Coronavirus

[maxjam]

6 hours ago, Albert said:

Yes, well spotted. That implies that it's the same order of magnitude, so the scale of the significance of the results will be similar as uncertainty tends with sqrt(n). 

You can believe what you want, but given the blunders you've made with grasping basic statistical principles...

Yes, because 11k is sufficient to judge safety of the drug. The point here is whether it's actually 100% protection across as a population of a few dozen million, which the sample size isn't sufficient to determine. 

11k trial is enough to judge the safety of the drug but 17k isn't enough to judge whether its 100% effective.  Okay well, I suggest you take that up with AstraZeneca, they are the ones putting out the information.

 

6 hours ago, Albert said:

Because 10k isn't a sustainable number for the disease. It has an infection fatality rate on the scale of 1%, while seasonal flus are well below 0.1%, implying that Covid-19 is at least 10 times more lethal than the worst seasonal flus, and usually far more than this. Even with a conservative estimate, just letting Covid-19 become endemic would be taking you 10-20k a year deaths for respiratory diseases, and magnifying that by at least a factor of 10 by letting it burn through the same population. That is, you'd be going from 10-20k deaths per year being the standard to it being over 100-200k deaths per year, once endemic. 

One of the issues a lot of people seem to brush over is that the deaths we're seeing with Covid-19 is without it being endemic, while seasonal influenza is. To understand the impact of such a disease being endemic, you have to consider the above. 

So no, it being at 10k per year just isn't realistic with it being endemic. Thankfully, it does appear to respond better to vaccination, at this time, than the flu, but this is largely as it is still one disease, with some variants. Allowing it to become like the flu, which is a collection of diseases with a myriad of variants, would lead to the same issues with vaccination, however. We're already seeing the beginnings of that, but thankfully we should, from here, have the tool so successfully prevent that.  

I don't care what the infection rates are, neither does 99% of the population imo.

If we've all been vaccinated and the number of deaths falls below annual flu deaths, my argument is that there is no reason not to reopen the economy and live normal lives again.

 

6 hours ago, Albert said:

...your article refutes the 50k flu deaths figure...

No it doesn't.

 

5 hours ago, Eddie said:

And still the 'flu comparisons come out in this asylum of a thread.

I'm not comparing covid to flu, I am arguing that after we've all be vaccinated if the annual deaths from covid fall below the annual deaths from flu the arguments to remain locked down is a weak one.

[Albert]

48 minutes ago, maxjam said:

11k trial is enough to judge the safety of the drug but 17k isn't enough to judge whether its 100% effective.  Okay well, I suggest you take that up with AstraZeneca, they are the ones putting out the information.

Yes, as the question for the 11k trial was whether there was a significant risk per person of adverse reaction, hence uncertainty of that finding tends with the square root of number of participants. 

In terms of the question of whether it's 100% effective, that can, paradoxically, only be judged by having a sample large enough to find cases where people with the vaccine have had it. If you're actually interested, we can discuss what actual confidence intervals we could calculate from their data. The fact that they didn't give them, however, shows they didn't see it as a finding of any real significance, and more a marketing point.

48 minutes ago, maxjam said:

I don't care what the infection rates are, neither does 99% of the population imo.

If we've all been vaccinated and the number of deaths falls below annual flu deaths, my argument is that there is no reason not to reopen the economy and live normal lives again.

My argument is the same, you just don't seem to actually understand the basics of what we're discussing, hence aren't really noticing that.

Number of deaths, etc, will fall below annual flu deaths if the reproduction number is kept below 1 from here. Allow it to rise will increase deaths again, which add up over a year. You'll note that most Covid deaths have come in waves so far, rather than the slow burn that the flu and pneumonia tend to do. That's the difference between pandemics and endemic diseases of course. 

You're also ignoring the risk of allowing more variants to develop, which is part of the process of the disease becoming endemic. The vaccines won't help if such is allowed to occur over the next few years. 

48 minutes ago, maxjam said:

No it doesn't.

