So when you say “upwards” of 5% of cases require the ICU, this is potentially very misleading. Stating that upwards of 5% of hospital cases require the ICU may be true, but now you must not try to extrapolate that value any further until you can tell me definitively what percentage of infections result in hospitalization? 1 in 5? 1 in 20?
I won’t get into debating R0 values because they are nearly impossible not to misuse. For example, COVID-19 does not have just one R0 value. R0 changes radically based on actual conditions on the ground, often day to day.
The R0 value of COVID in a cult church in South Korea may be actually 1,000 while the R0 globally over the entire course of this news cycle might be 1.2.
No, they are accurate as far as I know and I've been following this thing with some dedication since before it was discussed in the media because of relatively close ties to China.
> You forgot the most important qualifier, which is that the denominator is some fraction (potentially small fraction) of the total number infected.
You are misrepresenting the OP's words. He said 'of cases', that is the denominator right there. That we don't know all cases is something that everybody takes as a given. But just like we don't know all cases we also don't know all of the dead. The same happened with the swine flu, with SARS and with every other epidemic so far. Only afterwards, sometimes more than a year later do we have all the numbers. And sometimes those end up substantially higher than the original estimate both for the number of infected people as well as for the number of people that end up dead.
> So when you say “upwards” of 5% of cases require the ICU, this is potentially very misleading.
Upwards of 5% of the known cases. But that's assumed, if we don't know someone is infected then they by definition won't require ICU level care.
> Stating that upwards of 5% of hospital cases require the ICU may be true, but now you must not try to extrapolate that value any further until you can tell me definitively what percentage of infections result in hospitalization? 1 in 5? 1 in 20?
They don't need to tell you anything 'definitely' to be allowed to speak their mind. And as far as I'm informed that 5% rate is on the money. So, instead, the onus is on you to prove that less than 5% of the people infected that require care require ICU care. I don't think you are able to do that but I will happily take good news at this point if properly sourced.
> I won’t get into debating R0 values because they are nearly impossible not to misuse.
Agreed.
> For example, COVID-19 does not have just one R0 value. R0 changes radically based on actual conditions on the ground, often day to day.
> The R0 value of COVID in a cult church in South Korea may be actually 1,000 while the R0 globally over the entire course of this news cycle might be 1.2.
True, but there are averages and even the WHO seems to place the number at > 3 at this point in time. That is already a pretty high number, even when taking all the care in the world to ensure it doesn't spread. Of course the criminals in Korea that hope to bring about the end times in some kind of warped self fulfilling religious prophesy are going to skew that number upwards. But in Iran and Italy there was no such cult, and their numbers are very much in line with the development in Korea. And since we're mentioning Iran now, there are lots of informal reports that the situation there is much worse then officially reported, with the number of high officials there infected this - unfortunately - isn't hard to believe at all.
I think you may be giving OP too much credit with their use of “cases” implying hospital cases (which is the 5% ICU number, it is not 5% of infections require the ICU) I’m happy if we are agreeing and will leave it at that.
> Upwards of 5% of the known cases. But that's assumed, if we don't know someone is infected then they by definition won't require ICU level care.
You are still making too strong of a statement. In the study last week from NEJM of 1099 hospital patients around China that found that 90% plus were diagnosed with pneumonia, and 5% were admitted to the ICU. Case fatality rate was 1.4%. [1]
It is absolutely not 5% of “known” cases. China turned anybody not seriously ill from the hospitals back home. So we actually do know there are a large number of uncounted mild (and even asymptomatic) cases.
It is 5% of serious hospital cases going on to the ICU. This is not my claim, this is straight from the latest issue of the NEJM.
Also, I didn’t say anyone couldn’t speak their mind. But I said if you do try extrapolating a sub-population percentage and multiplying by world population, it would be a serious miscalculation. That’s why I think it’s important (crucial) to clearly identify the sub-population when you state the percentage. Calling a sub-population dealing with hospital admittances the “known” cases is not clearly stating identifying the patient population.
I don’t think HN commentators should play epidemiologist while making predictions of catastrophe.
Where did that happen? In this thread or elsewhere?
Because it leads to fear-mongering. For example, downthread, here;
But you did imply that. Look, I am all for keeping a cool head. But at the same time I'm not going to cool my head by sticking it in the sand. I think if you're going to participate in these discussions then you probably should do so by making your statements a lot clearer than you do.
If you - for instance - idd not intend to suggest that someone did try to extrapolate those numbers then you did a poor job of it. You can't just make up strawmen and then suggest people shouldn't do that.
The comment you linked to was properly qualified and is not fearmongering. If you think that is fearmongering then you possibly have a reading / comprehension issue.
Let me help you by picking that comment apart:
> The 1.4% apply to the general population.
You can't really argue with that I think,
> For people over 80, the fatality rate is 15%.
Not exactly correct, I have it at 14.8% once infected and showing symptoms. But close enough.
> There are roughly 13 million people over 80 in the US [1].
This is properly cited.
> If everyone gets infected, a fatality rate of 15% means 2 million deaths in that single bracket alone.
