Saturday 21 March 2020

"Coronavirus deaths: What we don't know"

An excellent article at the BBC, which has saved me the bother of writing something like this myself.

Worth reading in full, the upshot seems to be this:

... given that the old and frail are the most vulnerable, would these people be dying anyway?

Every year more than 500,000 people die in England and Wales: factor in Scotland and Northern Ireland, and the figure tops 600,000.

The coronavirus deaths will not be on top of this. Many would be within this "normal" number of expected deaths. In short, they would have died anyway.


One of the charts in that article says that about 9% of people over 80 who contract coronavirus die as a direct result. But the point is, at that age, you'd expect approx. one person in ten to die in the next twelve months anyway.

If you compare the coronavirus deaths table with the 'chance of dying anyway' table, they are pretty much exactly the same. I would assume that the overlap is very high, so additional deaths will be negligible. Worst case, there is no overlap whatsoever (highly unlikely), in which case your chance of dying in the next 12 months has doubled. That sounds worse than it is - if you are in your 30s with a 1-in-1000 chance of dying, that's now gone up to 1-in-500.

29 comments:

benj said...

I'm glad you kicked this one off because this needs to be aired.

Mitigation (done and dusted in 3 months) = 250K deaths
Suppression (2 years of periodic lockdowns) = 50K deaths.

Current death rate 1% pa. Extra deaths suppression vs mitigation 1.2% pa

As you say, of those extra 0.2%, all overwhelmingly over 80, how many would have died anyway?

Add to which, the economic cost + the welfare(happiness) costs.

Also, is it really a good idea that the over 70s remain isolated, without visitors until the suppression strategy is over.

I don't think our politicians, under pressure from the public, have thought this through.

Mark Wadsworth said...

B, thanks and exactly. Sounds hard hearted, but that's life.

A K Haart said...

Good points - that 500,000 figure bothers me too. Gives a sense of scale to what is actually happening now.

Mark Wadsworth said...

AKH, the numbers are bonkers. If you just assume we all get it and 5% of us die, that's 3 million people. That is a heck of a lot and a national catastrophe.

But...

1. These diseases have a funny habit of suddenly tapering off, because less virulent strains develop, and they spread faster then the lethal ones (people with a mild case go about their daily business as normal, the really ill ones are at home or in hospital, infecting fewer people), so most people get the mild version and are immune to the more lethal one.

2. If you approach it the other way then death rates will - worst case - double. So it's between zero and 500,000 extra deaths, and probably closer to zero than 500,000.

Piotr Wasik said...

Apparently, 70% infection rate can give us "herd immunity". 66.4 mln (UK population) * 0.7 = 46.48 mln. 6% of them need ICU (2.8 mln). I am assuming at these levels of demand those who get it (at the moment 800 free beds at ICUs) is a rounding error, like 1 in few hundred. It is safe to say that without ICU almost all of those who need ICU will die from respiratory failure. 2.8 million extra deaths in the UK is a lot - and it is similar number to your 3 million.

I would not count on less virulent strains to develop, we would need to know how quickly new strains develop, versus how quickly the disease spreads (we know it is quick).

Also, I am not buying the overlap argument with the current mortality rate for age groups and sex alone. These mortality rates were calculated over last decades I assume and there was no comparable pandemic in the past, so coronavirus deaths would happen on top of these normal, typical deaths. Flu will keep on killing, so will heart diseases, cancers and road accidents.

I took 6% from here https://medium.com/@tomaspueyo/coronavirus-act-today-or-people-will-die-f4d3d9cd99ca - this article has a link to a github repository with plenty of aggregated pandemic parameters.

Lola said...

Being in the insurance game I'd come to the same conclusion.

Piotr Wasik said...

