Covid-19 Gambles

[echoes]

As is this https://www.nature.com/articles/d41586-020-02801-8.

Again, a very common theme in these reviews is the citation of studies which are inconclusive and/or have acknowledged flaws.
Every RCT of masks worn by the general public has failed to confirm any significant effect. And these are the studies which actually count, and that actually measure the only variable of interest.

 

[echoes]

The early parts of the rises and the late parts of the falls have been essentially exponential, but the best fit to the (important) bit in-between has been essentially linear (both up and down). I don't think anyone yet fully understands this, but it has presented on of the headaches for modellers.

Michael Levitt has done a lot of work on this, and yes, it is a very hard thing to model!

 

[JohnW2]

Varients of concern tend to appear in immunocompromised people*, but this is not set in stone and is not a very well understood area ....

Interesting - but I find that rather surprising (aka difficult to understand!).

Mutation surely occurs as an essentially random event when a virus is replicating within a host cell and, as such, I don't see how it is influenced by the host's immune status. I can understand that immunocompromised people are at increased risk of having virus replicating within their cells, but such replication is occurring all the time in hundreds of thousands of people who are not immunocompromised - and, as above, once replication is occurring within someone's cells, I don't really understand why their immune status should influence the likelihood of a mutation occurring.

I attempted to contact you by PM about this, not the least because I was wondering whether I might be able to 'pick your brains' on technical issues which would presumably bore to tears most of those in an 'Electrics' forum However, it seems that you don't accept PMs.

Kind Regards, John

 

[JohnW2]

Michael Levitt has done a lot of work on this, and yes, it is a very hard thing to model!

It's easy enough to ('empirically') mathematically model what has been observed (in the UK) but the worry is that because we don't really understand why it has been as it has (and what factors have brought it about), we cannot be very confident that such empirical models will remain valid for future forecasting (since the unknown 'factors' may change).

It has been suggested above that one does not need to bring theoretical considerations into modelling when one has 'actual data' to work with. That is a mathematician's viewpoint, but assumes that everything in the future will remain the same as it was during the past period which resulted in the 'actual data' which we have to work with today. In many ways that is the basis of what I was writing about in my OP - our 'gamble' relies on the effects of ongoing further deployment of vaccination but, without invoking some theory and/or a crystal ball, the data we currently have cannot tells us for sure what will be the effects of "further deployment of vaccination".

Kind Regards, John

 

[JohnW2]

Interesting. Worldometer data that I was plotting at the time showed a clear peak in the rate of increase on or around 18 March.

I can't speak for Worldometer data but, as you can see, the WHO and Johns Hopkins University data are essentially identical ...

Kind Regards, John

 

[xm77]

What about the suggestion that it is the restrictions that are driving the evolutionary pressure on the virus to become more severe and or become more vax -evading.
Those variants which cannot survive whatever restriction (lockdown/vax etc) will die off leaving the more resistant strains to survive. Had there been no restrictions then maybe the more severe strains will/would not gain the upper hand.

Think of it like the antibiotic tank experiment shown here.

The ever higher levels of antibiotic are driving the evolution of the bacteria to become ever more resistant to it and it is the very fact that the restrictions are not 100% which are driving this.
Had our lockdowns been 100% effective which is impossible the virus would be exterminated by now. So what about the suggestion that the part-vax'd and those for whom the vax is very much not 100% are the ones in whom a new vax evading variant is likely to grow if they catch the bug and not the totally unvaxed. Perhaps we should be quarantining people between their two vax doses!

Also what about the possibility that the new variant is now SO infectious that any practically achievable lock down rules will not control it (ie the Ro figure is in the 6 to 10+ region) and nor with such a Ro figure is herd immunity by vaxination possible given the vax's likely real world efficiency even with all adults vax'd (which ain't going to happen either) so there is no real further point anymore in having any restrictions.

 

[Swwils]

Interesting - but I find that rather surprising (aka difficult to understand!).

