totm april 2020 Coronavirus

[tater]

39 minutes ago, Mikenike said:

False, the death rate of your flu is higher than the corona virus. The media has hyped up this virus more than it should, happened with SARS, swine flu, Ebola. I bet good money that the Yellow fever would kill more than corona.

This is abject nonsense. And this is coming from someone who has been saying the CFR that's published is too high because we don't know the actual number of cases.

Bad seasonal flu has a total fatality rate of ~0.1%. That's TOTAL, not case fatality rate. Estimated number of people infected (say 30M in the US in one season), divided by the number of deaths (say 30k in a typical year). Note that regular seasonal flu can be lower than 0.1, and minus the elderly, far lower.

The death rate of this is certainly higher than 0.1%, though the 3.4% number used by WHO right now is also clearly too high since the denominator in the CFR is confirmed cases, which is a small subset of the total cases. Countries with more testing (SK) have a lower CFR since they have a bigger denominator. It's a fair bet that the total number of cases is 10-20X the confirmed cases. So at the outside (3.4/20=0.17), the CFR is still greater than 0.1. What's gonna drive the mortality up will be people who otherwise would have lived (mostly elderly) who will not get a vent because we don't have enough beds. Some cases that might require little active care and a vent might not have one to get. It's gonna be a mess.

This is important: a hospital "bed" is not a room with a bed in it. It's not the bed itself. An ICU bed, or a ventilator is not just the physical items, it's those items with a team of nurses and docs associated with it. Yeah, the team has multiple concurrent patients—but not too many, that's the point (~10? per intensivist, but there are multiple nurses, and they are calling consults from multiple services as well). We could not make more ICU "beds" short of a training program that would push for more ICU docs and nurses in 10 years, maybe (assuming people want to become intensivists). If you add more patients per existing team, each patient WILL get worse care. Some ICUs don't use the intensivist model, and they have worse outcomes. So you can add less good ICU care, but a % will die because of this choice.

 

[VoidSquid]

According to this paper from the university of Oxford, updated March 11th:

In the period up to and including 9th March 2020, the global Case Fatality Rate for COVID-19 are as follows.

Case fatality rate globally = 3.48%
[based on 109,578 confirmed cases and 3809 deaths]

Case fatality rate in China: 3.86%
[based on 80,904 confirmed cases and 3123 deaths]

Case fatality rate for the rest of the world: 2.39%
[based on 28,674 confirmed and 686 deaths]

Source: https://ourworldindata.org/coronavirus

[tater]

A short blurb from John Torres about our reporting problems:

I think the last time I saw him was in our old house, lol. He went to med school with my wife, nice guy (AF pilot before med school), did ER residency.

3 minutes ago, VoidSquid said:

According to this paper from the university of Oxford, updated March 11th:

In the period up to and including 9th March 2020, the global Case Fatality Rate for COVID-19 are as follows.

Case fatality rate globally = 3.48%
[based on 109,578 confirmed cases and 3809 deaths]

Case fatality rate in China: 3.86%
[based on 80,904 confirmed cases and 3123 deaths]

Case fatality rate for the rest of the world: 2.39%
[based on 28,674 confirmed and 686 deaths]

Source: https://ourworldindata.org/coronavirus

Yeah, those are the numbers that will certainly be resolved downwards when they do a retrospective total fatality rate. We'll probably need serology tests for that so they can then get an idea how many of us had mild disease and never knew it (we'll have antibodies).

On a personal note, last night I was texting a couple buddies who like me have wives who are docs... we all expect to get this, because we expect that everyone in healthcare will get it.

https://www.unz.com/jthompson/coronavirus-dilemmas/

[tater]

BTW, the point of the transmissibility WRT interventions is to drop it below 1 so it peters out like seasonal flu. If it's 3.5 we have to nearly halt any transmission between people. It's probably more like 2.x, and we need to move the dial by more than half.

[VoidSquid]

21 minutes ago, tater said:

Yeah, those are the numbers that will certainly be resolved downwards when they do a retrospective total fatality rate. We'll probably need serology tests for that so they can then get an idea how many of us had mild disease and never knew it (we'll have antibodies).

Agreed. The reason I was posting this was that in my opinion, and I've seen in here too, some people are just throwing numbers around without having any solid knowledge and data based on. I do not like that.

[RCgothic]

No evidence of UK govt making significant emergency preps. We're headed for Italy.

Edited March 12, 2020 by RCgothic

[DDE]

2 minutes ago, RCgothic said:

No evidence of UK govt making significant emergency preps. We're headed for Italy.

