PlagueWatch 2020-23 - Coronavirus

[budwom]

meanwhile, it sounds like Russia may have spiked the Covid-19 football a bit prematurely.

[Dr. Rosenrosen]

I find Gov Cuomo’s daily press conference to be factual and reassuring in spite of the challenge they’re facing daily. Helps that he is transparent about challenges and concerns but also what they are doing to address the most urgent issues. Worth checking out, in my opinion. Political affiliations 100% to the side... he is doing what a leader should do in crisis time.

[rsvman]

Appreciate the detailed response, but it really doesn't explain if this is capability, logistics, or will. Do you know the answer to that?

At peak monthly production, Ford rolled out 428 B-24 Liberators from one plant. (Edit: Another source says they put out about 1 per hour. Greatest Generation indeed.)

Probably a combination of a lot of things. Not enough stockpile of supplies, as people apparently never anticipated that demand would spike up like this. Everybody wants the supplies that are being made, and not everybody can get what they want.

I'm a doctor, so I don't really know a lot about supply chain economics and so forth.

[swood1000]

That is not at all what BostonDevil said. He or she made no statements about the validity of the 1 in 90, nor any definitive statements about the invalidity of the 0.06%. Just that it is convenient that the authors ignored it…

He said “Also, Vo had experienced 2 deaths already. 2 deaths in 90 cases. The article doesn't mention that. That's ~2% mortality rate. Jus' sayin.” And your position is that his purpose in pointing out a 2% mortality rate in Vò (later reduced to 1%) was not to contrast it with the 0.06%, nor to imply that the authors intended to hide this? Come now.

Although, given that the fatality rate is already over 0.01% of the total population in Italy, I would suggest that their 0.06% case fatality rate is probably an underestimate.

The population of Italy is just over 60,000,000. 0.06% of that number is 36,000. The deaths in Italy, as of March 26, are 7,503.

Which is precisely why their extrapolation is flawed. Because (a) there is no evidence that the 3300 sample of Vo is representative of the entirety of Italy, or even the entirety of Padua

I will grant you that the 3300 sample probably would not qualify as a representative sample for all of Italy. They used Vò because it was a large group, all of whom had been tested. So it’s an “unscientific” sample but that was clear to the reader. Why the umbrage and outrage? So what? It’s a back-of-the-envelope calculation not pretending to be something other. It shows an infection rate different from the official rate by a factor of more than 100. It suggests that there’s something seriously wrong with the official rate and calls for a more accurate survey. Anybody who reaches such a conclusion should be condemned? Why?

and (b) there is no evidence that the 90-person sample of infected cases ISN'T representative of the entirety of Padua or the entirety of Italy. Yet the authors have decided to suggest that the 3300 sample IS a representative sample while the 90 is not.

How do they suggest that the 90 is not?

One shouldn't point to a group of 3300 people in which 90 were infected and proclaim therefore that the nationwide case fatality rate of 1% or greater is false.

The fatality rate being given for Italy is roughly 7%.

[TeacherTom]

If this has been answered, I apologize. I’ve been busy creating virtual lessons for the remainder of the school year. Does having had the pneumonia vaccines offer any protection if one has the virus?

[Bob Green]

Today’s cumulative numbers for Virginia are 460 positive results out of 6,189 tests for 7.4%. Hospitalizations are at 65.

The most recent 24 hours saw 69 positive results out of 819 tests for 8.4%.

[rsvman]

If this has been answered, I apologize. I’ve been busy creating virtual lessons for the remainder of the school year. Does having had the pneumonia vaccines offer any protection if one has the virus?

The "pneumonia vaccines" are specific for Streptococcus pneumoniae, so they do not prevent lung infection with any other organism, bacterial or viral.

I'm not hearing that Strep pneumoniae superinfections are common or are the cause of death in the patients with severe COVID19 so far.


Based on those two things, I'd say the pneumonia vaccines likely won't provide any measure of protection at this time (although they will still help to prevent Streptococcus pneumoniae infection). For the record, Strep pneumoniae also causes ear infections, sinus infections, and blood stream infections; the vaccine protects against all of these even though they market it with the rubric "pneumonia vaccine."

[MChambers]

https://www.usatoday.com/story/money...ds/5076705002/

Not sure FEMA will function under these circumstances.

[bundabergdevil]

Lots of places to see the unemployment data visualizations (charts) but NYT has up an article showing just how dramatic this week's unemployment claims were relative to history. It's bonkers.

