Talk:Placebo effect (disambiguation)

I'm somewhat concerned...

I'm somewhat concerned this article doesn't conform to NPOV. The section titled "does the placebo effect exist" comes down pretty hard against the existence of the effect, seemingly on the basis of th authors. Does anyone else-- and especially User:Cacycle, who originally wrote that section-- have comments about this? I'm inclined to temper the conclusions of the discussion, but also shade it back somewhat toward the earlier scientific consensus. Tobacman 19:34, 6 Dec 2004 (UTC)

Actually I haven't read those two articles and the published replies to them, but I will do that soon. Both publications are metastudies that merely review previously published studies. To me it looks like there was no real scientific consensus before based on facts and data. It was more an unsupported and unquestioned belief into this effect. Those metastudies seem to be the first attempts to verify the effect.

If there are serious reasons to question the importance of the placebo effect then this must be adequately reflected in the article. Somehow 'softening' the well substantiated arguments of this party would clearly be against the NPOV. Moreove, in its current revision this sceptical standpoint is clearly separated from several other sections that assume the existence of the placebo effect. There is also no biased final conclusion.

Cacycle 21:59, 6 Dec 2004 (UTC)

Good!-- It sounds like we agree on the criteria for what should be included and why. Now we both need to do our homework...

But in the past there certainly has been a scientific consensus about the existence of the placebo effect, cf the fda, the nih, the skepdic, a book by a Harvard prof, and the fact that placebo controls are generally still required in clinical drug trials. Interestingly, Anne Harrington indicates in her book that there was a long period of denial about the placebo effect before people began to control for it.

Three last comments:

1. Some recent research (and cites therein) identifies neurological mechanisms that could underlie the placebo effect.
2. One way that perhaps ALL of these researchers (ie, Hrobjartsson and Gotzsche, and the others cited in the article) could be right, is if the size of the placebo effect has fallen over time. That would also be consistent with the sensible conjecture that patients now expect more treatment and medication than they used to. (I'll look for a citation for this hypothesis, since this speculation is obviously inappropriate for the article.)
3. Last, it seems plausible that there are some circumstances in which large placebo effects can reliably be generated and other circumstances where they can't; and there are some methodologies that are useful for measuring placebo effects and others that are deeply confounded. In the end, presumably a really good article will discuss size of placebo effect as a function of these characteristics, rather than (as we've both been doing) just asking whether or not the effect exists. Tobacman 09:37, 7 Dec 2004 (UTC)

One possible improvement would be to move up the sections on 'Placebo and Pain' and 'Placebo and Depression' so that they come before 'Does the placebo effect exist?'. This would emphasise the relative importance of the two points of view within the relevant scientific community. The section questioning the orthodoxy would thus remain, but would not have undue importance. --John G Walker 14:26, 8 December 2005 (UTC)

Sounds like a good idea. --JWSchmidt 15:23, 8 December 2005 (UTC)

I've changed the order. Does this look more NPOV now? --John G Walker 09:45, 2 January 2006 (UTC)

More content

Please take a look at this new content. [1]

Living in Denmark, I have followed this research from when it was in progress. Very interesting! It explains a lot of things that the old views didn't do very well. To maintain a NPOV, I have attempted to limit my comments to explanations of that research. -- Fyslee 22:23, 2 January 2006 (UTC)

I am not a doctor

I am not a doctor, but I think it should be clear that for some illnesses the placebo effect can not possible be effective. The organism can only do so much to heal itself. There must be some problems that can't be fixed, no matter what signals the brain sends out. Paranoid 10:30, 19 Mar 2005 (UTC)

I agree that this article is a bit POV. My impression (true or not) is that after most of the article was written, someone who believe the placebo effect does not exist came by and added the "does the placebo effect exist" section. There is absolutely nothing wrong with having the info in the article, but that section should also refer to studies that support the existance of the placebo effect (for example: [2] [3]). As it is now, the article feels like a debate rather than an encyclopeic article. --Apoc2400 19:23, 25 August 2005 (UTC)