It may be much, much worse than merely no emergency preps.

https://www.itv.com/news/2020-03-12/british-government-wants-uk-to-acquire-coronavirus-herd-immunity-writes-robert-peston/

In other news, Russia may or many not have its first COVID death.

https://ria.ru/20200312/1568510703.html
Rumoured to be a 75-year-old who'd been to Italy; confirmed only to have died in Moscow's designated COVID containment hospital.

[ProtoJeb21]

Looks like my school is shutting down because of the Coronavirus. I don’t know the full story yet, might post more later depending on what is announced. 

[StrandedonEarth]

The NBA and NHL have now suspended their seasons. So much for going to a game next week....

Edited March 12, 2020 by StrandedonEarth
Proofreading before posting is overrated

[RCgothic]

34 minutes ago, DDE said:

It may be much, much worse than merely no emergency preps.

https://www.itv.com/news/2020-03-12/british-government-wants-uk-to-acquire-coronavirus-herd-immunity-writes-robert-peston/

Herd Immunity is absolute rubbish when it comes to Covid19.

In the absolute most charitable interpretation, it's basically "vulnerable people need to hide until Coronavirus has run it's course through the healthier and more resilient segment of population".

UK govt policy is to stand back and watch people die. "Oh dear, so sad, never mind."

[tater]

1 hour ago, RCgothic said:

Herd Immunity is absolute rubbish when it comes to Covid19.

In the absolute most charitable interpretation, it's basically "vulnerable people need to hide until Coronavirus has run it's course through the healthier and more resilient segment of population".

UK govt policy is to stand back and watch people die. "Oh dear, so sad, never mind."

Herd immunity is sort of a thing, the more people who have had the disease before, the fewer people can be infected. Ie: you're in a group, and would infect 4 people, but if 2 of them already had it and are immune, you only infect 2. The R0 for you just dropped in that case from 4 to 2.

There's an old idea, Farr's law of Epidemics that says that all are bell curve shaped (in given wave, anyway).

 

The "on order of common flu is literally correct (same order of magnitude), but it's high for flu at 0.2%.

The principle concern still remains our parents (for those of us here who are older), or our grandparents (for you young 'uns).

Edited March 12, 2020 by tater

[RCgothic]

54 minutes ago, tater said:

Herd immunity is sort of a thing, the more people who have had the disease before, the fewer people can be infected. Ie: you're in a group, and would infect 4 people, but if 2 of them already had it and are immune, you only infect 2. The R0 for you just dropped in that case from 4 to 2.

There's an old idea, Farr's law of Epidemics that says that all are bell curve shaped (in given wave, anyway).

 

The "on order of common flu is literally correct (same order of magnitude), but it's high for flu at 0.2%.

The principle concern still remains our parents (for those of us here who are older), or our grandparents (for you young 'uns).

Herd immunity means if a virus normally spreads to 5 other people, but 90% are immune, then each infectee only infects half another person and the infection dies or over time. Which means the 10% who aren't immune mostly never encounter it.

For Covid19 nobody is immune. In order to reach herd immunity first a massive number of people need to get infected. Which is a disaster for a disease with 1-2% mortality.

[tater]

2 minutes ago, RCgothic said:

For Covid19 nobody is immune. In order to reach herd immunity first a massive number of people need to get infected. Which is a disaster for a disease with 1-2% mortality.

The mortality is not going to be that high. The number of actual cases is a multiple of the number tested.

Each year in the US the CDC reports several hundred thousand positive flu tests. The US has several 10s of millions of actual cases. The ascertainment rate is really low. Additional cases are "diagnosed" short of testing (tell your doc you have some symptoms that include body ache, and "you have the flu.").

My guess is multiply the number confirmed someplace with decent numbers tested by at least 10. That means divide the fatality rate by 10 as well. This will be substantially worse than flu for old people, but bad flu season-ish for everyone else. Say it's 0.2%, I'll be exactly 2X as worried about death as I was last flu season (which was 0% worried, so still not worried).

[tater]

This is interesting.

So their current ascertainment rate is 0.004%

Since most people who die do so within 2-3 weeks, we can look at the deaths in that time frame and divide by 117,000. Unless Ohio has 3900 deaths in a few weeks, their mortality is not 3.4% (WHO). If they have ~400 deaths a few weeks from now? Then they are 0.3% territory. This total case estimate is exactly what seems to be missing from everywhere.

[sh1pman]

10 minutes ago, tater said:

which was 0% worried, so still not worried

0% times anything is 0%, so even if it had a true mortality of 10%, you wouldn’t be worried, right?

[tater]

13 minutes ago, sh1pman said:

0% times anything is 0%, so even if it had a true mortality of 10%, you wouldn’t be worried, right?

