Hacker News Reader

Comment by bransonf | Hacker News Reader

It’s hard for me to understand the value proposition of self experimentation.

Probabilistically, It sounds like best case you’re looking at a small level of immunity. Worst case, you’re looking at enhancement.

Doing this to yourself proves little more than you are willing to take a moderate risk. The behavioral changes noted at the end about a change in perception seem particularly dangerous. Assuming you might have immunity makes you behave somewhat riskier. And, if the experiment causes enhancement, combined with your lowered inhibition sounds like a death sentence.

I’m all for moving fast in many things, but I just don’t see how direct human experimentation provides any scientific benefit at this point. Besides, medical sciences that move fast Have the highest likelihood of additional harm/death.

View on HN

"Best case" is that you then follow with a challenge trial and demonstrate actual immunity much faster than a conventional vaccine, leading to months shaved off the time needed to immunize the general population. That clearly has value, particularly given that people are still dying at a high rate in the US and much of the developing world.

Now... does that mean this is going to work? No idea. But these folks aren't dummies either. I think it has a shot and is worth trying as long as everyone involved (e.g. these are all university scientists) has a clear and solid understanding of the risks.

Aren't there phase 3 trials already? Meaning that we are in fact already giving people experimental vaccines. Maybe if you had done it months ago it would have sped things up.

Phase 3 requires monitoring how many people get sick from the environment. That takes a LOT of test subjects and a decent amount of time -- even in major hot spots covid isn't THAT prevalent. The overwhelming majority of people at any moment, after all, aren't getting sick.

If you have well-informed volunteers like this willing to break with conventional rules, you can do a challenge trial where you deliberately expose them to live virus. That can work very fast indeed.

Or just give major league baseball players the experimental vaccines. We'd be done with phase three in a few weeks if every MLB player who agreed to play volunteered for one of the vaccines currently entering/in phase three.

I don't get it. Isn't "every MLB player who agreed to play" too small a sample size to be useful?

Short answer: No.

This question comes up a lot. Someone should really write a blog post about how big sample size you need, and the dangers of using a biased sample (fit men), and spam it to death here.

I think the "fit men" can be an issue when you are testing whether a drug works to reduce severity (treat) an infection. However, the question of whether the vaccine actually prevents infection should be relevant regardless of the level of fitness. You should probably only vaccinate half of them and when inevitably an unvaccinated person gets it, you should have very good data about protection two weeks later. This is all just a thought experiment of course (unless there are volunteers . . .)

Ah, challenge trials. That makes sense.

> I just don’t see how direct human experimentation provides any scientific benefit at this point.

While historically self-experimentation has led to great scientific leaps, here (from what I understand) the goal was not so much to drive the science forward, but to protect the self-experimenter from infection and harm.

Reply on news.ycombinator.com