Help the Fight Against COVID-19

Status
Not open for further replies.
Respectfully, the only estimates I gave were (a.) rough, intended only to casually indicate order-of-magnitude; and (b.) limited to the U.S. alone.

(Not, of course, because the rest of the world isn't important! People are people, everywhere...but my post addressed the U.S. because I'm ignorant of nearly all the relevant factors in every other country apart from the U.S. and thus have nothing to say, and also because I was making a point about local- and state-level decision-making in the U.S., and how we should expect it to look.)

Are you replying in kind? That is, are you predicting fatalities of "tens, if not hundreds, of millions" in the U.S. alone?

"Tens...of millions?" That's "tens," plural? In the U.S.?

That would be 20 million dead, minimum. Are you offering a prediction that this virus, this year, will kill 6% of the U.S. population? That seems wildly pessimistic given the publicly-available information.

You even raise the possibility of "hundreds," plural. If I take you literally, that'd be 200 million, in a country with a population of 350 million. Are you predicting this virus, this year, will kill 43% of the U.S. population?

You also say, "all the evidence points...." You must have evidence nobody else has. What's your source?

To make what you're saying more reasonable, I have to conclude you're talking worldwide.

In that broader scope, 20 million would be more reasonable (if that's the word to use, for something so horrific). But even at the worldwide scope, your "hundreds of millions" would seem unduly dark, given the progress that's being made. (That'd be 3%-ish of the world population, I think?)

Respectfully, please clarify your usage, and if you can offer a few bullet-points to highlight "all the evidence" that points to these outcomes, I'd sure like to see them.
330M * 80% infected before herd immunity * 15% hospitalization - 300k hospital beds available = 39.3M who need hospitalization who can't get it. + deaths of non-COVID-19 patients who need hospitalization. + impact on economy cus millions are dead, which also causes deaths as you've noted.
 
330M * 80% infected before herd immunity * 15% hospitalization - 300k hospital beds available = 39.3M who need hospitalization who can't get it. + deaths of non-COVID-19 patients who need hospitalization. + impact on economy cus millions are dead, which also causes deaths as you've noted.
Yeah, but @yeky83, I don't see anything like 15% hospitalization being required.

(An aside: I note that you're using 330M, which is the U.S. population, give or take a few million. So, I take it that you believe @FractalAudio's numbers were about the U.S., alone?)

I think the 15% number comes from treating the "people we've tested positive" figure as if it were the "people who actually got the virus" figure, and then extrapolating hospitalization rates and fatalities from that.

If I'm correct about that, then the actual hospitalization rates are well under 5%, and might be under 1%. Or to put it differently: I think we have 68,000 confirmed cases precisely because the virus has infected, oh, 500,000 people, plus or minus 100,000, and the vast majority of them have recovered asymptomatically or with only 2 days' of fever to let them know they were ever sick.

Now if you get 10,000-ish hospitalizations out of 68,000 confirmed cases, you're gonna say, "15% hospitalization rate."

But if it's 10,000-ish hospitalizations out of 500,000, then we're talking 2%. That's 7 million beds needed, for an average of 10 days each, for 70 million days' worth of hospitalization.

If those days of hospitalization all show up at once, then yeah, we're toast because we haven't got that many beds, or even close.

If, however, they're spread out over a year's time (if, in short, we "flatten the curve" like we've all been talking about for weeks), then we're talking about an extra 200K beds needed. That's still incredibly bad, but it's not 39.3 million sick-people-lingering-in-the-streets-bad.

Furthermore, consider what else is meant IF I'm right (yes, I acknowledge that it's an if) that the actual infections are some multiple of the confirmed cases: That would mean that, for every case that went to the hospital (and, a month later, had produced either a death or a recovery) there were 2, 3, 5, 8, or goodness knows how many cases that had produced a recovery and an immune individual.

