covid-19

[RandBlade]

Can you please direct me to this evidence you keep mentioning?

I've seen a lot of reports and data but don't have it to hand sorry. If I come across it again I'll share it here.

Yes, to verify something like this works, I want a placebo controlled RCT. That's the gold standard, and that's what was used to provide the EUA in the first place. If a company thought that a less stringent dosing schedule (or just one dose) would do a good job, they would have included that as an arm in a trial. And in fact that's precisely what JNJ (and, belatedly, AZN) are doing.

Fine for a study, this isn't a study. This is real life.

I believe your friend was referring to 'off label' use; generally the manufacturers of a drug or device get them approved under a specific indication - to treat specific diseases in specific ways, up to and including dosing regimens. Certainly physicians are generally allowed to use an approved device or drug in a manner not explicitly on the label provided that their expert judgment determines that to be the best approach for that patient. It is not the same logic being used here, however. No one is claiming that this changing of dosing regimens will be better for the patient being withheld the second dose; at best, they will not be harmed, but at worst they will suffer substantial harm from reduced protection. The logic here has to do with public health policy, arguing that overall outcomes would be better if second doses were withheld at least temporarily. Generally doctors are not allowed to use items off-label if they think it will help someone else but not their patient.

Yes thanks, that was it. Off label makes more sense than off book.

Well in this case the JCVI aren't treating a solitary patient, they're explicitly seeking the best protection for across the community. If millions more are protected via this regimen following the logic and numbers I posted before, and if tens of thousands of lives are potentially saved as a result (given we face losing over a thousand a day currently some days that's not an exaggeration), then that seems reasonable to me. Does it seem absolutely unreasonable to you?

It is only by having this regimen that they will be able to offer at least one dose of vaccine protection to everyone in the most clinically vulnerable groups by the middle of February, if a different regimen were followed then it would potentially be the end of March before the same people were getting a dose - which would include for instance all those under 65 on chemotherapy (priority group 4).

[wiggin]

I've seen a lot of reports and data but don't have it to hand sorry. If I come across it again I'll share it here.

You asked for my data for the range of efficacy I quoted; I referenced specific studies. I'm asking you for supposedly voluminous evidence of >50% reduction in viral load and hospitalizations after 1 dose. It's kinda central to your argument.

Fine for a study, this isn't a study. This is real life.

I recognize that the UK's strategy does not lend itself to a random assignment, blinded, placebo controlled trial. That was what I was lamenting. You think that because it's an EUA it's somehow okay to go ahead with a huge dosing change without data. But those EUAs weren't granted in the first place without a large placebo controlled RCT precisely because you can't just make an educated guess about whether the vaccine will work in a given formulation/dose/etc.

I asked it earlier, but I'll pull it out here: In the absence of solid data, how do you know that this change to the dosing will result in a net improvement as opposed to some other change in dosing that is not being contemplated here? There are many ways to skin this cat: reduce the quantity dosed at each dosage but give the same number/timing of doses. Or you could just stick with 1 dose and skip the second entirely - certainly you can get even MORE people vaccinated quickly. Or you could mix and match vaccines to reduce the administrative burden. So how do you know this will work better than the method that is known to work... and that other changes aren't better? Which untested change do you embrace?

Well in this case the JCVI aren't treating a solitary patient, they're explicitly seeking the best protection for across the community. If millions more are protected via this regimen following the logic and numbers I posted before, and if tens of thousands of lives are potentially saved as a result (given we face losing over a thousand a day currently some days that's not an exaggeration), then that seems reasonable to me. Does it seem absolutely unreasonable to you?

I have no issue with this basic theory from a public health perspective. My issue is arguing that this falls under off-label use provisions, which are generally related to individual patients, not broader public health.

