Number of Deaths

[Machiavelli]

This from the FT

Since the beginning of March, there have been 27,015 more deaths registered up to April 17 than the five-year average for the time of year.

With an average delay of four days between someone dying and their death being registered, the figures relate to the period to April 13, during which the government said there had been 11,408 deaths of people testing positive for coronavirus in English and Welsh hospitals ...

The official figures verified Financial Times modelling that suggested 41,000 people had died by last Tuesday either directly or indirectly as a result of coronavirus, with the death registrations higher than expected by the FT’s model.

With almost 30,000 excess deaths by mid-April across the UK, approximately two weeks ago, the number of total deaths now is likely to be in excess of 45,000, according to the FT model.

 

[Harry Wilson's tackle]

As far as statistical reporting and trends go, I quite agree.

However, as of yesterday, the Government told Britain that there were a total of 21,092 fatalities in Britain. This is despite the Government issuing detailed data yesterday that shows we actually passed the 21,092 figure on the 14th April, two weeks ago.

The true figure yesterday was over 35,095 fatalities.

I don't think that is small stuff. I believe that understating to this degree is, at best incompetence.

Now, where's my glasses

Hi. Johns Hopkins are still reporting 21157 deaths for the UK. Where did you get the 35K from? I am not saying the true figure isn't 35K, but the only thing that matters to assessment of statistical trends is if the rubric changes. It has changed in France (look at their reported numbers over time - big step jumps).

Yes, of course the absolute number is essential for knowing the absolute number, but that is a different issue.

Anyway if the Hopkins number jumps to 35K in the next 24 h I will cetainly boggle.

 

[WATFORD zero]

Hi. Johns Hopkins are still reporting 21157 deaths for the UK. Where did you get the 35K from? I am not saying the true figure isn't 35K, but the only thing that matters to assessment of statistical trends is if the rubric changes. It has changed in France (look at their reported numbers over time - big step jumps).

Yes, of course the absolute number is essential for knowing the absolute number, but that is a different issue.

Anyway if the Hopkins number jumps to 35K in the next 24 h I will cetainly boggle.

I completely agree about the difference between statistical and trend reporting and the absolute true number. However, nowhere have I actually seen the true number used, even some time after, when we know what the true number is. This gives us the situation yesterday, where the Government announced that the number of fatalities had reached 21,092 and, at the same time published evidence that we had exceeded 21,092 two weeks ago.

The ONS release the detailed figures of death by day of occurrence, weekly but 10-17 days in arrears. This includes all Covid deaths (not just hospital and also, because it's in arrears, covering the majority of lag in reporting of deaths).

Yesterday, they released the detailed figures up to 17th April showing a total of 23,916 (21,284 for England, 1,016 for Wales and 1,616 for Scotland.) It also showed that we exceeded yesterday's announced figure of 21,092 two weeks ago on 14th April.

From the actual deaths per day against Government announced Deaths per day, we can see how many more have died each day than the number being reported daily. Although these proportions have changed over the early stages of the pandemic, England have been averaging 63% announced to actual, Wales 52% and Scotland 58% consistently over the last 2 weeks of detailed statistics.

Take these figures and apply them to yesterdays announced figures 21,092 (which we already know was exceeded 2 weeks ago) and it gives you 35,095.

On 12th June, you will be able to check just how accurate I am.

 

[beorhthelm]

As far as statistical reporting and trends go, I quite agree.

However, as of yesterday, the Government told Britain that there were a total of 21,092 fatalities in Britain. This is despite the Government issuing detailed data yesterday that shows we actually passed the 21,092 figure on the 14th April, two weeks ago.

The true figure yesterday was over 35,095 fatalities.

I don't think that is small stuff. I believe that understating to this degree is, at best incompetence.

Now, where's my glasses

this can be fixed by government not reporting daily on NHS reported deaths. only have the final ONS numbers based on certificates and date of death, everything else is subject to revision. and even those numbers are open to interpretation.

 

[WATFORD zero]

this can be fixed by government not reporting daily on NHS reported deaths. only have the final ONS numbers based on certificates and date of death, everything else is subject to revision. and even those numbers are open to interpretation.

Well they have announced that they are going to include Care Home deaths in the daily figures, so by the time the ONS release their figures next week, I should be able to say fairly accurately how much closer this had made the daily announcements to the true figure.

 

[Harry Wilson's tackle]

Check out the Brazil numbers on the Johns Hopkins page. It is going bat shit mental there - most stark increase in numbers of cases since it went bat shit mental in Trumpyland.

So much for the Southern Hemisphere being resistant. Bit of a dampner on my 'seasonal' specuation too.....

 

[vegster]

Check out the Brazil numbers on the Johns Hopkins page. It is going bat shit mental there - most stark increase in numbers of cases since it went bat shit mental in Trumpyland.

So much for the Southern Hemisphere being resistant. Bit of a dampner on my 'seasonal' specuation too.....

Worldometers don't put it that high ?

https://www.worldometers.info/coronavirus/country/brazil/

 

[Harry Wilson's tackle]

Worldometers don't put it that high ?

https://www.worldometers.info/coronavirus/country/brazil/

 

[oldboyroy]

So the government are not telling the whole story then!!??

