Feels like there should be laxities on booster spacing given Omicron — was rejected from 2 clinics as my second dose was in July, making it 5 months, not 6 months. By the time I can boost seems the worst of this wave will be over.
As far as I'm aware, the observed decreased severity of omicron conditional on infection (which is undeniable at this point) is still consistent with being the effect of prior immunity only. It might be innately less severe as well, but this still seems up for grabs. Even in NSW where it looks like the comparison with delta is like-for-like there's a selection bias: among omicron cases, a larger fraction are vaccinated than delta, due to omicron's immune escape. With this effect alone an omicron wave will have fewer severe outcomes per case compared to delta in a highly immune population, but only because in the same population the equivalent delta infections were prevented outright by immunity. So anyone unvaccinated and who hasn't been infected before I think should be pretty concerned still.
The evidence about omicron reproducing more poorly in lungs is good, obviously, but I'm under the impression that none of the analyses of hospitalisation data etc are yet showing a decrease in severity *when controlling for prior immunity*. This seems to me like an important distinction that has been lost in the messaging. I mean, actually I've seen the point made constantly, but it seems like there are enough bare headlines just saying it's mild that people have started to assume this has been shown in the innate sense as well. Even metaculus seems to have forgotten this point:
Hey! So, what we're seeing in Gauteng (and now other SA provinces and London as well) is a decrease in the absolute *number* of ICU admissions, not just the rate, since Omicron hit. That could theoretically be a result of the previous (Delta) wave being so widespread that it left a large fraction of the population with some acquired immunity. But the drop in absolute number of ICU admissions when Omicron arrived is so abrupt that the acquired-immunity explanation is unlikely to be sufficient to explain it. That means Omicron is likely to be less intrinsically severe; it's not just a denominator effect.
But yes, the unvaccinated should still be quite concerned. Even if it's intrinsically half as severe as Delta, that's still going to kill a lot of people.
Over two years, the total number of people who died OF Covid or WITH Covid is less than 1/4 of 1% of the population of the USA. That's little more than a rounding error against total USA population.
I understand it's a large effect, but so is omicron's immune escape!
The UK Health Security Agency just released Technical Briefing 33 [1] on variants of concern, I suspect this is the state of the art on the matter at the moment. From that we have (see from 23 onward or so) a bunch of numbers from various analyses showing a large reduction in hospitalisation risk even when controlling for vaccination. That makes sense, but can't distinguish between the selection effect and it being innately milder. The only figure trying to get at the innate reduction in severity gives:
"Including the likelihood of previous infection, in addition to vaccination in their model, they have estimated the intrinsic risk difference between Delta and Omicron as between 0 to 30%"
And the updated UKHSA Omicron Risk Assessment [2] says:
"Available data suggests that the observed reduction in risk in the UK is likely to be partly a reduction in intrinsic severity of the virus and partly to protection provided by prior infection. We cannot confidently quantify the relative contributions of these 2 factors at present."
If it was obvious enough that this was innate mildness, I'd expect these documents to be saying so. The closest we get is the estimate of a 0-30% reduction.
We ought to get some good data from New South Wales, since they don't have much prior infection, so looking only at unvaccinated cases should be a good proxy for no prior immunity, whereas undiagnosed infections skew things in the UK.
Edit: I've realised I've got some confusion about what is meant by "controlling for" various factors. Obviously the two groups in a study will have different proportions vaccinated or whatnot, and "controlling" for this (presumably) means measuring the effect of that variable on the outcome, and then constructing a hypothetical population (with vaccination status the average of the two groups...?) and answering what *that* group's reduction in severity is. But that's still not the innate severity - to get at that you need to construct a hypothetical population with *zero* prior immunity and give an answer for them. So controlling for immunity isn't enough, the authors need to explicitly say they're estimating the innate reduction. Which at least the NSW analysis you showed indeed appears to be trying to do.
Sure, but if you're comparing with previous waves then there are other large numerator effects, primarily that those waves created the immunity that is protecting people in this wave.
If it was that easy, you would think the studies comparing like for like to try and get at innate severity would have a definitive result by now.
That's right, if immunity to Delta is very widespread, it could be significantly lowering the death rate from Omicron. But this effect seems much too big to be mostly due to that.
Yep! This was what I was talking about with regards to the super-optimistic case in which Omicron actually ends the pandemic. It's waaaaay too early to talk about that, though! :-)
Not sure you'll still check this thread, but I would love someone to look into how the Paxlovid roll out is going.