[CDu]

He said “Also, Vo had experienced 2 deaths already. 2 deaths in 90 cases. The article doesn't mention that. That's ~2% mortality rate. Jus' sayin.” And your position is that his purpose in pointing out a 2% mortality rate in Vò (later reduced to 1%) was not to contrast it with the 0.06%, nor to imply that the authors intended to hide this? Come now.

Having read the totality of his or her posts on the topic, I am quite confident that the gist of that was (1) to illustrate that the authors were indeed hiding that fact, but also (2) just an anecdotal contrast about the fallacy of their analysis. I don't think he or she was stating that the 0.06% was definitely wrong. In fact, I am pretty sure he or she has explicitly said that the 0.06% could be right or wrong, and that the 2 (and later 1) of 90 wasn't intended to be a definitive proof that the mortality rate is at least 1%. Isolating the entire discussion down to that one quote seems intentionally shortsighted.

The population of Italy is just over 60,000,000. 0.06% of that number is 36,000. The deaths in Italy, as of March 26, are 7,503.

The number of deaths currently represent ~.0125% of the total population. For the case fatality rate to be 0.06%, that means that roughly 20.6% of the population would have had to be infected. Considering that Vo, one of the first places to get the virus in Italy, had just a 2.7% infection rate as of March 6, and drastic social distancing measures were implemented I would posit that much less than 20% of the total Italian population has been infected, especially given the distribution of confirmed cases to date (still predominantly in the North of Italy per their country's health dept website).

Also, remember that that's 7503 cases and counting. Most of the cases haven't resolved yet. Deaths inherently lag case confirmation. We haven't yet hit the down side of the death curve there. There will likely still be many more deaths.

I will grant you that the 3300 sample probably would not qualify as a representative sample for all of Italy. They used Vò because it was a large group, all of whom had been tested. So it’s an “unscientific” sample but that was clear to the reader. Why the umbrage and outrage? So what? It’s a back-of-the-envelope calculation not pretending to be something other. It shows an infection rate different from the official rate by a factor of more than 100. It suggests that there’s something seriously wrong with the official rate and calls for a more accurate survey. Anybody who reaches such a conclusion should be condemned? Why?

Because it is bad science that even you didn't realize was bad science until it was pointed out to you (and that you have been arguing is fine science until just now). And because they use that bad science to suggest that the virus isn't that dangerous. Or, at the very least, they present it in a way that will allow people who are already inclined to want to believe the virus isn't that dangerous into being convinced the virus isn't that dangerous. Which will lead to risky behavior, which will lead to more deaths.

If smart people like you are taking that back-of-the-envelope calculation as strong evidence (as you presented it in this thread), imagine what dumb people are going to do with that information?

How do they suggest that the 90 is not?

By not even mentioning the 1 in 90. Not even with the caveat that we don't know if that 1 in 90 is really 1 in 90 or 1 in 9000. That seems an implicit suggestion that they don't think 90 cases is representative.

The fatality rate being given for Italy is roughly 7%.

And? The case fatality rate being given is the case fatality rate they have based on the cases they have confirmed. There really isn't anything more they can do. It's an agnostic measure. Is it an overestimate? Almost certainly. But it's the only factual data we have. Researchers can try to estimate from there, provided they don't half-arse it like these guys did.

And again, this is a whole lot of unnecessary discussion about the case fatality rate, which is failing to see the forest for the trees. The problem with coronavirus isn't simply the case fatality rate (which may or may not be greater than the flu). That is obviously a key part of the story, but perhaps the bigger problem is that we have zero immunity to it, unlike the flu, and that it appears to be very transmittable by asymptomatic people, unlike the flu. So thinking of it like the flu is a really bad idea, regardless of if the case fatality rate is ten times the flu rate, the same as the flu rate, or half of the flu rate. If, say, the disease is 3 times more likely to be transmitted than the flu due to no immunity, even with only half the case fatality rate that would mean a ~50% expected increase in deaths relative to the flu without additional preventative measures implemented.

So if someone is going to make the claim that the death rate is no worse than the flu, they better be doing so with VERY good information. Not this half-arsed back of the envelope calculation. And they better be VERY careful in how they say it, making clear that the danger is as much (or more) in the transmission rate as the case fatality rate. Anything less is reckless and dangerous. This bad-science opinion piece is dangerous to public health.

