doesn't explain all the increase - the percentage of positive tests has increased.  The current level of restriction & t&T effectiveness is not reducing the prevalence, and the opening of schools (and associated return to the workplace for parents) along with bad weather keeping people indoors is going to increase transmissions further still.

In my opinion the death rate is lower because we've learnt to keep it out of care homes whose residents are 400 times more likely to die than a 20-something, the reduced prevalence has reduced the viral load that kills younger people, treatment is somewhat better than it was, and (as per Wintertree's calculations) the numbers infected were massively under-reported due to a lack of testing in the spring.  We know from mass testing (e.g. Jersey's arrivals border testing) that the number of asymptomatic cases tend to outweigh the symptomatic heavily, and thus when only symptomatic patients were tested in the spring under half (at best) of cases were detected.

For Wintertree - Jersey have been testing on arrival virtually every single person entering the island since the start of July, most of which will be from the UK or locals visiting the UK. Stats are available to download from https://www.gov.je/Health/Coronavirus/Pages/CoronavirusCases.aspx

In reply to Toerag:

Regardless of anything else, that dashboard from Jersey (along with many other country's) puts the UK government's dashboard to shame.

And the ONS can't even include Scottish results (and NI) to get a complete UK picture - shamefully poor.

 DancingOnRock 02 Sep 2020
In reply to Michael Hood:

To give you an idea of scale those figures are 57,000 test in two months for inbound passengers. 
 

Heathrow alone normally does 10m people in that period.

In reply to neilh:

> I will stick with the major on this one as clearly he had far more experience of these situations than you and I 

To claim the Nightingale hospitals represent the best advice of experts misrepresents the situation.

If you ask an expert in the middle of a pandemic with hospitals already struggling what can be done they might well say 'use a conference centre'. That's not because using a conference centre is the optimal thing to do in their opinion, it is because using a large existing building designed to be fitted out quickly is the only practical option when you've only got a few weeks.   

The UK had pandemic planning exercises years before Covid and the Tories did not carry out the recommendations to save money.  China made pandemic plans after SARS and had a prefab custom designed fever hospital assembled on a parking lot near the main hospital in Wuhan and they brough in medical staff from other regions to staff it.  They could transfer resource from other regions because they isolated Wuhan and stopped it spreading.

 neilh 02 Sep 2020
In reply to wintertree:

You mean simple things like people washing their hands which Whitty has been saying since this started is the simpliest control measure.

Going back to the start of this I gave my staff a right earful over their hand hygiene as I was shocked that nobody had taken it on board.

I bet this is replicated across alot of companies etc. You do notice that people are washing their hands more frequently.

Usually the simpliest things are the best.

 DancingOnRock 02 Sep 2020
In reply to Toerag:

You cannot take two points and create an exponential curve. A linear growth will double every x days and then double again after another x days. It’s not growing exponentially in the U.K.  It may well grow exponentially where there is an outbreak but as soon as you contain that outbreak the growth will reduce in that area. 
 

Exponential growth over the whole country would require outbreaks to be uncontrolled. Which they aren’t. 
 

What you are seeing is a few individual events leading to a hundred or so new infections each time. Bear in mind these new cases only last for a fortnight or so if they are asymptomatic. 

Post edited at 10:13
2
 DancingOnRock 02 Sep 2020
In reply to tom_in_edinburgh:

>They could transfer resource from other regions because they isolated Wuhan and stopped it spreading.

 

And they are a massively socialist country whose citizens do exactly what they’re told when they’re told. 

 wintertree 02 Sep 2020
In reply to tom_in_edinburgh:

> They could transfer resource from other regions because they isolated Wuhan and stopped it spreading.

Bit of a pipe dream to apply that to the UK though.  By whatever means the virus came to be circulating in index case near Wuhan and spread out geographically making isolation and containment an effective strategy.  Compare to the UK where we had the virus landing all over the country in over 1,700 identified importation events.  It may have taken off in London first, but it was everywhere.  What we should have done was stopped all entries to the country in February except returning nationals and residents and had them self isolate for 14 days.

In reply to wintertree:

> What we should have done was stopped all entries to the country in February except returning nationals and residents and had them self isolate for 14 days.

oh... those carefree days in February, when Bojo and pals thought they'd got it all figured out

https://bylinetimes.com/2020/04/17/the-coronavirus-crisis-boris-johnson-said-superman-brexit-britain-would-take-advantage-of-the-pandemic-and-go-its-own-way-seven-weeks-before-lockdown/

 DancingOnRock 02 Sep 2020
In reply to wintertree:

Except in February we didn’t know 2 things. 
 

1. Asymptomatic transmission. 
2. People were asymptomatic for up to 10 days. 
 

So you’re trying to convince about 10m people returning to the U.K. to isolate for 14 days. 

2
 freeflyer 02 Sep 2020
In reply to Mike Stretford:

Perhaps it's time for BoJo and his motley crew to hands the reins over to local government, which should surely have happened months ago.

https://www.bbc.co.uk/news/uk-england-manchester-53995677

The US seem to be doing ok with this plan, as Meathead and team have failed to do anything except for sending out a few unmarked vans to remove anti-populist demonstrators.

On the OP topic, France (where I'm currently sitting) seems to be in a similar situation, with cities accounting for most of the cases. Irritatingly they only seems to report hospitalisations by department, not infections - if anyone can help with this I'd be grateful.

Where I am currently (Morvan) there's next to no cases, and even Savoie seems to be ok; good discipline with the masks everywhere.

In reply to DancingOnRock:

> You cannot take two points and create an exponential curve. A linear growth will double every x days and then double again after another x days. It’s not growing exponentially in the U.K.  It may well grow exponentially where there is an outbreak but as soon as you contain that outbreak the growth will reduce in that area. 

> Exponential growth over the whole country would require outbreaks to be uncontrolled. Which they aren’t. 

> What you are seeing is a few individual events leading to a hundred or so new infections each time. Bear in mind these new cases only last for a fortnight or so if they are asymptomatic. 


You don't understand what exponential growth is do you? Exponential means increasing at the same percentage of the updated value no matter what that percentage is. Compound interest is exponential. Infection spread when R>1 is exponential.   Exponential growth exhibits as a straight line on a log scale graph. Worldometers shows that straight line.  It might be a very shallow gradient (because R is only just above 1), but it's still a straight line.  The steepness of the graph depends on the growth factor (R value in this context). If a logscale graph curves upwards (like France's is now) the growth is still exponential but the R value is increasing and thus the doubling time is decreasing.  The two dates I gave were examples, I've been computing the rate of increase using stats from worldometers since March.   Example  - 10th July 288133 cases, an increase of 0.2% over the 287621 cases the day before, which was 0.2% above the 286979 the day before, which was 0.2% above the 286349 the day before that. 31st august 335873, 0.4% above the 334467 the day before , which was 0.5% of the 332752 the day before that, which was 0.3% of the 331644 the day before that.  Both date series demonstrate exponential growth, but the current rate of growth is roughly twice that of July's.

Linear growth means a constant x cases rise per day irrespective of the number of cases. For the July dates above the rise was ~550 cases per day. Now it's 1,100+.   Linear growth would mean 550 per day in July, and 550 per day now.

Post edited at 12:21
In reply to DancingOnRock:

> Except in February we didn’t know 2 things. 

> 1. Asymptomatic transmission. 

> 2. People were asymptomatic for up to 10 days. 

> So you’re trying to convince about 10m people returning to the U.K. to isolate for 14 days. 

It was towards the end of February but that's what South Korea did, and it worked for them. Rather an extreme example of a country who followed a 'precautionary principle', but our government followed the opposite approach. 

