How many lives is hospitalization saving in the pandemic?

Do we have evidence that hospitalization of COVID19 patients is actually saving significant numbers of lives?

I’ve now seen multiple studies suggesting that up to 80 or 90 percent of patients who end up on ventilators ultimately die.  At this point, I guess there’s no way to know if the other 10 percent would have lived without the ventilators.  From what I can tell, most other hospitalized patients are getting supplemental oxygen, IV fluids and antibiotics.  I have not seen any evidence on the effectiveness of these treatments.  Many of those patients live, but we don’t know whether they would have recovered without hospitalization.  It would obviously be impossible to do a RCT on that at the moment.

Answering the question about the efficacy of hospitalization would seem to be critical, though, since, as best I can tell, the main justification for shutting down society now is to prevent our health care system from being overwhelmed – especially the supply of ventilators.  If our hospitals are overwhelmed, not only COVID19 patients, but others with treatable injuries/diseases might die.  But if hospitalization is not actually saving COVID19 patients in large numbers, then all the costly social interventions we are implementing now are mostly just delaying the spread of infection.  Still, I recognize that it’s possible that this delay could save lives in one of two ways (or maybe there are more I’m not thinking of?).

1. We use the time to get better at testing.  Then, when we lift the social distancing measures in a month or two, we have the ability to quickly test and isolate infected individuals and their close contacts.  Maybe we also have anti-body tests so we can avoid quarantining immune individuals.  This keeps the rate of spread relatively low until we have better treatments or a vaccine for those who haven’t been infected yet.  It’s possible that “at-risk” groups will have to stay isolated during this time until we get effective treatments/vaccine.  I haven’t seen any estimates of how effective this kind of strategy might be – i.e., over a course of 18 months (the time to develop/deploy a vaccine) how many infections would this prevent?

2. We could keep the social distancing policies in place until we get a vaccine/treatment.  But if estimates of 18+ months to a vaccine are correct, I suspect the economic costs will be too high to bear to wait it out this way.  So this is probably not in the cards.

If the number of lives we can save with #1 is relatively low (I have no idea what the number is), and if #2 is off the table, then we are really just delaying most deaths, at great social cost.  It might be better to prevent our hospitals from being overwhelmed by doing better triage for admission – especially to ICU beds and ventilators (what percent of people over age 75 survive after going on a ventilator?), and working on getting people other treatments (oxygen, etc.) at home.  At a minimum, it seems like the intense energy and resources focused on ventilators now might be misplaced.

For what it is worth, I’m not a skeptic of the current social distancing policies.  I’m pretty sure I’d be doing all this and more if I were in charge.  But I’d also be looking for evidence that what we are doing is the best course of action, given the massive costs.

That is an email from a very smart person.  To that tally we also must add the negative that hospitals often become a vector for the further spread of the virus.

So what does the best evidence say here?


We can't randomize covid patients who need hospitalization to either being hospitalized or not. This doesn't meet the standard of clinical equipoise. If this person is very smart, they can find the evidence themselves.

If the current processes are trying to save people, we can evaluate them by their outcomes. So if those who have been identified as needing ventilators are dying after being put on ventilators, that should adjust how strongly we believe ventilators will help. *If* they are not really helping now should we expect that trend to continue in a regime of more ventilators? That's probably one of the most important questions of the current crisis because the working assumption has been that expanding hospital and ventilator capacity will be effective and we should shut the economy down so that we can expand the capacity. If that's not true, then lots of resources are being wasted.

This exact criticism could be applied to any deadly disease. Lung cancer is a death notice but we still treat those who have it and after six months or so they die. Some who choose to not take the expensive and uncomfortable treatment also die after six months. So why even have hospitals???

Good point. We should distinguish between palliative care (which is what we have for most terminal deceases) and "regular" care (which at least aims at recovery). So the question is: what sort of palliative care is available for covid19? I don't know if adding a respirator actually improves the quality of remaining life - it might actually worsen it. So I think the conversation is still worth it. The main problem it seems is that we cannot clearly define which cases of covid19 are terminal and which are not. In that case, our only option is to assume they all can recover and eat the costs...

Non-invasive ventilation using a CPAP or BiPap machine with a strap-on mask + Oxygen would be more pleasant palliative/hospice care for a Coronavirus patient than having a tube stuck down their throat - I am sure they would be more likely to die though. I am not a doctor - just a CPAP user for sleep apnea. CPAPs and BiPaps can still be had for $200 om craigslist btw - picked up an extra one just in case hospitals get overloaded and a family member gets covid and can't get in.

the Death conversation needs to shift perspectives...fear based ethics and science in past 70 years has left us hanging onto the inevitable moral and economic cliff. Death is not something to be afraid except for the people involved. Medical intervention that pumps you up with with life saving methods is what needs a new perspective Take a look at traditional civilizations, people lived long lives in many societies. Healthy rituals and ceremonies are missing. Healthcare has lost its way. Is this how you want to leave your body?

I'm not sure it's a critique of all disease treatments, but it is how one should begin to analyze treatments (especially if no RCT is available). If the treatment is ineffective, it shouldn't be performed. How is that even controversial? If it's only effective for a small percentage of patients, applying the same treatment to all probably doesn't pass a cost-benefit test. Also, in the lung cancer case (for someone with private coverage), it is that person's dollars (plus those others who have freely chosen that insurance) going to the treatment.

I doubt these numbers are true, but let's assume they are - only 10% of people who require a ventilator survive. And let's say we have enough observational data to accurately predict which patients would make up that 10%. How would you implement that into actual policy?

1) Would you be willing to tell society that based on best evidence, only people with/without variables A,B...C would be allowed to receive full life-sustaining therapy? 2) Do you think society would accept this policy, no matter how finely the message was crafted? 3) If you were included in the population that was deemed "low priority" b/c you have medical condition X (or are older than age Y), would you accept this protocol if you became ill? 4) If you were critically ill and the only way to survive, even for just a few days, was to go on a ventilator, would you say, oh not me?

As a palliative care physician who treats adults and children, I can say without a doubt that most people who have not been critically ill or seriously-ill for a prolonged period of time, would not be willing to do so. As ANON points out, people with terminal cancer continue to seek therapy, even when their own oncologists recommend against it. Families with patients on ventilators (non-COVID) choose to not withdraw care, even when medical providers explain there is no possibility of recovery.

So, unless our social and cultural mores drastically change over the next couple of weeks, and US society at large becomes willing to accept federal dictates about who can and can't receive life-sustaining therapy, this line of questioning seems silly to me. If 10% of people survive, everyone is going to think they should be given a shot of being those 1 in 10.

1) yes
2) yes
3) probably not, but that's why it probably shouldn't be left up to the patient who has nothing to lose
4) what do those few days cost me and those I am leaving behind?

"people with terminal cancer continue to seek therapy, even when their own oncologists recommend against it. Families with patients on ventilators (non-COVID) choose to not withdraw care, even when medical providers explain there is no possibility of recovery." Well, how much are all these decisions costing other people. I get people want to hold on to hope, but maybe when the costs on others is excessive, the patient or her family shouldn't have a large say in the decision.

Many people would agree with you that perhaps pts/families should not be the ones deciding. I think I would fall into that camp.

But I totally disagree that US society would all of a sudden be ok with "experts" telling them what they can and cannot receive bc of statistics. If you disagree with that, I would like to know a) what makes u convinced people would accept a national rationing policy, and 2) what would your message be? If u were DJTs speech writer, and he actually stuck to your script, what would you have him say?

