Talk:2726: Methodology Trial

Woah! This is the first time I've seen a new comic without an explanation. It's pretty weird. SilverTheTerribleMathematician (talk) 02:36, 19 January 2023 (UTC)

You just need to have checked at the right (or wrong?) time, which is different every publication day so you can't generally predict it. But at least three readers find empty explanations every week, and now it's your turn!. OR, maybe this is just the first time you've unknowingly checked the placebo wiki... ;) 172.70.91.75 02:54, 19 January 2023 (UTC)

04:10, 19 January 2023 (UTC)~ Comment on the title text. IRB is an Institutional Review Board, which I guess is a committee that decides if research is ethical and ok to do. The title text refers g to a placebo IRB, which I suppose is a fake IRB that can approve research as part of an experiment to determine the real effects of IRBs. Or something.

Why would this experiment be more unethical than any regular placebo trial? In either case you're telling patients they're getting actual medication when in reality they're getting sugar pills (or whatever you use as placebo). Bischoff (talk) 08:22, 19 January 2023 (UTC)

In a normal trial, the patients are told the researcher is giving out some candidate medicine and some placebo - typically they'd have a 50% chance of receiving a medicine that may help them. In this case, none of the treatments given will actually contain the active ingredient. I suppose it depends what you consider the "trial" (whether it includes all the test sites or just the one Cueball is running).172.70.85.46 10:40, 19 January 2023 (UTC)

Placebo-based trials may have a threshold at which particularly significant results (as detected upon the half-unblinded results by the number crunchers that monitor progress) bring the primary study to a quick close and perhaps prepare to roll out the more than sufficiently proven treatment to all participants, for equal benefit from this point on. And now monitor for Adverse Events on both "early" and "late" treatment groups. (It might not benefit those who are "late", compared to the initial enrollment criteria, it might actually show greater improvement to them or it could even exibit a surge of unexpected AEs from those who have been on longer-term dosage regimes; that all still needs teasing out from the stats.)

Probably, in such a case, the patients (and blinded doctors) are only told that the study phase is over, not whether they are now necessarily being switched treatments, for continued double-blindedness. Once the study is truly finished, it would depend upon what the participant actually signed up for as to whether they (or their next of kin) ever learn the historic details of their participation.

Conversely, significant AEs (by frequency or severity) that are identified as cropping up worse in the test treatment (vs placebo or in studies vs whatever prior treatment they are comparing against) should result in early ending and all those being changed to already accepted treatments.

(But, for the latter, it could also just initially be a pause. To review the circumstances and determine relevent details behind the bare data. Such as the unexpectedly recorded death in a small nasal-spray trial being 'just' from being a passenger in a car accident, unlikely to have any connection with which cohort the participant was in – absent of a reported spate of other unusual AEs that indicated hightened tendencies to dangerous physical impulses, anyway.)

The study design should have levels of significance that the core stats team (unblinded as far as "Patients X, Y, Z took the active treatment") will raise escalating concerns about continuing in the face of the more convincing or unconvincing results, even if the AE discrepancy is not as clear (or, in some cases, expected) as actual death. Like raised/lowered levels of reported bedsores, etc. With viagra being the obvious "secondary side-effect" examplar, detected as useful beyond the intended scope of the initial studies. 172.71.242.139 12:07, 19 January 2023 (UTC)

I'd sort of assume a placebo IRB would approve or deny projects randomly, where as a real one would "work" and actually analyse the projects being proposed. You could use this to see if the IRB is more ethical than a placebo, which you'd seriously hope. There'd obviously be a whole conversation on what constitutes more ethical, but you could prove that experimentally with a trial involving real and placebo philosophy and ethics departments 172.70.250.245 09:29, 19 January 2023 (UTC)