If that's right, then it's mathematically impossible that the (currently-exponential) growth rates don't flip from exponential to logarithmic well before you get to 50% of the population having been infected. (This shouldn't surprise us; it's normal for biological growth patterns.) After all, if you're still only testing dire cases, then you know the real cases are a multiple of the dire ones. Let us suppose that, by April 15th, some 10% of the population has tested positive with dire symptoms. That's bad, very bad. But what else should we conclude? We should conclude that, what, 40%? ...50%? of the population is already infected without dire symptoms, not requiring hospitalization, and by May 15th will be recovered-and-immune. And you can't sustain an exponential growth rate in a population where every other target-of-infection is immune.

Meanwhile we have (a.) treatment that looks likely to shorten hospital stays in advanced cases and (b.) make them unnecessary in non-advanced cases. And we have skin-prick serum-antibody testing -- a home test, at that -- that'll show us whether I'm right that the "confirmed cases" number is less than one-seventh of the number of actual infections...not to mention that it'll identify the "already immune" persons who're ready to go back into the workforce and the restaurants.

I don't think I'm being Pollyanna-ish, here, unless somehow that 15% number was calculated off actual infections, both known and unrecognized. (But I don't see how it could be.)
 
Last edited:
Yeah, but @yeky83, I don't see anything like 15% hospitalization being required.

(An aside: I note that you're using 330M, which is the U.S. population, give or take a few million. So, I take it that you believe @FractalAudio's numbers were about the U.S., alone?)

I think the 15% number comes from treating the "people we've tested positive" figure as if it were the "people who actually got the virus" figure, and then extrapolating hospitalization rates and fatalities from that.

If I'm correct about that, then the actual hospitalization rates are well under 5%, and might be under 1%. Or to put it differently: I think we have 68,000 confirmed cases precisely because the virus has infected, oh, 500,000 people, plus or minus 100,000, and the vast majority of them have recovered asymptomatically or with only 2 days' of fever to let them know they were ever sick.

Now if you get 10,000-ish hospitalizations out of 68,000 confirmed cases, you're gonna say, "15% hospitalization rate."

But if it's 10,000-ish hospitalizations out of 500,000, then we're talking 2%. That's 7 million beds needed, for an average of 10 days each, for 70 million days' worth of hospitalization.

If those days of hospitalization all show up at once, then yeah, we're toast because we haven't got that many beds, or even close.

If, however, they're spread out over a year's time (if, in short, we "flatten the curve" like we've all been talking about for weeks), then we're talking about an extra 200K beds needed. That's still incredibly bad, but it's not 39.3 million sick-people-lingering-in-the-streets-bad.

Furthermore, consider what else is meant IF I'm right (yes, I acknowledge that it's an if) that the actual infections are some multiple of the confirmed cases: That would mean that, for every case that went to the hospital (and, a month later, had produced either a death or a recovery) there were 2, 3, 5, 8, or goodness knows how many cases that had produced a recovery and an immune individual.

If that's right, then it's mathematically impossible that the (currently-exponential) growth rates don't flip from exponential to logarithmic well before you get to 50% of the population having been infected. (This shouldn't surprise us; it's normal for biological growth patterns.) After all, if you're still only testing dire cases, then you know the real cases are a multiple of the dire ones. If 10% of the population has tested positive with dire symptoms, then 30%? 40%? 50%? of the population is already infected without dire symptoms, not requiring hospitalization, and by next month is recovered-and-immune. And you can't sustain an exponential growth rate in a population where every other target-of-infection is immune.

Meanwhile we have (a.) treatment that looks likely to shorten hospital stays in advanced cases and (b.) make them unnecessary in non-advanced cases. And we have skin-prick serum-antibody testing -- a home test, at that -- that'll show us whether I'm right that the "confirmed cases" number is less than one-seventh of the number of actual infections...not to mention that it'll identify the "already immune" persons who're ready to go back into the workforce and the restaurants.

I don't think I'm being Pollyanna-ish, here, unless somehow that 15% number was calculated off actual infections, both known and unrecognized. (But I don't see how it could be.)
We can agree the current data is lacking, but we can also agree risk assessment shouldn’t be made on lower bound estimates. But I agree, it should prob be 5% hospitalization. 13M, still in the tens.

