> would change the output by tens of thousands of deaths?
Yes, with the modification of the statement to "I would not be surprised if it changes by many percentage points."
The bug report you linked to says only "This mostly manifests as a time- offset, but there are numerical changes beyond that." It shows a different by day 80 of 100,000 deaths, or 20% overall. It does not show if by day (say) 120 those numbers have gotten further together or closer.
If they both end at about the same numbers, separated only by a week, then it does not change the overall conclusions in the paper at https://www.imperial.ac.uk/media/imperial-college/medicine/s... . It does shift some of the dates, and I am annoyed with the lack of error bars or the like in their plots.
You asked "how could that model possibly be useful, given that none of the inputs are known to anywhere near that?" You check it for sensitivity across a range of values and see how the overall conclusions change.
Quoting that paper I linked to, "Such policies are robust to uncertainty in both the reproduction number, R0(Table 4) and in the severity of the virus ... However, there are very large uncertainties around the transmission of this virus, the likely effectiveness of different policies and the extent to which the population spontaneously adopts risk reducing behaviours. This means it is difficult to be definitive about the likely initial duration of measures which will be required, except that it will be several months."
When all the lines in a spaghetti map show the hurricane bearing down on you, you prepare for bad weather, even though the hurricane track is almost certainly not going to follow any of those paths.
You write "but the numerical inaccuracy is many orders of magnitude below that."
What I know about chaotic systems is that I cannot make that conclusion without much more study about the relevant phase space. If the end result is an attractor, then even intermediate chaotic behavior is not much of an issue, depending of course on the specific details of the policy guidance.
You write "I think both are true, so I'm not sure what we're arguing about?"
My comment started because of your comments at https://news.ycombinator.com/item?id=23213711 where you wrote that discussion on the GitHub comments wasn't "terribly clear" to you. I gave my interpretation. This specific thread started because you wrote "They seem to think the model is fine and just the "deterministic from seed" mode is broken", while I don't see in that bug report where they state the '"deterministic from seed" mode is broken'.
Is the GitHub commentary clear to you now, and if not, what is unclear?
Two of the most influential models in shifting public policy have been Ferguson and the IMHE's model.
> We do not consider the ethical or economic implications of either strategy here, except to note that there is no easy policy decision to be made. Suppression, while successful to date in Chinaand South Korea, carries with it enormous social and economic costs which may themselves have significant impact on health and well-being in the short and longer-term. Mitigation will never be able to completely protect those at risk from severe disease or death and the resulting mortality may therefore still be high. Instead we focus on feasibility, with a specific focus on what the likely healthcare system impact of the two approaches would be.
(From Page 4)
Now let's shift to the IHME model. They do make passing reference to the econonic impact, same as Ferguson, but don't go any further:
> The overall financial cost over a short period of time is likely to be enormous, particularly when juxtaposed against the substantial reductions in revenue for many hospitals due to the cancellation of elective procedures and the broader economic consequences of social distancing mandates.
Now as I said, it's not necessarily their job to forecast the economic harm. But unfortunately we have created this notion that being pro-lockdown means "believing science" and therefore thinking that the lockdown is a bad idea is being "against science". Our politicians use these exact words, and as I said their actions show that they are not holistically evaluating the downside risk. On the contrary it appears to be a very simple game-theory type calculation where their incentive structures are leading them to make irrational decisions. COVID-19 mortality is much more "visible" than the fuzzier and longer-term mortality caused by our (IMO misguided) response to COVID-19.
Let's back up a bit to talk about the two strategies.
(1) Containment lets you indefinitely avoid COVID-19-induced mortality in the short-medium term, at the expense of ongoing, mounting costs to wellbeing and the economy. These costs are certainly non-linear, for example businesses can generally only survive a given number of days/weeks based off their capital expenditure and thus it's not quite as simple as a linear relation. But for our purposes, it's easiest to think of the wellbeing and economic cost as being in direct proportion to how long we spend in containment.
The postponement of mortality only becomes the true avoidance of mortality when we get a "game-changer": a vaccine or a highly effective treatment that seriously improves outcomes.
Given that we must practice indefinite containment until we develop the "game-changer", we are executing a strategy which is based off a temporally unbounded future event. Therefore the potential drawbacks are unbounded given that the strategy involves waiting for a miraculous leap forward in COVID-19 vaccination or treatment.
