Talk:Placebo

Template:Vital article

Types of Placebos

From the very first sentence, this current definition of a placebo is very wrong on several points.

"A placebo is a simulated or otherwise medically ineffectual treatment for a disease or other medical condition intended to deceive the recipient" - placebos are seldom, if ever, prescribed or administered to deceive the recipient. Even in clinical studies, the recipient must be advised that some patient will receive a placebo. - placebos are often medically effective. Placebo effects are real improvements in the condition of the patient. Placebos are 'believed to be medically ineffectual', but in reality, they can be very effective. - placebos are treatments for an illness, a specific case of an illness. They are not treatments for a 'disease', a class of illnesses. The patient presents an illness. The physician diagnoses a disease, but treats the illness. This is an important, non-trivial distinction.

How to understand placebos 101

==================

There are two fundamental types of placebos, although this is seldom explicitly stated in any literature. The two types of placebos are prescribed with two different intentions, neither of which is 'to deceive the recipient'.

Placebos (real placebos) are medications or treatments prescribed by a physician with the intent to help the patient, when the physician does not know how to help the patient. The physician might believe that the prescription has no 'physical benefit' to the illness - but prescribes because he does believe it will benefit the patient in some small way. The physician's beliefs might be right, or wrong.

Note: The benefits resulting AFTER the placebo prescription has real causes. However, in many cases assigning them to the placebo causes much confusion and in many cases avoids actual investigation Calling them "placebo effect" is navel gazing, when the challenge is to understand what is going on outside of the medical system, outside of the medicine prescribed.

Clinical Placebos (fake placebos) are false medicines or treatments. They are not prescribed by a physician and there is no intention to provide any benefit to the patient. There is no intent to deceive the patient. Clinical placebos are used in scientific experiments to provide a statistical measure of the so called 'placebo effect'.

Note: The benefits resulting AFTER a clinical placebo is administered also have real causes. Ignoring these causes, naming them "placebo effects" assigning them to 'the mind of the patient' is simplistic nonsense, avoiding true investigation of the facts.

There are many fundamental difference between a real placebo and a clinical placebo. Unfortunately most references do not notice, much less attend to this distinction resulting in total nonsense and confusion. It's as if we used the same name for bears and teddy bears. Here's a comparative list of some differences:

Placebos - prescribed by a doctor - doctor doesn't know what is best - doctor intends to improve the health of the patient - prescribed to sooth the patient - can be active or passive. Active are more effective - works, statistically, but we don't understand (and seldom study) why or how - patient believes they are getting a medicine (in normal cases)

Clinical Placebos - administered by a scientist physician - researcher believes that the placebo is useless - researcher has no intention to improve the health of patients who receive a placebo - administered to measure statistically, the effects of medicine, by subtraction of 'clinical placebo effect'. - often specifically designed to simulate the activity of the drug or treatment being tested - when the placebo works, the research has failed - patient hopes they are NOT getting the placebo

When we assume that clinical placebos are the same as real placebos, we create nonsense because the assumption is nonsensical.

Because of this fundamental difference between real placebos and clinical placebos, it is very difficult to measure the effects of real placebos in a clinical study - and as far as I am aware this has never been accomplished. To create the real life placebo, it would require the doctor to be allowed to decide when to prescribe a placebo, and to choose the placebo being prescribed.

Thus, we can also see that 'real placebo effects' are quite different from 'clinical placebo effects', but that's another story....

Until this reality about, and distinction between a real placebo and clinical placebo is recognized, we will continue to publish nonsense about placebos, because we simply don't understand.

Another possible historical account

I found this in The Anatomy of Melancholy by Robert Burton (1621):

--- And sometimes a strong conceit or apprehension, as [1617]Valesius proves, will take away diseases: in both kinds it will produce real effects. Men, if they see but another man tremble, giddy or sick of some fearful disease, their apprehension and fear is so strong in this kind, that they will have the same disease. Or if by some soothsayer, wiseman, fortune-teller, or physician, they be told they shall have such a disease, they will so seriously apprehend it, that they will instantly labour of it. A thing familiar in China (saith Riccius the Jesuit), [1618]If it be told them they shall be sick on such a day, when that day comes they will surely be sick, and will be so terribly afflicted, that sometimes they die upon it. Dr. Cotta in his discovery of ignorant practitioners of physic, cap. 8, hath two strange stories to this purpose, what fancy is able to do... --- from: http://www.gutenberg.org/files/10800/10800-h/ampart1.html

Removal of section List of medical conditions

I believe the section List of medical conditions should be removed. Since placebo effects are noted in basically all clinical trials (as mentioned with citations elsewhere in the article), this section effectively amounts to a list of medical conditions that have been studied in a clinical setting.

Are there any objections?--Kavigupta 01:07, 19 May 2018 (UTC) — Preceding unsigned comment added by Kavigupta (talk • contribs)

It seems a bit redundant to include it to me. Looks like your removal was reverted with no discussion here. Maybe more input from others is needed. Perhaps a post on the appropriate project page, Wikipedia:WikiProject_Medicine, would lead to more discussion? Rap Chart Mike (talk) 18:00, 21 May 2018 (UTC)

I agree that this section is a) absurd and b) a ridiculously small selection of things that should be here, if we are going to include it. User:BullRangifer please justify keeping this. "it has been there a long time" is not a good enough reason. Thanks. Jytdog (talk) 04:12, 23 May 2018 (UTC)

If it's really bad, then remove it. Otherwise follow WP:PRESERVE and improve, not delete, it. -- BullRangifer (talk) PingMe 05:25, 23 May 2018 (UTC)

Definitely support removal; having a list of medical conditions where placebo testing has taken place would not only be excessive when the list is 'more complete' but also I don't see the purpose in it. --Treetear (talk) 11:18, 23 May 2018 (UTC)

I absolutely support removal as it seems this is out of place being in this article. RobP (talk) 12:25, 23 May 2018 (UTC)

• support removing section--Ozzie10aaaa (talk) 20:32, 23 May 2018 (UTC)
• I think you'll find that it's more complicated than that. First, placebo effects aren't found in everything. Placebos don't stop people from bleeding when they're cut, and they don't set broken bones, to name just two obvious examples. Second, and much more importantly, there is a difference between "an effect was seen in a person taking a placebo" and "the placebo effect". The placebo effect is far stronger and more important in subjective symptoms (e.g., pain and depression) than in objective conditions (e.g., the amount of blood that has to be mopped up after a car wreck). I think that it would be better to talk less about individual conditions and more about general themes here, but this list is not wrong, and it is most informative when you read it with an eye towards both what's present and what's absent. WhatamIdoing (talk) 23:50, 23 May 2018 (UTC)
  • The issue is that this list is a list of conditions for which there are trials that have some effect show up in the placebo wing, which is "an effect was seen in a person taking a placebo", so it doesn't really differentiate the two cases you have above. The only experiment that can draw that distinction is one in which there is a placebo and a no treatment wing (and even then there can be biased reporting), and that is not the kind of experiment that this chart is tracking. An attempt at a discussion of what you're describing would be beneficial (and is in fact the reason I started trying to improve this article, take a look at Placebo#Criticism). Kavigupta (talk) 02:54, 24 May 2018 (UTC)

Instead of telling us to read it "with an eye to" figuring out the inclusion criteria, could you just tell us what that criteria is? It looks like it's just "We found a meta-study about this"?

