The Lemon Fool

langley59 wrote:Finally...so can we treat it like the flu and now get on with life again please?

If it was like flu we'd already be treating it like flu.

Scott.

johnhemming wrote:

servodude wrote:Thanks I scraped them from the request to load the dropdown

Thanks. Do you easily have a doubling time for your best fits?

I think I should be able to if I can remember how the maths works

if y_1 = a e^(b.x_1) and y_2 = a e^(b.x_2)

we want y_2/y_1 = 2

=> 2 = a e^(b.x_2) / (a e^(b.x_1)) = e^(b ( delta_T))

=> ln(2) = b(delta_T)
=> delta_T = ln(2) / b

or something like that?

I'll see how that looks
-sd

THE JOHN SNOW MEMORANDUM

https://www.johnsnowmemo.com

"MORE THAN 4,400 scientists, researchers & healthcare professionals have now signed the John Snow Memorandum.
We vet every signature, so it may take 48 hours for your name to appear.
Thanks for your support, and please continue to share with your colleagues."

XFool wrote:THE JOHN SNOW MEMORANDUM

https://www.johnsnowmemo.com

"MORE THAN 4,400 scientists, researchers & healthcare professionals have now signed the John Snow Memorandum.
We vet every signature, so it may take 48 hours for your name to appear.
Thanks for your support, and please continue to share with your colleagues."

Ah that John Snow
- the cholera guy
- not the "Winter is coming" guy that "knows nothing"
- nor the news chap

makes sense now
- sd

servodude wrote:

johnhemming wrote:

servodude wrote:Thanks I scraped them from the request to load the dropdown

Thanks. Do you easily have a doubling time for your best fits?

I think I should be able to if I can remember how the maths works

if y_1 = a e^(b.x_1) and y_2 = a e^(b.x_2)

we want y_2/y_1 = 2

=> 2 = a e^(b.x_2) / (a e^(b.x_1)) = e^(b ( delta_T))

=> ln(2) = b(delta_T)
=> delta_T = ln(2) / b

or something like that?

I'll see how that looks
-sd

Try this:

- again the curves are for the sets ending 14 and 7 days back from the end of the data (they don't include the blue bits)
- I should probably smooth the data with a windowed average before I fit to try and remove some of the lumpiness (but I'm not sure how easy I've made that for myself)
- also the code and maths have been subject to both alcohol and caffeine.... so no guarantees


- sd

Thank you for doing this. If we accept for the purpose of discussion that fitting all the regions to an exponential curve is correct (which I think is valid for two regions at the moment anyway) we then have a variation in doubling time to consider

The original doubling time of infection in an entirely susceptible population was 5 days. The various restrictions will have some effect (increasing that time) although most infection is in the home and the schools being open will move infection around between households. Public transport will also move infection around, but I think that is reduced. I do think that the figures demonstrate that SAGE's assumptions on susceptibility are wrong, but a few more weeks data will provide more information.

I expect there to be systemic problems in the positivity testing (cases) that make it hard for the cases to reflect the true situation.

johnhemming wrote:Thank you for doing this. If we accept for the purpose of discussion that fitting all the regions to an exponential curve is correct (which I think is valid for two regions at the moment anyway) we then have a variation in doubling time to consider

The original doubling time of infection in an entirely susceptible population was 5 days. The various restrictions will have some effect (increasing that time) although most infection is in the home and the schools being open will move infection around between households. Public transport will also move infection around, but I think that is reduced. I do think that the figures demonstrate that SAGE's assumptions on susceptibility are wrong, but a few more weeks data will provide more information.

I expect there to be systemic problems in the positivity testing (cases) that make it hard for the cases to reflect the true situation.

Indeed
As it looks presently it's 30 days till you're back at the peak for admissions nationally
- feeling optimistic?

-sd

servodude wrote:As it looks presently it's 30 days till you're back at the peak for admissions nationally
- feeling optimistic?

The first test is to look as to whether there is a difference between regions. I think it is clear that there is.

Then we need to know why there is a difference I would contend

a) Weather/Climate (whether that be consequential behaviour) and
b) Prior infections ie level of herd immunity.

In theory the restrictions should reduce the spreading of the virus and hence the north should not be growing as much as it is because there are harsher restrictions in the north. One possible conclusion from this is that the harsher lockdown in the north has the effect of making the disease spread further. I don't, however, think that is true. Hence I think the main reasons for the difference are a) and b) above.

I don't think the weather is that different. It could cause the seasonal variation to kick off earlier and I think it did, but I don't think it is sufficiently different.

I do think the levels of prior infection (herd immunity levels) are higher than the North first in the Midlands and South West and then in the East, South East and London.

I think that is the primary reason that things are both not as high in the South Easterly regions and also less than the North in the Midlands and South West. Remember I am taking into account also the ratio between the peak admissions earlier this year and the peak admissions now. Those I think reflect the same factors.

The eternal question, of course, is when this is likely to peak. We need to keep an eye on the Midlands/South West as they could start accelerating, but I think the big constraint nationally is the North. NE has peaked at 41% of the peak earlier this year and NW at 57%.

