Daily numbers, graphs, analysis thread 5

[Barracker]

Welcome to thread 5 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
Google mobility stats

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

London’s hospitals now have fewer covid patients with each day

Really good news for London. I had expected the numbers to tail off fairly evenly across regions from what ever level they had got to at lockdown. I’m wondering why London has reduced before the others? The fact it was ahead of the curve to begin with shouldn’t have anything to do with it - or at least I can’t see why it would.

BigChocFrenzy I felt myself exhale looking at those graphs. I feel like I've been holding my breath for a month.

I know the care home situation is still a hidden nightmare. But this is a start.

Different regions have different starting times, curve gradient, peaks - factors like age & racial demographics, population density -

and the effect of lockdown may work more quickly / slowly in certain areas because of this

There may also be different rates of compliance

I think of the UK curve as the summation of all the regional curves, with their different characteristics
This is partly why the UK curve has so many blips & noise
and why we usually smooth it out to a rolling average of 7 days or so

Thanks, Barracker.

Looks promising, Barracker
I wish we could access care home data and deaths at home somehow too

Hospitalisation data doesn't fit handily onto a single chart to compare with other countries

However, imo it may be the best metric we have atm

Hospitalisations don't rely on a testing regime
and they are more "connected" to the "real" number of cases

They also happen earlier in the cycle than the deaths, so give more up to date info about trends

Of course, in a situation like Italy had - or Wuhan - where hospitals were overwhelmed,
hospitalisation data is less reliable,
as too many people just won't get there

John Burn-Murdoch@jburnmurdoch (FT stats geek)

Thu 16 April update of coronavirus trajectories

Daily deaths:
• US & UK may be peaking, though recent days show it’s still too early to say
.....
cumulative deaths:
• US death now highest worldwide and still rising fast

• UK curve still matching Italy’s

Subnational region daily deaths:
• NY daily deaths may be peaking (we’re excluding nursing homes here for comparability)

• London too may be around peak deaths

My worry about the numbers in hospitals dropping is that it may be because people aren't going to hospital. I'd prefer to see every bed full if it meant that those with early symptoms of pneumonia were being watched closely.

In relation to the numbers who are showing immunity, or the numbers who have been infected, can I ask for opinions.

dd has been working front line in London. First week, no ppe, second and third week occasional masks, this week paper masks (one per shift). In all cases working with people who are either tested positive or waiting for results (mostly come back positive).

She hasn't got Covid. Nor have any of her friends colleagues (apart from one) - that is upwards of twenty or more all working in the same conditions.

However, pretty much all of them have, at some stage over the last four weeks, had a period of feeling shit, of being exhausted, of showing what might now be considered "minor symptoms".

If someone wants to test for antibodies, surely she and her colleagues - the front line staff working with Covid patients (not the ICU with full PPE, but the ordinary ward nurses - should be the ones to test. If they haven't got it, they should be immune; test them for antibodies. There's a perfect cohort of young, fit, healthy people, who have had multiple contacts with the virus to follow.

Surely some money and time could go into following them.

Oxford study criticism was always the lack of spike in deaths and esp care homes.

Derby,

I agree with your point. The reduction speed (as in time taken to get to fewer hospital admissions) should, theoretically, have nothing to do with the size of the epidemic. And, London’s population density would tend to imply a higher R0 after lockdown, so taking longer to get new cases down.

So, what is going on? I think we (as in Londoners) have been very compliant and careful post lockdown. I have witnessed this personally and it also concurs with threads on here where certain areas seemed to treat lockdown as the start of the summer holidays. That just did not happen here.

I think the reason for compliance is a combination of Londoners, on average, being well educated, and a real fear factor. Most of us know people who have had Corona and been pretty ill with it. We also knew we were ground zero and there was a real chance for infection.

I agree newcorona, i cant really get my head around the uk stats, probably because they are not clear. Howl many people have been in hospital? How many are there now? How many in intensive care? How many have been discharged? I know there are some graphs but they dont give acutual figures i dont think.

France's hosptial no.s are only maybe now coming down, and seem much higher than the uk ( cant see exact no.s in the uk) no.s in intensive care have started to come down and once again seem much higher than the uk but the deaths are considerably lower. France is furher along lockdown was on the the 17th. Figures in france are clearer for hospitals.

Not sure what this means but seems a bit strange. Are people not going into hosital? Are deaths rates much higher then?
Wish they would provide clearer information.

US reported 4591 deaths for April 16. I'm not convinced that curve is flattening.

www.washingtonpost.com/graphics/2020/national/coronavirus-us-cases-deaths/?itid=hp_hp-banner-low_web-gfx-death-tracker%3Ahomepage%2Fstory-ans

French,

I think death numbers from all countries are hugely unreliable and not helped by counting methodology being changed. France, for instance, had 17,000 new cases yesterday, as it suddenly decided to include cases in care homes. Ditto with deaths.

The most reliable data seems to be hospital admissions and how full hospitals are with COVID cases.

This will be overly pessimistic, however, as, when the strain comes off hospitals, they will be prepared to try and treat older and sicker people, keeping the numbers up for longer.

Globally, however, there appear to be about 70-80k new cases per day and about 6,000 deaths. If you look at the log curve of deaths, the gradient is clearly flatter than it was 1-2 weeks ago.

That's a good point newcorona. I suppose they haven't found an antibody as yet? So they can't confirm they've had it?

Now they're testing nhs workers they might be able to if their symptoms were mild.

But surely in places where there's been lots of testing, they have these people, and as yet no antibody test?

If the USA has a population size 5 times the UK, Italy and Spain - I wonder if the USA curve will peak at 5000 deaths a day. Ours all peaked around 1000.

I believe they have identified several antibodies but no test so far.

I know they used antibodies to treat some people in Korea and it was successful.

I'm NHS & there are clinical trials underway: ISARIC and recovery - for those with a proven diagnosis who have been hospitalised. This might include staff.

But there is nowhere near enough staff testing, either in hospital or care homes.

Until there's a handle on that,.....

China has just revised Wigan's death toll upwards by 50% to take into account all the people that died but didn't receive healthcare

Wuhan, not Wigan!

@GlassOfProsecco is there any news of there being more tests being arranged?

I can't believe a month into this and there still aren't enough tests.

What is ISARAC?

ISARIC even, sorry, letters are hard

In defence of the Oxford study, I recollect it was reported on very badly- it didn't actually say the authors thought 50% of the population had been infected: their point was that our testing/death count/general data is so poor due to under testing that technically it wasnt impossible (at that point) that the infection rate could be anything within a massive range. The media jumped onto the top of that range but the authors always said that none of them actually thought it was at 50%, they were just demonstrating that it is impossible to know.

If 20% of cases need hospital treatment then surely we can just times that number by 5 to get a idea of community numbers?

Maths isn't strong point? Too simplistic??