The Anonymous Widower

The Government’s Terrible COVID-19 Statistics

In an article in The Times today, Sir David Speigelhalter lambasts the Government over their collection of COVID-19 statistics.

I have been involved in the analysis of several very large databases and four factors improve the quality of the answers you get.

  1. The more records or in this case tests you have, the better. So I back David Speigelhalter! We should stop people on the street and test them!
  2. The more fields or pieces of data in each test you have, the better. So perhaps each test should be linked to your NHS record.
  3. The quality of the data is important. In my experience NHS scores about two out of five for quality as a lot of medical staff, often fill it in badly. In one case, I was looking for reasons for low birth weight babies and in many cases, the field was filled with 9.99 Kg.
  4. And then there’s the output of the data. I was taught how to display data for idiots, by the former Chief Accountant of a FTSE 500  company, who was working as the Chief Management Accountant of a Clearing Bank. He had found showing scatter diagrams with each branch as a single point on a large sheet, really got Branch Managers to think about what they are doing, if they weren’t running with the herd. These diagrams made problem branches stand out like the sorest of thumbs!

So where are the scatter diagrams for all of the Health Authorities in the UK, so everybody can see how their area, is doing against everybody else?

By the way, I trust Sir David, as we have the same birthday and share it with James Cameron, Jeff Thomson, Katherine Hamnett, Lawrence of Arabia, Menachem Begin and Madonna.

May 11, 2020 Posted by | Health | , , | Leave a comment

What Percentage Of Over 65s In England Have Tested Positive For COVID-19?

On the official government web site, they now give a breakdown of lab-confirmed cases in England by age and sex.

This is their graph.

Note.

  1. From 0-60, there ae more female cases.
  2. From 60-85, there are more male cases.
  3. Over 85, there are more female cases.

I do find this unusual. Could it be that more healthcare, teachers, nursery and care home workers are women under sixty.

I’ll now look at the individual figures.

Male

  • 0-4 – 356
  • 5-9 – 157
  • 10-14 – 167
  • 15-19 – 310
  • 20-24 – 938
  • 25-29 – 2256
  • 30-34 – 2549
  • 35-39 – 2493
  • 40-44 – 2986
  • 45-49 – 3782
  • 50-54 – 4457
  • 55-59 – 4725
  • 60-64 – 4473
  • 65-69 – 4121
  • 70-74 – 5064
  • 75-79 – 5816
  • 80-84 – 6598
  • 85-89 – 5703
  • 90+ – 3929

Under-65 – 29649

Over-65 – 31231

Over-75 – 22046

Female

  • 0-4 – 274
  • 5-9 – 129
  • 10-14 – 178
  • 15-19 – 501
  • 20-24 – 2303
  • 25-29 – 4756
  • 30-34 – 4546
  • 35-39 – 3842
  • 40-44 – 4160
  • 45-49 – 5038
  • 50-54 – 5538
  • 55-59 – 5018
  • 60-64 – 3729
  • 65-69 – 2684
  • 70-74 – 3359
  • 75-79 – 4266
  • 80-84 – 5792
  • 85-89 – 6387
  • 90+ – 6622

Under-65 – 40012

Over-65 – 29110

Over-75 – 23067

Total Under-65 – 69661

Total Over-65  – 60341

Total Over-75 – 45113

 

Note.

  1. In an earlier analysis,  I thought these were UK figures, so discount those.
  2. There are no figures for other parts of the UK.

I can now calculate the percentage of Over-65s, who have had lab-confirmed cases in England.

  • According to Age UK, there are 11,989,322 who are over 65 in the UK.
  • As England is 84.3 % of the UK population according to Wikipedia, that means there are 10,106,998, who are over 65 in England.

Using the total number of over 65 cases in England, this gives a figure of 0.6%.

I can now calculate the percentage of Over-75s, who have had lab-confirmed cases in England.

According to Age UK, there are 5,400,000, who are over 75 in the UK.

As England is 84.3 % of the UK population according to Wikipedia, that means there are 4,552,200, who are over 75 in England.

