B.1.1.529 Omicron variant

[Woodchopper]

Latest HSA threat assessment for Omicron. https://assets.publishing.service.gov.u ... .1.529.pdf

Has ‘insufficient data’ for infection severity.

[Woodchopper]

Thread on vaccine effectiveness (VE) for Omicron: https://twitter.com/diseaseecology/stat ... 19776?s=21

Concludes:

Omicron's 30-40-fold reduction in neut. antibody titers will lead to ESTIMATED:
~10-14% reduction in VE HOSPITALIZATION
~40-60% reduction in VE SYMPTOMATIC INFECTION
Both of these can be partly mitigated by 3rd doses.

Seems consistent with what has been seen in South Africa with very high transmission but less severe illness than with delta (so long as we assume that there were very high levels of prior infection among the unvaccinated population and infection offers similar protection).

Omicron will be worse in any place where there are still large numbers of people who don’t have antibodies (perhaps Germany). The transmission advantage and lack of NPIs compared to the last two years will mean that it’ll probably infect them.

[Woodchopper]

My spreadsheet is obviously way off if the doubling time is 2-3 days. That's unbelievably fast.

And Plan B is obviously hilariously insufficient if the doubling time is 2-3 days. Nothing like what would be enough.

Haven’t had time to look things up. But as far as I remember 2-3 days is what has been reported in several countries. Though these are for fuzzy data and very small numbers of cases. South Africa may be slowing down a little by now. Cases can’t keep growing like that as people react and change their behaviour.

For some British data see here: https://assets.publishing.service.gov.u ... update.pdf

S gene dropout prevalence has quadrupled in a week.

Though again, it’s still small numbers and fuzzy data. But the shape of the curve looks similar to Gauteng cases.

See also this thread: https://twitter.com/jcbarret/status/146 ... 15651?s=21

[Woodchopper]

SAGE 98 minutes: Coronavirus (COVID-19) response, 7 December 2021

[…]

The doubling time for new Omicron infections is currently around 3 days in England and although some potential biases in the data lead to some uncertainty around this growth rate, it is unlikely to be any slower than a 5-day doubling time (high confidence). Omicron cases are making up an increasing proportion of SGTF cases, which will allow this estimate to be refined.

[…]

7. The proportion of Omicron infections that result in hospital admission cannot be known until there have been more hospitalisations (and analysis will need to consider the age profile of those infected and the immune history, such as vaccination and past infection status). The number of people in hospital at any time will also depend on the length of stay, which will depend in part on severity of disease.

8. Some early indications from South Africa suggest less severe disease in those hospitalised when compared to previous waves, though this likely reflects at least in part the characteristics of those being admitted to date, who are younger than in previous waves (low confidence). A modest reduction in severity would not avert high numbers of hospitalisations if growth rates remained very high.

9. Although there are several unknown factors, preliminary modelling suggests that without any changes to measures in place, the number of hospitalisations from Omicron may reach 1,000 per day or higher in England by the end of the year (and still be increasing at that point).

10. The overall scale of any wave of hospitalisations without interventions is highly uncertain, but the peak could reach several times this level. The peak is highly likely to be higher than 1,000 to 2,000 Omicron hospital admissions per day without intervention to slow the speed of increasing infections; for it to be below this level there would need to be only a small degree of immune escape and very high protection from boosters against Omicron.

11. The impact of changes in transmissibility and immune escape on overall numbers of admissions is likely to be much more significant than the impact of any changes in severity (high confidence).

12. There are other impacts from high levels of incidence aside from hospitalisations and deaths, including the morbidity burden in those who are not hospitalised and workforce absences. With very rapid doubling times a large wave could occur leading to synchronous absences from work.

13. If measures to reduce transmission were introduced, admissions would continue to increase for a number of weeks after this point. This is due to the lags between measures being introduced, these measures affecting incidence of infections, and then this being reflected in hospitalisations.

14. The faster the growth in infections at the point measures are introduced, the more admissions will increase in the period between action being taken and the number of admissions being affected. With lags of the order of two or more weeks, and doubling times of the order of three days, it is likely that, once hospitalisations begin to increase at a rate similar to that of cases, four doublings (a 16-fold increase) or more could already be “in the system” before interventions that slow infections are reflected in hospitalisations.

https://www.gov.uk/government/publicati ... ember-2021

Skip to 14 for the scary bit.

[Herainestold]

I hope the NHS is stocking up on oxygen.