(2) My proposal is an approach where we try to direct testing resources and governmental assistance to protecting the most at-risk members of society. These groups are encouraged to shelter at home and are supported in doing so.
Bans on freedom of movement, transaction, etc are lifted. Non-at-risk individuals are encouraged to return to work. Given that we inflicted psychological harm on millions of individuals, we also would probably want a policy where someone is allowed to not work, but they must formally quit their job in order to be allowed to collect unemployment for up to a year (we likely also need to adjust unemployment because it's just way too high relative to wage earners right now). What we need to avoid is a case where someone "chooses" (in scare quotes because we have done true psychological damage to people) not to work for a year but their company can't let them go, since otherwise the company cannot replace them with a working employee.
So in short, we let people do what they want, we strongly encourage the at-risk to shelter at home and put out appropriate public health messaging in proportion to the real risk (which means overall WAY less fearmongering since we're so out of whack currently).
The uncertainty benefit is something we should implicitly factor in as well. The ultimate end state of my proposal is much more "known" than with containment (because we have no bound on containment worst-case scenario but we can use Ferguson to get a decent bound for mitigation). We're not sure how much mortality we will see, but with a 0.9% IFR and 82% of the population being infected we get about 2.2 million deaths per Ferguson (https://www.imperial.ac.uk/media/imperial-college/medicine/s...). I think that's a great upper bound to use.
BTW I think once accounting for vector exhaustion (not everyone has the same risk of infection) and what I feel is a more realistic IFR, we'd be closer to 600,000 deaths in the "realistic" scenario IMO.
The last thing I want to say is that I actually think in terms of wellbeing-years or quality-adjusted-life-years, and not just "lives". My belief is that the life of a healthy 12 year old is several times more valuable than the life of an 80 year old with heart disease, to use an example.
So, while it's hard for me to give you a "real" number, I'd say if we could save 20 million wellbeing-years, then lockdown was probably worth it. But keep in mind that means that LOCKDOWN_COVID19_MORTALITY_REDUCTION - LOCKDOWN_EXTERNALITIES >= 20 million wellbeing-years.
The fact that he changed that without taking into account any of the other flaws in his work suggests that he is not, actually, open to criticism of his work.
Here are some examples. He shows charts estimating per age range fatality rates and leaves out the most vulnerable group. He estimates an IFR and thinks that is comparable with a CFR. (They are not.) He is apparently unaware of the fact that IFRs have long been thought to be a bit under 1%. (For example https://www.imperial.ac.uk/media/imperial-college/medicine/s... in mid-March was the paper that convinced the UK to do a lockdown - it used an estimated IFR of 0.9%.) He does his own IFR calculation for the least vulnerable groups without looking at research showing the full IFR. He complains about hospitals being underloaded while refusing to acknowledge that hospitals would be overloaded without lockdowns. He fails to admit that it is extremely hard to limit transmissions within managed care facilities, and in an environment with lots of COVID-19 around you, personal distancing measures provide very little protection. (In other words we cannot simply "isolate the old people" and expect it to work.)
In other words he is ignoring every widely known fact that undermines his position. Which not coincidentally are the facts that lead everyone else to the opposite conclusion from his.
Given that he can't fail to have seen many points like these, his leaving them out of his analysis shows dishonesty on his part. And no, he is not attempting to correct this fault.
The US beat every single expert pre-peak total deaths prediction; by a lot. There are supply chain issues to fix, and the few spots that could not deal with the case load need to be fixed, but by any scientific measure against the predictions, we did excellent.
p16: "In the most effective mitigation strategy examined, which leads to a single, relatively short epidemic (case isolation, household quarantine and social distancing of the elderly), the surge limits for both general ward and ICU beds would be exceeded by at least 8-fold under the more optimistic scenario for critical care requirements that we examined. In addition, even if all patients were able to be treated, we predict there would still be in the order of 250,000 deaths in GB, and 1.1-1.2 million in the US."
Like this one?
Is 1.1M is "holding up well"?
Have you read the Ferguson paper? From the re-tweet you posted, I doubt it. The 1.2 million figure was the absolute highest estimate in the whole paper, and was mentioned off-hand as a worst case scenario if no preventive measures were taken.