Also, I agree that it is illuminating to express probability of dying when I get corona relative to probability of dying due to my age (44 yrs old) in an alternative universe without corona. Now it is 0.21% per year (average between 1/663 and 1/279). It seems that if I catch corona, for my age group I have a chance of dying at 0.4% (from https://www.worldometers.info/coronavirus/coronavirus-age-sex-demographics/, not so sure - number of cases should really extrapolated from number of tests administered), so the same as a chance of dying over 2 years in an alternative universe 1 - (1 - 0.0021)^2. Acceptable or not, I want to avoid taking the hit. However, if I need ICU and I am denied it, which would happen at large number of cases, and my chance of dying is really 6%, then it takes whopping 14 years to build up to 6% in an alternative, corona-free universe.

To get 14 years, I did this:
(1 -
(1 - 0.002039983680130559) (1 - 0.002213368747233289)
(1 - 0.0024189646831156266) (1 - 0.0026666666666666666)
(1 - 0.0029708853238265003) (1 - 0.00335345405767941)
(1 - 0.0036948087936449292) (1 - 0.003937783028155149)
(1 - 0.004214963119072708) (1 - 0.004534119247336205)
(1 - 0.0049055678194751045) (1 - 0.0053433075073470475)
(1 - 0.005866823115283075) (1 - 0.0065040650406504065)
(1 - 0.007296607077708865)) --> 6.02%

So (1 - (1 - probability of not dying at the age of 44) * (1 - probability of not dying at the age of 45) * ... etc, total 14 years to the age of 58 )

The risk from a single corona infection under NHS overfill being equivalent to 14 years without corona is a lot! I am definitely NOT up to it.

Piotr Wasik said...

EDIT: " if I need ICU and I am denied it, which would happen at large number of cases, and my chance of dying is really 6%" -- I mean overall chance of dying from infection if ICU is not available, because literally what I wrote "if I need ICU and I am denied it", I believe the chance of dying is close to 100%, as I explained above. Just a clarification.

Piotr Wasik said...

and another EDIT, sorry, I am not too attentive to details now: "(1 - probability of not dying at the age of 44) * (1 - probability of not dying at the age of 45) "

should have been "(1 - probability of dying at the age of 44) * (1 - probability of dying at the age of 45)"
or equivalently
"(probability of not dying at the age of 44) * (probability of not dying at the age of 45)"

"not" was already expressed by 1 - x

benj said...

@ Piotr

This is what everyone is now working from.

https://www.imperial.ac.uk/media/imperial-college/medicine/sph/ide/gida-fellowships/Imperial-College-COVID19-NPI-modelling-16-03-2020.pdf

So your chance of dying if you get infected is 0.15%

benj said...

@ Piotr

500K deaths, if we did absolutely nothing.

Physiocrat said...

As someone approaching 80, my main anxiety is about ending up in a care home for an indefinite period, due to Alzheimers or other brain function impairment. Other risks are disability through stroke, as well as the normal risks of being disabled by being knocked down by a reckless cyclist or tripping over a paving stone. This seems to be a widespread view. So there is a lot to be said for quitting while you are still winning.

At the age of 80, the average live expectancy is a few more years (there is a figure somewhere), and a figure can be given for the risk of dying in the next twelve months. If I get the virus, that risk goes up.

If the health service is overwhelmed with a surge of cases, people will die who would otherwise have had the opportunity to receive treatment and survived for a few more years. However - a salutary tale: one of my acquaintances had a heart attack and died on the beach at Brighton, at about the age of 82. He was resuscitated and shortly afterwards had a heart valve replacement. About three years later he went into mental decline and for five years until his death at the age of 93 was distressed and confused due to a complete loss of short term memory. The kindest thing would have been to let him go when he had died first time round.

benj said...

@MW

Your last paragraph sums that up very neatly. With mitigation, instead of doubling, its 50%. The overlap must be large, say 25%.

So for one year, is it worth fucking the economy and our lives for a 25% extra chance of dying? It can't be. IMHO.

But politicians don't have the stomach for TV pictures of the elderly dying on trolleys. You and I are in the silent minority.

With the lockdowns, look out for some civil unrest.

Mark said...

These mortality rates were calculated over last decades I assume and there was no comparable pandemic in the past,

And a terrible assumption you have made there. 2009 was the last big one, a H1N1 variant.

And seasonal influenza kills a variable amount each year.

My prediction : this will turn out to be a bad year for death by illness, but nothing much more.

