That seems like a plausible explanation. I wonder how one could confirm it? If there were two sets of covid death counts based on strict and on loose criteria, one could see if there is a divergence in their rates of increase.
Your comment didn’t show up at my blog. Not sure why...
Fuirst, I don’t think looking at “number of cases identified” is useful at all for this analysis. Mainly because this is too much a function of how much testing you do that it’s not very meaningful.
Let’s write down the hypothesis: “the linearity observed is primarily due to the arrival of “deaths door” for those approaching end of life. A fixed fraction of these deaths have covid in any country because a fixed fraction are in living situations that give them some minimal exposure.”
So what are some consequences of this hypothesis?
We could look at the death rates of young folks, say those under 50. We’d assume that for that set, their death would not be caused by frailty with covid. If they die with covid and covid-related symptoms, this would not be a death due to frailty. We know that among those under 50 only a small percentage have complications, so this absolute number should be very small. However we know it is not 0.
What we should see is that THAT number should go up exponentially (though remain small) as reopenings begin. However, since that number will STILL be small even at full infection rates, this shouldn’t effect the totals too much because that will be dominated by the linear arrival of “imperfectly isolated frail folks dying with covid.”
As well, the test cases too look at would be cities and regions that have not seen major infection rates or death rates. We might expect that in a place like NYC (where serological tests are suggesting up to 20 percent infection rates), that the young folks who were going to die of this may have already died.
But perhaps the interior US will allow us to test this hypothesis.
My bet would be: a small absolute number of young folks dying of the disease, but that that absolute number will go up exponentially in regions with very small numbers of covid right now.
That seems like a plausible explanation. I wonder how one could confirm it? If there were two sets of covid death counts based on strict and on loose criteria, one could see if there is a divergence in their rates of increase.
Your comment didn’t show up at my blog. Not sure why...
Fuirst, I don’t think looking at “number of cases identified” is useful at all for this analysis. Mainly because this is too much a function of how much testing you do that it’s not very meaningful.
Let’s write down the hypothesis: “the linearity observed is primarily due to the arrival of “deaths door” for those approaching end of life. A fixed fraction of these deaths have covid in any country because a fixed fraction are in living situations that give them some minimal exposure.”
So what are some consequences of this hypothesis?
We could look at the death rates of young folks, say those under 50. We’d assume that for that set, their death would not be caused by frailty with covid. If they die with covid and covid-related symptoms, this would not be a death due to frailty. We know that among those under 50 only a small percentage have complications, so this absolute number should be very small. However we know it is not 0.
What we should see is that THAT number should go up exponentially (though remain small) as reopenings begin. However, since that number will STILL be small even at full infection rates, this shouldn’t effect the totals too much because that will be dominated by the linear arrival of “imperfectly isolated frail folks dying with covid.”
As well, the test cases too look at would be cities and regions that have not seen major infection rates or death rates. We might expect that in a place like NYC (where serological tests are suggesting up to 20 percent infection rates), that the young folks who were going to die of this may have already died.
But perhaps the interior US will allow us to test this hypothesis.
My bet would be: a small absolute number of young folks dying of the disease, but that that absolute number will go up exponentially in regions with very small numbers of covid right now.