[lpm]

The NHS problem isn't going to be objects, it's going to be people. Omicron will sweep through doctors and nurses. Understaffing can't be prevented by stocking up on nurses now.

The partial answer is to wander round hospitals with vaccines, jabbing any staff member you meet who hasn't had the 3rd yet. But VE is too low against Omi transmission for this to be the full answer.

[lpm]

This is what it says about replacement of Delta. Doesn't say anything because they don't know.

My thinking is that Plan B is pretty anaemic so if it had been introduced months ago it would have only slowed Delta modestly. Hence it simply continues while Omi establishes itself places like Northamptonshire. Then replacement happens in a rush - a person with Delta and a person with Omi enter the same nightclub and everyone leaves with Omi.

15. It is highly likely that Omicron will account for the majority of new SARS-CoV-2 infections in the UK within a few weeks. Omicron may partially or largely replace Delta over this period, but the extent of this depends on the degree to which they are infecting different cohorts of people, which is not currently known. It is possible that both could continue to spread concurrently, which would mean that the Delta infections and hospitalisations would be in addition to Omicron ones. It is also possible that Omicron will completely displace Delta

[lpm]

101 yesterday, 131 today. 3 day doubling period. No error bars needed, a couple of data points is enough.

Now 101 to 131 to 249. It's accelerating.

[lpm]

Anyone want to comment on my assumptions for my latest spreadsheet "model"? Or approach?

I'm starting from Delta world and I'm assuming:

1) Delta R=1.0 with the basic UK measures in place
2) Scotland and Wales have had Plan B, effectively, going for a while now with masks, WFH etc, and it's made no discernable difference you can point to
3) Omicron reduces vaccine's transmission benefit by 50% vs Delta
4) Omicron reduces vaccine's hospitalisation benefit by 10% vs Delta
5) Omicron reduces vaccine's mortality benefit by 5% vs Delta
6) Starting point 50,000 cases per day Delta and 1,000 Omi
7) Hospital patients start at 7,500 and NHS "breaks" at 60,000, i.e. three doublings

[lpm]

So, first up, Omicron appears very much milder than Delta for hospitalisations.

Not because of any change to its real hospitalisation impact, the implicit assumption I've made is it's identical.

Purely because of the assumptions that vaccines have their transmission benefit down 50% due to Omicron, but hospitalisations are down only 10%.

Which is obvious when you think about it, but playing around with the numbers makes it really clear - Omi sweeps through the vaccinated with a massive case load but proportionally far fewer of those cases end up in hospital. The previous ratios used for Delta are completely transformed.

Vaccinated people already land up in hospital with Delta, however, and the 10% worse protection of vaccines increases it. This is merely a ratio thing. We need to be very careful looking at any data from South Africa etc that compares cases to hospitalisations because it's absolute hospitalisations that matter. Ratios are a trap for the unwary.

Unvaccinated people land up in hospital at exactly the same ratio as with Delta, using my assumptions. Looks like they're f.cked.

I think I'll assume the 3rd booster doses restores all vaccine effectiveness levels to that of Delta.

[lpm]

Well, cases rise to infinity. Can't stop them.

Which is the inevitable product of my assumption of boosters only restoring to Delta vaccine effectiveness, plus only 20 million people with boosters at the moment, plus a rollout of boosters to the rest only over a few months, plus Delta never being better than R=1 to begin with.

I'd have to assume either:

- 3 jabs vs Omicron gives improved vaccine transmission protection compared to 2 jabs vs Delta - seems unlikely
- Plan B (or additional measures) are sufficient to overcome 50% cut in transmission effectiveness - not going to happen

[dyqik]

So, first up, Omicron appears very much milder than Delta for hospitalisations.

Not because of any change to its real hospitalisation impact, the implicit assumption I've made is it's identical.

Purely because of the assumptions that vaccines have their transmission benefit down 50% due to Omicron, but hospitalisations are down only 10%.

Which is obvious when you think about it, but playing around with the numbers makes it really clear - Omi sweeps through the vaccinated with a massive case load but proportionally far fewer of those cases end up in hospital. The previous ratios used for Delta are completely transformed.

Vaccinated people already land up in hospital with Delta, however, and the 10% worse protection of vaccines increases it. This is merely a ratio thing. We need to be very careful looking at any data from South Africa etc that compares cases to hospitalisations because it's absolute hospitalisations that matter. Ratios are a trap for the unwary.

Unvaccinated people land up in hospital at exactly the same ratio as with Delta, using my assumptions. Looks like they're f.cked.

