Fascinating Story in the FT

[Machiavelli]

Prof Neil Ferguson has also expressed he's sceptical of this study to the HoCs Science Select Committee

 

[Deportivo Seagull]

Prof Neil Ferguson has also expressed he's sceptical of this study to the HoCs Science Select Committee

It’s not been peer reviewed nor has the dataset and i so would hope so.


Sent from my iPad using Tapatalk

 

[Kalimantan Gull]

It would be great news, we can hope.

On the 8,000 out of 30,000, remember those tests are for virus present in the body. Some of those 30,000 may have tested negative in the same way that someone who has recovered tests negative. We need the antibody tests to be sure.

I don't know nearly enough about anything to comment on this meaningfully, but a couple of thought experiments (to assist me in my strong desire to be optimistic).

Why the number of deaths now if it's been around for a while? Well if we accept the proposition that infact the virus kills many many fewer of the people it infects, and we would probably have to accept the proposition that it is far more contagious, then it isn't that hard to understand. For example, at the time we currently assume there were roughly 100 cases, and there was 1 death, perhaps there were infact 1,000 cases. Today, we likely assume that there are (let's say) 100,000 cases in the UK, and we are looking at approx 1,000 deaths (1%). Perhaps there are 1,000,000 cases. My point is (if I am thinking about this correctly), the number of deaths now does not contradict the idea presented in the article. The problem would have been that the first noticable deaths would have taken time to show up, because under this model, even when we had, let's say 10,000 cases, there would have been only 10 additional deaths from pneumonia around the country. That could easily be missed, especially if they are mostly in people with existing conditions or older people. Maybe 1 out of the 10 did not conform, nobody would notice that.

I'm not sure if I'm making sense here (or if I am even being logical) but I (think I) can see how you could have the death rate and timescale we have had, and the suggestion of many more cases in the article could still be correct.

I'm not sure about 50% of the population, because that would mean what, 32,000,000 cases, which seems huge. But I would imagine that it's possible that we have 1,000,000 cases, and if we did, that would reduce the death rate considerably (0.1%).

Also, there are a fairly large proportion of people who (today) get sick enough to be hospitalized, but don't require critical care. I'm not sure exactly what their condition is, but it might be that they are hospitalized in part because they are at risk of becoming critical, and in part because they need seperating from the general population. I'm sure I've heard of patients in hospital having pretty mild symptoms (I'm not clear on what it means to be severe but not critical tbh), but still being admitted once they "qualify" for a diagnosis (persistent symptoms for 7+ days). Those cases would not have been picked up before we started testing, they would likely have just been very poorly at home, and only a small proportion of them would have progressed to hospitalization.

Not sure if all that makes sense. I'm tired and stressed (I have no idea how I am going to get through 12 weeks or more of this tbh) Anyway, I'm sure we would all be very releaved to discover that this virus is more common and less deadly than we thought.

Also, one additional point, this thing didn't start in January. It started in November. If you are asking yourself how it could be possible that there are so many more cases than we have been assuming, I have a very reasonable question for you. How could it be possible that cases only started in the UK at the end of January? On Jan 31st there were just under 10,000 confirmed cases in China. Those were the confirmed cases, so we can assume there were more than that, maybe it would even be reasonable to assume 50,000 if 4/5 are mild cases. With all the international travel between China and Europe, nobody had shipped it out already? How much travel do you imagine there was between China and Europe over Xmas and New Year? Those are peak times for international travel I would have thought. And those numbers would be the numbers to use for this question if the rates are what we have been thinking they are and the FT article is wrong. It wouldn't make any sense to me that we had our first case at the end of January.

Food for thought(?)

There is a lot of food for thought in what you write. But even if 10,000 infected only corresponds to 10 deaths (and another 50-100 in hospital?) - don't you think that pneumonia deaths and illnesses like that, in February, with the coronavirus in the news, you think no doctor might have put two-and-two together? Certainly worth a test one would think. Seems highly unlikely. Secondly, if 50% of the population have had it, at our current raw mortality numbers - well it can't get any bigger in terms of numbers of people, this must be the peak. And yet we see from Italy and Spain that we can accelerate what we are seeing by a factor of 10 at least.

Beyond that, this kind of modelling would have already been done by the Chinese, they would be perfectly capable of doing it. And if it could spread so widely so easily without detection, the mortality numbers we saw in Hubei would have been repeated all over their country. But as it happens the Wuhan lockdown has worked in China. I think the fact that that lockdown has worked and prevented the spread to other regions of China is the biggest argument against the conclusion of this study.

Finally, the Diamond Princess cruise ship was a petri-dish experiment, to a certain degree. 10 deaths out of 712 confirmed cases out of 3711 passengers and crew. If you agree that all the people could have been exposed and got it asymptomatically, then 10 out of 3711 is a low number (although I think the captain would be pretty shocked if he lost 10 passengers to flu on the same cruise). But everyone was tested, regularly, so it seems far more likely that the quarantine procedures meant that only about 20% got it. 10/712 is 1.5% mortality rate, and 115 people are apparently still sick, 15 seriously so.

