Meet the Other Phone. A phone that grows with your child.

Meet the Other Phone.
A phone that grows with your child.

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 3

992 replies

Barracker · 03/04/2020 18:10

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. Flowers

OP posts:
Thread gallery
56
DuLANGDuLANGDuLANG · 07/04/2020 12:27

The ONS are surely going to have to clarify?

That we, who are following this more closely than average, cannot reach the same consensus, suggests that the information isn’t being presented with and equate clarity.

We shouldn’t have to second guess and speculate on official figures.

DuLANGDuLANGDuLANG · 07/04/2020 12:30

*with adequate clarity

Should know better than to type on phone without reading specs 🤓

NewAccountForCorona · 07/04/2020 12:32

Sorry about that - I'm delighted to be wrong.

I don't understand, though, how the figures released by the ONS can be totals when they are lower than the hospital figures given Confused. Particularly as we know from the NHS figures that the hospital figures given daily (the ones that are on Worldometer) are lower than the actual hospital figures because of the lag.

I'm back to being completely confused.

LivinLaVidaLoki · 07/04/2020 12:33

@NewAccountForCorona
The nhs will report deaths with positive tests for CV, the ONS records deaths that have been registered, so they will never marry up as some may take a while to register.

NewAccountForCorona · 07/04/2020 12:38

Before all this started I thought I was quite good at maths, but I think I need a Stats 101 course. Yesterday I did a crash course (via YouTube) in Linear vs Logarithmic graphs and now some things make a lot more sense.

Apologies again; I need to read links much more carefully before posting.

Eyewhisker · 07/04/2020 12:43

I think that the only things we can say so far is that both the ONS and DoH data are an underestimate due to lags, but that total deaths year to date are not yet higher than expected, but this can be expected to change in the next few weeks.

We should expect some high figures from the DoH this week as delays in recording during the weekend probably lowered the numbers for the last few days.

WatchingTVagain · 07/04/2020 12:59

How are Sweden looking in comparison to other European countries? Are they still going for the herd immunity model?

Gutterton · 07/04/2020 12:59

Year to date comparisons need to take into account all of the deaths by RTA, other accidents (crime) which are now lower due to lockdown.

Barracker · 07/04/2020 13:00

Yesterday's numbers:

      • DAILY UPDATE * * * Monday APRIL 6th

Total UK cases: 51,608
New UK cases: 3,802
Total UK Deaths: 5,373
New UK Deaths: 439

OP posts:
BirdandSparrow · 07/04/2020 13:19

The death toll doesn't seem to be rocketing in the way I feared, thank goodness. Is that to do with these reporting lags or is the UK not doing as badly as people thought? I've seen graphs where the UK's trajectory looks really bad but the death toll doesn't seem as alarming (iyswim, obiously 400 odd in a day is absoltutely awful, 400 people), but not like Spains' recent 950 in one day. (I'm in Spain btw)

PrimalLass · 07/04/2020 13:21

I think we might get a shock today as the weekend numbers catch up. Hope not.

larrygrylls · 07/04/2020 13:25

Deaths don’t really matter as an indicator as to how the virus is progressing. It is like looking back 2-3 weeks ago and seeing how many got infected then.

If we can reduce the transmission rate below 1, it seems this will not kill a lot of people who would not have died anyway.

The important numbers are the new case numbers and total who are infected. If we can get over the peak in the first followed by the second, deaths will also decline with a lag of about 11 days.

Karwomannghia · 07/04/2020 13:30

The bbc is warning us not to take too
Much stock over daily death reports; I fear this is because there will be a leap today. There was in Scotland.

Puzzledandpissedoff · 07/04/2020 13:31

It seems the London School of Hygiene and Tropical Medicine said that the transmission rate has already dropped below 1%

www.standard.co.uk/news/health/coronavirus-infection-rate-drop-lockdown-a4404111.html

larrygrylls · 07/04/2020 13:34

Puzzled,

That is a very weak study based on surveying people.

I really hope it is true, though. If so, our case numbers and deaths would fall off a cliff in the next week or so.

Lockheart · 07/04/2020 13:40

I'm aware this is an unknown factor at the moment, but I wonder if the warmer weather we've been having over the last couple of weeks and which is forecast to last this week as well will contribute to a fall in new cases.

I know there has been some speculation that the virus won't do as well in warmer weather but pandemics don't often behave in ways we expect.

I suppose we can only hope! In the meantime I will be interested to see whether we get a post-weekend rise in the numbers as we've seen previously. If we don't then it might show that the lockdown measures are slowly having an effect.

HostessTrolley · 07/04/2020 13:41

The issue with looking at case numbers and total number of infected is that no one knows because there doesn’t appear to be any structured firm of testing...

larrygrylls · 07/04/2020 13:42

Hostess,

Yes, you can pick between an unreliable leading (or at least contemporaneous) indicator or an unreliable lagging indicator.

