Coronavirus - Data Analysis Thread

[B'stard Child]

Biker 1 said:

Terminator X said:

Elysium said:

An update with todays data. I think this may be the last one I do as I am now daring to dream we are finally on the other side of this.

Edited by Elysium on Friday 4th February 18:44

You have done a sterling job but let us hope so!

TX.

Seconded!

Thirded - thanks for all the graphs fishes

Next slide pls

[Elysium]

One more update given the now very real prospect of an imminent end date for COVID testing

1. Year on year comparisons of the last two winter waves. I have simplified the first graph which now only shows cases per 100k tests and deaths within 28 days. The second graph shows the three main hospital metrics. The conclusion is now very obvious. Cases are not resulting in anywhere near the same level of severe outcomes that we saw one year ago:





2. This is why. The CHR (Case Hospitalisation Rate) and CFR (Case Fatality Rate) have consistently declined since December 2020. Back then 12% of cases resulted in hospitalisation. Now it's less than 2%. Back then 3% of cases resulted in deaths, whereas it is now less than 0.2%

These are enormous differences



3. These graphs set out the same metrics for the whole pandemic. The most striking thing is the MV bed line on the final graph. There was no winter wave of people requiring ventilators. We had 3500 beds occupied last year and its about 10 times lower now.

[Elysium]

I think we now have evidence that 'zero COVID' or the eradication of COVID in a region through isolation and lockdown has failed.

New Zealand, which was once the poster child for this idea, has now exceeded the UK's recent peak of cases when corrected for population size. However, they may now benefit with a lower death rate due to the inherent lower severity of omicron in a heavily vaccinated population:

[CarlosFandango11]

Elysium said:

I think we now have evidence that 'zero COVID' or the eradication of COVID in a region through isolation and lockdown has failed.

New Zealand, which was once the poster child for this idea, has now exceeded the UK's recent peak of cases when corrected for population size. However, they may now benefit with a lower death rate due to the inherent lower severity of omicron in a heavily vaccinated population:

What was New Zealand’s strategy for Covid? I don’t think it was to keep at “zero COVID” forever. If it was to keep levels of COVID very low, or zero, until their population had a high level of vaccinations to minimise deaths, then it looks like they’ve been successful.

[Elysium]

CarlosFandango11 said:

Elysium said:

I think we now have evidence that 'zero COVID' or the eradication of COVID in a region through isolation and lockdown has failed.

New Zealand, which was once the poster child for this idea, has now exceeded the UK's recent peak of cases when corrected for population size. However, they may now benefit with a lower death rate due to the inherent lower severity of omicron in a heavily vaccinated population:

What was New Zealand’s strategy for Covid? I don’t think it was to keep at “zero COVID” forever. If it was to keep levels of COVID very low, or zero, until their population had a high level of vaccinations to minimise deaths, then it looks like they’ve been successful.

They had an elimination strategy, which they rightly abandoned last summer:

https://edition.cnn.com/2021/10/05/asia/new-zealan...

[CarlosFandango11]

Elysium said:

CarlosFandango11 said:

Elysium said:

I think we now have evidence that 'zero COVID' or the eradication of COVID in a region through isolation and lockdown has failed.

New Zealand, which was once the poster child for this idea, has now exceeded the UK's recent peak of cases when corrected for population size. However, they may now benefit with a lower death rate due to the inherent lower severity of omicron in a heavily vaccinated population:

What was New Zealand’s strategy for Covid? I don’t think it was to keep at “zero COVID” forever. If it was to keep levels of COVID very low, or zero, until their population had a high level of vaccinations to minimise deaths, then it looks like they’ve been successful.

They had an elimination strategy, which they rightly abandoned last summer:

https://edition.cnn.com/2021/10/05/asia/new-zealan...

Quote from link said:

Ardern said while the transition to living with the virus was always a move New Zealand was going to make

Quote from link said:

Speaking to reporters Monday, Ardern said she had no regrets about enforcing New Zealand's zero-Covid strategy over the past 18 months because it had saved lives.
"Elimination was important because we didn't have vaccines. Now we do," she said.

An increase in cases to a very high level, after New Zealand moved from a "Zero COVID" approach to a "living with the virus" approach, doesn't mean that Zero COVID hasn't been successful for New Zealand. The low level of deaths suggests that the strategy has been successful, based on the information in your link.


Perhaps other countries Zero COVID approaches might not be viewed to be as successful...

Edited by CarlosFandango11 on Thursday 3rd March 09:24

[Elysium]

CarlosFandango11 said:

Elysium said:

CarlosFandango11 said:

Elysium said:

I think we now have evidence that 'zero COVID' or the eradication of COVID in a region through isolation and lockdown has failed.

New Zealand, which was once the poster child for this idea, has now exceeded the UK's recent peak of cases when corrected for population size. However, they may now benefit with a lower death rate due to the inherent lower severity of omicron in a heavily vaccinated population:

What was New Zealand’s strategy for Covid? I don’t think it was to keep at “zero COVID” forever. If it was to keep levels of COVID very low, or zero, until their population had a high level of vaccinations to minimise deaths, then it looks like they’ve been successful.

