Hacker News Reader

Hacker News

Comment by Retric | Hacker News Reader

Parent Full thread

Retric

A large number of South Korea infections are even younger due to how it spread.

Italy: Above 80: 1,532 (18.4)% vs age 20-29: 296 (0.0)%

South Korea: Above 80: 243 (3.1)% age 20-29: 2,261 (28.7)%

https://en.wikipedia.org/wiki/2020_coronavirus_pandemic_in_S...

https://en.wikipedia.org/wiki/2020_coronavirus_pandemic_in_I...

Of note, neither country has had anyone under 30 die. Deaths in the 80+ age range are 8.23% in South Korea and 13.2% in Italy.

PS: South Korea saw 2 deaths under age 50 where Italy has seen 1. So this age is by far the largest factor.

View on HN

A large number of South Korea infections are even younger due to how it spread.

Another possibility is that Italy doesn't have a handle on who is or isn't infected so what they report are the obvious infections, who tend to be older.

>Another possibility is that Italy doesn't have a handle on who is or isn't infected so what they report are the obvious infections, who tend to be older.

This is very likely the case, as can be seen in the positive test outcome rates.

16.7% in Italy vs. 3.6% in South Korea

Italy has run around 60k tests.

South Korea are testing around 15k people/day and have tested 234k people in total. They've even setup drive-through testing centers where anyone can drive through, get tested, and get results via SMS in under a day.

The result of this is likely a much larger number of asymptomatic or mild cases in the South Korean numbers, which based on the China data, tend to be a lot younger.

https://www.cdc.go.kr/board/board.es?mid=&bid=0030

https://github.com/pcm-dpc/COVID-19/blob/master/schede-riepi...

Yup. Meanwhile, the New York Times has run an article about how northern Italy is "a warning to the world" with the eye-catching claim that "about 50 percent of the people who tested positive for the virus required some form of hospitalization": https://www.nytimes.com/2020/03/12/world/europe/12italy-coro... Between this and other reporting, I'm starting to wish they hadn't opened up their paywall for coronavirus articles.

This makes sense. Now that Italy has ramped up testing to 13k per day we should see a correction in due course.

So if we assume a similar death rate, we could use the deaths to estimate how many are actually infected in Italy?

Yes, not only in Italy, but in most other places. This article explains well how deaths (or hospitalizations) are correlated to actual infections (most of which go unnoticed), and also why reduction of social interaction is effective but its effects take a while to show:

https://medium.com/@tomaspueyo/coronavirus-act-today-or-peop...

Nice - or as satisfying as something can be under the horrible circumstances - this is exactly what I've been arguing here.

In many (technically former) "red zones", especially in Lombardy, they tested also asyntomatic family members of infected people, or their contacts. I'm not sure how inaccurate it is, but I have no idea what has been done in SK.

South Korea traces the movement of everyone who tests positive, puts the movements online and tests those who were exposed to someone positive (this is in the article I believe). It's a massive effort but with the world economy at a stand-still, what else do people have to do?

North Italy has a population of 27 people btw so the 10K that are registered as positive as a drop in the bucket.

I guess I'm lucky enough to know one of these 27 people.

Fine, 27M(million)

"Why must I always explain" Bob Dylan

It is possible you are right. IIRC, 17% of those tested were positive in Italy. That's a very high percentage. Very likely the actual infected number is much higher.

Yup, you probably only have to look at the counts of infected tourists coming back. In Norway that is 89 from Italy, and 112 from Austria (25% of the total infected as of writing...)

True, same here, in The Netherlands. Of the first 500 reported cases, around 25% were tourists returning from the Alps region (northern Italy, western Austria), and around 50% were traced to secondary infections from those first 25%. The other 25% were unexplained, hence we're now in the endemic spread phase.

FAFAIK, we've never had a policy to pre-emptively test every returning tourist, and even now we don't test everybody. The current policy is to test new cases only if they have no relation to an already-established case. For example, people living in the same house as a known Corona case will not get tested. If they start showing symptoms, they're automatically assumed to have contracted it.

Is Norway testing everybody coming back into the country?

No. In fact they just revised their testing criteria and will not test most people who are asked to quarantine at home. Testing is expected for people who have acute respiratory illness, people who are hospitalized, health care workers handling these cases, and immune-compromised people with mild symptoms.

It’s unlikely that 14,000 infected 20-29 year olds where missed as a significant fraction end up in the ICU. Which means even if the ratio is off, the breakdowns are still significantly different.

