COVID COVID-19 Coronavirus Containment Efforts

[PeroK]

https://www.newsweek.com/perfectly-healthy-girl-dies-covid-two-days-after-test-1589493
A fifteen-year-old girl from Illinois has died from COVID-19 just two days after testing positive for the disease. Her family said their daughter, who excelled in school and athletics, had no known pre-existing conditions and had been perfectly healthy until recently, although she had complained of a headache on Saturday

In this case, it doesn't appear that this girl died directly from COVID. The news item isn't specific on the cause of death, but a severe headache is not a typical COVID symptom.

Moreover: if the virus is now widespread in younger people in the US, then many young people who are hospitalised for any reason will test positive on admission to hospital. You need to be careful to distinguish between young people admitted to hospital because of COVID and those who have COVID in addition to whatever else requires hospitalisation.

Such news items, in my experience, are full of data analysis fallacies.

 

[morrobay]

In this case, it doesn't appear that this girl died directly from COVID. The news item isn't specific on the cause of death, but a severe headache is not a typical COVID symptom.

Unfortunately headaches can be symptoms. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7273035/

 

[Rive]

In this case, it doesn't appear that this girl died directly from COVID. The news item isn't specific on the cause of death, but a severe headache is not a typical COVID symptom.

There are/were reported cases among young about having negligible symptoms with already developing pneumonia

Also, there seems to be a percentage of victims having the first 'real' symptom as sudden death.

This Covid thing is just far too tricky to discuss individual cases without knowing the exact details.

 

[AlexCaledin]

- It might be simply probability of inhaling a droplet with many viruses deep into one's lungs - in such case, the unprepared immune system can hardly help much...

 

[Vanadium 50]

She had to have it for at least 3-5 days before onset of severe symptoms.

According to KTLA, the symptoms started "little more than a week after she turned 15". According to her mother's gofundme page (because when you lose a child, of course the first thing you need to do is set up a gofundme page) she turned 15 two weeks ago. So this is consistent with that. Her parents didn't think it was Covid then, and so didn't have her tested until it was too late.

I will be called cruel and heartless, but think it is a bad idea to formulate public policy based on anecdotes that tug at the heartstrings. (And Google tells me a similar story happened half a year ago with a girl named Sarah Simental)

282 children have died from Covid in the US. In that time, the number of "excess murders" in the same age bracket caused by the lockdown is above 500 and may be as high as 1000. So the "Just think about the children!" argument doesn't actually point in the direction of "Lockdowns forever!".

 

[PeroK]

Also, there seems to be a percentage of victims having the first 'real' symptom as sudden death.

If you have medical evidence that COVID may produce sudden, asymptomatic death, then let's see it.

 

[Rive]

If you have medical evidence that COVID may produce sudden, asymptomatic death, then let's see it.

Thought it's common knowledge by now.
random link (with good references included) by looking for 'covid sudden death'.

 

[PeroK]

Thought it's common knowledge by now.
random link (with good references included) by looking for 'covid sudden death'.

Far from compelling, I would say. The cases seem to be older people who suddenly have a cardiac arrest.

This isn't my area of expertise, so I can't contribute more than say I'm sceptical of the News Week story, which seems to contain no medical evidence that the cause of death was COVID.

And, the data from the UK seems clear enough: most deaths from COVID in the older age groups; most excess deaths in the older age groups. We've no evidence of significant excess deaths in children during the pandemic.

 

[Wrichik Basu]

This editorial published in The Lancet explains why the current Govt. is to be blamed for the second wave of the pandemic in India.

In the last 24 hours, we have had >401k new COVID patients, and >4k deaths in the country.

 

[Rive]

We've no evidence of significant excess deaths in children during the pandemic.

I don't really get where that come from.

If you have medical evidence that COVID may produce sudden, asymptomatic death, then let's see it.

 

[OmCheeto]

Such news items, in my experience, are full of data analysis fallacies.