...mate...

48 minutes ago, maxjam said:

I'm not comparing covid to flu, I am arguing that after we've all be vaccinated if the annual deaths from covid fall below the annual deaths from flu the arguments to remain locked down is a weak one.

Nobody is arguing that people wait so long that we'd actually know that the 'annual deaths' are below the flu. When people discuss a staged opening, we're looking on the order of months before the opening is in full swing, but maintaining an R number below 1. 

[maxjam]

8 minutes ago, Albert said:

Yes, as the question for the 11k trial was whether there was a significant risk per person of adverse reaction, hence uncertainty of that finding tends with the square root of number of participants. 

In terms of the question of whether it's 100% effective, that can, paradoxically, only be judged by having a sample large enough to find cases where people with the vaccine have had it. If you're actually interested, we can discuss what actual confidence intervals we could calculate from their data. The fact that they didn't give them, however, shows they didn't see it as a finding of any real significance, and more a marketing point.

...mate ...

 

9 minutes ago, Albert said:

My argument is the same, you just don't seem to actually understand the basics of what we're discussing, hence aren't really noticing that.

Number of deaths, etc, will fall below annual flu deaths if the reproduction number is kept below 1 from here. Allow it to rise will increase deaths again, which add up over a year. You'll note that most Covid deaths have come in waves so far, rather than the slow burn that the flu and pneumonia tend to do. That's the difference between pandemics and endemic diseases of course. 

I understand fully thank you very much ? and the discussion is now turning into one of semantics 

Assuming the number of deaths remain low - whether than by the current vaccine, booster jabs or whatever the argument to remain lockdown is a very weak one.  Given that vaccines also reduce transmission this will help keep the R rate low and allow us to return to normal.

 

18 minutes ago, Albert said:

...mate...

???

As you 'just don't seem to be able to actually understand the basics' of the link here is another one, although you could have just googled it yourself;

https://fullfact.org/health/flu-pneumonia-death-years/

 

20 minutes ago, Albert said:

Nobody is arguing that people wait so long that we'd actually know that the 'annual deaths' are below the flu. When people discuss a staged opening, we're looking on the order of months before the opening is in full swing, but maintaining an R number below 1. 

Good, well lets leave it here then.  The R rate in the UK is well under 1 now and all of the over 50s and vulnerable will have been vaccinated by April/May.  There will be no reason not to reopen over the warmer months whilst the rest of the population continues to get their jabs.  This will naturally help keep the R rate lower than pre-vaccination and I tbh don't really care if the R rate climbs again assuming the jab keeps people from developing a severe case of covid. 

[BIllyD]

40 minutes ago, Albert said:

In terms of the question of whether it's 100% effective, that can, paradoxically, only be judged by having a sample large enough to find cases where people with the vaccine have had it. If you're actually interested, we can discuss what actual confidence intervals we could calculate from their data. The fact that they didn't give them, however, shows they didn't see it as a finding of any real significance, and more a marketing point.

Ooo something I can understand working in Internal Audit we have to work on sample sizes on a regular basis. What are you saying here Albert, only two outcomes so you believe this confidence level to be ?

[Albert]

27 minutes ago, maxjam said:

...mate ...

Excellent retort to the point. I know you're making a play on my response to your blunder with that article before, but anyone who bothered to read it, yourself included, would see pretty quickly that yes, it refutes your claim in multiple ways. 

27 minutes ago, maxjam said:

I understand fully thank you very much ? and the discussion is now turning into one of semantics 

Assuming the number of deaths remain low - whether than by the current vaccine, booster jabs or whatever the argument to remain lockdown is a very weak one.  Given that vaccines also reduce transmission this will help keep the R rate low and allow us to return to normal.

...you've had this all explained to you already, and yet you persist with this line of reasoning. 'Assuming the number of deaths remain low', we're still getting over 500 deaths per day. It's not low in the slightest yet. This is the problem with your little fantasy argument here, the deaths aren't low yet. 