 

[vegster]

View attachment 122912

What I mean is that currently there are about 3,500 new cases a day, in a population of 200 million. We had a few hundred more in a population a third the size of Brazil..

 

[Weststander]

I don't think that we have the full picture from other EU countries yet - Spain and Italy at least.

According to the Guardian, Italy and Spain do not report non-hospital covid deaths in the figures released. Confirmed by Spanish and Italian sources.

The comparative graph produced at the daily Downing Street press conference each day, with two lines for the UK, only includes hospital deaths for Italy and Spain.

A separate Guardian article highlighted:
An Italian government survey of 10% of care homes found that 45% of covid deaths in Italy could be in care homes. If they’re correct, their true death toll is 27,359 * 100 / 55 = 50,000, plus covid victims who’ve died at home.

Similarly for Spain, a leaked Spanish government document suggested 57% of the country’s death toll was in care homes.

 

[Green Cross Code Man]

According to the Guardian, Italy and Spain do not report non-hospital covid deaths in the figures released. Confirmed by Spanish and Italian sources.

The comparative graph produced at the daily Downing Street press conference each day, with two lines for the UK, only includes hospital deaths for Italy and Spain.

A separate Guardian article highlighted:
An Italian government survey of 10% of care homes found that 45% of covid deaths in Italy could be in care homes. If they’re correct, their true death toll is 27,359 * 100 / 55 = 50,000, plus covid victims who’ve died at home.

Similarly for Spain, a leaked Spanish government document suggested 57% of the country’s death toll was in care homes.

Terrible figures but ones which should be referred to by those that are determined to make political capital from the numbers.

 

[Blue Valkyrie]

According to the Guardian, Italy and Spain do not report non-hospital covid deaths in the figures released. Confirmed by Spanish and Italian sources.

The comparative graph produced at the daily Downing Street press conference each day, with two lines for the UK, only includes hospital deaths for Italy and Spain.

A separate Guardian article highlighted:
An Italian government survey of 10% of care homes found that 45% of covid deaths in Italy could be in care homes. If they’re correct, their true death toll is 27,359 * 100 / 55 = 50,000, plus covid victims who’ve died at home.

Similarly for Spain, a leaked Spanish government document suggested 57% of the country’s death toll was in care homes.

So basically, we cannot compare with Italy and Spain, but can compare with France ?

 

[WATFORD zero]

So basically, we cannot compare with Italy and Spain, but can compare with France ?

Comparing any country against one another is always going to be near impossible given the various ways they collect and report data. But as I posted on this thread a couple of weeks ago

As I understand it currently, Spain, Italy and UK don't include Care Homes, France, Belgium and Sweden do. (Although that is only my understanding as getting definitive answers is not easy ).

And that's still unchanged AFAIK. The UK have said that they will be reporting Care homes as of today. Once that number is published I will be able to analyse the figures and see what impact it has on the UK figures. However, I can only comment on the UK as these are the only figures that I have looked in detail at (and know how to get).

 

[Blue Valkyrie]

Comparing any country against one another is always going to be near impossible given the various ways they collect and report data. But as I posted on this thread a couple of weeks ago



And that's still unchanged AFAIK. The UK have said that they will be reporting Care homes as of today. Once that number is published I will be able to analyse the figures and see what impact it has on the UK figures. However, I can only comment on the UK as these are the only figures that I have looked in detail at (and know how to get).

Do we know if the UK is adding in the care home previous 'backlog' figures all at once ?

In which case we expect a sudden massive increase in the total ?

 

[WATFORD zero]

Do we know if the UK is adding in the care home previous 'backlog' figures all at once ?

In which case we expect a sudden massive increase in the total ?

No idea, but we will find out this afternoon.

There are two aspects of the difference between the daily reported figure and the actual figure. Source of the information and Time lag. Depending on how it's reported this could take out one aspect of that difference.

 

[Harry Wilson's tackle]

Comparing any country against one another is always going to be near impossible given the various ways they collect and report data. But as I posted on this thread a couple of weeks ago



And that's still unchanged AFAIK. The UK have said that they will be reporting Care homes as of today. Once that number is published I will be able to analyse the figures and see what impact it has on the UK figures. However, I can only comment on the UK as these are the only figures that I have looked in detail at (and know how to get).

I have an idea. Can we assume that although every country measures cases and deaths differently, they generally use the same rubric on a day to day basis? I think we can (when they don't the absolute numbers jerk around like they have done a couple of times in France and once in China). Let us assume so for he moment. We can then calculate the ratio of cases to deaths in each country. These numbers will all be different. In themselves they don't mean anything much, and are not comparable between nations to any great deegree. However these numbers will stay constant if the rate of new cases matches the rate of deaths.

