Number of Deaths

[Weststander]

Yes in effect the the figures are from various dates going back to March in cases and suggests that many were before lockdown. God only knows what would of happened if we hadn't gone into lockdown!! Credit to the government for that but obviously many questions about how an earlier lockdown would of prevented possibly thousands of deaths.

Earlier mandatory lockdowns would've undoubtedly saved many lives. Here and overseas.

My only question is, very early doors, whether the public would've taken it seriously and adhered to it? Even today I've seen perfectly respectable extended families with grandparents out walking next to each, not a care in the world, and that's in the knowledge of c. 20,000 dead Brits, plus it's been drummed into people that they may live but they'll spread the virus to strangers. Arrogance, nonchalance, selfishness?

Future covid-19 waves and other pandemics, will be acted on immediately by all non-Trump administrations.

 

[Machiavelli]

Or more likely, many of those WERE cv-related, but either undiagnosed, or conveniently mis-diagnosed. Like my elderly neighbour, who died last week in a care home, of a ‘non-specific pneumonia’.

It’s statistically very unlikely that the numbers of cv deaths are not much, much higher than stated. People talk of the ‘figure above the normal numbers’ of winter flu deaths, to play down the cv figures. This ignores, that the unprecedented lock-down scenario, must mean that the spread of ALL viruses are controlled, so there must surely be far less flu about - very, very few can have died from infectious diseases other than cv.

Oh, I agree, and as revealing as that graph is (if it transpires to be accurate), it's merely for hospital deaths, and the news on the ground coming from care homes, etc is none too encouraging. I do think that, in the long run, those ONS additional deaths will be an important figure to look at, because alongside the undiagnosed and misdiagnosed, there's also the difficult to diagnose (eg, as I posted earlier elsewhere, cancer deaths caused by attention oriented towards CV19).

 

[WATFORD zero]

Oh, I agree, and as revealing as that graph is (if it transpires to be accurate), it's merely for hospital deaths, and the news on the ground coming from care homes, etc is none too encouraging. I do think that, in the long run, those ONS additional deaths will be an important figure to look at, because alongside the undiagnosed and misdiagnosed, there's also the difficult to diagnose (eg, as I posted earlier elsewhere, cancer deaths caused by attention oriented towards CV19).

What I don't understand is that, as of yesterday, 22nd April, the government statistics included a line 'UK 'all settings', showing all deaths up to April 10th.

From the ONS data published yesterday, all Covid deaths that had been registered up to 10th April as occurring up to April 10th, was 10,335 (yellow line on graph below). Further to that between April 10th and April 18th (the cut-off for ONS data published yesterday) a further 2,786 fatalities were registered as having occurred before 10th April, giving total fatalities of 13,121 (red line on graph below). I suspect that a large proportion of these are deaths outside of hospital as you would expect hospital reporting to be more efficient.

So, as of yesterday we know that the total number of deaths up to April 10th is at least 13,121 (and still likely to rise due to delays in reporting, but with no idea how much).



Yesterday evening, the Government on all it's presentations, and the BBC as highlighted by [MENTION=4573]Green Cross Code Man[/MENTION] above, then decided that the deaths registered between April 10th and April 18th, but occurring before April 10th, would be removed from the total fatality figures, leaving the 10,335 reported by April 10th as the total fatalities. UK (all settings) on the graph below.



What is puzzling me is why on earth would you remove 2,786 fatalities (over 25% of the total fatalities) that you know have occurred in the period you are reporting (which, for valid reasons, is already 11 days out of date), when you have known these figures 3 days before you have to report it. It just adds a further 7 days delay in the publishing of figures that doesn't need to be there and confuses the whole situation more

 

[Machiavelli]

What I don't understand is that, as of yesterday, 22nd April, the government statistics included a line 'UK 'all settings', showing all deaths up to April 10th.

