Scrutable

From the Rod Liddle thread but more suitable here (if already being covered, mods please move):

Lew Dolby wrote: Sat Dec 12, 2020 10:30 pm But, AIUI, there's no evidence (yet ??) that the vaccines do anything to stop people being spreaders of CoViD.

Anyone with expertise clarify that or not ??

Random news story: https://www.cbsnews.com/news/covid-19-v ... nfectious/

If you've had the jab, you can still get infected — the vaccine simply stops you getting ill. Am I getting that right? And vaccinated people can still pass it on to others, is that right? So the small number of people who for one reason or another can't be vaccinated will have to shelter forever while the rest of us are partying away?

You don't think maybe there's already been extensive discussion on this in "Developing the Covid-19 vaccine" and it might be worth reading the thread?

Brightonian wrote: Sun Dec 13, 2020 12:33 pm From the Rod Liddle thread but more suitable here (if already being covered, mods please move):

Lew Dolby wrote: Sat Dec 12, 2020 10:30 pm But, AIUI, there's no evidence (yet ??) that the vaccines do anything to stop people being spreaders of CoViD.

Anyone with expertise clarify that or not ??

Random news story: https://www.cbsnews.com/news/covid-19-v ... nfectious/

If you've had the jab, you can still get infected — the vaccine simply stops you getting ill. Am I getting that right? And vaccinated people can still pass it on to others, is that right? So the small number of people who for one reason or another can't be vaccinated will have to shelter forever while the rest of us are partying away?

We don't know because we haven't studied it yet.

https://www.businessinsider.com/vaccine ... 20-11?IR=T

You can't just vaccinate potential "super spreaders" since you don't know who they are so you'd end up just having to vaccinate everyone, and the plan is to vaccinate everyone eventually anyway.

But generally since covid has an R_0 of maybe about 2 (i.e. not 10) you can achieve herd immunity relatively easily such that the virus just can't circulate in the population anymore.

It wouldn't be so bad if we could anyway normalize hand-washing, mask wearing in certain situations, and staying home if you have general cold/flu symptoms.

shpalman wrote: Sun Dec 13, 2020 12:59 pm and staying home if you have general cold/flu symptoms.

All those arguments I had with my old management and colleagues about taking days off when someone “just had a cold” are utterly vindicated. My line was always “it might just be a cold to you, it could be fatal to someone else”.

Sniff Petrol once mocked up a fake Lemsip advert which said something like "Your job's on the line so get back to work motherf.cker".

Any update on this? What impact will the vaccine have on R0?

Boustrophedon wrote: Thu Dec 17, 2020 1:06 am Any update on this? What impact will the vaccine have on R0?

That depends on which vaccine you are using. In theory, a vaccine must protect over 1/R0 of the population to cause a disease to die out where "protect" means that a vaccinated person cannot infect anyone else. From https://en.wikipedia.org/wiki/Basic_reproduction_number we find that COVID-19 has R0 >= 3.28. That means a vaccine must protect 70% of the population. That can be a vaccine which is 70% effective administered to everyone, or a 90% effective vaccine administered to 78% of the population, or a 95% vaccine administered to 74% of the population, or a 100% effective vaccine administered to 70% of the population. Note that getting R < 1 is very often stated as a goal, but that is not really adequate. If you get R=0.99 and infection takes two days (i.e. the 1000 cases turns into 990 after two days) then it takes just over 19 weeks for the number of cases to be halved. However, if you can get R down to 0.5, then the number of new cases halves every two days.

Of course that's an over-simplification for real-world use, because we can use vaccination and social distancing simultaneously, thus reducing the required net effectiveness. Based on what I have seen happening so far, what will happen as more and more people get vaccinated is that people (quite likely including those not yet vaccinated) will use this as an excuse to relax measures to the extent that we see a succession waves of infection as people relax, drive up infections, panic and get stricter, driving down infections, and then get complacent and we're back to relaxing.

And just to add more complexity, preventing someone being infected, infectious, or sick are three different things. A hypothetical vaccine which prevented someone getting sick would still be very useful even if it provided no barrier to being infectious - in principle we could administer it to the whole population and then the disease woul be harmless so we could ignore it and let anyone catch it. This is, of course, extremely unlikely as the COVID-19 vaccines are designed based on how infection works, so are very likely to prevent infection or infectiousness.

Millennie Al wrote: Thu Dec 17, 2020 3:41 am That depends on which vaccine you are using. In theory, a vaccine must protect over 1/R0 of the population to cause a disease to die out where "protect" means that a vaccinated person cannot infect anyone else. From https://en.wikipedia.org/wiki/Basic_reproduction_number we find that COVID-19 has R0 >= 3.28. That means a vaccine must protect 70% of the population.