Unfortunately I tested positive for SARS-CoV-2 today, I have 2 risk factors but am otherwise middle-aged, vaxxed+boosted, and in decent health.
I managed to get a script for Paxlovid by the end of the day, but there were a ton of hoops to jump through - 2 virtual visits, a trip to an ER for blood work to check kidney/liver levels, and lots of self advocacy. The prescribing Dr (at one of NYC's top hospitals) commented that he believed I was the first script that they've written in that hospital. They also made some side comments that their EMR system made the prescribing confusing and difficult. If I hadn't been a loyal "This Week in Virology" listener throughout the pandemic there's no way I would have gotten through this process.
What makes me worry Paxlovid is not being deployed widely is they warned me that while they would write the script, it would go to some central prioritization machine where I would likely get downgraded as I am young and vaccinated/boosted. Instead, the prescription was filled immediately and is getting delivered tomorrow.
"The first piece of good news is that vaccine boosters offer substantial protection against Omicron. Studies are showing that a third dose of mRNA vaccine (Pfizer or Moderna) restores a substantial amount of immunity."
One of the major problems with pro-vaxxer's is that they refuse to acknowledge that having a reduced chance of getting SERIOUSLY sick after taking your shots isn't "immunity".
Because of this misuse, people were led to believe that if they took the vax shots, they were going to get total protection from infection by the Covid virus.
Now with many still getting sick from Covid via breakthrough infections, they are coming to realize that they got snookered.
I don't believe that's an at all honest representation of the situation.
Two shots of the vaccine against the original variant protected you pretty strongly (~80-90%) from being infected. That's pretty close to what most people understand "immunity" to mean. That protection dropped a bit with Alpha and Delta (~70%) but was close enough that it was still a reasonable rule of thumb, especially since breakthrough cases were typically very mild.
Only with Omicron (~20-30%) have things shifted so much that it's no longer a good approximation. And Omicron's been a serious consideration for most people for ~two weeks. People didn't "get snookered". The reality changed! Dramatically! In less than a month!
Sorry, like so many, a lot of sloppy language in your post. I'm happy that you defined "pretty strongly, which saves me from asking how much that is. However, the Covax shots do not stop people from becoming infected. They REDUCE the possible effects that occur when someone becomes infected fo most people.
I would be willing to wager serious money that if you asked a statistical representative sample of people what "immunity" means, they would say that it PREVENTED something.
We have no data on how mild or serious breakthrough cases are because the CDC instructed back in May 2020 that they were only interested in hearing about such cases that took people to the hospital or resulted in death.
If you were to read outside your echo chamber, you would find ancedotal stories of plenty of CoVaxxed people who have had serious breakthrough infections BUT didn't wind up in the hospital. So they do not exist per the CDC. If you bury your head in the sand, as the CDC chooses to do with numerous data points (Politico has written at least a few articles on how bad data collection is in the USA and how the CDC is the cause of much of it) then you might think nothing is happening but that wouldn't be true.
What people got with the CoVax shots and associated boosters to date is some limited protection against a specific spike protein configuration that is wrapped around the Covid virus. When you get a heavily mutated variant like Omicron, the immune system training (antibodies) don't match the current configuration of the spike protein so are next to worthless.
Viruses mutate regularly. A leaky vaccine like the CoVax shots, which don't "sterilize" (kill) the virus helps it mutate in the vaccinated. The vaccinated may not feel as sick as someone suffering a full blown illness but that is a problem because it allows them to go about their work while being infectious and capable of further infecting others. Omicron did not appear in the last few weeks. Its likely been around for 6 months or more but finally got enough infections that someone in SA finally did the work necessary to check and discovered a new variant.
Prepare? I already GOT sick. The worst part wasn't the one day of feeling shit, it was being quarantined for the better part of the last 3 weeks.
Feels like there should be laxities on booster spacing given Omicron — was rejected from 2 clinics as my second dose was in July, making it 5 months, not 6 months. By the time I can boost seems the worst of this wave will be over.
Hi Noah,
As far as I'm aware, the observed decreased severity of omicron conditional on infection (which is undeniable at this point) is still consistent with being the effect of prior immunity only. It might be innately less severe as well, but this still seems up for grabs. Even in NSW where it looks like the comparison with delta is like-for-like there's a selection bias: among omicron cases, a larger fraction are vaccinated than delta, due to omicron's immune escape. With this effect alone an omicron wave will have fewer severe outcomes per case compared to delta in a highly immune population, but only because in the same population the equivalent delta infections were prevented outright by immunity. So anyone unvaccinated and who hasn't been infected before I think should be pretty concerned still.
