I want to ask a question to which I really do not know the
answer. It may sound on first hearing
like a heartless question, but it is in fact exactly the opposite. It is a question asked out of anxiety
bordering on terror. Here it is:
In 2018 there were about 2.8 million deaths in the United States
from all causes, and I think the deaths in 2019 were comparable. Obviously, many of the people who died were
old people like me and everyone else who lives where I do. Dr. Fauci has offered the informed
guesstimate that between 100,000 and 200,000 Americans will die from
COVID-19. Does that mean that he expects
the 2020 death figures to be 2.9 million or 3 million rather than 2.8 million,
or are some of those expected to die from the disease people who, statistically
speaking, could have been expected to die from other causes?
Second question. By
the time we are done with this disease, it is estimated by many experts that
between 30% and 60% of Americans will contract it, which is to say between 100
million and 200 million people.
World-wide, the mortality rate for the disease seems to be 1% to
1.2%. If that is right, then we could
expect between one and two million Americans to die from COVID, not between one
hundred and two hundred thousand. Does
that mean that 2020 will see 3.8 to 4.8 million American deaths rather than 2.8
million?
Does anybody know?
21 comments:
I think one reason Italy is doing so poorly with regard to deaths is that the elderly population in Italy (over 65) is 23% of the population as compared to 16% in the US. Also, in Italy, there is much more intergenerational activity in closer quarters as well than in the US.
https://www.wired.com/story/why-the-coronavirus-hit-italy-so-hard/
So my guess is that in the US, most covid deaths will be "in addition to."
That is really acary, Jerry.
Just one short comment: I don't think the mortality rate for COVID19 is known w any certainty yet b.c for one thing the number of asymptomatic cases isn't really known. There seems to be consensus that it's higher than the seasonal flu but whether it's 0.5 percent or 1 percent or whatever is not clear. Still scary though.
We've only had 7 deaths from corona virus in Chile so far.
I may be wrong, but I believe all were over 65. I don't have the exact figure, but most of them had other very serious illnesses, for example, cancer in several cases. I'm not sure what the recovery rate for people over 65 without other very serious illnesses are, but recovery from corona virus is much likely for those with a pre-existing medical condition.
Here where I live (in the center of Santiago) we can only leave our homes (mine is a small apartment) to go to buy food, to a pharmacy, to a bank or for a doctor's appointment. We need to get permission to leave our homes for the above reasons: permission is available online, but those without internet have to go to a police station to get permission: as usual, life is harder for the poor. There are no permissions to take a walk or to jog, but you can get one to walk your dog. I have no dog, but so far I've gotten a permission daily to buy food and I walk to the supermarket and back (anyway, I don't have a car).
These strict measures are not in force throughout Chile or even in all the metropolitan area of Santiago, only in certain areas such as the one where I live.
my error:
in my first paragraph it should read "recovery from corona virus is much less likely"
Keep in mind that ICUs are typically used to save lives at risk from other, non-Covid 19 conditions. As ICUs are overwhelmed with Covid 19 patients, other conditions which in this day and age do not regularly end in death will end in death.
Along these lines, Josh Marshall posted about a news story on Bergamo: non-covid 19 deaths in Bergamo Italy in the first quarter have been running about 3-4 times the normal level when compared to the last few years. Presumably some of these are covid 19 deaths that were not identified as such while others were non covid-19 deaths from causes that do not typically result in deaths under normal circumstances.
US may not have as elderly a population as Italy, but it has the highest rate of obesity in the world which is associated with other underlying conditions that may severely complicate CV19 recovery.
Even the worst case number of total deaths seems to be very optimistic at this point given the preparedness of the medical profession.
I don't think anyone really knows right now- the data just isn't adequate (or at least the public data). There hasn't been enough testing even of some who have symptoms and as of yet there hasn't been anti-body testing that could give us an accurate idea of how many may have already had it and completely recovered. Or even if once a person got it if they would be immune to subsequent infection.
My brother the MD was just here visiting and I asked him to read your question. He runs several urgent care clinics in the area and has had several patients test positive. His opinion is that many people including himself have already had it and recovered. But there is no way right now we would be able to tell that. So there really isn't an accurate estimate of the mortality rate from this thing. He doesn't think a million Americans will die from this but again it is too early to say anything for sure. Because we haven't had the testing. Just can't answer your questions until more data is available.
