Placebo effect

[blutoski]

Are you claiming that it doesn't exist? You'd be waaaaaay on the wrong side of the evidence if so, unless you're attempting to make a much subtler point than you stated.

I disagree, and don't mind repeating myself on this issue because I think it's important.

I'm saying it's so weak as to be irrelevant. My concern is that there's a lot of hand-waving about the effect as if it is known to have medically significant results. The debate about whether it's ethical to use it as a substitute for other treatments, or as a solution in the absence of viable alternatives is jumping the gun, because it assumes that there's a there there.

It's like when I do ghost investigations and my fellow investigators head to the library to research previous owners of the property to see who might be haunting the place. They jump right over the 'is there evidence of haunting' step and lunge into the fun part: the explanation.

Is it immoral to lie to someone if doing so is known to increase their likelihood of becoming healthy? (And thereby making the lie true...)

(Not intended as a flippant reply - that's the crux of it after all.)

Fair enough. It does violate the current standards of medical ethics, which revolve around patient informed consent. Doctors can lie to patients 'for their own good' in lots of ways. They used to sterilize young women 'for their own good' and tell them it was an appendectomy *. Let's not restore this licence.

Further, on a personal basis, as a Kantian, my answer is that lying is always unethical.



* As some may have discovered, I have an interest in eugenics. I'm writing a book on Canadian eugenics programs, particularly in BC and Alberta, and the political environments that produced and eventually dissolved the legislation. There was a woman a few years ago who had appendicitis, which was a shock, since she had this appendectomy scar. Hello! Tubes tied when she was 12! The relevance is that giving practitioners the authority to circumvent informed consent is more than just a slippery-slope: it undermines the cornerstone of modern medical ethics.

 

[blutoski]

ok this is from the http://skepdic.com/placebo.html website

with some question as to the effectiveness of placebo

but it would suggest that this research whilst interesting, would require further investigation before concluding that placebo was "powerless"......

I've had a problem with Carroll's entry for quite some time. He's conflating two common uses of the expression 'placebo effect'. There are two common uses:

1. Baseline + placebo
2. placebo alone

The first part of his entry on 'placebo effect' refers to the first type, including thoughts about regression to the mean, &c. These are actually baseline phenomena that also appear in a non-treatment group that does not receive placebos.

The second part of his entry focuses on the other meaning, and the studies he cites are designed to specifically look for what, exactly, a placebo does in a study, over and above baseline. It looks pretty trivial. Most of the improvement in placebo groups appears to happen in non-placebo groups, which means that administering a placebo may not have much effect.



I think this has led to much confusion, and I'm not sure how to go about spreading knowledge of this so that the discussions can be more fruitful. Unfortunately, I'm wrestling with my server tech support people right now (they merged and relocated my account to another server park and the settings got botched), or I would post a link to the BCSkepticsWiki entry on the subject, which I had hoped would serve this purpose. When the technical problems are resolved, I'll follow up in this thread with a link.


Regarding ethics: since placebo groups by definition include effects from baseline recovery, not caused by the placebo itself, and since many studies with non-treatment groups show that the non-treatment groups can have a positive recovery rate, the state of the art is to be honest with the patient and say: I have nothing for you, but studies show that this problem resolves itself much of the time.

 

[blutoski]

How can you make a double blind study for placebo?
As long as the researcher doesn't know who is getting the placebo you get a single blind study, but How can you make the patients blind?

I don't think you can. That's fine, since the placebo is hypothesized to work *because* of what the patient expects it to do.

 

[andyandy]

Fair enough. It does violate the current standards of medical ethics, which revolve around patient informed consent. Doctors can lie to patients 'for their own good' in lots of ways. They used to sterilize young women 'for their own good' and tell them it was an appendectomy *. Let's not restore this licence.
.

As an interesting (well...maybe) aside, in Japan it is common practise for a patient diagnosed with a terminal illness not to be informed of the fact by the doctors. Instead the doctors will tell the next of kin - who then have to choose whether or not that information is in the best interest of the individual......

which i think is justified by the reasoning, well if you've only got 6 months left to live, you don't want to spend that time worrying about dying......
but seems (from a western perspective) to be a violation of doctor-patient trust. I guess as with placebo, you have to decide the greater good - but the difficulty comes as to from who's perspective the greater good is chosen.....

 

[Jeff Corey]

http://www.nytimes.com/library/magaz...g-talbot7.html

"Page not found".
Try again.

 

[Dancing David]

Woops. It appears the quote from grayman from another thread isn't printing, and the link isn't working. He said, basically, that studies have shown that 75% of the effectiveness of anti-depressants is due to placebo effect. Let's assume such studies do exist, for the moment.