Lockheart · 07/04/2020 13:46

Totally true @larrygrylls and @HostessTrolley.

Until we have widespread testing among the population I don't think we'll ever know the true figures.

Puzzledandpissedoff · 07/04/2020 13:49

That is a very weak study based on surveying people

You're right, and in fact the limitations were acknowledged in the report itself
Maybe I'm clutching at straws, but as with the number of deaths not being so much higher than normal up to now, I guess it's better than being the other way round

MarshaBradyo · 07/04/2020 13:51

Case numbers are lagged too as it’s tested in hospital.

But yes very hard to get reliable data. Maybe the upcoming antibody tests in Germany will shed a bit of light on proportion of population at least.

larrygrylls · 07/04/2020 13:58

Marsha,

Yes, case numbers lag infection date by about 12 days, deaths by about 23.

There must also be a ‘second wave’ effect after lockdown where infected people infect those that they are locked down with.

So I guess it takes a while to see the ‘clean’ effect of the lockdown.

Theoretically, though, it should work and has worked in China and now, seemingly, Spain and Italy. It also looks to me that Germany and Switzerland have peaked (though I know some disagree on here).

MarshaBradyo · 07/04/2020 14:01

Larry I do think lockdown will work too. What happens to exit it is another challenge. A very difficult one given the economic pressures.

ShootsFruitAndLeaves · 07/04/2020 14:05

A lot will depend on the next 2-3 weeks but it is still very likely that total deaths this year will be the same as in other years, even with the pandemic.

There's no evidence for or that.

The median death registration time was in 2018 5 days. But longer in certain regions i.e. the South East.

Current figures cover registrations for 21st to 27th March, which is therefore roughly deaths for the 16th to 22nd March, but perhaps a day or two earlier depending on current delays.

The covid-19 deaths reported for 16th to 22nd March:

16: 20
17: 16
18: 33
19: 40
20: 33
21: 56
22: 48

That is 246.

The reported number of deaths was 11,141 vs 10,130 average, which is 1011.

So for weeks 1-13 this year, versus 5 year average

1: +79
2: +236
3: -226
4: -904
5: -594
6: -939
7: -693
8: -707
9: -367
10: -603
11: -186
12: +72
13: +1,011

We are still only at 11,000 deaths, which considering that annual deaths range between about 8k in late August and 13k in January is not that bad.

At the moment we have about 1000 extra deaths where we were expecting only 250 (or less, if you consider that this is deaths with covid-19, not necessarily of covid-19).

We will get a better picture in a couple of weeks but for now the stats are quite old.

Just to clarify I think there was some confusion:

5 deaths were registered* in week 11 (to 13th March) with covid-19 on the death certificate, 103 in week 12 (to 20th March), and 539 (to 27th March). This is a total of 647

  • for the dates 28, 29, 30, 31 March 1 April (5 days), there were 1568 - 647 = 921 deaths registered with covid-19 on d. cert, where those deceased died UP TO 27th March (some deaths will have been registered in this time who died AFTER 27th March, but these are not counted, while others who died up to 27 March will not have been counted yet). We do not know when these people died, some will be more than 5 days before, so we should be careful about these data since they are not the same as the normal release.

However if these deaths occurred 23rd to 27th March (5 days lag), then the published daily deaths were:

23 54
24 87
25 41
26 115
27 181

= 468

Hence the number of deaths with covid-19 on death certificate is almost double (921 vs 468) the declared figures.

And the really interesting figure will be 'excess deaths'. Note that these fluctuate more than some people have suggested - this is not just killing people with days to live, respiratory illnesses shorten people's lives by at least a season

To week 13:

2020: 150k
2019: 146k (527k annual
2018: 165k (539k annual)
2017: 154k (533k annual)
2016: 144k (524k annual)
2015: 160k (539k annual)

So where you have a bad winter you will have more people die that year, their lives shortened by months or even years.

At the moment we are on 150k and should expect (minus covid-19) to get around 530k deaths this year on that trajectory.

That's not going to happen, so the question is how many extra deaths we do get. Week 13 shows 1000 extra deaths vs. 250 expected (from daily stats), so essentially almost none of the 5373 official deaths have gone into the ONS death stats.

It seems very obvious that the 5373 official deaths is significantly lower than the actual count of 'people dead while covid-19+', and also that since seasonal flu shortens actual people's lives that it's implausible that we won't have higher than usual annual deaths this year.

IMO if we keep it to 600k then that will be a good outcome.

ShootsFruitAndLeaves · 07/04/2020 14:06

We are still only at 11,000 deaths, which considering that annual deaths range between about 8k in late August and 13k in January is not that bad.

this should say 'weekly deaths', not annual.

Swipe left for the next trending thread