It's not what's right or wrong, it's about what's politically feasible. Americans, for better and for worse, is a hyper individualistic society. You think you could convince a 50yo affluent small business owner with hypertension, who has 2 kids, that they should forgoe life-sustaining therapy so that an unemployed 25yo with no children can receive tx? Our experiences in expanding Medicaid seem to say otherwise.

Crap... Misread the question...that's what happened on my macro comp. I will say that we have a lot of people sheltering in place on the advice of experts. Maybe now's the time to go for it.

Death panels! Death panels! I'm 74, active ,and have no underlying conditions.. GUESS what I want for myself?

Alex Berenson is saying similar things

"main justification for shutting down society now is to prevent our health care system from being overwhelmed" is wrong. We are shutting down society to lower the number of cases to a level where we can feasibly monitor/trace/etc. individual cases to keep R<1.

If not, we will have N*p*d deaths, where N is population, p is the herd immunity level (which just means enough people are immune that R<1 naturally) and d is the death rate. At p=.7 and d=.005 (probably on the low end), we have 1.15 million US deaths. To say nothing of millions more with permanent lung damage. That's slightly more than the US per capita death total in all 4 years of US involvement in WW2. We clearly are not going to let that happen.

We know from Taiwan and Korea and others that it possible to keep deaths low with test and trace without massive distancing. It is very labor intensive, and can't realistically be done with 100000s of cases. With hundreds of cases and quick testing, we now have many examples of it working.

That is to say, there is not only no tradeoff between getting the economy back and the social distancing we are doing now, but the social distancing we are doing and the ability to keep deaths low a month from now with much less severe policies are complements, not substitutes.

Way to move the goalposts.

The entire justification for the lockdown was to #flattenthecurve. No one knows if contact tracing scales to non island nations. We did not shutdown the economy of the world for your science experiment. We shut it down since the medical establishment and the media scared the populace that hospitals would be overwhelmed and the costs would be ‘dire’.

We should have questioned the ‘experts’.

By the time you finish questioning the experts, the virus would have decimated the country.

Innumerate statement.


+1, more accurate

Centimated is probably a worst case scenario. 3 million dead, 2/3rds of them over 60.

To be clear I'm not really picking sides in this debate. There's not enough good information to swing my opinion one way or the other.

There's probably a good case to be made for true lock downs in the hardest hit areas. I'm not sure half assed lock downs for 2/3rds of the country make sense. But they might be better than doing nothing, or they might be a lot worse.

I do know that within 4-6 weeks, your median person will be ignoring lock down orders on a routine basis through sheer boredom and being broke. So, we'd better pick our starting week carefully.

Just look at Northern Italy. The US is on par or even a worse path currently. With inaction hospitals will be overwhelmed for sure.

Bergamo had 4-5 times above the average amount of deaths, but only 30% due to covid19. The rest where other causes / illnesses, which could not be treated as usual (due to overload of the health system with covid19 patients).

If you let this run, good look. You will not save the economy, as mid term, much drastic measures will be needed, as the public will not tolerate the amounts of victims.

You can't assume that the remaining 70% in Bergamo were all due to other causes or illnesses. Probably most of those were also COVID-19, but there wasn't any time to test them, and not much point in doing so after they died.

Oddly enough, apparently the Italians are testing everyone who dies with respiratory symptoms to have a precise number of those with covid19. This might be because the Italians have been cremating such dead for a couple of weeks at this point. This contrasts with the Germans, who are not spending any effort on testing the dead, as they believe that if testing is necessary, they can dig up the bodies for samples.

That was information from just a couple of days ago - it may not be accurate today.

I have had a biopsy postponed because of COVID-19.

I had an MRI postpone due to COVID-19.

Is there a place with data on the progression of coronavirus in Lombardia, and not Italy as a whole?

"The US is on par or even a worse path currently."

Maybe in # of cases but since Italy's rapid rise they have had a pretty steady 10% death rate (currently a little higher) and the US is between 2-3%.

Death lag infections by 2 to 3 weeks. You can not compare the current death rates to Italy. The US will very likely exceed the Italy death rates in 2 to 3 weeks.

And the next time a duck farmer in India sneezes…we do this all again?

But with bravado!

After all, Trump did not close the borders soon enough! We did not go into lockdowns soon enough!

Halt commerce prophylactically or be accused of negligent mass manslaughter!


I understand hysteria. But hysteria as institutionalized policy?

See post 9/11.

Because Americans are hysterical. Asians are clinical. They did the job necessary to contain the pandemic without shutting down their economy. Western leaders have a lot to learn.

S.Korea, Taiwan, HK, had experience with SARS and other recent epidemics, therefore have built systems that allowed a more rapid & robust response. Also S.Korea, for example, has imposed tracking, quarantine and penalties that would likely to be found unconstitutional in the US. The dichotomy is not "clinical" vs "hysterical". Should the US prepare for every improbable contingency in advance ? That is a formula for a failed economy too.

Pestilence, war and famine brought down the Qing Dynasty. Chernobyl and oil price wars brought down the USSR. You don't have to prepare for every possible contingency. Just the ones that will bring down your society :-)

Not at all, clinic requires data or what's the buzxward now "data-driven decision making". Your cherry picked darlings simply got lucky compounded by positive factors in this case, i.e. Western medicine level, small population, authoritarian governments, easily isolateable terrain, and docile fearful citizens.

Get back to me when we see how Mongolia and Burma handle it once it starts to community spread or the easy Asian counterexample, Iran.

otoh so far it looks like
South Korea has probly had the most effective response to a viral pandemic in the history of viral pandemics. in the future it will be studied & copied

The data is whatever they collected during the SARS epidemic and the lived experience of seeing it destroy their communities. The West should have learned from this but apparently did not.

There probably are 100000s cases right now in the US, if not millions. The hope is then that lockdown can reduce this number over time to something which is manageable to track and trace.

But this will need lots of time for existing cases to recover and stop transmission, plus ongoing large scale testing probably well beyond SK, and lots of surveillance.

Although that said, breaking news (2 hour) -

"Singapore to close most workplaces except for essential services and key economic sectors in "circuit breaker" move to curb spread of #COVID19 spread, says PM Lee " Aka "If you're not a 'key worker', house arrest for you' comes to Sing.

Welp, maybe "Competent East Asian authoritarian technocratic track and trace" (paraphrasing MR admirers of 'efficient' pro-business authoritarianism) isn't really such and alternative to lockdown after all, and we need a new model for what comes after the the next two months...

First, we need to get to the point where the disease is not spreading before talking about the next couple of months. Singapore clearly wants to do everything possible to avoid turning into Italy or Germany or South Korea, which had 86 new cases today, according to

If you read Robin Hanson's post to which Tyler linked today, or just look at the effects of total lock down re slowing the virus, it appears it's correct to say that lock down is an alternative, but "effective"?

Immunity level is way too low. At equilibrium the ratio is about 40 immune to 1 infected, assuming we all see the virus eventually. An immunity level of .98 is better. This is seasonal immunity vs two week incubation. And they expect immunity to last more than one season for many.

How does Immunity level relate to the equilibrium ratio?
Just realized that I do not know.

Sorry A Fine Theorem, the whole justification (as others have pointed out) of all this social distancing was to "flatten the curve" so that hospitals wouldn't be overwhelmed. It was explicitly a mitigation (not a suppression) strategy. The worry about suppression is whether there will be a resurgence (like in the Spanish flu case). If you want to make the case for suppression, go ahead, but social distancing has always been about flattening the curve. Test and trace was never sufficient (though it does has its place). Singapore is shutting down offices, and South Korea has had its schools shut down for a month and (though there has been no massive government ordered closures) there have been reports of drastic reduction in economic activity due to people staying home ( You must also consider what the *excess* deaths from the disease is. How many of the people who will die would have died in the next year or two anyway? I've pointed this out several times, but this basic concept (thinking at the margin) seems to have alluded many people.