I don't have an agenda, I'm just asking questions.

1) Surely, all accounts from people taking anti-depressants supports the fact that these people feel different than without them. It's another question whether they are being cured of depression. Can we speak of placebo effect in cases where a drug is obviously psychoactive, as opposed to a sugar pill?

ADs are most likely somewhat akin to giving a diabetic insulin, it does not cure the condition, that is for sure.

2) I'm very surprised to hear that the very existence of placebo effect is being questioned. Aren't there any studies that test whether placebos work better than no treatment at all? (Maybe this is something I've just assumed, wrongly.)

i am being the devils advocate here, i just realized that i hadn't seen anything to support that the placebo effect is actualy significant.

So I asked, just like I would with any other issue.

I consider this a live issue for many people. The related issue of the effectiveness of psychiatric treatment is pertinent at JREF because so often, people who claim paranormal abilities are being told to consult a psychiatrist.
Yet there is some question what, exactly the psychiatrist could actually DO for them, other than keep them out of trouble....

Hope this isn't off-topic. Tried to look for a thread where this is being discussed...

Start a psychitric thread, i will be there, I suggest you look through Stamenflicker threads.

 

[Dancing David]

Well psychiatrists do more than just prescribe drugs, there is that whole thearpy thing that you could say is one big placebo

Only iof you want to pay for therapy with a psychiatrist, usualy about $300/hour on up!

 

[Dancing David]

just to add: Dancing David, your post came up while I was typing mine.
Here's the "plain language" summary:

"No evidence that placebo treatments in general have important effects, although they may have small effects on patient-reported outcomes, for example pain

It has been widely believed that placebo (dummy) treatments (for example sugar tablets) are associated with substantial effects on a wide range of health problems. However, this belief is not based on evidence from randomised trials that use a placebo treatment for one group of people, while another group receives no treatment. The effect of placebo treatments was studied by reviewing more than 150 such trials covering many types of health care problems. Placebo treatments caused no major health benefits, although they possibly had a small effect on outcomes reported by patients, for example pain."

I buy this. My question now is: in the case of depression, are you "better" if you think you are? Not necessarily, I guess.

I used all the techniques of CBT(cognitive behavioral therapy) for a long time to change my depression with OCD, I was able to keep myself functioning. But it did not change the early waking and the frequent rousing during my sleep cicle. So for me ADs are able to reduce the OCD to a very low level and they help me to not have disturbed sleep.

 

[Dancing David]

if placebo is a myth why is it standard practise to carry out double blinded trials? Is the medical establishment taken in as well?

I was playing devil's adovocate. If someone is suggesting that there is a therapuetic benefit to the placebo effect, then it would be good to demonstrate it is caused by the placebo, and not just the passage of time.

 

[drkitten]

So a Dr. knowingly selecting patients, excluding known subjects that don't perform the way he wishes to adjust the desired outcome is placebo effect?

No, that's "scientific misconduct" (aka fraud).

now, I'm just being silly

Nothing new there.

 

[blutoski]

As an interesting (well...maybe) aside, in Japan it is common practise for a patient diagnosed with a terminal illness not to be informed of the fact by the doctors. Instead the doctors will tell the next of kin - who then have to choose whether or not that information is in the best interest of the individual......

which i think is justified by the reasoning, well if you've only got 6 months left to live, you don't want to spend that time worrying about dying......
but seems (from a western perspective) to be a violation of doctor-patient trust. I guess as with placebo, you have to decide the greater good - but the difficulty comes as to from who's perspective the greater good is chosen.....

Mm. The problem isn't just about doctor/patient confidentiality: it's about respect for the patient's autonomy and right to relevant information. Patients have the right to know whether they need to get their affairs in order and say final goodbyes. What a waste of your final days if you never get around to doing the few things that you think matter, because you figure you'll have plenty of time when you get well.

It also puts incredible power in the hands of the relatives who are notorious for manipulating situations for share of inheritance.

The bottom line is that neither the doctor nor relatives are in a good position to decide what's in the patient's best interest in this case because of incomplete information and conflict of interest.

 

[joobz]

No, that's "scientific misconduct" (aka fraud).

Exactly, an issue that can be blocked by double blind studies. which goes with my original statement
Doubleblind isn't for just placebo-effect

----
Now to those who challenge the existance of placebo. There does seem to be some psychosomatic effect that exists. I can think of people who lower their heart rate through thought/relaxation. Same as the monks who raise body temperature.

a simple example is stress and heart conditions. To place yourself under mental stress/nervousness. Heart rate goes up, blood pressure goes up-heart gives out.

But to mentally avoid the stress in the first place, you'd avoid the state that brought on the disease.

