Daily numbers, graphs, analysis thread 7

[Barracker]

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.

@BigChocFrenzy from what I've heard, even HCPs aren't guaranteed to be swabbing far enough down.

Hi, I've been lurking from the beginning, and there is something I just don't understand.

Why has Scotland not peaked at roughly the same time as England? We locked down at the same time, and actually cancelled large events earlier. However, some areas may still not have peaked, and the overall peak is certainly later than for the UK as a whole.

On other threads, people are saying it is because we were behind the curve to start with, but surely that doesn't matter, and the thing that matters is the actions people take in lock down to reduce the R rate?

Is it that our care homes have been more badly hit? Or a lower natural immunity level meaning the same actions are giving a higher R number?

Some scottish graphs:
www.pressandjournal.co.uk/fp/news/politics/scottish-politics/2083381/coronavirus-in-scotland-track-the-spread-in-these-charts-and-maps1/

www.travellingtabby.com/scotland-coronavirus-tracker/?fbclid=IwAR1TZDwMG4oZugHfbSAusAsPOWCxkAfX_K_GKU50sDleJSZk2EyGMqosU0k

I don't think Scotland ever really 'peaked' - we plateaued at around an average of 50 deaths per day for 3 weeks.

Scotland is very gradually declining, but more gradually than England as the level of deaths is much lower than England's

We see this even more so in Wales and NI, which have lower deaths than Scotland

The nations / regions / cities / towns within a country can have very different seriousness of outbreak
e.g. Lombardy vs Rome
hence different curve characteristics - in start point, gradient, form, peak, decline etc

At the low infection levels of the 3 other nations, it may be more difficult to squeeze down infection rate below a certain level ?

Maybe analagous to losing weight ?
the pounds usually come off much more quickly when you are 20 stone than when you are 9 stone

Thanks, that makes sense to look at it that way round. It's not necessarily that the infection rate is higher than England now, it's that it was lower to start with, so not so far to drop.

John Burn-Murdoch (FT) Retweeted Henry Lau@henryjameslau

Deaths in care homes in Scotland are declining
but are still higher than hospital deaths for week 18 (week ending May 3 2020).

Here's my analysis of @NatRecordsScot weekly deaths statistics

Overall Scotland continues to see decline number of deaths
but this is still higher than in the previous 5 years

Updated FT excess mortality tracker

John Burn-Murdoch (FT) ‬

UK had 43,000 more deaths than usual in March & April vs 22,000 reported Covid deaths at the time

43,000 vs 22,000 would tie in broadly with most countries having 50% of deaths outside hospital, in care homes etc. I'm constantly surprised that UK official figures show much fewer that 50%, and wonder whether that is due to the official line of announcing "deaths of people who have tested positive" rather than the looser "probable" or "contributing factor" or "on death certificate" that is used in other countries.

Don't think that graph is very helpful, bigchocfrenzy. One cannot plausibly argue that a country with a billion population and 10,000 deaths is worse hit than one with a million population and 1,000.

There are 26k reg'd deaths in England, which is equivalent to 1,400 in Wales (1283 actual), 2,500 (2270 actual) in Scotland, and 900 (370 actual) in N Ireland

The truth is only N Ireland is significantly better off than England.

The metro Cardiff area is worse hit than most of England. Rural Wales not so much - but then neither is the Isle of Wight, or Hull.

Depends what one is looking for, Shoots

Normalised graphs are useful to assess impact, or who is "worse / better"
but imo non-normalised are still useful for assessing infection growth / decline and form of curves

For the 4 UK nations, we could keep a running table of raw totals of excess deaths in 1 column, then that number divided by population in the next column

I find the FT excess death charts in regions / cities is useful,
because it also gives them as a % increase of the 5-year average

This league table stuff is hard to do and (for me, at least) only moderately interesting.

What is interesting and somewhat concerning is the apparent lack of meaningful decline in cases and deaths in U.K. compared to other countries.

I know our data is appalling (understatement) and, in hospitals trend seems to be down but, still, is 6,000 new cases yesterday merely indicative of greater testing? Are they now testing all those reporting symptoms?

I am assuming that the lockdown is working and there are at most 50% as many infectious in the community than at the peak (and hopefully more like 30%), but it would be nice to have some hard data to confirm this, especially if we are going to ease the lockdown.

Larry. I agree. New infections just don't seem to be decreasing. I'm assuming this is down to care homes and workers and higher tests but like you I'd like some clarity on why?