Regarding the point of the email to Tyler, it seems a lot of resources are currently going into how to increase ventilator production. *IF* people who are put on ventilators are going to die anyway, we should rethink diverting resources away from that towards more productive uses.

Great question.

One of the many questions (eg: life expectancy of those saved) that has to be asked and should have been asked before starting this policy.

Another question is: Do we have evidence that hospitalization of COVID19 patients will actually end up costing significant numbers of lives?

And could this cost actually be higher than the benefit in terms of QALY? For example if you save an elderly COVID-19 patient now but lose a middle-aged person a year from now because chemotherapy and "elective" tumor removal surgery are now interrupted.

By the time you set up a 5000 bed hospital like in Madrid, the entire point is to keep the infected in one place. It also makes it much easier to deal with those who die in an efficient fashion.

+1 to Lukas since back in Jan/Feb many fringe sites and rumors were saying the Wuhan field hospitals were just death camps. Keep them all in one place so they don't become vectors for infection elsewhere. Harsh but true.

Bonus trivia: I think the de facto effect of "flatten the curve" will be to kill off the poor rather than the rich. After all, the rich like 1% me can afford home delivery, servants to cater to their needs (I have some now!), and a sheltered life in the countryside away from the maddening crowds (ditto for me)... speaking of maddening crowds, "28 Days Later" was a good movie, saw it the other day, and not scary either (I hate scary movies with fake blood, and this coming from a guy who slaughters farm animals)

A frequent literary faux pas:

From Wikipedia: Hardy took the title from Thomas Gray's poem "Elegy Written in a Country Churchyard" (1751):

Far from the madding crowd's ignoble strife
Their sober wishes never learn'd to stray;
Along the cool sequester'd vale of life
They kept the noiseless tenor of their way.

"Madding" here means "frenzied".[5], or, actually, "maddening".

I love it when a blowhard makes an elementary mistake and is taken down a peg. Good on you, Chuck

Don't celebrate too quickly. It was a mere typo. Even geniuses like me ocassionally make typos.

Are you testing your servants for COVID-19?

Bingo plus an increase in causes of death from bored people watched Netflix 18 hours a day for months. Alcohol and recreational drug sales are through the roof, my guess is pulmonary embolisms going to skyrocket, and ditto both maternal deaths and abortions. Wouldn't be surprised in the aggravate we lose more lives from secondary causes that had we did nothing.

My understanding is that hospitalization is primarily for patients who need supplemental oxygen (not a ventilator). This is quite important.

Aren't there portable oxygen devices that can be used at home?

Yes, but isolation is currently the only way to stop the spread of the pandemoc. It is much easier to have all the contagious people in one place than to attempt to support them spread out everywhere.

Isn’t it also to prevent people who are sick who then go to the ER? Those people have to be tested and that takes time and manpower. Lowering those numbers helps relieve the burden on the hospitals as well.

Also, is it really true they’re only admitting patients who need ventilators? And even if only 80-90% are dying, that still means there are patients who recover after being admitted and does anyone know if they would have lived without being admitted?

There are stages depending on disease severity/progression, something the person seems unaware of. A number of around 10% of confirmed infected persons need medical care, keeping a fever under control, being fed, being isolated until no longer contagious, etc. The number requiring any sort of breathing assistance is considerably less, though the disease can cause severe breathing problems. Such people are the ones that are truly hospitalised. Depending on age, and the state of the health care system, if they are over 80 or have other condition, it has been 10 days or more since such people have been provided breathing assistance in Lombardia. Grimly, the death rate of those put on ventilators in Italy has probably decreased at this point, due to that simple fact alone. And the French still seem to believe that it is worthwhile to keep 30-50 year olds on ventilators, as they send them to Germany or use a special TGV to send them to Bretagne.

Seems like this message has been lost in the shuffle. You should spread the word to GM before they become a ventilator manufacturer.

Well put.. great questions.

But if it were my mom or my wife or my son or my daughter or my sister or my brother???

Screw dad... haha, j/k. Per American society, I seem to have forgotten dad.

My kid would be sad for a month or two.

We’ll need to see more data. The Seattle article in NEJM suggests that at least 33% will live. While China did better in many ways, generally critical care is better in the US.

I'd add the following way a delay could save lives:

3. We get sufficiently many surgical masks / convince people to make their own, so that when the lockdowns are lifted the vast majority of people can wear masks everywhere.

I've read that a major benefit of masks is that they help protect *everyone else* from *the wearer* (more so than the other way 'round). If the vast majority can be persuaded to wear masks everywhere, transmission rates might be lowered. I also heard that Czechia is trying this strategy, so in a few months we'll see if it is effective.

This is the dismal science I signed up for!

Dismal pseudo-science.

Yeah...Dismal Pseudo-Science.

Or Pseudo-dismal science?

Also, while vaccines will indeed be 12+ months away, April will see a lot of results of pharmacotherapy clinical trials and possibly more data on convalescent plasma. Those we could ramp up quicker.

Yes, in Bale they are already testing this, to see how effective it is. The ratio is reported to be two convalescent patients are required for one infected patient. No results have been reported yet, but the testing is being expanded to other Swiss medical centers capable of doing such treatment.

> "Maybe we also have anti-body tests so we can avoid quarantining immune individuals. This keeps the rate of spread relatively low until we have better treatments"

If immunity lets you work and earn, this might backfire. You'll see people infecting themselves hoping for a mild illness. Maybe even antibody doping to cheat the test by buying plasma from a recovered person.

Isn't "antibody doping" a way to _actually_ create immunity?
(I am less enthusiastic about self-infection but at least it's conscious and calculated risk.)

Yes and the "libertarian solution" is to either transfuse sanitized blood with antigens into your body, or practice variolation by infecting yourself with Covid-19 virus, and then being "immune" to the disease and offering your services to society at a premium. Solve for the equilibrium, perhaps some company will figure this out, and with government approval (ha) market this to people?

Transfusing blood from those with antibodies is a method to fight an infection by pumping up the level of antibodies quicker than your own body is capable of. Adding antibodies to your blood does not train your immune system to make those antibodies, however. So a transfusion from a person who is immune does not "teach" the recipient's body immunity. It's a short term measure to help fight the infection.

Pulmonary/Critical Care doctor here. That is an absurdly high mortality for ARDS and also, from my so far limited but growing experience, for this disease. Hospitalization and medical intervention so obviously saves lives that any randomized trial would be unethical. Medical intervention in the ICU is so obviously life-saving, that when we withdraw it, say because the patient or family requests it, death usually follows in minutes to hours. COVID-19 is not that much different from other diseases that we routinely deal with. Overwhelming the medical system will rapidly increase mortality though. The data to back this up will come from the post hoc analysis of the case fatality rate under different care regimes (early, overwhelmed, controlled)

Here are some numbers from the UK
Of 165 patients admitted to ICUs, 79 (48%) died. Of the 98 patients who received advanced respiratory support—defined as invasive ventilation, BPAP or CPAP via endotracheal tube, or tracheostomy, or extracorporeal respiratory support—66% died.
I think In Wuhan, the statistics were worse.

What about the patients with covid who aren't admitted to the ICU? What are their mortality rates?

First of all, you're assuming that ICU=being on a ventilator. There is more that goes on in an ICU than that. Those who go onto a ventilator are some fraction of those in the hospital receiving something ranging from supportive care (without which they would have a significant mortality) to experiment medications to routine breathing treatments/supplemental oxygen, and so on.

