[bsder]

Except that the drug has negative side effects and does not work-not even with the most optimistic interpretations.

If you allow that, you're back to HCQ for Covid territory. Or all the autism "treatments".

The problem is that most people think of medical outcomes as having 2 results when there are 4.

The 2 everybody thinks about are "drug given->patient gets better" and "drug not given->patient gets worse". But there are two more cases.

"drug not given->patient gets better" is wonderful in the super-rare cases when it happens and is totally unproblematic.

The problem is that the case "drug given->patient gets worse" is common--doubly so for experimental treatments--exponentially so for experimental cancer treatments. And the problem is that it can be due to either the disease or the drug or both. This case is the one that you can't explain in a soundbite. And it's one of the big hurdles in drug discovery.

The vast majority of people simply will never get this. We know this. We watched it play out in real time--multiple times.

Sure, if you made everybody pass a Bayesian statistics exam before they could get their treatment, that would work. But then you'd be "gatekeeping a cure".

Stupidity is the norm, not the exception, and greedy, evil bastards are not unrare. Your rules have to, sadly, take that into account.

[krisoft]

> "drug not given->patient gets better" is wonderful in the super-rare cases when it happens and is totally unproblematic.

It is not actually super rare. One might even say this is the most common case. I guess what you can say is that spontaneous remission is super rare to happen for “serious” diseases. But that is circular reasoning because we kinda define serious diseases by the fact that it is rare to just get better without treatment from them.

Why is this important? Because this is one of the main wrinkles with drug testing. If some people just get better on their own then you have to be very carefull with statistics and double blind protocols. Because if not you might confirm your belief in an innefective, or maybe even harmful treatment by random happenstance. Therefore i would say this case is also one of the big hurdles in drug discovery.

Otherwise i agree with the general gist of your comment.

[ifyoubuildit]

I think this belief is how you end up with people who think you need some cocktail of OTC drugs every time you get a cold.

A child gets a run of the mill cold, parents give them all the OTC things, kid gets better and assumes it must have been because of the drugs. In reality, they didn't need any of it.

[solatic]

That doesn't sound like an argument against the right to try, that sounds like an argument for requiring specific, intentionally frightening language around informed consent. The thing is, probability is a bitch, and emotionally filtered probability is the devil. An experimental treatment for a terminal illness can be 99% likely to result in a quicker or more painful death, but is it really right to deny the 1% chance to someone who understands the statistical choice in front of them?

[amanaplanacanal]

Another facet of this is: should insurance pay for it? Or should the patient bear the financial burden when they want to try a shake-oil cure? Especially in this case when the drug is now FDA approved?

[mcherm]

> Except that the drug has negative side effects and does not work-not even with the most optimistic interpretations.

One of the difficult things about ethics (including medical ethics) is that everyone has slightly different opinions.

But in MY opinion, if someone's medical situation has an extremely high likelihood of extremely poor outcomes (terminal cancer and severe Alzheimer's both qualify), then I think it is perfectly acceptable to allow them to use a treatment which is well understood to have no useful effect and many harmful side effects.

I have an ethical problem with encouraging them to do this; I have an ethical problem with profiting from them doing this; but I think that prohibiting them from doing it (as we do today, mostly) is also wrong. EVEN IF the patient is "too stupid to understand it correctly".

[bachmeier]

> I think that prohibiting them from doing it (as we do today, mostly) is also wrong. EVEN IF the patient is "too stupid to understand it correctly".

There's a lot that can be said on this, but it's harder to defend that position if the decision has an effect on others. The obvious situation that arose during the pandemic is someone experimenting and then creating more work for an already overwhelmed hospital system. Then you have to ask about paying to clean up the mess - should it be health insurance, that is paid by everyone else? If you die, what are the effects on your family? How do you draw the line where "well they don't have much time left anyway" vs "they killed themselves experimenting with flu treatment"?

[gus_massa]

For a real case of a somewhat popular and harmful treatment https://en.wikipedia.org/wiki/High-dose_chemotherapy_and_bon...

> Clinical use of this treatment was driven by women with breast cancer and advocacy groups, such as ACT UP; they believed that the FDA treatment approval process was too slow.