I think I'll assume the 3rd booster doses restores all vaccine effectiveness levels to that of Delta.

There's another (I think it's slightly different to this) effect that can make Omicron look milder without it being so - see https://twitter.com/nataliexdean/status ... 89994?s=20

[lpm]

Outcome of "model":

1) Slower rise of Omi by 31 December compared to most of the other speculation. I only see a rise from current 50,000 to 119,000 cases at 31 Dec.

2) By this point Omi has taken over and Delta fades away. The exponential of Omi kicks in and it goes thumping up, despite increasing boosters giving more and more people with Delta-like immunity. As always, the exponential virus defeats the linear vaccine rollout.

3) Cases:
- 250,000 cases a day by 7 Jan
- 1,000,000 a day by 17 Jan
- Towards the final week of January we approach herd immunity via infection - over 90% of the population have been infected - and the epidemic will fizzle

4) Hospital numbers (two week lag for hospitalisations):
- Match Jan 2020 peak of 40,000 in hospital on 21 Jan
- Reach my breaking point of 60,000 on 25 Jan
- 150,000 in hospital by 31 Jan (which is based on case numbers of 17 Jan). I think there are nearly 150,000 beds in total in the UK, so long as you can turf out every single non-Covid patient...

5) Deaths:
I haven't done deaths. A lot of the unvaccinated are prior infections, which should save their lives. A prior infection isn't as good as two doses. A prior plus a single jab now will probably be as good as a three dose boosted person. And I don't know how much of a benefit the latest treatments will give to the severely ill. And the lags... I don't think it's the death toll that will force the government to act, it's too slow a metric. It's going to be hospital collapse.

Conclusions:

1) The devastation is going to happen very quickly in January. The 11 Jan case load delivers the NHS breaking point hospital numbers of 60,000. Delay lockdown just a few days beyond 11 Jan and the hospitals are truly f.cked.

2) Boosters can't possibly be fast enough to stop it. 0.5 million per day sounds like a lot. But there's literally only a few weeks till we're in f.cked territory. And presumably a dose now still takes a few days lag before it prompts the immune system.

3) If Omicron is naturally milder than Delta, rather than this merely being a ratio effect, then that could save us a doubling. But a doubling is only a few days, increasing to a week or so thanks to boosters.

4) Back to full lockdown by 26 December latest is required. Obviously this assumes the basic 2-3 day doubling estimate, plus all the other lesser assumptions. BUT...

5) There is going to be a serious perception problem. Total cases = very large constant Delta + very small exponential Omi. So it's a very slow curve upwards to begin with. But then you hit the point where it's large exponential Omi and the curve just goes crazy upwards.

It's like you get paid a constant £100 a day plus 1 pence, but the 1p doubles every day. After a week you are earning £100.64 a day. After two weeks you are at £182. But at three weeks you are getting paid £10,586 a day. The mathematicians here will have the jargon for this sort of exponential.

But what matters is there's zero hope of Tory MPs, the media and the general public understanding this dynamic. In two weeks time cases will have gone up a bit, but it'll be Christmas and no-one will be watching and the rise doesn't feel steep. And then suddenly it goes exponential but it's New Year and no-one will be watching. And hospitalisations will still be sluggish because they're the result of the older part of the curve, so everyone will still have an excuse to procrastinate...

I think this is my main take away from playing around with the numbers.

[lpm]

There's another (I think it's slightly different to this) effect that can make Omicron look milder without it being so - see https://twitter.com/nataliexdean/status ... 89994?s=20

Yep, that picture is perfect at expressing what I was clumsily trying to put into words.

The "severe" box on the right would actually expand a bit under the assumption there's a 10% lower effectiveness against serious illness - some of the extra re-infected jump into that bucket (as indeed they do with breakthrough/re-infection Deltas - currently 38% of Covids in hospital in the UK are vaxxed).

[Herainestold]

The NHS problem isn't going to be objects, it's going to be people. Omicron will sweep through doctors and nurses. Understaffing can't be prevented by stocking up on nurses now.

The partial answer is to wander round hospitals with vaccines, jabbing any staff member you meet who hasn't had the 3rd yet. But VE is too low against Omi transmission for this to be the full answer.

Valid point. Can we stockpile nurses from India and the Philippines now and keep them in readiness?

Seriously, we should be locking down now, cancelling public events, pushing vaccination, and working very hard on an Omicron vaccine.
For the individual, get boosted, double mask, avoid all social contact. Easy to say, harder to do.