I'm not sure how the media should report on deaths over 80, or even over 90.

It's not that old people aren't worthy of protection, but when you're 85 in the US you have a 10% chance of dying in the next year for any and all reasons.

So saying this disease kills 15% (for example) of 80+ year olds without context sounds horrific, but contextualized in the actual risks of just being very old to begin with, sounds much less scary.

If Italy, for example, were to report on the number of fatal infections in people younger than the natural life expectancy, the fatality rate would drop in half.

Fatality numbers won't reflect those who developed permanent lung damage or had to go into intensive care.

Best not to suggest anything that might down play the seriousness considering how infectious this virus is. (Otherwise it may encourage risky behavior.)

I think it's best not to overstate like the media currently has been (at least here in the USA) which is that everyone should be panicky and worried if they'll be alive next week.

Permanent lung damage? Link please?

It's far too early to draw any conclusions, but the HK Hospital Authority has suggested that they are seeing reduced lung function in some "recovered" patients.

https://www.scmp.com/news/hong-kong/health-environment/artic...

but what about the risks of over-selling this disease? Look at Italy for example. I've read self-reports from healthcare workers claiming 200% hospital capacity. I can only image the needless death resulting from such panic.

Mind you that isn't covid 19 putting them over-capacity. In the worst hit region in Italy (Lombardy) confirmed covid 19 cases represent enough cases to take up less than 10% of their hospital beds. So it seems incredibly unlikely anything but panic is to blame for such absurdly high over capacity.

> In the worst hit region in Italy (Lombardy) confirmed covid 19 cases represent enough cases to take up less than 10% of their hospital beds.

Do you have a source for this comment? This WaPo story from 6 hours ago says differently:

https://www.washingtonpost.com/world/europe/italy-coronaviru...

> Giulio Gallera, Lombardy’s health chief, said Thursday that the region would reach its capacity in “five, six or seven days,” even if it tried to add more beds in hospital “cellars.” In an interview with Italy’s La7 channel, Gallera described the possibility of adding 500 intensive-care beds at Milan’s expo center, the kind of rapidly assembled zone that China created in the hard-hit Wuhan area.

lombardy has 7280 confirmed cases [1] lombary has a population of 10MM (wikipedia) italy has 3.18 hospital beds per 1000 [2]

the math for that adds up to 31,800 hospital beds 7280 confirmed cases is 22% - so my sources were off or wiki is wrong as originally I recall reading lombardy has a 16MM population.

I will say that your article claims they haven't run out of ICU beds yet but that number is surprisingly high for ICU needs to cover covid 19. The article implies 600 ICU cases from covid 19 - that's over 12% ICU from confirmed covid 19 cases, which is more than twice what china reported.

It's also very surprising how few ICU beds they have allocated towards ICU....

The USA has something like 14% of hospital beds as ICU beds [3] but Italy appears to only have have less than 3% of their beds available for covid 19 ICU

[1] https://www.statista.com/statistics/1099375/coronavirus-case... [2] https://en.wikipedia.org/wiki/List_of_OECD_countries_by_hosp... [3] https://www.sccm.org/Communications/Critical-Care-Statistics

Most hospitals don't have all that much slack in bed capacity; especially at the tail end of flu season most of those beds are already filled. The hospitals may be extrapolating the growth rate of the disease and realizing how soon they will have problems. Or given the nature of the outbreak it could be that even within Lombardy there are areas with much higher concentrations of cases, like there are more cases, inc. per capita, in Wuhan than other cities in Hubei.

What's not plausible is ICU beds being taken up by people who don't have the disease (maybe miild-to-moderate flus or bad colds) and are just panicking. Emergency room lines, sure. Test shortages, absolutely. But hospitals will not put someone who's just scared and doesn't even need to be admitted at all into the ICU.

> Most hospitals don't have all that much slack in bed capacity; especially at the tail end of flu season

Remember that there's a difference between hospital beds and ICU beds.

In the UK, we maintain 80% utilisation of our ICU beds all year round[0], with very little change in the number of beds available in real number terms. We have ~4100 total ICU beds, which can be expanded to ~5000 if all operating theatres, etc are shut down and used as ICU equivalent instead.

While getting past flu season will help the total number of beds, it does nothing for the ICU.