To answer your question (which properly phrased would be, "What is the fatality rate for patients so sick they need to be hospitalized, but who don't end up on a ventilator?"): Much, much lower than 80-90%, and much, much lower than 66%. Here at Marginal Revolution, the smart folks seem oddly unable to understand the nuances of illness.

Thanks for the data you provided, which can take this discussion forward and provides information useful for policy formulation.

Not treating a disorder that in normal times is highly treatable seems wrong... but if there were instead a time limit, i.e. after 1 day or 10 days, 20 etc that mortality increased I could see that being not only effective but ethical. Is there any indication that the studies have done any time sensitivity analysis?

Why would we need to wash our hands, obviously from our observation people are dying from something else. A randomized trial of hand washing is absurd.

Do we have evidence that hospitalization of COVID19 patients is actually saving significant numbers of lives?

Yes, as being on a ventilator is the last stage of being hospitalized. Keeping a fever under control, for example, is no less important with this infection than with any other. Maybe such treatment is not hospitalization for this person, but it is certainly medical care.

And the main justification for shutting down society now is to prevent our health care system from being overwhelmed has to include how to handle 100,000 pandemic victims in the span of a few months. This may not be what most people consider the health care system, but morgues are also part of the total chain.

It is odd that the one country which has been fairly successful (up to now) at keeping deaths low is using extensive hospitalization for those sick enough to require it. The difficulty in making such judgments about hospitalization is the age profile - many young people put on ventilators do end up recovering, whereas many old people won't. The pandemic is basically compressing years worth of fatalities into a short time frame. That is what the health care system needs to deal with, along with thousands and thousands of contagious patients needing to be isolated while not able to care for themselves, a number that seems to be an order of magnitude higher than those put on ventilators. A number made up of a majority of those under 50 or so everywhere with a major outbreak. The giant 1000+ bed hospitals set up in conference centers in various European capitals do not appear to have ventilators, they are simply designed to keep people isolated (in the center, not from each other) from further spreading the disease and from dying due to a lack of basic medical care.

The pandemic is basically compressing year worth of fatalities into a short time frame.

Fixed that for you. And that is, by far, the upper case of estimates. Likely much lower. You are welcome.

Reported death rate in several EU countries from 2016 - At the opposite end of the scale, the lowest death rate across the EU Member States was recorded in Spain (829 deaths per 100 000 inhabitants), ahead of France (838), Italy (843), Malta (882), Luxembourg (905) and Sweden (913).

Current death rate from covid19 alone in 4 weeks or so from worldometer -
Spain - 22.1/deaths per 100 000 inhabitants
Italy - 23/deaths per 100 000 inhabitants

A total that rises every day. And weren't you one of the ones claiming just a week or two or so ago how Spain and Italy had clearly peaked? One certainly assumes as stricter public health measures are followed in both countries, the number of newly infected and dead will decline - cause and effect does actually have an impact on how a virus spreads though a population.

I was one of those, and it does look like they have cleared peak. Italy is still on course to have more deaths than a bad flu season, but it doesn't look (at the moment) as if it will be more than twice as many. To justify the policy response, the "natural" progression of the disease would have to be significantly worse than that (given the costs we undoubtedly will suffer in terms of number of deaths from impoverishment).

The stock response is, "yes, but without the extreme measures Italy took, it would have been MUCH worse." I have yet to see the evidence for this. Robin Hanson has suggested that these extreme measures are so leaky that they don't really have a significant effect on the spread of the virus (it slows down a bit, but, ultimately, most people who would have gotten infected anyway will get infected). The question of whether "the cure is worse than the disease" is far from a stupid question.

Using the above data, about 2.27 people out of 100,000 die each day in Spain. Though there is an overlap we will ignore it. The cases designated in Spain as caused by covif19 has been a total of 9,000 in two weeks, so let us be a bit sloppy, and use 20 deaths per 100,000. In that two weeks, 1.4 people per 100,000 per day has died from covid19. Do you reasonably expect that the current death rate in Spain has dropped to .8 people per 100,000 per day?

I expect the IFR to be less than 0.9 percent. Further I expect less than 100% of the population to get the disease.

The normal fatality rate of the population in on the order of 0.9 percent per year.

-all ventilator treatment is not equal
-while patients are being ventilated they are also receiving other
treatment along with the mechanical ventilation
the listed studies don't detail the other medical treatments
given to the ventilated patients
-this is part is sorta important
this email from the very smart person says
"From what I can tell, most other hospitalized patients are getting supplemental oxygen, IV fluids and antibiotics. I have not seen any evidence on the effectiveness of these treatments. Many of those patients live, but we don’t know whether they would have recovered without hospitalization."
in the context of a viral pandemic/full hospitals
we would bet bigly that not too many patients are being admitted
to the hospital unless it is necessary. respiratory failure/death happens
very quickly. if anything we need to guard against not admitting patients who need hospitalization

After a shamefully slow start, Australia appears to be on track to contain the virus. However, the numbers do jump around a bit, so it's not possible to be certain. Assuming the trend is correct it will mostly be contained in a month and deaths will hopefully be below 200.

So in a month or two restrictions can be eased but vigilance will have to be maintained due to the possibility of new outbreaks. The odds are good there will be an effective vaccine in sufficient quantities for mass use in 15+ months which will bring the crisis to a close.

If we hadn't shown such incompetence at the start our death toll wouldn't be so high and the economic costs would be less as it would have taken less time to contain the virus and we could have kept some geographic regions virus free.

But, given that we're idiots, that's how the economic cost of COVID-19 is limited. Containment. Elimination. Vigilance. There's no better alternative.

Having visited Australia once (Sydney and Melbourne) if any country can practice "social distancing" it has to be Australia, which, with the exception of the two cities above, seems to me to be sparsely populated and at the far end of the world. I got the same impression when I lived in Arizona. Nice places to visit though. Cheery-o mate (they need to do another Crocodile Dundee movie, updated somehow, sure some script writer has written one).

Bonus trivia: looking at new cases at the John Hopkins site for Australia,, indeed they seem to have peaked, unless this is measurement error in that, like in the USA, the number of new cases depends on the number of tests being conducted (since tests are limited to only sick people due to supply constraints).

Crikey mate, there is no way Australia will have below 200 deaths.

Yeah, that's probably way too optimistic. Realistically we would need majorly improved treatments to pull that off.

Australia will easily have less than 200 deaths, if indeed they have passed the peak as the John Hopkins statistics say. Covid-19 has a 2.5% death rate, extrapolate to 7k cases and do the math. They only have 28 deaths and they've already passed the peak.

Bonus trivia: Australia has the most violent crime rate of all OECD countries. When I was there, there was an undercurrent of menace. Saw a woman with a bloody nose and a pram, probably a domestic violence case. There were Lebanese riots in Sydney about the same time I visited. Possibly the convict roots of the native Australians showing? Good looking girls though, which they call "Sheilas" in their native slang.

There were Lebanese riots in Sydney about the same time I visited. Possibly the convict roots of the native Australians showing?

