So y’all are pretty much agreeing then? Initial IFR from the NYC explosion was 1.5% or more, now the population of freshly infected folks is skewing differently, maybe to .5% or less, which might bring the aggregate closer to 1% for now?
Aggregate IFR is an academic argument to me as it’s different in every region based on demographics, and we’re never going to get a whole population of a region caught unaware again.
IFR by age cohort, and then that combined with current case demographics in a region - would be a lot closer to accurate imo. There’s probably data out there to do it.
That’s why CFR is good. At least you can model that against deaths and look at past cases to try to predict the deaths in the near future.
The main reason to be interested in IFR imo is to try to guess at the ratio of actual cases to confirmed cases, in order to guess at the current level of herd immunity. But I’ve become convinced that’s a big mess of constantly shifting sands.
The DC data modeler has something to say about this:
July 27th COVID Fact Check: Did the change in reporting of hospital data from CDC to HHS affect case data?
In the screen capture below, a widely shared poster suggests that there is some illicit plan in motion by the Administration to reduce the number of tests published by the government through the movement of the hospital reporting system from the CDC to HHS. This SPECIFIC claim is clearly false.
The reason it is not possible for the move of hospital capacity data submission from the CDC to HHS to have caused the referenced changes in confirmed COVID-19 cases is that THOSE DATA ARE NOT CONTAINED IN THE HOSPITAL DATA. The hospital data that have been moved to the new HHS system include such metrics as hospital bed occupancy, ICU occupancy, PPE stocks, staffing availability and the like. They do not include test or death data. You can see what has been collected in this system here:
https://www.cdc.gov/nhsn/covid19/report-patient-impact.html
To reiterate, hospitals don’t contribute to the national case or mortality data through this system. Case data are collected directly from the testing labs (and occasionally local health agencies) who send them to the CDC and continue to do so. You can see how this works at the following link.
Hospitalization data are vital, and there is a lot to be concerned about if these data continue to be kept from academics, journalists, and the general public. That said, there is no apparent pathway I can see through which this change could be contributing to the changes in confirmed cases (mis)represented on the graphic attached.
I’ve encouraged the author to revise this post to indicate that the data represented are not being delivered to HHS as part of the July 16th change. I covered the video URL so as not to be contributing to its dissemination.
Also, please wear a mask.
8 Likes
The south tho
I would like to do that, but I don’t think I’m smart enough to do it on my own. Is there an example state you want to try? How would you want the data presented to try to do this? This was the project I was talking about a week or so ago that I wanted to try to figure out how to do but a bit differently than you described. I quickly saw I wouldn’t be able to do it at the day level, so the goal would be to try to figure out how to do this with the weekly stats and project about 2 weeks in advance to see if the hypothesis holds. I think within maybe 2 weeks I’ll be pretty confident I’m really close on target SDI guesses if trends continue to hold. I’m monitoring a few places I might change, but there are only maybe a couple I don’t feel relatively good about (DC being the main one).
The way I think of it is this:
5 points above target significantly slows spread
5 points below target significantly increases spread
on target just slows it somewhat
a little above or below slows or increases spread noticeably but not by a lot
10 points above or below target majorly increases or reduces spread
Technically, I think the true target SDI should be 10 points above the current target SDI guess but very few places are willing to do that. No one has gone below target SDI and gone back to 10 points above target even though nearly every place in the country was once 10 points above its target SDI in the height of the early shutdown.
I think North Carolina might be a pretty good test case for ‘predictions’ due to them being well below target SDI for a long period of time. My hypothesis is each place has a maximum caseload it can hit based on current testing (testing level can affect this obviously) and effects beyond a certain low SDI score won’t get worse, they’ll just sustain for a long period of time. If Arizona has a legit drop, that’s good data that suggests a place can hammer the curve down quickly if they did any decent mitigation originally. That’s why it’s really important to hope Texas does start reducing quickly after it got back to the target SDI for 4 weeks straight. We should start seeing Texas have a reduction in cases from 7/22-7/28, and then a significant reduction from 7/29-8/4 if this hypothesis holds (and if the Texas SDI holds).