From the ONS data published yesterday, all Covid deaths that had been registered up to 10th April as occurring up to April 10th, was 10,335 (yellow line on graph below). Further to that between April 10th and April 18th (the cut-off for ONS data published yesterday) a further 2,786 fatalities were registered as having occurred before 10th April, giving total fatalities of 13,121 (red line on graph below). I suspect that a large proportion of these are deaths outside of hospital as you would expect hospital reporting to be more efficient.

So, as of yesterday we know that the total number of deaths up to April 10th is at least 13,121 (and still likely to rise due to delays in reporting, but with no idea how much).

View attachment 122627

Yesterday evening, the Government on all it's presentations, and the BBC as highlighted by [MENTION=4573]Green Cross Code Man[/MENTION] above, then decided that the deaths registered between April 10th and April 18th, but occurring before April 10th, would be removed from the total fatality figures, leaving the 10,335 reported by April 10th as the total fatalities. UK (all settings) on the graph below.

View attachment 122628

What is puzzling me is why on earth would you remove 2,786 fatalities (over 25% of the total fatalities) that you know have occurred in the period you are reporting (which, for valid reasons, is already 11 days out of date), when you have known these figures 3 days before you have to report it. It just adds a further 7 days delay in the publishing of figures that doesn't need to be there and confuses the whole situation more

You've asked me all sorts of questions I can't answer. I think I follow what you're saying, and it's just left me more confused! I suspect that the answer might come from the fact that there are different graphs being used, and we're/you're/I'm making assumptions with one based upon another. Apologies, but I don't have the time at present to attempt to do any better.

 

[WATFORD zero]

You've asked me all sorts of questions I can't answer. I think I follow what you're saying, and it's just left me more confused! I suspect that the answer might come from the fact that there are different graphs being used, and we're/you're/I'm making assumptions with one based upon another. Apologies, but I don't have the time at present to attempt to do any better.

Sorry, it wan't a question, it was a statement.

Yesterday at the daily briefing the Government understated the total fatalities up to 10th April by over 25% from the actual fatalities by introducing a further 7 day delay in reporting. (The long winded explanation being above).

The question I am asking is if anyone would have any idea why they would do this ?

 

[beorhthelm]

What is puzzling me is why on earth would you remove 2,786 fatalities (over 25% of the total fatalities) that you know have occurred in the period you are reporting (which, for valid reasons, is already 11 days out of date), when you have known these figures 3 days before you have to report it. It just adds a further 7 days delay in the publishing of figures that doesn't need to be there and confuses the whole situation more

the number in the COBR report is actually 11413 so that'll confuse you even more right? we shouldnt try and pick holes in these data, the complete numbers come out in the end.

 

[dazzer6666]

the number in the COBR report is actually 11413 so that'll confuse you even more right? we shouldnt try and pick holes in these data, the complete numbers come out in the end.

Agreed - the key stats re deaths is the ONS one pinned to actual date of death (confirmed by registration) but these have anything up to a 2 week time lag it seems. It's only this one which will unequivocally confirm where the peak is/was and how we're progressing - other indicators are useful (in particular the number of people in hospital being treated for CV19) but not comprehensive and prone to some quite wild fluctuation as we've seen. Problem is no-one wants to hear two-week old data...................

 

[WATFORD zero]

the number in the COBR report is actually 11413 so that'll confuse you even more right? we shouldnt try and pick holes in these data, the complete numbers come out in the end.

I'm not sure understating the known fatalities by 25% is 'picking holes'.

We know all the figures from our discussion yesterday where you helpfully pointed that the difference of 2,786 (>25% of total) being deaths reported between 10th and 18th April, but having occurred before 10th April

So UK deaths known to have occurred before 10th April are at least 13,121.

So, why would the Government remove over 25% of the known fatalities for the reporting period, before publishing ?

I have serious concerns about Care Homes and I believe these particular statistics are heavily influenced by Care Home reporting. The sooner we get accurate figures for periods (even in arrears), the better we can see any discrepancy in our early estimates and then use those figures for estimating (and reacting to) trends going forwards. I would rather this took priority over any other interests and therefor, can't see any benefit in introducing delays to the process.

