Hacker News Reader

Comment by finolex1 | Hacker News Reader

Anyone knows why they insist on both doses given the limited supply of the vaccine and the marginal increase in effectiveness with the second dose? I would imagine every single day counts.

https://www.lesswrong.com/posts/Rvzdi8RS9Bda5aLt2/covid-12-1...

View on HN

From the linked article[0]:

“Yes, Moderna’s vaccine prevents transmission. One dose is good for reducing infection by 63%, two by over 90%.”

The number of people you need to vaccinate with a 63% effective vaccine versus a 90% effective vaccine is a huge gap.

There are two issues with vaccines that staring at these numbers won’t tell you:

1. Not everyone is going to get vaccinated.

2. The number of people who will get vaccinated is directly correlated with public trust.

If you release a vaccine that is 63% effective, the “this vaccine doesn’t work crowd” and their hugely amplified voices on YouTube and social media will be exponentially worse.

That snowballs into fewer people being vaccinated. Unlikely for us, with a less effective vaccine we need far more people being vaccinated.

So you want a more effective vaccine because fewer people need to get vaccinated, more people will trust the vaccine, and more people will get vaccinated.

(To pre-empt the bad-faith replies, this does _not_ mean we would sacrifice another six months to get a 98% effective vaccine (if it were possible) versus starting now. It’s a balance, of course.)

Also very important is that a second dose isn’t just for effectiveness, it’s to boost the immune system’s response so that the conferred immunity lasts significantly longer.

[0]: I struggled through this article. I find this self-congratulating “I’m smarter than everyone and I told you so in this other blog post” writing insufferable.

The lesswrong quote seems to be misleading. It links to an article that says the vaccine may prevent / reduce transmission but that it hasn't been proven yet. Did I miss something?

The British Medical Journal has an article that makes the point, including quotes from Tal Zaks of Moderna, that the studies aren't designed to be statistically significant indicators of disrupting transmission, preventing hospitalisation, or even preventing death:

https://www.bmj.com/content/371/bmj.m4037

So, from what I can tell. It isn't proven to disrupt transmission or reduce hospitalisation/death. It would require much larger or longer studies, which they have opted not to do.

I think lesswrong is wrong here. I looked at the document that lesswrong links, and I couldn't find any statement about effectiveness of the vaccine in preventing infection after two doses.

Also the “I don’t trust a vaccine approved by the Trump administration” crowd. It was hugely irresponsible to politicise the vaccine.

I'm not American so I could be way off on this, but I can't help but wonder if the politicization might've actually helped in this case? My big assumption is that Trump supporters would normally be more skeptical of vaccines. If non-supporters would trust the vaccine regardless, then perhaps politicizing it just means that more Trump supporters would also take it?

This will be overcome. It was about trusting the scientists instead of the White House. I'm fine with it now. But didn't they pressure that CDC director to approve it that day or get fired? They were probably going to approve already. But what if they weren't far enough along and some of this nonsense would've played out like that? It'd have been very harmful for the trust in the vaccine.

Also the "don't trust big pharma crowd", but apparently big pharma is the best thing ever this year.

I trust the scientists and doctors who've said it's safe and explained the reasons it was approved faster than normal that have nothing to do with Trump, aside from his botched preparing for and handling of the pandemic necessitating it.

To be fair, the type of people that vacation in Martha’s Vineyard might also be anti vax / anti gmo.

Viewpoints like theirs have been completely eradicated off YouTube.

This argues against only giving one dose instead of two. But no one is proposing that.

The argument is that while there is a vaccine shortage, it is better to give as many as possible one dose, instead of giving only half that many people two doses, leaving the rest unprotected.

When available, everyone gets the second shot, which should work just as well 6 months later.

> which should work just as well 6 months later

Do you have clinical evidence for that? I doubt it. OP still has a more valid point. Public trust in the vaccine and its apparent effectiveness ASAP after the vaccination campaign begins will be gigantic factors in its success.

