Register a SA Forums Account here!
JOINING THE SA FORUMS WILL REMOVE THIS BIG AD, THE ANNOYING UNDERLINED ADS, AND STUPID INTERSTITIAL ADS!!!

You can: log in, read the tech support FAQ, or request your lost password. This dumb message (and those ads) will appear on every screen until you register! Get rid of this crap by registering your own SA Forums Account and joining roughly 150,000 Goons, for the one-time price of $9.95! We charge money because it costs us money per month for bills, and since we don't believe in showing ads to our users, we try to make the money back through forum registrations.
Which horse film is your favorite?
This poll is closed.
Black Beauty 2 1.06%
A Talking Pony!?! 4 2.13%
Mr. Hands 2x Apple Flavor 117 62.23%
War Horse 11 5.85%
Mr. Hands 54 28.72%
Total: 188 votes
[Edit Poll (moderators only)]

 
  • Post
  • Reply
Slow News Day
Jul 4, 2007

The problem with the Israeli study about waning immunity is one of looking at statistics in a vacuum, without intuition from the real world.

The study examined people who had been vaccinated in the winter and those who had been vaccinated in the spring. It found that the former group was more likely to contact the virus in the summer than the latter group.

This of course made scary headlines around the world, but here's the thing: the study's findings by themselves do not offer conclusive proof of waning immunity. In order for that to be the case, the two groups that were compared would need to have been sufficiently similar to each other (so as to eliminate potential confounding factors). But they are different. The former group, for example, consists of people who are more affluent and educated, and also happened to be among the first that were exposed to the Delta variant, possibly because they were more likely to travel for work, leisure, etc. So their higher infection rate may have stemmed from the new risks they were taking, rather than any change in vaccine protection.

Statisticians call this the Simpson's Paradox: when topline statistics (i.e. "winter group appears to be getting infected more frequently than the spring group") point to a false conclusion (i.e. "vaccine immunity wanes after 6-8 months") that disappears when you examine subgroups. This paradox would also explain some of the US data that the CDC and the FDA have used to justify booster shots. A lot of people resumed indoor activities this spring after getting vaccinated, and the fact that more were catching Covid may simply be explained by their increased exposure, combined with the arrival of Delta.

The other thing that should make us question the waning immunity hypothesis is that it is directly and strongly contradicted by other data. For instance, the ratio of positive Covid tests among old adults (who got vaccinated early on) and young kids (who haven't yet been vaccinated) has remained stable. Dowdy is an infectious disease epidemiologist at Johns Hopkins, and he said:

https://twitter.com/davidwdowdy/status/1431113092583370755

If waning immunity was a major issue, then we would see a faster rise in Covid cases among older people, but that doesn't appear to be the case, at least at this juncture.

Adbot
ADBOT LOVES YOU

haveblue
Aug 15, 2005



Toilet Rascal

Gotta admit, as a solution to the problem of unvaxxed covid patients taking treatment resources away from non-covid care, this isn't half bad

Solkanar512
Dec 28, 2006

by the sex ghost
So what the gently caress are we supposed to do about these massive anti-vax groups that are pulling their family members out of ICUs and brain poisoning as many as they can without anyone on Twitter or Facebook stopping them?

https://www.wsj.com/podcasts/google...e7-c9dc558da5dd

There’s a full transcript with this as well.

Main Paineframe
Oct 27, 2010

Solkanar512 posted:

So what the gently caress are we supposed to do about these massive anti-vax groups that are pulling their family members out of ICUs and brain poisoning as many as they can without anyone on Twitter or Facebook stopping them?

https://www.wsj.com/podcasts/google...e7-c9dc558da5dd

There’s a full transcript with this as well.

It's too loving late to do anything about it now. Like COVID itself, the best approach was to head it off early, because giving it time to spread would make it much more difficult to rein in. And it's not like we can hang all our hopes on someone inventing a vaccine against bad ideas.

Much like with COVID, the only hope is interrupting transmission to stop the problem from spreading further. But your link is pretty clear about the fact that social media companies aren't going to intervene in ways that might hurt the underlying metrics they use to sell ads.

quote:

Zoe Thomas: In some of the other Facebook file stories that we've been talking about, the fact that Facebook focuses so much on connections and engagements seems to be the cause of some of its problems. Was that the case here too?

Sam Schechner: When you read these documents, you see that with any intervention that they're making, they have to do a sort of impact assessment about what kind of impact is this going to have on some of the core engagement metrics that they're interested in, like MSI, meaningful social interaction. And so you see that in some of these documents, it's like, "Well, this will reduce health misinfo by 6.7% and it doesn't have much top-line impact in MSI, so that's good. But if it does reduce that, then there are issues." And so they're constantly trying to balance a bunch of different priorities. They want to create legitimacy for the platform among users, they want to make sure that engagement is rising. There's also a lot of attention in these documents towards avoiding potential public relations crises.

Zoe Thomas: Sam, here's what I'm struggling with. Mark Zuckerberg is the CEO of Facebook and he has this goal of helping 50 million people get a step closer to getting a COVID-19 vaccine, and yet it seems like he's limited in what he's been able to do.

Sam Schechner: I think that gets to the core of the issues facing Facebook. On one hand, it's a social media company so its content is not written in-house, it's not like the Wall Street Journal, he's not the Editor-in-chief, he can't tell people what to write. But beyond that, the system itself has certain incentives built into it that can help, as we've seen in some of the other reporting, give a benefit to polarizing material. There's certain things built into the system that actually maybe make it harder to reign this stuff in, and rather than tuning those down, and they did for some health stuff, but rather than turning it off altogether, they instead invest on trying to figure out, "Okay. Well, we'll target these individuals or these types of posts and try to clean it up on the other end." And so that's potentially a harder thing to do. In the end, it's a very good question because we don't know the answer. Why is it that Facebook has seemingly repeatedly been caught playing catch up on problems that its own internal research has foreseen? Is that a problem that gets solved through regulation? Is that a problem that gets solved through change in philosophy? Is it something that requires a change of leadership? Only time will tell.

