Re: Coronavirus - Numbers and Statistics

[jfgw]

The Milton Keynes lab is finding about 85% of positives are compatible with the Kent variant.

https://twitter.com/The_Soup_Dragon

Julian F. G. W.

[jfgw]

London, East of England, and South East admissions still falling, South West seems to have peaked,

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/

Fewer people in hospital,

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/

Mechanical ventilator bed occupation is starting to go 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/


Julian F. G. W.

[jfgw]

England, Scotland and Northern Ireland admissions are following similarly-shaped curves. The curve for Wales is a lot flatter.

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/

Data are not updated every day by all four nations and the figures are not comparable as Wales include suspected COVID-19 patients while the other nations include only confirmed cases.

https://coronavirus.data.gov.uk/details/healthcare?areaType=overview&areaName=United%20Kingdom

Scotland's figures are subject to change; data going back more than two months had changed.


Julian F. G. W.

[jfgw]

New cases (people testing positive for the first time),


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


Julian F. G. W.

[jfgw]

New cases possibly 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/

Hospital admissions still falling

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/

Hospital bed occupancy falling

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/

Mechanical ventilator beds, slight fall

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.

[Sorcery]

Sorry if this has been asked before, is the "new cases" adjusted for count of tests, an implied percentage of population would be the most useful.
If so it's a little strange that the new cases graph is asymptoting, other graphs look more promising.

[jfgw]

"New Cases" is everyone who has tested positive for the first time. It is not adjusted for the number tests.
Someone infected (and testing positive) for a second time is only counted the first time - a potential issue if mutations (such as the SA variant) result in reinfections.

The ONS survey gives estimates of numbers infected,
https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/conditionsanddiseases/bulletins/coronaviruscovid19infectionsurveypilot/5february2021 .


Julian F. G. W.

[Sorcery]

"New Cases" is everyone who has tested positive for the first time. It is not adjusted for the number tests.
Someone infected (and testing positive) for a second time is only counted the first time - a potential issue if mutations (such as the SA variant) result in reinfections.

The ONS survey gives estimates of numbers infected,
https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/conditionsanddiseases/bulletins/coronaviruscovid19infectionsurveypilot/5february2021 .


Julian F. G. W.

Thanks for that, it means the cases graph could be subject to wobbles say if testing effort was diverted to vaccinations. So perhaps treat it as indicative while taking the latter 3 graphs as more important.

[jfgw]

One thing to I didn't mention about the ONS survey is that it only covers infections in the community, not, for example, care homes or hospitals.

Thanks for that, it means the cases graph could be subject to wobbles say if testing effort was diverted to vaccinations. So perhaps treat it as indicative while taking the latter 3 graphs as more important.

It is certainly not perfect, which is why admissions data have been touted on these fora as giving a more accurate indication of infection trends. Testing has increased over time, https://coronavirus.data.gov.uk/details/testing?areaType=nation&areaName=England .

I would consider it likely that vaccinations and, possibly, mutations would alter the fraction of those infected who require hospitalisation. Seasonality could also be a factor.


Julian F. G. W.

[jfgw]

These are absolute values (not per capita) so the heights of the different plots are not comparable. It is the shapes that are relevant.

There is little evidence of the vaccines working in the 75 to 84, and 85+ age groups as yet. Admissions are possibly falling slightly faster than other age groups but not enough to draw firm conclusions. Vaccines take time to work, and there is a delay between infection and diagnosis. Maybe I need to wait a little longer.


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]

I have calculated admissions for each age group per 100 000 based upon 2019 ONS population statistics,


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]

Deaths by age group. These are absolute numbers (not per capita) for England. I have just plotted 60+ age groups to keep the graph clean. I would consider deductions drawn from this graph regarding the effects of vaccinations to be somewhat speculative. Data separated into vaccinated/not vaccinated would, I expect, provide evidence of vaccine efficacy.


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.

[funduffer]

Deaths by age group. These are absolute numbers (not per capita) for England. I have just plotted 60+ age groups to keep the graph clean. I would consider deductions drawn from this graph regarding the effects of vaccinations to be somewhat speculative. Data separated into vaccinated/not vaccinated would, I expect, provide evidence of vaccine efficacy.


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.

Perhaps showing the proportion (or percentage) of deaths in each age category might show an effect?

[Itsallaguess]

An update from the BBC here, discussing the recent and welcome drop in the regularly measured R-rate -

As we told you earlier the reproduction number, or R number, across the UK has fallen below 1 for the first time since July.

Latest estimates for the R number across the UK put it at between 0.7 and 0.9, suggesting the epidemic is shrinking.

The Office for National Statistics says infection levels are going down in all four UK nations, but experts warn infection levels still remain high




Source - https://www.bbc.co.uk/news/live/uk-56037505

Given that we're probably just moving into the area where the initial UK vaccinations might be starting to produce real-world results, then to see this happening before the bulk of those benefits flow through is quite encouraging...