[RandBlade]

You asked for my data for the range of efficacy I quoted; I referenced specific studies. I'm asking you for supposedly voluminous evidence of >50% reduction in viral load and hospitalizations after 1 dose. It's kinda central to your argument.

https://assets.publishing.service.go...ter_14a_v6.pdf

I recognize that the UK's strategy does not lend itself to a random assignment, blinded, placebo controlled trial. That was what I was lamenting. You think that because it's an EUA it's somehow okay to go ahead with a huge dosing change without data. But those EUAs weren't granted in the first place without a large placebo controlled RCT precisely because you can't just make an educated guess about whether the vaccine will work in a given formulation/dose/etc.

We aren't operating with no data, we're operating with limited data. There's a difference.

The EUAs were granted because the placebo controlled RCT showed the vaccines were safe and efficacious. After that it becomes a question of how best to use it.

I asked it earlier, but I'll pull it out here: In the absence of solid data, how do you know that this change to the dosing will result in a net improvement as opposed to some other change in dosing that is not being contemplated here? There are many ways to skin this cat: reduce the quantity dosed at each dosage but give the same number/timing of doses. Or you could just stick with 1 dose and skip the second entirely - certainly you can get even MORE people vaccinated quickly. Or you could mix and match vaccines to reduce the administrative burden. So how do you know this will work better than the method that is known to work... and that other changes aren't better? Which untested change do you embrace?

This change is based on evidence that it works.

We know for a fact that supply is the constraining factor.

We know for a fact that the choice is 2 doses for half as many people, or single doses for now for twice as many people, since all supply is being used.

We know from the data that exists that a single dose works after a fortnight with a Confidence Interval over a 50% from one dose which means that twice many people getting one dose provides more community protection than 2 doses for half as many people.

We don't know that a reduced dosage at each dosage works. We don't know that mix and match vaccines works, which is why its only authorised as a last resort if it is necessary for operation reasons but is not recommended.

As for a single dose alone, a booster dose probably gives more protection than just one dose so there's no reason to permanently skip the second dose. Given the risk profile too it makes sense to be giving second doses for the eldest and clinically vulnerable before first doses for the youngest. Given the vast majority of the deaths are from the first 4 priority groups, while under 1% of deaths are from the under 45s, a second dose for the first 4 groups makes sense before the first dose for people like me. But a first dose for all the vulnerable, makes sense before a second dose for half of the vulnerable.

Ultimately there's no guarantee which method is best, we're operating in an imperfect world not a trial. A reasonable best guess is required - and a reasonable best guess that twice as many vulnerable people with 1 dose > half as many vulnerable people with 2 doses is well reasoned.

I have no issue with this basic theory from a public health perspective. My issue is arguing that this falls under off-label use provisions, which are generally related to individual patients, not broader public health.

The principle is the same, just its off label based on an overriding public health emergency.

Since it takes about a fortnight for vaccine protection to take effect, then a week to two weeks to go from infected to hospitalisation, it takes about a month from vaccination to an avoided hospitalisation - so its a bit early to see the vaccines feeding through to hospitalisation figures in the UK. However there's already some early evidence that vaccinations are having an impact on hospitalisation numbers. The age-related hospitalisation figures have gone up and down together to scale, they're starting to go down now thanks to lockdown but there are early signs the over 85s admissions are declining faster than other admissions are.

[wiggin]

https://assets.publishing.service.go...ter_14a_v6.pdf

The only vaguely relevant information in that link is the phase III results from Pfizer and AZN, which no one is disputing. Those data do not, however, represent comprehensive data on 'lower viral load and lower hospitalizations after 1 dose' by > 50% especially when I have already shared data that disputes that, and the CIs on the Pfizer and AZN data were enormous given the small window for infections. You have yet to share additional data beyond what can be inferred from the original phase III trials.

The EUAs were granted because the placebo controlled RCT showed the vaccines were safe and efficacious. After that it becomes a question of how best to use it.

No, a question on 'how best to use it' might be something along the lines of who gets it first rather than 'let's just completely change the dosing regimen that was clinically tested because our gut tells us it'll be fine'.

This change is based on evidence that it works.

Evidence that does not exist past 3 or 4 weeks and is thin before then.

We know for a fact that the choice is 2 doses for half as many people, or single doses for now for twice as many people, since all supply is being used.

