Corona Virus main thread

[Van wink 2,994]

2 hours ago, Mark .Y. said:

You can ask the responders to do anything, Badger.

I'm one and we were given links to tell us how to deal with tasks such as shopping for other people ie how you take their card/cash, how you leave their shopping at the door etc etc

Google it up and give it a try  🙂

Just seen your last post about getting priority access - glad you're sorted

 

Good on you Mark, we have a great local response team here and a really good facebook group for support. Hopefully we will come out of this with a much stronger sense of community .👍

[Van wink 2,994]

15 minutes ago, Bill said:

because many of the ,ost vunerable have died to to allowing it to spread before lockdown

What are you talking about?

[TIL 1010 4,773]

1 minute ago, Van wink said:

What are you talking about?

Dear God winky what have you now gone and done ?

[ricardo 7,395]

51 minutes ago, Yellow Fever said:

 

Exactly Bill - far too much accepting of bald numbers without doing any 'basic reality' check. Engineering course 101.

Easy to criticise others i.e China or USA and their numbers but our own don't really stack up either or compare as with say Germany.

So - let's not blame anybody or claim we are any better/worse but get real hard facts. Meanwhile err on the side of caution as per the government extended lock down until we know better.

If any European nation is fiddling the figures then they are all fiddling much the same.

Spain and Belgium looking worse than average, Germany and Austria better than average.

not much more than a fag paper between the rest.

Deaths: Western Europe: Italy, Spain, France, Netherlands, Switzerland, UK, Austria, Germany, Belgium

Deaths: Western Europe: Italy, Spain, France, Netherlands, Switzerland, UK, Austria, Germany, Belgium

 

[Yellow Fever 3,810]

1 minute ago, ricardo said:

If any European nation is fiddling the figures then they are all fiddling much the same.

Spain and Belgium looking worse than average, Germany and Austria better than average.

not much more than a fag paper between the rest.

Deaths: Western Europe: Italy, Spain, France, Netherlands, Switzerland, UK, Austria, Germany, Belgium

Deaths: Western Europe: Italy, Spain, France, Netherlands, Switzerland, UK, Austria, Germany, Belgium

 

 

The question was 'hospital admissions' data which I am suspicious  about as it open to all sort of subjective judgements and indeed forms of negative feedback and doesn't stack up with 12-15% infected (was it 1 in 7 in Germany as well  - glanced at earlier - could be mistaken) as quoted earlier. Perhaps if we did hospitalize 1 in 7 of the 4% infected (if we could) we might be on the German curve ?

I have stated earlier that the 'deaths' I tend to treat as more definitive and a better if lag indicator.

[Bill 1,788]

4 minutes ago, Yellow Fever said:

 

The question was 'hospital admissions' data which I am suspicious  about as it open to all sort of subjective judgements and indeed forms of negative feedback and doesn't stack up with 12-15% infected (was it 1 in 7 in Germany as well  - glanced at earlier - could be mistaken) as quoted earlier. Perhaps if we did hospitalize 1 in 7 of the 4% infected (if we could) we might be on the German curve ?

I have stated earlier that the 'deaths' I tend to treat as more definitive and a better if lag indicator.

it is all pretty academic (meaningless) as long as the lockdown continues as there is no real measure of how well that is being observed or what is the vunerability levels of those still working with the public

 

[ricardo 7,395]

1 minute ago, Yellow Fever said:

 

The question was 'hospital admissions' data which I am suspicious  about as it open to all sort of subjective judgements and indeed forms of negative feedback and doesn't stack up with 12-15% infected (was it 1 in 7 in Germany as well  - glanced at earlier - could be mistaken) as quoted earlier. Perhaps if we did hospitalize 1 in 7 of the 4% infected (if we could) we might be on the German curve ?

I have stated earlier that the 'deaths' I tend to treat as more definitive and a better if lag indicator.

Where do you get this 4% infected from? We have no idea how many are infected because large numbers show no symptoms and probably never know if they have had it. Also those infected will have their infections over a long period of time and not all at once.