Cheers,

Itsallaguess

[jfgw]

Perhaps showing the proportion (or percentage) of deaths in each age category might show an effect?

I have plotted the deaths for each age range as a percentage of all Covid deaths (all age ranges).


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/

I will tentatively suggest that deaths could be just starting to fall quite noticeably in the 80+ age groups. This may be more evident in a couple of weeks' time.


Julian F. G. W.

[Itsallaguess]

Some interesting news coming out of King's College regarding the very up-to-date figures that they're seeing from their ZOE symptom tracker app -

Coronavirus infections will be back to summer levels by early March, experts predicted as they said restrictions should be lifted "sooner rather than later".

The latest forecasts from the King's College ZOE symptom tracker app show that cases are falling dramatically and will soon be at the same levels as when the first lockdown was released.

The team estimates that there are just 14,818 new daily symptomatic cases of Covid in Britain on average, compared with 20,360 a week ago, a decrease of 27 per cent. By March 8, when it is hoped that schools will be able to open again, the ZOE app predicts that there will be 3,373 new daily cases and an estimated one in 780 active cases – the same levels last seen at the beginning of June.

However, Britain will be in a far better position than at the end of the first lockdown because around 20million people are likely to have been vaccinated by then.

Tim Spector, professor of genetic epidemiology at King's College London and the lead scientist on the app, said: "We are soon to be in the same place we were in early June, with the advantage of having a large proportion of the population vaccinated which could mean good news in terms of lifting some restrictions sooner rather than later.

"By March 8 we should have fewer than one in 780 people with symptoms, allowing us to get kids back into the classrooms and starting to allow people to exercise and meet, at least outdoors, where the risk of transmission is much lower.

The ZOE symptom tracker app combines data from one million weekly reports of symptoms and more than 15,000 swab tests conducted each week. While most other surveillance groups, such as the Office of National Statistics (ONS), have a time lag of cases, the ZOE app is a better indicator of what is happening in real time.

https://www.telegraph.co.uk/news/2021/02/12/coronavirus-infections-will-back-summer-levels-early-march/

Cheers,

Itsallaguess

[GeoffF100]

Here is an interesting report:

https://www.youtube.com/watch?v=wvTz24XmPA4

Their graph for the UK does not show any additional drop in hospital admissions as a result of vaccination. That is in contrast to Israel, where they have exclusively given two does of the Pfizer vaccine.

[Itsallaguess]

From the Times this morning -

Experts predict huge fall in English Covid patients -

Leaked data shows virus cases plummeting and hospitalisations to halve within a month

The number of coronavirus patients in hospitals in England will more than halve over the next month, according to internal government projections seen by The Times.

Hospital admissions and deaths are predicted to fall to October levels, according to estimates presented to No 10 by its scientific advisers.

They said that infection rates were falling faster than anticipated and that they were increasingly optimistic about the reopening of schools on March 8 and the relaxation of other restrictions in April.

https://www.thetimes.co.uk/article/experts-predict-huge-fall-in-english-covid-patients-7xxn8qrmp

Cheers,

Itsallaguess

[servodude]

From the Times this morning -

Experts predict huge fall in English Covid patients -

Leaked data shows virus cases plummeting and hospitalisations to halve within a month no

The number of coronavirus patients in hospitals in England will more than halve over the next month, according to internal government projections seen by The Times.

Hospital admissions and deaths are predicted to fall to October levels, according to estimates presented to No 10 by its scientific advisers.

They said that infection rates were falling faster than anticipated and that they were increasingly optimistic about the reopening of schools on March 8 and the relaxation of other restrictions in April.

https://www.thetimes.co.uk/article/experts-predict-huge-fall-in-english-covid-patients-7xxn8qrmp

Cheers,

Itsallaguess

I'd be really interested to see what their criteria for relaxation is
- whether it's driven by statistics, and which, and how measured?
- or by opinion and whose?

I think it would help if they could be clear about how and why they will make the decision
- more so than choosing a date and trying to make stuff fit to it

-sd

[Mike4]

I'd be really interested to see what their criteria for relaxation is
- whether it's driven by statistics, and which, and how measured?
- or by opinion and whose?

I think it would help if they could be clear about how and why they will make the decision
- more so than choosing a date and trying to make stuff fit to it

-sd

My impression is thay have no idea, other than balancing media pressure to drop all restrictions and get back to 'normal', with hospitals being rammed with dying patients.

Once the death rate is demonstrably on a downward trajectory, I'm expecting a headlong rush to cancel all restrictions just in time for the summer reduction in infection to be attributed to the vaccination programme. Rinse and repeat next winter as I'm really not convinced vaccinating 50% of the population is gonna get us to "herd immunity".