Discendo Vox
Mar 21, 2013

We don't need to have that dialogue because it's obvious, trivial, and has already been had a thousand times.
I'll resume work on the antivax book project, which can inform this issue. There is also a bill likely to be introduced either late this year or sometime next year that can indirectly massively enervate the antivax movement. I'll speak on it in greater detail (and advocate for goons to contact their members of congress about it) when it goes fully public.

HelloSailorSign
Jan 27, 2011

whiskey patrol posted:

Do you have any studies or anything to point me to for this, especially the forward looking protection?


Fritz the Horse posted:

HelloSailorSign has a better understanding of immunology than I do, but I believe they're referring to how memory B cells work. When you get a covid infection or the vaccine, your body creates memory B cells that remember how to make antibodies to SARS-CoV-2. Over time your circulating antibodies will wane so you're more likely to be reinfected, but your memory cells can crank out fresh antibodies so you'll still mostly be prevented from severe disease. However, memory B cells don't just store antibodies from infection/vaccination, they also produce a bunch of slight variations on the original antibodies. In that way your immune system anticipates viral mutation.

Caveat: immunology is complex wizardry but that's my simplified understanding.


Fritz the Horse, who is occasionally harvested for immunoglobulin therapy :v:, got it right - as I understand it, and while I know more about immunology I am far from an expert.

This study https://www.medrxiv.org/content/10.1101/2021.08.24.21262415v1 went into the infection acquired vs. vaccine acquired. Now, I don't agree with their end conclusion that infection acquired is better, I think they've got some dataset/demographic issues that could explain the difference, but they still demonstrate evidence that infection acquired immunity is at least robust against future issues.

Look up things on somatic hypermutation.

buglord
Jul 31, 2010

Cheating at a raffle? I sentence you to 1 year in jail! No! Two years! Three! Four! Five years! Ah! Ah! Ah! Ah!

Buglord

Discendo Vox posted:

I'll resume work on the antivax book project, which can inform this issue. There is also a bill likely to be introduced either late this year or sometime next year that can indirectly massively enervate the antivax movement. I'll speak on it in greater detail (and advocate for goons to contact their members of congress about it) when it goes fully public.

What would this be? Or is that part of the book?

e: Also FritzHorse, do you live on an indian reservation? I saw you made a post about it and wanted to know more about your experience there if you do. I don't really hear a whole lot about native communities in the US and how they handled it, alls I know is that there's a lot of common comorbidities such as diabetes/alcoholism/obesity.

buglord fucked around with this message at 20:23 on Sep 24, 2021

mawarannahr
May 21, 2019

Mr. Pardiggle posted:

e: Also FritzHorse, do you live on an indian reservation? I saw you made a post about it and wanted to know more about your experience there if you do. I don't really hear a whole lot about native communities in the US and how they handled it, alls I know is that there's a lot of common comorbidities such as diabetes/alcoholism/obesity.

Yes please, I appreciated the post and would read more. Most of my awareness had to do with Ireland raising money for tribes in 2020 and that certain tribes had reached 100% vax.

Professor Beetus
Apr 12, 2007

They can fight us
But they'll never Beetus

mawarannahr posted:

I know the supply shortage is bad but we can’t even make IK stars in this country anymore?

Today I couldn't get vanilla at my local grocery store; poo poo's bad.

mawarannahr posted:

Yes please, I appreciated the post and would read more. Most of my awareness had to do with Ireland raising money for tribes in 2020 and that certain tribes had reached 100% vax.

Also this. Really happy to have other perspectives and something I was hoping to see more in the rebooted thread. Big thanks to Fritz for continuing to participate.

Fritz the Horse
Dec 26, 2019

... of course!

Mr. Pardiggle posted:

e: Also FritzHorse, do you live on an indian reservation? I saw you made a post about it and wanted to know more about your experience there if you do. I don't really hear a whole lot about native communities in the US and how they handled it, alls I know is that there's a lot of common comorbidities such as diabetes/alcoholism/obesity.

mawarannahr posted:

Yes please, I appreciated the post and would read more. Most of my awareness had to do with Ireland raising money for tribes in 2020 and that certain tribes had reached 100% vax.

Yes I've lived here most of my life and worked here for the last ~5 years. I posted some in this thread about the pandemic here I guess I can try to summarize our experience over the last year and a half.

There are a high rate of comorbidities, yes. Diabetes, heart disease, obesity, smoking, substance abuse. A major concern was also that many/most of the households here are overcrowded and multi-generational. Most people live in government housing and there's a severe shortage. Multi-generational households are traditional but also a result of overcrowding. So you often have young kids, parents, grandparents, maybe an aunt/uncle or cousin all under the same small roof. Access to healthcare is also very limited. I think we have a handful of ICU beds available for a population in the tens of thousands? And it's 3-4 hours to the nearest regional hospital.

Keep in mind that every reservation and tribal nation is very distinct and the pandemic response to each is going to vary greatly. Navajo Nation (Dine) was hit very hard early on and then made news for mounting a fantastic vaccination campaign and getting to >90% vaxxed quite quickly.

Early on the regional tribes all instituted mask mandates, curfews, and for some time they also had checkpoints at the major entrances/exits. Tribal police at the checkpoints would ask vehicles their destination and request that they continue through without stopping. One reservation even restricted access entirely for a while, only allowing residents to enter the reservation. Meanwhile the GOP state governors took no precautions whatsoever. Kristi Noem in South Dakota even threatened to take some of the tribes to court over their restrictions. Her pandemic policy was "I trust local counties, towns, and school districts to make responsible choices." Not tribes though I guess, they're not allowed to enact mask mandates or restrict access.

The pandemic hit significantly but the local IHS system was not overwhelmed and we didn't have as horrific a situation as the Dine. There was a lot of trouble enforcing quarantines and contact tracing, a lot of people were not cooperative. But overall the tribal government and agencies did a good job.

