Main Coronavirus / Covid-19 Discussion Thread

[Green Cross Code Man]

Totally and utterly false, nonsense numbers. Genuinely how anyone or any program can ‘predict’ anything about this with regards to numbers is utterly ridiculous and should not be allowed. Terrible.

Prof Neil Ferguson in the Guardian:

Ferguson said he did not think the predictions could be relied on. “This model does not match the current UK situation,” he said, adding that the numbers used by the IHME were at least twice as high as they should be for current bed usage and deaths in the NHS. “Basically, their healthcare demand model is wrong, at least for the UK,” he said.

The IHME said its model was designed to be updated from day to day as the pandemic goes on. For a country such as the UK, which is quite early on in its outbreak, the uncertainty was higher and the headline numbers might change over the next few days as more data is collected.

https://www.theguardian.com/world/2...ropes-worst-hit-by-coronavirus-study-predicts

Looks to me as though there will be some adjustment in their figures over the next few days.

 

[Machiavelli]

From today’s conference it seemed as though we are not reporting care home deaths.

Yes, there was a big disjuncture between the hospital deaths, and the total deaths.
The ONS are also reporting that the actual number of deaths in the UK might be substantially high, perhaps double the amount. See:

https://www.channel4.com/news/new-data-suggests-coronavirus-deaths-could-be-higher

 

[Green Cross Code Man]

Yes, there was a big disjuncture between the hospital deaths, and the total deaths.
The ONS are also reporting that the actual number of deaths in the UK might be substantially high, perhaps double the amount. See:

https://www.channel4.com/news/new-data-suggests-coronavirus-deaths-could-be-higher

That page loads up for me with only:

New data suggests coronavirus deaths could be higher
Victoria Macdonald
Health and Social Care Correspondent
Every day we’ve been given the official figures for the number of people who’ve died after testing positive for coronavirus from the Department of Health.

Is there more?

 

[jimhigham]

Totally and utterly false, nonsense numbers. Genuinely how anyone or any program can ‘predict’ anything about this with regards to numbers is utterly ridiculous and should not be allowed. Terrible.

I certainly hope that you are right. The numbers are alarming and it would be a huge relief if they ultimately turn out to be overly pessimistic or just plain wrong. I remain concerned given that the institute is drawing comparisons with other nations but without a political agenda (at least that we're aware of). We'll know soon enough I suppose.

 

[Triggaaar]

... poster’s claim that the UK could be heading for the highest per-capita death rate from Covid-19.

That seems extremely unlikely. We have decent health care (not necessarily the best in the world, but sure as hell not the worst). We are also on lockdown, and following social distancing advice. And our numbers aren't worse than Italy, Spain, or France, not to mention countries with poorer health care and poorer political leadership. On that basis, I imagine the whole thing is made up nonsense.

 

[Jimmy Grimble]

https://twitter.com/chrisgiles_/status/1247458186300456960?s=21

This is an interesting Twitter thread re: reporting of death numbers

 

[jimhigham]

That seems extremely unlikely. We have decent health care (not necessarily the best in the world, but sure as hell not the worst). We are also on lockdown, and following social distancing advice. And our numbers aren't worse than Italy, Spain, or France, not to mention countries with poorer health care and poorer political leadership. On that basis, I imagine the whole thing is made up nonsense.

Well the numbers are made up in a much as they are projections based on a mathematical model. We’ll only know if they’re nonsense once time has passed.

Our numbers aren’t worse than Spain or Italy, as they are ahead of us in the cycle of transmission. The model suggests that we are heading towards a high per capita death rate. It doesn’t suggest that this is our position right now. Again, time will tell of this proves to be correct or not.

Our own prime minister has stated that testing is the key to our pathway out of the crisis yet we appear to be far behind other countries in this respect. This is only one factor but could explain why we are projected to have a steeper curve than other nations going forward.

I’m not trying to score any political points. I’m genuinely scared of what’s to come. I don’t want to lose loved ones or succumb myself. For this reason, I find the IHME numbers concerning. I don’t dismiss them simply because I do not like them. What concerns me most is that the authors are apolitical and drawing comparisons between developed nations for which relatively reliable data exists. Nobody that has dismissed the numbers out of hand has yet suggested why the data shows the U.K. in such a poor light save to say that it is simply unreliable. It might be. Bloody hell, I hope that it is, I really do. I’ll be delighted to come back here in a fortnight and say I was hopelessly wrong.

 

[atomised]

https://twitter.com/chrisgiles_/status/1247458186300456960?s=21

This is an interesting Twitter thread re: reporting of death numbers

That's a worrying set of charts

 

[Harry Wilson's tackle]

OK. This is bad...

https://twitter.com/globalhlthtwit/status/1247551346095644678?s=21

Checking the sources, the poster of the tweet is a doctor and former member of the World Health Organisation. He is quoting research by the Institute for Health Metrics and Evaluation (IHME). I wasn’t familiar with the IHME so I looked it up. It was formed in 2007, largely via a grant from the Bill & Melinda Gates Foundation. Wikipedia gives its purpose as “accelerating global health progress through sound measurement and accountable science”. As such, it wouldn’t seem to have a political agenda.

I wasn’t immediately able to find the research to bear out the poster’s claim that the UK could be heading for the highest per-capita death rate from Covid-19. It’s probably in there. I shall look again when I don’t have an early shift the following morning. What I did find was a projection that suggests our peak demand is 10 days away and that we could have just shy of 3,000 deaths on April 17th. The projection admits to a sizeable margin of error so we can only hope it will not be so bad. The bed shortage numbers are quite eye popping too.

