The Unz Review: An Alternative Media Selection
A Collection of Interesting, Important, and Controversial Perspectives Largely Excluded from the American Mainstream Media
 TeasersiSteve Blog
Your Odds ...
🔊 Listen RSS
Email This Page to Someone

 Remember My Information



=>

Bookmark Toggle AllToCAdd to LibraryRemove from Library • BShow CommentNext New CommentNext New ReplyRead More
ReplyAgree/Disagree/Etc. More... This Commenter This Thread Hide Thread Display All Comments
AgreeDisagreeThanksLOLTroll
These buttons register your public Agreement, Disagreement, Thanks, LOL, or Troll with the selected comment. They are ONLY available to recent, frequent commenters who have saved their Name+Email using the 'Remember My Information' checkbox, and may also ONLY be used three times during any eight hour period.
Ignore Commenter Follow Commenter
Search Text Case Sensitive  Exact Words  Include Comments
List of Bookmarks

From a new study (PDF) of several thousand New York patients, “Factors associated with hospitalization and critical illness among 4,103 patients with Covid-19 disease in New York City.” The first tree is for chances of hospitalization. Don’t think of these as actual decision trees used to make decisions, but as post-hoc reconstructions of what happened. (On the other hand, they may have some self-fulfilling prophecy aspects.)

How to read this: 3,282 people were considered for admission to the hospital. 49% were selected, including 87% of those over 65 and 37% of those under 65.

Second, who got critically ill:

 
Hide 44 CommentsLeave a Comment
Commenters to Ignore...to FollowEndorsed Only
Trim Comments?
  1. Sung to the tune of “Germ-Free Adolescent” (X-Ray Spex):

    His phobia is infection
    With Coronavid-19
    He wears masks for protection
    And keeps his meat-hooks clean.

    He’s a germ-free geriatric
    Drinks chloroquine prophylactic
    Disinfects ten times a day
    Rub away
    Rub away
    Rub away
    The Purell way.

  2. Print on those graphics is too small for these eyes.

    • Replies: @Redman
    This is also going to need a lot of interpretation and explanation. Can anyone provide insight on what this is saying?

    Like who got critically ill?
    , @Dr. Doom
    Your browser should have a magnification feature built it.
    , @Pericles
    Right click on the image and open it in full size in a new tab.
    , @donut
    Predict admission .
    , @The Alarmist

    Print on those graphics is too small for these eyes.
     
    It's just as well: If you could read the chart, you might find you are doomed.
  3. So that guy in London was right way back in 2015. We all need to get Beach Body Ready.
    https://www.theguardian.com/media/2015/apr/29/beach-body-ready-ad-faces-formal-inquiry-as-campaign-sparks-outrage

    • Replies: @ThreeCranes
    Wonder what would have happened if in response The Weight Loss Collection had changed the ad to show a 350 lb. butterball woman and asked the same question “Are you beach body ready?”
  4. @Henry's Cat
    Print on those graphics is too small for these eyes.

    This is also going to need a lot of interpretation and explanation. Can anyone provide insight on what this is saying?

    Like who got critically ill?

    • Replies: @Meretricious
    Fat, stupid, poor, sedentary, bad diet.

    Stacey Abrams isn't poor any more so she might have a fighting chance.
    , @The Anti-Gnostic
    Don't be fat or old.
    , @Hypnotoad666
    I think this might have worked better as a table rather than a flow chart. But basically each box in Fig. 2A is the number of people with all the characteristics based on following the "Yes" and "No" divisions in each of the previous white boxes. Each box is labelled with the percentage of that sample who were admitted.

    The white boxes are the "decision trees" points that sort the designated samples. The red and green boxes are certain, more-or-less arbitrary end-point samples that, based on their collection of characteristics, had an admission rate over 50% (Red) or under 50% (Green).

    So, for example, the first white box says that 49% of the original sample of 3282 candidates (who were all presumably presenting with symptoms) were ultimately admitted. However, the next step states that 813 of this group were over 65, and that 87% of this over-65 group were admitted. Being over-65 thus gets you into a "Red Box" category of being highly likely to be admitted.

    The next white box notes that 37% of the 2469 people who are under-65 were admitted. However, 653 of this group were also obese, and these under-65 obese people were admitted 65% of the time. Next, 541 of the 653 under-65 obese people are also over 35. This group of obese people aged 35-65 was admitted 70% of the time and get its own Red Box as the prize.

    And so on.

    You could extract a lot of information from this chart, but you have to do most of the math on your own. It's basically like the set-up to a battery of SAT questions.

    , @Pericles
    Hospitalize ...
    if age > 65
    or obese and age > 35
    or (age less than 20 or greater than 44) and male and (diabetic or not white)
    or (age less than 20 or greater than 44) and female and diabetic

    (As you can see, they should have asked the question of 'diabetic' before that of 'male', which would have yielded a simpler decision tree.)

    I await the complaints regarding why an imaginary racist construct like whiteness is considered, and why non-binaries are not considered. Are they just supposed to DIE??

  5. Hypertension (yes) -> no critical illness ??

  6. Your Odds

    The fatality rate for “normal flu” was widely quoted as 0.1% when the Corona Panic started to take shape. Viewed as odds, this seems kind of alarmingly high. 1 in 1000 to flu? But normally we don’t think about it.

    An increasing number of pro-hysteria voices in the media began lecturing us on how we should cave into the panic, let the Panic assimilate us, and most importantly PANIC NOW (think of the toiler paper! won’t anyone think of the toilet paper!), because the coronavirus fatality rate was at least 3%, maybe 5%, possibly 8%! Fear! Apocalypse! All power to the government and to the Holy Media!

    Then the more actual data we got, the more the experts kept finding a strange thing, namely that this much-trumpeted Apocalypse Virus actually also had a fatality rate probably down around 0.1%, maybe 0.2% or in that vicinity, with substantial room to argue over definitions and probably reducing the magnitude to <0.1% by some counts. One way or another, this was, it turned out, not-particularly-remarkable virus.

    In other words, one's Corona Odds are actually very, very good, and were all along.

    Here the Economist summarizes the latest study to endorse the 0.1% estimate published this week in the US:

    The paper reckons that 7m Americans were infected from March 8th to 14th, and official data show 7,000 deaths three weeks later. The resulting fatality rate is 0.1%, similar to that of flu.

    (The paper is: “Using ILI surveillance to estimate state-specific case detection rates and forecast SARS-CoV-2 spread in the United States.” Authors: Justin D. Silverman, Nathaniel Hupert, Alex D. Washburne.)

    Even this 0.1% estimated fatality rate is arguable, given the deaths-with-vs.-deaths-from problem. What is also increasingly clear is there are many marginal/excess deaths occurring that are corona-negative; these are direct, immediate-term victims of the CoronaPanic itself. And that’s before the recession hits.

    • Troll: Kyle
    • Replies: @Hail

    there are...excess deaths occurring that are corona-negative; these are direct, immediate-term victims of the CoronaPanic itself
     
    On which, it came out this week that the UK government had not even bothered modeling marginal lives lost to the effects of the 'Lockdown':

    [T]he government decided on a policy that could potentially cause hundreds of thousands of unnecessary deaths without weighing those lost life years against the potentially saved life years of the adopted policy
     
    This was confirmed by Matt Hancock, the UK Health Secretary, speaking April 10.

    Re: Your Odds,

    If you are in a position of real or potentially compromised health, what are your odds:

    - To catch and die of the effects of the Virus?
    - To die of the effects of the Panic? (as in, untreated heart attacks [the well-documented huge drop-offs in heart-attack admissions reported in hot-spots] and scheduled surgeries or other treatments that are going undone; people terrified of going to hospitals, a central symptom of CoronaParanoia)

    , @Achmed E. Newman
    Hail, thanks for these 2 comments. There's something I'm going to write about later today on my blog about the insurance situation and Kung Flu mission creep. Again, it comes down to who died of the virus and who died with the virus.

    Let me put it this way: the insurance information I just read (just one company, but a big one) says that co-pays are eliminated for cases of COVID-19. I assume that's from people who have test information that proves they have the virus in them, but that's all.
    , @Michael S
    Funny how all of the actual data shows a CFR of anywhere from 2-20%, and all attempts at random testing show that even in the worst-hit areas fewer than 15% were ever infected and in most areas it's less than 1%, but the "estimates" constantly being put out by "economists" all manage to conclude that the actual data must be off by a factor of 100 or maybe more.