Anyway, point being frame of mind does impact our physiology. Why couldn't this adjust the outcome?

 

[ponderingturtle]

Exactly, an issue that can be blocked by double blind studies. which goes with my original statement
Doubleblind isn't for just placebo-effect

No you are wrong. Doing that unconsiously is a large part of the placebo effect, and that is why there are double blind studies. Intentionaly messing with data and groups is not what double blinding is for(peer review and such is what controls outright fraud)

Remember that the homeopathic studies with animals do not require fraud to get results that homeopathic techniques work, just poor design

 

[saizai]

Placebo treatments caused no major health benefits, although they possibly had a small effect on outcomes reported by patients, for example pain

Since when is a reduction in pain not a major health benefit?

 

[ponderingturtle]

Since when is a reduction in pain not a major health benefit?

It is, the problem is that pain is very very subjective and expectaion on the behalf of the participant or the doctor can lead to a percieved improvement where there is none.

 

[saizai]

ponderingturtle - You just defined the placebo effect.

Except, with pain, perceived improvement == real improvement.

Unless you're of the (somewhat perverse) sort of philosophy that says you can be in pain but not know it?

 

[ponderingturtle]

ponderingturtle - You just defined the placebo effect.

Except, with pain, perceived improvement == real improvement.

Unless you're of the (somewhat perverse) sort of philosophy that says you can be in pain but not know it?

The problem is that the perception of improvement might not be on behalf of the patient but in the side of the observer. So if I think you are doing better does that equal a real improvement?

 

[joobz]

No you are wrong. Doing that unconsiously is a large part of the placebo effect, and that is why there are double blind studies. Intentionaly messing with data and groups is not what double blinding is for(peer review and such is what controls outright fraud)

How can I be wrong when every dictionary out there agrees with me. (see previous posts)

Placebo effect = real outcome due to the expectation of that outcome.

This doesn't state the placebo effect is real, it just states what it is. It has been defined as such. end of the discussion.


Your definition of placebo effect is beging the question.
You describe the placebo effects as all of the things that are spoilers to it. It is these spoiler reasons that are controlled for in a double blind study.

Now, unless you go back in time, change the original definition of the placebo effect and/or rewrite all accepted definitions of it, you are simply wrong.


---
As an aside. Double blind trial does help control against fraud. It prevents fraud the researchers conducting the actual experiment from faking results. You are right however taht Peer-review and external oversight controls against fraud, but it is the fraud committed by the person analyzing the final data set and writing the report.

 

[saizai]

The problem is that the perception of improvement might not be on behalf of the patient but in the side of the observer. So if I think you are doing better does that equal a real improvement?

You conflated.

If *I* think *I* am doing better, then by definition I am (at least for mental or experiential measures).

If *YOU* think *I* am doing better, then there might be observer bias involved that needs to be eliminated (eg via double blind).

Edit: On rereading your statement, I think we were in agreement after all.

 

[ponderingturtle]

How can I be wrong when every dictionary out there agrees with me. (see previous posts)

Becuase you stated that if I think you are feeling better that is an improvement irregardless of how you actual feel.

Also dictionaries are not the best choice for technical meaning, unless they are focued on technical meanings

Placebo effect = real outcome due to the expectation of that outcome.

And this is wrong, you can not easily differeniate a real outcome from a percieved outcome, especialy when the one perciving the outcome might not be the patient.

This doesn't state the placebo effect is real, it just states what it is. It has been defined as such. end of the discussion.

No is defines it as a real outcome, not a error of perception and observation, this it says it is real. End of discussion.

Your definition of placebo effect is beging the question.
You describe the placebo effects as all of the things that are spoilers to it. It is these spoiler reasons that are controlled for in a double blind study.

A placebo is a research technique to remove random variance and observer error. The point is that the technique for double blinding is useing placeboes, and you can not tell if a change is through error or real change. So you can not be sure if there was a real improvement or if there was only a percived improvement.

As I said you feel like crap, I give you something, I think you are feeling better after even if you say you are not. That is also the placebo effect.

Now, unless you go back in time, change the original definition of the placebo effect and/or rewrite all accepted definitions of it, you are simply wrong.

And you seem to have little understanding that you can not seperate out any of the many effects that create the apearance of change in the placebo group.

---
As an aside. Double blind trial does help control against fraud. It prevents fraud the researchers conducting the actual experiment from faking results. You are right however taht Peer-review and external oversight controls against fraud, but it is the fraud committed by the person analyzing the final data set and writing the report.

Protection agenst fraud is not what it is for though, it is to protect against observer errors from expectation. There are some sorts of fraud that it would make more difficult, but that is not why it is done.