Daily numbers, graphs, analysis thread 2

[Barracker]

A follow on thread from here

Please try to keep it data driven, factual and civil.

www.worldometers.info/coronavirus/country/uk/

Great thread, thank you

the figures always dip at weekends
Sorry took an age to respond, but the figures that I can find from last weekend fri to sat the infection rate went up by 71 on the Friday, 321 on the Saturday and 4648 on the Sunday.
This last two days look to be the first it has slowed/reduced two days consecutively at all.
I know you cant read too much into a few days at the moment but lower rates of people getting sick now is a good thing right?

Re the dataset that was posted on the previous thread

www.icnarc.org/DataServices/Attachments/Download/b5f59585-5870-ea11-9124-00505601089b

It appears men are statistically more likely to become critically ill than women (70.5% of UK critical care patients are/have been male vs 29.5% female).
Any ideas as to why this could be?

Thank you for the new thread

@BayHorse Men are much more likely to die from covid-19 too. There are various suggestions as to why, probably a mix of behavioural, lifestyle and biological factors (handwashing, smoking, differences in immune systems). Men were also disproportionally affected by SARS and MERS:
www.theguardian.com/world/2020/mar/26/men-are-much-more-likely-to-die-from-coronavirus-but-why
and
www.nytimes.com/2020/03/20/health/coronavirus-italy-men-risk.html

Fascinating thread thanks.

Thanks for the new thread. I’ve been watching avidly.

Anyone else think that New York’s figures should be reported separately? It’s something DH and I were discussing earlier, but I can’t decide if I’m for or against.

A reporter asked an interesting question today in the daily briefing, saying that trusts are actually having more deaths than are being reported in the official figures. It is looping on sky atm. Is that what concerned you re the Cambridge source, @Barracker?

Today's graph.
We're still sitting right on top of Italy's trajectory, 16 days behind. Marginally faster growth than France.
Only Spain is accelerating faster than we are.

Marmite27

New York is looking very scary compared to other cities/regions.

NY looks like it’s beginning to tend towards the deaths doubling every day trajectory which would be awful.

Wow, that graph

You might be interested to look at the stats for Iceland - we are currently leading the world on testing per capita I believe so our data is probably providing a more accurate picture than many other countries (though we are way behind the southern European countries in terms of when it got here).

About 4% of the population have been tested so far. The health service is testing people exhibiting likely symptoms (cough, fever, aches) and a private biomedical research company are testing people with no symptoms (self selected volunteers).

www.covid.is/data

We had our first case confirmed on 28 February and 2 fatalities to date (one tourist who almost certainly had it already when he came into the country and one citizen).

Positive tests = 6.6% of tests taken.
0.2% of confirmed cases have led to death so far.

That graph seems to show that the Uk is following the same path as Italy. Spain looks even worse :(

Is that what concerned you

Yes.
I looked at the breakdown by trust.
And what I saw, was that many of the trusts were reporting deaths that dated much earlier than the 28th or the 27th. Some were reporting deaths that arrived in today's figures, but were dated the 21st, 25th, 16th. I'd expect one or two earlier deaths to be swept into each day's stats, but this is a really large proportion, and some are two weeks old.

The one that struck me especially was the Countess of Chester hospital. It reported one death into today's stats, and the death was dated 27th. I'm aware of one death at this hospital that occurred on March 14th. It was reported widely in the news. So it would seem that the death in today's stats can't be that same one.
But I can't find any reports of another death. And I can't see the stats for all of the trusts other than the list published today, for which I can't see the source data.

It's making me wonder how long the lag is before deaths are showing up in stats, and what proportion are delayed.

Judging by the list published today, there are a LOT of deaths not making the stats until several days after they happen.

So if anyone can find the source data from NHS England that the media is reporting today, that would be very helpful to understand how reliable the reported daily stats are.

Thanks, Puffin. Interesting.

In the Guardian live blog: Patients in the UK dying now were infected around a month ago, weeks before the country started social distancing, an expert has warned.

Professor James Naismith, director of the Rosalind Franklin Institute at the University of Oxford, spoke out after a further 209 people died after testing positive for coronavirus in the UK.

Although this is a lower number than the 260 deaths the previous day, he warned it is too early to draw conclusions, underlining that “daily numbers can be lower (or higher) than expected due to random factors”.

Prof Naismnith said: “I am confident that provided we follow social distancing the increase in deaths will stop and this will be followed by a decrease in the daily numbers.

“It can take up to two weeks from infection to onset of symptoms. For over 80% of people these symptoms are mild, for the much smaller number who develop the most serious illness, intensive care is needed around 10 days later.

“For those who do not recover but die, data from China suggests this takes around another four days. The tragic deaths today will be from people infected roughly a month ago. It is inevitable that there will be more tragedies for families ahead.

“We can reduce the final death toll only by following current Government advice. We must expect and welcome changes in Government advice as science, medicine and social science work together to learn more.”

Why is the percentage of deaths for cases with an outcome, 90% in the UK? WTF!

Italy = 45%
Spain = 32%
Germany = 5%
Iran = 18%
France = 27%

www.worldometers.info/coronavirus/country/uk/

About a month ago, the handwashing advice was in place, I think - Boris Johnson's tv interview on that was 28 Feb.

My not very accurate memory is that at the beginning of the following week (beginning 2 March) a few people had started to be cautious about, say, going on the tube, and some events were being cancelled. By the end of that week people were much more nervous. So this week to come the figures will reflect the period when social distancing increased fairly rapidly but was obviously on nothing like the scale we have now.

Because at the earliest part of the process, there are more deaths happening early, which inevitably close the case outcome, than recoveries, which take longer. Only when the entire epidemic has run its course do the numbers stabilise, as all cases reach one conclusion or the other.
Also, every single case in the UK represents someone ill enough to be hospitalised. The death rate in this cohort will be significantly higher than the death rate in a country which includes mild illness cases in its cohort of total cases tested

Following

Following. Thanks Barracker.

My not very accurate memory is that at the beginning of the following week (beginning 2 March) a few people had started to be cautious about, say, going on the tube, and some events were being cancelled. By the end of that week people were much more nervous.

Yes, that's my experience too. We were talking about whether a work conference we were organising in the third week of March would go ahead on 2 March and some people thought it was ridiculous. I flew abroad on 2 March and there was no social distancing in evidence then in either country. Then back on 5 March and away that night at a work event and people did bring it up in conversation. But it wasn't a thing most people in my circles thought about much until early the next week with Nadine Dorries etc.

Sorry if this has already been covered, but is it true that only deaths following a positive test in hospital are recorded as official COVID deaths here in the UK? So for example an elderly person who does at home or in a nursing home may or may not have had it but won’t be counted? What if they were known to have been symptomatic before death? My fear is that the death rate isn’t accurate since people are largely being encouraged to stay at home if they think they have it.

@Barracker, aren't the 'cases that had an outcome' for each country on worldometer only the hospitalised cases? Germany is on a comparable time frame to the UK too. Maybe I'm missing something?