Number of Deaths

[KZNSeagull]

Thank you, so despite many protestations from a single poster, we were correct in assuming that the number would increase but some of the deaths, albeit extremely sad and regrettable as any death is, would be within the Govts numbers.

Yes and importantly we can now see how serious this outbreak actually is in that it has increased the 5 year weekly death rate average by over 50%.

 

[dazzer6666]

Unfortunately it is the not until very much later bit that bothers me.

I understand there is little that can be done at the moment, as there is no vaccine, but I do worry that those being asked to "shield" will eventually be forgotten. Those bleating on about being in lockdown for 3 weeks really should try this shielding lark!

No chance of getting out for a once a day exercise, that's against the rules and yet some people just moan on about it...

Our household isn't probably one that the government had thought about when they asked people to shield, I'm guessing they had in mind old people, either on their own, or in a couple. We though are a family of 5, a 60 year old (me), my 59 year old wife, who has Lymphoma and is very vulnerable. We also have 2 teenage grandchildren living at home and a 3 year old.

The government advice is that all, bar my wife, can go outdoors and follow "stringently" the social distancing and hygiene rules. Unfortunately that is the problem in our household, the 2 teenage girls both have learning difficulties and as such do not have a clue about personal hygiene and wouldn't understand the concept of social distancing! Consequently we have had to take the decision to ALL follow the shielding process.

This is fine at the moment, where the government is being quite understanding and flexible with regards giving people advice to stay indoors unless you really have to go out, but at some point that attitude will change and the expectation would be for the elder one to return to work, or stop claiming Universal Credit and the same for the younger teenager, as she has just finished 6th form college, so should be looking for work! Our 3 year old is also due to start school in September - if they are back.

Basically, until this is sorted out there is no way that any of us are going out of our house.

I couldn't forgive myself if either me, or any of the 3 children in our house brought the virus back and signed my wife's death warrant...!

I really don't know where or how this is going to end for us, and I really do fear that 1.5 million most vulnerable people are going to be largely forgotten, or expected, as I have said, to sign their own death warrants!

Life clearly is always about risk management, but the risks have never been so high...

I feel for you - that's sounds incredibly difficult. Part of the problem is defining rules and criteria that suit all cases - which is pretty much impossible so first focus will naturally be on trying to deal with large chunks of the population. As lockdown/shielding eases (and it will) more and more sets of circumstances will be captured - conscious this won't necessarily help you though and there will be a degree of risk in whatever form of easing of lockdown/shielding applies. Hope you're getting the support you need.

 

[Harry Wilson's tackle]

Thanks. Yes, I took that number of 8 and worked out up to 5 million deaths based on it and a 60% infection rate but the very worst case in government projections was 250k so hopefully that combination of increased testing and better reporting will get that number down.

It also teaches me not to trust the Chinese - their number, early on, was around 1 in 100 which is still a fair few if infection goes uncontrolled but, with controls, isn't (comparably). As someone who tended to look on the bright side (note past tense) that was the reason for my initial "not sure what all the fuss is about" stance. Italy obviously changed that. Other data at that time was also dissimilar to our situation as you had South Korea who controlled the whole thing perfectly and Iran who are also less than open but now seem to be less affected than the UK.

One in eight would be horrific. But I'm still hoping that control and testing will bring that the correct way. Accurate antibody testing is really the only way to know for sure though?

Yes. Antibody testing. But unless we can test huge numbers of people who have never been ill, and can show that a huge proportion of them have been infected even though they never expressed symptoms, then we won't know what the true mortality rate is. The true mortality rate is the true number of people who have been infected (regardless of whether the got ill) divided by the true number that die of it (expressed that way, the bigger the number, the less lethal).

As I said, the present number for the UK, France, Spain and Italy is around 8 - which is horrifying. We are all hoping the true number is much bigger - ideally over 100, and the number 8 reflects the fact we are not testing enough people who have not been ill to find out the rate of symptomless infection. As I said I will keep my eye on the numbers published by the BBC and in a week's time will be able to comment on things again based on how the numbers change. Give me a poke if I forget

 

[pb21]

This will be a potential useful, albeit grim, way of tracking this kind of data:

[tweet]1250007002165755905[/tweet]

 

[Guinness Boy]

Yes. Antibody testing. But unless we can test huge numbers of people who have never been ill, and can show that a huge proportion of them have been infected even though they never expressed symptoms, then we won't know what the true mortality rate is. The true mortality rate is the true number of people who have been infected (regardless of whether the got ill) divided by the true number that die of it (expressed that way, the bigger the number, the less lethal).

As I said, the present number for the UK, France, Spain and Italy is around 8 - which is horrifying. We are all hoping the true number is much bigger - ideally over 100, and the number 8 reflects the fact we are not testing enough people who have not been ill to find out the rate of symptomless infection. As I said I will keep my eye on the numbers published by the BBC and in a week's time will be able to comment on things again based on how the numbers change. Give me a poke if I forget

Will do squire.