 

[Mellotron]

There is a lot of food for thought in what you write. But even if 10,000 infected only corresponds to 10 deaths (and another 50-100 in hospital?) - don't you think that pneumonia deaths and illnesses like that, in February, with the coronavirus in the news, you think no doctor might have put two-and-two together? Certainly worth a test one would think. Seems highly unlikely.

I feel like everyone keeps ignoring the article from January - totally unrelated to CV at the time - confirming flu cases in the UK being EIGHT times higher than normal.

Why isn't this far bigger news?!

 

[Kalimantan Gull]

I feel like everyone keeps ignoring the article from January - totally unrelated to CV at the time - confirming flu cases in the UK being EIGHT times higher than normal.

Why isn't this far bigger news?!

have you got the article? I just had a look at the latest National Influenza report https://assets.publishing.service.g...al_influenze_report_19_March_2020_week_12.pdf which states that:

"A total of 1,775 new admissions (158 influenza A(H1N1pdm09), 357 influenza A(H3N2), 1,153 influenza A(not subtyped) and 107 influenza B) and 101 confirmed deaths have been reported in the UK since week 40 2019."

and all the associated graphs in the report show flu numbers in line with seasonal expectations and substantially below the 2009-10 levels.

 

[Mellotron]

have you got the article? I just had a look at the latest National Influenza report https://assets.publishing.service.g...al_influenze_report_19_March_2020_week_12.pdf which states that:

"A total of 1,775 new admissions (158 influenza A(H1N1pdm09), 357 influenza A(H3N2), 1,153 influenza A(not subtyped) and 107 influenza B) and 101 confirmed deaths have been reported in the UK since week 40 2019."

and all the associated graphs in the report show flu numbers in line with seasonal expectations and substantially below the 2009-10 levels.

https://www.dailymail.co.uk/health/article-7810129/Flu-cases-EIGHT-TIMES-higher-point-UK-winter.html

Admittedly it's DM so no doubt a fair bit of sensationalising, just surprised it hasn't been made more of.

 

[dingodan]

There is a lot of food for thought in what you write. But even if 10,000 infected only corresponds to 10 deaths (and another 50-100 in hospital?) - don't you think that pneumonia deaths and illnesses like that, in February, with the coronavirus in the news, you think no doctor might have put two-and-two together? Certainly worth a test one would think. Seems highly unlikely. Secondly, if 50% of the population have had it, at our current raw mortality numbers - well it can't get any bigger in terms of numbers of people, this must be the peak. And yet we see from Italy and Spain that we can accelerate what we are seeing by a factor of 10 at least.

Beyond that, this kind of modelling would have already been done by the Chinese, they would be perfectly capable of doing it. And if it could spread so widely so easily without detection, the mortality numbers we saw in Hubei would have been repeated all over their country. But as it happens the Wuhan lockdown has worked in China. I think the fact that that lockdown has worked and prevented the spread to other regions of China is the biggest argument against the conclusion of this study.

Finally, the Diamond Princess cruise ship was a petri-dish experiment, to a certain degree. 10 deaths out of 712 confirmed cases out of 3711 passengers and crew. If you agree that all the people could have been exposed and got it asymptomatically, then 10 out of 3711 is a low number (although I think the captain would be pretty shocked if he lost 10 passengers to flu on the same cruise). But everyone was tested, regularly, so it seems far more likely that the quarantine procedures meant that only about 20% got it. 10/712 is 1.5% mortality rate, and 115 people are apparently still sick, 15 seriously so.

Oh I don't believe it's 50% of the population. But I wouldn't be surprised if it was in the millions. The Diamond Princess is a good example, but if the virus came to the UK in November or December it would have had a long time to spread without being noticed (if it really has a lower mortality rate). Also, I'm not sure about the timeline on that ship, but is it possible that some passangers got the virus, recovered from it (with mild or no symptoms) before they ever got tested for it? (I don't know, but more food for thought - and know I am being optimistic, these are ponderings, they are definately not assertions!)

 

[Kalimantan Gull]

https://www.dailymail.co.uk/health/article-7810129/Flu-cases-EIGHT-TIMES-higher-point-UK-winter.html

Admittedly it's DM so no doubt a fair bit of sensationalising, just surprised it hasn't been made more of.

Thanks. If you look at the graph in the DM article, which was from December



and then the same graph extended through to early-March:



You'll see that the overall number of cases in 2019-20 is almost identical to 2018-19 - its just that the peak came much earlier in the season this winter compared to last winter.

As has been pointed out earlier on this thread, our newspapers are terrible at interpreting science properly.