Hi Bob. For the first question I suspect many of the people who end up dying from covid would have passed away this year from other causes. The reason is that the danger of covid increases with age so that the oldest people that are at risk from covid are (relatively) likely to die from other causes too. To get a handle on it you can probably find statistics on "Of all Americans of age A what fraction will die in the next year". That seems like it would be a standard actuarial quantity that various agencies would like to keep track of. That would be an estimate of the chance that a covid victim would have died from something else this year. It might also be enlightening to compare with some expectation for covid deaths in the same age group: (fraction of americans of age A)*(probability of catching covid)*(probability of dying from it).
The other thing I would do is pull up the total mortality numbers from the past bunch of years (say ten). How much do these numbers fluctuate up and down? The year-to-year scatter should tell us whether 200,000 is a drop in the bucket or a significant additional cause of mortality compared with everything else.
For the second question, I think the 30-60% is the naive estimate if there was NO intervention. I.e. if each case generated 2-3 new cases you get exponential growth and it would quickly affect almost the entire population (you can read about "SIR models" on wikipedia). The point of the lockdowns and distancing is to halt the exponential expansion (as in China) until such time as widespread testing can isolate cases to extinguish pockets of outbreaks or a vaccine can be produced.
Anyway, my feeling is that all of these models give numbers that should be thought of as coming with huge huge uncertainties. These models all involve exponentials and we know that tiny changes in initial conditions or parameters can change the predictions by huge amounts. I.e. (1e5 cases)*exp(0.23 increase/day * 30 days) will be 60 million cases more than if the 0.23 were 0.2, and there's no way people know that number to such accuracy. A single number like 200,000 is useless on its own. When jiggling the parameters in the models you can change the expected number of deaths by orders of magnitude I'm sure (in either direction).
You can only control so much and maybe it's best not to worry about the things we can't control. I think a lot of us can't help indulging in these morbid calculations. But it sounds like you're doing the right things now to keep isolated. I see no reason why that wouldn't that be effective to keep you and your wife safe (given what we know about how it spreads).
I'm going to suggest that the best source for statistics on the likely death rate from covid-19 is South Korea which instituted a vigourous & successful testing regime from fairly early on, as also a vigorous program of tracing and isolating potential victims. They have probably detected a very high proportion of those infected. South Korea also has a reasonably good health service. In South Korea they have a death-rate as compared with verified infections of nearly 1.6%. Because of the vigorous measures they took, the health system has not (yet) been overwhelmed. This means that people with Covid-19 got more-or-less adequate treatment, as did people suffering from *other* life-threatening complaints.
Now let’s make an assumption which in the context of my current argument is highly optimistic. Let’s say that the South Koreans have undercounted their infections by 60%, that is that the number of people with Covid-19 in South Korea is 60% more than the official count, these cases presumably being asymptomatic or minimally symptomatic. That gives us a death-rate per infection of 1%. But note that this is the death rate with a functioning health system in which Covid-19 victims and other people with other life threatening complaints can get adequate treatment. This would not be the case if an already inadequate health system was overwhelmed by an influx of Covid-19 victims. Many more people would die for the lack of adequate treatment, not just the victims of coronavirus, but patients with other life-threatening conditions who could not find either a bed or a doctor. In that case the death rate would probably jump to two, three or even four percent, with a much higher death rate among the sick and old. Besides this there would be a good deal of ‘collateral damage’ due to an overstretched health system’s being unable to save a set of acutely sick people who under normal conditions would be perfectly salvageable. Now the question is , ’Is that likely to happen in the USA?’ And the answer is *Yes vey much so.* There are probably two or three times as many people with the virus as the number of officially confirmed cases and there is community infection in virtually every state. So unless an extremely strict lock-down regime is enforced almost everywhere, the chances are that in the course of this year about two thirds of the population gets the virus. That’s 2/3 x 329 million = 219 million. With a 1% death rate that gives 2.19 million deaths. With a 2 % death rate (much more likely given the inadequacies of the US health system) that’s 4.2 million deaths and that’s not counting the collateral damage from diseases and injuries that are likely to go untreated. In this connection it is worth checking out the following article: Coronavirus: The Hammer and the Dance; https://medium.com/@tomaspueyo/coronavirus-the-hammer-and-the-dance-be9337092b56 What it says in effect is this. The UK and in a more half-assed way the US are following what the author calls the ‘Mitigation strategy’ (See Chart 5) . That strategy could lead to over a million deaths in the UK alone. Currently 1 in 3 Americans remain under state or local government orders to stay at home to slow the spread of the virus. That means the two thirds are not. In New Zealand *everyone* has been mandated to stay at home except healthcare workers, postal workers, the police and people in the food industry which basically means supermarkets and their suppliers. (Fast food outlets are shut). Suppose I am being pessimistic and the current social distancing policies in the USA work a lot better than I suspect they will. Let’s suppose that only a third of the US population gets infected. That’s 109 million. With a death rate of 1% that’s 1.09 million. But the death rates will almost certainly be higher, at least 2% . That’s 2.18 million.
