Coronavirus - Modelling Aspects Only

[servodude]

Hard to tell now they have cancelled the daily press briefings...

Well I suppose anything that reduces the influence in politics in this has to be applauded

The numbers are published but they're rarely in the headlines - and if they are I consider them spin

- sd

[johnhemming]

The numbers are published but they're rarely in the headlines - and if they are I consider them spin

The case numbers are pretty useless statistically, but the hospital admissions are published by NHS regions. They indicate that the disease may be becoming slightly more prevalent in the North West.

[servodude]

The numbers are published but they're rarely in the headlines - and if they are I consider them spin

The case numbers are pretty useless statistically, but the hospital admissions are published by NHS regions. They indicate that the disease may be becoming slightly more prevalent in the North West.

They are what they are
- while they might not be calibrated to give a accurate absolute value that doesn't mean they lack enough precision or context to infer relative trends

- sd

[dealtn]

It means the UK is now seeing four times as many cases on average as it was in mid July.

Do we know what the number of tests per day was in July, and compared to now?

Poor form in replying to myself, but an easy to read article from the BBC that shows how its not as simple as just comparing the data now, with that published previously.

https://www.bbc.co.uk/news/health-54064347

[johnhemming]

- while they might not be calibrated to give a accurate absolute value that doesn't mean they lack enough precision or context to infer relative trends

However, without better analysis they are not sufficient information to infer very much. Given that there is the alternative of reliable information I would rely on that.

[servodude]

- while they might not be calibrated to give a accurate absolute value that doesn't mean they lack enough precision or context to infer relative trends

However, without better analysis they are not sufficient information to infer very much. Given that there is the alternative of reliable information I would rely on that.

If latency doesn't matter to you that's fine
- but the last time I looked viruses spread geometrically
- and this one has a bit of a lag before you get admitted to hospital

-sd

[johnhemming]

If latency doesn't matter to you that's fine
- but the last time I looked viruses spread geometrically
- and this one has a bit of a lag before you get admitted to hospital

I did consider latency. It is arguable that the latency is about the same as people show a positive test result on the PCR for a period of time. In fact there are arguments that the PCR test is positive after people have stopped being infected or being infectious depending upon the sensitivity level.

That is one of the reasons why the PCR test does not produce reliable information.

The latency for admissions is about 2 weeks. (those people who suffer badly tend to do so quickly).

[Itsallaguess]

It means the UK is now seeing four times as many cases on average as it was in mid July.

Do we know what the number of tests per day was in July, and compared to now?

Poor form in replying to myself, but an easy to read article from the BBC that shows how its not as simple as just comparing the data now, with that published previously.

https://www.bbc.co.uk/news/health-54064347

An interesting read, and touching on similar points that have been made here, but again, these points will be known to Van Tam and Whitty, so if they are still now urging a change in approach due to the rising number of UK cases, then they are either truly concerned anyway, or simply acting out of a sense of precaution, and it's not clear at this stage where we are on that spectrum of possibilities...

Cheers,

Itsallaguess

[Itsallaguess]

Chris Whitty has just used this slide on tonight's COVID broadcast to explain that the rising UK case numbers are not directly related to the higher number of tests being carried out -



Source - https://www.theguardian.com/politics/live/2020/sep/09/uk-coronavirus-live-matt-hancock-covid-testing-shortage-boris-johnson

Cheers,

Itsallaguess

[johnhemming]

In the mean time the key figures in many ways are what the proportion of hospital admissions were on Monday for each region as a proportion of the peak admissions. (percentage figures)
England 5.55
East 4.76
London 3.51
Midlands 4.64
North East 7.50
North West 9.85
South East 2.55
South West (admissions zero).

That is consistent with the idea that the level of infection earlier this year was greater in London, Midlands and the South East and since then the North East and North West have been catching up.

That also fits with the comparative curves I posted earlier (where you see London and the Midlands going down much faster than NE/NW).

[johnhemming]

Chris Whitty has just used this slide on tonight's COVID broadcast to explain that the rising UK case numbers are not directly related to the higher number of tests being carried out -

I would suggest that they are directly related to the higher number of tests, but also related to the proportion of tests that are positive.

Additionally, however, if tests are restricted to people with symptoms then we will get a higher proportion of tests being positive (one would expect).

[johnhemming]

This looks at the hospital admissions data from August (when it started going up IMO as a result primarily of the weather and secondarily of further infection).