The only certainty we have is that those diagnosed as infected will follow two pathways, roughly 80-85 % have a very mild condition and will never need hospitalisation, roughly 15-20% need some form of hospitalisation  and of those some 3-5% will need intensive care. These figures seem to be consistant across nations and there is more than enough data to confirm this.

The variations we see in deaths will have many causes, variable health care, social conditions and living in close proximity to large conurbations are only a few of many. Inevitably some countries will fare worse than others.

[Bill 1,788]

1 minute ago, ricardo said:

The variations we see in deaths will have many causes, variable health care, social conditions and living in close proximity to large conurbations are only a few of many. Inevitably some countries will fare worse than others.

Which is why I question the reason behind these alsmost endless posts - a few days ago the 'end was in sight' as it appeared the daily death toll was down

the real figures that are of any importance/relevance are the risk over return of introducing a 'vaccine' that may at this stage not be fully tested

similarly what work is going to become 'important' and who decides - therefore who will return to work

in the US sort rightwing nut (there are others?) has claimed that schools need to re-open though they will probably face a 3-4% mortality rate

it is the decisions that we take and have to take rather than unrelated numbers that is important

[Yellow Fever 3,810]

5 minutes ago, ricardo said:

Where do you get this 4% infected from? We have no idea how many are infected because large numbers show no symptoms and probably never know if they have had it. Also those infected will have their infections over a long period of time and not all at once.

The only certainty we have is that those diagnosed as infected will follow two pathways, roughly 80-85 % have a very mild condition and will never need hospitalisation, roughly 15-20% need some form of hospitalisation  and of those some 3-5% will need intensive care. These figures seem to be consistant across nations and there is more than enough data to confirm this.

The variations we see in deaths will have many causes, variable health care, social conditions and living in close proximity to large conurbations are only a few of many. Inevitably some countries will fare worse than others.

The 4% came from the Dutch studies as reported on here (was it Mark Y) and from memory correlated with other data elsewhere. I actually halved it to 2% for the sake of argument and assumed population of 50M - makes my numbers easy (1M 'infected' to date) for 10% (not the 12-15 quoted) then needing hospital give 100K hospital patients.

Clearly 100K patient are not in hospital (but could be at the stroke of the pen if we of took in all the care home cases !).

As to the timing - given that it's a fast exponential growth the vast vast majority of such patients should be recent - last few weeks - not in February or even March so the approximation is valid.

So something doesn't stack up with the numbers or more likely interpretation given. 

[Badger 2,407]

15 minutes ago, ricardo said:

Where do you get this 4% infected from?

My guess is that it is from perfect mathematical modelling of inadequate data! 

We will need to time to get the data right, unfortunately, but in the meantime, what we have is still of some use to inform practice. In the long term the data will be better but my guess it will have a wide enough range of probabilities to fuel much political debate.

[ricardo 7,395]

2 minutes ago, Bill said:

Which is why I question the reason behind these alsmost endless posts - a few days ago the 'end was in sight' as it appeared the daily death toll was down

the real figures that are of any importance/relevance are the risk over return of introducing a 'vaccine' that may at this stage not be fully tested

similarly what work is going to become 'important' and who decides - therefore who will return to work

in the US sort rightwing nut (there are others?) has claimed that schools need to re-open though they will probably face a 3-4% mortality rate

it is the decisions that we take and have to take rather than unrelated numbers that is important

There is no end in sight until there is a vaccine.

If the lockdown holds the number infected will decline as will deaths

If we go back to some sort of normal the number infected will rise as will deaths.

It shouldn't be a difficult concept to understand that the object is to control the infection rate at a level where it doesn't overwhelm us.

[Yellow Fever 3,810]

1 minute ago, Badger said:

My guess is that it is from perfect mathematical modelling of inadequate data! 

We will need to time to get the data right, unfortunately, but in the meantime, what we have is still of some use to inform practice. In the long term the data will be better but my guess it will have a wide enough range of probabilities to fuel much political debate.

