4 ms·

People keep repeating this idea that there's no basis for guessing the denominator of the CFR. This is silly. The denominator of CFR is the number of people po

People keep repeating this idea that there's no basis for guessing the denominator of the CFR. This is silly. The denominator of CFR is the number of people positively identified as having the disease. This isn't something that needs guessing. The problem is that the general population doesn't understand that you can't project from a CFR at one instant in time to the actual fatality rate in the population. But there's no need to criticize epidemiological terms of art just because people don't understand them and/or misuse them.

jMyles 7y ago
You are responding to arguments that nobody (at least that I've seen) are making. This has nothing to do with terms of art. The WHO used the CFR at a moment when every expert was pointing out that the vast majority of cases were unknown and unreported. If the WHO wanted to publish a specific report, outlining their methodology, and including the (not very useful) CFR, that'd have been fine, and in keeping with what we might expect from the practices of a world-class public health organization. But they didn't. They held a press conference, intended for the general public, during which the director read a prepared statement that said, "Globally, about 3.4% of reported COVID-19 cases have died. By comparison, seasonal flu generally kills far fewer than 1% of those infected," with no further clarification of what those numbers mean or what it means to compare them in this fashion. > The problem is that the general population doesn't understand that you can't project from a CFR at one instant in time to the actual fatality rate in the population. I don't think this is the problem. I think that the problem is that the WHO chose to use this number without explaining it, and used it in a comparison that makes no sense. > But there's no need to criticize epidemiological terms of art just because people don't understand them and/or misuse them. Just to be clear: you are criticizing John Ioannidis for his lack of understanding of the terms of art of epidemiology because he thinks that the WHO press conference was misleading? Do I have that right? How exactly do you suggest he frame his argument so as not to run afoul of this rule? Do you make the same argument of all of the experts around the world pointing to "the denominator problem"?
- mercutio2 7y ago
  Sure, there's a sizable denominator problem. That people misunderstand what CFR is is a problem, too. If we want to know how lethal the virus is, we need sampling of who actually has contracted it, which is tough in the early stages of a pandemic.
  - jMyles 7y ago
    > If we want to know how lethal the virus is, we need sampling of who actually has contracted it, which is tough in the early stages of a pandemic. But several studies have done the best job possible, given different pieces of the dataset, to arrive at a more accurate IFR (see the links in the above content). I haven't seen anybody, in any published paper, defend the 3.4% as a likely indicator of the actual fatality rate. So why are we still crafting policy as if this virus is between 10 and 100 times more deadly than it actually appears to be, according to the best data available today?
    - mennis16 7y ago
      Because it becomes 10x or so more deadly when proper care is not available, as is happening in areas where too many people have it at once. Letting this run wild would cripple hospitals. Many NYC area hospitals are already out of ventilators and that situation is just ramping up.
      - jMyles 7y ago
        > Because it becomes 10x or so more deadly when proper care is not available Source? > Many NYC area hospitals are already out of ventilators I don't think that's true. The governor predicted it last week, but it hasn't happened yet, right? I can't find evidence of this happening in any hospital, much less many hospitals.

jMyles 7y ago

You are responding to arguments that nobody (at least that I've seen) are making. This has nothing to do with terms of art. The WHO used the CFR at a moment when every expert was pointing out that the vast majority of cases were unknown and unreported. If the WHO wanted to publish a specific report, outlining their methodology, and including the (not very useful) CFR, that'd have been fine, and in keeping with what we might expect from the practices of a world-class public health organization. But they didn't. They held a press conference, intended for the general public, during which the director read a prepared statement that said, "Globally, about 3.4% of reported COVID-19 cases have died. By comparison, seasonal flu generally kills far fewer than 1% of those infected," with no further clarification of what those numbers mean or what it means to compare them in this fashion. > The problem is that the general population doesn't understand that you can't project from a CFR at one instant in time to the actual fatality rate in the population. I don't think this is the problem. I think that the problem is that the WHO chose to use this number without explaining it, and used it in a comparison that makes no sense. > But there's no need to criticize epidemiological terms of art just because people don't understand them and/or misuse them. Just to be clear: you are criticizing John Ioannidis for his lack of understanding of the terms of art of epidemiology because he thinks that the WHO press conference was misleading? Do I have that right? How exactly do you suggest he frame his argument so as not to run afoul of this rule? Do you make the same argument of all of the experts around the world pointing to "the denominator problem"?

mercutio2 7y ago
Sure, there's a sizable denominator problem. That people misunderstand what CFR is is a problem, too. If we want to know how lethal the virus is, we need sampling of who actually has contracted it, which is tough in the early stages of a pandemic.
- jMyles 7y ago
  > If we want to know how lethal the virus is, we need sampling of who actually has contracted it, which is tough in the early stages of a pandemic. But several studies have done the best job possible, given different pieces of the dataset, to arrive at a more accurate IFR (see the links in the above content). I haven't seen anybody, in any published paper, defend the 3.4% as a likely indicator of the actual fatality rate. So why are we still crafting policy as if this virus is between 10 and 100 times more deadly than it actually appears to be, according to the best data available today?
  - mennis16 7y ago
    Because it becomes 10x or so more deadly when proper care is not available, as is happening in areas where too many people have it at once. Letting this run wild would cripple hospitals. Many NYC area hospitals are already out of ventilators and that situation is just ramping up.
    - jMyles 7y ago
      > Because it becomes 10x or so more deadly when proper care is not available Source? > Many NYC area hospitals are already out of ventilators I don't think that's true. The governor predicted it last week, but it hasn't happened yet, right? I can't find evidence of this happening in any hospital, much less many hospitals.