Friday 10 April 2020

Weekly deaths, England & Wales, 2018 to 2020 (weeks 1 to 13)

Statistics from ons.gov.uk

I will update this monthly.

14 comments:

Bayard said...

Total deaths from all causes up to week 13 2020 = 150,057
Total deaths from all causes up to week 13 2018 = 164,625

The next four weeks' statistics are going to be interesting.

Piotr Wasik said...

Mind however it is under lockdown condition. If we let the disease run, it would be a disaster. So be careful not to arguing like: "look, we've extinguished fire in the kitchen, nothing happened. Therefore fires in houses are not dangerous."

Sobers said...

"Mind however it is under lockdown condition. If we let the disease run, it would be a disaster. So be careful not to arguing like: "look, we've extinguished fire in the kitchen, nothing happened. Therefore fires in houses are not dangerous.""

Nonsense. We've only been in lockdown for 18 days. Infection is initially symptomless, for about 5 days. Then once symptoms arise it takes about 7 days to pass. If after 7 days you still have symptoms, then thats when the trouble starts and you get hospitalised, probably a few days later. A few more days getting worse in hospital, admitted to ICU, ventilator, a few more days at least before death. Whole process takes about 20 days minimum. So almost everyone dying up to today was infected BEFORE the lock down started.

Take Boris - his symptoms arose around the 27th March. He was thus infected around the 22nd. He was admitted to hospital on 5th April, he was moved to to the ICU on 7th. He responded to treatment and was taken out of the ICU on the 9th. Thats 17 days. Had he not responded to treatment he would presumably have been put on a ventilator and may have died over the next few days. 20-ish days from infection to death.

So unless the death rate plunges dramatically after the weekend (after 3 weeks of lock down) then it has all been for nothing.

benj said...

Yep, after this is all over, the excess deaths will be the true cost of the virus and the measure of whether the projected drop of 20% GDP was worth the effort to lower it by less than half (50% of all ICU patients die anyway).

What this is really about is "our NHS" being overwhelmed and the public/media hysteria this causes. Psychologists are going to have a field day picking over this.

A news anchor I watched the other day thought the reproduction rate R for each person infected was per day, not per case. No wonder people are losing the plot.

Humans are not good at risk assessment and it's mitigation. Ask the nuclear industry.

Can we have a "fun" poll predicting the final infection fatality rate? Started off at 0.9%. I'm going for 0.2%

benj said...

What happened in week 9 2018?

Piotr Wasik said...

@Sobers, fair enough - I jumped to conclusion now; I was quoting average 17 days from infection to death myself some time ago. I was also using doubling time in uncontrolled spread as 3 days (from some Chinese data). Surely the point of lockdown is to lower infection rate so an infected individual infects on average fewer than 1 person, so the disease contracts, not expands. It would mean that doubling time is infinite. Okay, I will repeat my argument in 2 weeks then. I will write "see, if it was not for the lockdown, we would have 2^(14/3) more cases and (pick your death rate estimate) more deaths now".

By the way, what happened 18 days ago exactly? I see school closure was announced on the 18th of March, so 24 days ago. https://www.theguardian.com/world/2020/mar/18/coronavirus-uk-schools-to-be-closed-indefinitely-and-exams-cancelled

Mark Wadsworth said...

Bayard, yes, I'm going to re-run this exercise every month.

PW, well, yes and no. Point is, even fairly extreme lock downs can only slow the rate of spread, not eliminate the disease. If the disease is going to kill you, it's going to kill you sooner or later. So we have to look at the longer term picture.

Benj, we'll never know the fatality rate because we don't know how many people had only a very mild form of it and never got counted.

"What happened in week 9 2018?" Loads of stuff happened! Do a Google search and use tools to narrow the search dates.

S, PW I think the lock down started Friday 20 March, the last day pubs were open.

Everybody can have their definition of when the lock down started, but using mine, the lock down started three weeks ago.

And, your estimates of approx 3 weeks from infection to death can't be far off, so if the lock down is having an effect, we should be seeing a slow down in new cases/deaths fairly soon. The compound growth rate seems to have dropped off a bit, doubling every week instead of every three days.

Piotr Wasik said...

MW, "If the disease is going to kill you, it's going to kill you sooner or later." - not necesserily. If lockdown or at least social distancing, any measures are taken so the disease is not running wild, number of infectious people running around is much lower, (1) I may not catch it at all, or not for years at least (*), because probability of catching it is proportional to number of infectious people around. (2) If I catch it much later, medical practictioners are more experienced at handling it so I expect mortality rate to drop as time passes by - they learn even if no dedicated drug is invented. And (3), if I catch it it when NHS ICU beds are available, my initial estimate was that chances of dying are 6 times lower than if I am refused ICU treatment. For what it is worth, I think the number of deaths (8k or 9k) so far versus initial data on ICU beds available (800) when the pandemic started and still no reports of running out of beds - together it means to me that they reject people people who qualify for ICU treatment already.

(*) within years I expect vaccine to be available

benj said...

"we'll never know the fatality rate because we don't know how many people had only a very mild form of it and never got counted."

Porton Down in the UK and I believe other countries too, are conducting randomised anti-body testing. So we should know. Quite soon hopefully.

Mark Wadsworth said...

PW, yes, those are all fair points. I replied in haste. Point (4) is that these diseases tend to become milder over time.

B, yes, the base line appears to be about ten per cent. There's a weak negative correlation between number of people tested in a country and the per centage of positive results.

Mark Wadsworth said...

B... In the countries with mass testing, about ten % have it.

Bayard said...

B, AFAICS, the purpose of the mild lockdown we have in the UK is simply to slow down the rate of infection so that there are no avoidable deaths caused by lack of NHS resources. The restrictions on movement are not severe enough to stamp out the disease by containment in the way that bubonic plague was contained in the Great Plague of London.

Mark Wadsworth said...

B, that's a fair point. If they are going to do the job, do it properly. Tell everyone to stock up on enough food for two weeks and then impose a total 100% ban on any movement whatsoever (except for police, NHS and utilities) for two weeks. See what happens.

Bayard said...

If we'd been doing Track 'n' Trace like the Irish, we could have limited the restrictions to areas of known infection and made them more severe. If we are going to behave as if COVID-19 was the bubonic plague, we might as well use the same methods as was used against the bubonic plague and worked.