This is the other reason why it's increasingly important to https://www.FlattenThesCurve.com and employ social distancing techniques. The rate of patients being admitted into ICUs needs to be slowed as much as possible. Once we're out of capacity, every additional patient has a much higher probability of dying than the overall fatality statistics indicate.

[0] https://www.england.nhs.uk/statistics/wp-content/uploads/sit...

There's a typo, it's https://www.flattenthecurve.com/

The NHS regularly has to cancel routine operations in order to free up space for flu patients during the flu season, which might be one reason why the bed usage seems so uniform. Obviously if they're already doing that for flu it leaves less slack in the system if something else comes along.

1. Not every hospital bed is the same as a bed for a highly infectious disease

2. Other things still happen to people and those people need to have their hospital beds.

3. I'm not a hospital planner but I think there may be other things that determine capacity than just hospital beds. I worry that the hospital bed metric is actually a bad metric. Obviously adding hospital beds implies adding people to staff those beds, but also adding people probably also means adding all sorts of other infrastructure to support those people.

As a developer it reminds me of a project that goes way off schedule, you don't bring the project back on track just by adding developers because of the overhead more developers add to your system. Probably most hospital administrators have never actually experienced operating at capacity, so estimations of what would happen could be off and are being corrected now.

Italy has about 12.5 ICU beds per 100,000 people. ICU beds are not the same as hospital beds.

https://www.statista.com/chart/21105/number-of-critical-care...

Maybe you shouldn't transfer every confirmed case into hospital since most cases will do just well ... unless they don't. (Not questioning the danger of this virus.)

Your 10% premise is wrong because you're conflating intensive care capacity and regular hospital beds. You can't just put Covid patients in with the general hospital population and it isn't trivial to convert other hospital beds to the requirements of Covid while ensuring you keep everyone else in the hospital safe from infection. Lombardy does not have anywhere near enough intensive care capacity to handle the cases they've already seen, much less anything further.

You can see that in the death rate, it's happening because they can't care for the volume of older intensive care patients properly. They don't have the medical infrastructure capacity to do it. That's why Italy and other European nations such as France have an extreme variance in mortality rates, despite France and Italy both having similarly old populations. Germany is even older than Italy and I'm skeptical we're going to see anywhere near a 6% mortality rate there (their rate is very low so far). Italy's mortality rate has been abnormally high from the first few thousand cases in, when it quickly began overloading their healthcare system in Lombardy.

A quick google suggests around 15% of hospital beds are ICU and covid 19 cases that end up in the hospital, only 5% end up in ICU. So that's an interesting line of thinking but it lacks supporting evidence that covid 19 is overwhelming ICU.

https://www.nejm.org/doi/full/10.1056/NEJMoa2002032

edit to address your edit: The mortality rate is often wildly inflated with new diseases. That's because most people don't seem to understand real mortality rate vs confirmed cases mortality rate. Obviously when a disease is new the confirmed cases will be confirmed against the people showing serious symptoms. Then when you pretend deaths/confirmed cases == mortality rate you end up with a wildly inflated mortality rate. Until we understand how many people are asymptomatic we won't have a real mortality rate that is reliable.

Percentage of ICU beds varies widely. Minnesota, for instance, a state in the US, has about 500 ICU beds [1] and over 10,000 hospital beds total [2]. ICU beds are already heavily used and sometimes there simply are not enough, before COVID-19 has even shown up. Average inpatient stays right now are 4.21 days in Minnesota [3]; I can't find a reliable number for length of stay for COVID-19 patients, but it is longer.

There is plenty of evidence that in Italy COVID-19 has overwhelmed ICU beds. They are trying to transport people to ventilators via helicopter. We do have at least twice the ICU bed capacity in the US (compare info from [4], [5]). But many other people need these ICU beds -- it's not like other illnesses will just stop for our convenience.

[1] https://minnesota.cbslocal.com/2020/03/08/talking-points-pre... [2] https://www.ahd.com/states/hospital_MN.html [3] https://www.mnhospitals.org/mn-hospitals/quick-facts [4] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4351597/ [5] https://link.springer.com/article/10.1007/s00134-012-2627-8

Edited to add: remember, included in most totals of ICU beds are neonatal and pediatric ICU beds, which won't help much with COVID-19.

The ICU beds are also taken by all the other sick people. Only a minority is actually free. And people don’t end up in the ICU because they panic: I spoke to several Italian hospital doctors and they are overwhelmed by the troves of patients who can’t breathe.