It's pretty obvious that government and Sage members were trying to 'guess' what the virus would do and what people would do. Plan A was 'herd immunity', a completely untested and unproven approach that would have led to disaster. Thankfully, people and organisations started to make their own decisions and government ended up following them.

Post edited at 12:24
 DancingOnRock 02 Sep 2020
In reply to Toerag:

I have a degree in electronics. I fully understand what exponential growth is. 

Worldometer shows daily new cases on a linear scale.

500 cases one day and 1000 cases the next just means doubling of cases between two dates. If it’s 1500 cases the next day the growth is linear with m=500. If it was 500 cases one day and 500 cases the next day. It’s linear but with m=0 and no growth. 
 

Post edited at 12:34
In reply to DancingOnRock:

> >They could transfer resource from other regions because they isolated Wuhan and stopped it spreading.

> And they are a massively socialist country whose citizens do exactly what they’re told when they’re told. 

The UK is now doing regional lockdowns.  Italy had regional restrictions around its hotspot in the north.

People are scared of catching Covid and willing to accept restrictions.  What it needs is a competent government with communications skills explaining why they are imposing restrictions and why the restrictions will make it less likely you will catch Covid.   

The Tories are not competent,  they have god awful communications, they don't want to engage with Covid because it distracts them from Brexit which is the only thing they care about and they are corrupt as f*ck taking every opportunity to give hundred million quid contracts to their mates.

2
In reply to wintertree:

> Bit of a pipe dream to apply that to the UK though.  By whatever means the virus came to be circulating in index case near Wuhan and spread out geographically making isolation and containment an effective strategy.  Compare to the UK where we had the virus landing all over the country in over 1,700 identified importation events.  It may have taken off in London first, but it was everywhere.  What we should have done was stopped all entries to the country in February except returning nationals and residents and had them self isolate for 14 days.

There was no excuse for letting people go on holiday to Italy and letting people walk through airports with no testing or enforced isolation.    They had seen the pictures of what was going on in Wuhan and Italy, they knew it was a sh*t storm and they did absolutely nothing to protect the UK, which, as an island, could have been protected.   There should never have been 1,700 importation events.   If it started to kick off in London anyway they should have had a local lock down really fast, like they are finally doing now.

 DancingOnRock 02 Sep 2020
In reply to tom_in_edinburgh:

It has nothing to do with government. It has to do with main stream media who have no science based journalists and are only interested in selling copy. 
 

The government could have been the most competent going - the daily press briefings involving Jenny Harries were clear and concise and she did not let herself be drawn by the press questions - but the press quite frankly were appalling. 
 

If you’re going to blame anyone, blame main stream media, who are still doing a very poor job of communicating facts. 

2
 baron 02 Sep 2020
In reply to tom_in_edinburgh:

Lockdowns and travel restrictions only work when either the government has the power and resources to enforce them or the population sees the need for them and cooperates.

Neither of these was going to happen in February and the early part of March.

2
In reply to DancingOnRock:

> It has nothing to do with government.

Ok I'll post this again.

https://bylinetimes.com/2020/04/17/the-coronavirus-crisis-boris-johnson-said-superman-brexit-britain-would-take-advantage-of-the-pandemic-and-go-its-own-way-seven-weeks-before-lockdown/

The government clearly had preconceptions about this which obviously steered the early policies.

This current right wing politics of deny everything, take no responsibility, blame anyone and everyone else....... really is pathetic. I'm not surprised most Tory supporting posters are anonymous.

Post edited at 13:03
 DancingOnRock 02 Sep 2020
In reply to Mike Stretford:

Did you not listen to the daily briefings? 
 

They were perfectly clear to anyone who listened. It was almost as if the press had written their questions in the morning and were going to ask them anyway, regardless of what was said in the briefings. And when they didn’t get a reply that fitted their agenda, wrote something else instead.

3
In reply to DancingOnRock:

> Did you not listen to the daily briefings? 

Sure did, there was this one on the 3rd March

https://www.theguardian.com/politics/2020/may/05/boris-johnson-boasted-of-shaking-hands-on-day-sage-warned-not-to

“I was at a hospital the other night where I think there were a few coronavirus patients and I shook hands with everybody, you will be pleased to know, and I continue to shake hands,” he said. “People obviously can make up their own minds but I think the scientific evidence is … our judgment is that washing your hands is the crucial thing.”

Do you actually believe your own bullshit?

 baron 02 Sep 2020
In reply to Mike Stretford:

> Sure did, there was this one on the 3rd March

> “I was at a hospital the other night where I think there were a few coronavirus patients and I shook hands with everybody, you will be pleased to know, and I continue to shake hands,” he said. “People obviously can make up their own minds but I think the scientific evidence is … our judgment is that washing your hands is the crucial thing.”

> Do you actually believe your own bullshit?

Weren’t their claims at the time that Johnson hadn’t actually shaken any COVID patients hands?

2
 DancingOnRock 02 Sep 2020
In reply to Mike Stretford:

The advice at the time was to wash hands. 
 

Do you think that publishing photographs of him shaking hands and highlighting the fact was ‘in the public interest’ at the time, or a dangerous thing to do and likely to undermine public adherence? 
 

It seems to me that you’re agreeing with me on the front that the media are only interested in selling copy and are not interested in helping stop the spread of Coronavirus.

2
In reply to DancingOnRock:

> Do you think that publishing photographs of him shaking hands and highlighting the fact was ‘in the public interest’ at the time, or a dangerous thing to do and likely to undermine public adherence? 

You are trying to shoot the messenger, and I'm actually pretty alarmed that Tory supporters are at that stage.

The PM is responsible what what he says and does, and the consequences of that. Expecting the media to filter what they report of the PM, to make him look better, is pretty chilling.

 wintertree 02 Sep 2020
In reply to DancingOnRock:

> Except in February we didn’t know 2 things. 

> 1. Asymptomatic transmission. 

The studies on asymptomatic transmission from back then that I have repeatedly linked to in the past for you to oddly dismiss as hindsight, despite them being medically evidenced cases of such dating back to early Feb...

In reply to DancingOnRock:

> I have a degree in electronics. I fully understand what exponential growth is. 

> Worldometer shows daily new cases on a linear scale.

It's the total number of cases you should be looking at. Change the scale to Log.

> 500 cases one day and 1000 cases the next just means doubling of cases between two dates.

Correct. If it goes to 2000 the next day, and 4000 the day after, that's exponential, with the daily increase percentage = 100, yes?  If it was doubling every other day, the daily increase percentage would be 50%, yes?  What we're currently seeing is a daily increase percentage of 1.8%, resulting in a doubling every 39 days. In June it was negative i.e. live cases were reducing, but since the 16th July it's been positive, and as high as 3%. A regular positive percentage increase = exponential growth, a regular positive numerical increase = linear growth.

If it’s 1500 cases the next day the growth is linear with m=500. If it was 500 cases one day and 500 cases the next day. It’s linear but with m=0 and no growth. 

Correct. But that's not what we're seeing though. The number of new cases each day is generally higher than the previous day's increase, and expressed as a percentage the percentage remains constant or increases.  Let's look at the raw increases from the dates I listed:-

10th July 288133 cases, an increase of 512 (551 7 day average) over the 287621 cases the day before, which was 642 (552) above the 286979 the day before, which was 630 (543) above the 286349 the day before that.

31st august 335873,  1406 (1323) above the 334467 the day before , which was 1715 (1244) above the 332752 the day before that, which was 1108 (1164) above the 331644 the day before that.  The trend is clear to see on the new cases graph on worldometers - the number of new cases per day has been slowly increasing since early July. France and Germany show the same trends, but at higher percentage rates.