Speaking of analysis fallacies, the other day I did some of my suspicious maths, and ended up with my state(Oregon) having the worst increase rate of cases in the nation. And there, after being the Australia of America as far as total death rates go... (#5 from the bottom)

Did I do this wrong?

 

[Vanadium 50]

Did I do this wrong?

This is PF, where we ask you to show your work.

I don't see wherre you get 36.2%. If you are comparing 186344/4217737 and 166822/4217737, you are comparing 4.42% with 3.96%. That looks to me more like 11% than 36%.

​

 

[OmCheeto]

This is PF, where we ask you to show you r work.

I don't see wherre you tget 36.2%. If you are comparing 186344/4217737 and 166822/4217737, you are comparing 4.42% with 3.96%. That looks to me more like 11% than 36%.

​

The very bottom of the image shows my work.

 

[Vanadium 50]

The very bottom of the image shows my work.

So that's what that is!

That expression is extremely sensitive to the value on 4/18. If for whatever reason those numbers go in one day early or one day late it will change your 36.2% by 28% (absolute) or 79% of its value.

 

[PeroK]

The very bottom of the image shows my work.

It's a fairly meaningless figure, which I don't think is telling you anything important. Not directly anyway.

That calculation is very sensitive to small variations. Imagine moving $2,000$ cases from the middle fortnight to the last. Overall that's not a big deal, but it would make a huge difference to your calculation.

 

[OmCheeto]

It's a fairly meaningless figure, which I don't think is telling you anything important. Not directly anyway.

That calculation is very sensitive to small variations. Imagine moving $2,000$ cases from the middle fortnight to the last. Overall that's not a big deal, but it would make a huge difference to your calculation.

I think the figure got us from "never being average" to "exceeding the average", for the first time in this pandemic.

This is the "Chubris" I was talking about the other day.
Fortunately, our vaccinations are getting up to snuff, and the death age demographic rates are to the point where they are what I would call 100% noise.

 

[Ygggdrasil]

I don't really get where that come from.

Here's a figure from a CDC publication on excess deaths in the US during the pandemic:

FIGURE 2. Percentage change in the weekly number of deaths in 2020 relative to average in the same weeks during 2015–2019, by age group — United States, 2015–2019 and 2020
https://www.cdc.gov/mmwr/volumes/69/wr/mm6942e2.htm

The data goes only up to Oct 3, 2020, so it misses the major surge in Winter 2020-2021 as well as later data where the newer variants account for most infections.

 

[AlexCaledin]

Even as a deadly second wave of Covid-19 ravages India, doctors are now reporting a rash of cases involving a rare infection - also called the "black fungus" - among recovering and recovered Covid-19 patients.

https://www.bbc.com/news/world-asia-india-57027829

This aggressive infection affects the nose, eye and sometimes the brain.

 

[OmCheeto]

Even as a deadly second wave of Covid-19 ravages India, doctors are now reporting a rash of cases involving a rare infection - also called the "black fungus" - among recovering and recovered Covid-19 patients.https://www.bbc.com/news/world-asia-india-57027829

Wrichik mentioned the fungus a few days ago.

Not COVID-19, but definitely related to it:

https://www.hindustantimes.com/citi...-among-covid-19-patients-101620320873994.html

I was surprised to see that it causes blood clots, and was curious if it might be the cause of the breakthrough deaths in young women.

"One such sign is fungal invasion into the blood vessels which results in the formation of blood clots ..." ref: https://en.wikipedia.org/wiki/Mucormycosis​

 

[Wrichik Basu]

Even as a deadly second wave of Covid-19 ravages India, doctors are now reporting a rash of cases involving a rare infection - also called the "black fungus" - among recovering and recovered Covid-19 patients.

https://www.bbc.com/news/world-asia-india-57027829

This aggressive infection affects the nose, eye and sometimes the brain.

That fungus is deadly. The BBC article mentions surgical removal of the eye; I have read in other newspapers that jaw bone and nose had to be removed in many cases. The BBC article wrote that the intravenous injection that can treat the fungal infection costs about $48; but I have read in several other papers that it costs around $123 (INR 9000), and it has to be taken daily for 21 consecutive days. Diabetic patients are at the most risk.

