This has been linked a few times here, but just in case people haven't seen the report. I'll be using the "Current Best Estimate" numbers. https://www.cdc.gov/coronavirus/2019-ncov/hcp/planning-scenarios.html

IFR of 0.4% all ages if symptomatic
35% of cases asymptomatic
IFR of 0.26% for all ages

Now we have a little over 100k deaths attributed in the US, so I'll round. 100k deaths with an IFR of 0.26% would work out to having 40,000,000, 26M being symptomatic, positive cases in the US. We have ~1.7M confirmed positives.

I realize there was a lack of testing in the beginning, but I'm having a hard time thinking we've had 26M with symptoms and so little activity in hospitals. I'm not sure where numbers may be wrong or why, but interested in discussion. A few thoughts:

• They estimate 3.4% of symptomatic cases result in hospitalization. Have we really had 800k hospitalized? According to the IHME model we peaked with about 60k hospital beds needed in a single day. I'm struggling to find numbers nationally on total hospitalized.
• ~25 million symptomatic cases wouldn't have required hospitalization. That's pretty massive and makes me wonder how mild the symptomatic yet no hospital cases are and how many are going untested.
• 40M just seems so massively higher than we've thought about I'm not sure what to make of it. That is about 12% of our population.
• Do number of positives matter at all when they account for such a tiny percent of those who have had it? Even if we have massive spikes
• We are saying about 12% has it, but most states are only having 3-5% positive rates on testing. That would imply that a number of people being tested already have the antibody but yet are likely sick enough to warrant a test. Seems strange.

Anyway, curious other thoughts.

Those numbers at the top of your post have got to be wrong. I thought that when I first heard them the other day (I think Ben Shapiro talked about them on his podcast). I went through the same mental exercise that you did when I heard them and they just don't pass the reasonableness test. I would be very curious to hear how the CDC arrived at them because clearly it wasn't through testing, because like you said only New York has been anywhere close to having that percent positive on antibody testing.

If they are wrong then it's on the CDC and it would mean the IFR should be higher. I suppose my extrapolation assumes equal distribution based on age and if clusters in nursing homes are causing the most vulnerable to also being getting it way more frequent than others. But, not sure I'd by that.

Also, does R0 of 2.5 seem really low if compared to 40M? I tried to find a calculator to examine how it would have spread with R0 of 2.5 but couldn't find anything good.

It seems to me either death rate is higher or R0 is higher. Not sure which would be better.....

Carl Bergstrom had a twitter thread about this:

The CDC's "best" estimate of the IFR does seem implausibly low. Putting in the NYC number of deaths (currently almost 21,500), and population size (8.4 million). An IFR of .26% would mean that over 98% of residents of NYC were infected. That is almost 5 times higher than the (late April) estimate of infection rates in the city.
The crowded hospitals could have elevated the IFR somewhat, but it seems very implausible that it inflated it by a factor of 4 or 5.

ORAggieFan said:

---

If they are wrong then it's on the CDC and it would mean the IFR should be higher. I suppose my extrapolation assumes equal distribution based on age and if clusters in nursing homes are causing the most vulnerable to also being getting it way more frequent than others. But, not sure I'd by that.

---

Yes sorry if I was unclear, I meant that I think the CDC numbers have to be wrong - not your numbers. I think your extrapolations based on their numbers are correct.

CompEvoBio94 said:

---

Carl Bergstrom had a twitter thread about this:

The CDC's "best" estimate of the IFR does seem implausibly low. Putting in the NYC number of deaths (currently almost 21,500), and population size (8.4 million). An IFR of .26% would mean that over 98% of residents of NYC were infected. That is almost 5 times higher than the (late April) estimate of infection rates in the city.
The crowded hospitals could have elevated the IFR somewhat, but it seems very implausible that it inflated it by a factor of 4 or 5.

---

Is anywhere else reporting numbers like NYC? Perhaps the numbers are inflated?

GAC06 said:

---

CompEvoBio94 said:

---

Carl Bergstrom had a twitter thread about this:

The CDC's "best" estimate of the IFR does seem implausibly low. Putting in the NYC number of deaths (currently almost 21,500), and population size (8.4 million). An IFR of .26% would mean that over 98% of residents of NYC were infected. That is almost 5 times higher than the (late April) estimate of infection rates in the city.
The crowded hospitals could have elevated the IFR somewhat, but it seems very implausible that it inflated it by a factor of 4 or 5.

---

Is anywhere else reporting numbers like NYC? Perhaps the numbers are inflated?

---

Nothing points to NYC being inflated, but you can remove NYC and the scientific consensus across the globe is still pointing to an IFR in the 0.5 to 1.5% range. His point was NYC would have to be so absurdly off / an outlier for this to make any sense.

Here you have the CDC claiming a 0.26 IFR and a S-IFR of .4%. Insane.

[keep the political comments on the Political forum and not here. - Staff]

Sarduakar said:

---

Why cant the 100k deaths be wrong? You assumed it was correct and tried to extrapolate back. Could it be much lower from actual Covid, the CDC knows this?

---

And the rest of the World doesn't? I'm no betting man, but...

CT75 said:

---

BQ_00 said:

---

ETFan said:

---

Sarduakar said:

---

Why cant the 100k deaths be wrong? You assumed it was correct and tried to extrapolate back. Could it be much lower from actual Covid, the CDC knows this?

---

And the rest of the World doesn't? I'm no betting man, but...

---

Maybe the rest of the world does know this, it just doesn't get as much attention as the doomsday reports.

---

"The people who die are essentially those who would statistically die this year anyway, because they reached the end of their lives & their weakened bodies immune systems can no longer cope with any random everyday stress (including ~ 150 viruses currently in circulation).

IMO ....that seems to be an accurate observation and description. Of course there are outliers....always are.

---

A CDC database issued in mid-April actually supports this thesis.In the 65+ age group, the percentage of COVID deaths and the percentage of deaths from all causes were materially the same. End of life is a stat that not a single human has avoided to this date (outside of Transylvannia)...