In reply to Mike Stretford:

> It was towards the end of February but that's what South Korea did, and it worked for them. 

It's what the government here did too. 125 days virus-free now.

 DancingOnRock 02 Sep 2020
In reply to Mike Stretford:

I did not hour for this government and ‘to make them look better’ is a far reach from publishing things that are not in the public interest. 
 

There is a pandemic on and the press are reporting negatively. That is not in the public interest and risks damaging the message. They are the messengers and they have a duty to ensure that the message is delivered correctly and fully without missing out parts in order to sell more papers. 
 

The media in this country are biased. The Telegraph and/or the Guardian, just to mention two, either of them. They have an agenda. The fact that you use the main stream media to reinforce your own dogma, so much so that you actually believe I’m a Tory supporter, is also chilling.  

4
 DancingOnRock 02 Sep 2020
In reply to wintertree:

I’m not denying there were no studies supporting it. ‘Studies’ and ‘knowing’ are two different things. As you keep saying - you would rather err on the side of caution. That’s not the way the world works. 

In reply to DancingOnRock:

> I’m not denying there were no studies supporting it. ‘Studies’ and ‘knowing’ are two different things. As you keep saying - you would rather err on the side of caution. That’s not the way the world works. 

The 'precautionary principle' is how some of the world works.....it's a matter of public record so don't bother denying it. Those countries have fared much better.

Post edited at 14:54
 DancingOnRock 02 Sep 2020
In reply to Toerag:

No. It’s daily cases that show the rate of growth. The total number of cases never goes down. It’s accumulative, always increases and irrelevant in this context. 
 

After 21 days, people either recover or they die. 

In reply to DancingOnRock:

> I did not hour for this government and ‘to make them look better’ is a far reach from publishing things that are not in the public interest. 

> There is a pandemic on and the press are reporting negatively. That is not in the public interest and risks damaging the message. They are the messengers and they have a duty to ensure that the message is delivered correctly and fully without missing out parts in order to sell more papers. 

And they did deliver the message 'correctly and fully', including the part where Bojo said he'd been shaking lots of hands.

In a national emergency it is up to the PM to deliver the right message. Expecting the media to decide which bits to filter out, even in the interests of public health, is unrealistic.

Post edited at 14:55
 wintertree 02 Sep 2020
In reply to DancingOnRock:

> I’m not denying there were no studies supporting it. ‘Studies’ and ‘knowing’ are two different things.

There were multiple independent studies and there was the implicit evidence for widespread asymptomatic transmission in just how rapidly and well it had spread in Wuhan and then Northern Italy.  

> As you keep saying - you would rather err on the side of caution. That’s not the way the world works. 

It's the way many parts of the world responded to this pandemic, and it oddly enough those parts of the world don't have 60,000 dead and the most damaged economy in the G8.

In reply to DancingOnRock:

> There is a pandemic on and the press are reporting negatively. That is not in the public interest and risks damaging the message. They are the messengers and they have a duty to ensure that the message is delivered correctly and fully without missing out parts in order to sell more papers. 

And anyway, this was a live televised press conference

https://www.theguardian.com/world/video/2020/mar/27/i-shook-hands-with-everybody-says-boris-johnson-weeks-before-coronavirus-diagnosis-video

what did you expect the broadcast media to do? Cut him off air?

In reply to DancingOnRock:

> No. It’s daily cases that show the rate of growth. The total number of cases never goes down. It’s accumulative, always increases and irrelevant in this context. 

The rate of increase in total cases is the result of the type of growth though. so if the total case numbers go 1,2,4,8,16,32 then it's exponential growth @100% per day. Exponential growth is doubling every defined time period. The current growth rate is doing that, it's just that the time period is long.

Didn't manage to reply to all your points before

> Exponential growth over the whole country would require outbreaks to be uncontrolled. Which they aren’t. 

The rate of exponential growth varies depending on controls. It would appear that an uncontrolled situation (e.g. when a population was first infected) has a doubling time of about 3 days. Currently in the UK it's 39 days.

> What you are seeing is a few individual events leading to a hundred or so new infections each time. Bear in mind these new cases only last for a fortnight or so if they are asymptomatic. 

Experience here was that it was 2 weeks for most cases regardless of whether they're symptomatic or not. 

 DancingOnRock 02 Sep 2020
In reply to Toerag:

July 28th 696

Aug 6 833

Aug 12 945

Aug 20 1050

Aug 21 1155

Aug 28 1190

Thats an increase of less than 100 cases every week. It fits the straight line y=100x+700 very closely where X is the week. It’s linear. It’s not exponential for the simple reason that the growth is due to separate isolated outbreaks. It’s not due to the same breakout controlled by pressures. When there is an outbreak they move to control it and the new infections are reduced. 
 

It will only be exponential if you are not apply pressure. Stop looking at the cumulative cases graph. It’s meaningless in this context. 
 

>Experience here was that it was 2 weeks for most cases regardless of whether they're symptomatic or not. 

 

Not always. Some people were in hospital fighting the infection for a lot longer than two weeks. 

Post edited at 16:21
2
 wintertree 02 Sep 2020
In reply to DancingOnRock:

> It’s linear. It’s not exponential 

It's hard to say scientifically which form of growth we currently have.  Looked at over a time period shorted than it's half life or doubling period, an exponential and a linear curve are very similar (Taylor's theorem and all that).  Without a quantification of the uncertainties on the data it is basically meaningless to argue over which better fits the data.

So, we must wait for longer to have enough data to make a solid proclamation.  But - as we wait for longer, all sorts of things change R, so the separate data periods aren't compatible.

The important - and problematic - thing is that infection numbers are rising.   This is bad.  Very bad.

However...

> It’s linear. It’s not exponential for the simple reason that the growth is due to separate isolated outbreaks. It’s not due to the same breakout controlled by pressures. When there is an outbreak they move to control it and the new infections are reduced. 

This makes no senses.  

If the control was working, there would be no growth.  If there was no control, their would be exponential growth.  

True linear growth of a virus requires a very specific mechanic; few things can naturally produce this - the only natural one I can think of is when it spreads radially from a fixed point and that expanding front increases in perimeter linearly with time. This only translates to cases if people are only infected by the expanding perimeter as it passes, and not through local chains of infection - it's totally incompatible with everything that's know about covid.  An artificial mechanic that could produce linear growth would be control measures that are always improving at an ever decreasing rate nudging R down towards 1 in ever smaller steps , Zeno's paradox style.

1
 Blunderbuss 02 Sep 2020
In reply to tom_in_edinburgh:

> There was no excuse for letting people go on holiday to Italy and letting people walk through airports with no testing or enforced isolation.    They had seen the pictures of what was going on in Wuhan and Italy, they knew it was a sh*t storm and they did absolutely nothing to protect the UK, which, as an island, could have been protected.   There should never have been 1,700 importation events.   If it started to kick off in London anyway they should have had a local lock down really fast, like they are finally doing now.

Did the SNP now have the power to do this in Scotland?....and was Nicola pushing the UK government to do this?

 elsewhere 02 Sep 2020
 DancingOnRock 02 Sep 2020
In reply to wintertree:

It’s because we are only applying hard lockdown in areas that get more than 20 cases per 100,000. So it’s increasing exponentially probably at R very slightly above 1 until it hits that trigger in those areas only. At which point a two week lockdown drops the infections right down. Meanwhile it starts rising somewhere else. This combined effect is causing that radial effect you describe.  
 

 wintertree 02 Sep 2020
In reply to DancingOnRock:

> his combined effect is causing that radial effect you describe.  