London is picking up a bit at the moment (and that may be weather related), but it has peaked (so far) at 10% of the previous peak.

I don't think London will beat the previous peak for London, but there is a chance that NE and NW do this although I am not sure at the moment.

I think the analysis above is correct, but I don't really have sufficient information to forecast with any precision what is going to happen. We know that a high proportion of cases of infection in hospital are found over a week after admission which implies they are nosocomial and that this is a particular problem in the North West.

Chap on Times Radio this morning - specialism infection control in epidemics - commenting that when hospitals are over-full, infection control procedures within the hospital begin to break down.
So not desirable to run near peak capacity.

TR doesn't seem to have a 'listen again' at present.

V8

servodude wrote:

XFool wrote:THE JOHN SNOW MEMORANDUM

https://www.johnsnowmemo.com

"MORE THAN 4,400 scientists, researchers & healthcare professionals have now signed the John Snow Memorandum.
We vet every signature, so it may take 48 hours for your name to appear.
Thanks for your support, and please continue to share with your colleagues."

Ah that John Snow
- the cholera guy
- not the "Winter is coming" guy that "knows nothing"
- nor the news chap

makes sense now
- sd

And can you IMAGINE the furore of objections that would have happened when when he disabled that water pump, had they had Twitter, FB,TLF et al back then? Demands for evidence, assertions his evidence was flawed, this is no different from 'flu, etc etc...

Mike4 wrote:And can you IMAGINE the furore of objections that would have happened when when he disabled that water pump, had they had Twitter, FB,TLF et al back then? Demands for evidence, assertions his evidence was flawed, this is no different from 'flu, etc etc...

Indeed because he was going against the establishment consensus of the time, so he would be in the GBD camp wouldn't he?

langley59 wrote:

Mike4 wrote:And can you IMAGINE the furore of objections that would have happened when when he disabled that water pump, had they had Twitter, FB,TLF et al back then? Demands for evidence, assertions his evidence was flawed, this is no different from 'flu, etc etc...

Indeed because he was going against the establishment consensus of the time, so he would be in the GBD camp wouldn't he?

Yes, makes exciting telly. Meanwhile, back in the real world people are dying.

Scott.

Mike4 wrote:this is no different from 'flu, etc etc...

Luckily so far it is not as bad as spanish flu. (50,000,000 deaths around 45 as many deaths as Covid ... so far. )

langley59 wrote:

Mike4 wrote:And can you IMAGINE the furore of objections that would have happened when when he disabled that water pump, had they had Twitter, FB,TLF et al back then? Demands for evidence, assertions his evidence was flawed, this is no different from 'flu, etc etc...

Indeed because he was going against the establishment consensus of the time, so he would be in the GBD camp wouldn't he?

What is the "GBD camp"?

Googling it seems to suggest it is a firm of architects.

Mike4 wrote:What is the "GBD camp"?

Great Barrington Declaration

johnhemming wrote:

servodude wrote:As it looks presently it's 30 days till you're back at the peak for admissions nationally
- feeling optimistic?

The first test is to look as to whether there is a difference between regions. I think it is clear that there is.

Then we need to know why there is a difference I would contend

a) Weather/Climate (whether that be consequential behaviour) and
b) Prior infections ie level of herd immunity.

In theory the restrictions should reduce the spreading of the virus and hence the north should not be growing as much as it is because there are harsher restrictions in the north. One possible conclusion from this is that the harsher lockdown in the north has the effect of making the disease spread further. I don't, however, think that is true. Hence I think the main reasons for the difference are a) and b) above.

I don't think the weather is that different. It could cause the seasonal variation to kick off earlier and I think it did, but I don't think it is sufficiently different.

I do think the levels of prior infection (herd immunity levels) are higher than the North first in the Midlands and South West and then in the East, South East and London.

I think that is the primary reason that things are both not as high in the South Easterly regions and also less than the North in the Midlands and South West. Remember I am taking into account also the ratio between the peak admissions earlier this year and the peak admissions now. Those I think reflect the same factors.

The eternal question, of course, is when this is likely to peak. We need to keep an eye on the Midlands/South West as they could start accelerating, but I think the big constraint nationally is the North. NE has peaked at 41% of the peak earlier this year and NW at 57%.

London is picking up a bit at the moment (and that may be weather related), but it has peaked (so far) at 10% of the previous peak.

I don't think London will beat the previous peak for London, but there is a chance that NE and NW do this although I am not sure at the moment.

I think the analysis above is correct, but I don't really have sufficient information to forecast with any precision what is going to happen. We know that a high proportion of cases of infection in hospital are found over a week after admission which implies they are nosocomial and that this is a particular problem in the North West.

Thanks for that.