They had an elimination strategy, which they rightly abandoned last summer:

https://edition.cnn.com/2021/10/05/asia/new-zealan...

Your link supports the view that New Zealand's Zero COVID approach has been successful:

Quote from link said:

Ardern said while the transition to living with the virus was always a move New Zealand was going to make

Quote from link said:

Speaking to reporters Monday, Ardern said she had no regrets about enforcing New Zealand's zero-Covid strategy over the past 18 months because it had saved lives.
"Elimination was important because we didn't have vaccines. Now we do," she said.

Perhaps other countries Zero COVID approaches might not be viewed to be as successful...

Although this is intended as a data thread rather than a discussion one I am going to reply, because I think your comments are relevant to the analysis of the New Zealand data in the coming weeks.

Zero COVID is an elimination strategy. The openly stated objective is that by maintaining NPI's for a lengthy period you can eventually eliminate the virus from entire regions or countries.

This is the strategy that New Zealand were pursuing until it became apparent that it was not working. Prof Mark Woolhouse, who was part of the SAGE modelling team SPI-M, has recently published a book that discusses the flaws in this idea. Not least of these is the characteristics of exponential decay, which means that universal NPI's would need to be applied for a very long time indeed.

You are arguing that their approach may ultimately have led to the best outcome. That is a different argument to the idea that the zero COVID policy 'worked', because if that were true it would have achieved its stated aims and the virus would have been elminated from New Zealand.

New Zealand is now experiencing its first coronavirus wave. I think they will benefit from having delayed it, but that was a gamble and only time will tell.

[the-photographer]

Cases generally down, admission ticking up



Waning, more mixing (with reduced testing capturing less positives), something else?

UKHSA yesterday

[CarCrazyDad]

the-photographer said:

Cases generally down, admission ticking up



Waning, more mixing (with reduced testing capturing less positives), something else?

UKHSA yesterday

Don't think it's anything to do with Covid.

It's just people with positive tests going into hospital for anything.

[NoddyonNitrous]

Here are two interesting questions from the Scottish statistics.

What happened on February 26th? Relaxation of restrictions effect?



Why is the length of stay for Covid patients increasing?

Hospital admissions are flat:



Number of inpatients is increasing, which can only be because they are staying in longer:



In ITU, the number of long stay (>28 days) are also decreasing, but a slight increase in shorter stay. Our pre-Covid average length of stay in ITU was 2.5 - 3 days for comparison. My ex-colleagues report that they have virtually no COVID coming to ITU now, and are not seeing pneumonitis except in unvaxed patients.

[url]

[the-photographer]

Its a puzzle

Antibodies stable
https://www.ons.gov.uk/peoplepopulationandcommunit...



But





More graphs here https://twitter.com/VictimOfMaths/status/150151155...

[NoddyonNitrous]

An increasing proportion of inpatients testing positive are coincidental rather than prime cause of admission. Maybe the combination of covid and other disease is slowing recovery.

[Zad]

I have a couple of theories regarding the recent increase in infections. Firstly, the weather in February was especially nasty. Lots of high winds for a long time. This sort of thing deters people from socialising, and so drove the numbers down quite quickly. Secondly, the school half term holidays were (varying on local authority) at the end of Feb. This pushed the stats for under-19s down further.

Then the schools went back, and the weather improved and infections started increasing again. This might explain why the up-tick is so geographically widespread. The recovery has been rather counter-intuitive; rural areas seem to be slower than urban ones. Presumably because omicron burned through the population so quickly in cities that it created a sort of fire break. For example, several inner city areas of Bradford are reporting under 3 cases now (which shows as white on the case maps).

I wonder if a difficulty in obtaining free lateral flow tests is having an influence too. People going back to work / school / using public transport when they are still infectious.

[vaud]

Lateral flow tests are still free to order?

[Elysium]

Zad said:

I have a couple of theories regarding the recent increase in infections. Firstly, the weather in February was especially nasty. Lots of high winds for a long time. This sort of thing deters people from socialising, and so drove the numbers down quite quickly. Secondly, the school half term holidays were (varying on local authority) at the end of Feb. This pushed the stats for under-19s down further.

Then the schools went back, and the weather improved and infections started increasing again. This might explain why the up-tick is so geographically widespread. The recovery has been rather counter-intuitive; rural areas seem to be slower than urban ones. Presumably because omicron burned through the population so quickly in cities that it created a sort of fire break. For example, several inner city areas of Bradford are reporting under 3 cases now (which shows as white on the case maps).

I wonder if a difficulty in obtaining free lateral flow tests is having an influence too. People going back to work / school / using public transport when they are still infectious.

I see some problems with these theories:

1. Transmission tends to increase in bad weather because people socialise more indoors. That is the basis of seasonality for viruses.

2. It is now pretty much concluded that schools are not a major driver of transmission.

There is a bit of a clue in the 'Victim of maths' twitter thread quoted above, which observes that cases and hospitalisations have increased simultaneously when there is usually a 7 day lag between them.