ApLundell (talk) 05:00, 24 May 2018 (UTC)

My comment isn't about the inclusion criteria for the list. It's about what's in the list (and what you would expect to find in any typical list about placebos) and what's not in the list. So, what's there? Lots and lots and lots of things involving pain and distress. It's all disorders that are primarily characterized by subjective, self-reported symptoms and conditions that are affect and are affected by conscious behavior. What's not there? Infectious diseases, conditions easily treated by surgery, and other disorders primarily characterized by objective problems with clear, direct etiology. WhatamIdoing (talk) 02:33, 25 May 2018 (UTC)

If that's the purpose, I think it's handled well in the text immediately above the chart, and doesn't need a chart (which only partially satisfies your criteria: food allergies, hypertension, and heart failure aren't on there). Also, you wouldn't expect many diseases generally treated by surgery to be on there because placebo testing has only recently become a thing for surgery. [Again, not claiming that there is a real placebo effect for any of these conditions, just that some of the linked studies don't really show one either way]. Kavigupta (talk) 03:01, 25 May 2018 (UTC)

Hypertension and heart failure are affected by behavior, and behavior is affected by placebos (also by the clinical trial effect). A remarkable number of people who believe themselves to have food allergies actually don't (or don't have a clinically significant allergy any longer), so it's not surprising that people who have a reason to be calm about the challenge (namely, that they think the "drug" will protect them) turn out to not report a significant reaction.

Placebo-controlled surgical trials have been done, but nearly all of them have been done for conditions already on this list. (See this report if you're interested.)

I think that the first two sentences of the ==Symptoms== section should be expanded significantly. We should spend more time talking about the general factors (e.g., works better for subjective than objective; works better if the placebo seems expensive or difficult or special) than we do on detailing just exactly how big it is in CFS/ME. Whether we namecheck things like asthma and irritable bowel syndrome – eh, I don't really care either way. But we need to beef up the overall explanation, and to more clearly explain the difference between spontaneous remission, unintentional behavioral changes, and the effects actually produced by placebos. WhatamIdoing (talk) 06:33, 25 May 2018 (UTC)

I could not agree more that we need to "clearly explain the difference between spontaneous remission, unintentional behavioral changes, and the effects actually produced by placebos". In fact, that is the reason I originally got interested in improving this article, before realizing that I should first try to fix what I viewed as some of the organizational difficulties first. I don't think that this chart really communicates the message of which areas are more affected by placebos than others, however. — Preceding unsigned comment added by Kavigupta (talk • contribs) 17:34, 26 May 2018 (UTC)

• Support removing section or importing all of Category:Human diseases and disorders. Natureium (talk) 21:46, 24 May 2018 (UTC)

unsupportable

The judgement "unsupportable" is not in the source. This article was changed to win a debate in homeopathy. I guess this comment will be reverted and I will be blocked for troublemaking. If anyone sees this before it's beign reverted and I am blocked, please keep arguing for neutrality. Read this piece from Larry about this small group of users that has managed to replace the principle of neutrality by a principle of scientism and spread the word. Best wishes. --rtc (talk) 11:20, 16 June 2018 (UTC)

Probably because they don't want to alienate themselves from the possibilities of future funding from Alt-Med supporters. -Roxy, the dog. barcus 11:23, 16 June 2018 (UTC)

I watched that thread develop. There is a reason Larry is no longer involved with this project. -Roxy, the dog. barcus 11:34, 16 June 2018 (UTC)

Yes, because the dignity of his neutrality policy was taken away by the edits of group of naive and radical adherents of scientism. Probably? I mean seriously? We can use probable cause to put our own judgements, unsupported by the source, into the article? --rtc (talk) 11:41, 16 June 2018 (UTC)

Science is very important—this is realism, not scientism. Tgeorgescu (talk) 12:07, 16 June 2018 (UTC)

Please focus on the issue at hand, which is the use of a judgement that is unsupportable given the source cited. --rtc (talk) 12:15, 16 June 2018 (UTC)

I guess that point has been answered more than a hour ago. Tgeorgescu (talk) 12:25, 16 June 2018 (UTC)

No it has not been answered in any way whatsoever. The word "unsupportable" is nowhere in the source, that's the only fact. The rest is editors arguing for unsupportable judgements with the use of further unsupportable judgements. The reverter says '"no evidence was found of any placebo effect in any of the studies cited by him" - it's fine' No it's not. This is a very old study that's being criticized, and it is based on even older sources. The critic's paper says that older study is fallacious and the effects observed by its sources can be attributed in obvious ways to explanations which do not rely on the placebo effect. The authors did not even look at data from the many more recent sources that were published since the 1950 paper they criticize. The author's actual conclusion is that "the existence of therapeutic effects of placebo administration seems questionable", and not that it is unsupportable. The word "unsupportable" was defended above with the ridiculous claim that the authors did not use that word merely because they feared that this could risk future funding. --rtc (talk) 12:37, 16 June 2018 (UTC)

The change has clearly improved the article, though it is still erroneously focusing on what the study has to say about the ancient evidence from before 1955. The actually relevant judgement of that paper is what it has to say about the placebo effect itself, and, as already stated, in this respect it says "the existence of therapeutic effects of placebo administration seems questionable". There are more recent papers with more citations than this one on placebo effects, why is it mentioned so prominently? It merely showed severe errors in interpretation of ancient sources by the popular Beecher paper, and did not contain any new reserach on the actual question. --rtc (talk) 14:10, 16 June 2018 (UTC)

Do you have a MEDRS-compliant source for the change you are proposing? Brunton (talk) 15:07, 16 June 2018 (UTC)

dafuq? The quoted passage is from the same paper. --rtc (talk) 15:29, 16 June 2018 (UTC)

What change are you now proposing? Brunton (talk) 15:47, 16 June 2018 (UTC)

As I already said, what the study has to say about the ancient evidence from before 1955 is rather irrelevant. Iif the paper is to be used at all its bottom line should rather be quoted, which is, as already stated "the existence of therapeutic effects of placebo administration seems questionable". And, as already stated, it should be called into question whether this paper should be cited at all in this context, rather than one of the more recent ones on the actual issue (about which this paper has nothing actually new to say), of which there are abundantly many in the article, such as http://cochranelibrary-wiley.com/doi/10.1002/14651858.CD003974.pub3/abstract or at least the one I ref'd here (unfortunately it didn't make it very long). As you can see I am mostly repeating myself. IMO this article was actually fairly fine before it was changed to win the discussion on homeopathy. The passage "However, some recent research has called into question..." has everything there is to say. Apparently the way it said it was not strong enough for my opponents from the homeopathy article who are apparently not satisfied with the placebo effect being merely called into question, but want it to be presented as a complete myth, such as to make completely sure it cannot possibly be used as a rationalization for homeopathy. --rtc (talk) 15:56, 16 June 2018 (UTC)