Mutation surely occurs as an essentially random event when a virus is replicating within a host cell and, as such, I don't see how it is influenced by the host's immune status. I can understand that immunocompromised people are at increased risk of having virus replicating within their cells, but such replication is occurring all the time in hundreds of thousands of people who are not immunocompromised - and, as above, once replication is occurring within someone's cells, I don't really understand why their immune status should influence the likelihood of a mutation occurring.

I attempted to contact you by PM about this, not the least because I was wondering whether I might be able to 'pick your brains' on technical issues which would presumably bore to tears most of those in an 'Electrics' forum However, it seems that you don't accept PMs.

Kind Regards, John

Essentially if a beneficial mutation occurs in a healthy person there is a higher likelyhood that their immune system deals with it before it has a chance to infect someone else.

Often this can be a patient being treated with rituximab, or something else that depletes antibody-producing B cells - which result in long term immune dampening or anyone with low virus neutralization titres.

Memory b cells are very interesting. After they are generated after initial infection they don't just sit around - they undergo somatic hypermutation. For months and years after an infection, memory B cells hang out in your lymph nodes, and their genes that code for antibodies acquire mutations themsleves. These mutations result in a more diverse array of antibodies with slightly different configurations.... Which is why the human immunity responce range is up in the quintillion district antibodies when your genome only has 20,000 or so genes.

If total knowledge of human immunity was a marathon, we have done about 3 meters.

 

[Swwils]

What about the suggestion that it is the restrictions that are driving the evolutionary pressure on the virus to become more severe and or become more vax -evading.
Those variants which cannot survive whatever restriction (lockdown/vax etc) will die off leaving the more resistant strains to survive. Had there been no restrictions then maybe the more severe strains will/would not gain the upper hand.

Think of it like the antibiotic tank experiment shown here.

The ever higher levels of antibiotic are driving the evolution of the bacteria to become ever more resistant to it and it is the very fact that the restrictions are not 100% which are driving this.
Had our lockdowns been 100% effective which is impossible the virus would be exterminated by now. So what about the suggestion that the part-vax'd and those for whom the vax is very much not 100% are the ones in whom a new vax evading variant is likely to grow if they catch the bug and not the totally unvaxed. Perhaps we should be quarantining people between their two vax doses!

Also what about the possibility that the new variant is now SO infectious that any practically achievable lock down rules will not control it (ie the Ro figure is in the 6 to 10+ region) and nor with such a Ro figure is herd immunity by vaxination possible given the vax's likely real world efficiency even with all adults vax'd (which ain't going to happen either) so there is no real further point anymore in having any restrictions.

Fortunately covid19 is a virus and antibiotic resistance is not a factor.

Also being a Coronavirus, its capacity for varients is slightly less concerning than say the flu, which has strains. (Depending on who you speak to).

Yes the new varient is very infectious and that will have a basic effect on control measure effectiveness balance. Some data also shows that's seasonality is far less impactful (5%) than hoped(10%).

We have experience with very contagious disease. The measles. Fortunately the vaccine is very effective, even with that vaccine we still see sporadic outbreaks in the UK - hence why vaccine uptake needs be so high to truly eliminate such diseases when other controls measures are not in place. (Even then some outbreaks might not be avoidable).

 

[JohnW2]

Essentially if a beneficial mutation occurs in a healthy person there is a higher likelyhood that their immune system deals with it before it has a chance to infect someone else. .... Often this can be a patient being treated with rituximab, or something else that depletes antibody-producing B cells - which result in long term immune dampening or anyone with low virus neutralization titres. ....

Fair enough, thanks.

So, if I understand correctly, you're not suggesting that (as I said, which would have surprised me) mutation is more likely in an immunocompromised person but, rather, that a person with normal immunity would be more likely to 'deal with it' (deal with any variant) before it had a chance to infect someone else. Is that correct?