What I was arguing was that reopening should be done in a controlled manner with an eye on that R number, you seem to agree at some level, but don't want to admit it on those terms. 

27 minutes ago, maxjam said:

???

As you 'just don't seem to be able to actually understand the basics' of the link here is another one, although you could have just googled it yourself;

https://fullfact.org/health/flu-pneumonia-death-years/

I'm not sure if the issue is that you don't bother reading sources you post, or if you think simply linking things shapes them to your argument. Your current choice of source demonstrates that 50k deaths in a year from flu and pneumonia is not normal in the UK, and no year has been over 25k in the last two decades. 

The great irony is that you're using this an attempt to make it seem like people accept the numbers of deaths from the flu and pneumonia, but the reason for those drops is that people didn't. In effect, you've destroyed one of the pillars of your own argument, twice, with your choices of sources, yet don't seem to have the self-awareness to realise it. 

27 minutes ago, maxjam said:

Good, well lets leave it here then.  The R rate in the UK is well under 1 now and all of the over 50s and vulnerable will have been vaccinated by April/May.  There will be no reason not to reopen over the warmer months whilst the rest of the population continues to get their jabs.  This will naturally help keep the R rate lower than pre-vaccination and I tbh don't really care if the R rate climbs again assuming the jab keeps people from developing a severe case of covid. 

Reopening has been a sliding scale in countries with a competent response. It's not just 'oh look, under 1, throw the doors open', but rather, a step by step process assuring that things don't get out of hand. The UK should follow the same, particularly given the rise of variants with the potential to render the vaccine less effective. 

9 minutes ago, BIllyD said:

Ooo something I can understand working in Internal Audit we have to work on sample sizes on a regular basis. What are you saying here Albert, only two outcomes so you believe this confidence level to be ?

The issue here is that while the sample size is large, the claim is 100% effective. We cannot evaluate that claim directly with that amount of data, as determining the uncertainty from it would require that there actually was a serious case or death. Let's say that the actual chance of serious illness is 1 in 20k with the sample size of 20k. The probability that nobody in this sample got seriously ill would be around 37%, which is still quite likely. 1 in 20k doesn't sound like a lot, but the UK was at one point getting ~70k new cases per day, even with restrictions. Opening up on the assumption such protection is enough would lead to 3-4 seriously ill patients per day on those figures, which doesn't sound like a lot, but would represent over 1200 per year. That's the sample without taking into account the subset of the vulnerable that cannot be taken into account, which is the key worry. 

That said, 1200 per year when compared to flu and pneumonia deaths most years would actually represent a small number, unlike what Covid had shown before. The other concern with it all those is the question of whether allowing the disease spread with less severe symptoms would lead to further variants, particularly given we know some aren't impacted as much by the vaccine. That, in effect, is why keeping the reproduction number below 1 is so key in all of this. If the vaccines are effective, it'll happen without much issue and the reopening will happen quite rapidly. Doing so in that manner, however, provides a safeguard in case things don't go so well. 

It's interesting to me that the very people who were throwing cold water on the prospect of a vaccine only a few months ago, on virtually the same data we have now, are the same ones who are so keen to throw all their faith behind it, without those very same questions being answered. It's almost as though there exist some on here that are just arguing to demand the restrictions end, and don't actually care about the impacts on lives and livelihoods as long as that happens. 

[SSD]

Watched a report last night on how a number of people in EU countries are now starting to question whether being in the EU is the right thing because of the slow speed of the vaccine across these nations. Don't want to turn this into a political conversation by all means, thought it was interesting how people in the EU seem to be looking to the UK as a benchmark for how quickly we are vaccinating the vulnerable groups. For once it feels like Europe doesn't actually hate us. Granted, we've paid more per vaccine than the EU, however we seemed to have planned months before they got their procurement sorted out. 