Today's figures are pretty good, so maybe we can keep it to under 200 deaths. But other countries have had temporary improvements which haven't panned out, so we may not be that fortunate. It also remains to be seen how deadly it will turn out to be here. But, yeah, I'll cross my fingers and hope for under 200 deaths. (Or should I be social distancing those fingers?)

i think it is too hard with limited data to answer these questions in real time. Is it worth it ? How many lives saved ? how many years of life saved, how effective are the treatments. How costly is a shutdown ? How many other excess deaths are incurred from this shutdown ? What is the cost benefit?
By the time we can offer an answer, we have chosen a course of action lest we shall be seen as indecisive. After that it’s hard to change course.
Any politician doing the cost benefit calculus with lives on one side, will be accused of being callous and sociopathic.
As a result, we err too much on the side of let’s do all we can to save these people and we’ll figure out all later.
later , the touted answer will probably be : without the shutdown , 2 M would have died but the true answer might not be forthcoming for many years.
Undoubtedly lives will be saved. Lives would be saved if we all stopped driving but we choose not to.
I think poorer countries have an easier choice , daily life is precarious for many citizens and the extra risk from Covid isn’t worth making life even more precarious with a serious disturbance of. economic life

Arrogance is too mild a word for one who assumes that life is less important for a poor person than for himself. Or lack of real life experience.

Nobody has ever claimed that eugenicists are humble.

Irrelevant straw man. It’s not life that’s less important to a poor person in a poor country. . It’s Covid-19 and its exaggerated risks. It pales vs other risks they routinely face. In Africa they call it the “ white man virus “. We all manage risk in our lives and to a poor person in a poor country, Covid is hardly a bigger threat than lack of income needed for daily survival. See the daily laborers in India now deprived of income.

Please, surely you're not one of those poor people advocates like they have in the US/UK? So you propose infecting us 1% so some ignorant daily laborer can continue to make their minimum wage? There has to be a better solution, like a 30 day lockdown across the USA to flatten the curve to South Korean levels, then test test test. For the poor, you can have drone-delivered bread and soup lines.

Ray, the curve was at the South Korean level at the beginning, we let it run away. What will be different next time ? Will the infected running around by May 1st be manageable/traceable or will it be another mess with more quarantines? what's the endgame ?

Lots of smart ppl say and think stupid things. Even fucking China was forced to shut down an entire province. Maybe Smarty is missing some important data. Or maybe he's worried about his portfolio. But he's miles off the mark. This is the same ignoramus bullshit that caused this mess in the first place

Not really ignorant lower-case jim. Let me educate you. Italy tried "total lockdown" and if you do the math you get 528 Italians per infected Covid-19 person, while Singapore tried "partial lockdown then business as usual" and has 5433 citizens/infected (10 times better, higher is better). Do the math yourself. Which country is better off? Singapore. And they're closer to China than Italy, and presumably also had daily flights to China (though perhaps not like Milan to Wuhan).

After three weeks here, the grocery stores are getting slow. Mortgages and rents, and a couple weeks of grocery purchases and people are starting to get very tight with money, either because it ran out, or they are seeing that it is going to have to last till the middle of May at the earliest.

There is the beginnings of an underground service economy starting to be established. People needing services and people needing money.

I watched a public health official check a grocery store the other day. They are limiting the number of people in the store, making sure there are hand cleaners everywhere, making sure there are marks on the floors to keep people separated, and making sure that the signage for all this is made of materials that keep the virus active for 3 days.

Because if there is one people do with fine signage, it is run there bare hands over it.

And just wait until you learn how the virus lives on all surfaces, including the item you just picked from the shelf and will bring home.

@ Fine Corinthian signage - you're almost as ignorant as lowercase jim above, see how I kept him in their lower class place. FYI the C-19 virus lives on plastic, like Rich Corinthian Leather, only for a few days (paper only a few hours). So if you wash your hands there's nary a risk in catching the disease say the experts. And you do trust the experts don't ya?

Bonus trivia: Rich Corinthian Leather...what a classic marketing phrase that was. Who here is old enough to remember? Da plane, da plane!

And surprisingly affordable.

There is the beginnings of an underground service economy starting to be established. People needing services and people needing money.

That economy has always existed. Those that have been members in the past will have a leg up on the newcomers. It appears that the incredibly complex society that now dominates most of North America may well be unable to persist in the face of multiple challenges. The most inventive humans will take advantage of the situation to not only survive but prosper. The longer the elites fail to come to grips, the more likely dramatic changes will occur.

We have medical experts in charge of this. Their bias is to talk in terms of people cured, of lives saved, period.
None of them will do a cost benefit analysis and say : how many lives saved for what economic cost.? Economists do this but they’re not in charge of pandemic response.

Do you honestly think medical experts are in charge in Italy or Spain? Or that at this point, they are talking about people cured and lives saved, as separate from stopping the ongoing spread of a pandemic?

"If the number of lives we can save with #1 is relatively low (I have no idea what the number is), and if #2 is off the table, then we are really just delaying most deaths, at great social cost. "--TC

I would have said "great social and economic cost."

And even so, there is an abyss, a chasm and a gulf between people who are thrown out of work, and the relatively protected class that is making public policy and commenting on it.

If a vaccine is 18 months away and perhaps two years, then I say we voluntarily sequester old people and and the rest of us run the gauntlet.

By the way, I am 65.

Public policy is not always what you want to do, sometimes it is what you have to do.

Weren't you in your 50's last week?

I’m confused by this...being put on a ventilator isn’t the only treatment these patients are getting right? What about the patients who receive other treatments and recover? These patients are certainly taking up beds and other valuable medical equipment.

I somehow don't trust these P(death|ventilator) numbers. Because the decision to put someone on a ventilator is based on different criteria in each case and is dependent on the situation in the particular hospital/city/country.

Nobody over 80 has been put on a ventilator since March 21 at one major Strasbourg hospital, according to various media reports. Death rate from such cases? 100% of course. Likely increasing the survival chances of the people (under 40 years old and no pre-existing conditions, of which there are now hundreds in ICU in France) still being put on ventilators, or being transported to regions where ventilators are still available. The worldometer number that is useful is the in the overview - which in France is 6,399 (15%) in serious or critical condition, a percentage that seems quite constant over all countries with tens of thousands of cases in Europe. And at least in Strasbourg, a number that basically includes no one over 80. They are in another category, and it is very rarely recovered/discharged.

Smartest blog post I read here in a long time

But in a couple of weeks, the sheer ignorance of the actual effects of the disease on patients and how they are treated will be plain, and no one will remember this.

+1. If people here scanned back through the last two months of MR posts, and comments, about this pandemic, they would read a lot that they would like to forget. Like today's post and most of the the comments.

From my experience here, most people will pull-a-Sumner and claim their words in the past actually mean something else.

The Tyler Cowen Progress Studies Prize in Straussian Historical Interpretation

Anyone remember Tyler's initial soothing post about the Chinese Virus (as it was known at the time... back when you were racist for being concerned about it)?

He basically said "the good thing is, it only seems to be killing the elderly."

Well now -- that fact is still true. Even more true, actually.

But you won't find him soothing anyone with this truth any more. Not when there are political motivations to suppress it.

Part of the answer is viral load. If we just let it spread and did nothing, the proportion of people who would die would be much higher than with periods of social distancing. With high viral load a healthy 40 year old will have a profile of a 70 year old.

60 somethings who used public transit everyday and went to crowded places would die in high quantities due to the high viral load they get from everyone.

Second the doctors and nurses would get it and our healthcare system wouldn’t function. And a huge number of people would die. The providers would truly be at risk, because they’re exposed to highest viral load. And even if you are callous and don’t care too much about their lives, training new RNs and MDs takes enough time that we would go several years without a functioning healthcare system. Mortality would skyrocket that’s what happened initially in Wuhan).

Remember that without social distancing the demand for PPE would be much greater leading to even more vulnerable medical providers.

Third, what someone above said: lots of super ill patients who cannot care for themselves until they somewhat recover. They’re not in the ICU. But without a hospital, they’d die.

Anyone who is put on a ventilator would die if they weren’t. (My parents are doctors, and they don’t have any doubts about this).

How many would eventually die if they were?

