5 ms·
relatedly, the only good reason to stop a trial early is that it becomes unethical not to treat the control group because the effect size in the treatment group
by hprotagonist 4y ago
relatedly, the only good reason to stop a trial early is that it becomes unethical not to treat the control group because the effect size in the treatment group is so huge. And it does happen, sometimes.
- civilized 4y agoI sometimes wonder if the exhilaration of such a result comes with a twinge of regret that the result could not have been foreseen before the science reached it, and more people given the lifesaving treatment immediately. But that is the human condition, I guess. Scientific progress and learning brings regrets, often very momentous ones in retrospect.
- carapace 4y agoRegret, sure, but since we can't change the past that regret should motivate us to work harder to make the present and future better. We're a young species, and part of growing up is looking back with chagrin at how foolish we seem in the light of our new growth and learning.
- chrisbrandow 4y agoWhile I totally get what you mean, I’d guess for scientists, the answer is generally, “no”. Expected outcomes for trials like this are a whole lot less certain to the people doing the work, that than it seems on the outside, so i think it wouldn’t even occur to the scientists that the downside of “withholding” treatment from the tiny (relative to the population) control group comes close to the upside.
- civilized 4y agoYeah, it's just a wistful thing, and most of us laymen don't even experience the part where you want to know something but don't know it yet.
- aperson_hello 4y agoThe number of trials that don't work in humans when it worked in every pre-clinical trial up to that point is enough that it makes sense to be extra cautious.
- qgin 4y agoIt’s true, but there’s no alternative. 99% of things don’t work.
- hprotagonist 4y agogenerally speaking, “holy shit it worked?!!!” drowns out a lot. It’s not so much exhilaration as it is a kind of astonished joy.
- orblivion 4y agoI mean, treat everybody in the world with that condition at that point, right?
- thematrixturtle 4y agoThe treatment costs ~$100k over half a year.
- gumby 4y agoThe other reason is futility: the treatment has no benefit or any improvement is grossly outweighed by side effects.