In fact, if you read his paper and look at his estimated deaths under various different interventions and time period, and the current death toll, he actually may have underestimated the death toll.
Here, read the paper: https://www.imperial.ac.uk/media/imperial-college/medicine/s...
As best I can tell, the Imperial College report (https://www.imperial.ac.uk/media/imperial-college/medicine/s...) got its numbers from this study: https://www.medrxiv.org/content/10.1101/2020.03.09.20033357v...
The derived IFR is still <1% (0.66%), but I'll admit it seems unjustifiably high given the inconsistent reporting out of China. It would have been better characterized, at least qualitatively, as somewhat pessimistic, given all the assumptions.
>Only around 10,000-20,000 otherwise-healthy would die even in the worst case
>Nothing much, since 98% of those "millions" would be folks already above age 75 and with 2+ serious underlying conditions.
This is at odds with what actual experts say and actual reality as observed by people on the ground now.
Non-mitigation was projected to result in 2.2 million fatalities in the US and to date 46% of fatalities are under 75. This would be 968k under 75. We would in fact lose 75k under 44. 23k under 34.
>American meat processing has been permanently closed already.
How is it "permanently" closed?
>We are looking at 100 million or more dead due to starvation in the coming year,
Where are you getting this number? The labor required for our agricultural industry is much smaller than the total labor pool. Why are you assuming that we can't reasonably discover enough people to keep this minority of our economy functioning enough to keep everyone from starving? We could run the sector with the military if we had to if the alternative was everyone dying. We can also run it with people that have already had the disease and come through it. On the one hand you predict that almost all young people come through it just fine and on the other you are predicting that we can't find enough young survivors of covid to run the minority of our economy for our population to survive. Which is it?
We cannot make life and death decisions based on your unsupported and unsupportable fancies we have to deal with real numbers, projections, and models.
In reality letting it "wash over us" like you and the president want would have killed millions including far more of otherwise healthy folk than you imagine. Digest more information and come to better conclusions.
Oh, it'll be more than two waves if we're lucky. See figure 4 from the the Imperial College paper.
Note that that paper assumes R0 of 2.2. A paper in CDC's Emerging Infectious Diseases written by researchers at Los Alamos National Laboratory shows evidence that the real R0 could be as high as 5.7.
This guy's argument is wildly reckless, he's basically advocating for the rejected "herd immunity by infecting everyone at once" strategy that would've killed millions. He gives absolutely zero thought to the health care workers who are already dying en masse even with a quarantine strategy.
This is such an outrageous thing to say. Do yourself a favour and go read the paper:
Researchers go and model data, making every caveat imaginable along the way, and then predict outcomes based on possible actions taken, and your response is to label them as spreading "bad" and "dangerous" information. You're literally condemning them for doing the science at all. You're inadvertently screaming at models...
The virus will kill 2 million in the US if nothing is done and only 100k with shutdowns. But note that we may need 400 days of shutdowns over a 2 year period. 
Economies with good human capital recover after disaster, like the Japan and Germany did after WWII.
: See Page 7 and 13 https://www.imperial.ac.uk/media/imperial-college/medicine/s...
: Second paragraph of page 15
There are alternatives to locking the whole population down: For example the Imperial College study predicts less deaths if only the 70+ population is locked down.
Also look at page 10: If you lockdown for only 5 months, the virus just comes back and still kills 1 in 500 people.
If you want to minimize the deaths, the study suggests that you need 400+ days of lockdowns over a 2 year period.
Let me rephrase it: If governments does nothing, 1 out of 200 people will die.
This study expects 510,000 people to die in the UK (depending on the R0 value) and 2200000 in the US. That's 1 in 135. https://www.imperial.ac.uk/media/imperial-college/medicine/s...
It will be between 1 in 200 and 1 in 100.
It’s not relevant; yes, it is wrong. Firstly, we have a lot more than the one datapoint. Secondly, we can look at what’s happened in other countries (like China). But, most importantly, the models don’t work like this, they’re more like a simulation. Take a look at the “Methods” section of this report, for example. It’s quite easily readable.