If we spread the curve over a year’s time, we probably don’t get to save the economy.

And you’re talking about HCQ/CQ right? Doesn’t seem to be as effective as you make it out to be, but maybe we’ll figure it out.
 
Nobel Prize winner is basically saying what I'm saying wrt the importance of seeing new cases slowing down: https://www.latimes.com/science/story/2020-03-22/coronavirus-outbreak-nobel-laureate
Hmm.

I wish I could confidently add Michael Levitt's analysis to my list of "reasons we're not all going to die." (Well...actually, we are, just not all this month.)

I wish I could wave him around and say, "See...?" But, I worry because he's leaning partly on numbers from China, and I don't trust those numbers as far as I can spit a rat. It's Potemkin villages, all the way down.

Now, South Korea's numbers, I trust much more. And they're suggestive of a slowdown.

But I expect Italy's numbers to be pretty realistic, too, so it's a bit disheartening to see they're still on the upswing.

If all these countries (and my own) really had widespread-and-statistically-valid testing, I'd be happier with our ability to predict. As it is, though, most of the U.S. is still rationing testing only for the folk with the direst life-threatening symptoms, and we're all absurdly trying to use that data to figure out how bad the disease is. That's like trying to figure out what the average life-expectancy for a human being is, and taking your entire sample-set from dark alleys in Detroit.

And, now that I come to think of it...I wonder whether South Korea's numbers, or Italy's numbers, really mean what we think they do? Or are their numbers all just one giant artifact of how they select whom to test, and not strongly correlated to what the disease is really doing?
 
Belgium and Bahrain report that the Trump pills are working.

Be VERY careful with messages like these.

Your statement is simply not correct.

Belgium is testing and states:

“Use of (hydroxy)chloroquine is off-label and with risks (interactions, serious side effects), no effect has yet been demonstrated for curing COVID-19 and certainly not preventative. Pay attention with azithromycin: no proven efficacy yet against COVID-19 and numerous risks!”
 
Be VERY careful with messages like these.

Your statement is simply not correct.

Belgium is testing and states:

“Use of (hydroxy)chloroquine is off-label and with risks (interactions, serious side effects), no effect has yet been demonstrated for curing COVID-19 and certainly not preventative. Pay attention with azithromycin: no proven efficacy yet against COVID-19 and numerous risks!”

How do you know it's "not correct"?

Quote from the Bahrain News Agency (BNA): "
The drug is reported to have had a profound impact when used to treat the symptoms exhibited by active Covid-19 cases.
Bahrain first used the drug on Feb 26, following the registration of its first Covid-19 case on Feb 24."

Bahrain has 419 cases. 4 (four) deaths, and 190 recoveries.

Belgium: "Belgium’s Federal Agency for Medicines and Health Products (FAMHP) is reserving hydroxychloroquine, an anti-malarial drug, for patients who really need it in light of the new coronavirus (Covid-19)." Looks like, unlike Bahrain, they've just started testing it, and their death rate so far is pretty bad. We'll see what happens when they start giving it to more patients.

They wouldn't be "reserving it" if it there was evidence that it does nothing or causes harm. They'd instead ban it for off-label use. Neither would India. It's also confirmed in-vitro.
 
Last edited by a moderator:
How do you know it's "not correct"?

Quote from the Bahrain News Agency (BNA): "
The drug is reported to have had a profound impact when used to treat the symptoms exhibited by active Covid-19 cases.
Bahrain first used the drug on Feb 26, following the registration of its first Covid-19 case on Feb 24."

Bahrain has 419 cases.

Belgium: "Belgium’s Federal Agency for Medicines and Health Products (FAMHP) is reserving hydroxychloroquine, an anti-malarial drug, for patients who really need it in light of the new coronavirus (Covid-19)."

