[cryoshon]

it's very early in terms of the data that's out there, but so far it looks quite bad.

the ELI5 is that it's expanding really fast, and it is probably at least somewhat resistant to our immune system's first line of defenses that are formed from vaccination or a prior infection. the details on all of those points are very much in flux still, so be aware that the story could change as we learn more.

in more detail:

the biggest piece of worrying evidence is that it's growing as a proportion of cases in south africa at such an intensely fast rate that it's outcompeting the delta variant as though it isn't even there, which means it's far more transmissible than any other variant.

as far as i know, there isn't any information on whether it has higher lethality compared to delta, but my hunch is that it does.

the reason why i think it will end up having increased lethality is that it appears to have a complex of mutations that are associated with resistance to antibodies, likely including those generated by vaccination. that will make it take longer for the immune system to form an effective response to infection, allowing the virus to replicate rampantly and cause organ damage for longer.

keep in mind every other variant has also exhibited this property at least to a small amount, and in most cases, regardless of vaccination, our bodies are still able to clear the infection.

the issue is that this time around, the variant doesn't have just one or two tropisms associated with lower antibody binding efficiency, but perhaps a couple dozen. so, it'll likely be more effective at reinfecting people who have already had covid, too. and even though the variant may be "vaccine resistant", i'd bet that vaccinated people will still fare better than unvaccinated people. unfortunately, our antibody therapies probably won't be as helpful at saving people who are hospitalized.

but, our antiviral therapies (including those still in development) shouldn't be any less effective against this variant, which is a very faint silver lining.

in conclusion: buckle up, we're probably in for a rough ride.

[whimsicalism]

> biggest piece of worrying evidence is that it's growing as a proportion of cases in south africa at such an intensely fast rate that it's outcompeting the delta variant as though it isn't even there, which means it's far more transmissible than any other variant.

OTOH - cases in SA were very low, so we would see this sort of growth of a variant very quickly just due to founder effects without anything nefarious going on.

Delta arrived when the background number of cases was much higher so it took longer to become a high proportion of cases.

[baq]

cases in SA are low because the testing infrastructure there has been complete garbage. real numbers are unknown.

[_djo_]

That's objectively false, seeing as though the test infrastructure has been able to detect hundreds of thousands more cases during the country's waves. If your claim was accurate, South Africa's case rate would have plateaued at a level and just stayed there.

Sure, the testing infrastructure is not as good as in a developed country, but it has many orders of magnitude more capacity than the current case rate.

[dredmorbius]

If testing is limited but effectively randomly sampling, then measured incidence will be proportionate to actual incidence.

If testing is targeted toward regions of higher interest and likelihood, then test positivity should skyrocket out of proportion to actual positives. (I'm not sure what the case is for South Africa.)

Using deaths as a lagging indicator of cases, South Africa reports about a 3% CFR (based on reported cases and deaths), as opposed to about 2% for the US. This would suggest a somewhat lower testing prevalence in ZA, by about a third, but not an especially bad record. This does assume that Covid deaths are being accurately assessed and reported. Total excess mortality is the usual check for that.

My read is that ZA's testing infrastructure is reasonably good, and that the B.1.1.539 variant's growth is extraordinary.

I'm relying on Worldometers data:

https://www.worldometers.info/coronavirus/country/south-afri...

https://www.worldometers.info/coronavirus/country/us/

[whimsicalism]

But have they become relatively more garbage in recent weeks? If not, then the relative number of cases is much lower.

[onlyrealcuzzo]

> the biggest piece of worrying evidence is that it's growing as a proportion of cases in south africa at such an intensely fast rate that it's outcompeting the delta variant as though it isn't even there, which means it's far more transmissible than any other variant.

Can anyone ELI5 how this is working?

If the Covid #s SA is reporting are accurate, the entire country is only at around ~2.5k cases per day.

Since Delta has an R0 between 6-7, for this to be outcompeting it so substantially, it would need to have an R0 of 8-9 (if not higher) -- at which point it would be almost as contagious as Measles. For it to be this contagious, wouldn't there already HAVE to be ~10k+ cases per day in SA?

The original Covid had an incubation period of 5.4 days. Delta dropped to 4.

If this has been around for weeks, with an incubation period of 4 days, shouldn't this have already infected close to ~100k people? And shouldn't there be 10s of thousands of infections per day?

To be fair, the growth rate South Africa IS reporting is 10x in 4 days. If that trend continues for even three weeks, then it would infect the entire country...

Does anyone know how reliable South Africa's #s are?

[ericmay]

> Does anyone know how reliable South Africa's #s are?

This is what I was going to comment on as I was reading your post but glad you brought it up here at the end.

I would have a difficult time believing the COVID numbers in South Africa are being reported reliably (regardless of the reason).

SPECULATION:

I'd love to know more if this is incorrect but I think even in countries like the United States, or Denmark, or Germany, or wherever the numbers are likely to be undercounted based on people just getting sick and not doing anything about it. My intuition is that numbers in countries like South Korea, Singapore, and perhaps Israel are more likely to be closer to the "ground truth". Other countries in Asia I have less confidence in (Japan and China). We probably need to develop and deploy more rapid at-home testing.

[_djo_]

There's always going to be some undercounting, as in any country, but in terms of being able to test enough cases and be consistent over time South Africa's numbers are reliable.

During earlier waves the testing infrastructure has detected orders of magnitude more cases, and the current lull in cases is following what epidemiologists predicted and expected before a fourth wave in December/January.

[whimsicalism]

I think it is almost certain that China has successfully controlled the virus up til now.

[ericmay]

I’d say sure it’s under control, I question the infection numbers and deaths. Then again maybe they’re posting the real numbers because they’re so good. But I’m not sure I buy that.

[whimsicalism]

I think if infection numbers were wildly off we would be able to see it in antibody tests of people coming from China.

Given that tons of people were wearing masks even prior to the pandemic, I don't find it that ridiculous.

[ericmay]

Sure. I think with China and COVID-19 so far my stance is guilty until proven innocent. Not really getting a lot of antibody testing done from people who don't travel or who have died. The CCP is inherently incentivized to fake numbers, underreport (this is a China-wide problem and you can see it manifest in the debt crises unfolding there), and downplay any problems so that the CCP looks strong.

And this has nothing to do with being pro or anti-China. I think it's just an obvious recognition of incentives and current state.

[VectorLock]

>the issue is that this time around, the variant doesn't have just one or two tropisms associated with lower antibody binding efficiency, but perhaps a couple dozen.

Can you share a source for this? I'd like to hear more.

[cryoshon]

general source to understand the phenomenon in prior variants: https://www.frontiersin.org/articles/10.3389/fimmu.2021.7571...

mutation profile of this particular variant: https://assets.publishing.service.gov.uk/government/uploads/...

summary of the phenomenon in layman's terms: when binding at 100% efficiency, antibodies bind to portions of the spike protein like a key fits into a lock, so it's easy to "unlock" the lock, thereby neutralizing the viral particle. when there are mutations which affect the shapes of the different portions of the spike protein, it's like the pins in the lock shifting so that the key you used before might not work without quite a bit of jiggling, assuming you can get it to unlock the lock at all. the more mutations that increase the amount of jiggling it takes, the more the pins in the lock become intractable with the key you have.

this set of analogies is imperfect in a few ways, but hopefully it helps you to understand the gist of the problem.

[edmcnulty101]

That sound really bad.