If so, I still find it hard to understand, numerically, your suggestion that variants are more likely to originate in an immunocompromised person. You presumably aren't suggesting that people with normal immune systems are more able to 'deal with' new variants than with previous ones (or that mutations are less likely to arise within their cells)? For every immunocompromised person, there must be tens or hundreds of thousands of other people who are not immunocompromised but whose immune systems are nevertheless not able to deal with the virus(es) within them (of whatever variant) before they infect someone else. So, even if the immunocompromised are more likely to 'let out' a new variant which arises within them, I would still expect the great majority of new variants to originate in people with normal immune systems. Am I missing something?

Whatever, I still think that it must remain the case that the chances of a 'nasty' variant emerging within any population in any given time period will be greatly increased if there are large numbers of infected people around - do you agree with that?

Memory b cells are very interesting. .... If total knowledge of human immunity was a marathon, we have done about 3 meters.

Probably a bit more than 3 metres. Even though we have a very long way to go, 'we' have moved on a lot - in my youth (longer ago than I care to recall ), "Immunolgy" was a couple of lectures tagged onto the end of a totally different course (and much the same was true of "Virology"!).

Even in a forum famous for 'thread drift', I do fear that this discussion may have slid in technicalities which might well be pushing forum members' tolerance to its limits! If you would feel inclined to move it into some other medium (e.g. e-mail), just send me a PM (which I do accept!).

Kind Regards, John

 

[Swwils]

Not a simple one to answer in a Coronavirus, as always!

We currently think the relatively slow mutations rate of C19 is because it has some ability to proofread newly made RNA copies, which is unusual for a RNA virus.

In any case, we may never know the true mutation rate since so many of them make the virus unviable and we never know about them.

Instead we measure a fixation rate over time. So yes, the more a virus spreads, the more opportunities it has to replicate, the higher its fixation rate will be, and the more the virus will evolve.

For C19 this is around one mutation every 10 days. Which is why when the Kent varient appeared with 26 mutations, it was proposed that such circumstance was from an immunocompromised person.

The good news is coronavirus are one long string of RNA. So possess no physical segmentation to undergo reassortment like the flu can.

 

[JohnW2]

We currently think the relatively slow mutations rate of C19 is because it has some ability to proofread newly made RNA copies, which is unusual for a RNA virus.

Intriguing.

In any case, we may never know the true mutation rate since so many of them make the virus unviable and we never know about them.

That's presumably just a general truth in relation to all viral mutation, isn't it? However, I wouldn't think that really has much practical (rather than theoretical) importance, does it, since I would have thought that our primary (practical) interest is only in variants which are not only viable but which are already spreading in the population?

Instead we measure a fixation rate over time. So yes, the more a virus spreads, the more opportunities it has to replicate, the higher its fixation rate will be, and the more the virus will evolve.

Fair enough.

For C19 this is around one mutation every 10 days.

What does that mean - that the replication of one virus particle will result in a mutation about every 10 days, that a mutation will arise in the 'entire population' of the viruses in about 10 days, or what?

... Which is why when the Kent varient appeared with 26 mutations, it was proposed that such circumstance was from an immunocompromised person.

I'm afraid that you have rather lost me there! Are you suggesting that it is thought that an immunocompromised patient survived for at least about 260 days (about 8.5 months) whilst the virus was replicating within them (and then started transmitting this variant with 26 mutations)?

Kind Regards, John

 

[JohnW2]

Again, a very common theme in these reviews is the citation of studies which are inconclusive and/or have acknowledged flaws. Every RCT of masks worn by the general public has failed to confirm any significant effect. And these are the studies which actually count, and that actually measure the only variable of interest.

Maybe, but I don't think that does much to move my "better than nothing" gut-feeling (unless someone can present me with convincing evidence that wearing masks can cause 'harm').

I don't think that you should read too much into the RCTs which have not shown (statistically) 'significant' effects, since many of them have far too low a statistical power (i.e. have been too small) to show 'statistical significance' unless the effect (of masks) were very large (which I don't think any of us believes to be the case).

