Saturday, April 18, 2020

New Data Suggest the Coronavirus Isn’t as Deadly as We Thought

A study finds 50 to 85 times as many infections as known cases—meaning a faA study finds 50 to 85 times as many infections as known cases—meaning a far lower fatality rate lower fatality rate.


Trish Keenan, the singer/ song writer in the British avant-garde band, Broadcast, was touring Australia 9 years ago when caught the flu and died of pneumonia. She was only 42.

Covid-19 may be very contagious but could possibly less lethal than the flu, according to new research. The high numbers of fatalities are due to the high nunber of people getting infected. That's bad, but Dr. Jay Bhattacharya says we may be doing uneccessary harm to our economy. Although the real problem is that our hospitals can't cope with the amount of people becoming seriously ill.

It seems that 50 % of people get no symptoms, and so they are helping to bring about herd immunity because they are out and about infecting people. 

Covid-19 shutdowns have been based on the premise that the disease would kill more than two million Americans absent drastic actions to slow its spread. That model assumed case fatality rates—the share of infected people who die from the disease—of 1% to 3%. The World Health Organization’s estimated case-fatality rate was 3.4%.

Some experts—including in these pages—have questioned this assumption. They argue that known cases are likely only a small portion of the true number of infections, and thus high case-fatality rates could be off by orders of magnitude. We don’t know what portion of infections have gone undetected for a lack of tests, restrictive qualifications to get tested, and a potentially large incidence of mild illness or asymptomatic infection. 


23 comments:

Matt Franko said...

Yeah but Kevin the point is how many hospitalizations per unit time are required even if you dont die? Call that X..

Then you have to compare that to how many patients your system can accommodate per unit time for that type of required care... call that Y...

If X > Y then its probably gotta be on like Donkey Kong...

This is what Greg has been banging the table about here for months...


Matt Franko said...

another one in Mass this week:

https://www.foxnews.com/science/third-blood-samples-massachusetts-study-coronavirus


33% of those tested ...

Kaivey said...

Yes, you're right, of course. The hospitals can't cope with the amount of people becoming very ill. Perhaps I should add that in. My article wasn't endorsement of Dr Jay's views, it was just something to consider.

Kaivey said...

I've added to the article.

Matt Franko said...

10-4

Greg said...

This is still pure speculation on his part. Until there is mass antibody testing to see who has already had it there will be no settling the “how many people have really had it” argument.


One just needs to look at Northern Italy and New York City and ask have you ever in your lifetime seen their hospital systems strained like this? I haven’t and I’m 61. Regardless of what final mortality ends up being measured as this thing has potential to wreck tightly packed areas

The hospital I was at in Ga now has 16 patients on ventilators. They had 24 ICU beds when I left and they were usually near capacity while I was there. It was not unheard of for us to have to use our post anesthesia recovery area as an ICU for a night or two during busy times.
If they had not compressed their usual surgery schedule the last month and had carried on as usual there would have been a lot of very bad outcomes. Many would not be getting what they needed when they needed it.

Matt Franko said...

“ I’m 61.“

I had you at 30-35 years old...

Greg said...

61.“

I had you at 30-35 years old...




I’m not sure that is good or bad.

I have often told people I can’t stay young but I don’t have to grow up

Matt Franko said...

You come across (I guess to me anyway...) as young...

I think probably good no? contemporary... contemporaneous...

Matt Franko said...

I was getting ready to lecture you one of these days about how f-ed up the Clinton years were... lol...

S400 said...

“This is still pure speculation“

Yes, that’s the problem everywhere one looks. There’s always someone who knows best about this situation and use there title whatever that is to push their speculation.

Peter Pan said...

World population: 7.7 billion
Coronavirus deaths: 161,128
Death rate: 0.002%

With most countries at or past peak (new cases per day), death rates per population are 1/10th that of the flu. Are we expecting 90% of deaths to occur on the downward slope of the curve?

Sweden: 10.23 million
Coronavirus deaths: 1511
Death rate: 0.01%

United States: 328.2 million
Coronavirus deaths: 39,015
Death rate: 0.01%

Canada: 37.59 million
Coronvirus deaths: 1470
Death rate: 0.004%

Taiwan: 23.78 million
Coronavirus deaths: 6
Death rate: 0.000025% - winning!

https://www.worldometers.info/coronavirus/country/taiwan/
There are stats and graphs galore at this website, with their sources.

Greg said...

When have we had this level of mitigation strategies used world wide for a flu epidemic? Never in my lifetime. To think that these 4-5 weeks of serious change of social behavior hasn’t made a difference is nuts. Even if some places have been less strictly enforced most everyone on the planet has refrained from gathering in very large groups tightly packed, improved their personal hygiene beyond where it was 6 weeks ago and is much more careful of where and what they touch......... and it’s working.