Those are hardly our only choices! Why couldn't we also include 2x half doses for twice as many people? Other doses were used in the phase I/II trials and some showed modest efficacy, albeit underpowered. Why couldn't we push for true herd immunity (albeit shittier than a full regimen) for the entire population by giving everyone one dose and figuring that will provide better overall protection for at risk groups than giving them the second dose before everyone else gets their first? There are perfectly reasonable justifications for any of these strategies, but without testing (or at a bare minimum, really rigorous modeling) we don't know which are awful ideas and which are winners.

We know from the data that exists that a single dose works after a fortnight with a Confidence Interval over a 50% from one dose which means that twice many people getting one dose provides more community protection than 2 doses for half as many people.

We also know from data that exists in a bigger and more realistic dataset that a single dose of the Pfizer vaccine works with 33-60% efficacy after 10-14 days until 3 weeks. Which means, well, it's hard to know without a lot more information, but it's certainly not proof for your conclusion.

The principle is the same, just its off label based on an overriding public health emergency.

Just don't use the word off-label use, it means something different and has different legal and ethical underpinnings than what you're talking about.

[RandBlade]

The only vaguely relevant information in that link is the phase III results from Pfizer and AZN, which no one is disputing. Those data do not, however, represent comprehensive data on 'lower viral load and lower hospitalizations after 1 dose' by > 50% especially when I have already shared data that disputes that, and the CIs on the Pfizer and AZN data were enormous given the small window for infections. You have yet to share additional data beyond what can be inferred from the original phase III trials.

That's not the only evidence but it is substantial evidence and what I was able to find from a quick search. As I've said other evidence is out there too but that is significant evidence.

No, a question on 'how best to use it' might be something along the lines of who gets it first rather than 'let's just completely change the dosing regimen that was clinically tested because our gut tells us it'll be fine'.

Not just gut, evidence and reason.

Evidence that does not exist past 3 or 4 weeks and is thin before then.

Yes it does. Evidence is not proof. If there's evidence that the vaccine provides protection at week 3 and week 4 then given everything we know about how vaccines work that is evidence it will be comparable in weeks five and week six. Immunity does not just magically vanish overnight, it doesn't work that way.

This seems to be the biggest difference between a few of you guys and any of the scientists in the UK I've heard speak on it. You are acting as if the second the trial data ends that is the end of what we know and the end of any protection then that's it - the great unknown and zero there. Whereas the scientists here are using a well-reasoned understanding of what we know about vaccines to extrapolate past the three week window. The PIII trial revealed that immunity was acquired, it won't be gone overnight after that.

We don't know for certain what the level of protection is - but we do have well reasoned evidence. What we do know for certain is the level of protection the unvaccinated get: None whatsoever.

Those are hardly our only choices! Why couldn't we also include 2x half doses for twice as many people? Other doses were used in the phase I/II trials and some showed modest efficacy, albeit underpowered. Why couldn't we push for true herd immunity (albeit shittier than a full regimen) for the entire population by giving everyone one dose and figuring that will provide better overall protection for at risk groups than giving them the second dose before everyone else gets their first? There are perfectly reasonable justifications for any of these strategies, but without testing (or at a bare minimum, really rigorous modeling) we don't know which are awful ideas and which are winners.

For what possible reason would we do 2x half doses for twice as many people, when we don't know if half doses work, rather than 2x full doses spaced just a little bit further apart?

We don't have time to test every strategy. We need to make a decision sometimes. This isn't a scientific endeavour, it is real life. We know what the consequences are for not bothering to vaccinate people.

We also know from data that exists in a bigger and more realistic dataset that a single dose of the Pfizer vaccine works with 33-60% efficacy after 10-14 days until 3 weeks. Which means, well, it's hard to know without a lot more information, but it's certainly not proof for your conclusion.

Its not proof but it is evidence.

Just don't use the word off-label use, it means something different and has different legal and ethical underpinnings than what you're talking about.

I didn't use off-label, I said off-book originally because I misquoted the person who used it - who was himself using an analogy. I made clear I was quoting someone else, but from memory.