As someone who have experienced the healing effects of self-suggestion and meditation (immideate healing of minor physical diseases, such as sore throat and cardiac symptoms), there can be no doubt that the so-callled placebo effect indeed exists, but that the strenght of the effect varies greatly between individuals, wherein those with the most pronounced efficiency (and thus important in medical treatment) usually comprise a definite minority (such as those with IQ's above 120, or below 60, etc.). I think thus it will be impossible to draw clear conclusions from the trials referenced in the article, as most of the casual study participants would statistically have somewhat under-developed psychic healing abilities. Sort of conducting a casual population intelligence trial and expecting "significant" percentage of IQ's above 140, for instance. -- User:152.94.19.164 13:17, 26 January 2006

NPOV

As you can see, I made a bunch of edits (one I accidentally marked as minor, sorry). This is an initial shot at NPOVing the article without getting into a massive edit war. The most basic problems are that: a) the placebo effect is NOT controversial, it's widely accepted in the scientific community and b) the "placebo effect is fake" view is represented even more than the prevailing scientific view (that 99% of scientists support). That's a clear violation of NPOV.

NPOV can preserve old errors

If the conclusions of this large study of the literature [4] is correct (and there is nothing to indicate otherwise), then it should be better represented. Otherwise a strict adherence to the NPOV will be a biased support for old misunderstandings, IOW it won't be neutral at all, but simply conservative resistance to change, and impediment to progress.

If correct, it should, with time, become the predominant viewpoint, with the old viewpoint remaining in its invaluable role as a tool for quackery. Quacks will not relinquish the old viewpoint, for obvious reasons. -- Fyslee 22:45, 2 January 2006 (UTC)

continued...

At some point later I may expand the section supporting the placebo effect, because I know there have been NUMEROUS studies on it on many scales in many countries. The attribution of it mostly to a single 1955 study is just a *wee bit* misleading.

• side tangent--you may ignore*

I've looked at the abstract for the article in question, and I suppose it should be kept in to support the minority view, but I must say, it is very pseudo-scientific and its conclusions don't support its data at all.

Many large scale studies have shown improvements from the placebo groups. These are for disorders people have had for years, so it can't be explained away simply by saying that that they simply naturally healed or any such nonsense. The hawthorne effect also can't seem to explain away the nocebo effect, nor why such a large number of people in a large study would be eager to please researchers to the point that they'd lie.

Not just that, but biofeedback techniques have been shown to reduce severity many kinds of symptoms, including pain. Since biofeedback is entirely mind-controlled, it makes no sense to discount the phenomenon.

It seems to me it's just a few quacks who are trying to become famous that are trying studies like this.

• end tangent*

Nathan J. Yoder 04:17, 22 October 2005 (UTC)

My first time using thie edit function, appoligies if I am making a mistake. I wanted to point out that the assertion that a placebo, or anything at all can result in "objectivly quantifiable analgesia" is incorrect. There is no objective measure of pain in existance. All a doctor can do is ask a patient about their pain, but he can never measure it directly as he can measure swelling or inflamation directly. One of the central problems regarding pain management is the fundamentaly subjective nature of pain. I susspect what was meant was merely that the patients reported less pain, which is subjective and therfore impossible to prove or disprove.

Major rewrite

Major rewrite I thought there were POV problems, although it was hard to exactly put my finger on it. I started a minor rewrite and it turned into a major one, with a lot of checking the PubMed database to make sure I was up to date. I found it very difficult to find a way that felt right to keep all the pro-placebo stuff at the top and all the anti-placebo stuff at the bottom, as was suggested here previously. I think I'm being up front and NPOV to say that many studies see a placebo effect, but few studies have studied the placebo effect, and then to lay out the pros and cons. (In fact, I am a medical researcher and I would never think of doing a study without some kind of control group.) For better housekeeping I incorporated the outside links into the article rather than a separate section at the end, and I pared down the References a bit as it seemed topheavy with placebo effect opponents. Thatcher131 03:50, 12 February 2006 (UTC)

"Major"....? No joke! Even if you are 100% right, and your rewrite were a 1000% improvement, it's still a bad idea to do such major rewrites, and it could be considered vandalism of other's work. You are undoing the work of many others who have used weeks, days, and hours to arrive at agreement on the existing article, often struggling over details of exact wording and shades of wording. Please show a little consideration for others next time (hopefully no "next time"). I see that you are doing this on other articles, and you're going to create a lot of hard feelings. You aren't the only one here, nor the only expert.