Mark Wadsworth said...

PW, that is all possible. But for the sake of humanity, I hope you're wrong and I'm right.

Bayard said...

"Add to which, the economic cost + the welfare(happiness) costs.
I don't think our politicians, under pressure from the public, have thought this through."

Oh yes they have. The course of action with the lower economic cost (Plan A) had a much higher political cost. Not only would the NHS been overwhelmed, increasing the number of more avoidable deaths, but every death from COVID-19 would have been blamed on the government, whether or not these people would be likely to have died anyway. So, starting with an under-equipped and under-staffed NHS, the current course of action (Plan B) is the only one that makes political sense. Sod the economy, that was going to go to ratshit anyway and sod the expense, it's not as if it's coming out of the pockets of the people in power, is it?

L fairfax said...

I do wonder if we didn't have 24 hour news and social media, if the reaction would be very different.

benj said...

https://www.nytimes.com/2020/03/20/opinion/coronavirus-pandemic-social-distancing.html

https://www.statnews.com/2020/03/17/a-fiasco-in-the-making-as-the-coronavirus-pandemic-takes-hold-we-are-making-decisions-without-reliable-data/

Mark Wadsworth said...

B, the second article points out that the data is very patchy. This is true, but it's the only data we've got.

Lola said...

Bayard. My thoughts exactly. Does not 'public choice theory' make this very point? And that 'public choice theory' also applies to state bureaucrats. Plus they are solely incentivised not to let any crisis go to waste.

Bayard said...

L, It's too late now. The "God-shaped hole" that was occupied by Warmenism is now occupied by COVID-19. You have to believe the new truth, that which the authorities tell you, otherwise you are a heretic, sorry, denier, sorry cruel and heartless person who doesn't care about the elderly.

Mark Wadsworth said...

B, re this: "The "God-shaped hole" that was occupied by Warmenism is now occupied by COVID-19"

Exactly that!

I follow some Alarmists and some Skeptics on Twitter. The Alarmists are predicting doom and disaster, calling for total shut downs and bewailing the underfunded NHS. The Skeptics reckon the whole Covid 19 thing is being blown out of proportion and will sort itself out in a few months.

Bayard said...

One thing these new restrictions is going to highlight is just how much work can be done from home. I wonder what the knock-on effect will be on the rental value of office space.

Lola said...

B - hopefully! I need to move offices in the next 21 months. A nice big rent crash would suit very well.

Mark Wadsworth said...

B, I said the same thing to the Mrs. This is one of The Stigler's favourite topics, why don't we all WFH? Maybe 'the team' can meet up once a week somewhere.

Furthermore, our IT is usually crap. Normally crashes once or twice a day. But with remote log in, it seems to work a lot better, in three days of WFH, it hasn't crashed once.

Bayard said...

I've been thinking about this and I reckon that the reason why more people don't WFH is that two days a week of their spouse and immediate family is about as much as they can take. OTOH, if anyone could do it, it would be the British, a culture to whom "social distancing" is normal - just look at a railway carriage that's only a quarter full, everyone is sitting as far as possible from everyone else.

Mark Wadsworth said...

B, our house is big enough for all four of us to have their own room with desk and PC etc. We pretty much ignore each other until 5 o'clock, then evenings it's normal family again.

But what about all the younger people sharing a house etc? Some of them at work told me that their "work space" is a basically a lap top balanced on their knees while they sit on the bed. A couple sharing a room in a shared house, which happens in London a lot, are pretty much f***ed.

Bayard said...

I had a room like that in a shared house when I was at college. There was room for a single bed and that was all. However, I worked out that the ceiling was high enough that, if I raised the bed to a level where I could walk underneath it and constructed a ladder at one end, there was just room for me to get into bed without contorting myself too much, which left room below for a chest of drawers and a desk. Also it was a lot warmer up there.

benj said...

@ MW

Not carrying out large randomised testing seems an inexplicable policy failure to me. Could prove very costly. This is what journo's should be reporting.

https://www.youtube.com/watch?time_continue=1655&v=ZEr4rmjwd0g&feature=emb_logo