This is the report credited in the media for changing the government response in a number of countries
The analysis that is claimed (1) to have affected the change in the Covid-19 policy of the British government:
is by Neil M Ferguson et al.
16 March 2020 Imperial College COVID-19 Response Team
"Impact of non-pharmaceutical interventions (NPIs) to reduce COVID-19 mortality and healthcare demand"
In my opinion, the experts were aware of the problems in approach of many governments, but the governments managed to do what they wanted (pretend nothing is happening) for too long. Also in my opinion, the smaller players believed too much to the most influential one, which has a specific history of bluffing. That even works in social contexts, but, in the words of Richard Feynman eventually:
"reality must take precedence over public relations, for Nature cannot be fooled."
1) E.g. The Guardian: https://www.theguardian.com/world/2020/mar/16/new-data-new-p...
I'm not an epidemiologist, or an MD, so with a large grain of salt:
Containment: Testing & contact tracing - you try to contain the disease before it widely spreads. Usually one of the early stages of fighting.
Mitigation: You can't contain any more, and you're trying to slow down the progress to avoid a large peak. Test & treat those with severe symptoms, encourage people with mild symptoms to stay home, encourage people to keep distance.
Suppression: Things have hit the fan. You need to drastically halt the progress of the epidemic. This is shelter-in-place, lockdown, quarantine etc. #staythefuckhome has become a bit more mandatory. That's pretty much where we are right now. You want to drastically reduce the number of infections in a short amount of time.
"Meaningful supply" was in the context of suppression actually taking hold. At some point, you're hopefully down to illness levels where containment or mitigation make sense again. But for that to happen, you need tests, you need PPE, you need infrastructure so you actually can contain. We're at suppression/lockdown because we failed at that the first time round.
So, it's not about avoiding the lockdown now.
The goal is lockdown now to prevent catastrophic overload and buy time to get supplies in place for later containment stages.
Hope that clarifies? But, of course, containment is not guaranteed to work, so we might be cycling back and forth between those measures
The report #9 from the Imperial College of London details the ideas behind that cycling approach: https://www.imperial.ac.uk/media/imperial-college/medicine/s...
> Infection rates are already significant, there just aren't enough people to sustain two months' growth.
This is wrong, of course there are enough people for this to grow for months even with no containment efforts.
Well, it sort of did:
Hence the change in tact.
e^(ct)/e^(c(t - 14days)) = e^(c*14days) = k = constant, with c the growth rate. While you are right that, in an exponential growth scenario, the number of recovered cases is smaller than the number of currently active cases, the factor is proportional to the growth rate which is worst case constant, but likely going down over time due to quarantine measures. As such we have theoretically enough recovered cases if n_infected/n_critical > k. For the most critical age group an estimated 10% of infected require ICU care . Thus we need k < 10. With a 1:2 ratio for serum we get k < 20. Likely not everyone will donate though and constant factors in this equation will end up mattering but the exponential nature does not kill the idea in the manner you indicate. However, it still requires scaling production exponentially. This is true for all other drugs as well though.
> Which experts are proposing the way forward is actually a total lockdown?
The Imperial College COVID-19 Response Team, for example: https://www.imperial.ac.uk/media/imperial-college/medicine/s...
In Germany, the Society for Epidemiology has a similar analysis with similar conclusions: https://www.dgepi.de/assets/Stellungnahmen/Stellungnahme2020...
The Imperial College report found suppression to be effective even with much less compliance. Here are the different measures they modeled and the assumed compliance, taken from the table on page 6 of the report and edited out the info about assumed impact on transmission, for brevity.
> Case isolation in the home — Assume 70% compliance — Symptomatic cases stay at home for 7 days. Household contacts remain unchanged.
> Voluntary home quarantine — Assume 50% compliance — Following identification of a symptomaticcase in the household, all household members remain at home for 14 days.
> Social distancing of those over 70 years of age — Assume 75% compliance
> Social distancing of entire population — Assume 75% reduced contact outside household, school or workplace (the wording here is different, does not mention "compliance")
> Closure of schools and universities — 100% of schools, 75% of universities
I would recommend reading the report. Unlike certain summaries, the report itself is very clear about all the assumptions baked in, and so you won't get a false sense of "This is 100% how it will be". (That said, I'm not questioning their numbers since I think they're likely more accurate than any I could come up with myself).