    If the data disagrees with the estimates, the estimates must be right! Millions of Americans were already infected and didn't die! No wait, tens of millions! No wait, hundreds of millions!
    , @moshe
    Good comment.

    Though I agree that the corona panic is, and always has been, irrational, I think that humanity's decision to "take a break" makes some sense and that Corona Panic is simply the easiest way to describe their rational for the new "Quarantinism" religion.

    Like all religions, its causes and outcomes are complex. And as a friend of religions I'm glad to generally give them the benefit of the doubt while still poking them with, "so you really think that God Almighty had a baby with a lady and gave birth to himself so that he could die for our sins, ey?"


    Quarantinism or Coronaism is a complex social phenomenon and worthy of being addressed as such. Though that doesn't mean that we shouldn't necessarily mock the Fundamentalist True Believers who think that this is all really about Corona.

    As time passes, there will be less fundamentalists and history teachers in grade school will still teach the santa klaus version (WWI happened because of Gavrilo Princip and the allies fought WWII to save the Jews) while university professors will explain this phenomenon in a somewhat more rational light (humanity needed a Jubilee year to reasses things, like their relationship with the outdoors, with work, with their smartphones, etc.)

    What uosets me is that when this all becomes obvious, I won't be getting any credit for having seen and described it for what it was at the very beginning. Rather the credit will go to the same people who are so very wrong now as they retrofit their rationals from Saddam helped Bin Laden to WMDs to Operation Iraqi Freedom, etc.

    Whatever good and rational things come from this quarantimism (less car accidents, UBI, etc) will be trumpeted by its original trumpeters (while they poo poo the downsides of The Quarantine) as though they knew and trumpeted those benefits all along, rather than that they were fraidy cats soiling their diapers out of some sort of deep religious fear, who sought penance in Worldwide Quarantinism under the original claim (pre "flatten the curve", then, "time to find a vaccine", then "herd immunity", etc) that Covid 19 will therefore have nowhere to go and simply vanish from the Earth.
    , @Hypnotoad666
    New York City death numbers are getting the most press on the alarmist side. They reported about twice the usual 3,000-ish death toll for March-April 10. They are basically counting every single "excess death" as either confirmed "with" Cov-19 or "presumed" to have Cov-19.

    Three thousand more people died in New York City between March 11 and April 13 than would have been expected during the same time period in an ordinary year, Dr. Oxiris Barbot, the commissioner of the city Health Department, said in an interview. While these so-called excess deaths were not explicitly linked to the virus, they might not have happened had the outbreak not occurred, in part because it overwhelmed the normal health care system. https://www.nytimes.com/2020/04/14/nyregion/new-york-coronavirus-deaths.html
     
    Of course, the medical system has not really been "overwhelmed." Rather, it was the panic that gave everyone the perception that it was overwhelmed. Nevertheless, the officials won't be able to resist the temptation to just include the entire "excess death" count as if it were caused by Cov-19.

    They have just barely begun the process of randomized antibody testing in NYC, and that should shed light on the infection rate. However, the bug has a long incubation period and is super-contagious in the crowded NYC eco-system. It probably ripped through 25% of the city before anyone knew what hit them. Thus, even if the entire excess death count were all Cov-19, it would still probably be consistent with a mere .1% CFR.

    Keep an eye on Sweden also. Critics are saying that their model is dubious because infections and deaths are going up there rapidly. But that is the whole intention for them -- to get it over with in a sudden spike that confers herd immunity as quickly as possible. If they are correct, their numbers will spike shortly and then drop precipitously.

    Super-contagious and not very deadly is exactly the combination that makes an economic shutdown expensive and ineffective. That reality should be starting to penetrate the Hive Mind soon.

    , @Polynikes
    Which is all kind of funny, because that more or less correlates with what Ionnadis extrapolated from the Princess cruise ship data. i think he was at about a 10-20% overall infection rate with about a 0.1%, or less, CFR. Just goes to show you with good extrapolation skills you go a long way with a little data. And with bad skills, but just the faintest whiff of academic credibility, you can lead a country down a dangerous path.
    , @Anonymous

    ... pro-hysteria voices in the media began lecturing us on how we should cave into the panic, let the Panic assimilate us, and most importantly PANIC NOW ...
     
    Quite. They had their Ethiopian errand boy at WHO declare a PANdemIC because they needed the PANIC.

    Ron Unz pointed out earlier that the precise timing of first appearance is consistent with human agency.

    The characteristics of COVID-19 are consistent with it having been created to do maximize the slow-motion coup d'etat we are now witnessing.

    Imagine a James Bond-type bad guy had asked his science wizards and Georgetown-trained political science "nudging" specialists WHAT VIRUS would best support his dastardly plan for a quiet take-over of power in Western countries. WHAT WOULD BE DIFFERENT FROM WHAT WE ARE SEEING?

    COVID-19 characteristics:

    1. Slow-burning epidemic, including a modest number of fatalities and serious cases.

    2. Symptoms similar to regular flu for deniability. The populace would be much more suspicious if the disease produced unusual or novel symptoms as did traditional epidemics like measles, cholera, bubonic plague, etc.

    3. Disease symptoms include unusual attacks on oxygen transportation system (hemoglobin) that are difficult to treat. Ditto for cytokine storms.

    4. Serious outcomes most likely in the elderly and those with existing health problems, much lower risk for those more likely to resort to energetic, organized resistance.

    5. Initial release in China - could be an elaborate ruse to launch a bioweapon developed for purposes of advancing a CIVIL WAR or coup d'etat in the West.


    BTW we seem to hear very little about the different strains of COVID-19 detected in different areas/countries.

  7. @Henry's Cat
    Print on those graphics is too small for these eyes.

    Your browser should have a magnification feature built it.

    • Replies: @Achmed E. Newman
    Let me help a bit more: ctrl and "+" sign together, each hit gives another level of magnification. ctrl and "-" sign bring it down in the same size steps. In this case, the good resolution of the graphics leaves plenty of room for magnification.
    , @Known Fact
    With mine you just say, "Magnification five, Mr. Sulu"
  8. @Hail

    Your Odds
     
    The fatality rate for "normal flu" was widely quoted as 0.1% when the Corona Panic started to take shape. Viewed as odds, this seems kind of alarmingly high. 1 in 1000 to flu? But normally we don't think about it.

    An increasing number of pro-hysteria voices in the media began lecturing us on how we should cave into the panic, let the Panic assimilate us, and most importantly PANIC NOW (think of the toiler paper! won't anyone think of the toilet paper!), because the coronavirus fatality rate was at least 3%, maybe 5%, possibly 8%! Fear! Apocalypse! All power to the government and to the Holy Media!

    Then the more actual data we got, the more the experts kept finding a strange thing, namely that this much-trumpeted Apocalypse Virus actually also had a fatality rate probably down around 0.1%, maybe 0.2% or in that vicinity, with substantial room to argue over definitions and probably reducing the magnitude to <0.1% by some counts. One way or another, this was, it turned out, not-particularly-remarkable virus.

    In other words, one's Corona Odds are actually very, very good, and were all along.

    Here the Economist summarizes the latest study to endorse the 0.1% estimate published this week in the US:


    The paper reckons that 7m Americans were infected from March 8th to 14th, and official data show 7,000 deaths three weeks later. The resulting fatality rate is 0.1%, similar to that of flu.
     
    (The paper is: “Using ILI surveillance to estimate state-specific case detection rates and forecast SARS-CoV-2 spread in the United States.” Authors: Justin D. Silverman, Nathaniel Hupert, Alex D. Washburne.)

    Even this 0.1% estimated fatality rate is arguable, given the deaths-with-vs.-deaths-from problem. What is also increasingly clear is there are many marginal/excess deaths occurring that are corona-negative; these are direct, immediate-term victims of the CoronaPanic itself. And that's before the recession hits.

    there are…excess deaths occurring that are corona-negative; these are direct, immediate-term victims of the CoronaPanic itself

    On which, it came out this week that the UK government had not even bothered modeling marginal lives lost to the effects of the ‘Lockdown’:

    [T]he government decided on a policy that could potentially cause hundreds of thousands of unnecessary deaths without weighing those lost life years against the potentially saved life years of the adopted policy

    This was confirmed by Matt Hancock, the UK Health Secretary, speaking April 10.