Presumably we also need to test those who have had symptoms but not bad enough to go to hospital to get that rate too though, right? There have been a few NSC users who have reported symptoms but not been tested so this is another cohort. I'd love to know if the cough I've had for two weeks that will not eff off and gets better or worse randomly is "it". Given I've not had a temperature, am eating like a horse and can work I'm assuming not but it's the weirdest damn bug I've had in a while. I'd like to know.

Regards from self-isolation
GB

 

[darkwolf666]

I feel for you - that's sounds incredibly difficult. Part of the problem is defining rules and criteria that suit all cases - which is pretty much impossible so first focus will naturally be on trying to deal with large chunks of the population. As lockdown/shielding eases (and it will) more and more sets of circumstances will be captured - conscious this won't necessarily help you though and there will be a degree of risk in whatever form of easing of lockdown/shielding applies. Hope you're getting the support you need.

Thank you for your reply - at the moment we are getting all the support we need - after nearly 3 weeks of not being able to get an online food delivery, I had emails from Asda, Iceland, Sainsbury's and Tesco, all offering us priority slots for deliveries of home shopping, help, it appears comes along like buses. Fortunately, before the lockdown I had the foresight to book a recurring slot with Asda, so we are OK for food now!

As for the other stuff - like I said our teenage girls (19 & 17) have learning difficulties - but we can't keep them imprisoned forever, although it wouldn't actually bother them as they are happy in their bedrooms, we just know it's wrong! But what can we do?

Like I said, I really don't know how this is going to play out - but I am quite confident that we will be where we are until September - and likely longer!

 

[dazzer6666]

Will do squire.

Presumably we also need to test those who have had symptoms but not bad enough to go to hospital to get that rate too though, right? There have been a few NSC users who have reported symptoms but not been tested so this is another cohort. I'd love to know if the cough I've had for two weeks that will not eff off and gets better or worse randomly is "it". Given I've not had a temperature, am eating like a horse and can work I'm assuming not but it's the weirdest damn bug I've had in a while. I'd like to know.

Regards from self-isolation
GB

Yes - we need a reliable, simple and fast antibody test (there isn't one yet) and then blitz vast numbers of the population asap (contract Amazon to deliver them to every household etc)...........far more important now for future strategy than testing people that 'have it' is finding out who has 'had it' (assuming prior infection and recovery confirms immunity, and no ability to carry/spread). Various bits of research around the globe at the moment suggesting numbers are miles higher than antigen ('have it now') testing confirms - this has to be the case in the UK where our testing is limited to those hospitalised with severe symptoms. If the c40% numbers that I've seen in some reports are true then we might have had 25 million+ people infected already.

 

[Harry Wilson's tackle]

Will do squire.

Presumably we also need to test those who have had symptoms but not bad enough to go to hospital to get that rate too though, right? There have been a few NSC users who have reported symptoms but not been tested so this is another cohort. I'd love to know if the cough I've had for two weeks that will not eff off and gets better or worse randomly is "it". Given I've not had a temperature, am eating like a horse and can work I'm assuming not but it's the weirdest damn bug I've had in a while. I'd like to know.

Regards from self-isolation
GB

Yes. In order to know what the true mortality rate is, we need to test people regardless of whether they have been ill, think they have been ill, have been diagnosed as ill down the phone by a nurse practitioner, or have been tested before and found to have actually been ill. We need a massive and properly random sample of the population to get a clear picture. Randomized testing. Randomization is how we do research. If it isn't randomized it is biased and will tell you nothing useful.

And separately we need to test bloody everyone once we know how lethal this virus is so we can protect those at risk. If it is as lethal as the raw UK figures suggest, those currently unexposed (no antibodies) need to stay isolated until there is a vaccine.

 

[KZNSeagull]

This will be a potential useful, albeit grim, way of tracking this kind of data:

[tweet]1250007002165755905[/tweet]

I would say very useful as it provides context.

 

[Weststander]

Belgium's under the media radar, where the death toll is horrendous, already over 4,000 deaths in a small country of just 11.5m. Per capita just behind Spain, with the equivalent of over 1,000 new deaths announced today.

 

[Uter]

This will be a potential useful, albeit grim, way of tracking this kind of data:

[tweet]1250007002165755905[/tweet]

Fascinating. I never knew there was such a big variability from year to year. Presumably seasonal flu is the main driver of the big variations in Jan/Feb.

 

[Green Cross Code Man]

Fascinating. I never knew there was such a big variability from year to year. Presumably seasonal flu is the main driver of the big variations in Jan/Feb.

Another factor is how cold the weather is in a particular season.