We can take the age profile of Venezuela and the estimated hospital rates and general virus profile from the Imperial College report.
Assume 80% infection, and assuming that 100% of people who would have needed hospitalization die.
This is 720,000 people, or 2.5% of the population. However, if we just take the working-age population (65 and lower), and assume 50% survival rate instead of 0%, the deaths are just 0.79% of the total population.
So decimation is absolutely inaccurate and the higher-up poster is correct. Developing countries like Venezuela cannot afford to and should not shutdown their economies or societies.
Here is a report from a team of epidemiologists at Imperial College, led by Neil Ferguson. The title speaks for itself: Impact of non-pharmaceutical interventions (NPIs) to reduce COVID- 19 mortality and healthcare demand. The paper is sobering reading. But of course, all the numbers plugged into the model are saddled with huge uncertainty.
There is no need to imagine, we have the forecasts from the Imperial College report:
There are 40 million 40-somethings in the USA, you're proposing that 100% of them will have lung damage.
Let's imagine that 80% of the population is infected, that 80% of cases are symptomatic, and that 80% of the 4.90% of hospitalised cases (this 4.90 is direct from the report) get damaged lungs. This is one million 40-somethings, not the entire population.
1 million sounds like a lot, but its only 2.5% of the population at that age range, and assume worst-case.
You know what causes reduced lung functionality in adults? Asthma, driven by air pollution from vehicles and coal power plants. 7.7% of adults have Asthma. Where is the war-scale drive to eliminate fossil-fueled vehicles and power generation?
The Imperial College study, which seems well-received, and which caused the UK government to change strategy, estimates a hospitalisation rate of 4.4%.
This is on page 5 of the paper.
US DHHS: "18 months or longer" https://int.nyt.com/data/documenthelper/6819-covid-19-respon...
UK government: "at least a year." https://news.sky.com/story/coronavirus-social-distancing-nee...
Modelling: "2 months on 1 month off until a vaccine is developed". https://www.imperial.ac.uk/media/imperial-college/medicine/s...
It's unprecedented and I tend to agree with you that at some point the economic carnage is going to force alternatives, but right now that's the plan. Not the plan they're telling you publicly, but the real one they're planning for and reacting to.
At the end of the day if that's the plan you'll go along with it because the men with guns will make you.
As I've commented elsewhere, an approach using contact-tracing via app and GPS data, backed by widespread testing (millions per week) and small-scale isolation of affected, along the lines of China or South Korea, seems to be more effective.
https://twitter.com/CarolYujiaYin/status/1239583581325778944 (deffo not propaganda /s)
The US population is going to hate hate hate that one. But it's that or stay inside except for going to the grocery store for the next 12-18 months.
If we don't choose option A, and we don't choose option B, we implicitly choose option C: 20% of our seniors die in a single year. Choosing not to choose is still making a choice. I believe there's a song about that.
Well, never, but we have 20 ventilators per 100k people, "flattening the curve" so strongly that we can treat everybody was always a bit of a pipe dream. You need to flatten the curve almost to zero to get to a "sustainable" level of disease. They are nearly the same goal. Any non-trivial level of disease that is not controlled will rapidly push you past the "flatten the curve" threshold. And we will certainly be here for longer than 2-3 weeks.
It is a little white lie that is important to tell the public to help build compliance, the lower you can push it the fewer people will die, but hospitals simply are going to be overloaded regardless. The choice we can make is one between "very overwhelmed" and "overwhelmed to the point of non-functionality".
"shelter for 12-18 months" is a little bit of an exaggeration but not much. Simulations suggest that after an initial harsh quarantine period (up to several months) to get transmission under control, we will need to shelter for about 2/3rds of the time (2 months on, 1 month off) for 12-18 months, and other measures for 100% of the time. The vaccine will be what finally terminates the situation. https://www.imperial.ac.uk/media/imperial-college/medicine/s...
Even then that may be optimistic (especially if the population is not fully compliant). Italy is not just working from home and social distancing, they are straight up quarantining everyone, and they still have overloaded hospitals and bodies piling up faster than they can cremate them. Presumably the quarantine will slow things down on the order of weeks to months, but right now things are brutal.