    Re: Your Odds,

    If you are in a position of real or potentially compromised health, what are your odds:

    – To catch and die of the effects of the Virus?
    – To die of the effects of the Panic? (as in, untreated heart attacks [the well-documented huge drop-offs in heart-attack admissions reported in hot-spots] and scheduled surgeries or other treatments that are going undone; people terrified of going to hospitals, a central symptom of CoronaParanoia)

  9. My odds? I’ve been down in Florida since January 11 this year taking care of my brother who’s been battling cancer for roughly three years. So, I’m 62. I smoke, I drink, I cheat at cards. In and out of various medical situations, hospitals, doc’s offices and the rest, I haven’t experienced C-19 (so far). Grocers, pharmacies, gas stations, car dealers for a repair, I’ve been traveling throughout without masks or consequence. My region has three deaths from C-19 so far, with a couple of hundred cases total in the Penellas and Pasco counties region, not sure about greater Tampa. The whole thing is a fizzle down here.

    Meanwhile, I gotta get back home to Boston to renew my DL. I ain’t flying, I’m renting a car and driving back up so I don’t have to quarantine for 14 days upon arrival. So, who dreams this shit up? Who’s to enforce it? Who decides?

    So fuck em. I’m going on with my life. I’m Serv-Pro. Like it never even happened. Because it never did. I’m convinced 90% of C-19 deaths were deaths that were gonna happen anyway. The Covid-19 is one more bit of the criminal enterprise. Besides, a quarter of our GNP goes to medical. If they can’t handle it, we pay them too much, too. Research, Uni’s, CDC, NIH, the pharmas. Nurses, doctors, they’re raking in overtime like no one’s business. They aren’t heroes, they’re beneficiaries. If they can’t solve this as a profession, then they’re one more God Damned element of the thievery of profit-based medical-care delivery.

    First responders and medical-delivery elements have the best jobs in the U.S. economy. My ass just bleeds for all of them. That their minions in the media pretend to worship them doesn’t mean WE have to. To me, they’re no more important that the folks that gin up toilet paper, the butchers, the bakers or the candlestick makers. Just one more profit center is all they are.

    • Replies: @Reg Cæsar

    ...I cheat at cards.
     
    a

    Is anybody still playing cards? Texas Don't-Touch-'Em, perhaps?

    Napoleon at St Helena is recommended these days.
    , @North Carolina Resident
    You can renew online.
    Also, most states are extending expiration dates.
  10. @Dr. Doom
    Your browser should have a magnification feature built it.

    Let me help a bit more: ctrl and “+” sign together, each hit gives another level of magnification. ctrl and “-” sign bring it down in the same size steps. In this case, the good resolution of the graphics leaves plenty of room for magnification.

  11. @Redman
    This is also going to need a lot of interpretation and explanation. Can anyone provide insight on what this is saying?

    Like who got critically ill?

    Fat, stupid, poor, sedentary, bad diet.

    Stacey Abrams isn’t poor any more so she might have a fighting chance.

    • Replies: @Louis Renault
    She'll get Trump pills early and complain about the side affects.
  12. @Louis Renault
    So that guy in London was right way back in 2015. We all need to get Beach Body Ready.
    https://www.theguardian.com/media/2015/apr/29/beach-body-ready-ad-faces-formal-inquiry-as-campaign-sparks-outrage

    Wonder what would have happened if in response The Weight Loss Collection had changed the ad to show a 350 lb. butterball woman and asked the same question “Are you beach body ready?”

    • Replies: @Louis Renault
    I'm sure it would generate a lot of publicity. Protein World ought to try that out this year.
  13. @Redman
    This is also going to need a lot of interpretation and explanation. Can anyone provide insight on what this is saying?

    Like who got critically ill?

    Don’t be fat or old.

    • Thanks: Redman
  14. @Hail

    Your Odds
     
    The fatality rate for "normal flu" was widely quoted as 0.1% when the Corona Panic started to take shape. Viewed as odds, this seems kind of alarmingly high. 1 in 1000 to flu? But normally we don't think about it.

    An increasing number of pro-hysteria voices in the media began lecturing us on how we should cave into the panic, let the Panic assimilate us, and most importantly PANIC NOW (think of the toiler paper! won't anyone think of the toilet paper!), because the coronavirus fatality rate was at least 3%, maybe 5%, possibly 8%! Fear! Apocalypse! All power to the government and to the Holy Media!

    Then the more actual data we got, the more the experts kept finding a strange thing, namely that this much-trumpeted Apocalypse Virus actually also had a fatality rate probably down around 0.1%, maybe 0.2% or in that vicinity, with substantial room to argue over definitions and probably reducing the magnitude to <0.1% by some counts. One way or another, this was, it turned out, not-particularly-remarkable virus.

    In other words, one's Corona Odds are actually very, very good, and were all along.

    Here the Economist summarizes the latest study to endorse the 0.1% estimate published this week in the US:


    The paper reckons that 7m Americans were infected from March 8th to 14th, and official data show 7,000 deaths three weeks later. The resulting fatality rate is 0.1%, similar to that of flu.
     
    (The paper is: “Using ILI surveillance to estimate state-specific case detection rates and forecast SARS-CoV-2 spread in the United States.” Authors: Justin D. Silverman, Nathaniel Hupert, Alex D. Washburne.)

    Even this 0.1% estimated fatality rate is arguable, given the deaths-with-vs.-deaths-from problem. What is also increasingly clear is there are many marginal/excess deaths occurring that are corona-negative; these are direct, immediate-term victims of the CoronaPanic itself. And that's before the recession hits.

    Hail, thanks for these 2 comments. There’s something I’m going to write about later today on my blog about the insurance situation and Kung Flu mission creep. Again, it comes down to who died of the virus and who died with the virus.

    Let me put it this way: the insurance information I just read (just one company, but a big one) says that co-pays are eliminated for cases of COVID-19. I assume that’s from people who have test information that proves they have the virus in them, but that’s all.

  15. “On which, it came out this week that the UK government had not even bothered modeling marginal lives lost to the effects of the ‘Lockdown’:”

    Fire the lot of the idiots. The UK possesses significant amounts of expertise in complex systems science, simulation, modeling, etc, but apparently the government never bothered to take advantage of it.

    Any basic text on policy analysis would have been sufficient.

    • Replies: @Anonymous

    “On which, it came out this week that the UK government had not even bothered modeling marginal lives lost to the effects of the ‘Lockdown’:”
     
    You are assuming that the objective of the UK government is to preserve "marginal lives" yet their actions bespeak callous disregard.

    Perhaps they have other objectives they don't care to talk about.

  16. @Hail

    Your Odds
     
    The fatality rate for "normal flu" was widely quoted as 0.1% when the Corona Panic started to take shape. Viewed as odds, this seems kind of alarmingly high. 1 in 1000 to flu? But normally we don't think about it.

    An increasing number of pro-hysteria voices in the media began lecturing us on how we should cave into the panic, let the Panic assimilate us, and most importantly PANIC NOW (think of the toiler paper! won't anyone think of the toilet paper!), because the coronavirus fatality rate was at least 3%, maybe 5%, possibly 8%! Fear! Apocalypse! All power to the government and to the Holy Media!

    Then the more actual data we got, the more the experts kept finding a strange thing, namely that this much-trumpeted Apocalypse Virus actually also had a fatality rate probably down around 0.1%, maybe 0.2% or in that vicinity, with substantial room to argue over definitions and probably reducing the magnitude to <0.1% by some counts. One way or another, this was, it turned out, not-particularly-remarkable virus.

    In other words, one's Corona Odds are actually very, very good, and were all along.

    Here the Economist summarizes the latest study to endorse the 0.1% estimate published this week in the US:


    The paper reckons that 7m Americans were infected from March 8th to 14th, and official data show 7,000 deaths three weeks later. The resulting fatality rate is 0.1%, similar to that of flu.
     
    (The paper is: “Using ILI surveillance to estimate state-specific case detection rates and forecast SARS-CoV-2 spread in the United States.” Authors: Justin D. Silverman, Nathaniel Hupert, Alex D. Washburne.)

    Even this 0.1% estimated fatality rate is arguable, given the deaths-with-vs.-deaths-from problem. What is also increasingly clear is there are many marginal/excess deaths occurring that are corona-negative; these are direct, immediate-term victims of the CoronaPanic itself. And that's before the recession hits.