Vaccine rollout was fast. I got shots in Feb-March. IHS has a poor reputation but they did a great job with mass vaccination clinics. Unfortunately, our vaccination rate is still quite low. Last numbers I saw were about 30% of eligible population fully vaccinated. Thankfully my workplace decided to require vaccinations for all employees back in July, and our facilities are restricted access. Public are not allowed in without proof of vaccination.

I've been asking community members and coworkers the last few days about why vaccination rate is low and I don't hear a lot about misinformation or access issues. It's more apathy, from what people tell me. The poverty and related issues here is so severe that people are a lot more focused on daily survival than finding time to get shots or even to follow/care about pandemic news. That and we weren't hit very hard so there isn't a sense of urgency. The tribe recently released $2,000 per capita (children included) CARES Act stimulus plus an additional $500 for every person who is fully vaccinated. I had several people tell me they think the vaccination incentive isn't working well because the additional $500 is not very enticing when everyone just got an automatic $2,000 which is a huge amount of money around here. Maybe in a few months the $500 incentive will be more attractive.

So I guess as things stand we're in okay shape in terms of healthcare capacity and vaccine availability, but vaccination rates are still low. Our case numbers have been going up significantly, Delta is starting to hit. Some of the local schools are back to in-person but everyone did online all year last year and it sounds like they're considering going back to online education if the situation continues to worsen.


Personally I thought misinformation on Facebook would be something people mentioned more but I didn't hear that from my informal poll. I see a lot of bullshit on Facebook, distrust of government because of smallpox blankets, I'm not going to take the vaccine because I heard bear root is a great cure, etc.

edit: I know some people would have liked the tribe to make more of the relief stimulus contingent on vaccination and that's a good idea in a vacuum. However the per capita annual income for reservations in our region is like $7,000-$10,000. It's pretty cruel to withhold stimulus equivalent to several months' income in that situation.

edit2: oh also, fwiw the preferred label is "tribal nation." I know I use "tribe" a lot because it's common everyday usage but in more formal contexts most prefer "tribal nation" since it reinforces that they're sovereign nations. Canada of course goes with First Nations.

Fritz the Horse fucked around with this message at 21:02 on Sep 24, 2021

Arc Hammer
Mar 4, 2013

Got any deathsticks?
Have there been any double vaccinated breakthrough cases that put someone into the ICU?

UCS Hellmaker
Mar 29, 2008
Toilet Rascal

Sidenote you cannot leave ama if you are on high oxygen levels of assistance, period. Because unless you have someone that brings in tanks and has them going, you physically cannot leave the hospital because you won't make it past the nurses desk at the levels that put you in the icu. And no doctor will prescribe you oxygen (which btw needs a prescription! It's a drug!) So that you can leave ama, especially because at the high rates they use they need full tanks every few hours. Oxygen concentrators only work to 6lpm which is not nearly enough for highflow nasal cannula and will suffocate you if you use a non rebreather.

Also other caveat, at least in Ohio a doctor can declare you unable to make rational decisions or pink slip you if it will lead to you dying from a medical condition. I've seen it a handful of times (mainly involving seizures) but a covid patient on high level o2 they definitely could do it because you suffer from hypoxic delirium as you become more and more unable to do anything without desatting harshly.

So yeah, by the time these people are in icu they aren't able to actually leave. Not unless they leave to the morgue or somehow not end up dead after weeks.

Mr Luxury Yacht
Apr 16, 2012


Arc Hammer posted:

Have there been any double vaccinated breakthrough cases that put someone into the ICU?

Yes but they're super rare by comparison and concentrated in the older population. Up here in Ontario there's only been a single one under 50 (and they didn't die).

Arc Hammer
Mar 4, 2013

Got any deathsticks?

Mr Luxury Yacht posted:

Yes but they're super rare by comparison and concentrated in the older population. Up here in Ontario there's only been a single one under 50 (and they didn't die).

That's reassuring.

Cases projected to spike in the province after the holidays being announced the same day that capacity limits on sporting events are being raised in Ontario... not so much.

Mr Luxury Yacht
Apr 16, 2012


Arc Hammer posted:

That's reassuring.

Cases projected to spike in the province after the holidays being announced the same day that capacity limits on sporting events are being raised in Ontario... not so much.

Yeah things have weirdly stabilized here. The daily cases never cracked 1000 in this wave and the average has been slowly dropping despite schools being back for three weeks and no other additional NPIs other than vaccine passports that only came into effect two days ago. Same with hospitalization and ICU numbers. If you had told me that would be the state of things now a month ago I would have never believed you. They were projecting like, overwhelmed ICUs by now. We do have a much higher vaccination rate than a lot of places though.

Why they're possibly jeopardizing that over the loving Jays and Leafs is beyond me.

Epinephrine
Nov 7, 2008
The thread has moved on from the below study, fortunately, but I'm going to make note of it now just in case someone brings this up again in the future.

Discendo Vox posted:

https://www.authorea.com/users/421653/articles/527590-ffp3-respirators-protect-healthcare-workers-against-infection-with-sars-cov-2

The above appears to be the "study" referenced in the article, which as far as I can tell was never put through peer review (but somehow got an opinion piece in bmj). It looks like all involved just ran with the promotional material Cambridge put out.
This study has some problems and I hope this doesn't pass peer review without at least some major revisions. The authors at least acknowledge some the problems in their discussion, to the extent that they clearly say that the strongest statement you can make about the data is that it warrants further study (this is the least-nice thing you can say about a set of findings while still saying something nice).

Here we have a quasi-experimental study of two wards in the same hospital. Pre-intervention, the red ward was way worse off than the green ward. Post-intervention, both wards have a level of extra infections statistically indistinguishable from zero. There are several "buts" here though; here's the biggest:

So do you remember what happened with the "broken windows" policy? It got attributed by many as ending the crime wave in NYC, but in fact the crime wave was already on its way out by the time Giuliani was elected. The broken windows policy wound up taking credit for a change that was already happening independently of and before the policy was enacted.