Again, this is properly qualified by the leading 'If'
> With the 1.4% figure it's 4.6 million over the entire US population.
Again, with the implied assumption that people are all infected, something that is unlikely to happen.
> As a comparison, the number of yearly deaths in the US caused by influenza is in the 5 digits.
This is correct as far as I know, between 10 and 60K cases annually ranging from very good years to very bad ones.
> World war 2 killed 3% of the world population over a range of multiple years [2].
Can't argue with that.
> If the coronavirus kills only 1% of the current world population, its death toll in absolute numbers will be equal to WW2.
Again, properly qualified by a conditional.
> I think at this point, the disease can't be contained, its spread only slowed.
I don't think anybody is arguing that it can be contained still, even the WHO has moved on this, though they kept it up for longer than I personally think was responsible.
> Everyone will be exposed to it sooner or later.
Probably correct, similar to the flu, most people get exposed to it in their lives and usually once per year.
> Fortunately though there are treatments, and there are already candidates for vaccines. The question is only how quickly and how well those treatments can be deployed.
If there is one part of this comment that I have issues with then it is this one. There are - as far as I know - no vaccine candidates that have been trialed. I'm happy to be proven wrong here so any link is welcome.
> The less people have the disease at the same time, the better.
And that is the key insight: the whole thing revolves around managing the capacity to deal with the people that are ill. That and effectively reducing R0 < 1 are the main tools that we have at this stage to deal with this particular virus.
Now, you want to keep playing that game.
To extrapolate over the general population with the 1.4% case fatality of serious hospital cases, would be wrong and would be fear mongering. The general population does not have a infection rate of 100% nor a case fatality rate of 1.4%.
So the premise is wrong, the assumptions are wrong, and the result are wrong.
>> If everyone gets infected, a fatality rate of 15% means 2 million deaths in that single bracket alone.
> Again, this is properly qualified by the leading 'If'
Qualifying a statement ending with ”means 2 million deaths in that single bracket alone.” with a qualifier that is false is what I would call fearmongering.
>> With the 1.4% figure it's 4.6 million over the entire US population.
> Again, with the implied assumption that people are all infected, something that is unlikely to happen.
No, it’s not an “implied assumption” it’s faulty epidemiology. First, it is faulty to assume that everyone who is exposed will be infected. Second, it is faulty to apply the 1.4% case fatality rate of serious hospital infections to the entire US population.
The assumptions are wrong. The figures are wrong. The result is wrong. It is putting false information based on poor understanding of the facts into the discussion and it’s gone totally off the rails. Please, you should seriously reconsider why you are supporting this.
I am not playing games. This stuff is far too serious to play games over.
> To extrapolate over the general population with the 1.4% case fatality of serious hospital cases, would be wrong and would be fear mongering. The general population does not have a infection rate of 100% nor a case fatality rate of 1.4%.
No, but the Spanish Flu for instance had an infection rate of about 25%, and the general infected population could very well be above that 1.4% (in fact, there is every indication so far that it will end up somewhere between 2 and 3%).
> So the premise is wrong, the assumptions are wrong, and the result are wrong.
The premises isn't nearly as faulty as you make it out to be, it might be off by a factor of four, hopefully much less than that. But so far - taking Wuhan as our baseline - unchecked this thing will spread very fast and kill lots of people. The only reason the onslaught was halted is because the Chinese took a bunch of draconian counter measures, the likes of which very few countries in the world will be able to copy.
> Qualifying a statement ending with ”means 2 million deaths in that single bracket alone.” with a qualifier that is false is what I would call fearmongering.
That's fine by me. But it isn't. As a whole it isn't and in the details it isn't either. We actually have precedent for this whole situation, it is the 1918 pandemic and the world apparently hasn't learned its lesson fully. Which may mean we are about to re-learn it the hard way. Maybe this time it will stick. But then - as now - there were plenty of people like you who were willfully - given the way this conversation is progressing I am assuming you are doing this on purpose - ignoring the cold hard facts. And that made the pandemic much worse than it had to be.
Large events were not called off and led to the disease spreading much further than it had to.
Absent an effective anti-dote to this virus and with an overwhelmed health care system (note the qualifiers!) we might as well be in 1918.
> No, it’s not an “implied assumption” it’s faulty epidemiology.
Not as far as I can see. People talk about hypotheticals all the time. You can't really come out like this against a pretty clear comment and at the same time make up stuff you rail against that never happened in the first place. Fix your own issues first.
> First, it is faulty to assume that everyone who is exposed will be infected.
Agreed, so let's divide the number by 4 and call it a day. That's the worst this could possibly get. Right? Or do you think that given a sample of '1' the 1918 flu might not be the best guideline for precision and that it might be better or worse? We don't actually know the answer to that.
> Second, it is faulty to apply the 1.4% case fatality rate of serious hospital infections to the entire US population.
That's true. But for all the same money we end up with a higher case fatality rate. So that may end up affecting the equation in a negative way.
But assuming 10% of the country becomes infected and 5% of those end up with symptoms serious enough to warrant ICU care there is a serious problem.