The evidence about omicron reproducing more poorly in lungs is good, obviously, but I'm under the impression that none of the analyses of hospitalisation data etc are yet showing a decrease in severity *when controlling for prior immunity*. This seems to me like an important distinction that has been lost in the messaging. I mean, actually I've seen the point made constantly, but it seems like there are enough bare headlines just saying it's mild that people have started to assume this has been shown in the innate sense as well. Even metaculus seems to have forgotten this point:
https://www.metaculus.com/questions/8766/omicron-variant-less-deadly-than-delta/
(Though disagreeing with the metaculus community *and* Noah Smith definitely makes me reconsider if I'm missing something!)
Hey! So, what we're seeing in Gauteng (and now other SA provinces and London as well) is a decrease in the absolute *number* of ICU admissions, not just the rate, since Omicron hit. That could theoretically be a result of the previous (Delta) wave being so widespread that it left a large fraction of the population with some acquired immunity. But the drop in absolute number of ICU admissions when Omicron arrived is so abrupt that the acquired-immunity explanation is unlikely to be sufficient to explain it. That means Omicron is likely to be less intrinsically severe; it's not just a denominator effect.
But yes, the unvaccinated should still be quite concerned. Even if it's intrinsically half as severe as Delta, that's still going to kill a lot of people.
Define "a lot" please.
Over two years, the total number of people who died OF Covid or WITH Covid is less than 1/4 of 1% of the population of the USA. That's little more than a rounding error against total USA population.
I understand it's a large effect, but so is omicron's immune escape!
The UK Health Security Agency just released Technical Briefing 33 [1] on variants of concern, I suspect this is the state of the art on the matter at the moment. From that we have (see from 23 onward or so) a bunch of numbers from various analyses showing a large reduction in hospitalisation risk even when controlling for vaccination. That makes sense, but can't distinguish between the selection effect and it being innately milder. The only figure trying to get at the innate reduction in severity gives:
"Including the likelihood of previous infection, in addition to vaccination in their model, they have estimated the intrinsic risk difference between Delta and Omicron as between 0 to 30%"
And the updated UKHSA Omicron Risk Assessment [2] says:
"Available data suggests that the observed reduction in risk in the UK is likely to be partly a reduction in intrinsic severity of the virus and partly to protection provided by prior infection. We cannot confidently quantify the relative contributions of these 2 factors at present."
If it was obvious enough that this was innate mildness, I'd expect these documents to be saying so. The closest we get is the estimate of a 0-30% reduction.
We ought to get some good data from New South Wales, since they don't have much prior infection, so looking only at unvaccinated cases should be a good proxy for no prior immunity, whereas undiagnosed infections skew things in the UK.
Edit: I've realised I've got some confusion about what is meant by "controlling for" various factors. Obviously the two groups in a study will have different proportions vaccinated or whatnot, and "controlling" for this (presumably) means measuring the effect of that variable on the outcome, and then constructing a hypothetical population (with vaccination status the average of the two groups...?) and answering what *that* group's reduction in severity is. But that's still not the innate severity - to get at that you need to construct a hypothetical population with *zero* prior immunity and give an answer for them. So controlling for immunity isn't enough, the authors need to explicitly say they're estimating the innate reduction. Which at least the NSW analysis you showed indeed appears to be trying to do.
[1] https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/1043680/technical-briefing-33.pdf
[2] https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/1043756/22-december-2021-risk-assessment-for-SARS_Omicron_VOC-21NOV-01_B.1.1.529.pdf
Again, remember to look at *total* severe cases and deaths, not just *rates*. If the numerator falls, then it can't just be a denominator effect.
Sure, but if you're comparing with previous waves then there are other large numerator effects, primarily that those waves created the immunity that is protecting people in this wave.
If it was that easy, you would think the studies comparing like for like to try and get at innate severity would have a definitive result by now.
That's right, if immunity to Delta is very widespread, it could be significantly lowering the death rate from Omicron. But this effect seems much too big to be mostly due to that.
