Hacker News Reader

Comment by adrianN | Hacker News Reader

It kills 1% if ICU care is available. It kills a lot more if the healthcare system is overwhelmed. Look at Spain or Italy. Case fatality rate is around 10%.

Letting it run its course it needs to infect 60-70% of the population before herd immunity kicks in. 1% of 60% of the population of the US is 2 million people. World-wide you're looking at 45 million people for a very optimistic scenario. The numbers could easily be 5 or 10 times larger. If we're lucky and the number of asymptomatic cases is very high, the numbers could also be lower. At this stage we really don't know, but I don't think it's ethical to take that risk.

View on HN

Spain and Italy have fatality rates of 10% because the infection exploded far before they could ramp up testing capacity, so they were only testing the people who were in serious medical distress.

This is obvious if you compare the numbers between different countries versus their relative level of hospital capacity.

The comment you're replying to referred to "1% of the people who get it", which has basically nothing to do with CFR. Getting the virus is necessary but far from sufficient to become a case.

Not quite. 1% of people who get it don't survive when healthcare resources are not strained. But when ICUs fill up, that number starts to skyrocket, because people who need intensive care cannot get it.

You're really missing my point. CFR is not the percentage of "people who get it" that don't survive. It's the percentage of "people who get it, get tested, test positive and are thus confirmed as a case" that don't survive. The actual mortality rate, whether ICUs are working or not (they mostly are), isn't known.

That was a prediction CDC made in early Feb for broader China. More recent CDC reports for United States (https://www.cdc.gov/mmwr/volumes/69/wr/pdfs/mm6912e2-H.pdf) say mortality is somewhere between 1% (lower bound) and ~30% (depending on age) based on calculations they did w/ cases having "known outcomes". The CDC says they only have 44 such samples at this time, and that ICU status data is missing or unknown for the several thousand or so other cases. So we really just don't know the facts yet.

The case fatality rate is meaningless when testing isn't done extensively.

I know. But we don't know the true fatality rate yet and we won't know it before it's way too late to act. I don't think it's okay to literally risk millions of lives on the off chance that we missed 95% of the cases with our testing.

That is a reasonable line of thought, but incomplete. How many lives are we risking by tanking the global economy? That seems like it should be a more quantifiable risk. It becomes a trade-off between the amount of lives we can calculate will be lost via our pandemic-prevention measures (and to what degree of certainty) and the number that might be lost without them (and to what degree of certainty).

The trade-off is between loss of lives one way versus loss of lives another way.

It is quantifiable, however it appears that periods of economic decline are strongly associated with decreased mortality rates, at least according to the best evidence I can find from Nature.

Okay, but you have to look at both sides of it. How many people aren't dying in road traffic accidents that don't happen because of isolation? How many people aren't dying of diseases they won't develop because reduced economic output has also reduced the output of carcinogens and toxins produced by manufacturing that's not being done? And like that - those are just the first two examples off the top of my head; there are more.

I don't see much mention of this among the "back to work at any cost" crowd, and I wonder why. It seems like a pretty obvious consideration.

Exactly this! The WHO estimates that air pollution kills 7M people annually[0]. A really big range of estimates suggest 20-50% decrease in airborne pollutants since the start of the pandemic. Even a 10% decrease, sustained, could save 700k lives/year.

Yes, bad economies harm people. But they also reduce harm. Pointing simply at directionality due to a bad economy is disingenuous.

[0] https://www.who.int/health-topics/air-pollution#tab=tab_1 [1] https://www.bbc.com/future/article/20200326-covid-19-the-imp...

Now do a calculation that includes a world war with nuclear weapons, because that's what happens when you hit 30% unemployment.

How on Earth do you figure that?

A natural outcome when you have 30% unemployment in the worlds largest economies. Who knows, maybe we will get lucky and it will only be civil wars in China and India with only hundreds of millions dead.

No one in Italy has ever been denied an ICU bed yet.

I'd like a source for that because it directly contradicts news reports I've seen.

Source is Ministry of Health, data is collected by each Region. Specifically this is the free icu rate per region https://infogram.com/tasso-di-occupazione-posti-letto-1h7k23...

Unfortunately it's live without the per-day history, but you should consider that we're at the peak right now.

Considering the many people downvoting I feel like US news suggest otherwise. Weird!

In Italy you see 10% mortality because Italy tests only (really) sick people. It's really likely that millions of italians have covid. This is not really demonstrable at the moment, but still https://www.wantedinrome.com/news/coronavirus-report-six-mil...

Italy is treating every single patient with high standard healthcare. They did not reach the "0 free ICUs" moment yet.

What we can say as of today is that the high number of deaths do not depend on overwhelmed hospitals, but likely just because there are A LOT of covid patients. Also a greater average age and the fact that older people are much more integrated in the society played a role affecting mortality, but still those are only theories and we can't say how much they affected the total.

sick of hearing this silliness: https://www.ecodibergamo.it/stories/bergamo-citta/coronaviru... so yeah, they barely test half the dead people (which are still double the normal deaths...). And A LOT of covid patients means a lot of deaths as well, bc. at 1.3% (Diamond Princess, where people are still not through it/dying) to 5% letality, the thing IS BAD.

But again, what was the age group on Diamond Princes? Because it matters, the death rate goes up with age. 5% says the average age was something like 50.

A cruise ship skews older and most people from Diamond Princess should be out of it by now.

with emphasis on "should". Most people started to argue with the low number of deaths, when there were 5. Now there are at least 10 and about 100 people unaccounted in statistics...

Hm - 10 as of a few weeks or so ago, I doubt we will see any more and certainly not 2-3x more. That 10/800 = 1.2% and the cruise ship skewed older. 100 people entirely unaccounted for would only impact spread rates, not CFR, unless the unaccounted people are more likely to have died than those from the ship writ large.

I don't see where people are getting 5% from. 1% is already 10x worse or so than the flu, I don't see the need to exaggerate the facts.

There was an 11th death Apr 1, and 113 of the 712 are still active, not recovered. Didn't find from a quick search how critical or mild they are, but it is concerning that it's been this long...

That doesn't appear to be entirely correct - from my research, it appears to actually have been about 12 deaths (the last on March 28th from https://www.mhlw.go.jp/stf/newpage_10599.html) but there appear to be multiple sources on this.

Regardless, I'll admit that I was wrong - there are still people dying. Regardless, we'd have to get about 3-4x the number of deaths we've had so far to reach 5% - and from the sources I can find most of those remaining cases are non-critical (though there could be even a doubling).

The US media reporting has also been terrible. CNN even noticed that a lack of testing was correlated with a higher reported percentage of deaths... and then went and concluded that the reason for this was that more testing caused fewer people to be infected, even though this wouldn't directly affect the mortality rate and there was other evidence indicating that the lack of testing had screwed with the denominator in the calculation.

Reply on news.ycombinator.com