[swood1000]

Today’s cumulative numbers for Virginia are 460 positive results out of 6,189 tests for 7.4%. Hospitalizations are at 65.

The most recent 24 hours saw 69 positive results out of 819 tests for 8.4%.

I’m not sure what these percentages tell us, beyond how many people felt concerned enough to go get tested and how many of those tested positive. Doubtless there have been many more than 6,189 infected people in Virginia. They just haven’t come in to get tested.

[CDu]

I’m not sure what these percentages tell us, beyond how many people felt concerned enough to go get tested and how many of those tested positive. Doubtless there have been many more than 6,189 infected people in Virginia. They just haven’t come in to get tested.

I agree. Or, have been turned away via triage because of the shortage of tests available. No idea what the true infection rate is.

[cato]

I hear it commonly reported, and it makes sense to me. It goes further than just suicide. Loss of jobs means loss of health care, inability to buy medicine, starvation, malnutrition, increased drug and alcohol use (abuse), increased violent crime, domestic violence, and probably much more than is occurring to me right now.

I doubt there really could be a comprehensive study. Perhaps someone could convince me the reasoning behind it is flawed.

FWIW: I'm not advocating an immediate return to our "pre-covid" lives. We NEED to allow our medical system to catch up and resupply, and to ensure we are taking care of the health care workers on the front lines. On the other hand, we cannot shut down the economy until we have zero reported new cases. There must be a balance struck.

I hear commentators opining about it, and it does make sense that economic impact has bad non-economic follow on effects. Just like closing the schools (which also has economic impacts). What I have not seen is any analysis to back it up, so I was wondering if anyone else had.

I am not advocating for any position.

[cato]

Probably a combination of a lot of things. Not enough stockpile of supplies, as people apparently never anticipated that demand would spike up like this. Everybody wants the supplies that are being made, and not everybody can get what they want.

I'm a doctor, so I don't really know a lot about supply chain economics and so forth.

If I may rephrase for you:

[McCoy]I’m a doctor, Jim. Not a supply chain expert![/McCoy]

[Bostondevil]

The example said that if the prevalence of the infection in Italy is the same as its prevalence in the town of Vò, then comparing the total number of deaths in Italy to the extrapolated number of infected people gives a fatality rate of .06%. It’s just a rough estimate, intended to show that the officially reported numbers differ from this number by over two orders of magnitude. Would you have expected that to be the case? It no doubt causes the public to ask for an explanation. What’s wrong with that? Maybe Vò is two orders of magnitude different from the rest of Italy.

You argue that the authors, out of malign motives, hid the fact that one of the 90 infected people in Vò died (actually, he was the first person in Italy to die and that is why they tested the entire town). You suggest that 1/90 gives a proportion of roughly 1%, not .06%, thus demonstrating that a country-wide fatality rate of .06% must be wrong. But .06% relates to the entire population. Some segments of the population will have a much higher fatality rate and some lower. The decedent was a 78 year old man, and people in that category will have the highest rate. It could be 1% or higher, depending on his pre-existing conditions. Younger people have a much lower rate of death. In the U.S., for example, the death rate from the annual flu of those ages 5 – 14 is 0.3% of the death rate of those aged 65 and up. When we average all these rates together we get an overall rate for the entire country. One couldn’t point to a group of 100 people in which a high-risk person died, and proclaim that therefore a country-wide average of less than 1% is shown to be false.

(Responding to bold) I didn't suggest that in the slightest.

What I suggested was that people who do not understand statistics will not see the flaws in the argument put forth by Professors Bendavid and Battacharya. Did I claim they have malign motives? No, I did not. I pointed out that some actual empirical evidence does not support their hypothesis but I also said, multiple times, that it is possible they are correct. It is too early to tell. I maintain, and it is only my opinion, that convincing people who do not understand statistics that "the real" mortality rate is 0.1% is irresponsible right now, maybe even dangerous. I think we all want to believe that the mortality rate is really 0.1%, less than "the flu", but until we have more evidence, it is nothing more than a belief.

(Responding to bold #2) One couldn't point to a group of 100 people and proclaim ANYTHING about a population wide mortality rate - although one could report a point estimate with confidence intervals. I imagine a mortality rate of 0.1% is within the boundaries of a 95% confidence interval at this point. But, given the data from Italy, I imagine so is 8%.

