Daily numbers, graphs, analysis thread 3

[Barracker]

Welcome to thread 3 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.

I know this is pedantic but viral load is the amount of virus in the blood and viral dose refers to the amount of virus you have been exposed to.

I think viral dose seems to be of some importance both because it mechanistically makes sense and because it provides a reason as to why young healthcare workers seem so at risk. In addition research has shown that higher doses provoke more serious diseases in many viruses (mainly animal research for ethical reasons..).

Encouragingly it does look like new cases have at least stabilised in U.K. with today’s numbers being lower than those on April 1st, 2nd and 3rd. Before lockdown this was increasing by at least a factor of 4 weekly.

What would be even more encouraging would be seeing a meaningful decline, and it would be nice if the NHS would sometime update the ‘recovered’ data.

Attached are case doubling times for several countries
As Whitty said, about 5 days for the UK, so I'd have thought some way to go yet to target.

Austria is starting to relax measures very cautiously at only 8 days, but their cases are at a far lower level than neighbouring countries

Germany has a high case level, with 5,000 in hospital;
hence Merkel originally said doubling at 10 days and then later said maybe even 13 or 14.

Merkel: ‘We Have to Live with Pandemic for a Considerable Time’

Lockdown continues, but she says no tightening needed atm

https://berlinspectator.com/2020/04/09/merkel-we-have-to-live-with-pandemic-for-a-considerable-time/

Presumably with high viral dose, you are at risk of high viral load - which is what is causing the most serious cases in the younger patients -
unless PPE or at least basic masks are used

Thank you for being pedantic @larrygrylls, I’ve been scratching my head at all the talk of viral load over the last week because I always thought it was more how a body reacted to a reproduce a virus; I.e antivirals for hiv lower viral load regardless of how the virus was contracted. Viral dose makes much more sense.

Dr Hannah Fry doing us all proud on Horizon now, BBC 2.

I'm so pleased to see the University of Bonn study is out - but it raises lots of questions.

Can the results be extrapolated (however tentatively) to other countries? If 15% of the German population (or that bit of Germany, at least) have COVID-19 antibodies, does that mean that the UK, which has had (according to Worldometers) 4 times as many deaths per million population, has had 4 times as many people infected...? Or does it not work like that?

Also, I thought a study came out yesterday showing that lots of people who have had clinically diagnosed covid are not showing antibodies...? So, does that mean the German study might be underestimating infections?

We must see the full report with their findings and conclusions
I expect they'll say some aspects need further study

What is interesting is that they confirmed their earlier findings that CV doesn't live long on surfaces

• they found lots of dead virus, but not much alive

So they are emphasising "hygiene " - hand washing, avoid touching your face ... and masks to avoid droplets in / out

Masks suddenly very much in the spotlight here, which is a big change

Even talk here of setting up a separate ministry to organise the production of umpteen million masks per day, for everyone to have one per day

Doesn't live long?? Did they specify whether they meant hours, days or what?

If you don't show antibodies do you really have any Immunity?

I'm no immunologist Midgebabe but I suspect no antibodies = no immunity. However, if you've been infected once and not mounted an antibody response, yet been recovered well, I'm guessing chances of you being fine second time round are pretty good.

I've found the Uni Bonn COVID report, which is brief and only interim, as the study has some way to go yet

However, they already made recommendations that relaxation of measures, when decided, should follow the existing DGKH strategy

https://www.land.nrw/sites/default/files/asset/document/zwischenergebniscovid199casestudyygangelt_0.pdf

We therefore strongly recommend implementing the proposed four-phase strategy of the German Society for Hospital Hygiene (DGKH).

This provides the following model:

Phase 1: Social quarantine with the aim of containing and slowing down the pandemic and avoiding an overload of the critical care structures, in particular the health care system

Phase 2: Beginning withdrawal of the quarantine while ensuring hygienic framework conditions and behavior.

Phase 3: Abolition of the quarantine while maintaining the hygienic framework

Phase 4: State of public life as before the COVID-19 pandemic
(status quo ante).

Also Midgebabe, knowing number of true infections is critical to understanding the true mortality rate.

Thanks for clarification of dose/load, larrygrylls.

We never normally know the "true" mortality rate of a disease, only the % of deaths of reported cases,
so the official % is always much higher than reality

e.g. flu mortality is only the % deaths of those taken to hospital or who went to a doctor,
whereas there are a large number of people whose symptoms were "bad cold" and / or didn't need a sicknote for work if retired, SAHP etc

What's the reason for the lag in reporting deaths? For example there were a lot in the figures today but were really from the previous week

That's great BigChoc - thanks for the translation and summary - better than I've found on the news sites.

Thanks for the links BigChocfrenzy and sexnotgender. I'll keep hunting for those daily England stats.

Here's the volcano, itsgettingweird.
I've added blue bars for the announced England deaths, which gives an apples to apples comparison against the volcano.

I've no clue what has happened to the data on March 31st, it's a very weird dip, and it isn't getting backfilled. I can only guess that data got shoved somehow into the days either side.

I think we can assume with a degree of confidence that all of March is pretty much complete data now.
April continues to be backfilled each day.

Dr Streeck says it is time to move to Phase 2

• but I think he just means for Heinsberg,

as this area is êstimated to be about 2 ½ weeks ahead of the rest of Germany in the progress of CV

The idea seems to be to use Heinsberg as a canary down the mine for the rest of Germany

So they try things out bit by bit and hope that Germany as a whole can follow their "hygeine model" sufficiently well, but 2-3 weeks later

The volcano is here !
That's excellent, barracker 👏🏼

but it does show what a mix of data is issued on any date

• which would also be the case in other countries too Comparisons and conclusions need to be done with great caution

What I'm noticing, as the body of data grows, is that each day will probably see at least 10 noticeable layers of 'lava' before the incremental increases become insignificant.

For example, if you look at April 1st, you can see 8 shades of blue, 8 noticeable layers of 'lava', each representing one of the last eight days of NHS timestamped data getting properly backfilled retrospectively into the date they occurred, April 1st. And there will probably be another two noticeable layers laid before it's done.

This 'ten layers till done' gives a rough idea of where the true trajectory will end up.

Everything to the left of the blue marker on the graph is complete-ish.
What you see there, to the left, is the true trajectory of England's deaths, timestamped into the days they occurred.

Everything to the right of the marker is not yet complete, and will continue to grow until you can see 10 layers or so at which point it will be the final picture.

That's fascinating about the lava layers, like a contour map over time

31 March just looks like a big chunk of data couldn't be processed, maybe something like an IT outage in reporting or collating systems ?

It's weird, because when you look at the spreadsheet of numbers every day the backfilled numbers for March 31st keep being smaller than they 'ought' to be.
These numbers are compiled from all the trusts, it makes no sense that they all should see the same large dip on that day.
There must be an explanation for it, it's too odd.

When did they change the reporting cut off time? Is it that?

I don't know, but it ought not to be. These are timestamped deaths. They should eventually be allocated to the correct day regardless of reporting vagaries.
I'm befuddled.