I'm far from convinced.  As you suggest, with locally responsive lockdown the situation is very complex.  

I don't think linear vs exponential is worth arguing about as the net result is the same, cases are rising and  I think everything is about to change significantly with the return to schools, the related return of parents to the workplace, the moving of > 1 million young adults around the country into largely self-policed halls of residence and HMOs etc, and the onset of colder, rainier weather and then flu season.

 climbercool 02 Sep 2020
In reply to ChrisLeigh19:

0.3% todays seven day averaged covid death rate in france

27% April 8th seven day averaged covid death rate france

that makes it something like 91 x less deadly.

obviously it is ludicrous to think that the virus has changed by anything like this amount, but given that testing has only increased from around 0.4 per thousand in april, to around 1.82 per thousand at end of august and the best treatment Dexamethasone only reduces deaths by around 20% I feel that if people dont believe the virus is now inherantly less deadly than there should be some ideas far more robust than the ones i have so far seen mentioned on this thread as to why there is a reduction of this scale.   Yes we are protecting nursing homes better now, and its supposed to be spreading more among young people but again I cant see how this can account for a 91x reduction.

Frace is far from unique in this reduced mortality, most countries experiencing a second wave are seeing death rates orders of magnitude lower than their first.

can someone please point me in the direcion of the data which shows how the infected demographic has changed over time so i can have an idea of what sort of impact this could be having on the reduction.

 DancingOnRock 02 Sep 2020
 wintertree 02 Sep 2020
In reply to climbercool:

I think testing prevalence has increased way more than a factor of 4.5 since April; by my estimates testing in the UK was catching 2.6% of new infections at the peak in the UK, and is now catching around 50% of new infections and so has increased by about 20x - that makes your 91x reduction more of a 4.5x reduction which is largely explained by the reduced age of the infected and improved clinical care.

  • So a pretty basic flaw in your assumptions I should say.  I use two independent methods to estimate that 2.6% - one is using estimates of true fatality rate and is a bit self referential, the other is using seroprevalence data and is truly independent of testing data except for poor testing meaning deaths were likely 30% under reported as covid deaths in PHE data.  Both of these are no surprise - the number of tests daily has scaled from something like <10 in Feb to > 100,000 now.  

> can someone please point me in the direction of the data which shows how the infected demographic has changed over time so i can have an idea of what sort of impact this could be having on the reduction.

The mean age of detected infections has decreased a lot in the UK - the data is available from PHE in some random and coarse age bins; UKC user "Victim of Mathematics" has made a much better plot than the one in the weekly PHE Surveillance Report.  I have read similar claims for France in the media but I haven't dived down in to the detail.  If anything, given the improved clinical care and reduced mean age, and the likelihood that contact tracing is now catching > 50%  of infections, my question is "why is the fatality rate so high in France and Spain?"

https://twitter.com/VictimOfMaths/status/1297242516467712000

Post edited at 19:19
 jkarran 02 Sep 2020
In reply to cp123:

> It also does a disservice to the various organisations which pulled together at short notice, including military and NHS planners and logistical staff.  These people, as their day job, either do contingency worst case planning (military) or have specialist knowledge in organising healthcare (NHS).

Sure but you do seem to be forgetting most hospitals were, at peak, before they needed to second staff to the Nightingales, pretty severe staff and equipment shortages. I'm sure there was a plan to staff the Nightingales but it was cut very fine and it wouldn't have looked anything like the professional health service we're used to. The sustainability of the staffing in the absence of sufficient PPE also has to be wondered at.

Jk

Post edited at 19:00
In reply to climbercool:

> obviously it is ludicrous to think that the virus has changed by anything like this amount, but given that testing has only increased from around 0.4 per thousand in april, to around 1.82 per thousand at end of august

You are assuming that both of these are random sampling.

In April the population being tested was those who were presenting as seriously ill.

In August the population being tested is much closer to a random sample of the UK.

Hence you would expect a much higher number of deaths/test in April - simples 😁

Post edited at 19:59
In reply to DancingOnRock:

> After 21 days, people either recover or they die. 

Don't understand your comment, some people are ill for way longer than that and end up recovered or dead.

 climbercool 02 Sep 2020
In reply to wintertree:

> I think testing prevalence has increased way more than a factor of 4.5 since April;

but it hasn't, not in france anyway, even in the uk which had the famously slow start and is now testing more than pretty much anyone in the world it has only increased by a factor of 10 from April 8th to end of august.

https://ourworldindata.org/coronavirus-testing

>by my estimates testing in the UK was catching 2.6% of new infections at the peak in the UK,

If the U.K was catching 4000 ish a day and this was just 2.6% during the peak of april that would mean around 160,000 infections a day, and 4.8 million in April alone, you really think it was that high?

and is now catching around 50% of new infections and so has increased by about 20x - that makes your 91x reduction more of a 4.5x reduction which is largely explained by the reduced age of the infected and improved clinical care.

Even if it is just a 4.5x reduction, I haven't seen any evidence that clinical care is improving survival by anything like this amount. As for the reduced age of infected, this tells you nothing, of course it has decreased,  if in April we were only testing a select few of the sickest patients they will all be elderly people, now we are testing nearly everybody sick or not of course we will start catching the younger people in massifly increased numbers.

Thanks for sharing! cool that a dude of ukc did this.

 climbercool 02 Sep 2020
In reply to Michael Hood:

> You are assuming that both of these are random sampling.

> In April the population being tested was those who were presenting as seriously ill.

> In August the population being tested is much closer to a random sample of the UK.

> Hence you would expect a much higher number of deaths/test in April - simples 😁


yes thats all obvious, but not 91x higher is my point.

 climbercool 02 Sep 2020
In reply to wintertree:

Even if France are currently catching 100% of infections, which obviously they are not, the death rate would still just be 0.3% this is lower than we ever thought  a few months ago.  In terms of infections recorded and deaths reported what sort of figures would you need to see before you started thinking that the virus is not as deadly as it once was? 

 Cobra_Head 02 Sep 2020
In reply to ChrisLeigh19:

It's not just about dying though is it, many people are left with serious health conditions, we're just not hearing much about that at the moment, because everyone is worried about dying, or surviving, not all the stuff in between.

 Cobra_Head 02 Sep 2020
In reply to DancingOnRock:

> After 21 days, people either recover or they die. 

But recovery doesn't mean they're back to where they were before catching it.

 DancingOnRock 02 Sep 2020
In reply to Michael Hood:

The population being tested now is just those showing symptoms or those that have been in contact with those tested positive. Isn’t it? 
 

Also in April, allegedly,  they weren’t even testing a large proportion of people who died. How did that figure in the new cases figure? 

 wintertree 02 Sep 2020
In reply to climbercool:

> but it hasn't, not in france anyway

As Michael Hood said, as well as the increase in testing numbers it has moved from being targeted to hospital admissions (those most likely to die) to effectively random sampling.

> >by my estimates testing in the UK was catching 2.6% of new infections at the peak in the UK,

> If the U.K was catching 4000 ish a day and this was just 2.6% during the peak of april that would mean around 160,000 infections a day, and 4.8 million in April alone, you really think it was that high?