    Funny how all of the actual data shows a CFR of anywhere from 2-20%, and all attempts at random testing show that even in the worst-hit areas fewer than 15% were ever infected and in most areas it’s less than 1%, but the “estimates” constantly being put out by “economists” all manage to conclude that the actual data must be off by a factor of 100 or maybe more.

    If the data disagrees with the estimates, the estimates must be right! Millions of Americans were already infected and didn’t die! No wait, tens of millions! No wait, hundreds of millions!

    • Replies: @Hail

    Millions of Americans were already infected and didn’t die!
     
    That does indeed appear to be correct:

    New antibody data from the Italian community of Robbia in Lombardy shows that about ten times more people had the corona virus than originally thought, as they developed no or only mild symptoms. The actual immunization rate is 22%.
     
    If Robbia's rate is the same as Lombardy's as a whole, granted a big assumption, that is 2.2 million present-positives plus past-positives, the latter being those whose immune systems easily (in some cases after mild symptoms) batted away the virus at some point in the past few months and had no idea.

    Of the perhaps 15,000 Lombardy coronavirus-positive deaths that will be recorded by the time the epidemic ends (currently 11,600 with numbers in steady decline), there remains a significant problem of "deaths from vs. deaths with." Between late February and late April, 4,500 people in Lombardy are statistically expected to die, the natural death rate absent a flu spike; with a flu-spike, it could be 6,000 (as observed several times in the 2010s, without causing a mass-shutdown-style Panic).

    What we know is a very large portion of the deaths are nursing home patients, reportedly about half. We know that Italy counts every corona-positive death as a coronavirus death, without distinction.

    We also know that during the CoronaPanic, nursing home staff abandoned their posts to a large degree and there were unnecessary deaths. To some extent nursing home patients are going to be among the highest mortality group in any society anyway, and also the first group at risk in a Panic. A report was that a lot of Ukrainian staff at these nursing homes and Ukraine recalled all their citizens when the Panic began. But others, including Italians, also began a very high degree of absenteeism.

    The nursing home deaths are, therefore, definitely not all unambiguous coronavirus-attributable deaths but some deep mix of 'virus,' 'Panic,' and 'natural deaths,' with some of the latter pushed ahead some months before they'd otherwise maybe have gone. The other deaths have to at least some degree a similar problem.

  17. @Hail

    Your Odds
     
    The fatality rate for "normal flu" was widely quoted as 0.1% when the Corona Panic started to take shape. Viewed as odds, this seems kind of alarmingly high. 1 in 1000 to flu? But normally we don't think about it.

    An increasing number of pro-hysteria voices in the media began lecturing us on how we should cave into the panic, let the Panic assimilate us, and most importantly PANIC NOW (think of the toiler paper! won't anyone think of the toilet paper!), because the coronavirus fatality rate was at least 3%, maybe 5%, possibly 8%! Fear! Apocalypse! All power to the government and to the Holy Media!

    Then the more actual data we got, the more the experts kept finding a strange thing, namely that this much-trumpeted Apocalypse Virus actually also had a fatality rate probably down around 0.1%, maybe 0.2% or in that vicinity, with substantial room to argue over definitions and probably reducing the magnitude to <0.1% by some counts. One way or another, this was, it turned out, not-particularly-remarkable virus.

    In other words, one's Corona Odds are actually very, very good, and were all along.

    Here the Economist summarizes the latest study to endorse the 0.1% estimate published this week in the US:


    The paper reckons that 7m Americans were infected from March 8th to 14th, and official data show 7,000 deaths three weeks later. The resulting fatality rate is 0.1%, similar to that of flu.
     
    (The paper is: “Using ILI surveillance to estimate state-specific case detection rates and forecast SARS-CoV-2 spread in the United States.” Authors: Justin D. Silverman, Nathaniel Hupert, Alex D. Washburne.)

    Even this 0.1% estimated fatality rate is arguable, given the deaths-with-vs.-deaths-from problem. What is also increasingly clear is there are many marginal/excess deaths occurring that are corona-negative; these are direct, immediate-term victims of the CoronaPanic itself. And that's before the recession hits.

    Good comment.

    Though I agree that the corona panic is, and always has been, irrational, I think that humanity’s decision to “take a break” makes some sense and that Corona Panic is simply the easiest way to describe their rational for the new “Quarantinism” religion.

    Like all religions, its causes and outcomes are complex. And as a friend of religions I’m glad to generally give them the benefit of the doubt while still poking them with, “so you really think that God Almighty had a baby with a lady and gave birth to himself so that he could die for our sins, ey?”

    Quarantinism or Coronaism is a complex social phenomenon and worthy of being addressed as such. Though that doesn’t mean that we shouldn’t necessarily mock the Fundamentalist True Believers who think that this is all really about Corona.

    As time passes, there will be less fundamentalists and history teachers in grade school will still teach the santa klaus version (WWI happened because of Gavrilo Princip and the allies fought WWII to save the Jews) while university professors will explain this phenomenon in a somewhat more rational light (humanity needed a Jubilee year to reasses things, like their relationship with the outdoors, with work, with their smartphones, etc.)

    What uosets me is that when this all becomes obvious, I won’t be getting any credit for having seen and described it for what it was at the very beginning. Rather the credit will go to the same people who are so very wrong now as they retrofit their rationals from Saddam helped Bin Laden to WMDs to Operation Iraqi Freedom, etc.

    Whatever good and rational things come from this quarantimism (less car accidents, UBI, etc) will be trumpeted by its original trumpeters (while they poo poo the downsides of The Quarantine) as though they knew and trumpeted those benefits all along, rather than that they were fraidy cats soiling their diapers out of some sort of deep religious fear, who sought penance in Worldwide Quarantinism under the original claim (pre “flatten the curve”, then, “time to find a vaccine”, then “herd immunity”, etc) that Covid 19 will therefore have nowhere to go and simply vanish from the Earth.

  18. @Jim Christian
    My odds? I've been down in Florida since January 11 this year taking care of my brother who's been battling cancer for roughly three years. So, I'm 62. I smoke, I drink, I cheat at cards. In and out of various medical situations, hospitals, doc's offices and the rest, I haven't experienced C-19 (so far). Grocers, pharmacies, gas stations, car dealers for a repair, I've been traveling throughout without masks or consequence. My region has three deaths from C-19 so far, with a couple of hundred cases total in the Penellas and Pasco counties region, not sure about greater Tampa. The whole thing is a fizzle down here.

    Meanwhile, I gotta get back home to Boston to renew my DL. I ain't flying, I'm renting a car and driving back up so I don't have to quarantine for 14 days upon arrival. So, who dreams this shit up? Who's to enforce it? Who decides?

    So fuck em. I'm going on with my life. I'm Serv-Pro. Like it never even happened. Because it never did. I'm convinced 90% of C-19 deaths were deaths that were gonna happen anyway. The Covid-19 is one more bit of the criminal enterprise. Besides, a quarter of our GNP goes to medical. If they can't handle it, we pay them too much, too. Research, Uni's, CDC, NIH, the pharmas. Nurses, doctors, they're raking in overtime like no one's business. They aren't heroes, they're beneficiaries. If they can't solve this as a profession, then they're one more God Damned element of the thievery of profit-based medical-care delivery.

    First responders and medical-delivery elements have the best jobs in the U.S. economy. My ass just bleeds for all of them. That their minions in the media pretend to worship them doesn't mean WE have to. To me, they're no more important that the folks that gin up toilet paper, the butchers, the bakers or the candlestick makers. Just one more profit center is all they are.

    …I cheat at cards.

    a

    Is anybody still playing cards? Texas Don’t-Touch-‘Em, perhaps?

    Napoleon at St Helena is recommended these days.

  19. @Jim Christian
    My odds? I've been down in Florida since January 11 this year taking care of my brother who's been battling cancer for roughly three years. So, I'm 62. I smoke, I drink, I cheat at cards. In and out of various medical situations, hospitals, doc's offices and the rest, I haven't experienced C-19 (so far). Grocers, pharmacies, gas stations, car dealers for a repair, I've been traveling throughout without masks or consequence. My region has three deaths from C-19 so far, with a couple of hundred cases total in the Penellas and Pasco counties region, not sure about greater Tampa. The whole thing is a fizzle down here.

    Meanwhile, I gotta get back home to Boston to renew my DL. I ain't flying, I'm renting a car and driving back up so I don't have to quarantine for 14 days upon arrival. So, who dreams this shit up? Who's to enforce it? Who decides?