The same thing is happening in this study. Almost the entirety of the data in this study claiming a problem in the red ward rests on a spike in cases in weeks 5 and 6. The spike ended and cases returned to baseline (see Figure 3) two weeks before the FPP3 protocol was enacted. In other words, we're looking at a kind of reversion to the mean problem here. The return to baseline could easily be due to the hospital or staff independently taking other measures in response to the spike. The authors, to their credit, acknowledge that they haven't considered confounds such as other changes in red ward behaviors. Not to their credit, this is a really basic thing that needs to be considered.

And then there's the statistical treatment of the green ward. The study assumes no change in the green ward pre- and post-intervention. This is bad statistics because, among other things, it assumes there's no possible source of shared variance between the green and red wards (and its not hard to see in there figures that there is shared variance). If things get better, generally, in the hospital, it would wind up being taken up by the pre-post effect in the red ward stats, even though a better explanation would at least include a general improvement. I highlight this specifically, because it's very clear in figure 3 that (although the error bars are at ceiling) the proportion of cases that were community acquired (higher numbers here mean fewer extra infections) increased over time in the green ward and was effectively at 100% in the green ward post-intervention. This would also be consistent with, as before, people enacting more safety protocols prior to FPP3 introduction in red ward.

So, to go back to the earlier discussion of preprints: take them with a big grain of salt. I can see these issues because I have years of training that helps me know what to look for; most people don't have that. However, peer reviewers do! Absent field expertise (at which point you already have the skills to take a deep dive into the data) relying on the basic things like "has this passed peer review?" is the best policy.

PT6A
Jan 5, 2006

Public school teachers are callous dictators who won't lift a finger to stop children from peeing in my plane

Mr Luxury Yacht posted:

Yeah things have weirdly stabilized here. The daily cases never cracked 1000 in this wave and the average has been slowly dropping despite schools being back for three weeks (and passports only just coming into effect). Same with hospitalization and ICU numbers. We do have a much higher vaccination rate than a lot of places though.

Why they're possibly jeopardizing that over the loving Jays and Leafs is beyond me.

Because Alberta's done it with the Flames and Oilers and god knows there's no good reason to be suspicious of our excellent judgement (pay no attention to the overflowing ICUs).

Solkanar512
Dec 28, 2006

by the sex ghost

UCS Hellmaker posted:

Sidenote you cannot leave ama if you are on high oxygen levels of assistance, period. Because unless you have someone that brings in tanks and has them going, you physically cannot leave the hospital because you won't make it past the nurses desk at the levels that put you in the icu. And no doctor will prescribe you oxygen (which btw needs a prescription! It's a drug!) So that you can leave ama, especially because at the high rates they use they need full tanks every few hours. Oxygen concentrators only work to 6lpm which is not nearly enough for highflow nasal cannula and will suffocate you if you use a non rebreather.

Also other caveat, at least in Ohio a doctor can declare you unable to make rational decisions or pink slip you if it will lead to you dying from a medical condition. I've seen it a handful of times (mainly involving seizures) but a covid patient on high level o2 they definitely could do it because you suffer from hypoxic delirium as you become more and more unable to do anything without desatting harshly.

So yeah, by the time these people are in icu they aren't able to actually leave. Not unless they leave to the morgue or somehow not end up dead after weeks.

Thanks for clarifying here!

Professor Beetus
Apr 12, 2007

They can fight us
But they'll never Beetus

Epinephrine posted:

The thread has moved on from the below study, fortunately, but I'm going to make note of it now just in case someone brings this up again in the future.

This study has some problems and I hope this doesn't pass peer review without at least some major revisions. The authors at least acknowledge some the problems in their discussion, to the extent that they clearly say that the strongest statement you can make about the data is that it warrants further study (this is the least-nice thing you can say about a set of findings while still saying something nice).

Here we have a quasi-experimental study of two wards in the same hospital. Pre-intervention, the red ward was way worse off than the green ward. Post-intervention, both wards have a level of extra infections statistically indistinguishable from zero. There are several "buts" here though; here's the biggest:

So do you remember what happened with the "broken windows" policy? It got attributed by many as ending the crime wave in NYC, but in fact the crime wave was already on its way out by the time Giuliani was elected. The broken windows policy wound up taking credit for a change that was already happening independently of and before the policy was enacted.

The same thing is happening in this study. Almost the entirety of the data in this study claiming a problem in the red ward rests on a spike in cases in weeks 5 and 6. The spike ended and cases returned to baseline (see Figure 3) two weeks before the FPP3 protocol was enacted. In other words, we're looking at a kind of reversion to the mean problem here. The return to baseline could easily be due to the hospital or staff independently taking other measures in response to the spike. The authors, to their credit, acknowledge that they haven't considered confounds such as other changes in red ward behaviors. Not to their credit, this is a really basic thing that needs to be considered.

And then there's the statistical treatment of the green ward. The study assumes no change in the green ward pre- and post-intervention. This is bad statistics because, among other things, it assumes there's no possible source of shared variance between the green and red wards (and its not hard to see in there figures that there is shared variance). If things get better, generally, in the hospital, it would wind up being taken up by the pre-post effect in the red ward stats, even though a better explanation would at least include a general improvement. I highlight this specifically, because it's very clear in figure 3 that (although the error bars are at ceiling) the proportion of cases that were community acquired (higher numbers here mean fewer extra infections) increased over time in the green ward and was effectively at 100% in the green ward post-intervention. This would also be consistent with, as before, people enacting more safety protocols prior to FPP3 introduction in red ward.

So, to go back to the earlier discussion of preprints: take them with a big grain of salt. I can see these issues because I have years of training that helps me know what to look for; most people don't have that. However, peer reviewers do! Absent field expertise (at which point you already have the skills to take a deep dive into the data) relying on the basic things like "has this passed peer review?" is the best policy.