We all know that a hard lockdown brings down the new-infection numbers WEEKS after beginning the lockdown, yet there are still a lot of people on Twitter, YT and at demonstrations (at least here in Germany) where folks are angry that the lockdown doesn't work 2 DAYS after its beginning.

The amount of truly wary/ignorant people out there is around 5-15% of the entire population in Germany and handing out a 63%-effective vaccine will be fuel to their bullshit stories.

I wish people would start taking this more seriously. We have a huge chunk of the population (5-15%) where discussions don't work anymore because their arguments aren't based in a common reality (e.g. "viruses exist and may pose a threat" is surprisingly often not a supported opinion).

> Do you have clinical evidence for that?

I'm no epidemiologist, but I understand this is the how booster vaccine shots typically work.

I care less about public trust and the PR angle, and more about getting as many people protected as possible. If dumb people say dumb things matter much less.

> We all know that a hard lockdown brings down the new-infection numbers WEEKS after beginning the lockdown, yet there are still a lot of people on Twitter, YT and at demonstrations (at least here in Germany) where folks are angry that the lockdown doesn't work 2 DAYS after its beginning.

In the spirit of understanding, in your words what are the biggest arguments against forcing a hard lock down?

As someone that has been preparing for an influenza pandemic since 2005 I can tell you that the yearly tri or tetra valent influenza vaccines each year are often, in the last decade mostly, down near 50% protective. And you don't get 2 influenza shots.

This is not something new or special. And people being ignorant should not change how things are properly done. Yes, they're going to freak, but it's because they're idiots. Don't base behavior on what they think. Accomidating them will not change their behavior at all. Their behavior does not reflect external realities.

It's not an unreasonable suggestion (and a lot of people are making it now), but the risk with a single dose is low persistence. We know that the first dose has high effectiveness in the first month, but it may trail off sharply. Or it might not. Since we haven't tried a single dose, we don't know.

In animal studies, the second dose was required for a durable response, that's why Pfizer decided to go with it in their trial.

They played it save when designing the protocol. Better to err on that side, than to risk failure because you chose a dose that's too low.

Given that's what they trialled, there really isn't a mechanism that would allow either them or the FDA to change the protocol from what was tested.

Yes, this is annoying. You can make that educated guess and not show up for your second shot and you'll probably help someone.

But medicine has a long history of people being rather convinced of theories that made an awful lot of sense and killed an awful lot of people.

At some point, they noticed. The double-blind trial became not just the preferred method or something like that. It became absolute gospel. Anything else is considered the GOTO of injecting people with... stuff: Even if it's exactly what you believe is needed right now, it's just not going to happen.

See also: "masks don't work" ( = "even though it sounds like a good idea, there is just as much actual evidence for their usefulness in preventing viral diseases right now, January 2020, as there is for crystal healing. Give us a month and we'll have data")

"But medicine has a long history of people being rather convinced of theories that made an awful lot of sense and killed an awful lot of people."

That works for both sides, both the true believers and the skeptics. Both sides are absolutely convinced that they are right and the other side's belief will kill people. At least one side is wrong. Both have their reasons, both appeal to history, both appeal to science. It is a confusing time to live.

There is a lot of real evidence that masks do work

https://www.thelancet.com/journals/lancet/article/PIIS0140-6...

My impression is that the consensus is reasonably strong that masks work in and of themselves, that is - they do act as a physical barrier against transmission, but, most infections are happening at e.g. family gatherings and in households, so the actual effects of masks writ large are more muted than proponents argue. Basically, it's like saying seat-belts work, which is obviously very true, but regarding them in a kind of isolation that forgets about setting speed limits.

Could it be that most transmissions happen with family and not shopping simply because the majority wears masks?