Yes.  My attached image is the seroprevalence data from the latest PHE surveillance report.  The dashed red vertical guide lines were added by me, as were a couple of  dashed extrapolation of curves into the period where they had no data.   Here I summarise the rise in seroprevalance in various regions in terms of start and end % values and the increase in % value,  and I give the population of the region and the number of people who went on to gain detectible antibodies in this period

  • London : 2.5% > 13.0% : 10% increase in 8.9m people > 0.89m people
  • Midlands: 2% > 7% : 5% increase in 10.7m people > 0.53m people
  • North West: 1 % > 7%: 6% increase in 7.3m people > 0.43m people
  • North East & Yorks: 3% > 5%: 2% increase in 8.5m people > 0.17m people
  • For the regions (east of England, south east, south west),  there isn’t any data in the window so I’m taking a cruder stab at the rise in that time.  I pick 2% similar to other regions at the start, and a crude average of 5.5% at the end off the time window, giving a 3.5% rise in a population of 21.0m  = 0.74m people

Sum the numbers - 0.89m + 0.53m + 0.43m + 0.17m + 0.74m = 2.76m people infected in 28 days = 98,500 per day being infected enough to go on and develop antibodies.  That is for England; scaled to the UK that is about 120,000 people per day.  But it seems milder asymptomatic infections don't generate a detection threshold of antibodies so this test misses perhaps the mildest 20% to 40% of infections - with 160,000 per day landing in the bounds of those rates.

This is a super-quick analysis...

> Even if it is just a 4.5x reduction, I haven't seen any evidence that clinical care is improving survival by anything like this amount.

Nor has anybody claimed that it has.

> As for the reduced age of infected, this tells you nothing, of course it has decreased,  if in April we were only testing a select few of the sickest patients they will all be elderly people, now we are testing nearly everybody sick or not of course we will start catching the younger people in massifly increased numbers.

What is the evidence for your "of course"?  To me that is totally unjustified - in March/April there was a critical shortage of PPE and care homes were scrabbling to develop and implement ways of working that protected high dependancy residents from the virus.   In the UK at least we had an insane process of discharging untested "bed blockers" from hospitals back to their care homes to create space in hospitals.  You are right that testing was biassed towards the elderly early on but that is part of the targeted vs random issue that combined with the very high death rate clearly explains a lot of the lowering ratio of measured cases to recorded deaths as Michael Hood noted - you can't play that argument in both directions simultaneously.

> Even if France are currently catching 100% of infections, which obviously they are not, the death rate would still just be 0.3% this is lower than we ever thought  a few months ago.

Is it?  The true fatality rate population wide has estimated by various credible authorities to be in the range 0.1% to 1.0% going right back to March.  

>  In terms of infections recorded and deaths reported what sort of figures would you need to see before you started thinking that the virus is not as deadly as it once was? 

I have seen no genetic or structural evidence that the virus has mutated to be less lethal.  

I see credible evidence that the average age of infection has altered downwards in the last month (the link I provided up-thread) in a period where testing has been widespread and not ramping up (hence not confusing the issue).  Given the massively non-linear link between age of infection and fatality this has a massive downwards effect on fatality rate.  There's suggestion that the reduced viral loads resulting from risk control measures are reducing severity of infection.

I see absolutely nothing but unsubstantiated and wishful navel gazing in looking at these numbers and concluding that the virus has become less lethal.  The virus is having a less lethal effect, and that is as a result of the various risk control measures that have been implemented across society.  But it has the potential to be just as lethal as it was if we are premature in dropping those risk control measures.    This is why threads started by an OP such as this one make me very concerned - things are predominantly good because we are taking risk control measures, so to argue that we should drop risk control measures because things are currently good is a rather ill thought out stance.

To look at it another way, given the rate at which this virus can spread, a lower bound of the fatality rate emerging in France and Spain of around 0.2% is still really bad news if those fatalities predominately land in a condensed time period, and if they represent a source of high viral load infection to health workers.

I do not think that simple detected infections vs deaths figures can possibly prove what you seem to want them to prove.  Much more information is needed - a full breakdown by age and severity of exposure at infection for starters.  Short of incredibly unethical human experimentation we just aren't going to get that data, and so we have to work with what we have, and to take the time to understand the limits of what it is, and is not, just to infer from that.

Post edited at 21:25

 DancingOnRock 02 Sep 2020
In reply to Cobra_Head:

I’m talking about the number of active cases. You’re no longer an active case if you don’t have the virus but you still appear in the ‘total cases’ figure. If you’re not an active case, you can’t transmit the virus (and according to all the present data, you can’t be susceptible) 

We we’re discussing the maths and the growth rather than the medical aspect. 

Post edited at 21:03
 Si dH 02 Sep 2020
In reply to climbercool:

There is a good chart of infection fatality rate by age group at this link if you scroll to the right place:

https://ourworldindata.org/mortality-risk-covid#case-fatality-rate-of-covid-19-by-age

A mean IFR of 0.3% doesn't seem unlikely with the infections skewed very heavily to people under 60, which they are. Many people over 60 are still taking minimal risks.

I think wintertree's analysis of this seems reasonable at the moment.

If you want to look for evidence of reduced mortality would it be better to look at countries where the phases of the pandemic so far have been less discrete than in UK or France? Maybe Sweden, US, India, Brazil?  I'm not familiar enough with their data but if you could find evidence of changing IFRs with time there, I think it might be more convincing.

Post edited at 21:03
 Si dH 02 Sep 2020
In reply to DancingOnRock:

> The population being tested now is just those showing symptoms or those that have been in contact with those tested positive. Isn’t it? 

You only get tested, theoretically, if you have symptoms yourself. Having been in contact with, or indeed living with, a positive case does not mean they advise you be tested. Of course if you are concerned about whether you have caught the virus from a family member you can still get a test by claiming you have a cough; the guidance on what symptoms 'count' is difficult to interpret anyway. 

 wintertree 02 Sep 2020
In reply to DancingOnRock:

> and according to all the present data, you can’t be susceptible

I believe there are two published studies of an individual becoming reinfected, so no, not “all” present data.  It’s early days yet as well.

 DancingOnRock 02 Sep 2020
In reply to wintertree:

My colleague has had it 4 times. I’m sure there will be people who are outliers with odd immune systems but 1 documented case of millions is pretty slim. We know many others have had it on and off for months. But not healthy individuals. 
 

There are also two slightly different mutations. 

Post edited at 21:56
 Si dH 02 Sep 2020
In reply to DancingOnRock:

> My colleague has had it 4 times. I’m sure there will be people who are outliers with odd immune systems but 1 documented case of millions is pretty slim. We know many others have had it on and off for months. But not healthy individuals

I think there is a significant difference between suffering from recurring symptoms after a single infection, and actually getting infected a second time. It seems like your post confused these.

I believe the first is unfortunately common whereas the second is still only very small (known) numbers at present.

 DancingOnRock 02 Sep 2020
In reply to Si dH:

He says he had it 4 separate times with multiple positive and negative tests. He’s been in and out of hospital since last October with a compromised immune system due to cancer treatment and various operations. Apparently, under certain conditions, it can stay with you for several months at undetectable levels though. 
 

I’d be very wary of any reports of people having it twice without knowing their full medical history. 

In reply to wintertree:

> > It’s linear. It’s not exponential 

> It's hard to say scientifically which form of growth we currently have.  Looked at over a time period shorted than it's half life or doubling period, an exponential and a linear curve are very similar (Taylor's theorem and all that).  Without a quantification of the uncertainties on the data it is basically meaningless to argue over which better fits the data.

OK, so currently we're both right and both wrong .  Let's come back to this at the start of October DoR

 DancingOnRock 03 Sep 2020
In reply to Toerag:

Why? That won’t change the last 4 weeks of data. Just stick the values in Excel and do some simple Regression analysis on it and see how close the data fits the line. 
 

From next week the schools and universities go back so you won’t be comparing apples with apples.