Thank you so so much for posting this kind of detailed analysis. This is the kind of post I love seeing in this thread, as a layperson, so thanks for jumping in.

enki42
Jun 11, 2001
#ATMLIVESMATTER

Put this Nazi-lover on ignore immediately!

Mr Luxury Yacht posted:

Why they're possibly jeopardizing that over the loving Jays and Leafs is beyond me.

Go to reddit if you want to see what the average Ontarian thinks about this. Not only is this totally no problem at all, it doesn't go far enough, since the second you have a vaccination passport all other restrictions are completely pointless apparently. And the cases have gone marginally down for a few days, so COVID is clearly over.

Platystemon
Feb 13, 2012

BREADS

Slow News Day posted:

https://twitter.com/davidwdowdy/status/1431113092583370755

If waning immunity was a major issue, then we would see a faster rise in Covid cases among older people, but that doesn't appear to be the case, at least at this juncture.

This ratio has a ton of confounders of its own. It’s weird to present it as good evidence immunity isn’t waning.

Inferior Third Season
Jan 15, 2005

UCS Hellmaker posted:

Also other caveat, at least in Ohio a doctor can declare you unable to make rational decisions or pink slip you if it will lead to you dying from a medical condition.
How does this work with things like DNRs and terminal patients deciding they just want to go home to live out their final days?

Mr Luxury Yacht
Apr 16, 2012


Inferior Third Season posted:

How does this work with things like DNRs and terminal patients deciding they just want to go home to live out their final days?

There's a reason you're heavily advised to fill out a DNR well in advance if you think you might want one.

Professor Beetus
Apr 12, 2007

They can fight us
But they'll never Beetus

Inferior Third Season posted:

How does this work with things like DNRs and terminal patients deciding they just want to go home to live out their final days?

I don't know how it works but I know my state has a compassionate euthanasia law, so I imagine this might look a little different from state to state.

Stickman
Feb 1, 2004

Slow News Day posted:

The problem with the Israeli study about waning immunity is one of looking at statistics in a vacuum, without intuition from the real world.

The study examined people who had been vaccinated in the winter and those who had been vaccinated in the spring. It found that the former group was more likely to contact the virus in the summer than the latter group.

This of course made scary headlines around the world, but here's the thing: the study's findings by themselves do not offer conclusive proof of waning immunity. In order for that to be the case, the two groups that were compared would need to have been sufficiently similar to each other (so as to eliminate potential confounding factors). But they are different. The former group, for example, consists of people who are more affluent and educated, and also happened to be among the first that were exposed to the Delta variant, possibly because they were more likely to travel for work, leisure, etc. So their higher infection rate may have stemmed from the new risks they were taking, rather than any change in vaccine protection.

Statisticians call this the Simpson's Paradox: when topline statistics (i.e. "winter group appears to be getting infected more frequently than the spring group") point to a false conclusion (i.e. "vaccine immunity wanes after 6-8 months") that disappears when you examine subgroups. This paradox would also explain some of the US data that the CDC and the FDA have used to justify booster shots. A lot of people resumed indoor activities this spring after getting vaccinated, and the fact that more were catching Covid may simply be explained by their increased exposure, combined with the arrival of Delta.

The other thing that should make us question the waning immunity hypothesis is that it is directly and strongly contradicted by other data. For instance, the ratio of positive Covid tests among old adults (who got vaccinated early on) and young kids (who haven't yet been vaccinated) has remained stable. Dowdy is an infectious disease epidemiologist at Johns Hopkins, and he said:

https://twitter.com/davidwdowdy/status/1431113092583370755

If waning immunity was a major issue, then we would see a faster rise in Covid cases among older people, but that doesn't appear to be the case, at least at this juncture.

I've talked about this before but Simpson's paradox isn't actually a problem with the Israeli study, or at least not to the degree that you're implying. Epidemiologists are well aware of the potential for confounding (the larger principle underlying Simpson's paradox; reference just for anyone who might not know the term) and design their sampling and analyses to address them as much as possible. Israel's waning study only looks at infections in the Delta-dominant period from July 11-31 to avoid differences between Delta and Alpha infection. The analysis is stratified by age category, and further adjusted for specific age, gender, COVID testing propensity, major demographic group, and week of infection.

That doesn't mean that the possibility of confounding is entirely eliminated, of course - that's an ever-present risk in observational research. For example, there may be behavioral differences between people in the same age/demographic/risk category who vaccinated early vs those who vaccinated later (though this would likely bias results towards an appearance of increasing effectiveness, since those with the greatest time since vaccination are also likely the most behaviorally risk-adverse). I would have liked to see adjustment for comorbidities, but that's more likely to affect serious disease analysis and probably not much of an issue for detected infection. Overall Israels' analysis does a decent job of addressing the primary known sources of confounding.

Dowdy's tweet is pretty old so it didn't have the benefit of the multiple corroborating studies that have been published/pre-printed since late August, but it's a prime example of exactly the thing you are talking about - it's dangerous to make assumptions about underlying trends from aggregate data because they may be obscured by confounding effects. Dowdy's trend could easily explained by increasing (relative) risk of exposure in young children as schools/preschools/etc reopened offsetting declining protection against infection in older adults. I haven't been able to find a nice graph of rates stratified by age in LA County or California so it's just a theory, but something is surely confounding the relative rates because Kaiser Permanente Southern California released an extensive analysis of SoCal vaccination effectiveness through August 8th and found evidence of significant decline in effectiveness vs infection (Lancet preprint page and non-paywalled manuscript):



Waning effectiveness has been found in every study that I'm aware of:

- Pfizer's phase 3 follow-up (randomized controlled trial). Vaccine efficacy vs infection declined from an average of 96% two weeks to two months post-vaccination, to an average of 84% 4 to 6 months post-vaccination. All of the follow-up was pre-Delta.