From feedback I've heard from family, the organisation of the vaccines centres is superb. No hassle, everyone is helpful and eases people's worries. I'm right at the bottom of the list and I've already had covid so I doubt I'll be getting my jab anytime soon. Even though we've lost far too many loved ones and there is multiple reasons why we've let this virus take control and killed 1000s, at least we're now trying to exit this horrible road as quickly as possible. 

[BIllyD]

13 minutes ago, Albert said:

The issue here is that while the sample size is large, the claim is 100% effective. We cannot evaluate that claim directly with that amount of data, as determining the uncertainty from it would require that there actually was a serious case or death. Let's say that the actual chance of serious illness is 1 in 20k with the sample size of 20k. The probability that nobody in this sample got seriously ill would be around 37%, which is still quite likely

For this review a sample size of 17k was taken, this is quite high and based on a 60m population would be considered 99% confidence with only a 1% error level. With this review stating a 100% confidence level would be the norm, based upon nobody getting seriously ill or death, your just adding extra probabilities into the equation that are not there. 

Are you  just picking 1 in 20k, or was that something that came back in the report ?
 

 

[maxjam]

6 minutes ago, Albert said:

Excellent retort to the point. I know you're making a play on my response to your blunder with that article before, but anyone who bothered to read it, yourself included, would see pretty quickly that yes, it refutes your claim in multiple ways. 

Yup, I treated your post with the same disdain as you treated mine.

And the link showed exactly what I wanted it to show, anyone that bothered to read it, yourself included, would quickly see that.

 

7 minutes ago, Albert said:

I'm not sure if the issue is that you don't bother reading sources you post, or if you think simply linking things shapes them to your argument. Your current choice of source demonstrates that 50k deaths in a year from flu and pneumonia is not normal in the UK, and no year has been over 25k in the last two decades. 

When have I ever said that 50k deaths is normal?  My actual statement was along the lines of the UK experiences 10k-50k annual flu deaths which is exactly what the graph shows (to be more precise its 15k-45k but there ya go). 

I don't even care about the upper limit, fixating on that is deflecting from the point I'm making in that assuming we get annual covid deaths down to below annual flu deaths - and I've quoted 10k previously as a ballpark figure - the argument to stay locked down is weak.

 

16 minutes ago, Albert said:

The great irony is that you're using this an attempt to make it seem like people accept the numbers of deaths from the flu and pneumonia, but the reason for those drops is that people didn't. In effect, you've destroyed one of the pillars of your own argument, twice, with your choices of sources, yet don't seem to have the self-awareness to realise it. 

Any different to dying with or from covid?  

 

Okay, this will be my last reponse to you *awaits snide comment*.  I have made my argument and stand by what I have said, you obviously disagree and won't let it drop - we're both just repeating ourselves now. 

I cba with the snide comments that tbh I don't receive from anyone else, I had a decent chat last weekend I think it was with someone else that was perfectly civil even though we agreed to disagree on some aspects.  Its just not worth getting into a conversation with you that ends up semantically going round and around in circles achieving nothing. Enjoy your weekend.

[Albert]

25 minutes ago, BIllyD said:

For this review a sample size of 17k was taken, this is quite high and based on a 60m population would be considered 99% confidence with only a 1% error level. With this review stating a 100% confidence level would be the norm, based upon nobody getting seriously ill or death, your just adding extra probabilities into the equation that are not there. 

Are you  just picking 1 in 20k, or was that something that came back in the report ?

20k was picked as rounding to the nearest 10k, as we're doing napkin statistics. 

As to your claim that it's "99% confidence" on the sample, I'd be curious to know your methodology given the information given. As noted, an rate of 1 in 20k serious disease in a sample of 20k has about a 37% chance of showing no cases. 

Edit: Also, they didn't claim 100% confidence for no deaths. You're mixing terms here. 

14 minutes ago, maxjam said:

Yup, I treated your post with the same disdain as you treated mine.

And the link showed exactly what I wanted it to show, anyone that bothered to read it, yourself included, would quickly see that.

I read it, it disagreed with your claim. You can't pretend otherwise, the information is right there. 

