Science: Health: Placebo
Placebo with plausible side effects   (+13)  [vote for, against]

The point of giving placebos in an experiment is to prevent the placebo effect. But what if a participant in an experiment can detect whether or not they are getting a placebo? People are not dumb. After all, a sugar pill is likely to have very different side effects (hardly any) compared to the real thing. Test subjects will be actively looking for signs to uncover whether they are getting placebos or the real thing. Almost everyone has an instinct for wanting to know the truth, and those who don't think about it consciously may do so subconsciously looking for the slightest evidence of whether or not they are taking placebo or the real drug. Because of the heightened attention the effects of a pill, the site effects could be extremely mild and still trigger the placebo effect because the patient assumes that side effects = real medication.

Imagine you have a serious disease. You take a pill nothing happens, you immediately think ... "Hmm, I must be in the control group. This sucks, I was hoping to get the experimental medicine to cure my disease". Compared to taking a pill and then feeling "different" couple hours later. You'll think "Hmm.. I think I feel the medicine kicking in, I really hope this helps to cure my disease".

At best, uncovering whether or not you're taking a placebo will introduce a mild placebo effect. At worst, you'll stop taking your medication because you believe it's a waste of time because you are in the control group. Either way, it really messes up the experiment.

To control for this, placebos should have plausible, random side effects consistent to those expected in the real medication.
-- ixnaum, Nov 16 2015

Why have placebos at all ? Shouldn't there be enough data sets on the delivery systems, placebo effects and on doing nothing for the medical problem in question, to analyse the trial against.
-- wjt, Nov 16 2015

[wjt] No, because you never know whether the placebo effect works better against some medical conditions than others, so you always need to control for it. For example a patient with a skin rash may have a very high level of confidence that their doctor will be able to give them some cream to clear it up, and so the placebo effect may be stronger than for another patient with advanced cancer where their confidence that a medical solution is available is low.
-- hippo, Nov 16 2015

Makes sense.

The only problem I foresee is that it's probably difficult, regulations-wise, to administer a substance which you know will have only adverse effects.
-- MaxwellBuchanan, Nov 16 2015

As an aside, what placebo is used in studies involving patients with diabetes? - sugar-free sugar pills?
-- hippo, Nov 16 2015


Cigarette, [Max]?
-- pertinax, Nov 16 2015

This is very clever.

But are you saying for instance, include a mild diuretic to make people pee more? Something that actually DOES induce some mild symptoms that can be assumed to be indicative of a real drug?
-- doctorremulac3, Nov 16 2015

// administer a substance which you know will have only adverse effects. //

Alcohol ?

// what placebo is used in studies involving patients with diabetes? //

Hydroxypropyl methylcellulose.
-- 8th of 7, Nov 16 2015

//Cigarette, [Max]?//

Thanks, but I've got my own. And, sadly, doctors are no longer allowed to prescribe them.

//Alcohol ?// I respectfully refer the questioner to my previous answer.
-- MaxwellBuchanan, Nov 16 2015

//are you saying for instance, include a mild diuretic to make people pee more?

That could be one of them. But ideally you would want to have something random that changes from study to study. That's because otherwise people learn what's a placebo and what isn't.

"Hmm.... I'm going to pee more often, this must be the placebo that I heard about on half bakery"

... the target has to be moving.
-- ixnaum, Nov 16 2015

Really clever.
-- doctorremulac3, Nov 16 2015

"Why is my pee blue doc?"

hmm, I wonder if there has ever been psyche profile comparisons of the differences between those for whom a placebo actually works, and those it doesn't effect.
-- 2 fries shy of a happy meal, Nov 16 2015

//The only problem I foresee is that it's probably difficult, regulations-wise, to administer a substance which you know will have only adverse effects.//

Of course the placebo has a positive effect (by definition), so that's okay.
Much harder to get clearance for the nocebo pill, though.

And actually, side-effects don't have to be bad. Tetracycline clears acne, niacin (taken for e.g. arthritis) lowers cholesterol levels etc.

...although, it's going to be an issue if the placebo is so successful that no study can find an improvement over it.
-- Loris, Nov 16 2015

Are you watching the second series of Fargo that's on at the moment? A storyline in that made me consider similar thoughts. Merely knowing you're in a trial and only 50% likely to be taking "proper" medicine must have a physiological effect itself. Ideally everyone should believe that they are taking the real thing, so a placebo that makes you woozy is a great idea. That, or keep a randomly selected portion of the population ignorant of these kinds of practices, and do experiments just on them.
-- zen_tom, Nov 16 2015

Placebos are one of those things that science worshipers use to lie to their followers.

Take, for example, a set of studies to determine the efficacy of a COX inhibitor in reducing discomfort (pain and inflammation) in arthritic knees. If I've been tasked with showing this product actually works, I'm going to run it up against a corn-oil or corn-starch based placebo. Corn is "well known" to be "neutral" in its effects, although it actually does cause low-grade inflammation in a large portion of recipients.

The COX inhibitor, although efficacious for pain, doesn't help the inflammation; but, since the chosen "placebo" causes some inflammation, the COX inhibitor appears to work better.

On the other hand, I could run the same clinical study, same design and doses and controls, except the "placebo" is now olive oil or olive extract. It exhibits very little inflammatory effect, so the COX inhibitor comes off not looking nearly so good. So I sell that study to the competitor.

The *really* cool part is there's no requirement to specify what placebo was used in any given study. In my last employer's library of around 22,000 studies on vitamins, minerals and supplements, about 8 percent stated the composition of controls. This number didn't change much when the studies were grouped by to-show versus to- disprove.

I've seen numbers that indicate pharma studies are pretty much in the same boat as VMS studies in this matter, but I don't have the personal experience with those.
-- lurch, Nov 16 2015

What I was thinking about, was the need for any dual testing with anything other than the drug on trial.

The truth, as I see it, is getting a baseline in the recipient, anything you do after that, will show. Modern sensing and data acquisition should make this possible.

Though, a placebo with tweaked side effects would make a nice drug for data trial.
-- wjt, Nov 17 2015

I think it should give you rainbow coloured shit.
-- 4and20, Nov 17 2015

// rainbow coloured shit// This means a 'white hole' T shirt is fully deserved.
-- wjt, Nov 18 2015

[zen] The second series of Fargo is excellent. I thought the same thing about the placebo storyline.
-- hippo, Nov 18 2015

random, halfbakery