Number of Deaths

[Steve in Japan]

Suspect the voluntary mass use of masks is a big one.

I think so too, but our scientists seem not to agree. Or maybe we just haven't got enough supply to cope and so they don't want to mandate it.

 

[RossyG]

That's a tough one but probably not materially different. But they are certainly bigger users of hand sanitisers due to prior experiences with SARS etc.

Thanks, that’s interesting. A senior Swedish epidemiologist said last week that hand hygiene is the key. I wonder if this was the case here?

 

[Harry Wilson's tackle]

Apologies for the mess above. I preface this ramble with the health warning that it is just a back of the enveope exercise that may interest a small number of you. My inferences are guesses, because the sources of error in every number are not identifiable. If a particular person feels inclined to stand on high and piss all over this post, there won't be any replies from me because I have my 'ignore' umbrella up.

First, attached is a comparison of April 14 and today's Johns Hopkins data.

View attachment 123231


Compared with Apr 14, the fall in ratio of total number of recorded cases to total number of recorded deaths (green column) indicates that more of the people who catch Covid are dying from it now compared with Apr 14. This, if we accept all the numbers as accurate and also the result of the same type of testing and data processing from then to now, at least within nations, means the virus is becoming more deadly, or has passed from the young to the old and vulnerable so killing more of the people it affects. In every nation.

Either may be true (especially the second possibility) but there are other better explanations for the apparent increase in lethality of the virus worldwide. For example, we may have changed the way we diagnose the disease. Actually we are now diagnosing better due to better testing (allegedly) and better awareness that Covid may be a cause of a death. But that would actually increase the ratio of cases to deaths, not reduce it. And yet the ratio of cases to deaths has actually fallen everywhere in the world, even in Russia (where we expect there is mass lying and faking). I am prepared to accept that the ratio of cases to deaths has indeed fallen, partly because the data says this is the case globally. There is no evidence the virus has mutated. I don’t think the argument that it has taken longer to get to the elderly in care homes can explain the numbers worldwide. Do Brazil, Turkey or India even have care homes?

What about the possibility that we were at the start of the epidemic on Apr 14, and deaths took a while to ‘catch up’ with cases. This may well be part of it. It would mean that countries that were ahead of the curve (early cases) should have smaller numbers in the yellow column. OK, Spain. Italy, France, UK, Sweden…..that fits. Germany and China don’t, though. And Germany really should have an increase in cases to deaths owing to their legendary increase in testing (from a standing start – like us all), not one of the bigger increases in deaths to cases (or falls in cases to deaths) – the negative number in the yellow column. If you test more, are willing to diagnose Covid more readily and are more proactive and professional the number of cases will increase disproportionately to the increase in deaths. Yet we don’t see this anywhere, let alone in Germany.

I don’t know the answer to this, other than that I assume we, in every nation, are getting closer to the true ratio of cases to deaths all the time, and all the numbers worldwide are heading in the same direction – towards smaller numbers than a month ago.

It certainly doesn’t mean that if you catch Covid you have a better chance of survival now than a month ago – no number anywhere in the world supports that. Also the argument that the virus is now picking off (killing) the more vulnerable doesn’t add up, because typically a virus will pick off the most vulnerable first. We made no attempt to restrict exposure in care homes initially so I would have expected deaths to be high there early on, not higher later on.

I still find this quite baffling – very clear patterns globally despite the wide variation in the provenance of the data, and yet a counterintuitive implication (that people are more likely to die of Covid if they get it now than if they got it a month ago – everywhere in the world). I just don’t buy that, which is why I have previously focused on the change in the ratio itself over shorter terms, with the hypothesis (or assumption) that the ratio may tell us something about whether the number of new cases is rising or falling (form its ratio to deaths). Over the last few weeks this number and my inferences have fitted together well with reported numbers of cases worldwide, but this doesn’t really tell us much and, although I accept the rate of increase in new cases from day to day is falling almost everywhere, I still don’t have much confidence in the accuracy of absolute numbers of cases, or absolute numbers of deaths.

As others have said, in the UK, the latter is best inferred from ONS data, and there are delays in the acquisition of this, and there is some uncertainty about the inferences. All that said, I am cautiously inclined to side with the optimists and suspect we are headed in the right direction. But that does not make it likely that we will be safe any time soon. Nor do I think we can predict the tolerable level of excess death either in terms of what this is or when it will be reached. Sorry to drone on but this remains more nuanced than a simple smile or frown at successive daily death counts.