The 1955 paper is historically important in the context of the article, and therefore so is the paper calling its conclusions into question, for that reason. The reviews you cite both fail to find anything beyond subjective effects. Again, what specific changes do you propose? Brunton (talk) 16:09, 16 June 2018 (UTC)

I won't repeat myself a third time. In doubt, the article should be reverted to the version from before the dispute. The actually relevant bottom line statement "the existence of therapeutic effects of placebo administration seems questionable" does not go beyond "some recent research has called into question" --rtc (talk) 16:20, 16 June 2018 (UTC)

@Rtc: "some recent research has called into question" is weasel-wording and downplayed to the extent it is wrong. The settled scientific consensus is that the so-called placebo effect either does not exist or is of no clinical value at all; it does however complicate reporting of various subjectively measured phenomena (e.g. pain) in a way which is clinically measurable. Alexbrn (talk) 16:14, 16 June 2018 (UTC)

Please provide reliable sources for your claims about the allegedly settled alleged scientific consensus. --rtc (talk) 16:20, 16 June 2018 (UTC)

See the Cochrane review PMID 20091554. This is a top-quality WP:MEDRS and so far as I know has not been challenged (except in various quack circles). Alexbrn (talk) 16:41, 16 June 2018 (UTC)

I have read that source (the full text of the monstrous full version can be found at [1]); in fact I pointed to it myself above stressing that it is much more recent than the 1997 source and actually discusses the core of the problem rather than merely reevaluating sources that are over 60 year old. This more recent source does not say "no clinical value at all", it says "in general [whatever that means] did not have clinically important [whatever that means] effects". It does not say "complicate reporting" either, it says "there were possible beneficial effects on patient-reported outcomes, especially pain" with a reservation to the effect that "it is difficult to distinguish patient-reported effects of placebo from biased reporting". The study lists the quality of the evidence as moderate at best. Further, the study contains the following important reservation: "It is a question of definition whether our review evaluates the 'placebo effect'. This term does not only imply the effect of a placebo intervention as compared with a no-treatment group, but is also used to describe various other aspects of the patient-provider interaction, such as psychologically-mediated effects in general, the effect of the patient-provider interaction, the effect of suggestion, the effect of expectancies, and the effect of patients' experience of meaning (Hróbjartsson 2002b). As patients in the notreatment group also interact with treatment providers, a no-treatment group is only untreated in the sense that they do not receive a placebo intervention (Hróbjartsson 1996). Our result is therefore neutral to many of the meanings of the term ’placebo effect’ cited above, and we do not exclude the possibility of important effects of other aspects of the patient-provider interaction" Fair enough, the author's ultimate judgement is "placebo prescription seems to lack both ethical and empirical justification", however what they really mean they stress just before making that statement: it holds within the bounds of their own ethical framework only, which they describe as "a clinical placebo intervention is ethically acceptable only if it fulfils two criteria. First, patients must be informed about the nature of the intervention. Second, the effect of placebo must be reliably demonstrated in trials that disclose to patients that they receive placebo." Thus they restrict practical treatment proposals and empirical evidence to what they consider ethical. It is obvious that such a position (or opinion, as the authors themselves call it), which is ultimately rooted in ethics, not in science, cannot be qualified at all as a "settled scientific consensus". In the homeopathy article, I gave management of moderate pain as an example (a fairly common condition). The authors seem not to have a truly convincing, purely empirical argument against a placebo effect in that case. They say they think it might be patient reporting error but that's really not much more than speculation. The fact is, if you tell the patient that the medication helps with the pain (which is ultimately not a lie, as it is kind of a self-fulfilling prophecy) this study does not deny that you may actually have a benefit, even if the authors consider it somewhat "unethical" (I guess they think it's more ethical to recommend iboprufen or even prescribe opioids) --rtc (talk) 18:40, 16 June 2018 (UTC)

The source doesn't mention homeopathy, so this would appear to be entirely your flight of fancy. We do know from various other sources that quacks latch on to the supposed "placebo effect" to further their frauds. This aspect could usefully be expanded in this article. Alexbrn (talk) 18:43, 16 June 2018 (UTC)

I didn't refer to homeopathy at all except by pointing to the discussion in that article. There I said objectively homeophatic medicine is way cheaper than conventional placebo. A conventional placebo costs almost twice as much. Many doctors are thus prescribing homeophathy, as a placebo. Howver, I can only speak for the situation in Germany. I do not know about other parts of the world. Plus, if you prescribe a conventional placebo and the patient figures out he tends to get pretty angry, you don't have those issues with homeophathy. --rtc (talk) 18:46, 16 June 2018 (UTC)

This is just utter nonsense (trolling?). So long as we point out that placebo prescription lack both ethical and empirical justification, without adding daft editorial interpretations, we shall be fine. You should probably look at WP:Lunatic charlatans. Alexbrn (talk) 18:51, 16 June 2018 (UTC)

"we shall be fine" No, we shall not be fine, as that statement clearly holds, as I said, within the ethical framework of the authors only. The authors themselves say that very clearly. I see we're wielding weapons again to ge the "lunatic charlatan" blocked. It's a pity. Happy blocking. PS: What is "utter nonsense" about what I wrote? Is anything wrong with it? Please use argument instead of thought-terminating cliché --rtc (talk) 18:58, 16 June 2018 (UTC)

Homeopathic medicine is an oxymoron. And no, homeopathy is sugar pills plus magic and marketing, so taking out the magic and marketing and leaving the conventional placebo is going to be cheaper. Plus, the doctor giving you the placebo probably won't lie about how the medical establishment is a giant conspiracy to suppress this wonderful medicine and next time you should go straight to the placebo salesman and eschew the medical profession, leaving you to, oh, I don't know, die of untreated cancer in horrible agony.

You claim that increasing numbers of doctors are prescribing homeopathy. That's what homeopaths claim, but there's no evidence for it, and it would be a fallacious appeal to popularity anyway. Most doctors don't prescribe homeopathy. Under 0.2% of British doctors did, before a recent review that advised them to stop entirely, homeopathic prescriptions have been in steep decline int he UK for decades, and Prince Charles' favourite quackademic medicine establishment, formerly the Royal London Homeopathic Hospital, no longer provides homeopathy at all. Regardless iof the fate of the numerous other pseudomedical placebos (acupuncture, for example), homeopathy is recognised by pretty much every scientific authority as complete bollocks and is in terminal decline in many countries. Reviews by the Swiss Federal Office of Public Health, the British Parliament, the Australian Healthcare Commission and the US Federal Trade Commission - all the government level reviews in recent years, in fact - have found no evidence that homeopathy works. Guy (Help!) 22:09, 16 June 2018 (UTC)

No, I have never claimed anyhting about any numbers being increasing. You claim "have found no evidence that homeopathy works". Source please for that statement -- does it really say homeopathy doesn't work or does it actually say it doesn't work better than a placebo? (though for you it may be the same since you believe a placebeo doesn't work either. yet I want to know what the source actually says, not what you believe) --rtc (talk) 22:26, 16 June 2018 (UTC)