I'd suggest you undo your major rewrite immediately and start editing more carefully, a little bit at a time.

Instead use this page to discuss such major editing (and minor editing) before deleting the work of other editors. -- Fyslee 11:22, 12 February 2006 (UTC)

I think the compare function is somewhat misleading and does not reflect that I did preserve nearly all of the concepts and thoughts expressed by others before me, if not their exact language. The article also had a NPOV tag for more than a year, and there were definitely still parts of the "anti" section that sounded like advocacy. I was also having problems with the WP servers so I marked up the article in a word processor and loaded it in all at once (that's not an excuse but an explanation). I will consider your recommendation but I would like to see other's comments first. Thatcher131 12:50, 12 February 2006 (UTC)

I agree about the "compare" function. It makes it very difficult to see precisely what's going on. This fact is sometimes used by unscrupulous (I am definitely not accusing you!) editors who wish to sneak significant (but small) changes "under the radar." They then claim that they have preserved everything, but a careful analysis reveals that they have changed significant words, or deleted viewpoints wish they find objectionable. To avoid such problems and suspicions, it is usually best to make edits gradually. Take a sentence or paragraph at a time, then wait for reactions before proceeding. Major edits should be discussed first here on the talk page.

I'm speaking from experience. When I started here I took the "edit boldly" policy quite literally. Not understanding the background for an article's creation, I unwittingly showed great disrespect for all the other editors who had sweat blood and tears in the creation of the article. Fortunately they usually simply reverted all of my edits and I had to proceed more carefully, sometimes arguing my position point by point. -- Fyslee 15:00, 12 February 2006 (UTC)

I don't really appreciate the use of the term "vandalism." After thinking about it I have decided to revert the article and work on it more incrementally. Thatcher131 19:46, 12 February 2006 (UTC)

I take my hat off to you! I do apologize for the use of the term, and I wasn't accusing you, just stating that "it could be considered"...... I am convinced that your major rewrite was done in good faith, and therefore the term vandalism certainly doesn't apply in this case. Your willingness to start over, and still contribute to this project, reinforces my opinion that you not only mean well, but can be a valuable contributor here. Thanks for your understanding. Good luck! -- Fyslee 20:15, 12 February 2006 (UTC)

Starting over

I reverted to the last version before I posted my major rewrite. I will do a series of incremental changes and explain any time I change someone else's comments. I will watch this space and respond politely to polite comments.

In the "Placebo and pain section" changed the intro sentence; "best research" is a matter of opinion. Removed the redundant statement about schizophrenia. Removed the statement added by DanielCD about needing to be able to control the body in order to control the mind as it seemed more like a personal conclusion/POV than an encyclopedia-style summary of the field. Thatcher131 19:59, 12 February 2006 (UTC)

OK, its 20:25 on Sunday and I added everything I thought was important and cleaned up the links. The only thing I have left to do is tackle the Hróbjartsson & Götzsche section. I will wait at least a day and really think about it. (Check my edits one at a time because the compare function makes it look like I took out other people's writing when I just moved it a bit.) Thatcher131 20:47, 12 February 2006 (UTC)

I can't put my finger exactly on why the H&G "no placebo" section seems wrong, maybe its not really an NPOV problem but it isn't really written like an encyclopedia. I will try a test rewrite and post it here first for consideration. Thatcher131 20:54, 12 February 2006 (UTC)

I didn't consider the formerly implied "either/or" situation as being viable. That was implied in the former heading - "Does the placebo effect exist?" - and was basically a straw man argument. I changed it to the current "Objective or subjective effects?"