In countries with actually working contact tracing you have regular cases from restaurants and doctor offices. For example

Quoting myself: right now, January 2020,

(meaning: I was trying to reenact a scene from the beginning of (gestures around) all this)

That paper: June 2020

If every single day counted, they would have done challenge trials. What counts is following the FDA's process, and the FDA's process says that you have to administer it in the same way as the trial. No one at the FDA is going to go out on a limb just to save a few thousand lives.

Challenge trials don’t work when 40% of infected are asymptomatic. And challenge trials especially don’t work when 15% of cases require hospitalization and ICU beds are already maxed out.

I don't see how the asymptomatic fraction makes a normal trial superior to a challenge trial.

As for the hospitalization and ICU, those numbers are inflated, but regardless, in order to pass the trial, the same number of people need to be infected in the trial. The only difference is that while you're waiting for a few dozen people to be infected naturally because you gave them a placebo, millions are being infected and thousands are dying daily.

I work in healthcare.

Where’s your source on them being inflated?

Citation needed for that 15% number. Even if it’s true, it absolutely isn’t true for all age groups. Give to 20-29 year volunteers. There were hundreds of thousands who volunteered.

What our (Lithuania) officials say is that about 5% of the cases are hospitalized. We are close to the top by the number of new infections per 100000 pop.

What doesn't work about it?

Being ethical and getting results are not the same thing.

Being ethical and being “ethical” according to professional expert ethicists is also not the same thing. If I learned anything this year, it’s that opinion of professional experts at ethics should not be considered when deciding who gets last piece of cake, much less in life and death scenarios.

> Later, a vaccine candidate that has been proven safe in clinical trials could be given to a group of healthy adults in this age range, who are then exposed to the virus in a controlled environment. They would then be closely monitored by medics and researchers to see if the vaccine is successful in preventing infection, as well as identifying any side effects.

> https://www.thelancet.com/journals/lanres/article/PIIS2213-2...

I imagine they will do what they did for the vaccine candidate ChAdOx1 nCoV-2019 and swab at least every week to test for infection.

Do you have any source to back up your claim that a challenge trial wouldn’t work for SARS-CoV-2?

When the asymptomatic rate is high, it’s hard to determine if the vaccine is truly works, and to what degree. If 40% of people are asymptomatic then that would theoretically mean that 40% of challenge trial participants won’t show symptoms. Does that mean the vaccine is working? Who knows. It could be working, it may also not work.

The only way to do it and get real results is to use a challenge trial only on the elderly population where the asymptomatic rate is low and where the symptoms are exaggerated. This would mean though that if the vaccine didn’t work you would overwhelm the hospital system with a lot of sick elderly people who require the most care, when the healthcare system is already overwhelmed.

Are challenge trials for dearly diseases even legal?

I'm not an expert, but my understanding is that the error bars on efficacy for the single dose are quite large, just because of how the studies are conducted. Moderna's lower bound was below 70% with a single dose, and Pfizer's is below 30%. That's not to say that a single does is definitely not effective, just that there's a lot more uncertainty, and so the bounds are much looser. We also don't know whether giving a second dose a few months later rather than a few weeks later would work as well.

Depending on supply forecasts, it might be worth running a new trial to get a better handle on the efficacy of a single dose, but going that route now would definitely be a gamble. Especially since we're formulating this single dose hypothesis after having run the experiment and seen the results, which is always a dangerous approach.

Still, the data do seem to suggest that it's very plausible that a single dose would be sufficient. If things get bad enough, or supply is low enough, maybe that's a gamble worth taking.

One thing to consider is that vaccine production is going to scale up in time; since the second dose is a month after the first, giving two doses instead of one will not mean that half the number of people will be vaccinated. Assuming, for the sake of the argument, that vaccine production doubles every month, at any given point after the first month the amount of people that have received at least one does will be a bout 2/3rds of the single dose protocol.

Noone has tested single-dose.

In this case it may sound like too much nitpicking, but the general principle is that you don't roll out something you haven't tested in a trial. In general this is a good principle. I hope they'll do trials comparing single to double-dose soon.

Reply on news.ycombinator.com