"if you prescribe a conventional placebo and the patient figures out he tends to get pretty angry, you don't have those issues with homeophathy" In other words, homeopathy makes it easier to lie to patients. This is a good thing only for unethical paternalistic doctors and charlatans. --Hob Gadling (talk) 11:55, 17 June 2018 (UTC)

Your statement is rooted in controversial ethics, not in science. --rtc (talk) 13:01, 17 June 2018 (UTC)

Actually, those were two statements. The first was basic logic. As for the second - those who think dishonesty is a good thing have a bad standing among scientists and among Wikipedia editors. --Hob Gadling (talk) 19:15, 17 June 2018 (UTC)

Why is the cherry-picked statement "no evidence [...] of any placebo effect in any of the studies cited" misleading? It is misleading because it may be read as saying that the authors reevaluated the data but found no effect at all that could possibly be explained as a placebo effect. But they are not saying that. They are saying that effects are there, but that those effects "can be fully, plausibly, and easily explained" with explanations other than the placebo effect. So they do not at all rule out placebo effects, they merely say that it not the only, let alone the simplest explanation for the data. --rtc (talk) 19:36, 16 June 2018 (UTC)

Paging Mr. Ockham. Brunton (talk) 19:48, 16 June 2018 (UTC)

Yes, but as I said this paper is over 20 years old and it evaluates the evidence from 60 years ago so it is pretty nonsensical to apply heuristics like Ockham's razor to this ancience state of evidence instead of looking at the full evidence as known today. There are more recent papers, in particular the one discussed just above, which says, yes, "placebo prescription seems to lack both ethical and empirical justification", but with the reservation that "there were possible beneficial effects on patient-reported outcomes, especially pain"; the authors simply do not consider such empirical evidence as collected in an ethical way or such use of placebos as compliant with their ethical opinion, and they stress that those effects may be fully explained by reporting error. --rtc (talk) 19:52, 16 June 2018 (UTC)

Where are you going with this? What changes are you proposing? Brunton (talk) 20:11, 16 June 2018 (UTC)

Stop acting stupid; the changes I made that were reverted, of course. (I re-added them but obviously they will soon be reverted again, that's for sure) --rtc (talk) 20:14, 16 June 2018 (UTC)

Note that Placebo#Criticism brings up the same topic again, this time without even mentioning the 1997 methodological criticism. Also it notes that the study which in updated form resulted in the 2010 source discussed above "received a flurry of criticism". @Alexbrn: claimed above that the 2010 source "has not been challenged". This is a half-truth. As can be seen in [2], the original version of that study "occasioned a blizzard of criticism (13–26) and some support (27). It’s in the papers (28, 29)." The journal that published the original version published 9 dissenting opinions in the next volume. The more recent versions of the study seem not contain any reply to the critics at all. The introduction does not mention those facts and those facts alone refutes the claim that the opinion of those authors can be considered the "settled scientific consensus", especially given the fact that they employ ethical arguments rather than purely empirical ones. I don't see how this can be explained as the opinion of "various quack circles". This seems to be a highly controverstial issue rather than a settled one. --rtc (talk) 20:32, 16 June 2018 (UTC)

• The relevance of the 1997 study is clear from the 1955 study which has been in the lede for a very long time, and is almost certainly the origin of the pervasive placebo myth. If you read a study, and then go back and re-analyse the sources it purported to review, and you find no evidence at all to support the main conclusion of the study, which is what happened here, that's rather significant. Beecher has been cited over 2,000 times and is still being cited now even though his conclusions are definitively refuted by Kienle and Kiene. That's not unusual, by the way: there are articles that have been retracted and are still being cited. What is rather clear now is that most studies purporting to show that the placebo effect is real, are written by "integrative medicine" advocates, i.e. by people whose living depends on the idea that their favoured woo can maguically harness the placebo effect, as a backstop position when (inevitably) they run out of any actual evidence of any mechanism of effect. Beecher is the Andrew Wakefield of the placebo myth, and an authoritative refutation of Beecher is clearly significant in that context. Guy (Help!) 21:59, 16 June 2018 (UTC)
  • What the 1997 study means by "no evidence" is that nothing points towards a placebo effect as the only possible explanation. The other statements in that study which were reverted make that perfectly clear. The 1997 study is a purely a methodological one. Just because it debunks faulty reasoning that has been used as an argument for the placebo effect does not mean it debunks the placebo effect itself. It is completely uncontroversial that Beecher's work is faulty, however, that does not mean it is consensus that the placebo effect does not exist or is a myth. This is a highly controversial claim ("blizzard of criticism"). --rtc (talk) 22:15, 16 June 2018 (UTC)

Here is the paragraph, word for word:

Recently Beecher's article was reanalyzed with surprising results: In contrast to his claim, no evidence was found of any placebo effect in any of the studies cited by him. There were many other factors that could account for the reported improvements in patients in these trials, but most likely there was no placebo effect whatsoever.

How the fuck you spin that as "nothing points towards a placebo effect as the only possible explanation" is a mystery to me. It is absolutely unambiguous. You don't say "the original author's conclusions are a crock of shit" in a scientific paper, but this is about as close as you get. Guy (Help!) 22:28, 16 June 2018 (UTC)

Yes, it is indeed unambigous. Note the statement "There were many other factors that could account for the reported improvements". This is exactly what I said: The authors do not claim a placebo effect necessarily does not exist, they merely say a lot of other (more plausible) factors could account for everything at least as well. At least in the evidence used by the study, which is over 60 years old. The cherry-picked quote, taken out of its context, makes it seem as if there were no "reported improvements in patients in these trials", ie., no evidence at all, so a placebo effect could not possibly exist. But that's precisely not what the authors say. Note they say "most likely". They prefer the plausible factors as explanation over the placebo effect hypothesis. They don't say the assumption of a placebo effect is fundamentally ruled out, they just don't think that this explanation is the most likely one. --rtc (talk) 22:32, 16 June 2018 (UTC)

Which comes between In contrast to his claim, no evidence was found of any placebo effect in any of the studies cited by him and most likely there was no placebo effect whatsoever. I don't think anyone else here is going to interpret this as anything other than unambiguous refutation. Guy (Help!) 00:13, 17 June 2018 (UTC)

Well of course if you simply ignore the statement that contradicts what you say then everything look as if it supported what you say. See, the great thing is we don't need to interpret based on those statements. They are merely the abstract. We can have a look at the main text of the paper. --rtc (talk) 00:17, 17 June 2018 (UTC)

Kaptchuk

It's time to start reducing this article's reliance on Kaptchuk. He's an acupuncture shill engaged in policy-based evidence making, his studies are uniformly shoddy and the data never live up to the spin he puts out. If you can find a Kaptchuk study that's based on robust objectively measured data instead of subjective outcomes and/or self-reported questionnaires I'd be astonished. He is single-handedly trying to keep the placebo myth alive in the face of increasing evidence that it's baloney, and every time you hear some new hype around the magic of placebo somehow it always seems to be Kaptchuk behind it. Guy (Help!) 21:35, 16 June 2018 (UTC)

Kaptchuk is certainly a relevant and influential author on the subject and his views should thus be mentioned. If his publications are controversial, it should be easy for you to find publications that criticize him. Those can be cited to say his views are under attack, and his views can be attributed to him, thus presenting them not as the uncontroversial truth, but as Kaptchuk's opinion. NPOV permits presenting multiple views, it does not require The One True Scientific Viewpoint to be presented to the reader. -rtc (talk) 00:11, 17 June 2018 (UTC)

Well, he certainly publishes a lot, but consider:

Although albuterol, but not the two placebo interventions, improved FEV1 in these patients with asthma, albuterol provided no incremental benefit with respect to the self-reported outcomes. Placebo effects can be clinically meaningful and can rival the effects of active medication in patients with asthma. (https://www.nejm.org/doi/full/10.1056/nejmoa1103319)

What he actually found was no statistically significant difference in objective outcomes between placebo, sham acupuncture and no treatment, but a substantial objective difference for albuterol. He is then spinning this as a validation of his sham treatment, based on prioritising the self-reported subjective outcomes over the objective ones.

