B.1.1.529 Omicron variant

[Woodchopper]

... the latest epidemiological evidence suggests that Omicron can evade immunity from infection with earlier variants and is causing reinfections at three times previous rates.

I think this is the preprint:

Increased risk of SARS-CoV-2 reinfection associated with emergence of the Omicron variant in South Africa
https://www.medrxiv.org/content/10.1101 ... 21266068v2

And here’s a thread by one of the authors: https://twitter.com/sacemadirector/stat ... 51371?s=21

Thread uses several methods to estimate that in Gauteng the Rt for Delta in November was 0.8 [cases were declining due to immunity and NPIs] whereas for Omicron it was 3-3.5.

https://twitter.com/trvrb/status/146724 ... 10624?s=21

[Woodchopper]

And back to hospital admissions in Gauteng.

Thread provides data on use of oxygen, ICUs and age profile of patients: https://twitter.com/jburnmurdoch/status ... 87012?s=21

[Woodchopper]

UK, looks like “Omicron could be doubling share of cases in <3 days.” https://twitter.com/forensic_stats/stat ... 18666?s=21

[shpalman]

good job you brought in those travel rules then

[shpalman]

https://www.gov.uk/government/publicati ... se-numbers has 32 Omicrons in the 3rd December update (compared to ~84,000 new Delta and Delta+)

[lpm]

good job you brought in those travel rules then

To be fair, it was a fairly prompt stable shutting. But Omi seems to Usain Bolt it out the blocks with perhaps the ability to transmit a day earlier.

[Woodchopper]

There's been a suspected Omicron outbreak in Oslo. Someone travelled from South Africa to Norway (following all the rules), and later went to a work function involving 100 people. I assume it was a Christmas party, so probably lots of dancing, singing and talking loudly over the noise. The traveler later tested himself and got a positive result, and it looks like 30-40 others were infected (either that evening or through secondary contact over the following days).

Link in Norwegian but Google Translate is your friend:
https://www.vg.no/nyheter/innenriks/i/R ... nvarianten

Now estimated that 50-60 have been infected with more expected. The increase probably reflects secondary infections.

Omicron has been confirmed in one person and others are expected. All the people infected are vaccinated adults. Most of them are said to have symptoms but are not seriously ill. As it was a tech company Christmas party I assume that they’d probably be fairly young and generally healthy.

Link in Norwegian https://www.vg.no/i/g6R7VJ

Now looking like 90-100 were infected by two people who’d recently returned from South Africa. 13 have been confirmed to have Omicron, we’re awaiting test results for the rest.

Link in Norwegian https://www.nrk.no/norge/fhi_-19-tilfel ... 1.15757630

Now 120, with 70 who were guests at the party and the rest were also at the venue.

https://www.vg.no/nyheter/innenriks/i/v ... ker-brygge (Link in Norwegian).

[shpalman]

Two cases found in Britain.

Which implies twenty cases already here? Two hundred? Two thousand? What multiple should we guess at?

Well luckily thanks to your vaccination program plus the general use of masks indoors and vaccination/recovery/negativity protocols for crowded indoor events you don't have that many cases so you'll be able to effectively trace contacts and sequence a substantial proportion of the positives.

And since you weren't trying to be sure of killing 1000 people per week your hospitals aren't too stretched so you'll be able to cope for the 3-4 weeks over which hospitalizations will keep rising even after the lockdown you'll quickly implement to keep this under control.

The UK’s NHS will be in a “very, very difficult position” if the Omicron variant were to lead to a surge in hospital admissions in the UK

1. they mean "when the Omicron variants leads to a surge" but are just showing off that they know how to use the subjunctive (except that the main clause should probably be conditional rather than future).

2. this is the president of the Royal College of Emergency Medicine saying it, who is hardly likely to say that everything is fine, knock yourselves out.

[shpalman]

We’re in a much better place than a lot of countries right now. The decisions we made in the summer are paying dividends.

Probably translates as "our cases aren't going up (because they were already higher than everyone else)" and "we found out that we can let 800-1000 people die per week without anyone really caring".

link

[shpalman]

More morning news via the live blog

Prof Paul Hunter, from the school of medicine at the University of East Anglia, said there was current concern that Omicron “is spreading rather more quickly than the Delta variant” and there were probably more than 1,000 cases in the UK at the moment.

Meanwhile Kit Malthouse was also asked given that there appear to so far have been no deaths or hospitalisations involving the new variant, was the UK government over-reacting. I think a better answer would be to note that it takes a week or three for hospitalizations and deaths to happen after getting infected so if you wait for that data and then have a bit of a think about it before deciding what to do you're already several weeks too late. But then the government basically decided to completely not react to Delta over the summer and it turns out that was fine because people don't care.

[shpalman]

Now estimated that 50-60 have been infected with more expected. The increase probably reflects secondary infections.