Right: we do have those ICU beds because people are in them.

Toolz, I just don't understand your math. You seem to assume that a neonatal ICU bed is just as good for a 55-year-old man as an adult one, and that no one is going to have any other ICU-needing illnesses for the duration of this epidemic. That's quite puzzling. Can you support these ideas?

I just read an article that claimed Lombardy has 737 ICU beds available for covid 19. Given Italy has a 3.18 beds per 1000 people that means they have less than 3% of their hospital beds available for covid 19 patients. That's absurd. In relation the USA has roughly 14% of their beds as ICU beds according to a reference from https://www.sccm.org/Communications/Critical-Care-Statistics

You seem to think that for some reason, ICU beds go unused except in a situation like this? ICU capacity is probably built to meet some fraction of demand in normal circumstances. Those beds weren't all unused before this outbreak started.

737 ICU beds available isn't the same as the total number of ICU beds. Normal health problems that need hospital treatment (strokes, heart attacks, car accidents, etc.) are still happening during this outbreak (well, car accidents are probably reduced now with Italy's lockdown protocols).

Someone else posted else thread that in UK at any time 80% of all ICU beds are in use. It is likely a similar percentage in Italy.

Hospitals worldwide have very little spare capacity and even less spare ICU capacity. A normal hospital bed isn’t going to help much. They are very likely to be overwhelmed, especially in countries that have not been able to delay initial spread due to lack of testing and containment.

This probably won’t kill millions (though it could worst case) but hospitals being overwhelmed would lead to an exponential rise in deaths, and many deaths from other causes as other treatments are abandoned. The situation is serious and will require significant measures to bring it under control, it is not yet being taken seriously enough in the US.

Are ICUs even helpful against the disease? I wonder if any triage experiments are being run with randomization.

When you have ADRS ICU treatement is pretty much the only thing keeping you alive. "Randomized trial" in this context means russian roulette - if you don't get a ventilator you'll die almost certainly.

Evidence [0] (partly based on COVID-19, parly on MERS) also says that NIV has a significantly worse outcome than ventilation. People are already getting NIV in some hospitals because they ran out of tubed ventilators.

[0] https://www.who.int/docs/default-source/coronaviruse/clinica...

An additional factor in why the death rate is so high in >75 year olds is that most deaths happen when ICUs are overwhelmed, and older ages get triaged into not getting a ventilator. There is a much higher rate for them naturally, but this amplifies the disparity even more.

You will not feel this way when you're 80

The problem is that this disease forces everyone to pretend they are 80. Give the old the care they need but don't drag down the young.

you could also say that for people 80 and above, this epidemic will more than double their already significant chances of dying this year...

My parents are in the range

And they live in Italy

Still the heat wave in 2003 killed 18 thousands people in a month and 15k in France in only half a month.

Old people are vulnerable to many things, unfortunately

That's a really hard thing to say this early. For all we know many 80+ people are asymptomatic and most older people aren't negatively affected at all by covid 19. It seems really dangerous to conflate actual mortality rate with mortality rate among the confirmed cases. Those numbers will be drastically different with diseases that have many asymptomatic carriers.

https://www.worldometers.info/coronavirus/coronavirus-age-se...

80+ age category has a 21.9% death rate in confirmed cases.

Of those in that group what % had a pre-existing condition? That is, is it age or age + illness? My sense is it's the latter.

I don't have the numbers ready, but my guesstimate is that about 100% had the pre-existing condition of old age.

Also when 85+ year olds die for other reasons, they don't all occupy ICUs for days or even weeks as in the case of COVID-19.

I'm surprised to hear someone be so blunt about their ageism. This thing has brought about a great deal of it already, with many saying, " it only affects the elderly" as if that's a worthwhile argument for not being too worried. Reminds one of indifference to AIDS in the 80s because it "only affects homosexuals"

> It's not that old people aren't worthy of protection, but when you're 85 in the US you have a 10% chance of dying in the next year for any and all reasons.

Some of the people that die from this would have died anyway, but for the most part this is on top of that 10% chance. Same for young people, there's something like a 0.2 chance of death every year and this is basically doubling that.

The numbers for 20-29 are probably not comparable. South Korea has been testing a lot of people per day. Italy, like many other countries, is overwhelmed, so there are probably many healthy-feeling infected 20-somethings who haven't been tested and aren't counted.

Reply on news.ycombinator.com