 

[WATFORD zero]

Agreed - the key stats re deaths is the ONS one pinned to actual date of death (confirmed by registration) but these have anything up to a 2 week time lag it seems. It's only this one which will unequivocally confirm where the peak is/was and how we're progressing - other indicators are useful (in particular the number of people in hospital being treated for CV19) but not comprehensive and prone to some quite wild fluctuation as we've seen. Problem is no-one wants to hear two-week old data...................

I'm afraid that is the key stat I'm questioning, and it appears there are two answers, fairly significantly different.

actual date of death (confirmed by registration) Date of death before April 10th registered before April 10th - 10,335

actual date of death (confirmed by registration) Date of death before April 10th registered before April 18th - 13,121

Both of these figures were released by the ONS yesterday. The Government are using the first, I believe they should be using the more up to date one as it will include more Care Home figures (based on the assumption hospitals are more efficient in reporting)

 

[dazzer6666]

I'm afraid that is the key stat I'm questioning, and it appears there are two answers, fairly significantly different.

actual date of death (confirmed by registration) Date of death before April 10th registered before April 10th - 10,335

actual date of death (confirmed by registration) Date of death before April 10th registered before April 18th - 13,121

Both of these figures were released by the ONS yesterday. The Government are using the first, I believe they should be using the more up to date one as it will include more Care Home figures (based on the assumption hospitals are more efficient in reporting)

Ah Ok - sorry hadn't realised - I'd agree although the whole situation re stats - at least comparable stats - is a bit of a mess globally. There seems to be no consistency, which means many of the country comparison charts are fundamentally flawed. A Spanish colleague of mine thinks their care home numbers are massively understated.

 

[beorhthelm]

I'm afraid that is the key stat I'm questioning, and it appears there are two answers, fairly significantly different.

actual date of death (confirmed by registration) Date of death before April 10th registered before April 10th - 10,335

actual date of death (confirmed by registration) Date of death before April 10th registered before April 18th - 13,121

Both of these figures were released by the ONS yesterday. The Government are using the first, I believe they should be using the more up to date one as it will include more Care Home figures (based on the assumption hospitals are more efficient in reporting)

it doesnt make sense in the end. point made earlier is the COBR report number (11413) is neither of those numbers. looking across different data sets with different date ranges will raise questions. if you think this is masking care home fatalities, it will come out next weeks numbers.

 

[giveusfalmerplease]

call me a cynic but maybe they dont want to suddenly show a 25% increase in the figures so that we suddenly jump up to the top of the charts for deaths in Europe. . However, how many other countries also dont include deaths outside hospital. We already know both Spain and Italy ignore the deaths in Nursing homes.

 

[Harry Wilson's tackle]

The rate of increases is still falling (yellow highlight). The ratio of cases to deaths is harmonising across the world, with the liars Israel and Russia, the only major out of step reporters (with unfeasibly low deaths to cases). My suppositions and expectations still hanging in there, but it is still far too soon to know if we are approaching a steady state or a downturn. My money is on steady state. My employer (a uni) has instructed staff that there will be no physical lectures, practicals or tutorials till Jan 2021 at the earliest. We haven't told the students yet. As Elvis Costello sang, there's no money-back guarantee on future happiness.



Oh, and the ratio of cases to deaths should end up the same everywhere, albeit scaled and much higher in nations with good health care and social cohesion. Judging by current numbers I expect to see this ratio being about ten times bigger in Western Europe than in India, Pakistan, and Brazil, etc., and at least twice as big as Russia. At the moment the numbers are absolutely inverted, and I think this reflects hopeless reportaing in the third world, plus some lying. So in fact for the value of absolute numbers, only Western Europe counts (and even then with scaling due to poor diagnosis). Ooo look, all the numbers are the same (around 8) apart from Germany. Their number is bigger because they are more inclined to diagnose COVID on the basis of fewer symptoms. Cautious folk.