- UK prospective cohort study with monthly testing. Pfizer VE vs infection during the Delta wave declined from ~85% to ~73% over the course of three months post-full-vaccination. Effectiveness and longevity was better for those <35.



- Qatar (test-negative case control). Pfizer VE vs Delta infection (any or symptomatic) declined to ~50% after 4 months.

There are a couple more studies (like Mayo Clinic's Minnesota study) but I'd need to look through them again to remember if they adequately control for the potential effects of Delta.

Stickman fucked around with this message at 23:20 on Sep 24, 2021

UCS Hellmaker
Mar 29, 2008
Toilet Rascal

Inferior Third Season posted:

How does this work with things like DNRs and terminal patients deciding they just want to go home to live out their final days?

This needs to be done well in advance, have your DNR and medical power of attorney setup before you go into the AMS of delirium. Because at that point you realistically can not make sound decisions. It has to do with informed consent. And if you want to go hospice care which is what your asking, that's complicated and directly involved with hospice nurses and physicians. I haven't actually seen that in regards to our covid patients, and honestly haven't seen any go home because of the oxygen requirements or sedatives needed in order to help with the delirium. Realistically if we could we would send them home if they went hospice, but being off oxygen drops them sharply to 60% spo2 which is MASSIVELY bad, and runs into that whole problem that no one has the ability to have the required oxygen at home for the patient. Also since hospice largely means you aren't going to be on those high levels of oxygen, and instead be on a simple face mask or NC with morphine, Ativan, and other medications to help the patient feel comfortable as the body shuts down.

Also DNR works different depending on the hospital, we have
standard DNR, which is if I code do nothing
DNAR: do everything up till I code including tube me unless my heart stops
DNRCC: Do nothing if my heart or lungs are stopping except give me meds to help me go
DNI: do not intubate do everything else
DNE: do not escalate based on criteria decided before I was unable to specify ie no ICU care, no pressors, no CRRT, no dialysis and so on picked by what you decide.
CC: Do nothing that will prolong, focus on comfort solely, do not try to treat any condition unless it will effect comfort ie put a foley in because I cant pee. Give me meds for comfort let me die peacefully

UCS Hellmaker
Mar 29, 2008
Toilet Rascal
Another sidenote, Another thing families or patients like to do if they don't like their treatment or think they aren't getting cared for because (}insert rant) is to try and get transferred to another hospital or hospital system. Here's a funny thing, realistically you can't because if its not to a higher level of care that isn't available at the hospital your at, the patient eats the cost of everything for transport. And your at the whims of the transfer center there for getting a bed, which has one family found out repeatedly they don't give a gently caress what you want they fill their beds with the patients they have in there system first. So requesting transfer to another hospital like many of these antivaxers will try will realistically never happen and many times even attempting to can take days if not weeks right now, along side many IFT ambulances literally requiring upfront payments to transport because insurance flat out wont cover it. Seen that one before, and have been the IFT ambulance that was doing that transport before.

Phigs
Jan 23, 2019

Fritz the Horse posted:

That it's not peer reviewed means it should be taken with a grain of salt and it's not going to be used to shape official policy (by CDC or whoever). Why isn't the CDC using it as evidence to change their masking recommendations? Because it's a preprint, for one thing.

The Guardian article is publicity in the sense that they mostly interview the authors about their work.

We were talking about actual people on the street making decisions about their personal safety though and whether or not it was weird. The question was not "is this scientifically proven to be effective" it was "could a reasonable person come to the reasonable conclusion that the equipment is effective".

A lot of people died "waiting for more research" this pandemic. And very, very rarely would (or should) a person wait on peer reviewed literature before they make decisions about their life. I don't know anyone who waited for a peer reviewed paper on their potential spouse before deciding to spend the rest of their life with them for instance.

mawarannahr
May 21, 2019

Discendo Vox posted:

https://www.authorea.com/users/421653/articles/527590-ffp3-respirators-protect-healthcare-workers-against-infection-with-sars-cov-2

The above appears to be the "study" referenced in the article, which as far as I can tell was never put through peer review (but somehow got an opinion piece in bmj). It looks like all involved just ran with the promotional material Cambridge put out.


The Guardian article points that out and hedges all its claims

quote:

The research has not yet been peer-reviewed, but is being released early because of the urgent need to share information relating to the pandemic.

Discendo Vox posted:

You know that it's because the nominal research has not gone through peer review or publication, and was tied entirely to promotional coverage generated by the university. It's in my post. This has been discussed before. None of it's new. I'm not interested in doing the "make vox explain scientific literacy from first principles" schtick again.

You’re responding to a post about someone making a personal decision about wearing a mask based on what knowledge they have available in a situation where, as the article you totally read points out, there was and is a dearth of actionable information that can’t really wait for peer review. The hostility is bizarre.

Your hardline stance is relevant to scientific research in general and how it might be used for public policy but people are more complicated than that and nobody bases all their decisions on peer reviewed studies. Just as practically nobody reads EULAs.

mawarannahr fucked around with this message at 00:27 on Sep 25, 2021

fosborb
Dec 15, 2006



Chronic Good Poster

Fritz the Horse posted:

That it's not peer reviewed means it should be taken with a grain of salt and it's not going to be used to shape official policy (by CDC or whoever). Why isn't the CDC using it as evidence to change their masking recommendations? Because it's a preprint, for one thing.

Materials presented at the FDA advisory committee meeting last week referenced 2 preprints on vaccine effectiveness against delta

https://www.fda.gov/advisory-commit...ng-announcement

Materials presented to the CDC yesterday and the day before on evidence to recommendation framework for Pfizer boosters cited preprints 5 times.

https://www.cdc.gov/vaccines/acip/meetings/downloads/slides-2021-9-23/03-COVID-Oliver.pdf

Sir John Falstaff
Apr 13, 2010

Professor Beetus posted:

E: sorry for contributing to a bit of a derail, but I just don't think SA has made much of an impact wrt covid aside from some people staying in more and wearing different masks.