Quote

When have I ever said that 50k deaths is normal?  My actual statement was along the lines of the UK experiences 10k-50k annual flu deaths which is exactly what the graph shows (to be more precise its 15k-45k but there ya go). 

I don't even care about the upper limit, fixating on that is deflecting from the point I'm making in that assuming we get annual covid deaths down to below annual flu deaths - and I've quoted 10k previously as a ballpark figure - the argument to stay locked down is weak.

The graph showed very clearly that it's not 10k-50k annual flu deaths. It's about 20-25k annual flu and pneumonia deaths, with 40k+ not being seen in the last 2 decades. 

I like how now questioning your claims is 'deflecting from the point'. 

As noted, your point about 'under 10k deaths' is silly on a number of levels, and you failed to respond to any of the points raised. 

Quote

Any different to dying with or from covid?  

People didn't accept the high number of deaths from respiratory deaths, and changes were made to bring those down. These were endemic of course, so quite different to a pandemic scenario, for the reasons already discussed. 

Quote

Okay, this will be my last reponse to you *awaits snide comment*.  I have made my argument and stand by what I have said, you obviously disagree and won't let it drop - we're both just repeating ourselves now. 

I cba with the snide comments that tbh I don't receive from anyone else, I had a decent chat last weekend I think it was with someone else that was perfectly civil even though we agreed to disagree on some aspects.  Its just not worth getting into a conversation with you that ends up semantically going round and around in circles achieving nothing. Enjoy your weekend.

You haven't made an argument at all, you've thrown around a few uncited figures, which when you've attempted to cite them have actually linked to sources which disagree with those very claims. You can stand by your 'argument', but there's not a lot to really stand by. 

[BIllyD]

6 minutes ago, Albert said:

20k was picked as rounding to the nearest 10k, as we're doing napkin statistics. 

As to your claim that it's "99% confidence" on the sample, I'd be curious to know your methodology given the information given. As noted, an rate of 1 in 20k serious disease in a sample of 20k has about a 37% chance of showing no cases. 

Edit: Also, they didn't claim 100% confidence for no deaths. You're mixing terms here. 

I'm not quite sure what you are saying, the report I believe was a survey of 17k of which the results showed none of these ended up in hospital ? You added the 1 in 20k I believe, or is that from the report ?
 

I must admit not looking into any detail just skimmed read through the report and the methodology is based upon standard statistical when determining a sample size.

[maxjam]

24 minutes ago, maxjam said:

Okay, this will be my last reponse to you *awaits snide comment*.  I have made my argument and stand by what I have said, you obviously disagree and won't let it drop - we're both just repeating ourselves now. 

 

10 minutes ago, Albert said:

You haven't made an argument at all (I HAVE), you've thrown around a few uncited figures (I HAVEN'T), which when you've attempted to cite them have actually linked to sources which disagree with those very claims (THEY DON'T). You can stand by your 'argument', but there's not a lot to really stand by (YOUR OPINION). 

Haha.

I can't really say what I'd like to say on this forum otherwise I'd be enjoying a nice relaxing break from contributing.  So lets simply go with enjoy the rest of your weekend ?

[Albert]

7 minutes ago, BIllyD said:

I'm not quite sure what you are saying, the report I believe was a survey of 17k of which the results showed none of these ended up in hospital ? You added the 1 in 20k I believe, or is that from the report ?

I rounded 17k up to 20k for the calculations. 

7 minutes ago, BIllyD said:

I must admit not looking into any detail just skimmed read through the report and the methodology is based upon standard statistical when determining a sample size.

They didn't mention the "100% protection against severe disease and death" in their report, it was just headline marketing. It's not even a methodology thing, it's just someone trying to advertise the results by highlighting that nobody in the trial group got seriously ill or died. As noted, that's not really enough to determine the actual rate of severe disease or death for those with this vaccine for the reasons stated. 

[Archied]

22 minutes ago, Albert said:

20k was picked as rounding to the nearest 10k, as we're doing napkin statistics. 