Oh, and I heard today that diagnosis is still essentially clinical. We do not have sensitive tests to diagnose properly. So the number of cases is all over the place from country to country. Which is part of the reason for the mad numbers in 'cases' data. 'Deaths' data is also inaccurate for the same reason.

I would guess that the true number of cases per deaths will be around 100-200 (optimist) in the UK with a majority of cases not recorded. In India I expect it to be around 20. They are basically sending 150 million internal migrant workers home, where many will die on the 200 mile walk home (I kid you not) or will die of starvation along with their family when they arrive home with no income. None of that will be recorded as Cova.

 

[WATFORD zero]

it doesnt make sense in the end. point made earlier is the COBR report number (11413) is neither of those numbers. looking across different data sets with different date ranges will raise questions. if you think this is masking care home fatalities, it will come out next weeks numbers.

Do you want to show me where this 11,413 is in the report and more specifically, what this number is, as I can't find it

https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/causesofdeath/articles/comparisonofweeklydeathoccurrencesinenglandandwales/uptoweekending10april2020

From the Summary on the first page.

2.Main points

This short article accompanies the Weekly deaths release for the week ending 10 April 2020 and explains the differences between various data sources that report on coronavirus (COVID-19) deaths.

A total of 10,350 deaths involving COVID-19 were registered in England and Wales between 28 December 2019 and 10 April 2020 (year to date).

Including deaths that occurred up to 10 April but were registered up to 18 April, the number involving COVID-19 was 13,121.

Thanks

 

[WATFORD zero]

Taking the data released by the ONS and then comparing against the Government daily announcements I believe that we can now more accurately predict the effect of the time lag on the reporting of actual fatalities.

I have taken the figures from 17th March (when daily fatalities were 100) through to April 10th (last day for which the ONS have released detailed figures).

On average, over that period the Government announced daily death rate has been between 38% and 70% of the actual death rate (Average 52%).
The most recent 2 weeks of that period the accuracy has improved to an of Average 58%.

If the figures were to continue to become more accurate at the same rate yesterday's total fatalities announced would be 64% of the actual fatalities so the 18,100 announced would be in the region of 28,281. Hopefully, if numbers are falling, the effect of the time lag may not be quite as great, but I may be clutching at straws.

Sorry

 

[beorhthelm]

Do you want to show me where this 11,413 is in the report and more specifically, what this number is, as I can't find it

it was in the data set used for the COBR report.
https://www.gov.uk/government/publi...ny-coronavirus-press-conference-21-april-2020

 

[WATFORD zero]

it was in the data set used for the COBR report.
https://www.gov.uk/government/publi...ny-coronavirus-press-conference-21-april-2020

It's a number in the data, but didn't make it into the report ?

Is it possible that you have misinterpreted what that figure in the data is, as I would have expected it to be a headline figure, wouldn't you ?

 

[Soylent Blue]

View attachment 122628

What is puzzling me is why on earth would you remove 2,786 fatalities (over 25% of the total fatalities) that you know have occurred in the period you are reporting (which, for valid reasons, is already 11 days out of date), when you have known these figures 3 days before you have to report it. It just adds a further 7 days delay in the publishing of figures that doesn't need to be there and confuses the whole situation more

That graph is weird. If it showed deaths per capita, which it should, then USA and sweden would flip with the UK above both.


Also there appears to be falls in crime and road accidents(the later by 50% in Los Angeles), how does this impact the data in the form of an absorbtion effect?

Daily road deaths divide by 2 x time period?

 

[Titanic]

Sir Patrick Vallence just said that he expects the daily number of deaths to remain at the current levels for the next two weeks.... that is another 7000

 

[beorhthelm]

Sir Patrick Vallence just said that he expects the daily number of deaths to remain at the current levels for the next two weeks.... that is another 7000

here's a morbid observation. we are not reportedly anywhere near capacity of ICU of about 4000 (plus the unused Nightingales). with a over a week of ICU expected for a serious patient before recovery or death, and at least a majority recovering, how do we get to 7000?? how many are not going through the intensive care route?