The placebo effect exists, but its nature (objective or subjective) has been the subject of much confusion, with quacks claiming its physiological healing effects to be so strong as to justify its exploitation in the use of worthless and dangerous "remedies."

It's not so much that H&G's conclusions are at all new - scientists have been working with this assumption all along - but that they base their conclusions on a thorough analysis of the available literature, exposing the weak basis for many if not most of the claims made for placebo's "powerful"(!) effects. Those effects must be reckoned with, but being primarily subjective, should not be exploited in clinical practice, but be fully understood and accounted for in research settings.

You removed your important thoughts, which were good observations:

"....rather, let me say that I think the placebo effect seen in clinical studies is a combination of many factors, including regression to the mean, poor study design, the natural history of the disease, the attention/reward phenomenon identified by H&G, and (sometimes but not always) a real placebo effect."

You're quite right, and that is a problem, since any effect that can be explained by any other name shouldn't be classified as part of the placebo effect: regression to the mean is just that; poor study design is just that; natural history of disease is just that; etc.

When all those factors have been ruled out, what's left is the true placebo effect. In practice this is nearly impossible to achieve, but a serious attempt should be made to reach this ideal. Such efforts will be rewarded with more reliably consistent - and consistently reliable - results. -- Fyslee 21:28, 12 February 2006 (UTC)

I looked over the Hróbjartsson & Götzsche/subjective effect section, and I think there are basically two problems that relate to POV. First, the section gave the conclusion before the results. So I wrote it to describe clearly (I hope) what they did, and what they found (no objective effects, only subjective effects).

The second problem was that there was a lot of material that related to the fact that there was a controversy without actually saying there is a controversy. Especially in the section "Implications" where the text was really trying to draw conclusions and make statements about the use of placebos. What I intend to do is solve this problem by adding more factual information about the controversy. You can describe a controvsery from a NPOV if you fairly report facts from both sides. I have read the original articles and the followups and I think I can explain the issue fairly. Thatcher131 16:59, 14 February 2006 (UTC)

I will finish up placebo in medicine in the next 12 hours, with references. Thatcher131 17:23, 14 February 2006 (UTC)

I am very impressed with your thoroughness and attention to detail! This is going to be a good article, in large part thanks to your efforts, so congratulations are in order. It's looking really good. You are keeping a good balance and I haven't noticed any grating POV problems. -- Fyslee 23:24, 14 February 2006 (UTC)

I very much appreciate the kind words. I hope others agree. I'm pretty much done for now. I think there is a possibility to write a section on alternative medicine--for example, the assertion of some homeopaths that since homeopathy works no better than placebo, therefore they both work--but I'm not moved to write it at the moment. Thatcher131 00:54, 15 February 2006 (UTC)

Concerns

I have some concerns about one section:

• The use of placebos in medical practice

"The ethics of prescribing placebos in medical practice is highly debated. The use of placebos is often justified because it will do no harm and may do some good. With the publication studies by Hróbjartsson and Götzsche and others, the proposition that placebos may do some good is under fire."

Below is something I wrote in an article about so-Called "Alternative" Medicine (sCAM) some years ago. I believe that especially the second paragraph speaks to this point. Is there some way these concerns can be included in a NPOV way?

Poor ethics can reveal themselves in everything from the frequent and well-intentioned use of undocumented methods and ideas to direct swindle and deception. Self-deception, with a suppression of the conscience, is an important factor here. "The end justifies the means" is a common rationalization. As long as something "works" it makes no difference if it is objectively proven or not. Subjective experience is considered to be adequate, with abundant anecdotes considered as sufficient and absolute proof. Thus empiricism is weighted too heavily and the placebo effect is exploited for all it is worth. Such a person wouldn’t dream of refusing to fulfill a patient’s wish to receive an undocumented form of treatment.