For North Carolina, I think its caseload based on current testing maxes at around 2500 per day. Florida might be closer 13000, California, 11000 or 12000, and Georgia maybe 4500. If these places got above those numbers, I would want to see if testing really increased to cause that.
I have no idea how much sense any of this made, but hopefully it made some. If there’s some data you want to test this with, I probably have it so just let me know exactly what you need.
Something interesting is possibly going on with the umd website. They uploaded July 22-24 and SDI is way up nearly across the board during that time frame. I reached out to them, but got an out of office reply from the lab director, so it seems like it’s unlikely I’ll get a response.
It seems possible they’ve changed their methodology if these numbers are correct and aren’t an actual change in behavior (trash can for all my analysis…lol). The big difference is in the amount of people staying home from what I can see at a quick glance. I think this might be an error on their part, they’re changing methodology, or behavior is now significantly changing. I guess we’ll see soon enough.
No, he thinks NYC was 1% but I think it was higher and 1% is the average for a random person.
It will be higher in the 50M that are 65+ and lower in younger people.
I’m not trying to nail down the exact current IFR as we don’t know enough about the current population and it’s localized and constantly changing as the virus spreads so I’m just assuming 1% as a general rule unless talking specifically about a younger population (eg we won’t see 1% of pro athletes that get it die) or an older population (nursing homes will have over 1%)
I’ve had this discussion with Suzzer and others since March so I don’t go over it all every time, but he’s been consistently trying to move it down to the .3-.5 range and I’ve always maintained that 1% range is going to be the best guess for us to use for general population.
If he wants to see my work or my reasoning he can reference the last 5 times we had this argument in the prior covid threads as well as in the 2p2 threads.
In early July at a WHO expert forum, the average of IFR guesses was 0.6%. CDC’s current best estimate is 0.65%. Antibody testing in Germany suggested a national IFR in the range 0.12% to 0.87% with the best guess being 0.37%. 1% is a bit high even factoring in undercounting, but 0.8% would probably be reasonable. I’d bet money it was upwards of 1% in NYC given the aggravating factors.
Easily
Now do Florida right in July 2020. 
What should we be doing differently if the IFR is 0.5% vs if it is 1.0%?
Knowing IFR would help to try to guess at current level of herd immunity - which is the context that it came up itt.
I’ve come to the conclusion that IFR in a region right now is way too population-dependent and unknowable to be useful for deriving other metrics.
For a while the CDC IFR estimate was 0.27%, I think it was, and that Ethical Skeptic dude went bananas inferring case numbers on that basis and showing how NYC actually had eleventy billion cases or something. Then the CDC changed their IFR estimate to 0.65%. I have not seen any mea culpas from him about this, but obviously I’m not going to hunt through the avalanche of incomprehensible drivel on his feed very hard.
But anyway, don’t be that guy. IFR is too uncertain and varies too much depending on parameters to use it as the basis for analysis.
1 Like
I don’t know if 10x is a proper multiplier right now. Those states could be at 20% when this wave is done (if they stop it), but I don’t think they’re there yet. You’re also only talking about like 25% of the national population between NYC, FL, TX, AZ, LA County and New Orleans… And I don’t think there’s any indication that 20% herd immunity is going to lead to a significant reduction in R0.
You’re basically just guessing at a number here, and this percentage shifts over time. Shittier replacement for CARES Act increases it by forcing people back to work, back to job hunting, into homelessness, etc…
Precautions ebb and flow as well.
This is true but I think it’s important to note that:
-
That may not do enough to the R0 to save us.
-
20% is probably a low estimate of susceptible population once you factor in household transmission. If one spouse works from home and the other doesn’t, that entire household is at similar risk to the susceptible worker.
-
We have a lot of fall/winter risk of super spreader events; Labor Day parties, back to school, gatherings for fall sports, Halloween, Election Day, Thanksgiving, Christmas/Chanukah/other winter holidays, New Year’s Eve.