Omicron has been confirmed in one person and others are expected. All the people infected are vaccinated adults. Most of them are said to have symptoms but are not seriously ill. As it was a tech company Christmas party I assume that they’d probably be fairly young and generally healthy.

Link in Norwegian https://www.vg.no/i/g6R7VJ

Now looking like 90-100 were infected by two people who’d recently returned from South Africa. 13 have been confirmed to have Omicron, we’re awaiting test results for the rest.

Link in Norwegian https://www.nrk.no/norge/fhi_-19-tilfel ... 1.15757630

Now 120, with 70 who were guests at the party and the rest were also at the venue.

https://www.vg.no/nyheter/innenriks/i/v ... ker-brygge (Link in Norwegian).

THIRTY MORE PEOPLE became infected AFTER THE PARTY ENDED AND PEOPLE LEFT. (probably wrong, see replies)

[Woodchopper]

Now looking like 90-100 were infected by two people who’d recently returned from South Africa. 13 have been confirmed to have Omicron, we’re awaiting test results for the rest.

Link in Norwegian https://www.nrk.no/norge/fhi_-19-tilfel ... 1.15757630

Now 120, with 70 who were guests at the party and the rest were also at the venue.

https://www.vg.no/nyheter/innenriks/i/v ... ker-brygge (Link in Norwegian).

THIRTY MORE PEOPLE became infected AFTER THE PARTY ENDED AND PEOPLE LEFT. (probably wrong, see replies)

Yes, I've had a look. None of the Norwegian sources state that the infections of other people at the venue happened after the party had left.

I think that would be unlikely anyway. Around here the usual thing at a work a Christmas party is for people to stay in one place (the company is probably paying for the drinks). It would be odd if they all got up and left the venue at once.

All the people infected were double vaccinated though.

[shpalman]

Meanwhile, Japan has a third case from "a man in his 30s who had been in Italy". Reminds me of how in the first wave there weren't many cases in Milan itself but lots of tourists who'd been to Milan turned out to have it once they got home; they can't all have been visiting Codogno. So there's obviously a lot more Omicron in Italy, either as covid infections which haven't been sequenced, or as infections which haven't even been identified as covid.

[shpalman]

336 in the UK now

[shpalman]

336 in the UK now

... we can conclude that there is now community transmission across multiple regions of England.

[lpm]

Judging by this chart, Omi is well fit. I don't mean in the young person slang sense.

https://twitter.com/rid1tweets/status/1 ... CWAyg&s=19

[Brightonian]

Oh. "Only 30-35% of PCR tests for general population are processed at labs that look for a key indicator of new variant"

So multiply the 336 confirmed cases by 3, along with any estimates that people have come up with?

[shpalman]

Prof Tim Spector, from out of off of the Covid Zoe app (and professor of genetic epidemiology at King’s College London #secondmentions) said

“... it’s probably at least 1,000 to 2,000 I would guess at the moment... doubling about every two days”

Not clear to me how many of those 1,000-2,000 would be included in the ~50,000 official cases per day, not least because the news from the app is usually how many are currently infected, not the new case rate, while the official data only has the new case rate and there was never a protocol for counting people recovered from covid.

about half of all cases at the moment of Delta are being “missed” because they are not presenting with “classic” Covid symptoms

... and this might be even worse for Omicron.

Let's do the maths:

assumed it’s 1,000 at the moment, and you think it’s going to be doubling every two days, you can see that those numbers in about 10 days’ time are going to be...

scribbles on bit of paper... opens calculator app... erm...

pretty (high) certainly

[shpalman]

I would also have Edited To Add that your hospitalizations are at a low at the moment but this corresponds to the dip in the case rate a few weeks ago, and since then it's mainly been young people getting infected (like in July), so the lag to hospitalizations and deaths is longer because it has two components, being the time it takes for young people to infect older people, and then the time it takes the older people to deteriorate. Booster jabs may have taken the edge off it so it'll come up slowly but it will come up.

[lpm]

Multiply the 336 confirmed cases by 3 = 1,000

Use 2x factor for asymptomatic/not tested = 2,000

Doubling time is a few days and we're looking at starlight from a few days ago = 4,000

So not yet reached the sensitivity of the daily case range of a few thousand up or down from 50,000 per day. But soon will be. We should expect to see daily case numbers increase from now on.

[sTeamTraen]

Daily UK cases have been higher than the equivalent number from 20 days earlier for 19 out of the last 20 days. On the other hand, daily deaths have been *lower* than the equivalent number from 20 days earlier for 19 out of the last 20 days.

Whatever is happening in the UK, it looks very strange on a chart, when compared with almost any other country where waves go up like a rocket and then fall almost equally quickly as lockdowns are brought in. Since July there have been at least 30k cases every day but only in the last few days has the daily number crept above 50k. Maybe that's Omicron?