[lpm]

A second version to be more optimistic:

- Assume that Plan B is joined by voluntary action by the public to limit interactions, cancel Christmas parties, stop going to nightclubs. I think this voluntary partial lockdown would be pretty plausible if we were in January now, we're really unlucky that Christmas is upon us. But I'm pretty sure we'll see Plan B+ in reality, rather than the pathetic official Plan B.

1) Not much different at 31 December, 106,000 cases (previous version 119,000).

2) Omi exponential much more muted in January.

3) Cases plateau at 176,000 per day by late Jan because boosters have time to kick in. Then cases decline.

4) Hospitalisations peak at 29,000 patients. below the Jan 2021 peak of 40,000.

Let's face it, the fate of the UK is better off in the hands of the people and their voluntary action rather than the actions of Johnson and his band of c.nts.

[jdc]

A prior plus a single jab now will probably be as good as a three dose boosted person.

Prior infection plus vaccination gives you kick-ass immunity - "gold standard" according to Finzi. Presumably because prior infection gives you broader protection (antibodies to non-spike proteins etc) while vaccines give better specific immunity (superior antibodies to spike protein). The hope is that things like longer gaps between doses or booster shots will have similar effect to infection + protection:

Spoiler:

Fig. 1b https://drive.google.com/file/d/1YIy4Vo ... Y_wYg/view and discussion:

We then tested the ability of plasma from BNT162b2 vaccinated study participants to neutralize Omicron versus ancestral D614G virus in a live virus neutralization assay. We tested 14 plasma samples from 12 participants (Table S1), with 6 having no previous record of SARS-CoV-2 infection nor detectable nucleocapsid antibodies indicative of previous infection. For two of these participants, we used samples from two timepoints. The remaining 6 participants had a record of previous infection in the first SARS-CoV-2 infection wave in South Africa where infection was with ancestral D614G virus (Table S1). Geometric mean titer (GMT) FRNT50 (inverse of the plasma dilution required for 50% reduction in infection foci number) was 1321 for D614G. These samples therefore had very strong neutralization of D614G virus, consistent with sampling soon after vaccination. GMT FRNT50 for the same samples was 32 for Omicron, a 41-fold decline (Fig 1B). However, the escape was incomplete, with 5 of the participants, all previously infected, showing relatively high neutralization titers with Omicron.
Beta variant escape from BNT162b2 in a live virus neutralization assay has been reported to be substantial (3) and our own data confirmed these results (4), with about 3-fold reduction in FRNT50. The results we present here with Omicron show much more extensive escape. However, escape was incomplete in participants with higher FRNT50 due to previous infection. Previous infection, followed by vaccination or booster is likely to increase the neutralization level and likely confer protection from severe disease in Omicron infection.

Also https://www.nature.com/articles/d41586-021-02795-x

A third vaccine dose might allow people who haven't been infected to achieve the benefits of hybrid immunity, says Matthieu Mahévas, an immunologist at the Necker Institute for Sick Children in Paris. His team found that some of the memory B cells from naive vaccine recipients could recognize Beta and Delta, two months after vaccination9. “When you boost this pool, you can clearly imagine you will generate potent neutralizing antibodies against variants,” Mahévas says.

Extending the interval between vaccine doses could also mimic aspects of hybrid immunity. In 2021, amid scarce vaccine supplies and a surge in cases, officials in the Canadian province of Quebec recommended a 16-week interval between first and second doses (since reduced to 8 weeks).

A team co-led by Andrés Finzi, a virologist at the University of Montreal, Canada, found that people who received this regimen had SARS-CoV-2 antibody levels similar to those in people with hybrid immunity10. These antibodies could neutralize a swathe of SARS-CoV-2 variants — as well as the virus behind the 2002–04 SARS epidemic. “We are able to bring naive people to almost the same level as previously infected and vaccinated, which is our gold standard,” says Finzi.

And https://www.nature.com/articles/s41586-021-04005-0

By aggregating variant of concern-associated and antibody-selected spike substitutions into a single polymutant spike protein, we show that 20 naturally occurring mutations in the SARS-CoV-2 spike protein are sufficient to generate pseudotypes with near-complete resistance to the polyclonal neutralizing antibodies generated by individuals who are convalescent or recipients who received an mRNA vaccine. However, plasma from individuals who had been infected and subsequently received mRNA vaccination neutralized pseudotypes bearing this highly resistant SARS-CoV-2 polymutant spike, or diverse sarbecovirus spike proteins.