Please tell me if I'm wrong as I'm working at present but I think a similar number has come from UK studies and our own scientists. I don't think a number of few percent is in much doubt. 

[Bill 1,788]

1 minute ago, ricardo said:

There is no end in sight until there is a vaccine.

If the lockdown holds the number infected will decline as will deaths

If we go back to some sort of normal the number infected will rise as will deaths.

It shouldn't be a difficult concept to understand that the object is to control the infection rate at a level where it doesn't overwhelm us.

err, that is what I have been putting forward for weeks

we will have to back to some sort of 'normal' to keep the economy ticking over and tp process a few million more

the point I have made umpteen times is how best to reduce the pressure without blowing the cap off the radiator

trial and error

[Mark .Y. 352]

9 minutes ago, ricardo said:

There is no end in sight until there is a vaccine.

If the lockdown holds the number infected will decline as will deaths

If we go back to some sort of normal the number infected will rise as will deaths.

It shouldn't be a difficult concept to understand that the object is to control the infection rate at a level where it doesn't overwhelm us.

Yes, I don't think there is anybody on here who hasn't know that from the start Ricardo.

I personally like to look at the daily net number of hospital admissions because I know that if it continues it's current downward trend that :

1. The number of daily reported deaths will follow it down

2. It means we are freeing up the capacity to permit the first stage of easing of the restrictions

[ricardo 7,395]

7 minutes ago, Bill said:

err, that is what I have been putting forward for weeks

we will have to back to some sort of 'normal' to keep the economy ticking over and tp process a few million more

the point I have made umpteen times is how best to reduce the pressure without blowing the cap off the radiator

trial and error

As light a touch on the accelerator as possible would be my guess.

 

[Mark .Y. 352]

Thought this was quite interesting too, the data from the French aircraft carrier Charles De Gaulle :

Aircraft carriers are turning out to provide some of our best data. This is from France: 940 positive tests, half of the infected are asymptomatic, 20 hospitalized (down from 31 a couple of days ago), 1 in intensive care.

[ricardo 7,395]

1 minute ago, Mark .Y. said:

Yes, I don't think there is anybody on here who hasn't know that from the start Ricardo.

I personally like to look at the daily net number of hospital admissions because I know that if it continues it's current downward trend that :

1. The number of daily reported deaths will follow it down

2. It means we are freeing up the capacity to permit the first stage of easing of the restrictions

No doubt those that have to make that decision have more data than we do.

[Yellow Fever 3,810]

4 minutes ago, Mark .Y. said:

Thought this was quite interesting too, the data from the French aircraft carrier Charles De Gaulle :

Aircraft carriers are turning out to provide some of our best data. This is from France: 940 positive tests, half of the infected are asymptomatic, 20 hospitalized (down from 31 a couple of days ago), 1 in intensive care.

Interesting but hardly representative of any population by age/gender/fitness. The 50% asymptomatic is interesting though.

[ricardo 7,395]

4 minutes ago, Mark .Y. said:

Thought this was quite interesting too, the data from the French aircraft carrier Charles De Gaulle :

Aircraft carriers are turning out to provide some of our best data. This is from France: 940 positive tests, half of the infected are asymptomatic, 20 hospitalized (down from 31 a couple of days ago), 1 in intensive care.

Yes, there is similar good data from the cruise ship Diamond Princess.

[Yellow Fever 3,810]

847 deaths today (hospital). Hmm.

[Mark .Y. 352]

26 minutes ago, Yellow Fever said:

847 deaths today (hospital). Hmm.

847 deaths in hospital recorded in the last 24 hours recording period YF, the actual number of deaths in hospital in that period will almost certainly turn out to be quite a few less

 

Edited April 17, 2020 by Mark .Y.
corrected spelling

[Yellow Fever 3,810]

Just now, Mark .Y. said:

847 deaths in hospital recorded in the last 24 hours reocrding period YF, the actual number of deaths in hospital in that period will almost certainly turn out to be quite a few less

 

Agreed - and previous days have been corrected up. Same will happen to this data - it will be corrected with hindsight.