Once it's a macroscopic risk, I suppose my response would be non-linear, you're right of course.

I look around at Costco and ask, "Am I among the least healthy people 10 people of my age bracket out of 1000?" (1% risk) Yeah, no, I'm not on the least healthy 10 including fat kids in there. I'm not in the least healthy 10%.

The push on the CFRs is what bugs me most. Everyone serious knows the true number of cases if far higher. I actually assume I'm getting it. My wife's on the front lines, half our friends are docs, or one of them is for couples. Everyone will be more concerned because it has a name, but the same thing last year would have been described as "I have the crud that's going around."

[sevenperforce]

4 minutes ago, tater said:

Since most people who die do so within 2-3 weeks, we can look at the deaths in that time frame and divide by 117,000. Unless Ohio has 3900 deaths in a few weeks, their mortality is not 3.4% (WHO). If they have ~400 deaths a few weeks from now? Then they are 0.3% territory. This total case estimate is exactly what seems to be missing from everywhere.

These numbers are coming out so fast that I think we're treating deaths and infections as though they are linear when we should not. Take Italy (which, while more urban than the US, had an initial posture very similar to the US's). Right now they are reporting 827 deaths off 12,500 infections, which is 6.6%, but we do not know how many of those 12,500 infections are going to go on to die in the next few weeks, or how many total infections they'll have by the time the death toll ticks up. Pretty soon we will be running into appreciable fractions of the entire population.

Death rate is also extremely hard to pin down because it depends heavily on whether the hospitalization rate exceeds triage carrying capacity.

If we cannot get total case estimates, it is fairly straightforward to compare based on medical visits. We know, for example, that the seasonal flu's average death rate of 0.1% goes up to 0.2% among the population who require medical treatment for it. Conservatively, the numbers out of Wuhan suggest 2.3% of those who require treatment for coronavirus will die -- 11.5x more lethal.

During a mild seasonal influenza outbreak with ten million total infections, you end up with 13,000 fatalities. Based just on this, it's easy to estimate the number of U.S. deaths based on how much of the U.S. population becomes infected:

• 10%: 489,000 deaths
• 20%: 978,000 deaths
• 30%: 1,467,000 deaths
• 40%: 1,957,000 deaths
• 50%: 2,446,000 deaths
• 60%: 2,934,984 deaths
• 70%: 3,424,000 deaths
• 80%: 2,913,000 deaths

And that's if COVID-19 is no more likely to put you in the hospital than the seasonal flu.

COVID-19 will likely infect between 40% and 70% of the U.S. population.

[tater]

2 hours ago, sevenperforce said:

During a mild seasonal influenza outbreak with ten million total infections, you end up with 13,000 fatalities. Based just on this, it's easy to estimate the number of U.S. deaths based on how much of the U.S. population becomes infected:

Last year's flu season in the US had an estimated 35,520,883 cases, of which 16,520,350 were estimated to have sought medical care of some kind, about 490,561 were hospitalized, and 34,157 are estimated to have died.

The hospitalization rate is therefore estimated to have been 1.38%. Mortality 0.1%.

The current confirmed case stats of COVID19 are basically at best probably between the hospitalization and medical care numbers for flu for places with loads of testing (sick enough to go to ER vs scheduled appt). That's between 1.38% and 46.5%. The more people get tested randomly, the better the numbers will be. If you only test people sick enough to seek care/testing, the bulk of people with "a cold" never get counted. The medical care % would result in the mortality rates being below half what the CFR suggests. The Read paper suggested the Chinese stats from January only tested 5% of the infected population. I think the rate from SK is mostly measuring testing. They did loads of testing, and increased the denominator of mild cases. Their number is ~0.6%.

For 33,000,000 cases, that's 198,000 deaths.

 

Edited March 12, 2020 by tater

[sevenperforce]

1 hour ago, tater said:

The hospitalization rate is therefore estimated to have been 1.38%. Mortality 0.1%.

Mortality among those who sought medical care is 34K/16.5M or just over 0.2%, which is what I said.

1 hour ago, tater said:

I think the rate from SK is mostly measuring testing. They did loads of testing, and increased the denominator of mild cases. Their number is ~0.6%.

SK has socialized medicine and took proactive measures to prevent spread. The U.S. has neither.

[Cavscout74]

2 hours ago, sevenperforce said:

COVID-19 will likely infect between 40% and 70% of the U.S. population

The last numbers I looked at showed the starting area of Covid-19 (Wuhan, China) had ~67000 cases.  Despite this running basically unchecked there for at least a month, that works out to just ~0.6% of the population of Wuhan (~11 million) being infected.   Actual numbers are certainly higher since not everyone getting it is getting sick enough to seek medical attention, but I doubt it's higher by more than 10x.