 

[Rive]

Here's a figure from a CDC publication on excess deaths in the US during the pandemic:

Sorry, the question was not about sources. I was wondering about the reason why the existence of 'sudden deaths' was dismissed with such unrelated statistics.

 

[russ_watters]

There are/were reported cases among young about having negligible symptoms with already developing pneumonia

Also, there seems to be a percentage of victims having the first 'real' symptom as sudden death.

Thought it's common knowledge by now.
random link (with good references included) by looking for 'covid sudden death'.

Your source does not support your claim. Far from being "common knowledge", it looks to me like you're taking two extra steps of logic to reach a conclusion beyond what your sources says:

2. Your source was describing statistical associations between COVID infection rates and out of hospital cardiac arrest/sudden deaths (OHCA/OHSD) only. It makes no mention of whether any of the people who died suddenly were even tested for COVID much less presents data about those rates. It suggest a *possible* causal link between OHCA/OHSD and COVID infection, *or* between OHCA/OHSD and COVID lockdown.

1. As @Vanadium 50 keeps having to point out, there is a difference between dying *of* COVID and dying *with* COVID. A lot of people have gotten COVID over the past year and a lot of people die every year - more last year. Some of those deaths will overlap regardless of if COVID infection is the part of the cause of death or not. This is the same logical problem as having a blood clot after a J&J vaccination vs a blood clot from a J&J vaccination. Or even autism after MMR vaccination vs autism from MMR vaccination.

You can make the same spurious analysis with other causes of death, but the preposterousness of those shouldn't distract you from the logical problem being the same: does COVID infection cause blunt force trauma (car accidents) and penetrative injuries (gunshot wounds) too?

So again:
2. Statistical association does not prove causation.
1. Coincidence (positive test at the same time as the death) isn't even proof of causation.

 

[Rive]

2. Statistical association does not prove causation.

Sorry for the confusion. I did not intended to prove anything. I've just provided a starting point.

 

[russ_watters]

Sorry for the confusion. I did not intended to prove anything. I've just provided a starting point.

That's really not good enough. You made a very strong claim about something very dubious and you do in fact need to prove it (provide evidence for it). Your wording even implies it is beyond proven ("common knowledge"). You may retract it, but you can't just let it stand unsubstantiated.

 

[Vanadium 50]

This is the same logical problem as having a blood clot after a J&J vaccination vs a blood clot from a J&J vaccination.

Technically not. One is "Cum hoc ergo propter hoc" (with this, therefore because of this) and the other is "Post hoc ergo propter hoc" (after this therefore because of this). Learn those fallacies!

I think there actually is more to the J&J blood clot story than post hoc reasoning. All the victims were women under 50, which is not what one expects from random chance. I think the better argument is that the risk is very small compared to the Covid risk. To put it in perspective, vaccinating the entire US probably involves a billion or two miles of driving. That's maybe 20 deaths from traffic accidents compared to, I think, 4 from the vaccine.

 

[russ_watters]

Technically not. One is "Cum hoc ergo propter hoc" (with this, therefore because of this) and the other is "Post hoc ergo propter hoc" (after this therefore because of this). Learn those fallacies!

I can't tell if you're being facetious or not. "After vaccination" = "while vaccinated".

Latin is a dead language.

I think there actually is more to the J&J blood clot story than post hoc reasoning. All the victims were women under 50, which is not what one expects from random chance. I think the better argument is that the risk is very small compared to the Covid risk.

I agree, but when news of the association first came out it was just that, an association. We started running numbers on statistical likelihood that it was a coincidence/statistical anomaly. After some digging, it was found that certain specific types of blood clots are likely causally linked to the vaccine. We may yet find out that COVID infection causes heart attacks in otherwise asymptomatic people.