Consequently, it is reasonable to deduce that the rate of new cases will match the rate of deaths if we are in steady state, or plateau; the ratio (let's call it my statistic) will be constant in a country. Thus if we look at the difference in this statistic from one day to the next the difference should be zero if we are at plateau. Before we reach plateau the rate of increase of new cases will be greater than the rate of increase in deaths so the difference between consecutive days (monday minus Tuesday, say) will be positive. When we are over the plateau the rate of increase of deaths will exceed the rate of increase of cases for a few days (maybe a week) and my statistic will go negative until it raches a new steady state.

What I have found is this all hangs together. My statistic has been falling from positive values in every country higher up the cases table on a day to day basis for a couple of weeks, and in the last couple of days the daily differences have reached zero, and even gone negative in some countries. I have been posting the data.

So there is no need to look at absolute cases and absolute deaths. I doubt we will have a real clue about absolute numbers for months, if ever for some countries. But we can look at ratios and their trends and I think this is giving the clearest picture of what is happening worldwide in terms of predicting increases in cases, plateau and downturn.

At present there is clear evidence of downturn, started or about to start, worldwide. What we don't know yet is if this will sustain or bottom out. My statistic may detect early signs, though. Actually, let's not use English understantement and fake vagueness, the statistic will detect this.

 

[Green Cross Code Man]

I have an idea. Can we assume that although every country measures cases and deaths differently, they generally use the same rubric on a day to day basis? I think we can (when they don't the absolute numbers jerk around like they have done a couple of times in France and once in China). Let us assume so for he moment. We can then calculate the ratio of cases to deaths in each country. These numbers will all be different. In themselves they don't mean anything much, and are not comparable between nations to any great deegree. However these numbers will stay constant if the rate of new cases matches the rate of deaths.

Consequently, it is reasonable to deduce that the rate of new cases will match the rate of deaths if we are in steady state, or plateau; the ratio (let's call it my statistic) will be constant in a country. Thus if we look at the difference in this statistic from one day to the next the difference should be zero if we are at plateau. Before we reach plateau the rate of increase of new cases will be greater than the rate of increase in deaths so the difference between consecutive days (monday minus Tuesday, say) will be positive. When we are over the plateau the rate of increase of deaths will exceed the rate of increase of deaths and my statistic will go negative.

What I have found is this all hangs together. My statistic has been falling from positive values in every country higher up the cases table on a day to day basis for a couple of weeks, and in the last couple of days the daily differences have reached zero, and even gone negative in some countries. I have been posting the data.

So there is no need to look at absolute cases and absolute deaths. I doubt we will have a real clue about absolute numbers for months, if ever for some countries. But we can look at ratios and their trends and I think this is giving the clearest picture of what is happening worldwide in terms of predicting increases in cases, plateau and downturn.

At present there is clear evidence of downturn, started or about to start, worldwide. What we don't know yet is if this will sustain or bottom out. My statistic may detect early signs, though. Actually, let's not use English understantement and fake vagueness, the statistic will detect this.

Really excellent and thanks. I think everyone should bear that in mind, particularly when emoting about possible numbers. We need to be patient and then, eventually, we will have a clear picture. I am not aiming that comment at [MENTION=396]WATFORD zero[/MENTION] btw who is applying a scientific view, but at some posters who start to froth at the mouth a little and lose sight of objectivity.

I think you need to edit this part:

'When we are over the plateau the rate of increase of deaths will exceed the rate of increase of deaths and my statitsic will go negative.'

 

[Harry Wilson's tackle]

No idea, but we will find out this afternoon.

There are two aspects of the difference between the daily reported figure and the actual figure. Source of the information and Time lag. Depending on how it's reported this could take out one aspect of that difference.

If they add new deaths but not new cases this will make our cases to deaths ratio suddenly become smaller. If they add new cases from the care homes it should stay the same. I have a feeling they will count as new cases only those cases that resulted in deaths which is incorrect. My guess is we won't see a massive change in the ratio of cases to deaths.

Given that most if not all care home Covid deaths are diagnosed 'clinically' (which means a doctor is told on the phone that the patient died while they had flu like symptoms, and he/she says 'yep, put that down as Covid'), and those that are living have not been tested for Covid I am going to stick my neck out and suggest we are going to see a jump in deaths but an exactly proportionate jump in cases from care home data, as the only extra cases recorded will be the deaths. I can even see them forgetting to add all the care home deaths to the cases list. Anyway, I will be able to interpret this later when the numbers 'jump'.

 

[Harry Wilson's tackle]

Really excellent and thanks. I think everyone should bear that in mind, particularly when emoting about possible numbers. We need to be patient and then, eventually, we will have a clear picture. I am not aiming that comment at [MENTION=396]WATFORD zero[/MENTION] btw who is applying a scientific view, but at some posters who start to froth at the mouth a little and lose sight of objectivity.

I think you need to edit this part:

'When we are over the plateau the rate of increase of deaths will exceed the rate of increase of deaths and my statitsic will go negative.'

Cheers! Error corrected.

Well spotted, and thanks

Edit: [MENTION=396]WATFORD zero[/MENTION] has been very thoughtful about absolute numbers, and it is inevitable these will catch they eye, especially given the kerfuffle over care home data - not least because many of us have olds in care homes, and it's all a bit of a worry.