Using the total number of over 75 cases in England, this gives a figure of 0.99%.

I can now calculate the percentage of Under-65s, who have had lab-confirmed cases in England.

  • Using the previous figures, there are 45,870,180, who are under 65 in England.

Using the total number of under-65 cases in England, this gives a figure of 0.15%.

 

 

May 7, 2020 Posted by | Health | , | 3 Comments

Five Eastern Counties

The five Eastern counties around London have the following numbers of people, who have tested positive for COVID-19.

  • Cambridgeshire – 673 of 852,523 or 0.08%
  • Essex – 2320 of 1,832,752 or 0.13%
  • Kent – 3232 of 1,568.623 or 0.20%
  • Norfolk – 1355 of 903,680 or 0.15%
  • Suffolk – 936 of 768,556 or 0.12%

Note that the second figure is the population and the third is the percentage found positive.

Compare these figures with some other places, including Hackney where I live now.

Blackpool – 377 of 139,720 or 0.27%

Cumbria – 1734 of 498,888 or 0.35%

Hackney – 596 of 279,665 or 0.21%

You can make some interesting deductions.

April 28, 2020 Posted by | Health | , | 2 Comments

What Percentage Of People In The UK Survive COVID-19?

This is only a simple analysis based on the COVID-19 statistics published on Sunday, 26th April.

  • So far 152,840 people have been lab confirmed as having COVID-19.
  • There has also been 20,732 deaths in hospitals.
  • Suppose another 25% have died in care homes or in their own bed.
  • That would give a total of 25,915 deaths.
  • So rather crudely, if you get tested positive for COVID-19, you have a 17% chance of dying. What is the chance of dying from a serious stroke or breast cancer?
  • On the other hand 126,925 or 83% have survived.
  • Some, let’s say 20,000 are in ICU beds in hospitals, reducing the figures to 106,925 or 70% that survived.

We should be examining these seventy percent to see why they survived.

The official statistics concentrate on the negative side, but don’t publish figures like how many left hospital for convalescence at home or in an ordinary hospital ward!

Update – 27th April 2020

The actual figure of those in ICU beds yesterday was 18,667, which makes the figures 108, 258 and 71% have survived.

April 26, 2020 Posted by | Health | , , , | 4 Comments

Our Social Distancing Is Working, Says Sir Patrick Vallance

This is a headline on the front of today’s Times.

Sir Patrick is the government’s chief scientific adviser and I doubt he would use words like that lightly.

He noted these points.

  • Hospital admissions are increasing by a stable amount each day.
  • The NHS is coping as numbers have stopped accelerating.
  • He felt that reductions in travel were working.
  • Public transport usage has severely reduced.

He indicated that more of the same was needed.

I wouldn’t be surprised to see that by the end of this week, we’ll be able to make much better predictions on the progress of the COVID-19 virus through the UK population.

Here in the UK, we could be approaching the End of the Beginning!

 

 

March 31, 2020 Posted by | Health | , , | Leave a comment

Lockdown ‘Is On Course To Reduce Total Death Rate’

The title of this post, is the same as that of this article on The Times.

This is the introductory paragraph.

Britain is on course for an estimated 5,700 deaths from coronavirus, far lower than originally predicted, experts believe.

It’s all contained in a paper by Tom Pike at Imperial College,

Apparently, it is based on the premise that the UK follows the lockdown and social distancing.

Here’s hoping!

The story is also in the Mail and the Express.

March 28, 2020 Posted by | Health | , , | Leave a comment

All Along The South Coast

Taking the figures from this page on the BBC for the total number of COVID-19 cases and population for Health Authorities along the South Coast, we get the following.