No, but we haven’t had time for that yet either. Very few places in USA#1 have had a winter opportunity to fuck this all up.
As it turns out, “mah guns pertects mah freedumbs from tie-ruh-knee” works pretty good. It’s just that tyranny = whatever they don’t like, regardless of actual tyranny. And it’s not like they’re going to war with the feds, they just had to kill a few minimum wage retail greeters to terrorize us enough to give them their way.
In other words, all the violence directed at greeters and security guards was domestic terror.
This is a feature, not a bug. The GOP wants a testing backlog, and they’ve managed to impose one even on blue states. Less quarantining means more spread, which means closer to herd immunity. Don’t be surprised if more stories come out soon of the Trump Administration actively directing resources away from blue states or cities. We learned today that he only cares about stopping it in red states. (As if we didn’t know already.) The next step is trying to let it rip in blue cities - we’ve seen GOP governors block mask mandates in their own cities.
I wouldn’t be surprised if the history books in the free part of the world talk about how Trump used the pandemic as a sort of “soft” partisan genocide.
It depresses me to say this, but I’m extremely confident this one’s going down right next to the summer slowdown theory. I would love to be wrong, but I don’t see it.
The main key is not that 20% helps all that much, but having 90% of people masked up, sanitizing, hand washing, etc does go a long way to reducing R0. The problem is this seems to only be happening in liberal cities or places that have already been hit hard. So you still can import cases via travel, and we seem to be aiming for R0 of 0.9 instead of trying to really stomp it out.
So if winter makes it worse (it will), we’re going to see upward trends and then it’s a matter of controlling them.
The Ivory Tower class appears to be either stupid, delusional, or content to retreat to their luxury bunkers. Just a quick ballpark, you’re looking at 40-45% of renters being evicted. That’s a shit ton of people and doesn’t factor in homeowners getting foreclosed on. So you’re looking at literally tens of millions homeless or taken in by family/friends. Current homeless population is 500K.
They infected vaccinated monkeys who had antibodies.
My landlord currently refuses to accept less than a 27% rent increase even as they lose 1% occupancy PER DAY according to the doorman, and local competitors a block away charge half as much and struggle to get tenants. I’m wondering if they already know they’re getting bailed out on all their losses.
Or they’re just morons.
Odds Trump and Mitch put a federal express lane in and tear gas squatters?
I didn’t read it, but the really interesting implication of lower viral load = less severity is the possibility of infecting everyone with a really low viral load to trigger immunity if vaccines don’t work. The question is whether they can minimize long term effects.
Assuming everything is 50% worse than the widely accepted worst case scenario has been a pretty consistently accurate thing to do with the pandemic.
Only all of us realists.
NYC is closer to 1.5% currently. It was probably higher in the spring.
Better healthcare system, likely fewer old people getting it.
He’s not doing this.
You’re pretty close to doing the opposite at times, not always. But I think people cope differently. Some have to have the most optimistic outlook possible and keep hope alive. Others need to prepare for the absolute worst and not get their hopes up.
“Hope is a dangerous thing. Hope can drive a man insane. It’s got no use on the inside.”
- Red
It is insane to me that we stopped research and didn’t finish those. I think the SARS one was maybe two months away. I guess you need active cases to test.
I imagine the competition to emigrate to New Zealand will be significant. Most of the world will go for eradication. Countries that don’t will be pariahs, unable to travel without major quarantines.
But if we’re estimating based on antibody studies, which suzzer is, then you want total deaths not confirmed positive deaths.
This is why I get so pissed at my friends IRL about this stuff, which makes me an asshole in their eyes.
At first glance, yes, if that’s the date HHS took over.
They came down from a 38% increase to a 27% increase, I’m not paying it, they may try again. They have tons of people moving out. They’re probably around 30% vacant now.
Right, like most renters are paycheck to paycheck or have a couple month’s expenses saved. So when people are getting multiple months behind that’s never getting made up. They’re eventually getting evicted or the landlord takes the L.
