I don’t think it can be a rule, trivially so even, but maybe, like, 65% of the time? Not that it would make it a useful thing, just a curiosity.
Important to remember that each copy of the virus doesn’t/can’t give a fuck what happens to any other copy of the virus. It is just driven to produce offspring in the most efficient manner it can.
There is no global virus conspiracy.
Human rabies cases are almost entirely spillover cases from asymptomatic reservoir species, isn’t it? So it’s not a very useful comparison.
What about smallpox? Even in the 1700s the Swedes had a 0.7% population death rate from it on the peak year. And frequent numbers for North American natives are 30% of population.
Is there something different between Smallpox and the common cold? Or is the common cold not dangerous simply because we have familiarity with it? We changed, not the viruses.
Houngan
1914
Virulence (how bad it is for the host) and transmission rate are dance partners. Smallpox killed fast but spread hard, weeping sores and a primitive society. The cold just has a much better party trick, easy to transmit but benign enough we don’t do anything to prevent it. Viruses aren’t reaching for perfect transmission or massive virulence, they just faff around until something works well enough that new variants aren’t as effective because the mutation makes them less viable than the main variant, generally speaking. Sometimes they get lucky, though, and probably the best luck is to jump species, where the mutation strain that let them do that is offset by a massive new host pool, so it’s effectively bereft of competition at the start.
abrandt
1915
I don’t know that I’ve seen anything on this topic, but I’d be curious if the common cold caused by a coronavirus was originally as serious as the one that causes Covid but eventually mutated into a less dangerous form?
The better example I suppose is Influenza. There aren’t many viruses that have mutated as much as that one, and it still kills scores of people every year. In 1918, the second wave, which is thought to be a variant on the first, was far deadlier. If virus variants always became less lethal over time, then you wouldn’t have the kinds of Flu death numbers we have every year.
In general I have no problem believing that virus variants tend to tilt towards higher propagation and lower lethality, but it’s quite evidently not a rule. And really, I’ve never seen evidence to indicate that it’s even a genuine phenomenon.
Houngan
1917
All life and pseudo-life, which viruses fall mostly into, are about success at reproduction. That’s the end game for any self-replicating entity, including clay. The things we consider life just change much, much faster and can try out many more options against an equally variable host structure. You make the world warmer, the birds get smaller.
JonRowe
1918
I mean, in general, they do though. Flu death numbers have a lot to do with underlying conditions etc. There will always been variants year to year that are worse than others, but in general fitness is about one thing and one thing only, replication. There will always be exceptions to this rule, but those mutations go away. Killing the host does not help with transmission, unless the deceased host can infect others, even then, there is a limit as a deceased host isn’t moving around.
Of course, COVID-19 could be an exception, but in general mutations favor transmissibility over all else. It is important to remember that COVID-19 is a particularly non-deadly coronavirus (when compared to the other SARS viruses of the past) which has helped it become a worldwide pandemic.
We can just hope that COVID-19 continues the typical trend of increasing transmissibility without any other mutations, as the human population gains immunity levels through vaccine and infection.
jsnell
1919
Report on hospitalizations in South Africa (600+ hospitals), with data from last Saturday:
This seems likely to be the most we’ll find out about severity this week, and in particular has a ton of data bucketed by age. There’s no data bucketed by vaccination / previous infection status. Key points:
- Figure 10 shows the cases vs. admissions in Gauteng over time. The shape for the latest wave is visibly different, but not an order of magnitude different.
- Figure 12 shows 90% of the hospitalizations being for Covid symptoms, 8% incidental.
- Figure 13 shows the age breakdown in comparison with the previous waves. It had been previously reported that a lot of the patients were young kids. It had not been contextualized just how few kids were hospitalized in the previous waves.
- Figure 14 shows the severity of the cases by age in the early parts of each wave. It suggests about a 50% reduction in severity vs. Delta in the same age bracket, but still has the caveat that the level of immunity is far higher now than six months ago (the start of Delta would basically predate all the vaccination and half the cases). Figure 15 shows the same for deaths, which is about a 70% reduction.
If you accept the idea that ZA has basically complete population immunity at this point, this looks to me like the protection given by double-vaccine/previous infection against severe cases really is only about 50% which would be bad. If you think they’ve still got a substantial un-vaccinated population, then this would probably be good news.
- Figure 16 shows that out of the people who were released from the hospital within the first 25 days of the wave, far fewer were severe cases. I don’t really understand this and figure 17. Why would we expect severe cases to be released so quickly? This is a number I suspect will be quoted a lot, since it’s the most optimistic view.
Seems to me it’s closer to 66% reduction if you assume everybody has been infected/vaccinated previously -and weren’t in the previous waves-).