The Chinese strategy of requiring everyone to get a phone app, to track them via GPS, doing widespread testing and then quarantining the people who have been in contact appears to be a much more effective strategy than blind quarantine or social distancing. No app, no uber, no groceries, no public transit, etc. The US citizenry absolutely will not stand for the mark of the beast though.
The Korean thing about masks and gloves probably helps too. Too bad the hospitals are down to about 11 days of masks let alone for the citizenry.
I'm not trying to be a downer but if chloroquine works that's a really fucking big deal, because that would give us a fairly straightforward alternative to treat cases or even prophylactically treat vulnerable populations. Unless you have a tool in your toolbox we are mostly just moving the deck chairs here and choosing between "really bad" and "oh shit".
Vaccine estimate is 18 months according to the Imperial College:
Vaccines in human trials in the US now need to go through a 14 month observation period to understand effectiveness and side effects.
Influenza vaccines, if they are any guide, are only 50-70% effective and need to be renewed annually.
Ask people if they are willing to sacrifice 18 months of their life so that someone with an average age of 79 and existing health complications can live another couple of years and you will start to deal with the problem honestly.
By my calculations for every 1 year of saved life (average age 79) we are paying for it with 10 years of lockdown life (average age 35-40).
The report from Imperial College  indicates that it may be impossible to flatten it enough for it not to overwhelm medical care capacity.
It wasn't (and still isn't) wrong to try to flatten it - it's what gives time to investigate the appropriate course of action. But unfortunately it seems like flattening won't be enough and more extreme measures (called "suppression" in the report, whereas flattening strategy is called "mitigation") need to be taken.
The report also notes that such extreme measures have never been taken on the timescale they anticipate to be required to successfully suppress the disease (basically until the cure or vaccine is found and widely distributed - at least 18 month and possibly years), so we have no idea how that will turn out.
What you cite is from: https://www.imperial.ac.uk/media/imperial-college/medicine/s...
topic of which isn't the evaluation of the recorded growth across the world, but to show that lax measures are problematic even using the slower growth rates.
They specifically write just:
"Infection was assumed to be seeded in each country at an exponentially growing rate (with a doubling time of 5 days) from early January 2020, with the rate of seeding being calibrated to give local epidemics which reproduced the observed cumulative number of deaths in GB or the US seen by 14th March 2020."
That's what they use in their model to demonstrate the point.
They never claim it's accurate or quote any sources for that number, because they just use that very "optimistic" value for the demonstration purposes: to show that even with that number, it's not reasonable not to implement serious measures. That's how mathematical proofs are constructed: you construct the lowest bound which doesn't have to be accurate, just to be obviously lower than the actual numbers, and show that even then some assumption doesn't hold.
I'm claiming here that that is not a "current measured rate."
> But note that we expect it to take 3 weeks before quarantines show up in fatality figures.
That's why I've stated about Italy: "they started to quarantine municipalities since February 22." It's not that they let it waiting for "herd immunity" like UK or NL. It was less than 4 weeks ago.with 3 days doubling time they would have today 2048 dead. It doesn't appear to be possible to be slower than 3 days. With 5 days doubling time they would have now just 128 dead, with 4 days doubling time 256 dead. It's that easy to calculate.
2020-02-22 2 dead 2020-03-19 3405 dead
That's the Italian "current measured rate" for a month now.
That is growth rate as measured by confirmed cases. However given various restrictions on availability of testing, confirmed cases and real cases are likely to tell very different stories. Given that most countries refuse to test, refuse to test, hit an "oh shit moment" and start ramping up, the confirmed rate that your link uses measures testing more than actual infection rates.
References such as https://www.imperial.ac.uk/media/imperial-college/medicine/s... have used a doubling time of about 5 days. As https://www.statnews.com/2020/03/10/simple-math-alarming-ans... points out, studies of what happened in Wuhan suggest a real doubling time of about 6 days.
There are admittedly a lot of question marks about everyone's numbers.
Does anyone have access to good data? There is so much fake news that is hard to trust any articles. If somone is 30 years old, healthy, workout, don't smoke, what is the probability of hospitalization, serious injury, icu or death?
Please see the Imperial College London report, recently released, for a model that contradicts what you’re suggesting .