 

[rogersix]

Yes. In order to know what the true mortality rate is, we need to test people regardless of whether they have been ill, think they have been ill, have been diagnosed as ill down the phone by a nurse practitioner, or have been tested before and found to have actually been ill. We need a massive and properly random sample of the population to get a clear picture. Randomized testing. Randomization is how we do research. If it isn't randomized it is biased and will tell you nothing useful.

And separately we need to test bloody everyone once we know how lethal this virus is so we can protect those at risk. If it is as lethal as the raw UK figures suggest, those currently unexposed (no antibodies) need to stay isolated until there is a vaccine.

Obviously everyone being tested is best, but what percentage of the population would be statistically significant Mr. Tackle?

 

[RossyG]

Fingers crossed we're on the plateau.

 

[Harry Wilson's tackle]

Obviously everyone being tested is best, but what percentage of the population would be statistically significant Mr. Tackle?

Not a simple question to answer. We actually need to estimate not only the national risk of death, but more importantly the risk-stratified death, in other words the the extent to which risk is increase and decreased (it has to be both) by age and comorbidity (what other health issues you carry) and other factors such as social class. Unfortunately to do group size estimation you need to know a lot about the data itself. This is why I am an opponent of reliance solely on power analysis as the arbiter of group size - most of the input data are guesses and the output is the minimum size to get enough power to obtain 'significance' if everything is correct and goes to plan (a mahorreason why so much published research is false).

Failing that, the best way is to make repeated estimates based on emerging data and look how the power calculated group size changes over time - once it is stable by whatever definition then you add another 50% to be safe and bob's your uncle - you can set your testing budget and timescales.

But the data has to be a random sample of the population to give a national picture. That said, because having data which is 50% derived from London may well give a completely different outcome versus data derived entirely from Royston Vasey, for example,what I would do is simply 'bust the bank' to test like crazy, and sift the data retrospectively, daily, grouping populations by age, health, proximity and social class (we have seen how some chavs behave in the social distancing disco). If it is correct that for whatever reason the folk of the Faroes don't die when infected whereas other folk drop like flies it really is important to know this, find out how the chilly islanders do it, bottle it and dispense it, so to speak.

 

[Deportivo Seagull]

@HWT. Any thoughts on how obesity is affecting mortality rates among Covid 19 cases? Not very scientific (actually not scientific at all) but a large number of the younger individuals who have died with no apparent underlying health issues appear from photos to be on the large size?


Sent from my iPad using Tapatalk

 

[Kinky Gerbil]

@HWT. Any thoughts on how obesity is affecting mortality rates among Covid 19 cases? Not very scientific (actually not scientific at all) but a large number of the younger individuals who have died with no apparent underlying health issues appear from photos to be on the large size?


Sent from my iPad using Tapatalk

You would assume if they are larger they would potentially have some additional stress on the heart / lungs- depends how large is larger...

 

[RossyG]

@HWT. Any thoughts on how obesity is affecting mortality rates among Covid 19 cases? Not very scientific (actually not scientific at all) but a large number of the younger individuals who have died with no apparent underlying health issues appear from photos to be on the large size?

I haven't got any links, I'm afraid, but I have read that obesity and smoking are big factors.

 

[Harry Wilson's tackle]

I just checked the BBC pages and they are reporting the numbers of cases and deaths in the main table for April 9, values with are of course lower than the ones they were showing yesterday and which I posted on here. Don't know what that's all about. And I did click 'refresh'.

 

[rogersix]

Not a simple question to answer. We actually need to estimate not only the national risk of death, but more importantly the risk-stratified death, in other words the the extent to which risk is increase and decreased (it has to be both) by age and comorbidity (what other health issues you carry) and other factors such as social class. Unfortunately to do group size estimation you need to know a lot about the data itself. This is why I am an opponent of reliance solely on power analysis as the arbiter of group size - most of the input data are guesses and the output is the minimum size to get enough power to obtain 'significance' if everything is correct and goes to plan (a mahorreason why so much published research is false).

Failing that, the best way is to make repeated estimates based on emerging data and look how the power calculated group size changes over time - once it is stable by whatever definition then you add another 50% to be safe and bob's your uncle - you can set your testing budget and timescales.

But the data has to be a random sample of the population to give a national picture. That said, because having data which is 50% derived from London may well give a completely different outcome versus data derived entirely from Royston Vasey, for example,what I would do is simply 'bust the bank' to test like crazy, and sift the data retrospectively, daily, grouping populations by age, health, proximity and social class (we have seen how some chavs behave in the social distancing disco). If it is correct that for whatever reason the folk of the Faroes don't die when infected whereas other folk drop like flies it really is important to know this, find out how the chilly islanders do it, bottle it and dispense it, so to speak.

And there's me hoping that you would say 20%.
I got stats o level once, it's not much good here is it?