Also, very similar story happening in San Francisco, where Delta did not affect a huge % of the population: https://twitter.com/Bob_Wachter/status/1474514977650196480
20% of South Africa has HIV - are they capable of prior immunity? (however we don’t know how many of those are hospitalized)
There’s also this in the good news category:
https://pbs.twimg.com/media/FHV35I8UYAIURzh?format=jpg&name=large
Hey Noah, did you see this?
https://www.deseret.com/platform/amp/coronavirus/2021/12/21/22848453/fully-vaccinated-people-omicron-variant-super-immunity
Yep! This was what I was talking about with regards to the super-optimistic case in which Omicron actually ends the pandemic. It's waaaaay too early to talk about that, though! :-)
Hey Noah,
Not sure you'll still check this thread, but I would love someone to look into how the Paxlovid roll out is going.
Unfortunately I tested positive for SARS-CoV-2 today, I have 2 risk factors but am otherwise middle-aged, vaxxed+boosted, and in decent health.
I managed to get a script for Paxlovid by the end of the day, but there were a ton of hoops to jump through - 2 virtual visits, a trip to an ER for blood work to check kidney/liver levels, and lots of self advocacy. The prescribing Dr (at one of NYC's top hospitals) commented that he believed I was the first script that they've written in that hospital. They also made some side comments that their EMR system made the prescribing confusing and difficult. If I hadn't been a loyal "This Week in Virology" listener throughout the pandemic there's no way I would have gotten through this process.
What makes me worry Paxlovid is not being deployed widely is they warned me that while they would write the script, it would go to some central prioritization machine where I would likely get downgraded as I am young and vaccinated/boosted. Instead, the prescription was filled immediately and is getting delivered tomorrow.
Are we potentially making the same mistake (https://noahpinion.substack.com/p/the-vaccine-war) we made with overly restrictive vaccine requirements again?
"The first piece of good news is that vaccine boosters offer substantial protection against Omicron. Studies are showing that a third dose of mRNA vaccine (Pfizer or Moderna) restores a substantial amount of immunity."
One of the major problems with pro-vaxxer's is that they refuse to acknowledge that having a reduced chance of getting SERIOUSLY sick after taking your shots isn't "immunity".
Because of this misuse, people were led to believe that if they took the vax shots, they were going to get total protection from infection by the Covid virus.
Now with many still getting sick from Covid via breakthrough infections, they are coming to realize that they got snookered.
I don't believe that's an at all honest representation of the situation.
Two shots of the vaccine against the original variant protected you pretty strongly (~80-90%) from being infected. That's pretty close to what most people understand "immunity" to mean. That protection dropped a bit with Alpha and Delta (~70%) but was close enough that it was still a reasonable rule of thumb, especially since breakthrough cases were typically very mild.
Only with Omicron (~20-30%) have things shifted so much that it's no longer a good approximation. And Omicron's been a serious consideration for most people for ~two weeks. People didn't "get snookered". The reality changed! Dramatically! In less than a month!
Sorry, like so many, a lot of sloppy language in your post. I'm happy that you defined "pretty strongly, which saves me from asking how much that is. However, the Covax shots do not stop people from becoming infected. They REDUCE the possible effects that occur when someone becomes infected fo most people.
I would be willing to wager serious money that if you asked a statistical representative sample of people what "immunity" means, they would say that it PREVENTED something.
We have no data on how mild or serious breakthrough cases are because the CDC instructed back in May 2020 that they were only interested in hearing about such cases that took people to the hospital or resulted in death.
If you were to read outside your echo chamber, you would find ancedotal stories of plenty of CoVaxxed people who have had serious breakthrough infections BUT didn't wind up in the hospital. So they do not exist per the CDC. If you bury your head in the sand, as the CDC chooses to do with numerous data points (Politico has written at least a few articles on how bad data collection is in the USA and how the CDC is the cause of much of it) then you might think nothing is happening but that wouldn't be true.
What people got with the CoVax shots and associated boosters to date is some limited protection against a specific spike protein configuration that is wrapped around the Covid virus. When you get a heavily mutated variant like Omicron, the immune system training (antibodies) don't match the current configuration of the spike protein so are next to worthless.
Viruses mutate regularly. A leaky vaccine like the CoVax shots, which don't "sterilize" (kill) the virus helps it mutate in the vaccinated. The vaccinated may not feel as sick as someone suffering a full blown illness but that is a problem because it allows them to go about their work while being infectious and capable of further infecting others. Omicron did not appear in the last few weeks. Its likely been around for 6 months or more but finally got enough infections that someone in SA finally did the work necessary to check and discovered a new variant.
I appreciate the case you make for medium / long-run optimism, but what do you think of this? https://www.washingtonpost.com/outlook/2021/12/24/merck-molnupiravir-antiviral-covid-fda/
Any thoughts on long covid with Omicron?