I would post more but, as I said, I’m off to bed. I may revisit this tomorrow afternoon but if some of you are better placed to interrogate the data than I am right now, please do so.

I’m genuinely very scared now.

The recent fall in Spain and Italy has stopped. Will it go back to falling or will it go up again? The tiny number of deaths per cases in Germany seems to refect a greater willingness to claim a diagnosis coupled with better social discipline. Will the latter be effective?

I can't call this (over in June versus November) yet.

 

[Bold Seagull]

Yesterday, one set of scientists saying kids could go back to school, another set saying we’ll be the worse hit in Europe. Hard to see the wood for the trees.

 

[Triggaaar]

Well the numbers are made up in a much as they are projections based on a mathematical model. We’ll only know if they’re nonsense once time has passed.

Then the model is nonsense.

Our numbers aren’t worse than Spain or Italy, as they are ahead of us in the cycle of transmission.

No, even when date adjusted (to we all start from the 10th death, or 50th death) we're still behind them.

The model suggests that we are heading towards a high per capita death rate. It doesn’t suggest that this is our position right now. Again, time will tell of this proves to be correct or not.

I suggest they are wrong.

Our own prime minister has stated that testing is the key to our pathway out of the crisis yet we appear to be far behind other countries in this respect.

Behind all other countries? No.

I’m not trying to score any political points. I’m genuinely scared of what’s to come. I don’t want to lose loved ones or succumb myself. For this reason, I find the IHME numbers concerning. I don’t dismiss them simply because I do not like them.

I also don't want to lose loved ones. I also don't dismiss them because I don't like them. I dismiss them because they don't appear to be believable.

 

[jimhigham]

Then the model is nonsense.

No, even when date adjusted (to we all start from the 10th death, or 50th death) we're still behind them.

I suggest they are wrong.

Behind all other countries? No.

I also don't want to lose loved ones. I also don't dismiss them because I don't like them. I dismiss them because they don't appear to be believable.

5 point pedantry. Top work.

Again, I hope you’re right and I’ll be very happy to come back in a fortnight and admit I was unduly alarmed.

 

[Triggaaar]

So by August, the UK will have had 66,314 deaths, while Italy will have had 20,300, Spain 19,209, and France 15,058.

Currently Italy has had 17,127 deaths, Spain 14,045, and France 10,328. Meanwhile we've had 6,159

So we're to have an increase of 1000%, while Italy gets an increase of 18%.

Honestly I don't know how these data modellers justify their existence.

 

[Machiavelli]

That page loads up for me with only:

New data suggests coronavirus deaths could be higher
Victoria Macdonald
Health and Social Care Correspondent
Every day we’ve been given the official figures for the number of people who’ve died after testing positive for coronavirus from the Department of Health.

Is there more?

If you have access to such facilities, you'll be able to watch last night's C4 News, and the Victoria MacDonald piece was early in the programme. Alternatively, you could just go to their website, and find that item, or also you could go to the CV part of the ONS website from which the report was sourced, I'd imagine.

 

[Green Cross Code Man]

If you have access to such facilities, you'll be able to watch last night's C4 News, and the Victoria MacDonald piece was early in the programme. Alternatively, you could just go to their website, and find that item, or also you could go to the CV part of the ONS website from which the report was sourced, I'd imagine.

Thanks, last night I tried accessing the site directly, I found the page but for some reason it just displays weirdly. Perhaps it is my chromebook?

 

[Bakero]

Thanks, last night I tried accessing the site directly, I found the page but for some reason it just displays weirdly. Perhaps it is my chromebook?

Try loading it in an incognito tab.

 

[CHAPPERS]

https://www.reuters.com/article/us-health-coronavirus-britain-path-speci-idUSKBN21P1VF

 

[Green Cross Code Man]

https://www.reuters.com/article/us-health-coronavirus-britain-path-speci-idUSKBN21P1VF

The government will be judged by what happens in the coming months. If things go badly, there will be a lot of justifiable recriminations.

BTW, 29 minute read? Gosh, there are some slow readers.

 

[CHAPPERS]

https://www.reuters.com/article/us-health-coronavirus-britain-path-speci-idUSKBN21P1VF

Fair enough Bozza. Didn't want to select a section of text for fear of (unconcious) politicising.

 

[Harry Wilson's tackle]

So by August, the UK will have had 66,314 deaths, while Italy will have had 20,300, Spain 19,209, and France 15,058.

Currently Italy has had 17,127 deaths, Spain 14,045, and France 10,328. Meanwhile we've had 6,159

So we're to have an increase of 1000%, while Italy gets an increase of 18%.

Honestly I don't know how these data modellers justify their existence.

There are modellers and modellers. The above is certainly not what my modeller mate is predicting, but anything is possible when you reset the parameters. The credibility of the prediction is, however, Gaussian distributed.....

My pal's main gripe is the provenance of the data. We wrote a paper on modelling of something I work on a year or so ago and concluded the predictions had 'limited value' owing to the shitness of the wet data available. I'd suggest the COVID-19 'wet data' we are getting from around the world is a bit like shirt sizes - even my XXL Brighton shirts are all of vastly different dimension, from skin tight to down to my knees. I have a feeling that all the useful modelling will be done after the fact, later this year, when predicting how COVID-20 will manifest.

Like predicting GE outcomes, among all the predictors someone is bound to get the prediction correct, but here is a huge amount of dumb luck involved.