    So fuck em. I'm going on with my life. I'm Serv-Pro. Like it never even happened. Because it never did. I'm convinced 90% of C-19 deaths were deaths that were gonna happen anyway. The Covid-19 is one more bit of the criminal enterprise. Besides, a quarter of our GNP goes to medical. If they can't handle it, we pay them too much, too. Research, Uni's, CDC, NIH, the pharmas. Nurses, doctors, they're raking in overtime like no one's business. They aren't heroes, they're beneficiaries. If they can't solve this as a profession, then they're one more God Damned element of the thievery of profit-based medical-care delivery.

    First responders and medical-delivery elements have the best jobs in the U.S. economy. My ass just bleeds for all of them. That their minions in the media pretend to worship them doesn't mean WE have to. To me, they're no more important that the folks that gin up toilet paper, the butchers, the bakers or the candlestick makers. Just one more profit center is all they are.

    You can renew online.
    Also, most states are extending expiration dates.

    • Replies: @Jim Christian

    You can renew online. Also, most states are extending expiration dates.
     
    Well, that's supposedly. When I went to renew online, they want the registration number on my birth certificate. I am eligible for online renewal with my last picture and eye test, but I don't have my birth certificate with me. Massachusetts gives drivers with expiry dates in April a 60 day waiver, but haven't granted same to those who come up in May. Hence my rush to get back up.

    Thanks, though, NC..
  20. Wow.

    it’s almost like being old, fat, or diabetic might mean you’re at higher risk of complications from infections!

    Who knew?

  21. Here’s a cool graph I found at ZeroHedge. It goes into what industries/professions are safer to open after the economic shutdown.

    https://www.zerohedge.com/health/these-are-jobs-highest-covid-19-risk

  22. @Redman
    This is also going to need a lot of interpretation and explanation. Can anyone provide insight on what this is saying?

    Like who got critically ill?

    I think this might have worked better as a table rather than a flow chart. But basically each box in Fig. 2A is the number of people with all the characteristics based on following the “Yes” and “No” divisions in each of the previous white boxes. Each box is labelled with the percentage of that sample who were admitted.

    The white boxes are the “decision trees” points that sort the designated samples. The red and green boxes are certain, more-or-less arbitrary end-point samples that, based on their collection of characteristics, had an admission rate over 50% (Red) or under 50% (Green).

    So, for example, the first white box says that 49% of the original sample of 3282 candidates (who were all presumably presenting with symptoms) were ultimately admitted. However, the next step states that 813 of this group were over 65, and that 87% of this over-65 group were admitted. Being over-65 thus gets you into a “Red Box” category of being highly likely to be admitted.

    The next white box notes that 37% of the 2469 people who are under-65 were admitted. However, 653 of this group were also obese, and these under-65 obese people were admitted 65% of the time. Next, 541 of the 653 under-65 obese people are also over 35. This group of obese people aged 35-65 was admitted 70% of the time and get its own Red Box as the prize.

    And so on.

    You could extract a lot of information from this chart, but you have to do most of the math on your own. It’s basically like the set-up to a battery of SAT questions.

    • Thanks: Redman
  23. @Dr. Doom
    Your browser should have a magnification feature built it.

    With mine you just say, “Magnification five, Mr. Sulu”

  24. I’m in good shape but at 65 who knows? To pilfer from Vixen’s Love is a Killer,

    I got a target on my back
    for a COVID dressed in black

    • Replies: @Joe Stalin
    https://www.youtube.com/watch?v=LeBxtimkCyQ
  25. @Known Fact
    I'm in good shape but at 65 who knows? To pilfer from Vixen's Love is a Killer,

    I got a target on my back
    for a COVID dressed in black

  26. @ThreeCranes
    Wonder what would have happened if in response The Weight Loss Collection had changed the ad to show a 350 lb. butterball woman and asked the same question “Are you beach body ready?”

    I’m sure it would generate a lot of publicity. Protein World ought to try that out this year.

  27. @Meretricious
    Fat, stupid, poor, sedentary, bad diet.

    Stacey Abrams isn't poor any more so she might have a fighting chance.

    She’ll get Trump pills early and complain about the side affects.

  28. @Hail

    Your Odds
     
    The fatality rate for "normal flu" was widely quoted as 0.1% when the Corona Panic started to take shape. Viewed as odds, this seems kind of alarmingly high. 1 in 1000 to flu? But normally we don't think about it.

    An increasing number of pro-hysteria voices in the media began lecturing us on how we should cave into the panic, let the Panic assimilate us, and most importantly PANIC NOW (think of the toiler paper! won't anyone think of the toilet paper!), because the coronavirus fatality rate was at least 3%, maybe 5%, possibly 8%! Fear! Apocalypse! All power to the government and to the Holy Media!

    Then the more actual data we got, the more the experts kept finding a strange thing, namely that this much-trumpeted Apocalypse Virus actually also had a fatality rate probably down around 0.1%, maybe 0.2% or in that vicinity, with substantial room to argue over definitions and probably reducing the magnitude to <0.1% by some counts. One way or another, this was, it turned out, not-particularly-remarkable virus.

    In other words, one's Corona Odds are actually very, very good, and were all along.

    Here the Economist summarizes the latest study to endorse the 0.1% estimate published this week in the US:


    The paper reckons that 7m Americans were infected from March 8th to 14th, and official data show 7,000 deaths three weeks later. The resulting fatality rate is 0.1%, similar to that of flu.
     
    (The paper is: “Using ILI surveillance to estimate state-specific case detection rates and forecast SARS-CoV-2 spread in the United States.” Authors: Justin D. Silverman, Nathaniel Hupert, Alex D. Washburne.)

    Even this 0.1% estimated fatality rate is arguable, given the deaths-with-vs.-deaths-from problem. What is also increasingly clear is there are many marginal/excess deaths occurring that are corona-negative; these are direct, immediate-term victims of the CoronaPanic itself. And that's before the recession hits.

    New York City death numbers are getting the most press on the alarmist side. They reported about twice the usual 3,000-ish death toll for March-April 10. They are basically counting every single “excess death” as either confirmed “with” Cov-19 or “presumed” to have Cov-19.

    Three thousand more people died in New York City between March 11 and April 13 than would have been expected during the same time period in an ordinary year, Dr. Oxiris Barbot, the commissioner of the city Health Department, said in an interview. While these so-called excess deaths were not explicitly linked to the virus, they might not have happened had the outbreak not occurred, in part because it overwhelmed the normal health care system. https://www.nytimes.com/2020/04/14/nyregion/new-york-coronavirus-deaths.html

    Of course, the medical system has not really been “overwhelmed.” Rather, it was the panic that gave everyone the perception that it was overwhelmed. Nevertheless, the officials won’t be able to resist the temptation to just include the entire “excess death” count as if it were caused by Cov-19.

    They have just barely begun the process of randomized antibody testing in NYC, and that should shed light on the infection rate. However, the bug has a long incubation period and is super-contagious in the crowded NYC eco-system. It probably ripped through 25% of the city before anyone knew what hit them. Thus, even if the entire excess death count were all Cov-19, it would still probably be consistent with a mere .1% CFR.

    Keep an eye on Sweden also. Critics are saying that their model is dubious because infections and deaths are going up there rapidly. But that is the whole intention for them — to get it over with in a sudden spike that confers herd immunity as quickly as possible. If they are correct, their numbers will spike shortly and then drop precipitously.

    Super-contagious and not very deadly is exactly the combination that makes an economic shutdown expensive and ineffective. That reality should be starting to penetrate the Hive Mind soon.

    • Agree: Hail
    • Replies: @Meretricious
    Agreed, but that herd immunity thing is still speculative. We'll know before too long.
  29. @Hail

    Your Odds
     
    The fatality rate for "normal flu" was widely quoted as 0.1% when the Corona Panic started to take shape. Viewed as odds, this seems kind of alarmingly high. 1 in 1000 to flu? But normally we don't think about it.

    An increasing number of pro-hysteria voices in the media began lecturing us on how we should cave into the panic, let the Panic assimilate us, and most importantly PANIC NOW (think of the toiler paper! won't anyone think of the toilet paper!), because the coronavirus fatality rate was at least 3%, maybe 5%, possibly 8%! Fear! Apocalypse! All power to the government and to the Holy Media!

