Shockingly low. 0.5% seems like the absolute floor of remote possibility. I think 0.75% to 1.5% is probably a reasonable range without looking too deep.
It’ll be higher than that when care is rationed, and might get down to like 0.5% when cases are low, it’s caught early and monitored, quality of care is high, etc.
Lombardy was like 1.1% based on some towns where they tested everyone for antibodies. That seems to be something a worst case scenario given it was early when they didn’t know how to treat it as well, and no masks to mitigate viral load.
I agree it’s possible North Dakota might be .7% or something if there are really a large % of the populace taking zero mitigation measures.
Well if you’re using IFR to estimate how many people have had it to draw conclusions about herd immunity, you can’t use the IFR from a couple months ago when ICU’s weren’t slammed and care was better then apply them to case totals in the spring when shit was hitting the fan. You’ll drastically overestimate cases from the spring, and think we’re way closer to herd immunity than we are.
I do think there’s a real chance some states will hit herd immunity before the vaccine but that says more of their determination to spread the 'vid than it does about how we’re counting cases.
early in the pandemic 1.5% was looking like a strong possibility. since they figured out ventilator deaths and make enough regeneron, 0.5% isn’t out of the range. no?
I’m not sure how this matters. .2% of the NJ population is dead in 8 months. For the actual IFR to be .3% you would have to believe the 70%-80% not infected have like a .03% IFR or something. That isn’t realistic.
I think 0.5% is the low end of reasonable but it really depends what you’re asking. Are you asking our optimal condition IFR? For me and you or the president? Cause you and I are probably not getting Regeneron, especially not early on before it’s dire.
Are you asking our total IFR to date? What our total IFR will be in the end?
Suzzer is trying to guess total cases based off death tolls, in order to see how close we are to herd immunity. As a result that’s the total IFR from beginning til now, and it’s going to be a lot higher than 0.3% imo. 0.5% would be amazing news and I’d be pretty surprised but not totally shocked. 0.3% would be jaw on the floor holy shit.
Like the flu is 0.1%. If it’s 0.3% the “it’s the flu” people were way closer to right than we were. I don’t buy it.
.5% is possible. Lots of states, even later hit states are at or over .1% of their entire population dead in 3/4 of a year. I would say .5% is on the low end but not improbable.
Crazy how we go from 0 positives in a 3 month bubble to this!!! Shit, they might have herd immunity before the season even starts.
https://twitter.com/shamscharania/status/1334270996803620866?s=21
I’m saying the IFR was probably a lot higher back then for the reasons I mentioned. NJ’s IFR could have been 1% back in March/April, but I doubt it is now due to better treatments and mitigation helping keep initial viral load lower.
It’s like pressure on a leaky water connection. At low pressure it might now leak at all. (Essentially the bubble keeping community exposure at 0)
At high pressure The connection sprays out water ie any community exposure is going to lead to problems.
I can get behind that. Now do one of the recent comers like the Dakotas who have seen .1% of their population die from Covid in the last 60 days. .3% won’t be right there either.
Well .3% might be with the most vulnerable isolating. So it’s not really true IFR. You’ll never really know true IFR unless everyone just goes nuts and ignores the virus. Lombardy in the early days might be as close as we get to knowing.
But again - I think IFR has gone down since then due to better treatments and viral load mitigation. How much? Who the hell knows. Italy was just throwing everyone on a vent from what I read. And no one was wearing masks.
I agree with Wichita pointing out .2% is absolute floor for obvious reasons. So .3% if you assume full population is not realistic. But also NJ might be a lot less healthy than other states. Given my experience around the Meadowlands area and inland - I would strongly suspect this.
But maybe .5% is realistic in SD and ND if at least nursing homes are taking it very seriously there - which I assume most are.
True ifr is so tricky vs the flu because most almost everyone with the flu knows they have it. Between asymptomatic cases and testing limitations we don’t know with Covid.
Working backwards is fraught with peril because we know that simply by age there are drastically different ifrs. So unless we are carefully controlling for age, pre-existing conditions, and health care availability it can be a guess. Probably gets better as we aggregate the whole country to take out some of the variability. But as several have noted there is an arrow of time with improved treatments.
Under ideal conditions I’d use 0.5% to back calc. That compares to a a little under 2% cfr so something in the 4x range.
But as testing gets limited we know cfr will go up. As health care becomes limiting or we saturate old folks homes we suspect/know ifr will go up.
Again we should focus on the can’t be too far wrong concept. Use something in the middle of range of estimates. To me thats a ifr of 0.5, maybe even 0.4 if truly ideal conditions. Blended it’s going to be a bit higher- maybe up to 0.75.
Sadly we won’t know for sure until the cfr goes nuts. If we hold at 175k/day cases but the positivity continues up then we can be sure we are detecting a smaller fraction of true cases.
The shitty thing is the baked in number by the end of the month is 4K/day. Any surprises and/or further growth and that could be higher then and certainly higher into mid-January.
2650 today so far with LA county, a bunch of CA/TX counties and 3 states yet to report. The all time record is toast the first daily total higher than 9/11 is the real sweat.
IFR isn’t that useful a number really since it depends so drastically on the age distribution of a given population. The CDC isn’t even doing an overall IFR estimate anymore, they just give numbers by age cohorts.
A recent meta-analysis published in the WHO Bulletin gave a median IFR of 0.27% based on seroprevalence studies, but the range of included values was 0% to 1.63%. So if your reaction to the number is “that can’t be right” then that’s because it isn’t right for the areas you’re thinking of. In countries where life expectancy is low anyway, or in like university towns, it’s going to be even lower than 0.27% probably. There is no “real” answer without first answering the question “how many old people you got?”.
COVID being worse than flu isn’t just that IFR is higher, it’s also that R0 is a lot higher, so it rips through populations much faster.
I think this is a good post but we have large and disparate populations here in the US that have already seen 0.2%-0.3% dead. NJ and NYC for example. Now suzzer’s point that we have improved treating the virus is legitimate. Those places would have fared better with the same outbreak now vs. March. Which I guess proves your point that it isn’t an absolute number. We are closing in on .1% of the entire US dead from Covid. We are probably already there and then some based on excess deaths. I would bet a lot of money current IFR is between .3% and 1%. Wouldn’t you? Obviously if we totally overrun our hospitals that can get worse and if we continue to improve treatment that can get better.
I was promised by OAN that it was supposed to end on November 4th. WTF.