The charts are adjusted to have the peak value at the same point (1000 on the chart). That makes it easier to consider the "curve". I have not tried to fit any formula to any data.

[Itsallaguess]

I think it'll be interesting to see in the coming months how the younger population responds with regards to the impending tightening of UK protocols.

Given that most of the recent age-group charts show marked increases of infections in the younger age-brackets (see below), then I wonder if we might begin to see 'sub-level herd-immunity' benefits sooner than we might think, if, as I suspect, many of those younger age-groups may now carry on regardless to some degree, and persist in ignoring the guidelines to help prevent cross-infection.

If there are sub-groups where most of the cross-infection occurs, which in this example would most likely be sub-level age-groups no doubt closely related to regular social peer-groups, then I wonder if a level of herd-immunity might continue to build up within those sub-groups over autumn and into winter, to a level where we might not have to then rely on country-wide herd-immunity for large parts of the benefit of such to actually begin to flow through...

What strikes me about the current worst-case predictions is that they seem to assume a blanket coverage across the UK population, and I really do think that huge number of older people are quite willing to still take much stronger precautions than the younger sections of UK society seem willing to, and it doesn't seem clear to me that this is being taken into account when looking at the recent worst-case projections.

Perhaps sub-group, age-related herd-immunity may go on to help provide a quicker benefit to the UK as a whole than is currently being anticipated...



Image Source - https://www.thisislancashire.co.uk/news/18735617.updates-professor-chris-whittys-coronavirus-tv-address/

Cheers,

Itsallaguess

[UncleEbenezer]

This looks at the hospital admissions data from August (when it started going up IMO as a result primarily of the weather and secondarily of further infection).

You mean, as a result of the legal requirement to wear germ-incubators-and-spreaders, that was introduced for shops at the end of July, and as predicted at the time.

[johnhemming]

This looks at the hospital admissions data from August (when it started going up IMO as a result primarily of the weather and secondarily of further infection).

You mean, as a result of the legal requirement to wear germ-incubators-and-spreaders, that was introduced for shops at the end of July, and as predicted at the time.

That is a hypothesis for which I see no evidence that does not at first thought seem likely.

[johnhemming]

Coincidentally we have this story today which I think is more plausible.
https://www.telegraph.co.uk/global-heal ... searchers/

Face masks may be inadvertently giving people Covid-19 immunity and making them get less sick from the virus, academics have suggested in one of the most respected medical journals in the world.

The commentary, published in the New England Journal of Medicine, advances the unproven but promising theory that universal face mask wearing might be helping to reduce the severity of the virus and ensuring that a greater proportion of new infections are asymptomatic.

If this hypothesis is borne out, the academics argue, then universal mask-wearing could become a form of variolation (inoculation) that would generate immunity and “thereby slow the spread of the virus in the United States and elsewhere” as the world awaits a vaccine.

It comes as increasing evidence suggests that the amount of virus someone is exposed to at the start of infection - the “infectious dose” - may determine the severity of their illness. Indeed, a large study published in the Lancet last month found that “viral load at diagnosis” was an “independent predictor of mortality” in hospital patients.

[Itsallaguess]

This looks at the hospital admissions data from August (when it started going up IMO as a result primarily of the weather and secondarily of further infection).

You mean, as a result of the legal requirement to wear germ-incubators-and-spreaders, that was introduced for shops at the end of July, and as predicted at the time.

I know you're clearly not a fan of mask-wearing UncleEb, but there's actually evidence starting to emerge that whilst wearing masks may not fully prevent cross-infections, they may well provide an important method of keeping 'viral load' to a low enough state that it helps to prevent very serious symptoms developing in those that might still catch the virus whilst wearing them. As immune-responses have still been found in asymptomatic and non-serious sufferers, any mask-wearing benefit could then be used to help prevent the most serious symptoms becoming more prevalent..

This article in the New England Journal of Medicine thinks that this is so important, that in the absence of near-term vaccines, wide-spread mask-wearing might even help to provide a form of 'variolation' for this virus, similar to the way we dealt with smallpox initially -

As SARS-CoV-2 continues its global spread, it’s possible that one of the pillars of Covid-19 pandemic control — universal facial masking — might help reduce the severity of disease and ensure that a greater proportion of new infections are asymptomatic. If this hypothesis is borne out, universal masking could become a form of “variolation” that would generate immunity and thereby slow the spread of the virus in the United States and elsewhere, as we await a vaccine.