Faced with a plethora of inadequate (too small) trials that have failed to show a 'statistically significant' positive effects, one can attempt some sort of meta-analysis, which at its crudest informal form would involve just looking at the distribution of results from the various trials.

If there were truly 'no effect' of wearing masks, one would expect the results of the various trials to be fairly symmetrically distributed around 'zero effect'. Hence, if there were truly 'zero effect' but some trials showed small/modest (but not 'statistically significant') results in favour of mask-wearing (less infections in mask-wearers), then one would expect there to be a similar number of trials with similar small/modest results in favour of not wearing a mask. If most of the trials show small (but not significant) differences in favour of masks, but few (or none!) show similar difference in favour of no-mask, then this would be fairly suggestive that there is a true, even if small, beneficial effect of mask-wearing.

RCTs are, of course, not without their problems in this situation. For a start, it has not been possible to undertake them in the UK (or a good few other countries), since there has been a legal requirement to wear masks in at least what are deemed to be the highest risk 'contained' environments. However, even in countries which don't have such legal impediments, the trials will presumably be less than ideal (in relation to the general population) because the sample on which they are undertaken will probably not be very representative of the general population - one would imagine that those agreeing to take part in a study which meant there would be a 50% chance of their being asked not to wear a mask, even in high-risk environments, would tend to be less risk-averse than the general population, and hence may tend to behave in other 'higher risk' ways than the general population.

Kind Regards, John

 

[echoes]

Maybe, but I don't think that does much to move my "better than nothing" gut-feeling (unless someone can present me with convincing evidence that wearing masks can cause 'harm').

You are of course free to think with your gut, but I will remain unconvinced until there is evidence. How would you differentiate gut-feeling from superstition?

I don't think that you should read too much into the RCTs which have not shown (statistically) 'significant' effects, since many of them have far too low a statistical power (i.e. have been too small) to show 'statistical significance' unless the effect (of masks) were very large (which I don't think any of us believes to be the case).

We should attribute higher significance to the larger studies - agreed. The Danish mask study was large, quality RCT and found no significant link (and they believed they would find one). The Vietnamese study comparing cloth masks to clinical masks in a clinical settings showed cloth masks to increase the risk of infection - by a small margin, but it was a finding.

The claim that masks 'protect others' is bordering on the unfalsifiable (and therefore not a scientific claim) but fits very well with the propaganda campaign that we have been been consciously or, in most people - unconsciously, subject for the last 16 months.

There's a reason why the pandemic plan (abandoned amid panic in March 2020) advised against wearing masks (and lockdowns, and virtually every other NPI measures that the govt. has implemented) - it was that there was no evidence that it would would be effective.
I would also point out that the WHO changed its public stance on mask wearing due to political, not medical reasons.
I believe the real reason for mandating masks was:
a) to encourage the fearful to get out more in the belief that they would be safer.
b) "we must do something; this is something"
c) as a highly visible reminder that thing were not normal, so to justify all that had gone before and that which would come after.
d) to embed a sense of collectivism and easily identify and 'other' dissenters. There is evidence from members of the SPI-B committee that this is so. I'll dig that out in due course.
These reasons have everything to do with behavioural psychology and very little to do with public health.

 

[denso13]

but I will remain unconvinced until there is evidence

I posted this earlier https://www.pnas.org/content/118/4/e2014564118. Am I misinterpreting the conclusions? Genuine question.

 

[echoes]

I can't speak for Worldometer data but, as you can see, the WHO and Johns Hopkins University data are essentially identical ...

View attachment 238873

Kind Regards, John

I think this shows the same: the rate of increase had peaked before lockdown. Anyway, the determination of whether lockdowns significantly affected the 'case' rate (that's another tangent) or excess deaths (the only reliable metric) must be via comparison of societies which implemented lockdowns to varying degrees of strictness (including no lockdown) and conformity. There you will not even find correlation (let alone causation) to support the hypothesis that lockdowns work against an endemic disease.