Meet the Other Phone. Protection built in.

Meet the Other Phone.
Protection built in.

Buy now

Please or to access all these features

Covid

Mumsnet doesn't verify the qualifications of users. If you have medical concerns, please consult a healthcare professional.

Daily numbers, graphs, analysis thread 7

981 replies

Barracker · 28/04/2020 12:53

Welcome to thread 7 of the daily updates.

Resource links:
Worldometer UK page
Financial Times Daily updates and graphs
HSJ Coronavirus updates
Johns Hopkins Coronavirus Resource Centre
NHS England stats, including breakdown by Hospital Trust
Covidly.com to filter graphs using selected data filters
ONS statistics for CV related deaths outside hospitals, released weekly each Tuesday

Thank you to all contributors for their factual, data driven, and civil discussions.Flowers

OP posts:
Thread gallery
127
BigChocFrenzy · 29/04/2020 22:29

Daisy re Spain:

They reported 325 deaths, slightly up from 301 yesterday, but not statistically significant imo

However, they then added a batch of additional deaths from previous days, which had been delayed for some reason,
to bring the total to 453

BigChocFrenzy · 29/04/2020 22:39

Factors that increased the UK death toll:

  • Too late lockdown, due to dithering / confusion / policy debates
    The potential effectiveness of a lockdown is reduced when it is too late within exponential infection growth

  • Mass events were allowed that brought in infection from abroad and spread it around the UK

  • A policy choice to protect the NHS from being overwhelmed by restricting access to hospital until turning blue
    (This policy has since been relaxed to some extent, by lowering the criteria)
    Early O2 is known to reduce the risk of more serious deterioration

  • Very little treatment ofered at home for patients; few home visits, unlike Germany

  • Lack of PPE and other resources has meant that infections have spread through some hospitals,
    infecting former non-COVID patients already weakened by other ailments

clarexbp · 29/04/2020 23:06

Small glimmer of hope on the treatment front. Second randomised controlled trial of Remdesivir reducing duration of symptoms from 15 to 11 days. In the treated group, mortality was 8% compared to 11.6% in the control group, although this difference was not statistically significant - the authors put this down to lack of 'power' (i.e. not enough participants) but they had well over a thousand patients, so should have been quite well powered to detect a difference this large, I would have thought...? Full paper doesn't seem to be published yet, so I can't have a look.

www.bbc.co.uk/news/health-52478783

This compares with the Wuhan trial of Remdesivir which published last week and seemed to show a non-significant reduction in length of symptoms (but only for those treated earlier in their course of illness). Again no statistically significant effect on mortality, although they were severely underpowered because they ran out of patients to treat - it looks like they only managed to recruit a third of the cases they needed. Some interesting non-statistically significant patterns reported in the paper - which they downplay, appropriately - for those treated early in the course of illness, mortality in the Remdesivir group was 11% compared to 15% in the control group. This was reversed in those treated later in their illness so that Remdesivir group did worse. But again, a non-statistically significant pattern. Also worth nothing that the randomisation Gods were not on their side (yes, they do exist, and they are a malevolent bunch) - the control group patients were slightly less sick and had fewer severe comorbidities than the remdesivir group meaning that the drug had to work harder to show a difference in outcome.

www.thelancet.com/journals/lancet/article/PIIS0140-6736(20)31022-9/fulltext

So technically 'nothing to see here' but both reported a difference in mortality between treated and control groups of around 30%. If larger studies confirm a statistically significant difference of that magnitude, that's a big step forward.

DaisylovesDonald · 29/04/2020 23:08

@BigChocFrenzy sorry I meant new cases in Spain - today they’ve reported 4771 new cases as opposed to 2706 yesterday?

Daily numbers, graphs, analysis thread 7
BigChocFrenzy · 29/04/2020 23:22

Sorry, daisy I thought you meant deaths

I often see this effect in Germany on a Monday or Tuesday, because of weekend delays
Although it is a Thursday, maybe just some lab / reporting / collation delay of a batch of tests
or maybe a large increase in the number of tests done ?

One day means nothing much
See what happens Thursday and Friday.
If they are both higher, then we can check if Spain increased testing recently

BigChocFrenzy · 30/04/2020 00:26

Andy Brucee@BruceReuters*

Critical point on this chart shown at the daily UK press conference

  • Belgium,the top line, includes suspected cases in its daily deaths data, which the UK does not.

France also does this.