Okay let’s get back to Professor Wolff’s question. It is abundantly clear that those who are sick and old are much more likely to succumb to the virus that the young and healthy, and a proportion of these would have died anyway, virus or no virus. But those deaths would have been smeared out over several years. As an indication of how much higher the death-rates are given the pandemic than they would have been otherwise there is this from Trump’s latest press briefing:
Trump was clearly moved by the scenes from New York, particularly hard-hit Elmhurst Hospital in his native Queens. “I’ve been watching that for the last week on television,” he said. ”Body bags all over, in hallways. I’ve been watching them bring in trailer trucks — freezer trucks, they’re freezer trucks, because they can’t handle the bodies, there are so many of them. This is essentially in my community, in Queens, Queens, New York,” he continued. “I’ve seen things that I’ve never seen before.”
Now this is impressionistic . But if they are having to bring in freezer trucks and trailer trucks to dispose of bodies, *while the infection rate is still comparatively low*, this suggests that the death rate from the virus if it really took hold would be vastly in excess of the death rate for old sick people who would have died anyway in the absence of a pandemic. The normal death-rate does not require freezer trucks to dispose of the bodies. While the casually rates are sill relatively low, freezer trucks *are* necessary. So, once the virus really takes off the death rates will be much, much higher. How much higher? I don’t know. But the extra deaths will probably be calculated in millions.
Huge mistake it was to read your comments before trying to sleep Charles. I will try to relax by remembering you are creating a model with data that could possibly be off by a factor of 10 if we are really optimistic rather than just 60%. We don't have the data we need to project this very accurately. But certainly we need to do all we can in the meantime.
Well at least I don't have to get up for work in the morning now that the virus is here.
This appeared today in the Washington Post; it's a copy of a mirror, mirror on the wall press release via twitter that Trump composed himself, about himself, and cites the numbers he really cares about:
“President Trump is a ratings hit. Since reviving the daily White House briefing Mr. Trump and his coronavirus updates have attracted an average audience of 8.5 million on cable news, roughly the viewership of the season finale of ‘The Bachelor.’ Numbers are continuing to rise...
— Donald J. Trump (@realDonaldTrump) March 29, 2020
Well you heard it from Dr Birx.
Premise 1) [Birx] 200,000 deaths is the minimum that the USA can expect if everybody self-isolates efficiently. .
Premise 2) [Unstated but pretty damn' obvious] Everybody won't self isolate efficiently
Conclusion;
Hence there will be many more than 200,000 deaths.
And that's the news from Happyland (aka the Whitehouse]
Meanwhile in my native UK, the government is run by a parcel of clowns who have managed to give one another coronavirus.
A couple of observations:
- There is recently more reporting on deaths among younger folks, which appear to be higher than expected.
- I think that there will be continued transmission of the virus until herd immunity levels are reached. I understand that to be 80%.
- There is the possibility that the virus will hit the southern hemisphere, mutate, and have a second run at the northern hemisphere states. This is
what happened in 1918. The flu started in 1917, mutated and came back with a vengeance.
- how things proceed will depend on the actions of state governors, specifically republican governors. Florida governor DeSantis just instituted social distancing regs but excluded churches. There was a fundamentalist church service in AZ on March 7. Many Navajo attended and as of today, 4/1, there are 174 covid 19 cases, and 7 deaths in a population of 200,000.
oops, I made a mistake: population of the Navajo Nation is 357,000.
One success story to emerge from this is Taiwan, which took early and vigorous measures, albeit apparently involving a level of govt intrusion that might not have been feasible and/or tolerated in some other places. The NewsHour (PBS) had a piece on this yesterday.
Post a Comment