You're embarrassing yourselves by comparing Covid-19 to the "Seasonal Flu".
You need to compare apples to apples, IFR to IFR
If you get the seasonal flu your chance of dying is around .03% [1] [2]
If you get Covid-19 your chance of dying is around 1%. [1] [2]
That means Covid-19 is 33 times riskier than the seasonal flu. Russian Roulette is only about 17 times riskier than Covid-19. Maybe find a new comparison.
Thanks for stopping by.
FAQ
What about the .1% flu death rate I sometimes see?
Typically this is an “isolated demand for rigor”. The CDC website estimates deaths from flu and cases of flu each year, and if you divide the deaths by the cases you get .1%. But if you do this same process on CDC Covid statistics you will get a Covid-19 fatality rate of 7%.
If you’re like me, you’ve seen a lot of people cite the “.1%” number for flu and almost nobody cite the 7% number for Covid. Let’s demand rigor for both of our numbers, shall we?
Most of us don’t use the 7% number because we know there are a lot of coronavirus cases that don’t get officially tested. The gold standard comes from estimating total infections using blood serology to test for antibodies. This has been done for well for Covid and the results are usually what you see cited (generally showing 1% infection fatality rate, called an IFR.)
But this same issue applies to the flu. Have you ever been tested for the flu? I know I haven’t! But I’m quite sure I’ve had it. So we need the number of people infected with the flu.
Fortunately, blood serum work has been done for the seasonal flu too.
Here’s how many people get infected with the seasonal flu each year based on serology: 1,326,000,000 . Here’s how many people die from influenza-like illness every year: 389,000 . Divide deaths by illnesses, just like we did with coronavirus, and you get .03% IFR.
Voila, now we are comparing apples to apples. And what we learn is:
Coronavirus is 30 times as deadly as the seasonal flu.
What about the young people?
The NYTimes just ran a piece claiming that “young, healthy Americans have a fatality rate similar to that of the seasonal flu.”
Young people should be angry about BS like this.
The author means to say “young, healthy Americans have a fatality rate similar to that of the seasonal flu for old people”. Not quite so reassuring is it?
But even that statement isn’t true. Here are the number of people who have already died of SARS-CoV-2 in New York City:
Scan over to the column second from the right. Look at the entry three rows down. That says that .02% NYC residents between 18-44 have already died of coronavirus. The all-ages IFR for seasonal flu is .03%. Please take a second with this.
The flu is estimated to kill .03% of those it infects. Most of those it kills are old. Covid-19 has already killed .02% of the young people in NYC.
Do we overcount Covid-19 deaths?
To misattribute a death, first someone has to die. Look at the scale of death in NYC April 2020 compared to average:
Are some of those deaths in the giant spike on the right misattributed? I mean, why not. Go ahead and pick your favorite stubby orange line on the left side of the graph and imagine that we are ignoring 100% of all of those deaths. That means the spike on the right is artificially boosted by that little tiny stub. Shrug.
If you want to proceed further you need to account for that huge spike. And you should recognize that we have a very plausible-looking hypothesis for this spike.
There is a new type of virus called SARS-CoV-2 that has spread around the world.
This virus has caused countries like China, not known for humanitarian tendencies, to lockdown their society.
It started spreading seriously in NYC by March. Many people tested positive for it. Many people had serious symptoms. Some of these people died.
And all while this is going on we had that enormous spike in deaths in NYC. And the majority of those deaths (more than 3/4ths) also tested positive for the virus.
Sure seems to me that it might be that new deadly virus.
But could your personal favorite alternate explanation be the cause instead?
Technically, sure. Just like it’s technically possible that every dead soldier at the Battle of Gettysburg had a heart attack right before they were shot. But at this point you are basically doing a bad Descartes impression.
What’s actually behind the CDC .1% flu estimate?
Do some digging and you find that the .1% number comes out of this process: 9% of hospitals report their hospitalizations for flu tests to fluserv. CDC estimates that this number is too low and so it revises it up by some amount. Then this new number gets multiplied by the “rate of death per hospitalization” number that the CDC has estimated. This is the number of deaths.
Then the CDC goes back to that hospitalizations number (revised up to correct for “under-detection”), and multiplies it by an estimated “ratio of illnesses to hospitalizations” to give the number of illnesses.
That means the .1% estimate is derived from the raw number of hospitalizations in 9% of the hospitals in a given year. Which is then scaled up to a representative proportion, then scaled up again by a pre-existing parameter to correct for flu under-detection, then multiplied by two other pre-existing parameters—one to estimate deaths and one to estimate overall “symptomatic” illnesses.
If you want to compare the .1% number to SARS-CoV-2 you should run the same model on SARS-CoV-2. Otherwise you should use the IFR.
I've always found it shocking that these "CDC is lying to us" people never thought that the CDC might be exaggerating flu deaths. So crazy to treat that .1% as if it's handed down from Moses or something.
Nice to have put this info in one place. I'm going to send this around.
I think you could be more clear about the citations at the beginning. i think I figured out that they correspond to the numbers from later.