“The R number range for the UK is 0.9-1.1 and the growth rate range is -2% to +1% per day as of 28 August 2020. The R number range for the UK is 0.9-1.1 and the growth rate range is -3% to +1% per day as of 21 August 2020.”

https://www.gov.uk/guidance/the-r-number-in-the-uk

In reply to DancingOnRock:

With a compromised immune system it's possible insufficient antibodies are created to give meaningful long term resistance? 

There are also plenty of diseases that require multiple exposure to provide life long protection, from chicken pox, to tick borne encephalitis, or hep b... sometimes one shot just gives you a few weeks protection etc..  

 DancingOnRock 03 Sep 2020
In reply to summo:

It’s possible for everyone’s immune system to be slightly different. With millions of infections and only one case of reinfection I’d suggest that even if it has happened in a ’normal’ person, they’re an outlier. 

In reply to DancingOnRock:

> It’s possible for everyone’s immune system to be slightly different. With millions of infections and only one case of reinfection I’d suggest that even if it has happened in a ’normal’ person, they’re an outlier. 

There have been other cases of apparent reinfection reported, this is one just has very hard evidence so they aren't saying 'could be a false positive on one of the tests.'

If the immunity lasts for several months before fading and the infection rate is now low in countries with good testing there's not going to be many documented cases of reinfection yet.   We won't get an idea of how often it happens unless there is a second wave with a sufficient gap after the first one for immunity to fade.

In reply to baron:

> Lockdowns and travel restrictions only work when either the government has the power and resources to enforce them or the population sees the need for them and cooperates.

I remember seeing pictures of coffins stacked up in Bergamo and the streets of Wuhan completely empty.

There was ample evidence for a competent government to communicate the danger and the need for measures like border controls and local lockdowns.  Many people were arguing for it in the UK and other countries were actually doing it.

The people who were arguing against public health measures were the same demographic who supported Brexit and their enablers in the right wing press.   The Tories did nothing because it was their moronic supporters who were against it.

> Neither of these was going to happen in February and the early part of March.

How come other countries with competent governments did exactly those things? 

 Martin Hore 03 Sep 2020
In reply to ChrisLeigh19:

We're possibly at risk here of counting only deaths as negative impacts of COVID and assuming that all those who test positive but survive suffer no long term consequences.

That was my view back in March. Despite being very close (two months now) to reaching 70, I reckoned that I was of considerably above average fitness for my age and relatively unlikely to succumb if I caught it. My view has now changed completely having learned of the medium and longer term effects on their health that survivors are experiencing, particular those in older age groups. Short of isolating, I'm now taking considerable care to avoid becoming infected.

Martin

 Blunderbuss 03 Sep 2020
In reply to tom_in_edinburgh:

You avoided answering me a few days ago so I will try again....were the SNP pushing the UK government to introduce measures and lock down sooner than they did? Did the SNP have the power to lock down Scotland before England?

BTW - I totally agree that our action in March was too little and too late.....but if the SNP were behind this approach then you can't really lay into the Tories and not the SNP.

 neilh 03 Sep 2020
In reply to tom_in_edinburgh:

And meanwhile the EU was saying to its members, do not impose border controls, it is against Shengen.The countries within the EU that imposed them were the likes of Hungary etc.

Bit of a paradox for you.

 wintertree 03 Sep 2020
In reply to Martin Hore:

> We're possibly at risk here of counting only deaths as negative impacts of COVID and assuming that all those who test positive but survive suffer no long term consequences.

I totally agree.  Even if I bought the OP's argument (which I don't) that the pandemic has "brought forwards" most deaths from the approaching winter flu season, I worry that the pandemic has also made far more other people more susceptible to the approaching flu season.

> Short of isolating, I'm now taking considerable care to avoid becoming infected.

Good for you.  I'm sorry that you face this going in to winter.

Post edited at 12:28
 baron 03 Sep 2020
In reply to tom_in_edinburgh:

> I remember seeing pictures of coffins stacked up in Bergamo and the streets of Wuhan completely empty.

> There was ample evidence for a competent government to communicate the danger and the need for measures like border controls and local lockdowns.  Many people were arguing for it in the UK and other countries were actually doing it.

> The people who were arguing against public health measures were the same demographic who supported Brexit and their enablers in the right wing press.   The Tories did nothing because it was their moronic supporters who were against it.

> How come other countries with competent governments did exactly those things? 

As I’ve posted before, my family and many of my neighbours were all globetrotting well into March.

As another poster said, maybe we weren’t as well informed or as sensible as we thought. But we were as well informed as the vast majority of the population and better informed than some. We shared a ski chalet with 10 doctors from London, none of whom seemed in the least bit concerned despite COVID being a frequent topic of conversation.

No way, at that time, would we have agreed to forgo our holidays nor would we have supported a government imposed lockdown.
With hindsight we were wrong.

 climbercool 03 Sep 2020
In reply to wintertree:

> Sum the numbers - 0.89m + 0.53m + 0.43m + 0.17m + 0.74m = 2.76m people infected in 28 days = 98,500 per day being infected enough to go on and develop antibodies.  That is for England; scaled to the UK that is about 120,000 people per day.  But it seems milder asymptomatic infections don't generate a detection threshold of antibodies so this test misses perhaps the mildest 20% to 40% of infections - with 160,000 per day landing in the bounds of those rates.

yep, this all seems reasonable, im just a little surprised that you realise undetected infections can be this high but last week dismissed the idea that New Jersey which has tripple the death rate of the U.K could be beggining to benefit from herd immunity in surpressing it's second wave (sorry i've digressed)

> What is the evidence for your "of course"?  To me that is totally unjustified

I really dont't understand what you dont agree with here, to me it is so simple, old people are more likely to get really sick, back in April we were only testing really sick people so the  majority of positives were from old people, now we are testing everyone who potentially could be infections including 1000's of asmyptomatic people and as a result the proportion of young people testing positive has increased.

> I have seen no genetic or structural evidence that the virus has mutated to be less lethal.  

What is it you would expect to see?  if it did/has become less lethal, is their some structural tell that would alert scientists to it being less deadly or would you just be watching the outcomes?  ( i have no idea)

I agree this is certainly a long long way from  conclusive evidence but several doctors/researchers are suggesting the mutated D614G strain currently circulating in europe is less deadly.

https://www.newscientist.com/article/2252699-covid-19-is-becoming-less-deadly-in-europe- cubut-we-dont-know-why/

> I see credible evidence that the average age of infection has altered downwards in the last month (the link I provided up-thread) 

yes i agree this is credible evidence of a reduction in age because we are testing in the same way now as we were a month ago, unlike between now and April.

>     This is why threads started by an OP such as this one make me very concerned - things are predominantly good because we are taking risk control measures, so to argue that we should drop risk control measures because things are currently good is a rather ill thought out stance.

I have never mentioned reducing risk control measures and I dont think we should, the evidence is not good enough yet that it is less deadly now and even if it is much less deadly it's still capable of killing thousands more.

> I do not think that simple detected infections vs deaths figures can possibly prove what you seem to want them to prove.  Much more information is needed - a full breakdown by age and severity of exposure at infection for starters.  Short of incredibly unethical human experimentation we just aren't going to get that data, and so we have to work with what we have, and to take the time to understand the limits of what it is, and is not, just to infer from that.

so 2, 3 or 6 months from now what possible evidence could convince you that it has become less deadly?

