One factor I haven’t seen discussed much, but which may have a role to play in explaining the lower IFR we’re currently seeing is that it appears to be the case that a lower initial viral load is correlated with a less severe illness, and mask wearing, while it can’t protect you from being infected, may lead to lowering the initial load:
That wouldn’t explain Florida’s lower fatality rate, though, since I don’t think they were that good about mask wearing.
Masking is not the only behavior that leads to a lower initial dose exposure. Less time in close distance, less time in close distance indoors, so on, also do. Nor of course does anyone think that it is the only factor.
The actual NEJM perspective discusses the support for the idea in detail, and I have to say that I was a bit skeptical before reading it but they make the case very strongly. Animal models to epidemiology.
I also wonder if smaller initial dose correlates with a smaller antibody but bigger T-cell response …
I can name some possible contributors.
First, at least in my hotspot part of Florida we saw & still see a very bimodal distribution of behaviors. Folks over age 50 and especially those over 70 (WAG) masked up heavily and stayed / are staying mostly home. 20-30-somethings did not / do not; they’re living life pre-COVID with moderate masking compliance.
IMO that behavior split will contribute to an age skew of those infected towards the youthful that generally favors milder cases and more survivors among the fairly few who become badly infected.
A second issue is that even at the height of our summer peak in the hotspots, we never got close to, much less beyond, our ICU capacity. The medical industry was working hard, but wasn’t running flat out to exhaustion. The end result is better care. It got close, but the local governments in the hotspot areas acted aggressively when the spike became obvious and folks mostly complied.
As well, to the degree the older and the elderly were both shunning non-emergent medical care that freed up assets to further bolster the COVID response. Those folks are an unusually large fraction of the urban/suburban population of the hotspot counties. So they have an unusually large leverage on total demand vs. that seen in other parts of the country. And we’re somewhat oversupplied in high tech / ICU facilities versus less elder parts of the country.
The fact the last spike was during summer when all our part-time residents were up north further helped. Our medical facilities are sized for our winter population that’s 20+% higher. That effect may go into reverse if we get a winter spike.
Third, FL is a very outdoorsy place. The vast majority of maskless or otherwise incautious behavior is happening outdoors. Restaurants have lots of outdoor space and people use it. My personal rule is I sit indoors only in a restaurant that’s otherwise nearly empty. Me, crowded, indoors; pick 2 or more likely only 1. I’m far from alone in that attitude.
Fourth, hotspot FL is, corrected for age, a less obese place than much of the rest of the south and the USA. Fitness is a bunch more popular here, even among many of the older set.
Fifth, for whatever reason, our Trumply Governor DeSantis very early on took a very draconian approach to regulating the residents and the staff of communal elder homes. To this day we still can’t visit my aged MIL at her home, although starting a week or so ago we can now take her out of the facility and return her without her being forced into a 14-day in-room-only quarantine.
As such, FL has largely avoided the wildfire transmission in elder communal living that spiked the death rate so severely elsewhere in the early days of COVID.
I suppose the big point here is that in COVID, as in life and everything else, effects don’t have single simple identifiable causes. The effect we see is the result of several, perhaps hundreds, of different causes all pushing one way or another that net to the result we see.
Said another way, looking for the cause of darn near anything in society is a fools’ errand. We should look for the major contributing factors instead. And inevitably some of them will be contributing positively and others negatively. The net is what matters.
Thanks for this detailed response!
Once the dust settles, I imagine there will statisticians poring over all of this – it has been a massive science experiment with tons of data.
Sadly, tons of data with huge uncontrolled = unknowable noise bars. As long as who gets tested is chosen by the testee, and who gets good or bad health care or exposure is chosen by their SES, we’ve got more noise than signal.
And certainly enough legit noise that anyone who wants to dispute a finding can find a legit-enough objection to some statistical adjustment to the data that’s enough to shed legit doubt on the conclusions.