SA spawned 4chan, 4chan spawned QAnon, ergo . . .

lil poopendorfer
Nov 13, 2014

by the sex ghost

UCS Hellmaker posted:

Sidenote you cannot leave ama if you are on high oxygen levels of assistance, period. Because unless you have someone that brings in tanks and has them going, you physically cannot leave the hospital because you won't make it past the nurses desk at the levels that put you in the icu. And no doctor will prescribe you oxygen (which btw needs a prescription! It's a drug!) So that you can leave ama, especially because at the high rates they use they need full tanks every few hours. Oxygen concentrators only work to 6lpm which is not nearly enough for highflow nasal cannula and will suffocate you if you use a non rebreather.

Also other caveat, at least in Ohio a doctor can declare you unable to make rational decisions or pink slip you if it will lead to you dying from a medical condition. I've seen it a handful of times (mainly involving seizures) but a covid patient on high level o2 they definitely could do it because you suffer from hypoxic delirium as you become more and more unable to do anything without desatting harshly.

So yeah, by the time these people are in icu they aren't able to actually leave. Not unless they leave to the morgue or somehow not end up dead after weeks.

if the provider deems the individual to be of sound mind and able to understand the risks of discontinuing treatment, you can't really keep them against their will (not includig the five states that allow medical holds for non-psych reasons, as you mentioned). Attempting to do so can open you to liability for false imprisonment. It still happens, something like one out of five 5150 holds in California are for non-psych causes, but the providers don't have much to back them up if they were to be sued

The dilemma is: if you deem someone unable to consent to the risks of leaving ama, how can they consent to the treatments that will be provided during their medical hold. Its a little more complex than that, but that's the medicolegal point in favor of letting them go even if it will jeopardize their life.

Local standard of care, hospital policy, and the providers own principals will shape the decision but generally if a patient of sound mind and their surrogate/POA wanna go, we let them go. Never had it happen on a vent but hell, if it's just for airway pprotection and not oxygenation/ventilatory support, why can't they decide to leave. Anyways Here's a recent case where the guy was on hfnc and left only to die the next day :sad:

https://www.kevinmd.com/blog/2021/08/dying-after-leaving-ama.html

Platystemon
Feb 13, 2012

BREADS

Stickman posted:

Dowdy's tweet is pretty old so it didn't have the benefit of the multiple corroborating studies that have been published/pre-printed since late August, but it's a prime example of exactly the thing you are talking about - it's dangerous to make assumptions about underlying trends from aggregate data because they may be obscured by confounding effects. Dowdy's trend could easily explained by increasing (relative) risk of exposure in young children as schools/preschools/etc reopened offsetting declining protection against infection in older adults. I haven't been able to find a nice graph of rates stratified by age in LA County or California so it's just a theory, but something is surely confounding the relative rates because Kaiser Permanente Southern California released an extensive analysis of SoCal vaccination effectiveness through August 8th and found evidence of significant decline in effectiveness vs infection (Lancet preprint page and non-paywalled manuscript):

One of the potential confounders I was thinking of is that Delta’s household secondary attack rate is higher. If this shifts the ratio of in‐house:out‐of‐house transmission toward the in‐house side, we would expect four‐year‐olds to make up a disproportionate share of new cases. They don’t have jobs or solo errands that take them outside of the home, after all.

Stickman
Feb 1, 2004

Yeah, that makes sense too. Older folks in multi-generational homes would still be affected by increased secondary attack rates, but that just going a be a proportion compared to all kids 0-4. Hopefully they'll be some tracing analysis to break it down, but it's definitely clear that something is increasing the baseline infection risk of kids even more than the baseline risk of adults!

LionArcher
Mar 29, 2010


Just read the last dozen pages.

Bunch of dog whistling racist comments about China, as posters through up their hands pretending like we don’t live in a lovely police state here, but only about racist poo poo when police power is threatened instead of what would actual help people.

We could have stopped this. We still can too, but it involves actually aggressive measures, by an administration who has no interest in actually fixing it. and the liberals in D&D are mostly cowardly computer touchers who will defend Biden’s lovely lack of help.

This is the bad COVID thread, and I say that as someone who got dunked on in the good thread today for a dumb post about BMI (which is my own dumb fault and I was pissed and blowing off steam.)

mobby_6kl
Aug 9, 2009

by Fluffdaddy

Stickman posted:

I've talked about this before but Simpson's paradox isn't actually a problem with the Israeli study, or at least not to the degree that you're implying. Epidemiologists are well aware of the potential for confounding (the larger principle underlying Simpson's paradox; reference just for anyone who might not know the term) and design their sampling and analyses to address them as much as possible. Israel's waning study only looks at infections in the Delta-dominant period from July 11-31 to avoid differences between Delta and Alpha infection. The analysis is stratified by age category, and further adjusted for specific age, gender, COVID testing propensity, major demographic group, and week of infection.

That doesn't mean that the possibility of confounding is entirely eliminated, of course - that's an ever-present risk in observational research. For example, there may be behavioral differences between people in the same age/demographic/risk category who vaccinated early vs those who vaccinated later (though this would likely bias results towards an appearance of increasing effectiveness, since those with the greatest time since vaccination are also likely the most behaviorally risk-adverse). I would have liked to see adjustment for comorbidities, but that's more likely to affect serious disease analysis and probably not much of an issue for detected infection. Overall Israels' analysis does a decent job of addressing the primary known sources of confounding.