Consequently, the practitioner loses track of what it is that really "works" with a treatment. The patient is deceived and the practitioner deceives himself. This is not a professional way to act. By thus losing touch with reality, serious illness may continue to go undiagnosed and improperly treated, while both the practitioner and the patient put their confidence in the undocumented treatment. -- Fyslee 22:51, 15 February 2006 (UTC)

Quick change because what I wrote is not what I meant to write. I will think about it in depth later (I have about a half inch thick stack of articles on this now) Thatcher131 23:02, 15 February 2006 (UTC)

I have bestowed a barnstar on your blessed head! Look at your User page.....;-) -- Fyslee 23:14, 15 February 2006 (UTC)

self-proving!!??

in "Conditioning" section, there is an example for conditional responce.

For example, a person who routinely takes aspirin for pain may experience pain relief when given a pill that looks and tastes like aspirin.

isn't it a self-proving example for Placebo effect? a better example is needed.

--Hkchan123 13:59, 7 May 2006 (UTC)

• I took it out. It seems fine without any "layman's example" since the results of a study is described right after. Thatcher131 03:04, 9 May 2006 (UTC)

Objective or subjective effects?

In the article, it said:

They performed a meta-analysis of 156 clinical trials in....

However, according to the source quoted[5]

RESULTS: We identified 130 trials that met our inclusion criteria. After the exclusion of 16 trials without relevant data on outcomes, there were 32 with binary outcomes (involving 3795 patients, with a median of 51 patients per trial) and 82 with continuous outcomes (involving 4730 patients, with a median of 27 patients per trial).

130 vaild trials + 16 invaild trials....it should be146.

Here I corrected it. --Hkchan123 12:03, 6 May 2006 (UTC)

That may have just been a typo when I went over that section. [User:Thatcher131|Thatcher131]] 12:07, 6 May 2006 (UTC)

If I interpret it correctly, the actual number was 114 (130 minus 16, which leaves 32 plus 82 -- which equals 114):

We identified 130 trials that met our inclusion criteria. After the exclusion of 16 trials without relevant data on outcomes, there were 32 with binary outcomes (involving 3795 patients, with a median of 51 patients per trial) and 82 with continuous outcomes (involving 4730 patients, with a median of 27 patients per trial).

That number could be used with some rewording to account for the discrepancy between 130 and 114. -- Fyslee 17:48, 8 May 2006 (UTC)

156 is correct. If you look at the two studies cited in the footnotes, the first is 130 minus 16 unusable, for 114. The second adds 42 more valid studies, for a total of 156 in the second meta-analysis. Thatcher131 18:04, 8 May 2006 (UTC)

Once again you reveal your excellent grasp of these things. Good going! I have now fixed the links so they are proper references, and the References section actually works! -- Fyslee 19:12, 8 May 2006 (UTC)

Except that the reference section is now a mixture of styles. Although I am a supporter of cite.php, I think this article would do better with Harvard-style references. I can convert tonight if you don't object. The other option is to convert all the reference links to cite.php refs and then label the ones which are not cited in the article as "For additional reading" or some such. Thatcher131 19:15, 8 May 2006 (UTC)

Fine with me. We just need to get the references working so they don't fill as much in the text area and are clickable. I think we should also provide URLs as much as possible. The way I've done it is just copied from another article, but the Harvard system looks good too. Go for it! -- Fyslee 19:23, 8 May 2006 (UTC)

This is how the harvard style should be done in my opinion. There is a place for cite.php style notes when a comment needs to be made that doesn't easily fit into the author-date format, and then the references in author-date format organized alphabetically. They all have URLs or PMID links (that's PubMed ID; clicking on it will take you to the abstract in PubMed. The PMID is a permanent number assigned to each published manuscript, unlike urls that could change or get put behind subscription walls). I also reverified and added PMID and ISBN links to the additional references. I think Senn is way over-cited, but it's not a battle I want to pursue at the moment.