[bob sterman]

Prior infection plus vaccination gives you kick-ass immunity - "gold standard" according to Finzi.

What do people think vaccination + later (consequently mild) infection might give you? Not as good that way round?

[dyqik]

A second version to be more optimistic:

- Assume that Plan B is joined by voluntary action by the public to limit interactions, cancel Christmas parties, stop going to nightclubs. I think this voluntary partial lockdown would be pretty plausible if we were in January now, we're really unlucky that Christmas is upon us. But I'm pretty sure we'll see Plan B+ in reality, rather than the pathetic official Plan B.

It'll be interesting to see how this works in the US as well, with Thanksgiving as the trigger for a lot of cases (possibly with Omicron involved), as that's the holiday when you're more obliged to go an eat dinner opposite your Trumpist anti-vax Uncle/Step-father-in-law, and it's too cold to be outside.

[Woodchopper]

Several vaccine makers are racing to develop revamped Covid-19 shots against Omicron in case the highly mutated variant renders the existing shots less effective. But it isn’t yet clear that the immune response can be fine-tuned to the new strain by boosting with a tweaked vaccine.

A big hurdle for developing variant vaccines is what immunologists call “original antigenic sin,” a phenomenon documented in flu and other infectious diseases, where the body returns to the immune response mounted against its first encounter with a pathogen—or vaccine—when faced with a slightly different variant.

Evidence is building that this phenomenon, also known as immune imprinting, is at work in Covid-19. The implication: Boosting with an Omicron-specific vaccine might only reawaken earlier immune responses, whether they were spurred by vaccination or infection. In other words, an Omicron-specific vaccine may have no advantage over simply boosting with the original vaccines.

“You might be better sticking with the first one [vaccine] you had rather than chasing the globe for the next variant,” said Danny Altmann, professor of immunology at Imperial College London, who was one of the leading authors of a recent paper that provided evidence of immune imprinting with SARS-CoV-2, the virus that causes Covid-19.

His paper, published last week in the journal Science, found that a person’s first encounter with the virus’s spike protein—the part that it uses to enter cells—be it through infection or vaccination, shaped their subsequent responses to the virus.

It detailed how, in U.K. healthcare workers, immune responses differed depending on which variant of the spike protein they first encountered. Those who were infected with the Alpha variant last winter, then vaccinated, had different immune responses to those who had been infected with the Wuhan strain during the first wave, and later vaccinated. For instance, they had stronger immune responses to the Delta strain, and weaker responses to the Beta strain, than those who had only encountered the spike protein of the original Wuhan strain. All of the authorized vaccines are based on the Wuhan strain.

“The world of SARS-CoV-2 immunity has become a very complex landscape where we’re all a little bit different, so we don’t start our vaccine with a blank sheet,” said Prof. Altmann. “This stuff needs to be decoded.”

https://www.wsj.com/articles/covid-19-v ... 1639054806

[shpalman]

Prior infection plus vaccination gives you kick-ass immunity - "gold standard" according to Finzi.

What do people think vaccination + later (consequently mild) infection might give you? Not as good that way round?

Has anyone ever had a second infection which was worse than the first one? Is that one of the things which might happening with Omicron? (I mean, young people who had mild/asymptomatic infections which weren't officially detected so that now it's noticed that they're getting symptoms from Omicron those symptoms aren't too bad, and it's making Omicron seem mild because the previous infection isn't being taken into account. There were some drawings off twitter explaining this somewhere.)

Otherwise, well, the whole point of vaccination is to make your covid experience milder, but if your covid experience was already mild even with incomplete vaccination then I don't know how subsequent infections would be worse.*


* - unless your general health severely deteriorated in the meantime for other reasons or something.

[lpm]

Anecdotally, second infections were similar to first - people felt just as ill - but the time feeling ill was shorter and recovery easier.

[jdc]

I thought bob was asking us to compare quality of immunity from infection then vaccination versus vaccination then infection rather than relative severity of infection 1 and infection 2?

(Not that I know the answer, like. Sorry bob.)

[Woodchopper]

Signs that cases in Gauteng may be starting to peak. If so the wave may be similar to Delta. But it could be a consequence of testing capacity.

https://twitter.com/rid1tweets/status/1 ... 57090?s=21

[bob sterman]

I thought bob was asking us to compare quality of immunity from infection then vaccination versus vaccination then infection rather than relative severity of infection 1 and infection 2?

(Not that I know the answer, like. Sorry bob.)

Yup. That's what I was wondering about.