Hacker News Reader

Comment by arcticbull | Hacker News Reader

> The long-term side effects piece is still undecided and that's undebatable, simply because not enough time has gone by. Maybe I'll be open to changing my mind in 5-10 years. Even then, I'd win, no?

Like I said long-term side effects doesn't mean latent effects. It doesn't mean things that show up years after administration. It means things that happen right after administration but last a long time. If getting the COVID vaccine caused your arm to fall off 30 seconds after administration, that's a long-term side-effect because last I checked arms don't re-grow.

Yes, we do know what the long-term side effects are. You've confused 'long-term' and 'latent'.

Yes, it's decided. No, it's not debatable.

> ... simply because not enough time has gone by.

Again that (a) doesn't matter and (b) unless you're a vaccineologist then your opinion about whether sufficient time has passed is totally meaningless.

View on HN

Ok, so latent side effects are long term side effects that have no immediately observable impact. As in, we aren't aware of them until later. You're arguing over semantics, but latent effects absolutely do matter as well. And are just as important to consider in the risk profile. If I eat mad cow prions, I might not have any immediate issues, but those prions are folding around in my head. Is that a latent or a long term side effect? The only difference is whether it's observed or not. So does it matter? It's still an issue. With the mRNA vaccines, one of the only side effects that's acknowledged now in public is the cardiovascular damage they can cause. You can damage your cardiovascular system and not notice it for a long time. Likewise, the iGG4 antibody triggering for the spike proteins, this actually has an immediate side effect that can be mistaken as a beneficial side effect, less symptoms! But, now your infections are treated like common allergens by your immune system and a "mild" case of COVID causes you to die suddenly. Latent, long term... what's the difference?

> unless you're a vaccineologist then your opinion about whether sufficient time has passed is totally meaningle Also, outright dismissal of an argument based on appeal to authority is not productive.

> Latent, long term... what's the difference?

Most people conflate latent with long-term. They believe when researchers are looking for 'long-term' effects they're looking for things that magically show up somewhere down the line, and therefore 'we haven't waited long enough!!' and 'there's not enough data!!' - but that's not what they're doing, although they're open to being surprised.

We've studied mRNA for like 50 years now, we understand how it works enough to know that what you're suggesting just isn't a real outcome.

You are right it doesn't matter per se because outcomes are normally distributed in severity and time and if you give enough people the vaccine you can substitute that for waiting. And by vaccinating 70% of the world, we can be utterly, completely, 100% sure it's safe.

It's safe, end of story.

> Also, outright dismissal of an argument based on appeal to authority is not productive.

To invoke appeal to authority there would have to be an attempt to invalidate evidence on the basis that someone is or isn't an expert. The parent didn't provide any evidence whatsoever. If they had either evidence or experience in the area I'd take their claim more seriously (but still inconclusively). However at this point it's just some random person on the internet saying "we need more time!" without explaining why or why I should believe them in absence of data.

Sorry bud, it's MSG all over again. But like I said it's a personal responsibility thing now and if you want to be wrong you're free to do so.

> outcomes are normally distributed in severity and time and if you give enough people the vaccine you can substitute that for waiting

What an utterly fortunate universal truth of vaccines!

It's kind of true when you're looking for the long tail in anything.

The reality is most drugs and vaccines that you get are tested on a few hundred or a few thousand people over a long period of time. But mostly because it's very difficult to measure the efficacy of something with statistical significance if the prevalence rate is low.

So you can either give a bunch of people Ebola to check if the Ebola vaccine works, or you can wait a long time. Ethics boards frown on the former.

However with COVID like 1% of the population had it when we were ready to test vaccines so showing efficacy was very quick and very easy.

But aren't we talking about latent issues? It sounds like you are saying most drugs de-facto check this because the trials must go for so long. But that's the point isn't it? Because short term efficacy was proven so quickly the de-facto check on latent issues never occurred.

Let me ask you this.

What reason do you or anyone else on earth have to think there would be a mechanism that yields latent issues? We have studied mRNA for about 50 years now. And the adenoviral vector vaccines are a completely different mechanism studied for about 50 years too.

What exactly bad thing do you think would miraculously appear 3 years later showing zero signs, symptoms or evidence in 6 billion people? And why would that be different after 5, 10 or even 15 years? There's no law that says Tylenol doesn't cause your arm to fall off 110 years later either, but that's not the standard of proof. Nothing in life has zero risk.

That aside I think if your n is large enough, I think even issues that are ordinarily latent would be at least detectable. Certainly 3 years later. The normal development cycle is as short as 5 years, 2 of which are in regulatory approval.

At some point we have to be guided by our past experience. You can't ever prove a negative, but you don't have to prove a negative to move forward.

I'm not really worried about mRNA per-se I'm worried about an unvetted delivery mechanism. Did you know its rare but accepted that some dogs get injection site tumors and we don't really know why? That would surely be a latent effect in dogs, why do we think it could never happen in humans who live much longer?

> I'm not really worried about mRNA per-se I'm worried about an unvetted delivery mechanism.

mRNA and the adenoviral vector vaccine delivery mechanisms have been studied for 50+ years and are vetted.

> Did you know its rare but accepted that some dogs get injection site tumors and we don't really know why?

Did you know dogs aren't humans, and there's no evidence of this kind of thing in humans at all?

btw it's also extremely extremely rare in dogs, to the extent it's hard to find any kind of data on it. It wasn't thought to occur except in cats until the 70s, so I've written up the rest of this generously assuming cats.

Most study headlines you read are fun achievements that only work on the animals under study and they don't have the same effects in humans. You can significantly increase the life of worms by not feeding them, but humans not so much. Weird huh.

Again this brings us back to our 'substitution of T with N' conversation. It's always a normal distribution. While these VAS tumors develop a few years in, in some cases they develop within a few months. So if you give 6 billion people vaccines over 3 years, you would have definitely seen this kind of latent thing by now, even if its peak effect is a few years later.

> That would surely be a latent effect in dogs, why do we think it could never happen in humans who live much longer?

That's not a thing. Again you're asking to prove a negative, which is not how we do anything because it's impossible. The burden of proof is on you to explain why you think it could happen, not on me to prove why it couldn't.

Prove to me that eating a tomato won't cause turbo-cancer of the elbow. You can't.

I love the way this progressed from, latent disease doesn't happen to prove to me tomatoes don't cause turbo-cancer. There's so much text there that its really nice and smooth.

It really doesn't. I keep illustrating the premise that you can replace time with sample count to make a best-effort, reasonable estimate of the long-tail risk of normally distributed outcomes. The ones to which you are alluding. Ideally, in a way you appreciate - or at the least understand.

I'm also saying that beyond that reasonable estimate based on the 300+ years of vaccine development behind us (and 50+ years of development of mRNA and viral vector vaccines) you are asking scientists to prove a negative which nobody can do.

That's not the standard we hold literally anything else to, nor should we.

I'm sorry for the snark and I do really appreciate the back and forth. I guess what I don't grasp is how we are confident that these are all normally distributed. We don't really know what these distributions look like until we resolve them in the real world. For things that could irreversibly fuck up a population it would seem to me you'd want to take that process tediously slow. I guess what you are saying is "YES goddamnit we are already doin it!" But still we are testing the whole package live without much of a holdback. The whole process felt imprudent for the likes of Covid.

Reply on news.ycombinator.com