The line of reasoning starting from a vague statistical association to a possible causal link may end with a yes, a no, or a maybe. In that case it ends with a "yes". It may be a counterpoint, but it is also an illustration of how a proper endpoint of the investigation can be reached. I've seen no evidence so far that the necessary analysis has been done to lead to a conclusion for sudden death in otherwise asymptomatic people.

 

[Ygggdrasil]

I agree, but when news of the association first came out it was just that, an association. We started running numbers on statistical likelihood that it was a coincidence/statistical anomaly. After some digging, it was found that certain specific types of blood clots are likely causally linked to the vaccine. We may yet find out that COVID infection causes heart attacks in otherwise asymptomatic people.

The line of reasoning starting from a vague statistical association to a possible causal link may end with a yes, a no, or a maybe. In that case it ends with a "yes". It may be a counterpoint, but it is also an illustration of how a proper endpoint of the investigation can be reached. I've seen no evidence so far that the necessary analysis has been done to lead to a conclusion for sudden death in otherwise asymptomatic people.

I think this leaves out important sources of evidence that go beyond just statistical association and point to likely mechanisms behind the clotting, such as the observed similarities to heparin-induced thrombocytopenia (HIT) and finding similar anti-PF4 antibodies in those experiencing the vaccine-induced immune thrombotic thrombocytopenia (VITT) as those experiencing HIT (see https://www.nejm.org/doi/full/10.1056/NEJMoa2104840 and https://www.nejm.org/doi/full/10.1056/NEJMoa2104882).

As for the out of hospital cardiac arrest events reported in the link @Rive cited, recall that hospitals saw fewer admissions for conditions like heart attacks and strokes at the height of the pandemic (likely because people were hesitant to go to the hospitals for fear of getting infected and from news reports of the hospitals being overwhelmed), so it follows that there should have been more deaths from conditions like these outside of hospitals.

 

[russ_watters]

I think this leaves out important sources of evidence that go beyond just statistical association and point to likely mechanisms...

Understood; ultimately to prove a causal link you need a known cause (mechanism). The stats can help point to the issue being real (and may point toward the mechanism), but can't themselves prove it.

As for the out of hospital cardiac arrest events reported in the link @Rive cited, recall that hospitals saw fewer admissions for conditions like heart attacks and strokes at the height of the pandemic (likely because people were hesitant to go to the hospitals for fear of getting infected and from news reports of the hospitals being overwhelmed), so it follows that there should have been more deaths from conditions like these outside of hospitals.

Yes, I think that is a pretty obvious/likely explanation.

 

[Vanadium 50]

Latin is a dead language.

Horribile dictu!

An example of cum hoc is "he died after contracting Covid, therefore Covid killed him". An example of post hoc is "after she got the shot she died, therefore the shot killed her". The difference is subtle.

After some digging, it was found that certain specific types of blood clots are likely causally linked to the vaccine.

Is that known? Or do we have a chain of plausibility but nothing quite so direct. When I first heard women under 50, my first thought was birth control. (OK, that was my second thought - my first was "how much under 50"?) That's a known cause of blood clots in women in that demographic, especially if "under 50" means "well under 50". If that's the case, who is to say that vaccines cause blood clots in women on birth control or birth control causes blood clots in women who are vaccinated?

Even if this isn't the situation this time, I think it's a good illustration of the difficulty in assigning causality. Which cigarette was the one that gave someone cancer?

We may yet find out that COVID infection causes heart attacks in otherwise asymptomatic people.

We might.
Or we might find, as @Ygggdrasil suggests it is rooted in people's inability to see a doctor during the lockdown. (I have two friends with heart conditions, and both are having difficulty seeing their doctors)
Or we might find that it is rooted in the constant "we're all going to die!" messaging. (People's estimate of the risk of 45-and-unders is 10x worse than it really is.)
Or we might find that it is rooted in the stress of not being able to support one's family because one is non-essential and unimportant.
Or it might be something else.
Or a combination.

 

[fresh_42]

I (re)moved several posts about fauna. Please stay on topic.