This suggests a general increase in prevalence and a corresponding increase in 'incidental' admissions.

I think the explanation may actually very simply. Most people seem to have taken the lifting of restrictions on the 24th Feb as a 'green light' to return to normal. I have certainly noticed more people out and about, mixing, socialising and travelling for work. In fact the last week or so has felt almost entirely like 2019.

Older, more cautious people are the last group to return to social mixing. They are least likely to have been infected before and will have been boosted more than 3 months ago. So this group will also include a fairly high proportion of people who are susceptible to infection.

As this susceptible group returns to normal life transmission is likely to increase. Since this older group is also more likely to be admitted to hospital for other conditions it follows that incidental admissions might also increase.

[grumbledoak]

Elysium said:

I see some problems with these theories:

1. Transmission tends to increase in bad weather because people socialise more indoors. That is the basis of seasonality for viruses.

2. It is now pretty much concluded that schools are not a major driver of transmission.

Isn't school mostly indoors?


I suspect "seasonality" is dominated by other factors.

[Elysium]

grumbledoak said:

Elysium said:

I see some problems with these theories:

1. Transmission tends to increase in bad weather because people socialise more indoors. That is the basis of seasonality for viruses.

2. It is now pretty much concluded that schools are not a major driver of transmission.

Isn't school mostly indoors?

I suspect "seasonality" is dominated by other factors.

Of course schools are 'indoors', but as was suggested early on, the real world data continues to support the idea that schoolchildren transmit the virus less than adults and that child to child transmission is suprisingly rare.

Seasonality is actually quite interesting. Sunlight is clearly a factor as UV light is proven to degrade viruses, but it is thought that one of the main drivers is the way that peoples behaviour changes in response to weather.

An example of this is warmer sub tropical regions which have a winter AND summer season for respiratory viruses. That doesn't fit with the idea it is directly weather related, but it does support the behaviour idea. People are more likely to socialise indoors when it is cold OR when it is very hot.

This is an airbourne virus that transmits mainly as an aerosol, so ventilation is going to have a very significant effect.

[grumbledoak]

Elysium said:

Of course schools are 'indoors', but as was suggested early on, the real world data continues to support the idea that schoolchildren transmit the virus less than adults and that child to child transmission is suprisingly rare.

Seasonality is actually quite interesting. Sunlight is clearly a factor as UV light is proven to degrade viruses, but it is thought that one of the main drivers is the way that peoples behaviour changes in response to weather.

An example of this is warmer sub tropical regions which have a winter AND summer season for respiratory viruses. That doesn't fit with the idea it is directly weather related, but it does support the behaviour idea. People are more likely to socialise indoors when it is cold OR when it is very hot.

This is an airbourne virus that transmits mainly as an aerosol, so ventilation is going to have a very significant effect.

Don't forget how much some people have wanted to use this to nudge control behaviour. It doesn't need to be true to be useful to them. It just needs to be widely believed that we can "beat this virus" by everyone staying isolated at home and never meeting in person.

I would like to see proper investigation of human susceptibility related to Vitamin D and even simple temperature changes, and virus susceptibility to sunlight and atmospheric conditions like humidity. I bet we don't look in those directions.

[Elysium]

grumbledoak said:

Elysium said:

Of course schools are 'indoors', but as was suggested early on, the real world data continues to support the idea that schoolchildren transmit the virus less than adults and that child to child transmission is suprisingly rare.

Seasonality is actually quite interesting. Sunlight is clearly a factor as UV light is proven to degrade viruses, but it is thought that one of the main drivers is the way that peoples behaviour changes in response to weather.

An example of this is warmer sub tropical regions which have a winter AND summer season for respiratory viruses. That doesn't fit with the idea it is directly weather related, but it does support the behaviour idea. People are more likely to socialise indoors when it is cold OR when it is very hot.

This is an airbourne virus that transmits mainly as an aerosol, so ventilation is going to have a very significant effect.

Don't forget how much some people have wanted to use this to nudge control behaviour. It doesn't need to be true to be useful to them. It just needs to be widely believed that we can "beat this virus" by everyone staying isolated at home and never meeting in person.

I would like to see proper investigation of human susceptibility related to Vitamin D and even simple temperature changes, and virus susceptibility to sunlight and atmospheric conditions like humidity. I bet we don't look in those directions.

Thankfully, these factors are being looked at, but they don't catch the headlines:

https://news.yale.edu/2021/06/14/heat-humidity-and...

Vit D somehow became controversial when it is already well understood that Vit D deficiency negatively impacts our immune response. I've been taking a supplement for 2 years as it is a minimal cost precaution with zero downside.

[vaud]

Elysium said:

Of course schools are 'indoors', but as was suggested early on, the real world data continues to support the idea that schoolchildren transmit the virus less than adults and that child to child transmission is suprisingly rare.

I'm not so sure. It went rapidly through our school and currently 25% of Foundation year are out sick with it. With lowish levels in the community, it would seem odd for them all to get it in one week if it wasn't being transmitted between the kids. Seems mild, but maybe just the current variant is more transmissible between kids.