    Then the more actual data we got, the more the experts kept finding a strange thing, namely that this much-trumpeted Apocalypse Virus actually also had a fatality rate probably down around 0.1%, maybe 0.2% or in that vicinity, with substantial room to argue over definitions and probably reducing the magnitude to <0.1% by some counts. One way or another, this was, it turned out, not-particularly-remarkable virus.

    In other words, one's Corona Odds are actually very, very good, and were all along.

    Here the Economist summarizes the latest study to endorse the 0.1% estimate published this week in the US:


    The paper reckons that 7m Americans were infected from March 8th to 14th, and official data show 7,000 deaths three weeks later. The resulting fatality rate is 0.1%, similar to that of flu.
     
    (The paper is: “Using ILI surveillance to estimate state-specific case detection rates and forecast SARS-CoV-2 spread in the United States.” Authors: Justin D. Silverman, Nathaniel Hupert, Alex D. Washburne.)

    Even this 0.1% estimated fatality rate is arguable, given the deaths-with-vs.-deaths-from problem. What is also increasingly clear is there are many marginal/excess deaths occurring that are corona-negative; these are direct, immediate-term victims of the CoronaPanic itself. And that's before the recession hits.

    Which is all kind of funny, because that more or less correlates with what Ionnadis extrapolated from the Princess cruise ship data. i think he was at about a 10-20% overall infection rate with about a 0.1%, or less, CFR. Just goes to show you with good extrapolation skills you go a long way with a little data. And with bad skills, but just the faintest whiff of academic credibility, you can lead a country down a dangerous path.

  30. Hmmm… no cell in the decision tree regarding insurance? KEK

    Small-c cynicism aside, one very interesting thing about that PDF is the tables.
     • Table 2 purports to decompose the ‘admitted’ (N = 1999) column of Table 1, but doesn’t decompose the 417 admitted who are not critically ill (so far);
     • ”Temp > 38°C” – percentages in Table 2 are not proportions of the relevant Table2 ‘population’

    The image below corrects both those things, and consolidates relevant rows of Tables 1 and 2.

    Petrelli et al – Composite Table 1 and 2 (Table 2 corrected where required)

    What stands out concerns the key co-morbidities that everyone (including me) has been considering an important determinant of mortality risk. It’s kind-of still true, but it’s at more than one remove; on their own they’re useful identifiers of the probability of requiring hospitalisation, but they don’t help identify which hospitalised people likely to go critical.

    There’s a very strong requirement for an AND clause – in other words there’s an interaction term (which also means that a generalised linear model is not a sensible choice).

    The ‘AND’ term is AGE (probably still categorical, with the category border in the 70s).

    Also interesting: if you wind up critical, your odds aren’t good – 45% of people considered critical died (and much more than 50% of deaths were people on ventilators); another 38% are still on a ventilator (and therefore highly likely to die).

    Speaking of ventilators: in Figure 1, 445 patients were put on ventilators; 200 of those are no longer on ventilators.

    Of that 200, 162 (81%) are off ventilators as a result of being dead.

    By counting people who were still on respirators in the denominator for respirator cases, they report a lower number for the deadliness of respirators. It’s 81% of all cases where a respirator is used until ‘fixed-ish or dead‘. By contrast the text says

    The largest detailed case series published to date included 1,099 hospitalized patients with laboratory-confirmed Covid-19 infection in China, of whom only 25 (2.3%) [=25/1099] underwent invasive ventilation and 15 (1.4%) [=15/1099] died. By contrast, 28.1% [= 445/1583] of hospitalized patients with definitive outcomes in this case series have so far required invasive ventilation and 18.5% [= 292/1583] have died. – [italics mine]

    Why not mention that 81% of mechanically ventilated patients died? It’s the single biggest correlate to death.

    Seems deliberate.

    .

    There’s also something very weird about the Odds Ratios for the Age variables in Tables 3 and 4; the ones in table 3 bear little relationship to the unconditional observed values (using 19/44 as numeraire). It indicates that the regression almost certainly found a local maximum.

    Age-Group Pr(Admission) and Unconditional Odds Ratios (19-44 = 1)

    (Note: the high admission rate in 0-18 includes 19 newborns; none died)

    Sadly, they haven’t made the data or model available yet: if they had I wouldn’t have to pfaff around.

    The paper says that they used logistic regression – almost certainly a LOGIT model, which models Pr(RHS) and then converts that to a binary outcome (H/¬H) based on some selected threshold. It would be ideal if they used PROBIT, which is better suited to the problem.

    Both are part of R’s glm (generalised linear model) package, but determining which to use requires that on the user understands why probit > logit for categorical dependent variables and a binary dependent variable.

    Since they use VIF to test for collinearity, all bets are off. VIF is known to be no good for categorical variables.

    Plus, with as many variables as they bunged onto the RHS (about 25, which is ridiculous), they’re gonna need more than VIF < 2.5 to be assured of a full-rank regressor matrix.

    Why not just do a test of the regressor matrix? Farrar-Glauber works OK for categorical variables.

    The regressor matrix dimensions are O(10³)×O(2⁴), so they can’t claim it’s too computationally burdensome.

    Simple enough to go “old school” – generate X’X and calculate its determinant directly, or just try to invert it and see if you get a warning. Or get its condition number, or its smallest eigenvalue.

    All of those things are in R’s mctest package; omcdiag from that package will give ‘kitchen sink’ output and a reliable declaration at the end – 0 or 1. Narrow down the culprit with imcdiag.

    You would think that glm would generate warnings when the regression was run – but logit and probit models are weird critters: often they seem to converge but produce nonsense.

    Plus, warnings can easily be overlooked, especially if someone wants to have small log files to wade though… just set options( warn = -1 ), which is the sort of thing a lazy person would do (say, someone lazy enough to fail to type VIF categorical variables into a search box).

    .

    I’m making clear my annoyance at the authors for not providing deaths by age group and respirator cases by age group. (Or better yet: the R file and the de-personalised data).

    This study has a decent sample size, and those two tables would be really useful to confirm several hypotheses (to confirm them properly, not using ‘HelloUdemy‘ statistics – the H1Bindhi version of journomath).

    As it stands, the ORs for admission and progress-to-critical by age cohort result in a model that makes no sense (if you generate some pseudo-data with plausible values). Without the deaths by age group table, the regression results are suspect, even though the strong significance of “Age ≥ 75” conforms to my prejudices.

    This could be a really important publication if the regressor matrix doesn’t have the problems I’ve mentioned above: they really need to drop the newborns from the analysis – with disclosure, of course – because they totally fuck up the hospitalisation regression. Newborns get hospitalised regardless (and none of the newborns got critical) – nobody would criticise the authors if they dropped those 18 data points and started age groups at 15-44 (even though more age-group granularity – 5 year cohorts – would be better and no harder to organise).

    • Replies: @Meretricious
    Are you paid by the word ?

    What do you think the MORE tag is for?
  31. To everyone proclaiming “the whole thing’s a hoax” or “it’s just the flu” I say – put your money where your mouth is.
    Go out and catch COVID-19! Stand out front at a drive-thru testing site and invite everyone to cough on you. Volunteer to mop the floors in the hospital. Hang out in a homeless shelter. Go get it!
    Don’t infect just yourself, either – since you’re an expert and know all about SARS-CoV-2, why not bring it home to your family? Just think – as survivors, you’ll be first-in-line to get one of the upcoming “Immunity Certificates” releasing you from the lockdown that you so decry.
    I’ll be waiting to hear back from you – or not. Until then, you’re talking through your hat.

  32. There need to be autopsies and audits of all “COVID-19 deaths”, which I suspect will never happen. Physician negligence, medical errors, and a complete breakdown of the medical system has occurred. A treatment, and an urgent cover-up is needed, so we can bury/burn our dead, and swiftly move on. Do not ask for, or expect accountability this time! This time, it’s different.

    Did you know: COVID-19 is unique among all known viruses, for its ability to cause death from all other forms of disease, including a strange neurological illness which only affects bystanders? Fear, desperation, unemployment and vaccine-acceptance are its most striking features, and any suspicion of the virus is enough to warrant extreme social distancing measures, and a mass issuing of death certificates precluding all non-COVID forms of disease. If we were honest, we might consider placing it in the next edition of the Diagnostic and Statistical Manual of Mental Disorders (currently, the DSM-V), but then it would overlap too much with the definition of hysteria, which we had already deleted from the fourth edition.

    I wish I was joking!