Larry League tables are only of interest to judge whether UK measures are working as well as they could.
The time for seriously criticising anyone should imo be in the public enquiry that we know will take place after the crisis is over, after mass vaccination

With the greater number of tests in the UK, I'm looking at the % positive

The UK is still testing well below Germany,
but Germany - with 83 million people - is averaging not much above 1,000 new cases

Of course Germany's figures may shoot up if relaxing lockdown turns out to be a bad idea, or done too quickly !
which would be bad news for u all

Ditto, hospital+care home deaths in Germany average not much above 100 daily and have been below 100 some days
(death spike yesterday is because a couple of days were missing from the stats, some RKI late collation issue)

re deaths:
ICU numbers are decreasing, but I expect in all countries we will see more of the longterm ventilated dying, which may artificially raise deaths
Some of the ventilated have been on there 40 days !

I am assuming that it is just more testing. 6,000/100,000 rather than 6,000/20,000 and 80,000 with mild symptoms not tested but we need proper facts, not speculation.

Imperial have asked for volunteers to do home antibody test. The results of this should be helpful.

The ONS is superb, afaik the most up to date full stats of any country,
but it does mean we should be careful about comparing with stats from other countries that are a month old

John Burn-Murdoch@jburnmurdoch (FT)

Here’s the UK.

7 regions have excess deaths above 50%.

The utility of the "how many regions" measure is that
it shows whether a country managed to contain its outbreak in one broad area,
as Italy & China did.

Bad outbreaks in multiple regions suggest action was taken too late.

One other metric for whether a country handled its Covid crisis well is
how many regions suffer bad outbreaks.

Here’s Italy’s data at region level, sorted north to south.

Lombardy was hit very hard,
and two other northern regions saw excess deaths of 50% or more

There’s been a lot of talk lately about UK having Europe’s highest death toll.

The only definitive answer on this will come from excess mortality.

Right now, it’s impossible to say whether UK is the worst in Europe,
as Italy has only published excess deaths data for March.

Oops didn't show

Interesting article (it's in the Spectator but Matt Ridley, who usually writes sensibly). The sub-title should appeal to posters on this thread!
It is data, not modelling, that we need now

Lots of science and research, but some conjecture too. I thought the comments about the peak were particularly interesting.

If Covid-19 is at least partly a ‘nosocomial’ (hospital-acquired) disease, then the pandemic might burn itself out quicker than expected. The death rate here peaked on 8 April, just two weeks after lockdown began, which is surprisingly early given that it is usually at least four weeks after infection that people die if they die. But it makes sense if this was the fading of the initial, hospital--acquired wave. If you look at the per capita numbers for different countries in Europe, they all show a dampening of the rate of growth earlier than you would expect from the lockdowns

I'd assumed it was a behavioural effect because people in high-risk areas like London.

Surely it has to be that the UK was significantly behind in testing, so it's not so much that new cases are high now, but that they were artificially low for the last few months.

So they may be actually decreasing, but because there is no data as to what the levels of infection should have been during April, the decline isn't showing in the figures.

UK infections seem to be declining in the country,
since more tests didn't show an increase
and hospitals have more cases leaving than arriving

However, still higher imo than we want before releasing lockdown

• at least, other countries that have started to relax did so at a much lower infection level

6k for UK, 67 million pop - with fewer tests than most

1k for Italy, 60 million pop
1k for France, 66 million pop
1k for Germany, 83 million pop
2k for Spain, 47 million

Probably Whitty or Boris won't release their target numbers for cases & deaths, but they must have such targets

Cathy There may also be the "seasonal" factor

• some viruses are very seasonal, even allowing for stronger immune systems in summer

We don't know if COVID is, but we should be able to see this within the next few weeks

• in it is, we really could get back to almost "normal" for a few months

I am watching that with interest @ big choc, a doctor in france is convinced if will die out for the summer ( end of may) . I dont link any of his stuff because it is not robust enough and he is extremely controversial but he has clearly stated this.

Cathy There is likely more hospital spread in some countries with poor infection control

• caused by lack of separation, inadequate PPE & testing of staff

The theory has some merit, but does not take account of the large spread initially from returning skiers and large public events,
e.g. carnivals, exhibitions, football matches

Also, look at the age ranges in the charts of the infected
The main problem re the elderly is their death rate when infected

e.g. yesterday's RKi summary in Germany

"Of all deaths, 6,053 (87%) were in persons 70 years or older,
but only 19% of all cases were in this age group."