Are you sure 1/R0 is the right formula? That would suggest that if the virus was super-spready with an R0 of 10, you would only need to vaccinate 10% of people. Maybe you mean (1-(1/R0))? (I'm guessing so, from the 70% number.)

R_eff=R₀P_i Where P_i is the probability of becoming infectious post exposure.

Let U_p= the fractional uptake of the vaccination.
Let E be the probability the vaccination works in you, IE the effectiveness of the vaccine.

P_i=1-(U_pE) that is, 1- (probability of taking up the vaccine and it working)

Let R_eff=1, We actually want R_eff less than one but I'm not good with inequalities.

Then 1=R_o(1-U_pE)

rearrangement gives: U_p= (R₀-1)/(R₀E), U_p has to be greater than this.

So if E=90%, R₀=3.28 The U_p>0.77. IE better than 77% So we're still f.cked.
.

Seems like we'd need some of the less invasive measures to remain in place to reduce R anyway.

I am, by the way, slightly more relaxed about that. Obviously we'll need population level data but indications so far are that the vaccines greatly reduce either the severity of the infection or reduce the probability of developing a severe infection; this will mean that the covid which will continue to circulate amongst the 10% of the 77% who have been vaccinated but in whom the vaccine is not efficacious will be milder than it otherwise would have been.

My attitude towards individuals in the 23% who aren't vaccinated depends on the reasons for why they are not vaccinated.

A far bigger issue would be that the UK won't even reach anywhere near 77% if it only has a plan to perform a million jabs a week during 2021.

sTeamTraen wrote: Thu Dec 17, 2020 3:26 pm

Millennie Al wrote: Thu Dec 17, 2020 3:41 am That depends on which vaccine you are using. In theory, a vaccine must protect over 1/R0 of the population to cause a disease to die out where "protect" means that a vaccinated person cannot infect anyone else. From https://en.wikipedia.org/wiki/Basic_reproduction_number we find that COVID-19 has R0 >= 3.28. That means a vaccine must protect 70% of the population.

Are you sure 1/R0 is the right formula? That would suggest that if the virus was super-spready with an R0 of 10, you would only need to vaccinate 10% of people. Maybe you mean (1-(1/R0))? (I'm guessing so, from the 70% number.)

You're right steamy, it's (1-(1/R0)) rather than 1/R0 for the simple herd immunity threshold. Discussion of vaccine effectiveness, nonrandom mixing, and nonrandom vaccination in the linked article.

Fine, Eames, Heymann: "the simple herd immunity threshold for random vaccination of Vc = (1−1/R0), using the symbol Vc for the critical minimum proportion to be vaccinated (assuming 100% vaccine effectiveness)" https://academic.oup.com/cid/article/52/7/911/299077

Of course, that's only relevant once we've checked whether the vaccine provides sterilising immunity and actually stops people getting infected rather than just stopping them getting ill. As it stands:

shpalman wrote: Sun Dec 13, 2020 12:59 pm We don't know because we haven't studied it yet.

This is a nice explanation.

At least, having no specialist knowledge of my own, I thought it was nice when I read it. YMMV.

https://fivethirtyeight.com/features/on ... ct-others/

bolo wrote: Fri Dec 18, 2020 7:57 pm This is a nice explanation.

At least, having no specialist knowledge of my own, I thought it was nice when I read it. YMMV.

https://fivethirtyeight.com/features/on ... ct-others/

I'm not entirely sure how the virus can stay in your nose on the other side of the mucous membranes as all the actual cells, so that it doesn't infect you but your immune system can't get to it, but still "might be able to replicate". And the scientist quoted reckons anyway that "there will be enough antibodies made that will neutralize the virus even at the mucosal surface" so he doesn't believe that mechanism either.

Also, comparison with the 'flu vaccination seems unhelpful since that one is known to be only 40-60% effective, not 90-95%.

Re: https://fivethirtyeight.com/features/on ... ct-others/
where Deepta Bhattacharya is quoted:

If you have a vaccine that’s 95 percent effective at reducing symptoms, there is no universe in which it wouldn’t also reduce the likelihood of transmission. It’s just not possible,

is clearly wrong. It is a very strong claim, and we know that some people are asymptomatic but infectious so, even though it might be so unlikely that we can ignore it, it is still theoretically possible that a vaccine might turn every case into one of these asymptomatic cases which would have no effect on the likelihood of transmission. Given the rest of the article, I suspect that something was lost in transmission as the message passed from expert, via journalism, to us. The same must have happened to the claim about the virus reproducing without entering any cells - the inability to reproduce without using another cell is pretty much the definition of a virus.


And yes, further upthread I did mean (1-(1/R0)) but left out the "1 -" by mistake. And I agree with Boustrophedon's conclusion.