  • Plymouth – 11 – 263,100 – 23,900
  • Devon – 43 – 795,286 – 18,500
  • Dorset – 11 – 376,484 – 34,200
  • Bournemouth – 12 – 395,784 – 33,100
  • Southampton – 11 – 252,796 – 23,000
  • Hampshire – 156 -1,376,316 – 8,800
  • Portsmouth – 25 – 215,133 – 8,600
  • West Sussex – 27 – 858,852 – 31,800
  • Brighton and Hove – 17 – 290,395 – 17,100
  • East Sussex – 9 – 554,590 – 7,400
  • Kent – 48 -1,568,623 – 32,700

Note the last figure is the number of residents per confirmed case. I would have thought that as the areas are similar, the last figure would be fairly constant.

The mean of the last figure is 21,700, with a standard deviation of 9,900.

It is my view that the standard deviation is too high, if all the areas were using the same health policies.

So I think testing methods are a bit variable!

Hopefully, the figures will settle down or those clever guys at Imperial will make sense of it all, in the next few days.

 

March 23, 2020 Posted by | Health | , | 3 Comments

Tracking COVID-19 Statistics

This post contains COVID-19 statistics gathered from this page.

Total Number of New Cases Per Day In The UK

  • 13th March – 207
  • 14th March – 264
  • 15th March – 330
  • 16th March – 152
  • 17th March – 407
  • 18th March – 676
  • 19th March- 643
  • 20th March – 714
  • 21st March – 1000
  • 22nd March – 665
  • 23rd March – 967
  • 24th March – 1427
  • 25th March – 1542
  • 26th March – 2129
  • 27th March – 2885
  • 28th March – 2546
  • 29th March – 2433
  • 30th March – 2107
  • 31st March – 3009
  • 1st April – 4324
  • 2nd April – 4244
  • 3rd April – 4450
  • 4th April – 3735
  • 5th April – 5914
  • 6th April – 3802
  • 7th April – 3634
  • 8th April – 5491
  • 9th April – 4344
  • 10th April – 5190
  • 11th April – 5234
  • 12th April – 5288
  • 13th April – 4342
  • 14th April – 5252
  • 15th April – 4605
  • 16th April – 4618
  • 17th April – 5599
  • 18th April – 5526
  • 19th April – 5850
  • 20th April – 4676
  • 21st April – 4301
  • 22nd April – 4451
  • 23rd April – 4583
  • 24th April – 5386
  • 25th April – 4913
  • 26th April – 4463
  • 27th April – 2730
  • 28th April – 3996
  • 29th April – 4076
  • 30th April – 6032
  • 1st May – 6201
  • 2md May – 4806
  • 3rd May – 4339
  • 4th May – 3985
  • 5th May – 4406
  • 6th May – 6111
  • 7th May – 5614
  • 8th May – 4649
  • 9th May – 3896
  • 10th May – 3923
  • 11th May – 3877
  • 12th May – 3403
  • 13th May – 3242
  • 14th May – 3446
  • 15th May – 3560
  • 16th May – 3451
  • 17th May – 3142
  • 18th May – 2684
  • 19th May – 2412
  • 20th May – 2472
  • 21st May – 2615
  • 22nd May – 3287
  • 23rd May – 2959
  • 24th May – 2409
  • 25th May – 1625
  • 26th May – 2004
  • 27th May – 2013
  • 28th May – 1887
  • 29th May – 2095
  • 30th May – 2445
  • 31st May – 1936
  • 1st June – 1570
  • 2nd June – 1613

Note that the figures seem to be going up and down all over the place.

Cumulative And incremental Cases In London

  • 21st May – 26,628
  • 22nd May – 26,683, 65
  • 23rd May – 26,742, 59
  • 24th May – 26, 780, 38
  • 25th May – 26818, 38
  • 26th May – 26840, 22
  • 27th May – 26867, 27
  • 28th May – 26883, 16
  • 29th May – 26913, 30
  • 30th May – 26938, 25
  • 31st May – 26968, 30
  • 1st June – 26989, 21
  • 2nd June – 27021, 32