[bedeviled]

Not only do we not have enough kits, but we are also running low on the swabs necessary to obtain samples for testing...[SNIP]...If the outbreak continues to expand, the focus will be more and more on blocking testing on asymptomatic and marginally symptomatic people, as testing will be aimed at more severely symptomatic and those who are more likely to actually have the disease

Appreciate the detailed response, but it really doesn't explain if this is capability, logistics, or will

Probably a combination of a lot of things. Not enough stockpile of supplies, as people apparently never anticipated that demand would spike up like this. Everybody wants the supplies that are being made, and not everybody can get what they want

Agree w/ rsvman (no surprise), it's a combination. Here are some thoughts:
Certainly supplies - admin members I've heard from say that staff is constantly on the phone trying to identify sources for supplies. As soon as manufacturer or distributor is known, word spreads to multiple systems and supply is gone. Also, apparently, they are experiencing scams just like we do personally - overseas distrubutors making promises, requiring payment upfront, and never delivering. Best supply info appears to be from FEMA and military connections. With limited supplies, decisions have to be made to ensure the most benefit.

Another supply/logistics issue: admin complaining that their hospital has been flooded with ppl requesting testing, with ultimately negative results (unclear if ppl are embellishing symptom history to get tested). The amount of testing is stressing in-hand resources - staff & personal protective equipment are blown through due to the amount of testing. Also, institutions may need to shut down or relocate other services in attempt to satisfy influx of testing and presumed cases while protecting other services, which leads to economic cost/instability issues. It costs a lot to operate many practices & institutions, and even temporary issues can be devastatingly severe. Not all procedures/tests provide the same income to the economic balance.

It is also an issue of will as there are debates about testing philosophy (utility and meaningfulness of screening and test results), especially given the supply & logistics concerns.

[AustinDevil]

Apologies if this has been discussed and I missed it, but reputable media in Italy are now hypothesizing that the real death toll from COVID-19 is much, much higher--and that it is actually tied to virtually the entire population (in the town studied, Bergamo) having caught the virus.

https://www.corriere.it/politica/20_...?refresh_ce-cp

[CDu]

Apologies if this has been discussed and I missed it, but reputable media in Italy are now hypothesizing that the real death toll from COVID-19 is much, much higher--and that it is actually tied to virtually the entire population (in the town studied, Bergamo) having caught the virus.

https://www.corriere.it/politica/20_...?refresh_ce-cp

The counterargument to the "true cases are underreported": the "true death toll may also be underattributed". In the interest of fairness (vis a vis my discussion with swood), there are some real concerns with this analysis as well. Namely, generalizability concerns. But these researchers' approach is essentially the same as rsvman suggested: take the expected all-cause mortality data based on seasonality and population, then compare it with the confirmed dead. If it is wildly out of line without any other underlying cause, there is a good chance that most of the difference is due to the virus.

That said, I don't think we can trust the 6x estimate for a few reasons: (1) it is based off of one month's data in a few municipalities and doesn't account for any natural variability in general mortality risk (i.e., it's assuming that the mean expected number of deaths is absolutely correct and that all deaths above this are due to the virus), and (2) it is assuming the evidence from those 3 municipalities translate to the entirety of Italy. Neither is a safe assumption: the first one is fairly reasonable, but still noisy; the second is not sound.

It's probably reasonable to say that the "true death toll" in those particular municipalities could be 3-15 times higher than what is being reported. But that's probably only safe to say in those particular municipalities. If there are reasons why those municipalities are more susceptible to at-home or otherwise unattributed deaths, then the generalizability goes out the window. For example, if those municipalities have a higher concentration of cases, or a higher concentration of people living in facilities that aren't amenable to getting tested, etc., they may be more likely to have deaths that aren't caught by the system.

As the authors suggest, one would need to track this across all municipalities where there have been official deaths due to COVID reported and compare the total deaths in the municipality's population to what would have been the expected death rate over the same time period to get a better estimate of the "true death toll." That wouldn't give you a case fatality rate, of course, unless you assume everyone has the disease (which would then give you the floor case fatality rate). But it would give a more reasonable figure of the death toll caused by the virus in Italy (and would probably still only be specific to Italy).

[budwom]

https://www.usatoday.com/story/money...ds/5076705002/

Not sure FEMA will function under these circumstances.

The National Cleanliness Factor just shot up by about 150%.

[budwom]

I found this article to be most interesting, and am choosing to believe it's true unless one of our resident experts declares otherwise.


https://www.washingtonpost.com/opini...ting-packages/

The author has some pretty solid credentials, which is why I think it's worth consideration...