As an asthmatic, I think I prefer an inhaler that objectively improves lung function over one that does nothing but makes me feel as if I have better lung function. Because, you know, asthma can actually kill you, and the objective, not the subjective, outcome is what makes the difference between being objectively dead or not. And this is not a lone example. He has a vested interest in "proving" that a placebo effect exists and that acupuncture, specifically, harnesses it in a uniquely powerful way. Every paper he publishes has positive findings based on self-reported and subjective outcomes. When there are objective measurements, they show no effect, so he waves them away.

It's hard to see that as anything other than research misconduct. Guy (Help!) 00:22, 17 June 2018 (UTC)

You have to option of raising the issue of research misconduct with his employing unversity, or alternatively publish a criticism of his claims. I don't think wikipedia is the right place for making a judgement. If the reader is informed that the existence of placebo effects is controversial, then he will automatically read Kaptchuk in a critical way. If there's a publication that accuses Kaptchuk of misconduct it can be cited as criticism of his work. Kaptchuk is cited so often, it would simply not be due weight to make it seem as if he didn't even exist. --rtc (talk) 00:28, 17 June 2018 (UTC)

Or we could start applying WP:MEDRS and reducing the references to primary research like Kaptchuk's. Guy (Help!) 07:23, 17 June 2018 (UTC)

RfC on

Should the fourth paragraph in the lede of Placebo:--

• Option 1-As written in the current version, quote from the paper's abstract that:--

  • but a 1997 review of the study found "no evidence [...] of any placebo effect in any of the studies cited".

or

• Option 2-As written in this version, quote from the main text and the conclusion of the full-text which says

  • but a 1997 review of the study found a wide range of "conceptual and methodological mistakes" in the study as well as "a total of 800 articles on placebo". It noted that the reported outcomes could be "fully ... explained without presuming any therapeutic placebo effect" and concluded that "the existence of therapeutic effects of placebo administration seems questionable".

• And in general:-
  • Should the article present it as the settled scientific consensus that the placebo effect is more or less a myth?

• Initiated by --rtc (talk) 22:08, 16 June 2018 (UTC)Heavily refactored by ∯WBG^converse.

!voting

• This rfc is malformed and not neutrally stated. That said, of course it should reference Beecher, as the father of the placebo myth, along with Kienle and Kiene's refutation, which is unambiguous, and yes it should reference the fact that the scientific consensus is that the "placebo effect" is in fact a combination of a number of biases and confounders such as regression to the mean, natural course of disease, expectation effects and so on, and also that objective measurements repeatably fail to show any clinically significant effect size. Asthma studies, for example, show great improvement in patient-reported symptoms but not objective measurements. One of Kaptchuk's papers claiming marvellous effects of sham acupuncture on asthma has, as figure 3, a set of bar charts that unambiguously show identical zero effect on maximum forced respiratory volume from the two placebos and no treatment, versus a substantial effect from albuterol. Guy (Help!) 22:24, 16 June 2018 (UTC)
  • Feel free rephrase the rfc such as to make it more compliant with your views on neutrality and well-formedness. --rtc (talk) 22:27, 16 June 2018 (UTC)

Why would I? I don't see anything needing changing. Guy (Help!) 00:10, 17 June 2018 (UTC)

Well you said it was "malformed and not neutrally stated" and to me that seemed to be implying it could be improved. --rtc (talk) 00:12, 17 June 2018 (UTC)

It could best be improved by not existing. Guy (Help!) 07:22, 17 June 2018 (UTC)

Right, that's what you wanted to say. I got that. And me being blocked or topic banned, because of troublemaking and being a timesink. --rtc (talk) 10:12, 17 June 2018 (UTC)

Per revised question, option 1 is straight from the abstract so preferable to option 2, which quote mines the report to give wiggle room to something the 1997 paper clearly rejects. Guy (Help!) 07:24, 17 June 2018 (UTC)

As shown above ("simply ignore the statement") it is in fact option 1 that quote mines to make a biased statement. Even the abstract clearly contradicts that cherry-picked statement directly by qualifying it by mentioning the important fact that there actually, uncontroversially, were "reported improvements in patients in these trials", though according to the authors of the study not ones that they most likely attributed to placebo effects. --rtc (talk) 10:15, 17 June 2018 (UTC)

• Malformed RfC (especially asking for a vague "in general" ruling). WP:LEDEs should summarize the body, so cramming in a load of quotations from two papers, of 1955 and 1997, which is not also discussed in the body is quite wrong. It is also wrong to boil a Cochrane review down to a statement that it "concluded" just one thing, when that paper's conclusions are in fact extensive and more far reaching that the proposed text suggests. I propose getting this article's body in shape then having the lede follow it, rather than using an RfC to try and have one flawed lede "win" over another flawed lede. In general, this articles sourcing also needs to be brought into line with WP:MEDRS and fringey claims of placebo's power toned down to match current accepted reality.Alexbrn (talk) 04:21, 17 June 2018 (UTC)