    COVID-19 was/is a multi-trillion dollar fraud operation. The globalists have won, amazingly.

  33. @North Carolina Resident
    You can renew online.
    Also, most states are extending expiration dates.

    You can renew online. Also, most states are extending expiration dates.

    Well, that’s supposedly. When I went to renew online, they want the registration number on my birth certificate. I am eligible for online renewal with my last picture and eye test, but I don’t have my birth certificate with me. Massachusetts gives drivers with expiry dates in April a 60 day waiver, but haven’t granted same to those who come up in May. Hence my rush to get back up.

    Thanks, though, NC..

  34. @Hypnotoad666
    New York City death numbers are getting the most press on the alarmist side. They reported about twice the usual 3,000-ish death toll for March-April 10. They are basically counting every single "excess death" as either confirmed "with" Cov-19 or "presumed" to have Cov-19.

    Three thousand more people died in New York City between March 11 and April 13 than would have been expected during the same time period in an ordinary year, Dr. Oxiris Barbot, the commissioner of the city Health Department, said in an interview. While these so-called excess deaths were not explicitly linked to the virus, they might not have happened had the outbreak not occurred, in part because it overwhelmed the normal health care system. https://www.nytimes.com/2020/04/14/nyregion/new-york-coronavirus-deaths.html
     
    Of course, the medical system has not really been "overwhelmed." Rather, it was the panic that gave everyone the perception that it was overwhelmed. Nevertheless, the officials won't be able to resist the temptation to just include the entire "excess death" count as if it were caused by Cov-19.

    They have just barely begun the process of randomized antibody testing in NYC, and that should shed light on the infection rate. However, the bug has a long incubation period and is super-contagious in the crowded NYC eco-system. It probably ripped through 25% of the city before anyone knew what hit them. Thus, even if the entire excess death count were all Cov-19, it would still probably be consistent with a mere .1% CFR.

    Keep an eye on Sweden also. Critics are saying that their model is dubious because infections and deaths are going up there rapidly. But that is the whole intention for them -- to get it over with in a sudden spike that confers herd immunity as quickly as possible. If they are correct, their numbers will spike shortly and then drop precipitously.

    Super-contagious and not very deadly is exactly the combination that makes an economic shutdown expensive and ineffective. That reality should be starting to penetrate the Hive Mind soon.

    Agreed, but that herd immunity thing is still speculative. We’ll know before too long.

  35. @Kratoklastes
    Hmmm... no cell in the decision tree regarding insurance? KEK

    Small-c cynicism aside, one very interesting thing about that PDF is the tables.
     • Table 2 purports to decompose the 'admitted' (N = 1999) column of Table 1, but doesn't decompose the 417 admitted who are not critically ill (so far);
     • "Temp > 38°C" - percentages in Table 2 are not proportions of the relevant Table2 'population'

    The image below corrects both those things, and consolidates relevant rows of Tables 1 and 2.

    Petrelli et al - Composite Table 1 and 2 (Table 2 corrected where required)
    https://www.dropbox.com/s/8bi9zfhgvfje5pr/Petrilli_Table2Fixed_20200416.png?dl=1


    What stands out concerns the key co-morbidities that everyone (including me) has been considering an important determinant of mortality risk. It's kind-of still true, but it's at more than one remove; on their own they're useful identifiers of the probability of requiring hospitalisation, but they don't help identify which hospitalised people likely to go critical.

    There's a very strong requirement for an AND clause - in other words there's an interaction term (which also means that a generalised linear model is not a sensible choice).

    The 'AND' term is AGE (probably still categorical, with the category border in the 70s).

    Also interesting: if you wind up critical, your odds aren't good - 45% of people considered critical died (and much more than 50% of deaths were people on ventilators); another 38% are still on a ventilator (and therefore highly likely to die).

    Speaking of ventilators: in Figure 1, 445 patients were put on ventilators; 200 of those are no longer on ventilators.

    Of that 200, 162 (81%) are off ventilators as a result of being dead.

    By counting people who were still on respirators in the denominator for respirator cases, they report a lower number for the deadliness of respirators. It's 81% of all cases where a respirator is used until 'fixed-ish or dead'. By contrast the text says

    The largest detailed case series published to date included 1,099 hospitalized patients with laboratory-confirmed Covid-19 infection in China, of whom only 25 (2.3%) [=25/1099] underwent invasive ventilation and 15 (1.4%) [=15/1099] died. By contrast, 28.1% [= 445/1583] of hospitalized patients with definitive outcomes in this case series have so far required invasive ventilation and 18.5% [= 292/1583] have died. - [italics mine]
     
    Why not mention that 81% of mechanically ventilated patients died? It's the single biggest correlate to death.

    Seems deliberate.

    .

    There's also something very weird about the Odds Ratios for the Age variables in Tables 3 and 4; the ones in table 3 bear little relationship to the unconditional observed values (using 19/44 as numeraire). It indicates that the regression almost certainly found a local maximum.

    Age-Group Pr(Admission) and Unconditional Odds Ratios (19-44 = 1)
    https://www.dropbox.com/s/hznohpibv9hb0f8/Petrilli_AgeGrps_20200416.png?dl=1
    (Note: the high admission rate in 0-18 includes 19 newborns; none died)

    Sadly, they haven't made the data or model available yet: if they had I wouldn't have to pfaff around.

    The paper says that they used logistic regression - almost certainly a LOGIT model, which models Pr(RHS) and then converts that to a binary outcome (H/¬H) based on some selected threshold. It would be ideal if they used PROBIT, which is better suited to the problem.

    Both are part of R's glm (generalised linear model) package, but determining which to use requires that on the user understands why probit > logit for categorical dependent variables and a binary dependent variable.

    Since they use VIF to test for collinearity, all bets are off. VIF is known to be no good for categorical variables.

    Plus, with as many variables as they bunged onto the RHS (about 25, which is ridiculous), they're gonna need more than VIF < 2.5 to be assured of a full-rank regressor matrix.

    Why not just do a test of the regressor matrix? Farrar-Glauber works OK for categorical variables.

    The regressor matrix dimensions are O(10³)×O(2⁴), so they can't claim it's too computationally burdensome.

    Simple enough to go "old school" - generate X'X and calculate its determinant directly, or just try to invert it and see if you get a warning. Or get its condition number, or its smallest eigenvalue.

    All of those things are in R's mctest package; omcdiag from that package will give 'kitchen sink' output and a reliable declaration at the end - 0 or 1. Narrow down the culprit with imcdiag.

    You would think that glm would generate warnings when the regression was run - but logit and probit models are weird critters: often they seem to converge but produce nonsense.

    Plus, warnings can easily be overlooked, especially if someone wants to have small log files to wade though... just set options( warn = -1 ), which is the sort of thing a lazy person would do (say, someone lazy enough to fail to type VIF categorical variables into a search box).

    .

    I'm making clear my annoyance at the authors for not providing deaths by age group and respirator cases by age group. (Or better yet: the R file and the de-personalised data).

    This study has a decent sample size, and those two tables would be really useful to confirm several hypotheses (to confirm them properly, not using 'HelloUdemy' statistics - the H1Bindhi version of journomath).

    As it stands, the ORs for admission and progress-to-critical by age cohort result in a model that makes no sense (if you generate some pseudo-data with plausible values). Without the deaths by age group table, the regression results are suspect, even though the strong significance of "Age ≥ 75" conforms to my prejudices.

    This could be a really important publication if the regressor matrix doesn't have the problems I've mentioned above: they really need to drop the newborns from the analysis - with disclosure, of course - because they totally fuck up the hospitalisation regression. Newborns get hospitalised regardless (and none of the newborns got critical) - nobody would criticise the authors if they dropped those 18 data points and started age groups at 15-44 (even though more age-group granularity - 5 year cohorts - would be better and no harder to organise).

    Are you paid by the word ?

    What do you think the MORE tag is for?

  36. @Redman
    This is also going to need a lot of interpretation and explanation. Can anyone provide insight on what this is saying?

    Like who got critically ill?

    Hospitalize …
    if age > 65
    or obese and age > 35
    or (age less than 20 or greater than 44) and male and (diabetic or not white)
    or (age less than 20 or greater than 44) and female and diabetic

    (As you can see, they should have asked the question of ‘diabetic’ before that of ‘male’, which would have yielded a simpler decision tree.)