Which is why no one over 80 is being put on a ventilator in Strasbourg, they are being used to keep those more likely to recover alive.

Our artificial suppression of R is measurable but still insignificant.

In NYC we are near equilibrium, heading into the downward phase of the tall sharp spike. This all happened before our debate even started. It is chaos in the hospitals, unmeasurable chaos. It will be chaos in the remaining large second tier cities starting almost tomorrow. Nature is mostly driving R down, hospitals are paying a huge price in chaos.

Watch the numbers tested and numbers confirmed in California. That number has been holding for a few days at about 15%. Same as NYC. Already our immunity levels ill be .85, day after tomorrow, or the day following, in NYC. In two weeks, California's immunity level will be .85, and NYC closer to .65. The numbers are biased toward the symptomatic, but over the days testing becomes more random and widespread, not less. That number should increase in accuracy. But equilibrium is coming to a town near you.

You say the numbers are biased toward the symptomatic, but aren't they so biased to be useless for determining general population immunity?

The best analogy for death and hospitalization is the Civil War. Next week all 20 of the second tier cities will be facing a triage level akin to being in the thick of the Civil War. It will be over in a quarter, and then what remains are some skirmishes in the suburbs.

If you live in Wyoming, lucky you, your environment is sparse, R is small to begin with. It will take over a year to equilibriate Wyoming and Montana, sparse outbreaks in the small towns here and there. And the pandemic industry will be all scaled up for the rural states, treatments a lot cheaper.

'In NYC we are near equilibrium'

Keep telling yourself that for the next few weeks, it might be comforting.

Prior_approval, ladies and gentlemen.

Misanthropy personified

Also, hey Ken, remember when you went full ad hominem to defend this shit? This is you.

Good times.

Someone who believes that even the skeptical cannot deny reality eternally. NYC is not even close to the peak of its pandemic, but why trust projections concerning NY, considering the variations - "As for New York, the worst days are near. Deaths per day are predicted to max out on April 10, according to UW scientists; Gov. Andrew Cuomo predicts deaths peaking at the end of the month."

But don't stop making a catastrophic situation fit into whatever personal dramas seem to be constantly playing out in your head.

Deaths lag diagnoses which lag new infections. The time lag is significant. In a sophisticated forum like this, you'd think people would say which "peak" they are talking about.

We need to isolate the Covid-19 patients from other patients and everyone else.

We need separate hospitals, in the UK they are converting exhibition centres into huge covid-19 hospitals.

Until you can test everyone and particular the suspecting people who have covid-19 it's impossible to stop the spread. Isolation and social distancing is buying time until testing is ramped up.

The article on lessons learnt from Italy seems the minimum template countries should be using
The differences in approach to the Covid19 in regions of italy

Current leadership seems reactive and is being outpaced by the speed of the virus. In a crises situation, speed of reaction is vital and practically all countries did nothing in the early phases and is paying a huge cost now

Just an 'uptick' - "Authorities in the Paris region have converted a portion of the sprawling Rungis food market outside the city into a temporary morgue to accommodate the uptick in patients who succumb to covid-19."

"Spain’s coronavirus death toll rose to a total of 10,935 from 10,003 on the previous day, the health ministry said on Friday, but showed the first fall in a daily death toll since 26 March."

Spain's death toll on March 25 was 656, so the 'drop' represents an increase of almost 50% in one week between those two numbers, and means around a thousand more dead than if the number had remained flat at 650 per day, causing a total of 4250 deaths.

GIGO. Your cited reference is from the first month of the pandemic, and involves 191 cases. You jump from a low survival rate for patients who go on ventilators, to discounting the utility of in-hospital treatment.

"So what does the best evidence say here?" The best evidence says that there is much more to treating severe cases of COVID-19 than ventilators. Here's something from a Physiciansweekly article a few days ago:

"I posed the following question on Twitter: “What is the mortality rate for [COVID-19] patients who require mechanical ventilation?” and received answers ranging from 25% to 70% from people who have personal knowledge of outcomes in their hospitals.

Probably the best published information we have so far is from the Intensive Care National Audit and Research Center (ICNARC) in the UK. Of 165 patients admitted to ICUs, 79 (48%) died. Of the 98 patients who received advanced respiratory support—defined as invasive ventilation, BPAP or CPAP via endotracheal tube, or tracheostomy, or extracorporeal respiratory support—66% died."

This is my personal experience as well. Our N is still small, but we are not seeing 80% death rates on vents.

The professional figures I am seeing suggest that about 5% of Covid patients are "critical", which typically means respiratory failure, shock, or multiorgan dysfunction, and the same places are reporting CFRs of around 2.5%. In general, 0% of people in respiratory failure survive without hospital care, shock is exceedingly low, and multi-organ failure also has a brutal survival rate without hospital care. And remember it is not just the vent. If you go into shock you may need volume repletion, pressers, antibiotics (all pneumonias can turn into bacteremia too easily), and a variety of other things that may be required, and we have had patients who needed all of that without needing a vent. Frankly the biggest thing I see are patients whose O2 sats suck and, so far, are not doing too horrible on oxygen. If you can think of a better place for mass oxygen therapy than a hospital with oxygen literally piped through the walls, do let me know.

In any event, going from the Italian figures in mid-March, about 12% of their Covid patients needed to be hospitalized. At the time, their CFR was around 7 - 7.5%. This suggests that 46% of hospitalized patients, in an overwhelmed system, were surviving.

The other way I look at this things is that we have a lot of docs out there. If treatment were futile on average, I suspect you might be able to tell me that 95% of clinicians are stubborn, emotionally invested, or whatever; but in a time a of limited data I am not hearing this position from anyone, at all, on the ground.

And, again, going from the Italian data we saw what happened when they began turning patients away, a giant spike in the death rate.

You would need an awful lot of contortions to make hospital care as a whole a net negative and still match the reported numbers for CFR, percentage with respiratory failure, etc.

Remember, the normal ICU death rate for influenza patients is 30-40%. So if this thing is twice as bad, most of the arguments about how dangerous the hospital system are would also apply to annual flu deaths.

Now maybe, when we run out of PPE and everything else is social distanced like crazy, then the hospital is the only major vector left. I could potentially buy that. But the numbers from everywhere but China do not look quite that grim, suggesting, once again, that Chinese data was intentionally manipulated.

Always good to see reality checks from the field. Thanks for taking the time to write this.

Where we have good data, perhaps the question is, at what age is ventilation no longer an effective treatment option? Aggregating lots of age data obscures determining this useful heuristic for the labour it could free up.

Why would we do that? Lungs are vastly more viable within age cohorts than between them.

For instance, if you smoked and have COPD, your lungs will be terrible at oxygenation and it is quite likely that you are oxygen dependent at baseline. A 60-year-old COPD patient is likely going to have vastly worse lungs than an 80-year-old with otherwise healthy lungs. Likewise, if your immune system sucks or if you have heart failure, your odds of dying even on the vent are vastly higher than people older than you without such morbidities at your age.

Most of the time, we judge this by clinical values. How bad is the consolidation on CXR or CT? What is their PFT? What is their ejection fraction? Is the axis deviation on the EKG? What is their presenting O2 sat? What is their minute ventilation? We can pretty well predict who is going to be low odds on a vent from a lot of clinical data far better than using age as a proxy for that clinical data.

When you start using age, is when you literally cannot get that clinical data. When provider time is so precious (due to the patient count) that you cannot properly triage, then you use age because it is quick and cheap to find. Better to guess wrong a bunch of time than to have all the patients die while you do even a couple minutes of workup.