Discussion

• I'm with Alexbrn. Guy (Help!) 07:20, 17 June 2018 (UTC)
• I can only condemn the rush to create facts to present placebos as one more fringe science thing like homeopathy, before this RfC has actually seen some comments. The article is more and more getting into a state where it is completely and totally hopelessly biased. Everything that even slightly contradicts the opinion that the placebo effect is fringe science is systematically being eliminated with completely outrageous claims (just take an example this edit which in its comment claims "A flurry of criticism from Wayne Jonas, tireless defender of hoemopathy, the best-known placebo" while in fact the cited paper merely lists the dozens or so references that give dissenting opinions and criticize the study, mostly published in the next volume of the exact same journal that originally published the study). --rtc (talk) 09:35, 17 June 2018 (UTC)
  • Something I'm seeing in the sources is that just invoking "the placebo effect" as a kind of monolith is fringey, as there is no such single phenomenon - only localized placebo effects in some circumstances. Getting the endemic fringeiness out of this article is necessary work. The means is to use decent WP:MEDRS and cut out poorly-sourced material. Alexbrn (talk) 09:41, 17 June 2018 (UTC)
    • Fair enough, but the way the article was changed goes way beyond "there is no such single phenomenon". It goes into the direction of claiming that there cannot possibly be placebo effects, anywhere. And it is systematically downplaying the controversy by presenting the ethically biased claims of one single study as the accepted scientific consensus while not mentioning at all the many dissenting opinions that criticize that study. --rtc (talk) 09:50, 17 June 2018 (UTC)
      • Incorrect. We explicitly define what the response and and effect are. In clinical trials these are universally recognized phenomena. What is problematic is all kinds of of non-WP:MEDRS being used to imply that placebos are some kind of effective treatment. There is also a deafening silence about how CAM advocates have appropriated placebos as a kind of marketing ploy. All this needs to be fixed. Alexbrn (talk) 09:55, 17 June 2018 (UTC)
        • I agree. Novella makes this case well. Guy (Help!) 10:05, 17 June 2018 (UTC)
          • Look at the intro. It says "no worthwhile effect on clinical outcomes in general". This is wrong: The study says no important effect (whatever that highly subjective judgement means). It doesn't contain the word "worthwhile". That is one significant example of how the sources are twisted to make them fit the bias. Also, it is not mentioned that in special cases significant effects could not be ruled out. It is not mentioned either that the study has been criticized a lot and is controversial. It is further not mentioned that the study partly relies on ethics to make its point. --rtc (talk) 10:08, 17 June 2018 (UTC)
            • "Worthwhile" is a fine paraphrase of "important". We are supposed to use our own words. Alexbrn (talk) 10:28, 17 June 2018 (UTC)
              • It is certainly not fine to extend a statement more about quantity than practical value into one that is more about practical value than quantity. Those are subtle points and such a central highly nuanced judgement should never be paraphrase, but given as is. --rtc (talk) 10:32, 17 June 2018 (UTC)
                • Novella is a research scientist, a clinical neurologist. He has written and spoken extensively on what "clinically important" and "clinically relevant" mean, so we don't need to guess. He presents them as scientific terms of art that translate, in lay language, to "worthwhile". An effect that is not clinically relevant means the treatment is not worthwhile. Actually he goes further: if it has no clinically important effect, it doesn't work. We're fine with ditching real medical treatments on this basis, but purveyors of pseudomedicine have more of an issue with it. Guy (Help!) 10:40, 17 June 2018 (UTC)
                  • So you mean you can use the opinion of one research scientist as a justification for biasing the article and twisting the judgements of the papers cited? --rtc (talk) 10:43, 17 June 2018 (UTC)
                    • It's not twisting. Since you admitted above you didn't understand what the word meant I'm not sure how you can now argue about its meaning? Alexbrn (talk) 10:54, 17 June 2018 (UTC)
                      • I don't see any point in discussing this. In doubt, the wording of the source should be used, not wording based on linguistic theories of evidence-based research scientist bloggers. --rtc (talk) 10:59, 17 June 2018 (UTC)
                        • Do you actually understand the meaning of the term "clinically important"? As in "We did not find that placebo interventions have important clinical effects in general"? It means that the treatment has no demonstrable effect on clinical outcomes. It is clinically worthless - you might as well try tea and sympathy. Guy (Help!) 12:22, 17 June 2018 (UTC)
                          • What I understand is that "important" is a different word from "worthwhile" and has a lot of different nuances, and I understand as well that the word "important" is in the source while the word "worthwhile" is not. This discussion is not worthwhile. The core judgement of the source should be given as is, not in the way a philosophically biased wikipedia editor would have made it. --rtc (talk) 12:44, 17 June 2018 (UTC)
      • Is that how you perceive it? Odd, since nobody has actually said that. The question is not whether a "placebo effect" could exist, but whether there is good evidence that it does exist, as a concept separate from confounders in clinical trials. The interesting part for me is the timeline. Beecher, from the heyday of eminence-based medicine, established a mythos, and as it began to be challenged Kienle & Kiene went back and looked at his data and found that he had invented the entire thing from whole cloth, that there was no evidence of any "placebo effect" - and now we are into a situation with many parallels in the world of quackademic medicine where True Believers are churning out crappy studies with n=not many and the wider scientific population is largely ignoring them because they mainly view the scientific question as answered: observed effects in placebo arms of trials are due to biases and confounders and are visible only in subjective endpoints, objective measures fail to support the hypothesis of a placebo effect. I personally find that very interesting as a story of how science works, and how pseudoscience works, and the comparisons and differences between them. When you get studies saying that because there is no difference between sham acupuncture and real acupuncture, thus acupuncture uniquely harnesses the placebo effect and this proves the power of acupuncture, you know you have cranks at work, and comparing their reasoning with people like Gøtzsche, whose focus is not on "prove hypothesis X" but ""test hypothesis X", is fascinating. Guy (Help!) 10:02, 17 June 2018 (UTC)
        • "The question is not whether a "placebo effect" could exist, but whether there is good evidence that it does exist" Actually neither the one nor the other. The question is whether and which placebo treatments do have effects (above no treatment). Your statement presupposes the philosophical framework of "evidence-based medicine", which is really logical positivism applied to medicine in disguise and thus a fundamentally highly biased opinion. --rtc (talk) 10:08, 17 June 2018 (UTC)
          • There is no doubt in RS that such effects exist (as they can be measured), but they are effects of measurement resulting from adjustments to subjective assessment ("I think I feel better because of taking the pill"). There is no disease "healing" effect. Alexbrn (talk) 10:27, 17 June 2018 (UTC)
            • That is your opinion. The study does not contain a final judgement on whether those effects are a reporting error or real. A neutral intro should thus state that those effects are observed and that there are two significant positions on how they can be explained (1. reporting error, 2. real effect). --rtc (talk) 10:29, 17 June 2018 (UTC)
              • Two positions? No, because that would be a textbook WP:PROFRINGE trumpeting of the WP:GEVAL fallacy. Alexbrn (talk) 10:33, 17 June 2018 (UTC)
                • Right, that's always the excuse to violate NPOV. The study says both are possibilities and they thus have to be presented as such, and it is not okay based on philosophical presuppositions of the wikipedia users to reject the one as fringe and the other as scientific and thus not mention the allegedly fringe one. --rtc (talk) 10:39, 17 June 2018 (UTC)
              • Have you ever heard of the fallacy of reversed burden of proof? That's what you're doing here. A "placebo effect" is an extraordinary claim. Many mundane explanations exist and are covered in the literature. It is now up to those who advocate a separate and distinct "placebo effect" to prove their case. Which they are trying to do with tiny studies using self-reported subjective endpoints. For some reason they seem surprised that this is not working. Guy (Help!) 10:36, 17 June 2018 (UTC)
                • The fact is that the huge 2010 study says in some areas placebo effects cannot be ruled out. --rtc (talk) 10:39, 17 June 2018 (UTC)
                  • To be precise, they say placebos have no effect except for patient-reported outcomes. The idea that the "effect" is actually a "healing" effect (e.g. shrinking tumours) is on the extreme fringe. Alexbrn (talk) 10:51, 17 June 2018 (UTC)
                    • Wrong, they conclude that "in certain settings placebo interventions can influence patient-reported outcomes, especially pain and nausea, though it is difficult to distinguish patient-reported effects of placebo from biased reporting" They're explicitly admitting that both are possibilities, though with the reservation that "The effect on pain varied, even among trials with low risk of bias, from negligible to clinically important" which suggests their personal opinion is that those are biased reporting rather than the other possibility. --rtc (talk) 10:57, 17 June 2018 (UTC)
                      • Ah yes, the old "you haven't definitively proved a negative" ploy. Guy (Help!) 11:00, 17 June 2018 (UTC)
                        • Rather the old "you are twisting the source until it fits your presupposition" diagnosis. --rtc (talk)
                      • @Rtc: Again, I don't think their words mean what you think they mean. I agree we should reflect the Cochrane review's claim that placebo is either an influence on patient-reported outcomes, or maybe reporting bias. The paper nowhere states or implies anything about a real "healing" effect. Alexbrn (talk) 11:00, 17 June 2018 (UTC)
                        • You're again twisting the source. it says the placebo can "influence patient-reported outcomes". That is the fact. The possibilities are now either those are 1) "patient-reported effects" (my emphasis) or 2) merely "biased reporting". --rtc (talk) 11:04, 17 June 2018 (UTC)
                          • The source says: We did not find that placebo interventions have important clinical effects in general. However, in certain settings placebo interventions can influence patient-reported outcomes, especially pain and nausea, though it is difficult to distinguish patient-reported effects of placebo from biased reporting. So any assertion of clinically meaningful "placebo effect" is clearly into "god of the gaps" territory, based on this review. You do seem to have a tendency to misidentify direct quotations from sources as misrepresentation, and your own creative interpretation of selected text as the opposite. That's a bit of an issue. Guy (Help!) 11:06, 17 June 2018 (UTC)
                            • Sorry, but nowhere does the study talk about "god of the gaps". You're simply making this up. What I say is not "creative interpretation", but direct quotation, which is completely identical to what you quote (and then interpret creatively) --rtc (talk) 11:09, 17 June 2018 (UTC)
                              • I think you're having a comprehension problem here. Of course giving people with headaches pills (maybe labelled "pain-away") can "influence patient-reported outcomes". That is the essence of the placebo effect. Cochrane says so. The only person saying that in the process disease was treated, appears to be you. Alexbrn (talk) 11:16, 17 June 2018 (UTC)
                                • Certainly I am mentally retarded in understanding texts compared to your PhD in English. Yet, my basic understanding is that Cochrane notes the possibility that those observations may explained by "patient-reported effects" (direct quotation, my emphasis) as an alternative to mere reporting error. --rtc (talk) 11:30, 17 June 2018 (UTC)
                                  • I don't see your point. Of course they are patient reported effects ("The effect of taking the pill was my headache got a bit better"). What else could they be? I repeat the only person talking about actual healing is you - and it's not a view supported in any decent source so far as I can see. Alexbrn (talk) 11:33, 17 June 2018 (UTC)
                                    • From basic logic, either there is an actual effect or not. If the patient reports an effect, but it is actually not there, then this is, by definition, a reporting error. This is the second alternative mentioned by Cochrane. Now Cochrane's first alternative is "patient-reported effects" (my emphasis). If we make an assumption of sanity (that Cochrane doesn't intend to give two identical possibilities as hard-to-distinguish alternatives), we have to conclude that he means real effects correctly reported by the patient as such. --rtc (talk) 11:37, 17 June 2018 (UTC)
                                      • This is just wrong. "Biased reporting" refers to things like falsified/skewing results from the researchers. For a trial where the patient is asked "On a scale of 1 to 10, how bad is your pain?" the response cannot suffer from a disparity with reality because the response is in itself entirely subjective and ipso facto correct. Alexbrn (talk) 11:45, 17 June 2018 (UTC)

