If our policy and testing procedure were static for the duration of the pandemic at today's levels, I'd agree. We have to remember our official procedure from Mar-Jun or so was that if you showed up at a hospital with clear as day covid19 symptoms, but were 50 years old or younger, you got sent home without a test to ride it out rather than overwhelm the hospital and potentially infect more people. That's a massive asterisk in our denominator.
We're also clearly underestimating the number of asymptomatic positives. Even back in May/Jun they were reporting surprise test results from prisons where 75% of the population tested positive. They tested Yuma Prison on Tuesday and 655 of 1066 came back positive. And almost certainly the other 411 were exposed to a significant viral load, but for one of a million possible reasons are effectively immune(age, genetics, flu shots).
My point is that it's exceedingly unlikely we have only a low multiple of 15.7M cases (up from 15.5M in the 16hr since I posted) with a virus so infectious and with a longer incubation period than almost all other coronaviruses. Logic says anyone in the US with moderate and repeated contact with dozens of people has been exposed to a level that would likely infect someone with susceptibly to infection. To me that's probably about 100-180M people. Of those, about 80M likely can't or would have a very difficult time getting and actual diagnosable infection. Those would be mostly kids and annoyingly healthy <45 people.
I just don't see how any other scenario is possible given the factors at play and the lack of precautions exhibited. I'll keep harping on the Swine Flu point.... in the 2009 pandemic we assumed that maybe maybe 1-5% of the US population had been infected with H1N1pdm09 virus, turned out the real range was 43-89M by the time we had data clarity a year later. We are on the exact same path IMO, currently at the denial of widespread infection stage.
Like everyone else here, we wish you were right, but the data we can gather doesn't support it. I personally know a fair number of people who had classic COVID symptoms and never showed up in the statistics. I lost my sense of taste and smell for a few days with no sinus congestion. I also live in the Northwest where COVID was a problem early on.
When the data has a lot of noise, you need to approach things from different angles and look for correlations between results. Most of the estimates here and elsewhere come up with somewhere 40-60 million actual infections in the US with an outside of about 75 million.
As
@AlanSubie4Life pointed out, most estimates are that about 0.6% of people who get COVID die. The US currently is around 3%, which does show an under reporting in US figures. The reported figures from John Hopkins today is 15.78 million cases and 294,144 deaths. That's a death rate of 1.86%.
If 100 million people had actually been infected with a death rate of 0.6%, that would be 600,000 dead. The numbers for deaths are under reported,
@AlanSubie4Life explained why. They might be as high as 360,000 dead, but hiding almost half of all COVID deaths at this point would be impossible.
Another way to look at it is to take the 360,000 estimated dead based on 100 million cases, that's a death rate of 0.36%, which would give the US one of the lowest death rate in the world. The only countries that report that low are suspicious due to lack of infrastructure or an authoritarian government that wants to play down the virus. The US does have medical infrastructure that is on par with some of the best countries, but it has massively botched handling the pandemic and the US has a larger percentage of the population with known risk factors than most of the rest of the developed world.
At best the US fatality rate is on par with the 0.6% estimated.
I wish you were right, but I don't see data to support it. It does seem like some people seem to have some resistance to getting COVID. Rudy Giuliani is an example. People were wondering why he didn't get it because he was exposed several times, but he did eventually get it.
This is a good post, but you should use the term IFR (infection fatality rate) rather than CFR here.
The IFR has indeed dropped a lot since the spring, but unfortunately is climbing again as ICUs run out of space.
Death reporting has a much lower undercount than cases, but even here the reporting lags by several weeks. We know Texas is slow at reporting deaths, and probably has a large backlog of deaths that hasn't been reported yet. Over the summer, newspapers revealed that the backlog reached as many as 3800 people dead but not yet counted in the official stats. Florida and California are also slow, though not nearly as slow as Texas. The Dakotas shipped many severely ill patients out of state, many of whom have likely died, but it's unclear if these are included in the Dakotas' counts or are listed for MN, CO, and NE instead. Same with Wyoming.
Point taken. I keep getting IFR and CFR mixed.
If patients from the Dakotas were reported in other state totals, it would still show up in the national totals. But, yes we do have a national undercount. However, deaths is less noisy data than the number of positive cases reported. I was trying to extrapolate something from the data that might be more reliable. But every number has noise.