So something has been limiting exponential growth in a way that hasn't been operating in countries with not massively dissimilar vaccination numbers. Since the first peak of the current wave in mid/late July, the mean increase in UK case rates has been about 80 per day per day [sic], which suggests an R only very very slightly above 1.0.

[lpm]

Yep. That's the big mystery we haven't solved. In the UK Delta never managed to go exponential, despite pulses of cases running through schools etc. And yet in Holland a very similar picture of vaccinations and unlockdown suddenly turned into a full-on exponential.

Deaths have often diverged from cases temporarily, it's just the mix of school kids vs OAPs. The past month cases has been all about kids. The latest few days have seen that filter into older people though. We never escape the inevitable for long.

[sTeamTraen]

Yep. That's the big mystery we haven't solved. In the UK Delta never managed to go exponential, despite pulses of cases running through schools etc. And yet in Holland a very similar picture of vaccinations and unlockdown suddenly turned into a full-on exponential.

Deaths have often diverged from cases temporarily, it's just the mix of school kids vs OAPs. The past month cases has been all about kids. The latest few days have seen that filter into older people though. We never escape the inevitable for long.

I suppose a danger is that the exponential spread of Delta has been held back somewhat by substantial previous (sub-herd?) immunity, and that if Omicron bypasses that you could have exponential growth from a base of 400 cases/100k/week.

[shpalman]

It was exponential enough at the beginning of July as cases spread between newly-liberated young people. We still don't know why it peaked at around 50,000/day and came down then, or why it peaked at about 40,000 and came down in early September, or 50,000 again in mid October, and it may well peak at about 50,000 again this week before coming down again.

50,000 cases per day in the UK corresponds to about 500/100,000/week. Scotland's October peak was at 800/100,000/week but all the other UK peaks are basically dominated by England. (Obviously as you look at smaller and smaller regions you'll find more and more extreme case rates until you get to individual people who are either 100,000/100,000/week or 0/100,000/week. There's an MSOA in Lincolnshire which is at 865.6)

Meanwhile in Italy (170/100,000/week), Trieste still seems to be the province with the highest case rate what with being close to Slovenia and apparently having a strong antivax sentiment. From a minimum of about 70/100,000/week at the beginning of October it shot up to about 700/100,000/week by mid November, but it has been fairly stable at that level since then. The rest of Friuli-Venezia-Giulia is only at about 350/100,000/week and coming up more slowly. Yellow Zone rules really aren't that much of a big deal and they've only been implemented recently so I don't think that's it. The new Green Pass / Super Green Pass rules which just came into force may have persuaded a few more holdouts to get vaccinated - according to https://lab.gedidigital.it/gedi-visual/ ... ni-italia/ there's currently over 30,000 first doses being given per day, when that number was less than 20,000 per day a month ago. Hmm, FVG doesn't seem to have started to give boosters at all.

It seems like there's something which manages to self-limit covid outbreaks, possibly that super-spreading/clustering model which means lots of people in limited social groups infect each other quickly but infections spread less easily from one group to another. Alternatively, people take personal measures when they feel like there's lots of covid about. In the UK, the app pinging people might make a difference, but I really don't think our app here actually does anything.

But then Austria in lockdown peaked at over 1000/100,000/week and Slovakia, also in lockdown, is at a peak of about 1400/100,000/week. Switzerland is at 700/100,000/week and still heading up.

Yeah no don't know.

[lpm]

Delta was clearly exponential in unvaccinated young and partially vaccinated young, but not at the 75% vaccinated national level. Omicron is going to be exponential at the national level. Cases are going to go up.

But it's NHS capacity we care about, not cases. So it all hangs on whether the ratio of hospitalisations to cases is changed by Omi.

The UK has room for two doublings of hospital patients, with the third doubling taking the NHS to the edge of breakdown.

- currently 7,500 Covid patients in hospital
- two doublings 30,000, still less than the Jan 2021 peak of 40,000
- third doubling to 60,000 takes us to the point where people are dying in the carpark

In a bad scenario of a fit Omi with the same hospitalisation ratio, the first doubling due to Omi will be very slow. It takes weeks for Omi to grow from 1,000 cases and replace Delta. At a quick look I estimate the first doubling of cases will take till the end of Jan. But the second doubling is a lot quicker as it's all faster Omi. I expect patients in hospital to hit the 30,000 mark at the end of Feb. You've got to lockdown before this point otherwise the lags in the system will guarantee you exceed the Jan 2021 peak.

Everything seems to indicate that Omicron will return us to the exponential world. Delta was only ever in balance so it doesn't take much extra to knock us into a phase transition. Dubious that boosters will be enough without lockdown measures to help.

Optimistically there might be hints that Omi puts people in hospital at the same rate but they stay there for a shorter time and need less oxygen. Obviously if hospital admissions double but length of stay halves you've not got a change in usage.