And that 10% could very well be low. So as far as I'm concerned you should start taking this a bit more serious instead of trying to shout down the discussion. Stay dispassionate and try to argue your case with actual numbers rather than nay-saying. That way you can contribute rather than just irritate. Think of it as an opportunity to educate, and you just might learn a thing or two yourself.
I just want to say, it’s easy to take the wrong numbers, multiply together, and see 100 ways that millions of people die. That’s fear mongering.
What I believe is that it’s incumbent on anyone who wants to participant in predictive discussions to do so responsibly and using only the best numbers that we have on hand.
Anything could happen. And this is always true, whether it’s the Ebola discussion from a while back, or SARS, or MERS, or COVID today.
You still seem to insist on claiming that we could see pervasive worldwide infection, 5% ICU admits, and 1.4%+ general population fatality. I hope you will soon realize that is not actually representative of what we are facing, where in fact, the vast majority of cases are mild or even asymptomatic. I expect this will be clear even to the most avid doomers by mid-April.
Lastly, I don’t want to go back and forth on little things, but this;
> at the same time make up stuff you rail against that never happened in the first place. Fix your own issues first.
I think you need to re-read what I originally wrote. I never strawmaned.
I said the numbers were all wrong as they were stated due to lack of proper context. I could have used the word “useless” or “misleading” instead of “wrong” and that might have been better.
I then followed up by saying it’s imperative not to try to use those numbers to extrapolate in the general population. I could have said “one must not”. I was not accusing Op of doing this.
I was trying to say that the danger of not properly qualifying the context of a percentage like case fatality is that people will misuse it by improperly extrapolating.
Then we found an example down thread which does exactly that... and you defend it anyway. Which is fine, but I stand my ground on the argument (same as the NEJM doctors) that it’s improper to extrapolate the hospital case fatality to the general population.
If one “assumes that the number of asymptomatic or minimally symptomatic cases is several times as high as the number of reported cases, the case fatality rate may be considerably less than 1%.” (NEJM)
But you yourself are not doing that. You quote data that has been long ago overrun by the reality on the ground.
> You still seem to insist on claiming that we could see pervasive worldwide infection, 5% ICU admits, and 1.4%+ general population fatality.
We could. We also could not. Let's hope for the best and plan for the worst. That way we can all laugh at this a year from now.
> I hope you will soon realize that is not actually representative of what we are facing, where in fact, the vast majority of cases are mild or even asymptomatic.
I do not doubt that at all. But that does not rule out that the numbers could still be quite bad.
10% infected and 1.4% CFR would be very bad news.
> I expect this will be clear even to the most avid doomers by mid-April.
Maybe those 'avid doomers' aren't the doomers. The doomers are the cultists and the 'we're all going to die' people. We're not all going to die. But unless we take substantial countermeasures a lot more people will die than need to. Look at Iran for a taste of that. And that's the message that should go out: reduce the opportunity for the virus to spread so that we get to R0 < 1 as soon as possible and then mop up from there. Anything less and it will just get worse and worse.
BTW that NJEM paper you keep referencing is very out of date. Yes it passed peer review and was published last week, but it's using data from Jan 29. It was actually written nearly a month ago; I found a preprint of that paper posted Feb 3. You'd be aware of this if you were actively following the novel coronavirus situation.
There's a more up-to-date paper published by China CDC that covers 72,314 patients / 44,672 confirmed cases using data thru Feb 11. I would reference figures from that paper instead.
> I think you may be giving OP too much credit with their use of “cases” implying hospital cases (which is the 5% ICU number, it is not 5% of infections require the ICU) ... It is absolutely not 5% of “known” cases.
No, he's correct. Current evidence shows at least 5% of confirmed cases require ICU care. From the WHO China report, they note that 6.1% of all cases are critical. They define critical as the following: "Critical cases are defined as respiratory failure requiring mechanical ventilation, shock or other organ failure that requires intensive care."
These numbers are not just a China fluke. We're seeing similar critical percentages in other countries. For example, as of today, Singapore has 106 confirmed cases of which 7 are critical (~6.6%).
> I don’t think HN commentators should play epidemiologist
Isn't that what you're doing?
[0] http://weekly.chinacdc.cn/en/article/id/e53946e2-c6c4-41e9-9...
[1] https://www.who.int/docs/default-source/coronaviruse/who-chi...
> In China, the overall CFR was higher in the early stages of the outbreak (17.3% for cases with symptom onset from 1- 10 January) and has reduced over time to 0.7% for patients with symptom onset after 1 February (Figure 4). The Joint Mission noted that the standard of care has evolved over the course of the outbreak.
So to highlight that, going forward, CFR of 0.7% of lab diagnosed cases.
For instance, this week 4 South Koreans died while waiting to be hospitalized. An additional 600 sick are still waiting for hospital beds. They don't have enough beds! And this is in the country with the 2nd most hospital beds per person!
Many of the serious, and almost all of the severe cases will perish without medical care. Once the medical system is overwhelmed, the CFR could climb past 10%.
https://www.reddit.com/r/Coronavirus/comments/fbxm0s/south_k...
https://en.wikipedia.org/wiki/List_of_OECD_countries_by_hosp...