It was closer to 1.5%. (deaths / population * .2)
Orange County, FL, was actually in a straight up perfectly linear downward trend for a few days. I haven’t checked since.
Unless they’re undercounting deaths through any number of methods, which I would bet they are if it could be proven.
Illinois is well on its way. Michigan appears to be headed that way. I don’t think there’s any argument Louisiana is hammered in a second wave right now.
Also NYC life mostly involves standing near people and yelling at the.
and interrupting the.
2 Likes
News from around the world for those who are interested.
Germany and Belgium warn against travel to parts of Spain
Germany and Belgium have joined the UK in warning against travelling to Spain. But unlike the UK, both countries only warned their citizens against travelling to a limited number of regions. Germany said non-essential travel to Aragon, Catalonia and Navarra are currently discouraged. Belgium warned that increased vigilance is required for those returning from Aragon and Catalonia.
Germany expresses ‘great concern’ about a spike in cases
Lothar Wieler, head of the Robert Koch Institute, said: “The latest developments in the number of Covid-19 cases are of great concern to me and all of us at the RKI.” He added: “We don’t know yet if this is the beginning of a second wave but of course it could be.”
WHO: ‘pandemic will be one big wave not seasonal’
WHO spokeswoman, Dr Margaret Harris, said we are still in the first “big wave” of the Covid-19 pandemic, which will be “one big wave”. She warned arned against complacency about transmission in the northern hemisphere summer, saying that this virus did not behave like influenza that tended to follow seasonal trends. She said: “People are still thinking about seasons. What we all need to get our heads around is this is a new virus and…this one is behaving differently.”
Daily deaths in Iran hit record level
Deaths from coronavirus in Iran have hit a daily record of 235 over the past 24 hours, according to official health ministry figures released. The Islamic Republic is the country hardest hit by the coronavirus pandemic in the Middle East with 16,147 deaths.
New cases in south-east Asia continue to rise
The Philippine health ministry confirmed 1,678 new coronavirus infections, reporting more than 1,000 new daily cases for a 14th successive day. Hong Kong has reported 106 new coronavirus case, including 98 that were locally transmitted. It is the seventh day in row that Hong Kong has reported a triple digit rise in cases. Indonesia reported 1,748 new coronavirus infections, bringing its tally to 102,051 confirmed cases overall. Taiwan is investigating its first possible local coronavirus infection in more than a month. Vietnam has suspended all flights to and from the city of Danang for 15 days due to the new outbreak in the city
Covid-19 outbreak in Xinjiang prompts fears of spread inside China’s camps
Rising numbers of cases in the Xinjiang region has sparked fears the outbreak could reach the secretive internment camps where China is believed to have detained more than a million Muslim minority people. Chinese health authorities reported 68 new cases of Covid-19, including 57 in the far western region of Xinjiang, bringing the area’s reported total to 235. The region is home to China’s program of mass incarceration of Uighur and other Turkic Muslims, which has drawn international condemnation and accusations that the detention, abuse, surveillance and restrictions on religious and cultural beliefs amount to cultural genocide.
Bolsonaro: ‘I didn’t have any problems’
Brazil President Jair Bolsonaro took off his mask in public as he greeted supporters in Brasilia, days after saying he had recovered from the coronavirus, which he said had not had a serious impact on his health The right-wing leader tested positive for the coronavirus earlier this month and went into quarantine at his residence, but said on Saturday that his latest test had come back negative. “I didn’t have any problems,” Bolsonaro said on Monday.
Spain’s PM said the UK quarantine decision not justified
Britain’s decision to impose a two-week quarantine on people travelling from Spain is unfair, Pedro Sánchez said. He added that the Spanish government is in touch with British authorities in a bid to get the country to reconsider its position.
5 Likes
Shit!
So there’s a <10% chance Bolsenaro actually had covid right?
2 Likes
Either that, a light viral load or just a downright lie to boost his macho image I guess.