So reading between the lines you still attest that relaxing restrictions could have had no significant effect? (I remember you posit that "lockdown" didn't affect the peak)

From my mental model of this I think it has every opportunity of accelerating, by which I mean the growth rate will increase
- and unfortunately I don't think there are any levers available to "them in charge" other than reintroducing restrictions
- which I think probably won't work given the lack of appetite, or disbelief around
- I don't think that this can remain "localised" within the UK given there are no restrictions on travel

Going back to my mental model:
I think the "environmental conditions" are again becoming similar to those at the start of all this (totally agree with the seasonality observation)
- and that the difference in growth rate now vs then would just equate to 2/3 fewer opportunities to spread due to behavioural changes (interactions/opportunities for local outbreaks)

-sd

johnhemming wrote:
The first test is to look as to whether there is a difference between regions. I think it is clear that there is.

Then we need to know why there is a difference I would contend

a) Weather/Climate (whether that be consequential behaviour) and
b) Prior infections ie level of herd immunity.

In theory the restrictions should reduce the spreading of the virus and hence the north should not be growing as much as it is because there are harsher restrictions in the north. One possible conclusion from this is that the harsher lockdown in the north has the effect of making the disease spread further. I don't, however, think that is true.

Hence I think the main reasons for the difference are a) and b) above.

I don't think the weather is that different. It could cause the seasonal variation to kick off earlier and I think it did, but I don't think it is sufficiently different.

I do think the levels of prior infection (herd immunity levels) are higher than the North first in the Midlands and South West and then in the East, South East and London.

I think that is the primary reason that things are both not as high in the South Easterly regions and also less than the North in the Midlands and South West. Remember I am taking into account also the ratio between the peak admissions earlier this year and the peak admissions now. Those I think reflect the same factors.

Jonathan Van Tam, when asked at one of his recent televised briefings why there was an apparent disconnect between new infection numbers in Northern and Southern England regions, singularly stated that it was because during the summer months some of the Southern regions were able to get the level of infection down to such low numbers compared to the generally higher Northern regions, that it's simply a case of some Southern regions now starting from a 'lower-base' of any new re-infection curves.

Given that he gave that clear and direct answer to the same issue that you're also considering above, then I'm very surprised to see the lack of any recognition of Van Tam's specific reasoning in your own conclusions, or yours in his...

Why do you think that is?

Cheers,

Itsallaguess

Itsallaguess wrote:Jonathan Van Tam, when asked at one of his recent televised briefings why there was an apparent disconnect between new infection numbers in Northern and Southern England regions, singularly stated that it was because during the summer months some of the Southern regions were able to get the level of infection down to such low numbers compared to the generally higher Northern regions, that it's simply a case of some Southern regions now starting from a 'lower-base' of any new re-infection curves.

Given that he gave that clear and direct answer to the same issue that you're also considering above, then I'm very surprised to see the lack of any recognition of Van Tam's specific reasoning in your own conclusions, or yours in his...

Why do you think that is?

Cheers,

Itsallaguess

Do you have any idea why he thinks that to be the case. Not disputing it is true, or accusing him of being dishonest. How could one region "manage" to do that, but other regions not? Are we thinking the people in the regions acted differently, or that "regional" lockdowns were policed differently.

I can imagine scenarios where population densities, and proximities and interactions are different, but its less easy to see such differences in comparison between London and Manchester, say, than London and the South West.

Are there alternative suggestions?

Absent rational explanations it is difficult not to think that some element of the apparent disparities across regions between waves 1 and 2, might be down to how many of those populations were exposed to the virus in the first wave, and the relative difference between the regions. The argument would be as to how much of this "herd immunity" type explanation explains the disparity. It might be quite small, or alternatively could be a significant contributer.

Itsallaguess wrote:Jonathan Van Tam, when asked at one of his recent televised briefings why there was an apparent disconnect between new infection numbers in Northern and Southern England regions, singularly stated that it was because during the summer months some of the Southern regions were able to get the level of infection down to such low numbers compared to the generally higher Northern regions, that it's simply a case of some Southern regions now starting from a 'lower-base' of any new re-infection curves.

Given that he gave that clear and direct answer to the same issue that you're also considering above, then I'm very surprised to see the lack of any recognition of Van Tam's specific reasoning in your own conclusions, or yours in his...

I rely on the hospital admission figures. Those although not perfect statistically are much more reliable than the prevalence tests that the government tend to rely on.

As you can see from SD's figures London specifically is not growing that fast. Furthermore if you take the average hospital admissions per NHS region for August London was 7, Midlands 10, North East 10 and North West 12.

Hence using this Metric there was not that much of a difference between the Midlands and North East say. London was lower, but is not growing as fast as well.

Arguably the infection reproduction rate in the north should be lower because of the interventions, but in fact is is higher.

servodude wrote:So reading between the lines you still attest that relaxing restrictions could have had no significant effect? (I remember you posit that "lockdown" didn't affect the peak)

I have refined that more recently in that I think the peak in the North East and West was probably restrained by the lockdown, but not London and the South East. Given the number of deaths in London it is quite possible that this restrained the national (England) peak and all the figures point to that.

Were I to be in government I would
a) Do better research on how people get infected particularly those more vulnerable and use that information to
b) Focus a lot more effort on protecting those people who are like to suffer.

I think a lot of the restrictions are making little difference and the only conclusion you can come to if the restrictions are the only thing affecting the spread of the virus is that the areas with more restrictions have more spread.