I don't think this needs to be ignored - that would be flat out dumb - but the level of hysteria I see is way out of line with the severity of the outbreak.  Several years ago, the largest ebola outbreak ever occurred, but here in the US people were called racist and any other insult for suggesting travel restrictions to/from areas of the outbreak.  Despite the nearly 50% mortality rate of that disease.

Note: I can't find where I got that 67000 cases number now, but it is what I remember seeing a day or two ago.  If someone has better numbers, feel free to correct me.

[tater]

1 minute ago, sevenperforce said:

SK has socialized medicine and took proactive measures to prevent spread. The U.S. has neither.

~half of US Medicine is socialized.

Seeking medical care isn't useful, patients are morons for the most part. People show up all the time for flu, then likely complain the doc doesn't throw antibiotics at them. For a virus. Hospital admits are the useful metric here. ~7% of flu related admits died last year in the US.

So a good comparison is what % of COVID19 hospital admissions die, because they are at least sorted by how severe they are. Of course these figures might be best at some point before they get overfilled (care quality decreases), but not right at the beginning (because the first people who show up already have bad disease, likely). Still won't give the total fatality rate (deaths/actual cases detected or not).

It's important to note that even if the fatality rate is .34 instead of 3.4, that's really bad. Also, that's an average over the 90%+ of cases that are mild, and much higher rates for the elderly. So maybe the mortality for 80 YO+ is closer to 5%—50X worse then average flu (maybe 5X flu for 80 YOs, though).

[sh1pman]

https://www.unz.com/akarlin/corona-cost/

Corona & the Cost of Doing Nothing

Permanent 2.5 Year Drop in US Life Expectancy - It's Just Like the Flu, Brah!

Quote

Here is a summary of the results:

• US life expectancy at birth was 78.86 years in 2017 (via mortality.org). There is a minor discrepancy with the official CDC figure of 78.6 years.
• Modeling a typical flu epidemic “on top” of that (so, in practice, a ~doubling of the flu season severity) would reduce US life expectancy to 78.63 years, translating to a reduction of ~0.25 years (three months).
• Modeling a 10% COVID-19 infection scenario with Riou et al. (2020) age-specific mortality rates – the sort of numbers we may expect should it become endemic – reduces US life expectancy to 76.15 years , translating to a reduction of ~2.5 years.
• Modeling a 10 70% COVID-19 infection scenario with Riou et al. (2020) age-specific mortality rates reduces US life expectancy to 66.79 years, translating to a reduction of a cool ~12 years.

 

[tater]

1 hour ago, sh1pman said:

Permanent 2.5 Year Drop in US Life Expectancy - It's Just Like the Flu, Brah!

Except the death rate by age is the CFR, which is likely off by an order of magnitude.

That's my constant complaint in this thread. If the total mortality is much past 1 in retrospect for the first COVID19 season I will be surprised.  That's still 10X flu, BTW. Bad. Multiple things can be true at once. This can be really bad, but the death rate doesn't need to be 3% to be bad.

We were talking about healthcare rationing over dinner tonight. My wife says this happens literally every day. Many outside the US probably imagine this is about cost/insurance. Yeah, no. My wife doesn't know or care if people can pay, or do pay, she treats everyone the same way, they write off charges on unpaid whatever. Doesn't matter, you show up at the ER, you get fixed (if possible). Period. Anyway, she is often stuck telling families, "Look, I can do this surgery, but it's not gonna mean grandpa is going to be back to normal, he's going to die soon anyway of his other medical problems, and he's 90." Many times the family demands care. Useless care. In this COVID19 situation, if you asked my dad, or my father in law (84/89 respectively) if they wanted an ICU spot that would take it from a 50 year old, or a 30 YO mom, they'd walk out AMA before they let their lives be saved at the expense of someone that much younger (haven't asked either, but I'm 100% certain of this because I'd do the same if it was me or literally any child). Anyway, I suspect that this will happen with ICUs all over, yeah, it's triage, but they should take younger people over much older people (and they will, and would before COVID19).

[mikegarrison]

8 hours ago, StrandedonEarth said:

The NBA and NHL have now suspended their seasons. So much for going to a game next week....

NCAA has cancelled all winter and spring championships, effective immediately.

[scottadges]

Disney World (Orlando, USA) and Disneyland (Los Angeles, USA) are both now planned to be closed starting Monday, March 16th.

(Which sucks for my family... we were going to Disney World next week on the 18th... oh well, better safe than sorry)