Anyway, thanks for your replies

 wintertree 03 Sep 2020
In reply to climbercool:

> yep, this all seems reasonable, im just a little surprised that you realise undetected infections can be this high but last week dismissed the idea that New Jersey which has tripple the death rate of the U.K could be beggining to benefit from herd immunity in surpressing it's second wave (sorry i've digressed)

I'm surprised that you're surprised

  1. The under-detection of infection in immunity assays is because of a very weak or non-existent immune response; close the loop on that thinking.
  2. The scale of the infection in NJ is perhaps 20% of the population which is not high enough to make much herd immunity difference to the virus.  
  3. Plenty of other places that are following risk control measures as well as NJ is reported to be doing (including by some of this parish) are also having no problems holding off a "second wave"
  4. The reasons I gave on that thread (in more detail than here) - documented instances of outbreaks from parties and a significant baseline level of infections.

> > What is the evidence for your "of course"?  To me that is totally unjustified

> I really dont't understand what you dont agree with here, to me it is so simple, old people are more likely to get really sick, back in April we were only testing really sick people so the  majority of positives were from old people, now we are testing everyone who potentially could be infections including 1000's of asmyptomatic people and as a result the proportion of young people testing positive has increased.

So you do actually understand why the apparent fatality rate was so high back in April (only testing hospital admissions) despite arguing somehow that it's evidence that the virus has become much less lethal?  You seem to be very confused.

The reason I think your "of course" is not just justified by the changing test patterns is that we have also seen a massive shift towards better protecting the vulnerable - so testing or not testing, there are fewer infections in the elderly because of better risk control measures.

> if it did/has become less lethal, is their some structural tell that would alert scientists to it being less deadly or would you just be watching the outcomes?  ( i have no idea)

You'd expect to see a new mutation in the coding part of the viral DNA that becomes widespread that can be linked to a change in the mechanisms of action of the virus.

It is hard to imagine "outcome" evidence that gives us any clear proof that the virus has become less deadly because (a) we don't know enough about who gets infected, (b) we don't know enough about how strongly they are exposed to the infection and (c) risk control measures are always changing.  Short of deliberately infecting large groups of people to see how many die, we can't get the data and attempts to spitball it from the outcomes data are wishful in the extreme.  It might be possible down the line with some better structured and organised datasets than we seem to have now, but it'll need a phenomenal effort to properly control the analysis.  Perhaps once the ONS pilot survey is well scaled up...

> I agree this is certainly a long long way from  conclusive evidence but several doctors/researchers are suggesting the mutated D614G strain currently circulating in europe is less deadly.

We shall see.  It's early days to proclaim on this.

> yes i agree this is credible evidence of a reduction in age because we are testing in the same way now as we were a month ago, unlike between now and April.

If you look - as another poster has linked and suggested - at the age dependance of death you can see that there's a 1000x difference in fatality rate with enough of a shift in age in adults.  So a small change in mean age can have a massive effect on fatality rates.  Massive.

> so 2, 3 or 6 months from now what possible evidence could convince you that it has become less deadly?

A study of a large enough cohort to have good IFR data on existing strains and ones with an identified key mutation would be the most compelling evidence.  

Failing that, having a country with a significant baseline infection stop all risk control measures and have the infections explode upwards but not the deaths.  I just don't see that though; for example Florida is reaching UK levels of per-capita deaths and most of those have happened in the last month, not early on as in the UK; India is continuing on a near-exponential trajectory of infections and the deaths are tracking this with no sign of the fatality rate decreasing. 

Failing that, some compelling cellular-level or human-transgene adapted rodent studies on the different mutations.  But I don't see that happening any time soon or being that easy to do meaningfully - I may be wrong though!

Failing that, us getting through the winter without serious grief - but that may just be due to the enhanced risk control measures we stick with.

Something that new scientist article doesn’t touch on is viral load - with risk control measures in place the viral load causing infections is down and that likely translates into less severe infections - pushing the fatality rate down despite the virus remaining as lethal.  Then again the article barely discusses the age of the infected; I’ve yet to see a controlled analysis of IFR vs age vs time which is what would be needed to disentangle this - but viral load effects and improved clinical care remain uncontrolled for - then especially with NJ and northern Italy there’s healthcare overload at peak death rate to control for again.  

Post edited at 19:50
 climbercool 05 Sep 2020
In reply to wintertree:

> I'm surprised that you're surprised

> The scale of the infection in NJ is perhaps 20% of the population which is not high enough to make much herd immunity difference to the virus.  

I'm surprised you think it's this high in uk  because 160,000 a day means 7.3% of the uk population infected in April alone, given that we went into lockdown on 16h th march, which would have bought an immediate reduction in infections you must presume that the peak infections were in Mid march and it had been steadily declining for 2 weeks before we even got into April, therefore if you estimate 7.3% for April surely March must have been much higher, A conservative estimate for march would be 50% higher than april, so 11%, meaning 18% of uk infected in March/April alone, given that new jersey has a death rate triple that of the u.k you must either presume that they had a lot lot more infections than us, or that there hospitals handled it much much worse than ours did, I have heard what you think of our government/hospital response so i presume you think N.J must have had much more infectiotns than us ( i do).  Much more infections than 18% is the sort of territory where herd immunity would start having a noticeable impact.

> So you do actually understand why the apparent fatality rate was so high back in April (only testing hospital admissions) despite arguing somehow that it's evidence that the virus has become much less lethal?  You seem to be very confused.

 Yes you know i understand that, upthread i accepted your estimate that there has been a 20 fold increase from 2.5% to 50% of infections that are picked up by testing.   what is interesting is that France has an 80 fold decrease in death rate and I'm not sure the other factors can explain this difference.

> The reason I think your "of course" is not just justified by the changing test patterns is that we have also seen a massive shift towards better protecting the vulnerable - so testing or not testing, there are fewer infections in the elderly because of better risk control measures.

Yes but there are fewer infections and better risk control measures in the younger population too.  My guess is that yes older people are now making up a lower proportion of infections but my point was that even if there had actually been a huge increase in the proportion  of old people being infected c comparing the demographic of positive tests of than and now would still show that old people were getting infected at a lower rate, because we were basically only testing old people in April so that number could only come down.

> If you look - as another poster has linked and suggested - at the age dependance of death you can see that there's a 1000x difference in fatality rate with enough of a shift in age in adults.  So a small change in mean age can have a massive effect on fatality rates.  Massive

yes i agree with this and it might be the answer,  i'm just not sure we have the evidence to prove this change in demographic has happened on a big enough scale.

 wintertree 05 Sep 2020
In reply to climbercool:

> I'm surprised you think it's this high in uk

I think you're over interpreting things here and you've completely lost me in that paragraph. I'm giving my estimates based on the seroprevalance data in earlier versions of [1] which suggests the majority if infections happened over about 5 weeks with a rate of ~150,000 per day in England to a total of about 7% of the population infected.

>  Much more infections than 18% is the sort of territory where herd immunity would start having a noticeable impact.

Not really; given how infections this virus is, a much higher percentage of immune is required for a significant herd immunity effect.  On the other hand, as several posters went in to in detail on the NJ thread, the observance of risk control measures in NJ is very high.  I did on that thread say NJ would be top of my list for places to look for such immunity, but that I think a raft of reasons mean it's likely not in play to any great effect, and I went in to those reasons there.  At 20% it would only make a difference if there are significant risk control procedures in place such that without the immunity one person would on average infect 1.25 people - which may be the case now but it’s not useful here immunity, as as soon as risk control measures are dropped, you get one person infecting > 3 people and the small fraction of immune barely moderate the explosive growth in cases at all.  All of this assume that antibodies imply effective immunity which is far from certain - but others here know way more about immune mechanisms than I.

> yes i agree with this and it might be the answer,  i'm just not sure we have the evidence to prove this change in demographic has happened on a big enough scale.