                                        And that means that it is based not just on the patient's actual pain but also on the patient's expectation of what the pain should be after the treatment and the patient's expectation of what the doctor wants to hear. --Hob Gadling (talk) 11:59, 17 June 2018 (UTC)

                                        • • Right. Placebo is plausibly effective for diseases of the feels, it's objective effects that are in short supply. Guy (Help!) 12:05, 17 June 2018 (UTC)
                                            • So what you are trying to say is that diseases of the feels are illusions, as with all this nonsense about consciousness, not objective states of affairs but subjective chimera that objectively are nothing but the brain causing the patient reporting a metaphysical and thus nonsensical thing -- a feeling -- that does not actually exist. --rtc (talk) 13:16, 17 June 2018 (UTC)
                                        • "refers to things like falsified/skewing results from the researchers" No. It refers to the patient reporting in a biased way. It's "patient-reported", not researcher reported things that passage is talking about. "the response cannot suffer from a disparity with reality" That's blatant positivist nonsense. The patient has a feeling of pain. If the patient has no pain but says it's 10 that answer clearly suffers from a disparity with reality. If you kick the patient harshly and he reports 0 this answer apparently most likely also suffers from a disparity with reality. The response is by no means ipso facto correct. --rtc (talk) 12:39, 17 June 2018 (UTC)
                                          • You need to read, and understand, the paper. WP:CIR. There is really nothing more to say. Alexbrn (talk) 12:45, 17 June 2018 (UTC); amended 13:39, 17 June 2018 (UTC)
                                            • Now we're starting again with arguments from authority. You are competent with your PhD in English, I am not, so you must be right, despite objectively you suffer a huge disparity with the reality of the sources. What comes next now? Closing this discussion? Blocking or topic banning me because of troublemaking and being a timesink, and for sealioning? --rtc (talk) 12:49, 17 June 2018 (UTC)

Actually, I beg your pardon as the paper does gloss this particular mention of bias as "response bias" ("as polite patients may tend to report what they think socially most acceptable"). However, this does not bear on the question of "actual healing" as this merely means that the responses may be biased to over-state the effect the subject feels: e.g. they may say their headache feels better even when they don't really think it does! Alexbrn (talk) 13:39, 17 June 2018 (UTC)

Sigh. Yes. The authors say they don't know. It might be response bias. It might be, quote, "a true effect". The authors claim it's difficult to distinguish. I am not so sure. Brain scanners have been used to study the use of placebo for pain management, to distinguish real pain from reported one, if I remember correctly. The original version of the study was harshly criticized by a number of opposing letters in the next volume of the journal. This discussion is really too tedious to me. Have a nice day. --rtc (talk) 13:46, 17 June 2018 (UTC)

LOL - "Honest Doc, it doesn't hurt". "No, you lie! the machine says it must be hurting!". At this point we're into the astrology-like world of reading brain scans. But crucially, if we're to include material in the article we need WP:MEDRS to support it. Alexbrn (talk) 13:57, 17 June 2018 (UTC)

Just because computers have been used to compute horoscopes does not make computers artifacts astrology. [3] has some information ("Expectations and anticipation of pain are also known to be major contributors to placebo analgesia...") And I stand by the fact that if this one 2010 study is being given such an extreme weight, then so should be the criticism (which was raised with respect to the older journal version but nevertheless has not become any less relevant). --rtc (talk) 14:42, 17 June 2018 (UTC)

It's a fundamental tenet of WP:MEDRS that weak sources (like letters) are not used to undercut strong sources (like a multiply-updated Cochrane systematic review). Alexbrn (talk) 15:17, 17 June 2018 (UTC)