In particular, see Figure 3 for a direct refutation of the idea that simply applying suppression techniques to the infected or primarily-at-risk groups is adequate.
Anything other than immediate and severe suppression techniques puts us over ICU capacity in the next few months, and the estimated number of deaths skyrockets into the millions.
1. I think nearly everyone besides WHO has the consensus that China’s released numbers is too low, especially when you compare it to the numbers from Italy and South Korea. Whether that’s due to corruption or the swamped hospitals and a lack of test kits in the beginning is debatable.
2. This white paper may explain Newsom’s numbers.
The growth is exponential and not linear which throws off most people. Surprisingly, even though the paper’s R0 is on the conservative side (2.2-2.4) and so is their case fatality rate. However, their simulations still predict 2.2 million deaths in the US if there’s no gov intervention for social distancing
I’m more and more worried this is an overreaction, even ignoring the economic impact. Look at the projections on page 19:
They seem to have picked the green line and not the orange line. Hopefully they’re planning for intermittent social distancing, or have some other medium term plan.
Introducing social distancing for the entire population comes at huge economic and social cost.
It is more effective to follow these three simple measures:
1. Symptomatic cases stay at home for 7 days
2. All household members of symptomatic cases stay at home for 14 days
3. Social distancing for the over 70 population only
Social distancing of the entire population also slows the acquisition of immunity with zero and low risk age groups, which drags the whole situation out months longer.
It might even be logical with the above 3 steps in place for zero/low risk group to deliberately seek out the virus. This grants immunity quicker, and if they become sick, they can receive treatment whilst hospitals are still underwhelmed.
Speaking personally I would happily accept a strain of the virus considered to be low risk, then go into 14 day quarantine, if it meant I could afterwards work and socialise without any lockdown or restriction.
It's a very difficult problem. Here is a paper outlining the cost in lives of your proposal.
The Imperial College report:
Full suppression will require total shutdown for ~18 months.
Mitigation, specifically requiring the sick and elderly to self-isolate and quarantinging COIVD-19 cases and their families, will result in 2-4 million dead (depending on its effectiveness), with the situation over in 3 months.
We need to just accept 2-4 million sick and elderly Americans dying. That outcome cannot be any worse than complete economic and social destruction that we seem to be pursuing in the name of virus suppression.
Edit: I am specifically referring to the Imperial College report which outlines this scenario. Here is the original report: https://www.imperial.ac.uk/media/imperial-college/medicine/s...
And here is a summary: https://twitter.com/jeremycyoung/status/1239975682643357696
The last 24 hours have seen some dueling arguments, from respected sources, regarding how to model this.
This one (pdf) has been highly influential to US and UK governments and is pretty sobering and alarming: https://www.imperial.ac.uk/media/imperial-college/medicine/s...
This one is way on the other end: https://www.statnews.com/2020/03/17/a-fiasco-in-the-making-a...
I'm just hoping that we discover effective treatments soon, because with those we might be able to start treating this like a normal cold before any of the worst case scenarios would come to pass.
Here is the link to the report itself - https://www.imperial.ac.uk/media/imperial-college/medicine/s...
The report is published on March 16th. However, US and UK government started taking action on the first week of March, so I am not sure how the story says this report led "U.S" and "U.K" to take action.
> ...not even China...
For N95 masks as the article points out, yes, no one has the melt-blown fabric manufacturing machines that are a precursor.
However, if everyone around you wears surgical masks, then everyone is blocking ingesting about about 80% of particulate matter . For an emergency like the pandemic, we're also interested in the masks blocking what wearers breathe out to surgical standards, what the masks were designed for. That's a "not letting perfect be the enemy of good" level of "good enough" if your alternative is letting front line healthcare staff go without melt-blown fabric filtering PPE because everyone is scrambling for P95 or P100 masks.
The US military establishment after the pandemic is controlled will hopefully be given a hand in determining precursor tech tree bottlenecks like the OP article pointed out, and have Congressional authorization to perform economic disaster preparedness within its scope. Stockpile and maintain sufficient operational skills alongside key equipment, so our disaster preparedness is better than "duck and cover" (some interesting questions behind that door).
