Re: Coronavirus - Numbers and Statistics

[johnhemming]

Off topic, but potentially of interest to sd & johnhemming (exponential/gompertz)
https://www.quantamagazine.org/the-sum- ... r-20190206

Thanks for this. I don't quite see the link to Gompertz. It strikes me as the sorts of maths like aspects of Number Theory that were initially not thought to have a practical use, but then underpinned asymmetric key cryptography. (about 200 years later maybe more)

[johnhemming]

Hospital admissions 17/1 3424, 3569, 3295, 3678, 3840, 4134 all regions reduced over last day.

[servodude]

Off topic, but potentially of interest to sd & johnhemming (exponential/gompertz)
https://www.quantamagazine.org/the-sum- ... r-20190206

Thanks for this. I don't quite see the link to Gompertz. It strikes me as the sorts of maths like aspects of Number Theory that were initially not thought to have a practical use, but then underpinned asymmetric key cryptography. (about 200 years later maybe more)

I do like the writing on quanta (and I often don't find myself there) so thanks for posting the link

One relevance I can see to the issue we've been discussing is how hard it would be to infer which space you would be in if you were near the origins
- but I hope we knocked that one on the head back in August

Its also demonstrates quite nicely why I prefer to work in (suitably ranged) fixed point maths for critical systems

- sd

[daveh]

I would take the data on this site with a very large pinch of salt

https://coronavirus.data.gov.uk/

When searching for the local data for Aberdeenshire it tells me there have been over 300,000 cases in the last 7 days (Population of Aberdeenshire is 260,000) and that the 7day cases per 100,000 at over 400 (figure on the public health Scotland site https://public.tableau.com/profile/phs. ... 0/Overview is 140 cases per 100,000). It was also reporting 4.2million vaccinations for Scotland, which is the whole UK figure.

I reported it last night but so far they haven't corrected the data.

[swill453]

When searching for the local data for Aberdeenshire it tells me there have been over 300,000 cases in the last 7 days (Population of Aberdeenshire is 260,000) and that the 7day cases per 100,000 at over 400 (figure on the public health Scotland site https://public.tableau.com/profile/phs. ... 0/Overview is 140 cases per 100,000).

How are you finding the 300,000? If I click on Aberdeenshire on the interactive map I get cases - 390 and rate per 100,000 149.3.

EDIT: I get the same values if I put in an Aberdeenshire postcode.

Scott.

[daveh]

When searching for the local data for Aberdeenshire it tells me there have been over 300,000 cases in the last 7 days (Population of Aberdeenshire is 260,000) and that the 7day cases per 100,000 at over 400 (figure on the public health Scotland site https://public.tableau.com/profile/phs. ... 0/Overview is 140 cases per 100,000).

How are you finding the 300,000? If I click on Aberdeenshire on the interactive map I get cases - 390 and rate per 100,000 149.3.

Scott.

That's what I was expecting to see, but I went to https://coronavirus.data.gov.uk/ (the Daily Update) and put in my postcode for Aberdeenshire and got the figures I quoted. I was about to post a picture, but it would have included my postcode which is just a few houses so would have linked quite closely to me. So I put in another Aberdeenshire postcode and got the figures you quoted. That is really weird - different values for two different postcodes. Now I have no idea what is happening.

[daveh]

This is what I see when I put in my postcode


and I've tried a couple of other aberdeenshire postcodes and its just mine that's pulling up the data in this image. Very odd.

[johnhemming]

Hospital admissions 18/1 3711, 3424, 3569, 3295, 3678, 3840, 4134

[daveh]

I would take the data on this site with a very large pinch of salt

https://coronavirus.data.gov.uk/

When searching for the local data for Aberdeenshire it tells me there have been over 300,000 cases in the last 7 days (Population of Aberdeenshire is 260,000) and that the 7day cases per 100,000 at over 400 (figure on the public health Scotland site https://public.tableau.com/profile/phs. ... 0/Overview is 140 cases per 100,000). It was also reporting 4.2million vaccinations for Scotland, which is the whole UK figure.

I reported it last night but so far they haven't corrected the data.

Looks like its fixed now - using my post code is giving sensible data again - not sure why I got the odd data for 24-48hrs with my Aberdeenshire postcode and not with other Aberdeenshire post codes I tried when Scott said he wasn't getting the same data as me for Aberdeenshire. I like this site for Scottish Covid data:
https://public.tableau.com/profile/phs. ... 0/Overview and it seems to have been upgraded today and is giving additional info in the drop down boxes.