As to your claim that it's "99% confidence" on the sample, I'd be curious to know your methodology given the information given. As noted, an rate of 1 in 20k serious disease in a sample of 20k has about a 37% chance of showing no cases. 

Edit: Also, they didn't claim 100% confidence for no deaths. You're mixing terms here. 

I read it, it disagreed with your claim. You can't pretend otherwise, the information is right there. 

The graph showed very clearly that it's not 10k-50k annual flu deaths. It's about 20-25k annual flu and pneumonia deaths, with 40k+ not being seen in the last 2 decades. 

I like how now questioning your claims is 'deflecting from the point'. 

As noted, your point about 'under 10k deaths' is silly on a number of levels, and you failed to respond to any of the points raised. 

People didn't accept the high number of deaths from respiratory deaths, and changes were made to bring those down. These were endemic of course, so quite different to a pandemic scenario, for the reasons already discussed. 

You haven't made an argument at all, you've thrown around a few uncited figures, which when you've attempted to cite them have actually linked to sources which disagree with those very claims. You can stand by your 'argument', but there's not a lot to really stand by. 

Is there anything you are not an expert / authority on??‍♂️

[Albert]

7 minutes ago, maxjam said:

 

Haha.

I can't really say what I'd like to say on this forum otherwise I'd be enjoying a nice relaxing break from contributing.  So lets simply go with enjoy the rest of your weekend ?

To reiterate, you did indeed blunder by posting links to sources that actually refuted your point. You have since tried to deflect from this by arguing that even talking about your figures is 'deflecting'. You've made zero attempt to refute the other points made; that's not an opinion, you've just not done so. 

[BIllyD]

5 minutes ago, Albert said:

I rounded 17k up to 20k for the calculations. 

They didn't mention the "100% protection against severe disease and death" in their report, it was just headline marketing. It's not even a methodology thing, it's just someone trying to advertise the results by highlighting that nobody in the trial group got seriously ill or died. As noted, that's not really enough to determine the actual rate of severe disease or death for those with this vaccine for the reasons stated. 

Sorry to reiterate then, it's the content within the report your contesting. As you stated 17k was not enough of a sample to give the confidence level ?

Which was one it ?

[maxjam]

7 minutes ago, Albert said:

To reiterate, you did indeed blunder by posting links to sources that actually refuted your point. You have since tried to deflect from this by arguing that even talking about your figures is 'deflecting'. You've made zero attempt to refute the other points made; that's not an opinion, you've just not done so. 

Haha, to reiterate, enjoy the rest of your weekend ?

FFS.

[Albert]

13 minutes ago, BIllyD said:

Sorry to reiterate then, it's the content within the report your contesting. As you stated 17k was not enough of a sample to give the confidence level ?

Which was one it ?

They didn't give a confidence level for that claim, as it wasn't a claim in the body of the report. 

[BIllyD]

2 minutes ago, Albert said:

They didn't give a confidence level for that claim, as it wasn't a claim in the body of the report. 

You stated earlier to @maxjamthat 17k was not enough of a sample for the UK population. I'm confused now, are you saying that it is or isn't ?‍♂️
 

[Albert]

26 minutes ago, BIllyD said:

You stated earlier to @maxjamthat 17k was not enough of a sample for the UK population. I'm confused now, are you saying that it is or isn't ?‍♂️
 

I'm not quite sure what you think the contradiction here is. 17k is too small a sample to evaluate the claim that a vaccine offers 100% protection against severe disease, but it is more than sufficient to determine safety, for the reasons previously discussed. 

[BIllyD]

7 minutes ago, Albert said:

I'm not quite sure what you think the contradiction here is. 17k is too small a sample to evaluate the claim that a vaccine offers 100% protection against severe disease, but it is more than sufficient to determine safety, for the reasons previously discussed. 

The sample size is not to small though if the results show an outcome, that is not a reason to dispute the findings. 
 

No wonder maxjam was confused with what your trying to say.