Have a look at Figure 15 - respiratory outbreaks by institution - in the latest surveillance report [1] - a 15x reduction in outbreaks in care homes since peak, and a significant increase in workplaces.  Given how the death rate for people of care home age is > 100x higher than for  those of mean workplace age.  The majority of infections are now occurring in places where people are < 100x as likely to die.  

I put another thread up related to this [2] - I need to reply to Si dH on it; I'm pretty much in agreement with them.  The data that links through to for the UK is another compelling example that infection is rife in younger people and almost absent in older; and with what's know about death rates vs age I see no mystery here over the low death count in the UK or France - although I maintain that the fatality rate in France is actually quite high still.

There is far more evidence that the virus is barely circulating in an age range > 100x as likely to die, than there is evidence that the virus has become less intrinsically lethal.  

[1] https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/914813/Weekly_COVID19_Surveillance_Report_week_36_FINAL.pdf 

[2] https://www.ukclimbing.com/forums/off_belay/phe_covid_data_by_hotspot-724494

Post edited at 17:26
 elsewhere 06 Sep 2020
In reply to ChrisLeigh19:

https://www.theguardian.com/world/2020/sep/05/covid-19-could-be-endemic-in-deprived-parts-of-england

No sign of herd immunity in Manchester or Glasgow as number of cases bouncing back to about 2/3rds of first wave.

In the first wave infection rate was pretty much the same for all ages, hopefully this time round the elderly and vulnerable will do better. I don't really understand how infection in younger generation does not spread to their parents/grandparents.

Post edited at 17:04
In reply to elsewhere:

> I don't really understand how infection in younger generation does not spread to their parents/grandparents.

Depends how often they see the Grandparents. My Dad lives 8 minutes' swift walk away and is round 3 times a week.  My brother lives 5 miles from Dad and sees him once a week or thereabouts. My grandparents lived overseas in Europe and we only saw them once a year.

 neilh 08 Sep 2020
In reply to elsewhere:

You are assuming that the older generation are not being careful just as a matter of course. 

 wintertree 15 Sep 2020
In reply to ChrisLeigh19:

> For me, comparing the cases vs deaths in France only confirms theories that (very unfortunately) most of the people who would have died due to Covid in western Europe due to underlying conditions, age etc most likely already have and therefore returning to 'normal life' is the best course.

So a few times now and back in Feb/March I've cautioned against drawing conclusions from numbers in the early stages of an exponentially rising pandemic phase where deaths lag infections significantly. 

Here we are, a couple of weeks later on from your post.   Worldometer 7-day moving average for deaths in France for Sep 14 is 12 people.  Their 7-day moving average for detected cases 3 weeks previously on August 23rd was 1039 people.  

12/1039 = 1.15% fatality rate.   

Fatality rates on the order of 1% have been batted about since Jan/Feb.  It's close to the fatality rate excess deaths and seroprevalence give for the UK.

 With the "normal life" you advocate for leading to a doubling of deaths every 5-7 days or so, and with absolutely no sign that the population is now safe due to the cull you posited in your OP, I hope you'll reconsider your stance. 

 climbercool 16 Sep 2020
In reply to wintertree:

> Here we are, a couple of weeks later on from your post.   Worldometer 7-day moving average for deaths in France for Sep 14 is 12 people.  Their 7-day moving average for detected cases 3 weeks previously on August 23rd was 1039 people.  

You have confused the U.Ks data with Frances!  France had 3481 infections  and 32 deaths on the dates that you mention.  Giving a case fatality rate of 0.91%.  But why even bother mentioning this,  what is important is Infection fatality rate not case fatality rate. 

I couldn't find french Seroprevalence data but weeks ago you suggested the u.k was catching around 50% of infections and as cases have almost doubled since than and testing has barely increased surely a 10% drop in recorded infections is very conservative.  That would give you 40% of 0.91 and an Infection fatality rate of 0.36%.

I still think this is probably an over estimate because i think going back 21 days is a little too far, i couldn't find good data on the mean number of days after testing positive until death but many countries have had death peaks just 4-5 days after the peak infections and this article says median days after positive test is 18 days.  https://www.drugs.com/medical-answers/covid-19-symptoms-progress-death-3536264/

> 12/1039 = 1.15% fatality rate.   

> Fatality rates on the order of 1% have been batted about since Jan/Feb.  It's close to the fatality rate excess deaths and seroprevalence give for the UK.

0.36 is not really in the order of 1% and most estimates were greater than 1% back in feb.

 wintertree 16 Sep 2020
In reply to climbercool:

> You have confused the U.Ks data with Frances! 

Sorry, slow day.  I did the calcs for both and gave the wrong one. 

> France had 3481 infections  and 32 deaths on the dates that you mention.  Giving a case fatality rate of 0.91%.  But why even bother mentioning this,  what is important is Infection fatality rate not case fatality rate. 

Because it's the best data we have, and estimates are test and trace is catching ~ 50% of cases as I mention.  

> I still think this is probably an over estimate because i think going back 21 days is a little too far, i couldn't find good data on the mean number of days after testing positive until death but many countries have had death peaks just 4-5 days after the peak infections and this article says median days after positive test is 18 days

I have done an analysis for the UK fatality rate on 14-, 21- and 28- day lags.  This is attached.  It has a similar result for all 3 measures right about now.   

> 0.36 is not really in the order of 1% and most estimates were greater than 1% back in feb.

It's within a factor 3; are you going to claim your "orders of magnitude" are binary and not base 10?  Further, it's perhaps 0.36% when cases are almost confined to the young, which was not the case for the estimates of ~ 1% back in Feb.

Your claim in your OP was that "but even if true that then shows that the mortality rate of Covid is orders of magnitude less than we think".  Orders plural.  Credible estimates back in February were on the order of 1% with a range of perhaps 0.5% to 2% [1]; now with those must susceptible shielding away in the UK and infection rife in the almost unkillable younger people we still have a measured fatality rate of between 0.68% (the mean over my 7-, 14- and 21-day measures) and perhaps 0.34% if we assume 50% of cases are undetected (but this may also mean some unattributed deaths, and we know the fatality figures are a significant under-estimate of excess deaths).

The claim in your OP was for multiple orders of magnitude...

  • So, not 0.4% to 2%, as reported early on and entirely compatible with rates now in the UK and France even before accounting for the highly evidenced (in the UK at least) demographic shift to younger people. 
  • So, not 0.04% to 0.2% - one order of magnitude below early reports, and well below the calculated number for now in the UK or France and below numbers accounting for under-detection of cases given estimates of T&T effectiveness and independent data like the ONS surveys.
  • But your claim would be at least two orders of magnitude or more given your plural "orders", implying  - 0.004% to 0.02% or below.  
    • As claims go this is utter junk it transpires.

[1] Plenty of estimates at the population level in the range of about 0.4% to about 2.0% in here - https://www.cebm.net/covid-19/global-covid-19-case-fatality-rates/

I am curious as to why you push a totally unsubstantiated claim about the fatality rate, along with stating "Surely more damage is being caused not returning to school, work and getting businesses back on track?" when in fact at the time of your posting and since then, the plan clearly was, is, and remains a return to school and work.  Were you confused and not realising that there were widespread plans in place for a full return to school, or did you perhaps want to create the impression that normality was returning much slower than it was (false) because we believed the virus to be much more dangerous than it was (false)?

Post edited at 18:41

 wintertree 16 Sep 2020
In reply to climbercool:

Also, sorry I thought you were the OP when I replied to you; you are not. Sorry, as I said, slow day!  Some of what I wrote was assuming your message was by the OP not by you.  I stick by the points as a rebuttal to the OP however.


This topic has been archived, and won't accept reply postings.