Cumulative And Incremental Cases in Hackney

  • 15rh March – 22
  • 6th March – 22, 0
  • 17th March – 22, 0
  • 18th March – 29, 7
  • 19th March- 32, 3
  • 20th March – 45, 13
  • 21st March – 52, 7
  • 22nd March – 64, 12
  • 23rdMarch -72, 8
  • 24th March – 74, 2
  • 25th March – 85, 11
  • 26th March – 119, 34
  • 27th March – 134, 15
  • 28th March – 151, 17
  • 29th March – 177, 26
  • 30th March – 178, 12
  • 31st March – 213, 38
  • 1st April – 221, 8
  • 2nd April – 266, 45
  • 3rd April – 298, 32
  • 4th April – 304. 6
  • 5th April – 345, 41
  • 6th April – 362, 17
  • 7th April – 365, 3
  • 8th April – 412, 47
  • 9th April – 444, 32
  • 10th April – 470, 26
  • 11th April – 475, 5
  • 12th April – 498, 23
  • 13th April – 509, 11
  • 14th April – 503, -6
  • 15th April – 516, 13
  • 16th April – 524, 8
  • 17th April – 537, 13
  • 18th April – 543, 6
  • 19th April – 552, 8
  • 20th April – 555, 3
  • 21st April – 568, 13
  • 22nd April – 572, 4
  • 23rd April – 580, 8
  • 24th April – 586, 6
  • 25th April – 590, 4
  • 26th April – 595, 5
  • 27th April – 596, 1
  • 28th April – 597, 1
  • 29th April – 597, 0
  • 30th April – 600, 3
  • 1st May – 604, 4
  • 2nd May – 606, 2
  • 3rd May – 609, 3
  • 4th May – 611, 2
  • 5th May – 614, 3
  • 6th May – 615, 1
  • 7th May – 620, 5
  • 8th May – 622, 2
  • 9th May – 622, 0
  • 10th May – 623, 1
  • 11th May 624, 1
  • 12th May – 624, 0
  • 13th May – 624, 0
  • 14th May – 626, 2
  • 15th May – 626, 0
  • 16th May – 637, 11
  • 17th May – 639, 2
  • 18th May – 639, 0
  • 19th May – 640, 1
  • 20th May – 640. 0
  • 21st May – 641, 1
  • 22nd May – 644, 3
  • 23rd May – 645, 1
  • 24th May – 645, 0
  • 25th May – 645, 0
  • 26th May – 646, 1
  • 27th May – 646. 0
  • 28th May – 646, 0
  • 29th May – 647, 1
  • 30th May – 647, 0
  • 31st May – 647, 0
  • 1st June – 649, 2
  • 2nd June – 649, 0

Note.

  1. The increases are out of phase with the previous figures.
  2. 5th April seemed to be a bad day!
  3. 14th April saw a change of web site and hopefully better figures.
  4. What happened in Hackney on the 16th May?

March 22, 2020 Posted by | Health | , | Leave a comment

A Statistical Provocateur

Not by words, but they are used in this thought-provoking article on the BBC web site.

It shows how the divorce rate mirrors margarine consumption, according to a web site called Spurious Correlations.

The trouble is that people tend to believe these relationships or not, as it suits them.

Perhaps the most famous statistical correlation is that smoking causes lunch cancer. But people still smoke! Or should that be idiots still smoke?

One correlation used by the Police, may have been found using my software called Daisy. Analysis by the Metropolitan Police showed that if they found a car, had no valid excise duty, insurance or MOT, then there was a fifty percent chance that if they stopped the car, they’d find evidence of a non-motoring offence. The logic behind this, is if you are say a burglar, drug-dealer or other crimnal, you are less likely to make sure that your car is legal.

May 26, 2014 Posted by | News | | Leave a comment

A Must View Web Site For Savers And Investors

Out of curiosity, I wanted to find out, what was the rate you’d got on savings in 2008.

So I searched Google and found this page on a web site called Swanlowpark. This is the first statement on the home page.

This website centres on one theme: how well have your savings kept their value when compared to inflation.
Website is not-for-profit, there is no advertising.

Looking at the page for annual savings rates, there are some interesting figures. Draw your own conclusions or read the author’s comments here.

 

March 29, 2014 Posted by | Finance | , | Leave a comment