Note that this is about the worst assumption possible and still gives 66% protection against severe disease (although there’s a big caveat with the data, see below).
I also want to see this data split between vaccinated and unvaccinated, since previous infection protection and vaccine protection might not be equal.
It’s not that they are released quickly, only that there are fewer severe admissions than non severe admissions, contrary to previous waves.
As for protection vs Delta we can do some preliminary analysis.
For now we don’t know protection against infection by Omicron offered by vaccines/previous infections. Logic says that it should be lower than for previous strains due to the mutations at play. We know protection against Delta wanes to less than 50% after 6 months, so that’s actually quite low too.
Of those infected protection against death by Delta (after 6 months) was about 83% on average for <65 yo and 72.5% for >65 yo. Using the percentage of severe cases above, one guess is that protection against severe infection by Omicron for over 60s is 33,5%, while for under 60s is 71% (comparing against the previous waves).
Lots of caveats (the biggest one being that we might be seeing a lower percentage of severe admissions just because before non-severe cases weren’t being admitted in previous waves, because they were hitting hospital saturation -if admission criteria is radically changed, there’s little we can gather from this-; we don’t know if the severe infection/death ratio for Omicron will hold the same as for Delta… etc) but it seems like long term (past 6 months) vaccines/previous infection protection might be still decently effective against Omicron (compared to against Delta) for under 60 year olds, while it is significantly reduced for older cohorts.
A very good thread that explains that vaccination and previous infection can offer more protection from severe disease than might be otherwise apparent on first look – but also which explains why Omicron can be every bit as virulent/dangerous as Delta, too…and not necessarily milder.
I think it’s looking more probable that persons with a previous immunity to SARS-CoV-2 (previous infection or vaccination) still get some reasonably high degree of protection from serious disease…but that Omicron still will pack a dangerous, hospital-overloading wallop when it hits population centers that lack this immunity in great numbers.
It’s either some fairly significant degree of that…or as things currently sit, Omicron isn’t as virulent. Which I remain pretty skeptical of.
The numbers in RSA continue to look good from this perspective:
This still may be a too-optimistic data statement, but it’s worth a read nonetheless.
jsnell
1923
The table you’re looking at is the most skewed data set that they give. It contains the results for all patients with an outcome. For the earlier waves, everyone has an outcome by now, no matter the severity. For the Omicron wave, we only have an outcome for the people who were released very quickly, i.e. predominantly mild cases.
(That’s different from figure 16, which normalizes for this effect by not looking at all past cases but only the ones from the start of each wave, .)
Indeed, but we have to take what we can get. It’ll be a week before we get more from ZA, I guess the best bet for age+vaccination data before then is the UK or Denmark.
Figure 16 is explicitly only including patients with an outcome. I.e. released from hospital or died. Very few people will have had time to die, so it’s predominantly people who were released quickly.
I think Figure 14 is the right one to reason about even though we need to read the graph for the numbers. It’s stratified by age, normalized in a useful way, and doesn’t have the confounders from early outcomes.
The version of the data that just looks at the first 25 days of the data controls for that naturally, the hospitals would not have been at capacity for any of them.
Not disputing your back of the envelope guess, the results will depend a ton on the assumptions. But 50% protection from double-vaccination against severe Delta six months out seems pessimistic.
jsnell
1924
Scotland, the biggest growth graphs to date:
Denmark is now reporting Omicron hospitalisations. 1.4%, which seems a bit lower than their overall ratio of hospitalisations to cases (1.7%). No breakdowns on e.g. vaccinations, the numbers are probably too small still.
https://files.ssi.dk/covid19/omikron/statusrapport/rapport-omikronvarianten-10122021-ek56
More good news regarding immunity and fighting off severe disease/outcomes from Omicron:
jsnell
1926
I wasn’t expecting a UKHSA briefing this week due to the mini update they published, but they released one after all. Lots of new data there, but the big one is early vaccine effectiveness data against symptomatic infection. Better than I expected for two doses, worse than I expected for three.
Full report:
Yep, and still seeing that either protection from serious disease or less virulent. (I’m betting the former, that it’s hitting vaccinated or previously infected populations first.)
Which is good news, but as the technical briefing states: a small percentage of a very large number is still a large number.
Can someone ELI5 that chart; how can vaccine effectiveness be negative (does it mean they’re MORE prone to getting it?), and what is “2+” in the time since vaccine (weeks) axis when 2-9, etc. is already there?
abrandt
1929
Pretty sure the 2+ at the end is referring to time after the booster shot, whereas the first 5 data points are all related to original course of vaccine.