Of course when you are in those situations, the unfortunate truth is that an experienced physician's (or experienced nurse's) gut is likely better even than age. A thirty second size-up is likely to beat a strict age criterion easily. But we do not like having physicians literally deciding who lives and who dies so we use age because it provides a rationale rather than just being somebody's gut telling us they should die.

I appreciate there are always judgements involved, often with many factors. I was expecting, based upon some of the papers I've scanned, that there is an age after which the likelihood of survival is so low as to not be worth pursuing, when the overall system is under strain. In engineering, we commonly employ 'load shedding' when the demand exceeds capacity. Consider an aircraft taking off, when power is a premium. On modern aircraft (specifically the 787) designers ensure power is not applied to galley ovens during takeoff. I'm merely arguing there should be a form of 'triage' applied as system load approaches capacity, and this might be an example of one of those. It is better to have such policy decisions thought out in advance than in the heat of an emergency.

The media is full of stories about doctors and other health care workers being laid off or furloughed. This is supposed to be because treatments that can be postponed are and the space is being taken by Covid-19 victims. Wait a minute. What were the health care workers now serving in the fight against the virus doing before it arrived? How is it that hospital personnel, trained medical professionals, can't be employed in this situation? Something doesn't add up in those whole scenario.

An example of healthcare workers being laid off is a freestanding same day surgery center. Those centers cannot get their normal supply of PPE, drugs, or other supplies. However, the employees are not trained to work in ICUs, ERs, or acute care wards. Eventually some of them will be move to such facilities but one of the most underreported stories is how Covid-19 has totally overwhelmed the management abilities of healthcare. Before Covid-19 managers were rewarded for controlling costs and keeping patient satisfaction scores high. Now the job is one of logistics, emergency planning, being flexible, and not worrying about costs. Not the same skill set.

What does opening up the economy really mean? Does that mean everyone just goes back to work and accepts the 1% chance of dying a horrible death.

Or maybe that everyone who can work from home, stays home. That the education systems stays shut. That masks and distancing are used to open up any closed factories and the as much of the supply chain as possible.

However, is anyone's life worth going to the movies or hanging out at a bar. Is a fund raiser for the private school really worth 50 people being infected with Covid-19.

There are many things that can be done to help the economy without killing a couple of million people.

People are bad at isolation. You see them in groups on the street. You see them picking up just a couple things at the store. Running to In-n-Out.

My fear at this point is that if people don't know how to lock down, we won't ever see R < 1.

In that case opening up with limited protection, masks for people who will wear them, isolation for the stubborn, might be what we get.

And NY won't be the only run-away situation.

As you know if "the gubment" hadn't shut these things down, everyone would blissfully be piling into them like nothing was going on because they don't give a shit about dying, getting permanent lung damage, or transmitting those same risks to family and friends.

No, the economy would be EXACTLY the same as it was before the plague if only "the gubmint" would get out of the way.

It's not like an uncontrolled deadly plague would shut down the economy all on its own no matter what "the gubmint" did and trying to control it is actually the only sane way to keep the economy from collapsing.

It was the private sector that responded first not the government. Many meetings and events like SXSW were shut down because the private sector refused to send their employees. The NBA shut down on its own without being prompted by the government. Disneyworld and Disneyland shut down well before any idiot governor or mayor was telling companies to shut down.

That is why all of the idiots who say that we can "open the economy" are fools. Many people would refuse to do it and many companies that tried to forced their employees back to work would eventually be sued our of existence, have employees quit and will have a few employees attack their idiot bosses.

Does anyone think that any employers will send their employees to a convention while the number of Covid-19 infections is increasing? Does anyone really want to get on a cruise ship or commercial aircraft while depending on the flight crew to wear masks correctly and maintain a very high level of cleanliness?

Also, if someone gets Covid-19 at work, what will that do to the workman' comp premiums, liability insurance rates of the employers, and the legal risks. My guess is that the plaintiffs attorneys are just waiting for the pandemic to claim down so that they can start suing every stupid employer or organization.

Aka the "the government needs to shutdown the economy because otherwise the private sector would shut down the economy anyway" argument.

We could compare the up and coming hotspot map to the government action map:

I don't have a feel for the areas other than Colorado, but the "Hot Spots #2" that are weighted toward population are calling out a bunch of sparsely populated mountain towns. Most of the cases are not locals but people who brought the disease with them on ski trips.

For God's sakes, they list Hinsdale County as a HOT SPOT and it has, no joking, 794 residents over 1,100 square miles! 1 confirmed case. They've got to filter out the low volume noise.

I guess there would be some level of idiot bosses and owners who decide to alienate their employees and ruin their reputation in their own comunity to try to make a few dollars. Have you not noticed how many businesses have been shamed into doing the right thing?

What happens when some employees shoots their idiot boss when the boss tries to use threats and force to "open the economy" back up.

It is all the U.S. can do to keep food distribution and emergency healthcare open. Their are no cleaming supplies or PPE to try to get everything open quickly.

The problem with using early data is, the earlier it is on the learning curve.

But other than that, a lot of commoners seem to think that institution should easily abandon their missions, and unilaterally and individually adopt extreme triage.

They can't. In most cases it would be a crime.

Maybe you want acting president Jared kushner to institute martial law.

lol, "commenters"

Let’s not pretend you didn’t mean commoners ;)

Very Smart Person sees 2 possible reasons to "shut down society" if not mitigate hospital overload.

1. Get better at testing
2. Develop "vaccine/treatment"

But VSP only discusses vaccine ("18 months" though I'd offer 4-1 odds on sooner).

What about treatment? Hasn't evidence moved even in past week towards Plaquenil as a treatment?

Won't coming 4 weeks offer decent shot of additional treatments, ramped up production distribution of HCQ, and edits/updates to exactly when/dosage/alternate formulations for HCQ?

Why isn't "buy some time, even weeks, to develop off-label treatments and then produce/distribute those pills widely" a good rationale?

Right. I stumbled over my phone typing above, but this is learning curve I was talking about. For treatment.

I did mention this 28Mar:

A related economic claim:

"Wearing a cloth mask when you go out is probably worth $5,000 to the economy. Shocked? You shouldn’t be. Now go read this whole thread!"- @EggersMatt

No judgement one way or another on the cloth mask claim and $ saved. Could be true, potentially.

But his paper literally cites a tweet as evidence that N95 respirators don’t prevent infection.

Scott Alexander walker through the evidence. That combined with what we know about Covid and viral load and the conclusion is obvious.

I mean, it’s a free country. More N95 masks for me I guess

or the fashion and textile industries start retooling to produce a new line of business suit/clothes that integrate PPE functions. That creates the effected social distancing via a personal "environment" an allows normal working to resume largely eliminating the economic costs of avoiding infection.

These do not have to be level 4 bio quality things, just reasonable isolation and production costs, with some ability to for simple decontamination similar to what we do with our normal clothing (washable as we know soap will kill the virus).

In a lot of industries we already have a distribution channel for such things as uniforms and uniform cleaning services.

too bad there is not edit period/function....

I should also add it seems to be the case that we want to control two factors: length of time being exposed and the quantity in any initial exposure.

Not confirmed but it does make sense and some evidence seems to be pointing towards this.

Even homemade masks may serve the purpose of limiting both the initial "viral load" as well as the flow rate of virus during exposure to a contaminated environment. So we don't need N95 type solutions here.

We do need some work on getting a sense on just what flow rates and load sizes define the thresholds.

Some bad inferences here.

Nearly all (inevitable) deaths go down a pathway which includes a ventilator. Doesn't mean ventilator wasn't critical in the 20% of ventilator survivors.

Plus Wuhan comorbidity issues. Western ventilator survivor rates likely to be higher.