 

[Mellotron]

I think so too, but our scientists seem not to agree. Or maybe we just haven't got enough supply to cope and so they don't want to mandate it.

Bit of the former but more of the latter. I reckon once we’re in a situation where NHS no longer looks close to stretched PPE wise, masks in public will be mandated.

 

[Billy the Fish]

But they are certainly bigger users of hand sanitisers due to prior experiences with SARS etc.

Think this is a huge factor when comparing us to the SE Asian countries. They've already had to deal with epidemics in recent memory so governments and the public are ready and know what to do to fight it ASAP.

 

[vegster]

Apologies for the mess above. I preface this ramble with the health warning that it is just a back of the enveope exercise that may interest a small number of you. My inferences are guesses, because the sources of error in every number are not identifiable. If a particular person feels inclined to stand on high and piss all over this post, there won't be any replies from me because I have my 'ignore' umbrella up.

First, attached is a comparison of April 14 and today's Johns Hopkins data.

View attachment 123231


Compared with Apr 14, the fall in ratio of total number of recorded cases to total number of recorded deaths (green column) indicates that more of the people who catch Covid are dying from it now compared with Apr 14. This, if we accept all the numbers as accurate and also the result of the same type of testing and data processing from then to now, at least within nations, means the virus is becoming more deadly, or has passed from the young to the old and vulnerable so killing more of the people it affects. In every nation.

Either may be true (especially the second possibility) but there are other better explanations for the apparent increase in lethality of the virus worldwide. For example, we may have changed the way we diagnose the disease. Actually we are now diagnosing better due to better testing (allegedly) and better awareness that Covid may be a cause of a death. But that would actually increase the ratio of cases to deaths, not reduce it. And yet the ratio of cases to deaths has actually fallen everywhere in the world, even in Russia (where we expect there is mass lying and faking). I am prepared to accept that the ratio of cases to deaths has indeed fallen, partly because the data says this is the case globally. There is no evidence the virus has mutated. I don’t think the argument that it has taken longer to get to the elderly in care homes can explain the numbers worldwide. Do Brazil, Turkey or India even have care homes?

What about the possibility that we were at the start of the epidemic on Apr 14, and deaths took a while to ‘catch up’ with cases. This may well be part of it. It would mean that countries that were ahead of the curve (early cases) should have smaller numbers in the yellow column. OK, Spain. Italy, France, UK, Sweden…..that fits. Germany and China don’t, though. And Germany really should have an increase in cases to deaths owing to their legendary increase in testing (from a standing start – like us all), not one of the bigger increases in deaths to cases (or falls in cases to deaths) – the negative number in the yellow column. If you test more, are willing to diagnose Covid more readily and are more proactive and professional the number of cases will increase disproportionately to the increase in deaths. Yet we don’t see this anywhere, let alone in Germany.

I don’t know the answer to this, other than that I assume we, in every nation, are getting closer to the true ratio of cases to deaths all the time, and all the numbers worldwide are heading in the same direction – towards smaller numbers than a month ago.

It certainly doesn’t mean that if you catch Covid you have a better chance of survival now than a month ago – no number anywhere in the world supports that. Also the argument that the virus is now picking off (killing) the more vulnerable doesn’t add up, because typically a virus will pick off the most vulnerable first. We made no attempt to restrict exposure in care homes initially so I would have expected deaths to be high there early on, not higher later on.

I still find this quite baffling – very clear patterns globally despite the wide variation in the provenance of the data, and yet a counterintuitive implication (that people are more likely to die of Covid if they get it now than if they got it a month ago – everywhere in the world). I just don’t buy that, which is why I have previously focused on the change in the ratio itself over shorter terms, with the hypothesis (or assumption) that the ratio may tell us something about whether the number of new cases is rising or falling (form its ratio to deaths). Over the last few weeks this number and my inferences have fitted together well with reported numbers of cases worldwide, but this doesn’t really tell us much and, although I accept the rate of increase in new cases from day to day is falling almost everywhere, I still don’t have much confidence in the accuracy of absolute numbers of cases, or absolute numbers of deaths.

As others have said, in the UK, the latter is best inferred from ONS data, and there are delays in the acquisition of this, and there is some uncertainty about the inferences. All that said, I am cautiously inclined to side with the optimists and suspect we are headed in the right direction. But that does not make it likely that we will be safe any time soon. Nor do I think we can predict the tolerable level of excess death either in terms of what this is or when it will be reached. Sorry to drone on but this remains more nuanced than a simple smile or frown at successive daily death counts.