Dowdy's tweet is pretty old so it didn't have the benefit of the multiple corroborating studies that have been published/pre-printed since late August, but it's a prime example of exactly the thing you are talking about - it's dangerous to make assumptions about underlying trends from aggregate data because they may be obscured by confounding effects. Dowdy's trend could easily explained by increasing (relative) risk of exposure in young children as schools/preschools/etc reopened offsetting declining protection against infection in older adults. I haven't been able to find a nice graph of rates stratified by age in LA County or California so it's just a theory, but something is surely confounding the relative rates because Kaiser Permanente Southern California released an extensive analysis of SoCal vaccination effectiveness through August 8th and found evidence of significant decline in effectiveness vs infection (Lancet preprint page and non-paywalled manuscript):



Waning effectiveness has been found in every study that I'm aware of:

- Pfizer's phase 3 follow-up (randomized controlled trial). Vaccine efficacy vs infection declined from an average of 96% two weeks to two months post-vaccination, to an average of 84% 4 to 6 months post-vaccination. All of the follow-up was pre-Delta.

- UK prospective cohort study with monthly testing. Pfizer VE vs infection during the Delta wave declined from ~85% to ~73% over the course of three months post-full-vaccination. Effectiveness and longevity was better for those <35.



- Qatar (test-negative case control). Pfizer VE vs Delta infection (any or symptomatic) declined to ~50% after 4 months.

There are a couple more studies (like Mayo Clinic's Minnesota study) but I'd need to look through them again to remember if they adequately control for the potential effects of Delta.
I'm sure I'm not smarter than all these researchers, but how do these kinds of studies account for naturally increasing cumulative probability of encountering a spreader or significantly bigger viral load over time?

I'm unlikely to run into one in one month but eventually some unmasked infected rear end in a top hat will turn up.

enki42
Jun 11, 2001
#ATMLIVESMATTER

Put this Nazi-lover on ignore immediately!

LionArcher posted:

Bunch of dog whistling racist comments about China, as posters through up their hands pretending like we don’t live in a lovely police state here, but only about racist poo poo when police power is threatened instead of what would actual help people.

Is this in reference to enforcing lockdowns? I think it's just a reality of policing in North America that giving police broad powers to enforce lockdowns against individuals (which pretty much necessitates random stops of citizens and carding) is going to result in racialized people being disproportionately targeted, and those stops / carding used as an excuse for additional searches / harassment / brutalization. I know this because this is demonstrably what happened in North America when police had those powers, and normalizing them and making those powers more common will do nothing but amplify that effect.

This isn't using indigenous or black people as an excuse for why we can't do anything, it's a genuine concern that a wide swath of the population is thankfully against. Ontario tried this exact thing, and public response was strongly against, to the degree that it was rolled back almost immediately (not without tons evidence of black and indigenous folks being targeted in the 1 weekend it was in effect). I'd argue the opposite, that ignoring what effect broadly expanded police powers would have on groups that have been the targets of police brutality over and over again to have lockdowns more enforced is throwing those groups under the bus for your own comfort, especially given the relative risk when you're vaccinated and effective PPE is available to you.

Finally, can you point out where racist comments about China were made? I see a lot of comments about the authoritarian nature of their government and no comments as far as I can tall about Chinese ethnicity as a whole.

enki42 fucked around with this message at 08:32 on Sep 25, 2021

Platystemon
Feb 13, 2012

BREADS

mobby_6kl posted:

I'm sure I'm not smarter than all these researchers, but how do these kinds of studies account for naturally increasing cumulative probability of encountering a spreader or significantly bigger viral load over time?

I'm unlikely to run into one in one month but eventually some unmasked infected rear end in a top hat will turn up.

By comparison to the unvaccinated hordes. They face the same increasing probability.

Not that that doesn’t introduce its own problem—the unvaccinated are far from a perfect behavioral mirror of the vaccinated—but it’s the basis for comparison.

Charles 2 of Spain
Nov 7, 2017

There's also going to be another issue soon for studies in the US and UK where a lot of the unvaccinated have already been infected.

Adbot
ADBOT LOVES YOU

Phigs
Jan 23, 2019

enki42 posted:

Is this in reference to enforcing lockdowns? I think it's just a reality of policing in North America that giving police broad powers to enforce lockdowns against individuals (which pretty much necessitates random stops of citizens and carding) is going to result in racialized people being disproportionately targeted, and those stops / carding used as an excuse for additional searches / harassment / brutalization. I know this because this is demonstrably what happened in North America when police had those powers, and normalizing them and making those powers more common will do nothing but amplify that effect.

This isn't using indigenous or black people as an excuse for why we can't do anything, it's a genuine concern that a wide swath of the population is thankfully against. Ontario tried this exact thing, and public response was strongly against, to the degree that it was rolled back almost immediately (not without tons evidence of black and indigenous folks being targeted in the 1 weekend it was in effect). I'd argue the opposite, that ignoring what effect broadly expanded police powers would have on groups that have been the targets of police brutality over and over again to have lockdowns more enforced is throwing those groups under the bus for your own comfort, especially given the relative risk when you're vaccinated and effective PPE is available to you.

Finally, can you point out where racist comments about China were made? I see a lot of comments about the authoritarian nature of their government and no comments as far as I can tall about Chinese ethnicity as a whole.

Could always try a non-police enforcement mechanism. Create a temporary organization that works with minority communities to create enforcement units who are tasked with ensuring compliance in their own communities. Part of the problem of police is that they are the monolithic enforcement arm of the state and capital, of course just adding a new job to the tool used to oppress the masses is just going to cause problems. Another problem is they tend to police outside of their communities like an occupying force instead of being a community tool. You could solve both problems with a unit that recruits from the community with community support. They wouldn't be able to FORCE people at gunpoint like police do, but you can get a lot done just by being seen as a legitimate authority and being able to leverage the community to pressure people within it. Plus you can pay them for their time, utilizing some of the lost labor force due to closures and injecting money into the community. You could add community-wide incentives so if lockdown is obeyed everyone within that area gets a bonus payment, but only if lockdown is maintained, etc.

And that's just a random idea I had just now. There's plenty of room between SEND THE POLICE and ah gently caress it do whatever.

  • 1
  • 2
  • 3
  • 4
  • 5
  • Post
  • Reply