One important reason for population-wide facial masking became apparent in March, when reports started to circulate describing the high rates of SARS-CoV-2 viral shedding from the noses and mouths of patients who were presymptomatic or asymptomatic — shedding rates equivalent to those among symptomatic patients. Universal facial masking seemed to be a possible way to prevent transmission from asymptomatic infected people. The Centers for Disease Control and Prevention (CDC) therefore recommended on April 3 that the public wear cloth face coverings in areas with high rates of community transmission — a recommendation that has been unevenly followed across the United States.

Past evidence related to other respiratory viruses indicates that facial masking can also protect the wearer from becoming infected, by blocking viral particles from entering the nose and mouth.2 Epidemiologic investigations conducted around the world — especially in Asian countries that became accustomed to population-wide masking during the 2003 SARS pandemic — have suggested that there is a strong relationship between public masking and pandemic control. Recent data from Boston demonstrate that SARS-CoV-2 infections decreased among health care workers after universal masking was implemented in municipal hospitals in late March.

https://www.nejm.org/doi/full/10.1056/NEJMp2026913

Variolation wikipedia page - https://en.wikipedia.org/wiki/Variolation

Cheers,

Itsallaguess

[servodude]

Well it's Friday morning where I am and it's a bit early for the pub so I'll post this here.

Here's a graph and I'll mumble about it below



Background:
A couple of weeks ago there was weeks worth of hospital admissions figures published in the pub which if you looked at them in isolation might have suggested a plateauing (blue spots)

I wasn't convinced of this (mostly because in my mind this system seems to have a time constant that means a week's data is predominantly noise) and tried to put some trends to the data to try and help me see what might be going on.

At the time I took the last 30days admissions data and plotted both the linear (faint red line) and exponential (faint blue line) best fit curves

and mumbled out loud:

seeing how the best fit curve changes for a running 30day period of a series could be interesting

why 30 days?
- I think it might be long enough to give scope for averaging out noise (e.g. variations in reporting during the week) and fit a trend to
- and short enough for that trend to be relevant; given the 1-2 week from infection to hospital admission that might be expected and that restrictions/guidelines/behaviour are frequently changing
- and it's easier than adjusting for a rolling month

so we've had a couple of weeks since (green dots) and I thought I'd have a look at how the trends have changed

I've tried to take it out the "weekly" noise via a 7 day windowed average (orange and yellow squares)
- it's plotted against the middle day of the window and I've used 30 days worth for the plot
- for this averaged set I've plotted linear (solid orange line) and exponential (solid yellow line) best fit curves
- I left a week off the end of this average to see how these curves fitted (or did not fit) the recorded data

Looks like projecting this (10 days ahead) exponentially overshoots a bit and projecting linearly undershoots by a bit more

so what next?
- if I shift the window to end at the last day I can average, the projection of the fitted curves for the next two weeks looks like



I think it takes about 2 weeks for any intervention to show in any figures
- so it will be interesting to see how wrong this is then

time for another coffee
- sd

[johnhemming]

At the moment almost precisely 2/3 of the admissions are in the North and the trend in the north probably fits an exponential curve. (whereas the rest of the country doesn't) I do have a spreadsheet which I update every day with the regional and England admission figures. If you want a copy to look at the regional variations then email me at john.hemming@open-banking.net

There is the curious point that the regions with the most restrictions now also have the most hospital admissions. Whether making people stay at home more results in more infections in the circumstances is true I cannot say. It would be surprising.

It remains that I think the weather is part of this as well as infection. I think that is why it started plateauing for a bit. However, I don't have a good source of average temperatures so have not done any work on this point.

[supremetwo]

At the moment almost precisely 2/3 of the admissions are in the North and the trend in the north probably fits an exponential curve. (whereas the rest of the country doesn't) I do have a spreadsheet which I update every day with the regional and England admission figures. If you want a copy to look at the regional variations then email me at john.hemming@open-banking.net

There is the curious point that the regions with the most restrictions now also have the most hospital admissions. Whether making people stay at home more results in more infections in the circumstances is true I cannot say. It would be surprising.

It remains that I think the weather is part of this as well as infection. I think that is why it started plateauing for a bit. However, I don't have a good source of average temperatures so have not done any work on this point.

Any stats regarding infections from a recent holiday abroad, especially students?

Does track and trace ask where you have been in the last three weeks?