[lisper]

> each person needs 2 doses

That's not quite true. You need 2 for full effectiveness, but it's not like missing the second dose leaves you with no benefit at all. In fact, it's far from clear, from a global perspective, that having twice as many people with less-than-full-effectiveness would not better than half as many people with full effectiveness.

[valarauko]

That's true, but the FDA EUA specify that Pfizer & Moderna stick to the dosage regime as tested. It's very likely that a half dosage/reduced dosage might be deemed acceptable, but AFAIK no national authority has allowed a dosage regime deviation from what was tested in the Phase III trials. The closest would be the UK allowing upto a 12 week gap between doses of the Oxford vaccine. I agree that it's probably worth it to half the dosage and double the number of people who can be covered when initial supplies are so limited, but till the FDA makes that call, we have to stick to the dosage as specified.

[lisper]

No national authority has had to deal with a world-wide pandemic for the last 100 years. Treating this situation as if it were business as usual and following the normal rules may not lead to the best outcome.

[valarauko]

I don't disagree with you, but they also have to balance risk. They need numbers to work on, and those numbers barely even exist right now. The 2K in the Moderna trial who received the single dose is a statistical blip compared to the hundreds of millions for whom the FDA would make decisions.

The EUA is their attempt to manage the health risk to the population - an investigational vaccine using an entirely unproven novel technology stack poses some unbound risk (likely small), vs the benefit of an effective vaccine. Once we change the dosage, the effectiveness also become unclear, while the risk remains (eg, potential trace contamination with precursors in the lipid carriers triggering shock in a untested population subset. Perhaps prescription statins greatly reduce the effectiveness of the vaccine, etc etc). It's the job of the FDA to err on the side of caution and see risks where we would not. I agree with you that the situation is unprecedented, but I also do not blame the FDA for only working with the data they have. I suspect that Pfizer & Moderna will conduct Phase IV trials with different dosage regimes based on which the FDA can modify the terms of the EUA.

[mullingitover]

I'm not so sure that a less effective vaccine is better than no vaccine at all. If the virus has a less-effective immune response, it could have a better chance of mutating into something that's resistant to the vaccine, in which case all the work creating the vaccine goes out the window.

[andybak]

I vaguely remember reading something along the lines of "You're thinking of antibiotics. Vaccines and viruses are different".

Anyone got any solid insight into this?

[mullingitover]

I found the article.

> “My concern, as a virologist, is that if you wanted to make a vaccine-resistant strain, what you would do is to build a cohort of partially immunized individuals in the teeth of a highly prevalent viral infection,” Bieniasz told STAT. Even rolling out the vaccine at all when there is so much transmission occurring is far from ideal, he said, suggesting it would have been safer to beat down the amount of virus in circulation before beginning the vaccine deployment.

>“You are essentially maximizing the opportunity for the virus to learn about the human immune system. Learn about antibodies. Learn how to evade them,” he said.

[1] https://www.statnews.com/2021/01/04/britain-takes-a-gamble-w...

[jhowell]

> In fact, it's far from clear

Seems clear from my perspective. Two doses are recommend by manufacturers. Anything more or less is pure speculation at best. Why speculate in absence of a way to test your hypothesis?

[cltby]

> Anything more or less is pure speculation at best

There are two lines of evidence. The first is empirical. It It appears that one dose of Moderna _is_ highly effective, perhaps 85% or more. Note the treatment-control divergence after about fourteen days [1]. The second is theoretical. From our ample medical experience with other vaccines, there is strong prior reason to think that one dose is likely to work well, and that a second shot in the distant future would be even better than a second shot after three or four weeks. Booster shots are given a.) as backup for people who don't seroconvert after one dose b.) to trigger a secondary immune response. In light of a.), we shouldn't be surprised by the 85% number (most people seroconvert; there's no partial immunity, you either seroconvert or you don't). In light of b.), we should be very skeptical of the four week interval, especially since the secondary response takes about two weeks to develop. It's shorter than the interval for every other vaccine.

[1] https://arstechnica.com/science/2020/12/fda-releases-data-on... [2] https://www.cdc.gov/vaccines/schedules/hcp/imz/child-adolesc...

[Exmoor]

Agreed. It's also important to remember that the 3/4 week dates for the second dose were not chosen based on it being the most effective timeline, but because it was the minimum time needed to complete the studies as quickly as possible. If they'd chosen 8 weeks we wouldn't have US approval yet. If they'd chosen 12 weeks we wouldn't even know if they were effective yet.

[BurningFrog]

The manufacturer doesn't know anything about this that isn't public information.

Yes, two doses are almost certainly better than one. But those are not the choices.

When there are far fewer doses than the number of people to vaccinate, the question is if it's better to give N people 2 shots, or 2N people 1 shot.

The test data looks like 1 dose gives ~80% protection vs ~95% for 2 doses. The 2x80% scenario is clearly better than 1x95%.

And, of course, everyone would get the second dose as soon as production catches up.

[pavelrub]

The main unknowns aren't around short-term efficacy, but long term immunity. If it turns out that a single dose provides ~80% protection but only for 3 months, whereas the second dose provides strong long-term protection, the one-dose scenario would be significantly worse.

[BurningFrog]

Longer term, everyone gets a booster shot as soon as they're available.

[Bootvis]

What if it takes 9 months to have the booster shots available? People will be vulnerable for 6 then.

[Exmoor]

Vulnerability isn't binary. You're not immune one day and susceptible the next. It's much more likely that immunity to getting sick might start decreasing over time, but still remain relatively high. So, perhaps 80% in 3 months, 60% in 6 months, and 30% in 9 months to use made up numbers. Additionally, production is still ramping up so it's highly unlikely that we'll have less vaccine in 3 months than currently. We'll also likely have 1-3 more approved vaccines by then as well.

[lisper]

Because the manufacturer recommendations are based on achieving the best outcome for the individual receiving the treatment. But we're in a global pandemic, a situation the world has not faced in living memory. A different quality metric might be appropriate under these circumstances.

[mns]

So even though the manufacturers (the companies that developed the vaccines, the scientists, the people running the trials, the ones that know the data and went through a full year of tests) keep on recommending 2 doses and the advice given to the UK government was still to stick to the 2 doses, we're now going the way that "we think we know better and it's better to try something new and untested because it's a pandemic and it might be better to have less efficiency to more people, than full efficiency to less people", even though it's, again, not tested and a shot in the dark?

[jjeaff]

Was there any testing done comparing 2 doses to one? I think it was speculation that led us to test with 2 doses because we'll, we don't have much time, 2 must be better than one.

[lisper]

https://www.wbur.org/commonhealth/2020/12/18/coronavirus-vac...

[freeone3000]

The study we have of a two-dose COVID-19 vaccine with one dose administered showed 67% effectiveness, in line with vaccines for other diseases (including eradicated ones - polio vaccine is 80%).

[Bootvis]

Giving one dose increases the probability of someone catching COVID and the virus evolving resistance within this person. I don’t know how likely that is but the impact could be huge.

[valarauko]

Is there evidence of this?

[Bootvis]

I think it's too early to have evidence of this I think but knowledgeable people seem to have some concerns:

https://www.statnews.com/2021/01/04/britain-takes-a-gamble-w...

That the impact could be huge doesn't take much imagination: We could be back to square 1 with the vaccination process and the evolved virus might be more dangerous.