S
I probably overstated my case some about speculation. Some speculators have some data that others don’t Some are better at doing probability calculations, in rough form, tha others as well. So I don’t mean to lump all speculation together. There are unknowns and some people are better at assessing what’s known and giving a more informed guess. Others are lazy, dont check their sources, and shoot from the hip. Some voices I will immediately dismiss because I have a sense that their choice of an acceptable outcome does not align with mine.

Matt Franko said...

Makes me think the Taiwans were in on it...

Matt Franko said...

Greg do they have like a 'transmission index' or something for these? or 'transmissibility index'....

Quantify it...




Greg said...

I know you’ve seen that quantified Matt, it’s something like 2.4 or 2.6 vs 1.4 for flu. Which is a lot greater than that difference of 1 appears.

Thing is those numbers are a result of the interaction of virus and density of people. Those ratios are an attempt to put each pathogen on a common ground controlling for density but it is quite imperfect.

The 1918 flu pandemic had a higher than usual R in many places like train cars filled with young soldiers. The factors that are used to measure R are things like 1) how soon after exposure do we know we are sick. 2) how soon after exposure are we able to transmit it 3) how long do we stay sick and able to transmit.

As I understand it most flus are lower than covid in all above metrics and the most concerning was that this corona was felt to be novel....so everyone on earth was potential host. Not true of flu viruses

Another thing I’ve seen discussed in some medical circles is that the patients that become hospitalized have such a high viral load that even young health care workers are getting real sick.

Greg said...

So an index of 2.4 means you can give it to 2.4 people and they give it to 2.4 etc. etc. so after 5 ( 2.4 to 5th power) iterations you have 80 infections
Where as 5 iterations of 1.4 you have about 6 infections

Greg said...

Now I think the estimates for flus R value are better supported since we have much more data on flus but I say even if corona only infects half of what is estimated above, 40 is much higher than 6. And the difference gets greater and greater with each iteration.

Matt Franko said...

well that seems like a good approach...

they should just try to quantify each of these as early as possible then based on this number we have to increase mitigation efforts proportionally...

Other thing is we obviously have to work to increase facility reserve capacity for these viral pandemics going forward... even if we never have to use them we have to increase and train...

But then the whole crew of "we're out of money!" morons will have to be disappeared for that to happen ...

Greg said...

They are trying to quantify but you can’t count what you can’t measure and let’s face it , there is still a fear by many politicians and business leaders that the truth about some of this stuff is too scary.


In Ohio, DeWine was very clear from day one that the reason for issuing the stay at home orders ( which he made on consultation with Cleveland Clinic, Ohio St and UC virologists, epidemiologists).was to allow for a build out of capacity. Some was hospitals securing more space within their walls plus finding space outside hospitals to be used if necessary. Whenever he was asked a question out the virus he always said “that’s not my area, I’ll let Dr Acton answer that, I’m simply the one with authority to make decisions about stay at home orders and what businesses are deemed essential. That’s my job” But many medical officials have CEOs squawking in one ear while their professional duties urge something else

Now we have near the capacity but if we can’t keep enough trained people healthy we won’t meet needs either. We are probably going to have to have attitudes similar to ones we have for fire protection. Pay people to sit aroind and be ready and hope we never use them. Now this can be a rotated pool of workers, not having the same folks sitting around in wait but it still requires a significant increase in workforce and attitude towards their usage.

I’d say it’s not just the “out of money” crowd but equally the “ central planning” crowd


Peter Pan said...

Flattening the curve stretches out the downward slope temporally. There's still a minimum number of deaths that will occur over the extended period of time.

Canada vs. US death rate indicates Canadian lockdown is having an effect.
US vs. Sweden not so much, but at least Sweden didn't tank their economy directly.

Joe said...

Statisticians throw down
https://www.mercurynews.com/2020/04/20/feud-over-stanford-coronavirus-study-the-authors-owe-us-all-an-apology/


Joe said...

I'm *definitely* no statisticians (although I'm not bad with combinatorics and odds), just some simple arithmetic.
The test they used at a false positive rate of 2 per of 371 = 0.54%, which sounds pretty good. But..
The sample size was 3330, so you'd expect about 18 false positives. Doesn't sound like a big deal, but if the true prevalence of the disease is 1%, then you'd expect 33 or 34 true positives.. So you'd end up measuring 50 to 52 positives, right around 1/3 of your positive results are actually not positives. Throw in some self-selection bias and you're results could be way off. Of course, the authors know all this and put in various adjustments, but that's all based off various assumptions..

If true prevalence of the deal is below the false positive rate, you end up with more false positives that real positives. HIV tests work this way among non-high risk populations. Even if you test positive with a 99.9% accurate test, you're still more than likely to not actually be positive.