“ I’m pretty sure I’d be doing all this and more if I were in charge. But I’d also be looking for evidence that what we are doing is the best course of action, given the massive costs.”

And this is a very smart person? Really Tyler?
He has no idea whether this will work and knows that this has massive costs and he will do more of it?

What has the world come to? Where has all the courage gone?

A good primer is how the decision for ventilators is handle, one can watch Extremis (2016) on Netflix

It was nominated for an Academy Award. It includes a discussion about the use of ventilators for people who are probably never going to recover. The film also covers how people need to make decisions about how they want to be treated now rather than forcing it on relatives/others later.

If the writer of the e-mail is "very smart" he'd have sent it two or three weeks ago. So why have you been sitting on it, Mr Cowen?

Robin Hanson's great book "Elephant In the Brain" has a couple of chapters devoted to this very topic.

Hint: Futile interventions are a feature, not a bug. They tell the rest of us (family, friends, community) we are willing to do anything, especially when its futile.

I commented on this before - what is the price of turning hospitals into hospice centers for the contagious dying - rather timidly a few days ago because I was not sure it was something one could say aloud - and I assure you it isn't - elsewhere than a blog like this ... But one other result of the current policy is it seems it will kill a lot of frontline doctors and nurses, and near as I can tell that suits the media and the Greek chorus just fine. They can feel they are making a sacrifice, an effort, while displacing the real sacrifice onto "heroes" that they support with their homemade mask-making and their "let's light a candle at our houses tonight to make a beautiful display of light to support our medical professionals" stuff.

I guess it's not axiomatic to me that doctors should have to die for this.

Yesterday, an acquaintance walked by and hailed me. She was very steeped in coronavirus news, as I suppose all but the strongest characters among us are. She also seemed very energized by it. We chatted pleasantly, trading, no doubt, our mis-conceptions and fake news and imperfectly-recalled facts. I alluded to the dosage effect I learned about here. She knew all about it. I said, it seems like it might partially explain why doctors are falling ill and dying - their prolonged exposure.

No, she assured me, that was not why that was happening - that was because they lack enough masks, or have to use the same one all day.

That was interesting to me that she thought that; her own father headed a MASH unit ...

Important questions but you don't get to ask questions for free. Questions take time to answer and time is done at this point. Studies at this point on death rates are unreliable. Were those put on ventilators given optimal care or were they made to wait for other patients to die to free up a ventilator?

I've been debating with someone on Facebook whose been arguing the virus is only killing people over 80. After finally looking at his article, I saw it is several weeks old when Italy had only 6K deaths, now it has twice. If 5 people come in at 6 AM and go on a ventilator, one 90, one 80, one 70, one 60 and one 40....the fact that the 90 and 80 year old are dead at the end of the day doesn't tell you everything is fine for the ones 70 and under.

S. Korea hints that given optimal conditions, the death rate could be 0.6% or lower making this about 6 times as bad as a rough flu, but that should be manageable. But evidence from other countries hints at death rates of 5% or worse.

Long story short we don't know the death rate, the hospitalization rate under ideal conditions nor under varying conditions. More importantly we know we cannot reliability obtain this information until long after the dust has settled and a dozens of students have earned their PhD's in epidemiology by shifting thru the data. We must make decisions without this information.

From my point of view everything has to remain shut until mask production, testing and ventilator production is scaled up. We are now 3 months into this and tests are still being rationed and $0.94 masks are unavailable.

Stewart Brand say 66% dies:

The mortality odds on ventilator for Covid patients: "Of the 98 patients who received advanced respiratory support...66% died," reports a UK study cited in a paper linked below. Dying on ventilator also means dying alone and speechless in the ICU. Discussion...

It is becoming increasingly apparent Tyler's doctor friends are outpatient plastic surgeons or something. The 'smart person' email is insipid.

And its funny that right here we get actual informed comments debunking the 'facts' and assumptions of that email, and regulars here keep right on using those facts and assumptions.

A good perspective on their comments on other topics.

Yes, Cuomo prattling on about ventilators is so much public square bloviation. If you're sick enough to need a ventilator, you're already f**cked. Not so near term, plasma treatment; longer term, vaccine. We need to be ready when the next virus hits.

>So what does the best evidence say here?

Oh, I dunno..... Trump's fault?

22% at most of the people under ventilator died... by March 31. Since they do not tell us the number of discharged people, the death rate might be
significantly higher. Anyway, just another data point:

Considering that the number of deaths per capita appears to be much higher in areas where the hospital system is overwhelmed, I'm inclined to conclude that hospitalization appears to be helpful to a significant degree. For example compare NYC, New Orleans and Detroit to other large cities in the US like Chicago or LA which are not yet overwhelmed and the number of fatalities per capita is in the 20/100k range for the former while it's in the low single digits for the latter.

I picked US cities because the level of service is likely to be similar. You could make a similar case for Germany, Austria and Switzerland as compared to Italy, Spain and France, or Wuhan vs the rest of China.

Of course there are confounders - later start of infection, deaths of other causes being lumped in with Corona-related deaths and test capacity being limited to severe cases are three of them - but it seems to me that they would have to be unreasonably big to explain the whole effect.

Googling the topic it seems that between 50 and 95% of people die even though they get respirator treatment, so let's put it at maybe a 25% survival rate. So yes, respirators may get too much spotlight and we need better data, but overall the evidence I see points to the availability of hospital care still being quite helpful in reducing the number of fatalities.

I think the problem here is too many economists are looking at spot measurements as if they were fixed constants they could plug into their models when in fact they are variables.

The mortality rate, for example, varies with whether you're going to get high quality care or if you're in a hospital where the system is collapsing. There is evidence if you catch the virus in a hospital with a lot of very sick people the disease is worse than catching it 'normally'. That means mortality is going to be lower where there are ample PPE supplies and strict infection control is able to be maintained versus a free for all where people are using coffee filters and rubber bands as substitutes for PPE while trying to care for overcrowded wards of people coughing. The same probably applies to ventilators. Are you being checked on every 15 minutes and a respiratory specialist updating settings frequently or are you hooked to a machine that's hooked to 2 or 3 other people?

Even with all this the rates are probably not static over time. As more and more patients move through the system HCPs will notice patterns and are likely to be able to tease out better results with respirators (assuming they don't collapse first).

The type of static calculation Tyler is trying to do here is not going to work in a dynamic system.

For what it is worth, I’m not a skeptic of the current social distancing policies. I’m pretty sure I’d be doing all this and more if I were in charge. But I’d also be looking for evidence that what we are doing is the best course of action, given the massive costs.

If I were in charge, I would be strongly recommend that everyone going out of their home in close proximity to other people should wear washable gloves and some sort of mask or multi-wrap scarf. I'd further recommend that they wash those gloves and scarves or cloth masks every night.

I'd also recommend that if they were frequently in close contact with people, they should also buy and wear a head sock/balaclava and face visor (or glasses).

I'd recommend that nursing home might want to consider even stronger measures.

And then I'd comment every day on what percentage of people appeared to be following my advice, based on videos from around the country.

It's a valid question to ask but it's based on wrong data. To conclude that mortality is 80%-90% amongst the hospitalised cohort, the "very smart person" uses a tiny data set out of China based on only c. 100 people.

Here's a better study from the UK. About 50% of those admitted to intensive care survive. This is highly influenced by age as well - for under 50-yos, 75% of those admitted to intensive care survive.

Just came back 4 days later to tell the UK PM is hospitalized with supplemental oxygen.

I'd love to read the comments of the smart person on this event since there's no evidence on the effectiveness of these treatments (supplemental oxygen, IV fluids and antibiotics).

Comments for this post are closed