    I await the complaints regarding why an imaginary racist construct like whiteness is considered, and why non-binaries are not considered. Are they just supposed to DIE??

  37. @Henry's Cat
    Print on those graphics is too small for these eyes.

    Right click on the image and open it in full size in a new tab.

  38. @Henry's Cat
    Print on those graphics is too small for these eyes.

    Predict admission .

  39. @Henry's Cat
    Print on those graphics is too small for these eyes.

    Print on those graphics is too small for these eyes.

    It’s just as well: If you could read the chart, you might find you are doomed.

  40. The way the flow charts read is that if you have a pulse Ox below 88 when you get to the hospital, you are doomed to wind up in the ICU.

    Of course, that’s not a causal factor, that’s an attribute that may come from other causal factors — poor lung function to begin with, obesity to begin with, immuno-compromised to begin with, cardio-pulmonary problems to being with.

    I seriously doubt that individuals who have healthy pulse OX in the 98-99 range, should they get infected, are going to see their pulse OX drop a dramatic 12% in the course of a week. I suppose it’s theoretically possible, and may have happened in unusual cases, but they are almost certainly outliers.

    A true decision tree would have a history of pulse OX problems at the top, or have it set below the co-morbidities.

  41. Don’t think of these as actual decision trees used to make decisions, but as post-hoc reconstructions of what happened. (On the other hand, they may have some self-fulfilling prophecy aspects.)

    They would be a very poor basis for decision making. For example, the bottom line of Fig. 2A has:

    Red Box: admission 60%, predict admission
    Green Box: admission 45%, predict no admission

    If used as a decision tree for admission, the odds would be 100% and 0% respectively, when they should be 60% and 45%. The accuracy of using this part of the tree to make the decision is only slightly better than if you tossed a coin.

  42. Anonymous[240] • Disclaimer says:
    @Hail

    Your Odds
     
    The fatality rate for "normal flu" was widely quoted as 0.1% when the Corona Panic started to take shape. Viewed as odds, this seems kind of alarmingly high. 1 in 1000 to flu? But normally we don't think about it.

    An increasing number of pro-hysteria voices in the media began lecturing us on how we should cave into the panic, let the Panic assimilate us, and most importantly PANIC NOW (think of the toiler paper! won't anyone think of the toilet paper!), because the coronavirus fatality rate was at least 3%, maybe 5%, possibly 8%! Fear! Apocalypse! All power to the government and to the Holy Media!

    Then the more actual data we got, the more the experts kept finding a strange thing, namely that this much-trumpeted Apocalypse Virus actually also had a fatality rate probably down around 0.1%, maybe 0.2% or in that vicinity, with substantial room to argue over definitions and probably reducing the magnitude to <0.1% by some counts. One way or another, this was, it turned out, not-particularly-remarkable virus.

    In other words, one's Corona Odds are actually very, very good, and were all along.

    Here the Economist summarizes the latest study to endorse the 0.1% estimate published this week in the US:


    The paper reckons that 7m Americans were infected from March 8th to 14th, and official data show 7,000 deaths three weeks later. The resulting fatality rate is 0.1%, similar to that of flu.
     
    (The paper is: “Using ILI surveillance to estimate state-specific case detection rates and forecast SARS-CoV-2 spread in the United States.” Authors: Justin D. Silverman, Nathaniel Hupert, Alex D. Washburne.)

    Even this 0.1% estimated fatality rate is arguable, given the deaths-with-vs.-deaths-from problem. What is also increasingly clear is there are many marginal/excess deaths occurring that are corona-negative; these are direct, immediate-term victims of the CoronaPanic itself. And that's before the recession hits.

    … pro-hysteria voices in the media began lecturing us on how we should cave into the panic, let the Panic assimilate us, and most importantly PANIC NOW …

    Quite. They had their Ethiopian errand boy at WHO declare a PANdemIC because they needed the PANIC.

    Ron Unz pointed out earlier that the precise timing of first appearance is consistent with human agency.

    The characteristics of COVID-19 are consistent with it having been created to do maximize the slow-motion coup d’etat we are now witnessing.

    Imagine a James Bond-type bad guy had asked his science wizards and Georgetown-trained political science “nudging” specialists WHAT VIRUS would best support his dastardly plan for a quiet take-over of power in Western countries. WHAT WOULD BE DIFFERENT FROM WHAT WE ARE SEEING?

    COVID-19 characteristics:

    1. Slow-burning epidemic, including a modest number of fatalities and serious cases.

    2. Symptoms similar to regular flu for deniability. The populace would be much more suspicious if the disease produced unusual or novel symptoms as did traditional epidemics like measles, cholera, bubonic plague, etc.

    3. Disease symptoms include unusual attacks on oxygen transportation system (hemoglobin) that are difficult to treat. Ditto for cytokine storms.

    4. Serious outcomes most likely in the elderly and those with existing health problems, much lower risk for those more likely to resort to energetic, organized resistance.

    5. Initial release in China – could be an elaborate ruse to launch a bioweapon developed for purposes of advancing a CIVIL WAR or coup d’etat in the West.

    BTW we seem to hear very little about the different strains of COVID-19 detected in different areas/countries.

  43. Anonymous[240] • Disclaimer says:
    @jbwilson24
    "On which, it came out this week that the UK government had not even bothered modeling marginal lives lost to the effects of the ‘Lockdown’:"

    Fire the lot of the idiots. The UK possesses significant amounts of expertise in complex systems science, simulation, modeling, etc, but apparently the government never bothered to take advantage of it.

    Any basic text on policy analysis would have been sufficient.

    “On which, it came out this week that the UK government had not even bothered modeling marginal lives lost to the effects of the ‘Lockdown’:”

    You are assuming that the objective of the UK government is to preserve “marginal lives” yet their actions bespeak callous disregard.

    Perhaps they have other objectives they don’t care to talk about.

  44. @Michael S
    Funny how all of the actual data shows a CFR of anywhere from 2-20%, and all attempts at random testing show that even in the worst-hit areas fewer than 15% were ever infected and in most areas it's less than 1%, but the "estimates" constantly being put out by "economists" all manage to conclude that the actual data must be off by a factor of 100 or maybe more.

    If the data disagrees with the estimates, the estimates must be right! Millions of Americans were already infected and didn't die! No wait, tens of millions! No wait, hundreds of millions!

    Millions of Americans were already infected and didn’t die!

    That does indeed appear to be correct:

    New antibody data from the Italian community of Robbia in Lombardy shows that about ten times more people had the corona virus than originally thought, as they developed no or only mild symptoms. The actual immunization rate is 22%.

    If Robbia’s rate is the same as Lombardy’s as a whole, granted a big assumption, that is 2.2 million present-positives plus past-positives, the latter being those whose immune systems easily (in some cases after mild symptoms) batted away the virus at some point in the past few months and had no idea.

    Of the perhaps 15,000 Lombardy coronavirus-positive deaths that will be recorded by the time the epidemic ends (currently 11,600 with numbers in steady decline), there remains a significant problem of “deaths from vs. deaths with.” Between late February and late April, 4,500 people in Lombardy are statistically expected to die, the natural death rate absent a flu spike; with a flu-spike, it could be 6,000 (as observed several times in the 2010s, without causing a mass-shutdown-style Panic).

    What we know is a very large portion of the deaths are nursing home patients, reportedly about half. We know that Italy counts every corona-positive death as a coronavirus death, without distinction.

    We also know that during the CoronaPanic, nursing home staff abandoned their posts to a large degree and there were unnecessary deaths. To some extent nursing home patients are going to be among the highest mortality group in any society anyway, and also the first group at risk in a Panic. A report was that a lot of Ukrainian staff at these nursing homes and Ukraine recalled all their citizens when the Panic began. But others, including Italians, also began a very high degree of absenteeism.

    The nursing home deaths are, therefore, definitely not all unambiguous coronavirus-attributable deaths but some deep mix of ‘virus,’ ‘Panic,’ and ‘natural deaths,’ with some of the latter pushed ahead some months before they’d otherwise maybe have gone. The other deaths have to at least some degree a similar problem.

Comments are closed.

Subscribe to All Steve Sailer Comments via RSS
PastClassics
The unspoken statistical reality of urban crime over the last quarter century.
Which superpower is more threatened by its “extractive elites”?
How a Young Syndicate Lawyer from Chicago Earned a Fortune Looting the Property of the Japanese-Americans, then Lived...
Becker update V1.3.2