Everything we know and understand about healthcare is based on one real assumption that is genuinely correct, people die of illnesses and diseases but with each death we learn something ( usually !) Lot's has been learned about the virus and by previous experience we should expect the death rate go down as knowledge circulates in the medical world and countries learn new therapies and procedures...….. It's really worrying if that ratio is increasing.

 

[beorhthelm]

...It certainly doesn’t mean that if you catch Covid you have a better chance of survival now than a month ago – no number anywhere in the world supports that. Also the argument that the virus is now picking off (killing) the more vulnerable doesn’t add up, because typically a virus will pick off the most vulnerable first. We made no attempt to restrict exposure in care homes initially so I would have expected deaths to be high there early on, not higher later on.

I still find this quite baffling – very clear patterns globally despite the wide variation in the provenance of the data, and yet a counterintuitive implication (that people are more likely to die of Covid if they get it now than if they got it a month ago – everywhere in the world). I just don’t buy that, which is why I have previously focused on the change in the ratio itself over shorter terms, with the hypothesis (or assumption) that the ratio may tell us something about whether the number of new cases is rising or falling (form its ratio to deaths). Over the last few weeks this number and my inferences have fitted together well with reported numbers of cases worldwide, but this doesn’t really tell us much and, although I accept the rate of increase in new cases from day to day is falling almost everywhere, I still don’t have much confidence in the accuracy of absolute numbers of cases, or absolute numbers of deaths.

think flaws are appearing because too many assumptions being made, volitile data and policy changes. example, virus getting the vulnerable first only applies if they are exposed evenly. if, for example many are shipped out to care homes a couple of weeks into the study and then contract the virus, they will appear in the death count much later. and near all the victims are in a vulnerable catagory of some sort). i wonder if those in the deaths count may not even be in earlier cases count, at least not until they die and reported with covid. might only be a few % but its something to screw the numbers.

another one is if a country is adopting track and trace they will tend to pick up cases, which because of demographic similarities may be more like to die too. extreme example, if i test everyone in a care home with a case, i'll get a lot more additional cases and they are all far more likly to contribute to mortality, than say if i picked a football team to test, or if i pick same number from population at random. just thoughts, maybe wrong too, to highlight there are so many variables involved in the simple cases/death comparison. i dont think the ratio based on short window works on the longer time frame for these many reasons.

 

[Harry Wilson's tackle]

think flaws are appearing because too many assumptions being made, volitile data and policy changes. example, virus getting the vulnerable first only applies if they are exposed evenly. if, for example many are shipped out to care homes a couple of weeks into the study and then contract the virus, they will appear in the death count much later. and near all the victims are in a vulnerable catagory of some sort). i wonder if those in the deaths count may not even be in earlier cases count, at least not until they die and reported with covid. might only be a few % but its something to screw the numbers.

another one is if a country is adopting track and trace they will tend to pick up cases, which because of demographic similarities may be more like to die too. extreme example, if i test everyone in a care home with a case, i'll get a lot more additional cases and they are all far more likly to contribute to mortality, than say if i picked a football team to test, or if i pick same number from population at random. just thoughts, maybe wrong too, to highlight there are so many variables involved in the simple cases/death comparison. i dont think the ratio based on short window works on the longer time frame for these many reasons.

I know. It is totally weird. Umpteen different unknowns, many of which will be different in different countries, and yet the ratio of deaths to cases has got bigger across the world. I'm still scratching my head. If you look at smaller day intervals (, window, as you call it, i.e., every couple of days) the ratio of deaths to cases has stopped getting bigger in the last week, though. And that change is occurring evenly across he world. I paste below. So maybe it is better to look at changes in a shorter time window, and do that sequentially, rather than compare from the beginning of the pandemic to now. That, at least, fits with the general view that the rate of appearance of new cases is falling. It's all giving me a headache lol! Interesting the latest chatter from the government is that restriction easing is going to be extremely cautious. I think that's the right choice.

 

[golddene]

I know. It is totally weird. Umpteen different unknowns, many of which will be different in different countries, and yet the ratio of deaths to cases has got bigger across the world. I'm still scratching my head. If you look at smaller day intervals (, window, as you call it, i.e., every couple of days) the ratio of deaths to cases has stopped getting bigger in the last week, though. And that change is occurring evenly across he world. I paste below. So maybe it is better to look at changes in a shorter time window, and do that sequentially, rather than compare from the beginning of the pandemic to now. That, at least, fits with the general view that the rate of appearance of new cases is falling. It's all giving me a headache lol! Interesting the latest chatter from the government is that restriction easing is going to be extremely cautious. I think that's the right choice.