 

[Triggaaar]

over 70's are not a more risk because of age per se, its because they are much more likely to have other conditions.

Is that correct?

I thought I had read something that older people are more susceptible generally because of their older immune/physiological response to the virus itself?

Also, why is it that those under 10 appear to be doing better than those from 10 to 20 (who in turn fair better than those from 20 to 30)? It's not like those a little older have had a tough life with multiple health issues.

 

[Triggaaar]

The author is a professor of theoretical epidemiology which means things like this will be her bread and better. As with everyone else on here I do not have access to the dataset. However, she has a reputation that she will not want to destroy so I think it safe to assume that the dataset used will be as up to date and accurate as feasibly possible.
At the end of the day this is a model and so will change as the dataset is updated. Key word here is theoretical.
The Chinese only really admitted to the flu in January though it now transpires it was around in November.
In the uk we were only looking for the flu in late January so any other cases prior to this would have been attributed to seasonal flu only.
It appears to be very contagious as all members of society appear to be getting it.
The problem is that we only test on serious cases so we have absolutely no idea how many people have had it.
Bottom line is the author is not going to put a her reputation on the line with a bit of a guess ... it states that she will not publish because clearly more data is needed but for her to even come out with it is telling.

We'll see.

 

[Mellotron]

Thanks. If you look at the graph in the DM article, which was from December

View attachment 121461

and then the same graph extended through to early-March:

View attachment 121460

You'll see that the overall number of cases in 2019-20 is almost identical to 2018-19 - its just that the peak came much earlier in the season this winter compared to last winter.

As has been pointed out earlier on this thread, our newspapers are terrible at interpreting science properly.

So the statement “flu cases unusually high in Dec 2019” would be inaccurate?

 

[beorhthelm]

Prof Neil Ferguson has also expressed he's sceptical of this study to the HoCs Science Select Committee

of course he would, scienists are sceptical of unverified hypothesis, even moreso if it potentially undermines their work.

 

[Green Cross Code Man]

I feel like everyone keeps ignoring the article from January - totally unrelated to CV at the time - confirming flu cases in the UK being EIGHT times higher than normal.

Why isn't this far bigger news?!

It didn't state 8 times higher than normal. It said that they were 8 times higher than the previous year. Flu varies in terms of outbreak and each strain is different in terms of severity. So one would need to know the numbers from the previous year to see if January's figures were unduly high or not.

 

[Kalimantan Gull]

Oh I don't believe it's 50% of the population. But I wouldn't be surprised if it was in the millions. The Diamond Princess is a good example, but if the virus came to the UK in November or December it would have had a long time to spread without being noticed (if it really has a lower mortality rate). Also, I'm not sure about the timeline on that ship, but is it possible that some passangers got the virus, recovered from it (with mild or no symptoms) before they ever got tested for it? (I don't know, but more food for thought - and know I am being optimistic, these are ponderings, they are definately not assertions!)

I suppose we'd be working on the basis that on the ship the first people got it, had no symptoms, and beat it before being tested; whereas it was the latter people that got it were the ones that were sick. And this has been repeated in every country in the world.... it seems unlikely that this is true for this current coronavirus outbreak.

However I am going to give you my alternative, completely untested and quite-probably-completely-epidemiologically-unsound hypothesis that might, might, find a friend in this recent modelling..

It goes like this. A version of this coronavirus has been circulating the world for a while, many many months, perhaps much of last year. This version of the virus was not deadly. It could make you ill, but not much more than a cold. The completely anecdotal evidence for this is that my wife was struck down by an illness in August last year that has much similarity with coronavirus, notably the hacking dry cough. And this cough was going around all over the place last autumn, we knew of and have spoken to SO many people that talk of having this horrendous cough last year. A cough very different to any we had experienced before.

The hypothesis then goes that this cough made it to China, where it perhaps reconfigured itself by aligning with the animal-based virus, or otherwise mutated, and then became deadly. Back in the UK, perhaps 50% of people were exposed to the original virus, and have antibodies against that - this is enough to beat the new virus fairly easily. The other 50% of the population do not have these antibodies and find it highly dangerous. This could mean that even if half the population have been exposed, the other half will still see sky-rocketing hospitalisation and mortality rates.

We are constantly hearing that doctors do not know why some people get few symptoms and others are left fighting for their lives. We see very different mortality rates in different parts of Europe and the world. We might find that half the people on that cruise ship simply couldn't get sick by the coronavirus. Who knows, but its an interesting chain of thought

We shall find out very soon about how many people have had it when this antibody test is released, will be interesting results.

 

[Kalimantan Gull]

So the statement “flu cases unusually high in Dec 2019” would be inaccurate?

I have no idea, as I haven't (and am not going to) check back on the average number of cases of flu in December or whether the peak of the flu season is always at the same time of year.