BigChocFrenzy · 30/04/2020 00:33

John Burn-Murdoch@jburnmurdoch (FT)

To put it another way, the number of deaths registered in England & Wales in the week ending 17 April was the highest since this data was first recorded
The
The dots on this chart represent every week going back 50 years.

And remember: that huge Covid spike came despite lockdown.

And here are subnational charts for England & Wales.

By 17 April there had been 27,000 excess deaths, yet reported Covid death numbers were just 13,719.

And deaths in London over recent weeks have been more than double the recent historical average.

Daily numbers, graphs, analysis thread 7
Daily numbers, graphs, analysis thread 7
BigChocFrenzy · 30/04/2020 00:37

John Burn-Murdoch@jburnmurdoch (FT)
.......
148k more deaths than usual across the 15 countries we‘re tracking;
54k above reported Covid deaths at the time

One new addition is the US, where we’re already seeing 17,000 excess deaths nationwide,
and we know lags in death registration mean this number will rise even for the period already shown.

The spike in England & Wales also continues to rise very steeply

Here are subnational regions for the US.

The urban North East bore the brunt of the early outbreak,
with excess deaths spiking in several states and cities.

NYC the undoubted epicentre;
but deaths in New Jersey also doubled.

Daily numbers, graphs, analysis thread 7
Daily numbers, graphs, analysis thread 7
BigChocFrenzy · 30/04/2020 00:44

Death rates have soared in some cities around the world:
e.g. London, NYC, Bergamo, Madrid, Stockholm

See charts below

John Burn-Murdoch@jburnmurdoch (FT)

Another addition is the Brazilian city of Manaus, epicentre of the country’s outbreak.

Between 2015 & 2019 there were never more than 600 deaths registered in the city during April.
This year there were already 1,676 by 18 April,
when the official Covid death count was just 161

Ecuador’s Guayas province still looks grim.

Latest figures from Ecuador’s civil registry show more than 12,000 deaths in the province since start of March.
In recent years, average for the same period is just 2,000,
yet the official Covid death count still stands at just 245.

Daily numbers, graphs, analysis thread 7
Daily numbers, graphs, analysis thread 7
NewAccountForCorona · 30/04/2020 00:48

BigChoc, Ireland also includes suspected cases. Today's press conference isn't printed in full anywhere, but Kathleen MacLellan from the Public Health Emergency Team pointed out that we had followed care homes from the start, and that every death that was "probably" related to Covid was recorded as such.

They have done a mortality census of all care homes from the beginning of the year to ensure that all possible deaths are included in the totals, and have concluded that 1 in 5 deaths in care homes since the beginning of the year was Covid-related. This is expected to increase.

The data isn't directly online yet - this is a newspaper report www.irishexaminer.com/breakingnews/ireland/one-in-five-nursing-homes-deaths-since-beginning-of-january-resulted-from-covid-19-infections-996829.html

NewAccountForCorona · 30/04/2020 00:57

This is the excess mortality rate chart for countries from the financial times. I presume he is looking at the equivalent of the ONS figures for the various countries, so I don't know how up to date or accurate they are, but I presume eventually we will see more accurate figures than those announced each day.

Daily numbers, graphs, analysis thread 7
EducatingArti · 30/04/2020 08:12

If we hadn't had a lockdown, what would people's guesstimate be for total excess deaths in uk. 150k?

Choux · 30/04/2020 08:14

Those FT charts show the scale of the outbreaks really well. Lagging yes but it eliminates the impact on official stats of a country's testing strategy.

England and Wales shows 27k. This isn't even the full UK figure.

Spain also shows 27k but is further along its' curve than the UK.

It is noted in the footnotes that Italy's graph is not for the complete country as the data is not available for all regions. It would be good if they could find a way to make it complete.

wintertravel1980 · 30/04/2020 09:50

However, the Russian response under Putin has looked slow and disorganised, maybe concentrating too much on stirring up trouble abroad to look after his own country.

I am not sure I agree with "slow". Russian case is actually quite interesting. They recognised the risk of C19 ahead of many other countries. They expanded their testing capabilities. They introduced testing and mandatory self-isolation requirements for all new arrivals from Italy from very early on and quickly expanded the list to the rest of European countries. They actively traced contacts of all confirmed cases. All the self-isolating people in Moscow had police officers visiting them daily (unannounced) to make sure they are staying inside and following the rules. The response outside of Moscow was a bit more patchy but the self-isolation rule for all arrivals from Europe was consistent across the country.

