Covid 2.0

[GMTV]

Probably a good question on the percentages. Numbers get thrown around on the subject and it's hard to really work out the implications.

The government are monitoring so many parameters they could probably conjure up a couple of headline numbers to make whatever point they want to. And the media just seem to lap it up with the minimal of scrutiny.

It works the opposite way too, when the covid deniers say it only affects 97% or whatever percentage of people in a minor way... Or say 99% of people are absolutely fine with it, no worse than a cold. Well 1% of 60 million people is 600k so I think that would easily prove its a hugely dangerous virus.

[Zetetic]

The NERVTAG minutes indicate that the "70% higher than other variants" refers to the growth rate, not the reproduction ratio (R).

I believe this isn't a mistake, but it's sort of harder to make sense of because growth rate depends heavily on prior prevalence. A very high R (e.g. "2") doesn't actually result in very fast growth while almost no-one has the disease.

The high growth rate of this variant is particularly remarkable, because it's in the context of being barely present at all in late-November to accounting for a majority of cases in some places (SE England) and a decent proportion of cases - 20% - in lots of other parts of the UK (e.g. ██████) by mid/late-December.

The NERVTAG minutes then indicate that presence of the variant has been associated with an "absolute increase" in the reproduction ratio (R) of between +0.39 to +0.93.

[steveh]

The current R₀ number for the whole of the UK is around 1.1 to 1.2, so if 10 people are infected there would be a 11 further infections using the lower number.

The estimates for the increase in R₀ with the new variant are from 0.39 to 0.93. So if we take the lowest number, instead of there being 11 further infections there would be 15.

NERVTAG also yesterday increased their confidence in the transmissibility and the numbers.

[Zetetic]

I don't think those are estimates of increases in R0, I think they're estimates of increases effective R (as observed in parts of the UK, pre/post detection of the variant, or possibly comparing different areas).

[steveh]

Ah, I'm mixing up two things. However, this says:

There are various estimates of what that increase in transmission means, but researchers say the new variant could have increased the reproduction number — a figure known as R-naught, which is the average of how many people one person will infect — from 1.1 to 1.5.

[Zetetic]

I think, in fairness, any consistency has largely broken down and this is simply now old-and-busted prescriptivism:
https://www.cebm.net/covid-19/when-will-it-be-over-an-introduction-to-viral-reproduction-numbers-r0-and-re/

(I guess it's also arguable what R0 vs. Re would even mean for the new variant, as it's never existed outside of the current lockdown.)

[BlodwynPig]

Variant has been around since September.

[Zetetic]

If that's in response to my "barely present at all in late-November" comment, that's not meant to be a claim that it didn't exist earlier just that we're not seeing it[nb]Well, S-gene weirdness.[/nb] in huge numbers before then.

[Chedney Honks]

All the this talk is about its being uncontrollable. I mean, it isn't. Just everyone stay home for a month. Done.

[GMTV]

Variant has been around since September.

If its been around since September then what's driving the fairly recent increase in numbers across the country?

[Uncle TechTip]

Can't be the schools like the childfree claim. University students finally allowed home perhaps?

[jobotic]

I'm not childfree and I claim its the schools, mixed with the new variant here.

[George Oscar Bluth II]

It could be that the variant passes on better in kids and that kids are basically the one group of people being sent to sit in rooms together for 6 hours a day, which accounts for the effect it has during a notional lockdown. Bet we'll get the schools closed in January.

[Bence Fekete]

So if these things mutate so predictably and will, scientifcally, inevitably do so given X number of iterations, then why even use R as our main metric?

I.e. if we knew they turn badass given enough time/freedom then why even risk such high (below 1) infection rates as a strategy to begin with? Shouldn't we be prioritising stunting its evolution at a lower infection level?

Or is this hindsight? I guess easy to say now but that's why I'm asking.

[George Oscar Bluth II]

I mean, more chances of bad mutations was definitely a reason to suppress it as far as possible over the summer rather than, I dunno, paying people to go to sit indoors with others eating and drinking.