However the statement "flu cases unusually high in winter 2019-2020" would be highly inaccurate.

And if you are proposing that many of the flu cases in December 2019 were actually misdiagnosed Covid-19 cases, then I would say that is highly unlikely, because it would mean that the actual number of flu cases this winter was substantially LOWER than normal.

What the data in the graph seem to show is that the seasonal flu epidemic in the UK proceeded exactly as expected, just peaking earlier in the season compared to last winter.

 

[beorhthelm]

Finally, the Diamond Princess cruise ship was a petri-dish experiment, to a certain degree. 10 deaths out of 712 confirmed cases out of 3711 passengers and crew. If you agree that all the people could have been exposed and got it asymptomatically, then 10 out of 3711 is a low number (although I think the captain would be pretty shocked if he lost 10 passengers to flu on the same cruise). But everyone was tested, regularly, so it seems far more likely that the quarantine procedures meant that only about 20% got it. 10/712 is 1.5% mortality rate, and 115 people are apparently still sick, 15 seriously so.

the other point to note about the Diamond Princess is how it came to be infected in the first place. supposedly a passenger from Hong Kong who got on in Japan. at least he was the first reported case from the ship, we dont know the previous iternay of the other passengers.

...
It goes like this. A version of this coronavirus has been circulating the world for a while, many many months, perhaps much of last year. This version of the virus was not deadly. It could make you ill, but not much more than a cold. The completely anecdotal evidence for this is that my wife was struck down by an illness in August last year that has much similarity with coronavirus, notably the hacking dry cough. And this cough was going around all over the place last autumn, we knew of and have spoken to SO many people that talk of having this horrendous cough last year. A cough very different to any we had experienced before.

ah, now that chimes as the missus has made reference to a sore throat and dry hacking cough going round the school gates in November/December. i dismissed as its too early (Covid-19 ID is from mid-late Dec) but it does sound similar to mild symptoms. your circular route might be taking it too far, it gives more weight to the idea of weaker strained knocking about.

 

[Beach Hut]

Mentioned in the PM briefing today, at present the study is not scientifically proven (or words along that line)

 

[Green Cross Code Man]

Mentioned in the PM briefing today, at present the study is not scientifically proven (or words along that line)

But interestingly, and of course not unexpectedly, it was not pooh-poohed.

 

[Mellotron]

I suppose we'd be working on the basis that on the ship the first people got it, had no symptoms, and beat it before being tested; whereas it was the latter people that got it were the ones that were sick. And this has been repeated in every country in the world.... it seems unlikely that this is true for this current coronavirus outbreak.

However I am going to give you my alternative, completely untested and quite-probably-completely-epidemiologically-unsound hypothesis that might, might, find a friend in this recent modelling..

It goes like this. A version of this coronavirus has been circulating the world for a while, many many months, perhaps much of last year. This version of the virus was not deadly. It could make you ill, but not much more than a cold. The completely anecdotal evidence for this is that my wife was struck down by an illness in August last year that has much similarity with coronavirus, notably the hacking dry cough. And this cough was going around all over the place last autumn, we knew of and have spoken to SO many people that talk of having this horrendous cough last year. A cough very different to any we had experienced before.

The hypothesis then goes that this cough made it to China, where it perhaps reconfigured itself by aligning with the animal-based virus, or otherwise mutated, and then became deadly. Back in the UK, perhaps 50% of people were exposed to the original virus, and have antibodies against that - this is enough to beat the new virus fairly easily. The other 50% of the population do not have these antibodies and find it highly dangerous. This could mean that even if half the population have been exposed, the other half will still see sky-rocketing hospitalisation and mortality rates.

We are constantly hearing that doctors do not know why some people get few symptoms and others are left fighting for their lives. We see very different mortality rates in different parts of Europe and the world. We might find that half the people on that cruise ship simply couldn't get sick by the coronavirus. Who knows, but its an interesting chain of thought

We shall find out very soon about how many people have had it when this antibody test is released, will be interesting results.

Understandable theory, however I’ve read quite a bit now about lots of different epidemiologists thinking this virus either hasn’t mutated, or isn’t likely to significantly in the future. If anything, they were suggesting it was likely to become milder if altering at all.

 

[Kalimantan Gull]

ah, now that chimes as the missus has made reference to a sore throat and dry hacking cough going round the school gates in November/December. i dismissed as its too early (Covid-19 ID is from mid-late Dec) but it does sound similar to mild symptoms. your circular route might be taking it too far, it gives more weight to the idea of weaker strained knocking about.

Right, IF there is any truth in this theory, that weaker version would still be here, and would still be showing up as a positive test for coronavirus, and thus the modelling being done by this scientist could be completely accurate that 50% of the population have been exposed to this virus - but its not the deadly strain that has recently arrived and is tearing through the other 50% of the population.....