They wouldn't be "reserving it" if it there was evidence that it does nothing or causes harm. They'd instead ban it for off-label use. Neither would India. It's also confirmed in-vitro.

If you really don’t see the discrepancy between “the Trump pills are working” and the actual Belgian statement, it’s no use continuing this discussion.
 
Link to in-vitro study: https://www.ncbi.nlm.nih.gov/pubmed/15351731 as well.

We should know this by mid next week for sure. Cuomo got beat over the head with a shovel a bit, so he allowed the trial to proceed in NY. Why would anyone ban a drug that could help dying patients is beyond me. Nevada too banned the drug outright for off-label use, without any evidence for what looks like political reasons. It's illegal for them to ban it, since FDA approved it for off label use federally, but they're trying anyway, because god forbid it works, that'd be a real tragedy.
 
Isn't hydroxychloroquine touted as an effective treatment based off the French trial with the ridiculously low sample size (20 people) and skewed numbers? When I say skewed, I mean skewed as in - non-random, incorrect sample testing for lower viral counts and not counting the people dropped out of the trial due to side effects of the medicine and/or hospitalization from the virus, one of which who died! That said, if I'm in the hospital I'm taking anything they give me provided I'm coherent enough to make any decisions. Let's not get too excited about this treatment yet though.

source
 
Last edited:
New York newspaper finding the worst affected hospital in New York and making you believe that's how it is across the board is selection bias, however. That's not to say the situation is easy there. That is to say that if they had a choice between showing a hospital which is doing OK and one that's completely fucked, you know exactly what would get printed. That's selection bias. If they were concerned with presenting the truth, they'd show another hospital where, perhaps, things aren't so bad. But they don't really give a shit. Panic sells subscriptions.
You're displaying what is known as "normalcy bias". You refuse to accept the grim reality of what is unfolding.
 
Respectfully, the only estimates I gave were (a.) rough, intended only to casually indicate order-of-magnitude; and (b.) limited to the U.S. alone.

(Not, of course, because the rest of the world isn't important! People are people, everywhere...but my post addressed the U.S. because I'm ignorant of nearly all the relevant factors in every other country apart from the U.S. and thus have nothing to say, and also because I was making a point about local- and state-level decision-making in the U.S., and how we should expect it to look.)

Are you replying in kind? That is, are you predicting fatalities of "tens, if not hundreds, of millions" in the U.S. alone?

"Tens...of millions?" That's "tens," plural? In the U.S.?

That would be 20 million dead, minimum. Are you offering a prediction that this virus, this year, will kill 6% of the U.S. population? That seems wildly pessimistic given the publicly-available information.

You even raise the possibility of "hundreds," plural. If I take you literally, that'd be 200 million, in a country with a population of 350 million. Are you predicting this virus, this year, will kill 43% of the U.S. population?

You also say, "all the evidence points...." You must have evidence nobody else has. What's your source?

To make what you're saying more reasonable, I have to conclude you're talking worldwide.

In that broader scope, 20 million would be more reasonable (if that's the word to use, for something so horrific). But even at the worldwide scope, your "hundreds of millions" would seem unduly dark, given the progress that's being made. (That'd be 3%-ish of the world population, I think?)

Respectfully, please clarify your usage, and if you can offer a few bullet-points to highlight "all the evidence" that points to these outcomes, I'd sure like to see them.
Tens of millions worldwide.
 
Isn't hydroxychloroquine touted as an effective treatment based off the French trial with the ridiculously low sample size (20 people) and skewed numbers? When I say skewed, I mean skewed as in - non-random, incorrect sample testing for lower viral counts and not counting the people dropped out of the trial due to side effects of the medicine and/or hospitalization from the virus, one of which who died! That said, if I'm in the hospital I'm taking anything they give me provided I'm coherent enough to make any decisions. Let's not get too excited about this treatment yet though.

source

The expert I spoke to yesterday was very unimpressed by the potential of that drug for mitigating the problem.
 
Status
Not open for further replies.
Back
Top Bottom