However the Russian government has completely ignored one of the most important considerations - even the most stringent rules will not work if they are not supported by the general population. Many Russians do not trust the government and take pride in avoiding central rules and regulations. People were escaping from hospitals to catch up with friends. Others just ignored the rules. One of the super spreaders in South Russia was a middle aged professor of medicine. She was supposed to self-isolate for 14 days on arrival from Spain but she ignored the requirement and continued to actively socialise infecting at least 11 people.

Russian response completely disregarded behavioural science and, unfortunately, the country has now got to deal with the consequences.

QuentinWinters · 30/04/2020 09:55

I think Russia's border with china/n Korea is also a factor. If the virus isn't as controlled in China as they are saying, and if it's running unchecked in N Korea that will be having a big impact in Russia too, that other Eastern Bloc countries aren't necessarily subject to

wintertravel1980 · 30/04/2020 10:13

I think Russia's border with china/n Korea is also a factor.

It could be but most of the active C19 cases are concentrated in large cities with Moscow being the number one hotspot. Moscow transmissions are believed to originate from Italy and other European countries rather than from China.

QuentinWinters · 30/04/2020 10:44

That's interesting winter

NewAccountForCorona · 30/04/2020 11:03

The sociology of this is interesting isn't it? Russians don't trust the government so ignore instructions, Swedish people trust the government and so are doing what they are told (mostly) without it having to be enforced. I'm amazed by the compliance in Ireland; usually we rebel against anything forced on us, so I'm presuming the general agreement shows trust in Leo Varadkar as a doctor, rather than the government per se, especially as he lost the election so badly so shouldn't technically be in charge at all.

BigChocFrenzy · 30/04/2020 12:03

Several countries have reported that the Russian online troll / bot army are working flat out to disrupt and demoralise Western efforts to combat COVID

Putin in the early part of the crisis was claiming that COVID was not a problem for Russia
It could be that was only for external propoganda and that at the time he actually was organising measures to protect his own country

However, public lack of trust in a kleptocratic dictatorship is a feature, not a bug

BigChocFrenzy · 30/04/2020 12:13

The FT 7-day rolling average of daily new deaths show that several Wstern countries, including the UK, have passed their peak
and are slowly declining

Interesting to get a global perspective of disease progression as a flow switching from Asia to Europe, then to the USA

Daily numbers, graphs, analysis thread 7
Daily numbers, graphs, analysis thread 7
FingonTheValiant · 30/04/2020 12:22

Logged off early yesterday but just wanted to mention re schools : France is bringing us back into work with classes of 15, so somewhere there’s science saying 15 is acceptable.

There was a cluster in a French lycée (high school) in l’Oise, near Paris. They’ve made quite an interesting study of it. 41% of the school population (students, teachers and non-teaching staff) tested positive for Covid. In total 661 people tested positive as they also tested the “second line”, the households of everyone at the school. When they later performed blood tests for antibodies they got positive results for 171 people (25,9%). This worked out to 41% of infected first line, and 10,9% of infected second line. So the second line were less likely to have to have produced antibodies. Research into why is ongoing. The theory at the moment is that, much like in hospitals, those infected at school were exposed to multiple sources, while at home is was just the one student bringing it home and infecting the family.

In terms of infection, 38,3% of the students were positive, 43,4% of teaching staff and 59,3% of non-teaching staff.

Almost 88% of the people infected in the lycée cluster reported loss of sense of smell or sense of taste. 17% had no symptoms. 5,3% were hospitalised. The average age of those hospitalised was 49 years, average age of those not hospitalised was 18. It affected 7,2% of smokers and 28% of non-smokers.

There are all sorts of other statistics coming out from it. It’s fascinating as it’s a known, young population.

FingonTheValiant · 30/04/2020 12:30

also they found that if the high school student was infected, their parents had a 9%-17% chance of being infected by them at home, and their siblings had a 3%-21% chance of being infected. Seems inexact, but they’re still looking into that as well. It’s the only school where they were able to get in and test everyone to see what was going on. And that was after a couple of teachers died in other schools in the region.

Lumene · 30/04/2020 12:37

Thanks Fing. Do you have a link to the study?

Humphriescushion · 30/04/2020 12:48

I saw that study @fing, but would not have been able to translate it so well.

BigChocFrenzy · 30/04/2020 12:52

"Cluster of COVID-19 in northern France: A retrospective closed cohort study"
(Preprint before peer review, English lang.)

www.medrxiv.org/content/10.1101/2020.04.18.20071134v1.full.pdf+html

Swipe left for the next trending thread