[Zetetic]

I think the general consensus on coronaviruses is that they're more stable than, say, influenza. Coronaviruses' genomes are a lot bigger - but they have more mechanisms in place to reduce mutation because they're otherwise so rickety. (See, e.g. https://www.nature.com/articles/d41586-020-02544-6 , which arguably looks quite optimistic now!)

It's just that, yeah, we've maintained a decent reservoir of infection over a pretty long time and applied a moderate selective pressure, I guess.

(Noting the theory that steveh highlighted that perhaps we specifically incubated this one in an immunodeficient person receiving plasma treatment in a hospital: https://www.cookdandbombd.co.uk/forums/index.php/topic,84084.msg4410538.html#msg4410538 )

[steveh]

Similarly, in the South African variant they think it might have happened in an immunocompromised patient, perhaps connected to the very high rate of HIV in the country.

However, PHE's report yesterday said it's possible that the N501Y mutation may be solely responsible for the increase and that wouldn't need such a radical set of mutations to occur and has been already seen in small mammals.

[olliebean]

The NERVTAG minutes indicate that the "70% higher than other variants" refers to the growth rate, not the reproduction ratio (R).

Ah, I see. In that context it's difficult to see exactly what "70% higher" means. The growth rate is the percentage daily change in the number of new infections. According to the government's current data, this varies between 0% and 6% in the various regions. I presume "70% higher" doesn't mean that translates to 70-76%. Rather, I would assume that 0-6% would translate to approximately 0-10%, i.e., 1.7 times as high. But then, a negative growth rate means daily new infections are falling. What does "70% higher" mean in that case? Presumably not that new infections will fall by 70% more per day.

[DistressedArea]

I took it to be something like the effective contact rate or else the transmission risk as described in:

https://en.m.wikipedia.org/wiki/Transmission_risks_and_rates

In any case I'd expect these to be model parameters which would not correspond to a rate you'd actually measure directly. Also, any measurements direct or effective would be susceptible to all kinds of measurement errors and noise.

Don't know if 70% is a relative increase or absolute though.

[Zetetic]

What does "70% higher" mean in that case?

Not sure this came up, mind you.

The PHE report linked by steveh is worth reading, as it talks about estimating the impact of KentVid[nb]"Variant of Concern (VOC)"[/nb] on R[nb]Rt[/b] and growth rate.

It gives this in absolute terms:

A 10% difference in VOC frequency [i.e. if 10% more of your cases were KentVid[nb]Based on S-gene-dropout - i.e. someone is tested and we find COVID Classic genes except for the spike gene, because it's too different to what we're looking for.[/nb]] in mid-November corresponds to approximately 50 more weekly cases per 100,000 [population] in early December.

[BlodwynPig]

The government has ordered us to look for rogue Pangolins in South West England.

ps. don't go bat watching for a while, yeh?

[jobotic]

Is this a fair summary up to now then?

https://www.theguardian.com/commentisfree/2020/dec/22/new-variant-coronavirus-genomic-sars-cov-2-pandemic

[Zetetic]

I suppose one thing I'd add is that it's thought we're now seeing this variant account for a decent proportion of positive test results across the UK, not just South-East England.

Estimated 14% of positives in Scotland for the week beginning 9th December.

Estimated up to 60% of positives in Wales as of the last few days.

[Bazooka]

Good news for vape heads:


2017 e400 0 60

[Chedney Honks]

Wake me up for Corono 3.0.

[Cuellar]

2017 e400 0 60

What's this carry on

[dissolute ocelot]

Wake me up for Corono 3.0.

Covid 4.0: Live Free or Covid Hard.

[jobotic]

Wake up. Hancock says it's here.

[steveh]

I see there's now a Welsh variant with a 501Y mutation that's also racing away:

https://nextstrain.org/groups/neherlab/ncov/united-kingdom?c=gt-S_501,69&d=tree,map,frequencies&f_country=United%20Kingdom&label=clade:20B&m=div&p=grid

Welsh variant in blue, Kent variant in green.

[Zetetic]

Has there been anything suggesting that this is spreading at the same rate as KentVid?

(I don't actually know if it produces the same S gene dropout.)