[keelansgrandad 6,679]

To quote Churchill

This isn't the beginning of the end

It is the end of the beginning

[Mark .Y. 352]

7 minutes ago, Yellow Fever said:

Agreed - and previous days have been corrected up. Same will happen to this data - it will be corrected with hindsight.

Yes, from the NHS webite, the highest actual daily number of deaths was on the 8th April with a total of 792.

The reported figure for that day was 938, so even allowing for a few more "late recordings" the actual deaths, although a tragic figure, will get nowhere near the quoted "headline" figure.

[ricardo 7,395]

16 minutes ago, Mark .Y. said:

847 deaths in hospital recorded in the last 24 hours reocrding period YF, the actual number of deaths in hospital in that period will almost certainly turn out to be quite a few less

 

Yes 22 of them date from as far back as April 7th

[Mark .Y. 352]

1 minute ago, ricardo said:

Yes 22 of them date from as far back as April 7th

Yes Ricardo, the NHS spreadsheet cautions that the last 5 days of data is "likely to change" but there are the odd few from a bit further back and that has been happening throughout.

[Yellow Fever 3,810]

1 minute ago, Mark .Y. said:

Yes Ricardo, the NHS spreadsheet cautions that the last 5 days of data is "likely to change" but there are the odd few from a bit further back and that has been happening throughout.

Which is why we all agree not to take a single days figures  - or even a few days as gospel of any particular trend - especially with weekends and banks holidays.

[T 190]

I think the Dutch blood tests actually showed 3pc. The UK estimated 4pc The UK was doing general testing at the start so would have an idea of the overall infection rate compared to those being hospitalised.
 

No doubt scientists are also sharing unpublished data internationally eg Germany did a random sample laboratory blood testing of 3000 people in Munich recently for antibodies which hasn’t been reported but would be known in the scientific community.The sample size would have normally been selected to give you c95pc certainty of the  infection rate as this is the most efficient point between sample size and certainty Munich was presumably chosen as it  has been suggested it was the entry point for the disease into Europe from China. 

[Yellow Fever 3,810]

5 minutes ago, T said:

I think the Dutch blood tests actually showed 3pc. The UK estimated 4pc The UK was doing general testing at the start so would have an idea of the overall infection rate compared to those being hospitalised.
 

No doubt scientists are also sharing unpublished data internationally eg Germany did a random sample laboratory blood testing of 3000 people in Munich recently for antibodies which hasn’t been reported but would be known in the scientific community.The sample size would have normally been selected to give you c95pc certainty of the  infection rate as this is the most efficient point between sample size and certainty Munich was presumably chosen as it  has been suggested it was the entry point for the disease into Europe from China. 

Thanks T - seemed to get in a pointless argument there for a while over some basic numbers!

[Badger 2,407]

49 minutes ago, Yellow Fever said:

Please tell me if I'm wrong as I'm working at present but I think a similar number has come from UK studies and our own scientists. I don't think a number of few percent is in much doubt. 

TBH, YF, I don't have much faith in the data yet. What we have is useful to plan what we do but we won't have a full picture on the extent of the virus yet, nor reliable comparative data. My most confident guess, is that it has been under reported everywhere.

At a low point of my career I worked in local govt (relatively short  lived) and one of my duties was to sit on a data committee that use to try to model data. It is much harder than you would think because the data that you want largely isn't available. The data that is gathered is usually for a specific purpose which does not fit very neatly when you try to re-purpose it + there are varying practices of recording it which call into question the reliability of the data.

e.g. Doctor A may be happy to put covid on a death certificate without a blood test whilst Doctor B isn't (I made that up, as a possible example, don't know if it's true).

You then have to go through a statistical process to adjust different sources of data in different ways. Ultimately, I think that the most reliable way may come from comparing historical death rates over time, but I'm not sure what sources are available (apart from the obvious ones).

Sorry - a very, very long-winded way of saying I don't know! 😳