It is possible to convert the Harvard citations in the text to blue links that, when clicked, will take the reader to the full citation at the bottom of the page. I'm not sure that's necessary, since its a relatively short article. However, I can add them if you feel strongly it should have them. (There is an example at WP:HARVARD.) It should be less work than tonight since everything is basically formatted and its a matter of adding the right tags. Thatcher131 03:10, 9 May 2006 (UTC)

The article seems to cover the topic in an informative and balanced way. One comment: The first section states that "Recently, it has even been shown that 'mock' surgery can have similar effects". This isn't exactly correct. The study that examined whether surgeries should be subjected to placebo-controls -- the mock knee surgery study -- has been widely misinterpreted because of multiple/sloppy definitions of the placebo effect and some unfortunate (easily misinterpreted) language choices by the authors in the conclusion. The authors of the study allude to the Hrobjartssen results, pointing out that the improvements in the placebo group (which were as great as in the two groups given surgery) are not the result of the patient belief in the treatment but something else about the placebo. Well, interestingly, all three groups were given a fairly strict regimen of rest, analgesics, and rehabilitation, including the placebo group. There was unfortunately no comparison group given nothing, as it would have been interesting to know whether the "placebo" (the rest, analgesics, and rehabilitation) was the reason for the improvement. Reading the study leaves the strong impression (though not concluded) that the regimen of rest, analgesics and rehabilitation had a larger effect than the surgeries, hence the similar results for all three groups. Unfortunately, the language used by the authors in the conclusion, to say that the placebo effect was large in this study, has been misinterpreted by others as saying that patients' belief in surgery can cause a measurable clinical effect in surgical trials. The study doesn't say that at all. The authors seem to define the placebo effect as basically anything that happens to alter clinical condition that isn't the result of the actual procedure. This is a legitimate and important reason to use a placebo control, to root out researcher interpretation bias, patient reporting bias (which could be the underlying reason for both placebo and nocebo effects observed in other research), natural changes in condition, unknown influences, etc. However, this study scope was not sufficient to fully answer the question of whether placebos should be used in surgical trials, a question that should be answered more convincingly given the serious ethical considerations. That many people interpreted the study with old biases about the placebo effect and an unwillingness to fully consider the Hrobjarssen results objectively has led to widespread misinterpretation of the study's implications.

Spelling of Hrobjartsson, Methodology of Administration section

Good work, everyone! I learned a lot. I had no objection to the point of view, just the spelling of Hrobjarttson. Apparently it's an Icelandic name and in Icelandic, it's spelled Hróbjartsson. However, the scientist himself is lives in Denmark and Danish has no ó. Also, the Danish version of this article doesn't spell it with an accent, so I think it's more correct without.

I do have a problem with the "Methodology of Administration" section. It seems like an early version of this article. It makes some good points but jumps all over the place. Anyone interested in rewriting (or even deleting) it? Dblomgren 04:58, 11 November 2006 (UTC)

• Methodology of Administration

That section was a split from a similar artice Hawthorne Effect and probably does need some reworking. It was IMO out of place in the Hawthorn Effect article and I thought I would throw it too the wolves to break out what should be kept, modified, or deleted.M jurrens 23:03, 11 November 2006 (UTC)

References

This article is in desperate need of help as regards the references. I have added the references code, so now it is possible to turn all links and the existing notes into proper linkable references that don't take up space in the article. As much as possible we should also find URLs for references. One can look at other articles to see how it's done. More information here: WP:FOOT.

Here's an example of the format, taken from the subluxation disambiguation page. It shows the same reference being used later in the same article:

"Bla, bla, bla."

"Bla, bla, bla, bla, bla."

References

Here is what the code looks like:

"Bla, bla, bla."<ref name="WHO_guidelines">[http://www.who.int/medicines/areas/traditional/Chiro-Guidelines.pdf WHO guidelines on basic training and safety in chiropractic]</ref>

"Bla, bla, bla, bla, bla."<ref name="WHO_guidelines"/>

-- Fyslee 23:37, 11 November 2006 (UTC)