View attachment 123243

Would like to say thanks for all your hard work in analysing this b**stard virus H. I for one find your understanding and analysis very informative Top work H.

 

[Jim in the West]

I know. It is totally weird. Umpteen different unknowns, many of which will be different in different countries, and yet the ratio of deaths to cases has got bigger across the world. I'm still scratching my head. If you look at smaller day intervals (, window, as you call it, i.e., every couple of days) the ratio of deaths to cases has stopped getting bigger in the last week, though. And that change is occurring evenly across he world. I paste below. So maybe it is better to look at changes in a shorter time window, and do that sequentially, rather than compare from the beginning of the pandemic to now. That, at least, fits with the general view that the rate of appearance of new cases is falling. It's all giving me a headache lol! Interesting the latest chatter from the government is that restriction easing is going to be extremely cautious. I think that's the right choice.

View attachment 123243

This is fascinating. We seem to be bottom of the class (with Italy, France and Belgium)

 

[Harry Wilson's tackle]

Would like to say thanks for all your hard work in analysing this b**stard virus H. I for one find your understanding and analysis very informative Top work H.

Thanks. Apprectated.

 

[Harry Wilson's tackle]

This is fascinating. We seem to be bottom of the class (with Italy, France and Belgium)

On the face of it, yes. The main reason I think this may be the case is we all seem to have almost identical ratios, and the numbers are small and steady. But it may just be coincidence. And everone else seems to be getting closer to us, many from quite far away numbers back in April. The actual numbers of reported cases and deaths vary more than the ratios. I'm trying to focus on the direction of change of the ratios rather than their actual values, though, and not get too distracted. It's too easy to start seeing patterns that aren't there, invoking various prejudicial justifications for disregarding numbers that don't fit, or nit pick. All the best, mate.

 

[Swansman]

https://www.youtube.com/watch?v=bUCWcft6kao

This guy dont seem overly impressed by the Italian way of counting deaths. He also seem a bit tired of the lockdown.

Gotta love Italy sometimes, I'd like our parliament to be as fun.

 

[Billy the Fish]

Would like to say thanks for all your hard work in analysing this b**stard virus H. I for one find your understanding and analysis very informative Top work H.

I'll second that. Really enjoying the alternative analysis from [MENTION=1200]Harry Wilson's tackle[/MENTION].

 

[Harry Wilson's tackle]

I'll second that. Really enjoying the alternative analysis from [MENTION=1200]Harry Wilson's tackle[/MENTION].

Thanks

 

[oldboyroy]

Absolutely heartbreaking a 6 week old baby becomes UK youngest victim!! God bless RIP

 

[Machiavelli]

Thought this might be the best thread to flag this up, but there's a new bit of info on the local case updates, which indicates the rate of infections. B&H are doing comparatively well on this front. I'd imagine most of our cases are in care homes, but it does indicate that, despite having the first cases appear here, that the early test, track, trace, isolate policy yielded success. See:

https://coronavirus.data.gov.uk/#category=utlas&map=rate

 

[Bodian]

https://twitter.com/statstadium/status/1258794562505629696

(sorry - not sure how you link to a tweet image)

 

[Triggaaar]

If a particular person feels inclined to stand on high and piss all over this post, there won't be any replies from me because I have my 'ignore' umbrella up.

get over yourself

 

[WATFORD zero]

New ONS fatality figures out again today. First time I've noticed it, but the figures released from the ONS previously are still going up. I guess it's down to delays in reporting but the figure for 24th April has gone from 27,626 published last week to a corrected total of 28,030 this week and whilst not huge, is still fairly significant.

From last week's ONS figures I also reported that Care Home deaths went over 300 from 11th April, peaking at 415 on 15th April and only dropping back under 300 on 24th April. In this week's figures that has also been corrected to the 27th April (for under 300) and they were still over 280 a day on 29th April. So they do appear to have peaked but not as early and not dropping as significantly as it initially appeared last week.

I haven't seen the total daily fatality figures from the Government today, but I believe yesterday's was 32,065, a number that the ONS figures confirm was passed on 30th April.

Unfortunately, given the ONS figures released today I now believe yesterday's announced figure of 32,065 to be an actual fatality figure of > 39,624. This means that if there is any significant 'catch up' today as there has been previous Tuesdays, it will probably put us over 40,000.