[johnhemming]

Hospital admissions 20/1 3679, 3768, 3711, 3424, 3569, 3295, 3678, 3840, 4134

[jfgw]

The new variant is declining in England (but not yet in Scotland, Northern Ireland or Wales) according to the ONS survey,

https://www.ons.gov.uk/peoplepopulation ... uk-variant


Julian F. G. W.

[jfgw]

New Admissions falling or levelling off,


My graph. Contains public sector information licensed under the Open Government Licence v3.0. : https://www.nationalarchives.gov.uk/doc/open-government-licence/version/3/


Julian F. G. W.

[jfgw]

Hospital bed occupancy is going down,

My graph. Contains public sector information licensed under the Open Government Licence v3.0. : https://www.nationalarchives.gov.uk/doc/open-government-licence/version/3/

MV bed occupancy is levelling off,

My graph. Contains public sector information licensed under the Open Government Licence v3.0. : https://www.nationalarchives.gov.uk/doc/open-government-licence/version/3/

The percentage of Covid-19 patients in Mechanical Ventilator beds is going up. This would be expected given the pattern of admissions and the delay from admission to ventilation.

My graph. Contains public sector information licensed under the Open Government Licence v3.0. : https://www.nationalarchives.gov.uk/doc/open-government-licence/version/3/



Julian F. G. W.

[dealtn]

And the delay in coming off ventilators, ie. from ventilation to discharge (or death, sadly).

[johnhemming]

Hospital admissions 22/1 3016, 3237. 3679. 3768. 3711. 3424, 3569, 3295, 3768, 3840, 4134 South West still quite high (for the south west), first time back below peak of first wave (which was 3099)

[johnhemming]

Hospital admissions 23/1 2780, 3016, 3237
All regions are down. Whether we can see an early effect of vaccinating people over 80 or not I cannot say. The figures do tend to vary quite a bit during the week and may go back up by 10-20% within the normal pattern of movement.

[johnhemming]

I have looked at the age analysis of hospital admissions and the proportion of each age category for each day since October to 15th Jan.

Max	Min
Total reported admissions and diagnoses  0-5	2%	0%
Total reported admissions and diagnoses  6-17	2%	0%
Total reported admissions and diagnoses  18-54	24%	16%
Total reported admissions and diagnoses  55-64	17%	10%
Total reported admissions and diagnoses  65-74	21%	15%
Total reported admissions and diagnoses  75-84	29%	22%
Total reported admissions and diagnoses  85+	27%	17%

Those in the 85+ category are currently running at around 20%. I would expect to see any impact of vaccination first in that category.

20%	19%	20%	20%	20%	22%	20%	21%	19%	20%	20%

Looking further at hospital admissions the NHS do separate out admissions from the community from diagnoses (which including hospital acquired infections). I will continue to post the daily gross figures, but we should be aware that the admissions from the community are much lower.

[johnhemming]

Hospital admissions 25/1 2648, 2824, 2780, 3016

[johnhemming]

Hospital admissions 26/1 2799, 2648. 2824, 2780
I have got the trust figures to 24th, but the age analysis is not yet available (it is weekly some time). Up to 15th Jan there is no evidence of a change in age profile (from vaccination). Hence this all arises from prior infection limits. I will do the trust analysis and Gompertz fit for trusts some time in the next 24 hours).

[servodude]

This was an alarming read

https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(21)00183-5/fulltext

After initially containing severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), many European and Asian countries had a resurgence of COVID-19 consistent with a large proportion of the population remaining susceptible to the virus after the first epidemic wave. By contrast, in Manaus, Brazil, a study of blood donors indicated that 76% (95% CI 67–98) of the population had been infected with SARS-CoV-2 by October, 2020. High attack rates of SARS-CoV-2 were also estimated in population-based samples from other locations in the Amazon Basin—eg, Iquitos, Peru 70% (67–73). The estimated SARS-CoV-2 attack rate in Manaus would be above the theoretical herd immunity threshold (67%), given a basic case reproduction number (R0) of 3.
In this context, the abrupt increase in the number of COVID-19 hospital admissions in Manaus during January, 2021 (3431 in Jan 1–19, 2021, vs 552 in Dec 1–19, 2020) is unexpected and of concern