Some modeling suggests we need a minimum 3 month lockdown in the US to cut deaths in half , 18 months (or however long it takes for critical mass of population to develop immunity or vaccinated immunity) to approach a semblance of status ante quo death rate. I wish they would put their modeling source and data onto Github (or Wolfram Alpha?), so others can work with the model (I'm interested in its backtest results from previous epidemics). If we lockdown for even 3 months, then that's GFC effects; 18 months is Great(er) Depression.
> What no one is talking about is how much social distancing/isolation is needed to avoid hospital saturation.
People are talking about it, that talk is driving policy, and the answer is “with some local adjustments possible based on caseload, until there is an effective vaccine developed and deployed, which will probably be 18+ months”, and the alternative is a truly catastrophic number of fatalities.
That is incorrect. The US, UK, and others were taking a mitigation strategy until the last couple of days; the change wasn't driven by delayed blind copying of China, but by this analysis, which was shared with governments as a WIP before being published:
There may be legitimate criticism of the broad policy approach, but “blindly following China” isn't one of them.
Around 18 months (until sufficient quantities of vaccine) is this group of experts’ estimate
Around 18 months (until sufficient quantities of vaccine) is this group of experts’ estimate
Around 18 months (until sufficient quantities of vaccine) is this group of experts’ estimate
Around 18 months (until sufficient quantities of vaccine) is this group of experts’ estimate
Doomsday panic? Ill informed opinion? I'm literally looking at the chart used by the White House that comes from the data generated by the UK government. Look for yourself:
> Covid-19 is 99.9999% survivable for those under 70 with no pre-existing conditions.
No, your six 9s are wrong. So far, the best quality evidence we have says: https://twitter.com/ShaunLintern/status/1239693272311828483/...
If you are aged> If you have scientific evidence supporting mass isolation for a pandemic like covid-19 please share.
10-19 you have a 0.006% chance of death 20-29 0.03% 30-39 0.08% 40-49 0.15% 50-59 0.60% 60-69 2.20%
The original report from Imperial College COVID-19 Response Team:
Here's the paper by Ferguson  referred to in the article.
This was the UK government position till a few days ago. Then they ran the numbers again and changed their mind, because even with all mitigation measures in place they estimated around 250k deaths with this approach.
The UK is now heading towards a more comprehensive shutdown and I suspect will resemble France and Spain soon - the method referred to in this paper as suppression.
This will be really damaging to the economy though and probably lead to a global depression if it is done everywhere. I'm honestly not sure which is worse, because a global depression will lead to a lot of deaths as well.
Sadly that's not a possibility.
According to mathematical models the capacity of your hospitals will be exceeded by more than 30 times.
Give it another week or two.
Perhaps our most significant conclusion is that mitigation is unlikely to be feasible without emergency surge capacity limits of the UK and US healthcare systems being exceeded many times over. In the most effective mitigation strategy examined, which leads to a single, relatively short epidemic (case isolation, household quarantine and social distancing of the elderly), the surge limits for both general ward and ICU beds would be exceeded by at least 8-fold under the more optimistic scenario for critical care requirements that we examined. In addition, even if all patients were able to be treated, we predict there would still be in the order of 250,000 deaths in GB, and 1.1-1.2 million in the US.
If you want something more reputable, how about Imperial College London’s epidemiologists:
Yup. And the worst-case scenario is infections peaking next winter to coincide with regular 'flu season.
An Imperial College paper  out today predicts this under its most stringent suppression scenario; see the chart on page 10.
The logic of the UK position is well articulated in this very HN-friendly article by a UK epidemiologist unaffiliated with the govt scientific advice team [complete with graphs, the Python code of his illustratory models and an appropriate caveat about the magnitude of the uncertainties of the assumptions the scientific advisory team's models will have made]
nb the UK has since issued stronger isolation guidelines and downplayed herd immunity in the media. Longer paper about their current model https://www.imperial.ac.uk/media/imperial-college/medicine/s...
That is not what the math says. Even if people do become immune, it takes much much longer for the disease to spread through the population at a rate that won't overwhelm healthcare capacity. There was some uncertainty about this until now, but a paper just came out from the team at Imperial, whose advice was informing the UK govt.
You have to keep severe social distancing in place until there's a vaccine, or there are too many people to care for. Better just hope it's a worst-case scenario when they estimate 12-18 months for a vaccine.
It's here now : have a look.