It has indeed become fashionable to believe that "strong" scientific authority must be given undue weight and "weak" criticism must be kept out of the articles. I disagree with that distortion of the NPOV policy. --rtc (talk) 15:36, 17 June 2018 (UTC)

This is not the right venue to discuss changing WP:MEDRS. If you want to argue that letters may be used to undercut systematic reviews, argue that at Wikipedia talk:Identifying reliable sources (medicine). Alexbrn (talk) 15:41, 17 June 2018 (UTC)

Your use of the word "undercut" clearly shows your ideological bias. You know as well as me that the tight-knit gang of evidence-based enthusiasts defends this distortion of the NPOV to its teeth and it is currently basically impossible to change it. The gang is too strong and too established and has shooed out every critic so there's no chance whatsoever to undercut its strong and powerful but illegitimate authority. --rtc (talk) 15:48, 17 June 2018 (UTC)

That's consensus you're dissing, whose mighty tide keeps our content in good shape - and our medical content is probably Wikipedia's most highly-regarded. With more work this article can become part of that quality corpus. I think this conversation is now at an end. Alexbrn (talk) 15:58, 17 June 2018 (UTC)

consensus is just an euphemism for group think. Consensus is a bad thing. And the content is in very bad shape, too; it is clearly biased towards positivist ideology and this article is just about to become the next in a long series of articles mutilated by the gang such as to be in line with its naive, defective philosophy, which is "So confident. So clueless". What is still lacking is the application of the gang's tenet that "The pseudoscientific view should be clearly described as such", so I guess the statement "the placebo effect is pseuodscience" quickly needs to be added. --rtc (talk) 16:08, 17 June 2018 (UTC)

Clearly you are opposed to the aims of The Project. As for WP:PSCI it seems you get it! But no, placebos are not pseudoscience -- but there are some invocations of them in the altmed universe which are. We shall get there in time once the more fundamental aspects of this topic are sorted, I should think. aybe start thinking about sourcing starting here. Alexbrn (talk) 16:22, 17 June 2018 (UTC)

Clearly I am not opposed to the aims of the project, which is to build a neutral, freely licensed encyclopedia, not to build an instrument of propaganda for positivist philosophy. --rtc (talk) 16:24, 17 June 2018 (UTC)

You use the word positivism (aka falsificationism or empiric-analytic model) as a way of attacking science. An article about a scientific topic should be science based. Tgeorgescu (talk) 16:37, 17 June 2018 (UTC)

No I am attacking positivism to attack positivism. I know positivists believe their philosophy actually is identical to science, just a different word for it, and keep spreading the legend that it is identical to falsificationism. An article about a scientific topic should be as neutral as any other and should not give the alleged scientific point of view any undue weight. --rtc (talk) 16:47, 17 June 2018 (UTC)

Actually you are attacking editors, not positivism. You apply your idiosyncratic diagnosis of "positivism" in order to frame the discussion as one of evil positivists versus you as the force of righteousness. That's not working well for you right now. Guy (Help!) 20:57, 17 June 2018 (UTC)

Recent changes

These represent a huge improvement, especially the reorganisation to reduce back and forth with one section implying one thing and another something else, from the same studies.

Are there any review studies on fMRI and placebo? I'm concerned that fMRI is rthe medical equivalent of pareidolia, you can show pretty much anything including life after death in fish, so primary fMRI studies seem particularly dangerous per WP:MEDRS. Guy (Help!) 12:32, 17 June 2018 (UTC)

Giving up

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

---

Fighting this well-known, tight-knit gang of evidence-based enthusiasts without support from others is really consuming too much energy and time. Humanity is served in a better way if I put my energy into other, more productive things. I hope someone finds some time to ensure this article does not become another totally biased statement of faith in positivist philosophy. --rtc (talk) 13:25, 17 June 2018 (UTC)

I want to give you a piece of advice, it is not an attack. If you will listen to it you will have a good time around here.

Wikipedia is mainly a venue for parroting mainstream science and mainstream scholarship (and perhaps mainstream press, for certain subjects). Editors are supposed to understand this, to wish this and be competent at doing this.

So supporting mainstream science and mainstream scholarship is required of all editors, failure to do this leads to losing disputes, being blocked and eventually banned. Strong adherence to mainstream science and mainstream scholarship is what made Wikipedia one of the greatest websites. So, dissent from mainstream science and mainstream scholarship will be perceived as an attack upon Wikipedia itself. If you want to win a dispute, you have to show that your claims are mainstream science or mainstream scholarship. If you cannot honestly do that, then refrain from making that particular claim. And remember, Wikipedia is just a mirror, mainstream science and mainstream scholarship exist outside of Wikipedia and cannot be changed through editing Wikipedia, Wikipedia merely reflects them. So if you want to change science/scholarship, you have to be a scientist or a scholar, Wikipedia is not the venue for doing that. Tgeorgescu (talk) 16:40, 17 June 2018 (UTC)

Thanks for stating the prevailing ideology in such clear terms. Honesty is certainly a first step. I actually have a good time around here since 2005 and I have never changed my opinion that Larry's philosophy is right and that the prevailing philosophy about "Wikipedia is mainly a venue for parroting mainstream science" is naive and wrong and I will keep defending the view that Wikipedia is neutral, is not at all "a mirror, mainstream science" and supporting mainstream science and mainstream scholarship" is not required for any editor and any use of blocks and bans to advance the science-savvy ideology (which has been used against me a few of times already) is an illegitimate abuse of power. But, don't get me wrong, science is important. It is just not the definition of neutrality. --rtc (talk) 16:55, 17 June 2018 (UTC)

"All opinions are equal" has never been part of WP:PAGs, see WP:ABIAS. Tgeorgescu (talk) 18:55, 17 June 2018 (UTC)

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Important

Rtc is very keen to use the word important. In context his preferred text is:

However, research has found placebo interventions result in no important effect on clinical outcomes in general, and only in certain settings (especially for pain and nausea) they can influence patient-reported outcomes, though it is difficult to distinguish true placebo effects from biased reporting by the patient.

The actual authors' conclusions from the first cited source:

We did not find that placebo interventions have important clinical effects in general. However, in certain settings placebo interventions can influence patient-reported outcomes, especially pain and nausea, though it is difficult to distinguish patient-reported effects of placebo from biased reporting. The effect on pain varied, even among trials with low risk of bias, from negligible to clinically important. Variations in the effect of placebo were partly explained by variations in how trials were conducted and how patients were informed.

And the second:

We found little evidence in general that placebos had powerful clinical effects. Although placebos had no significant effects on objective or binary outcomes, they had possible small benefits in studies with continuous subjective outcomes and for the treatment of pain. Outside the setting of clinical trials, there is no justification for the use of placebos.

I would say that Rtc's wording is a clear misrepresentation of the overall tone of these sources, noting that while Hróbjartsson and Gøtzsche do speak and write excellent English, their language is very formal, and they are not native speakers. Hence the somewhat clumsy "We did not find that placebo interventions have important clinical effects in general" rather than "We found that placebos have no clinically important effect in general", which is almost certainly what a native English speaker would have written.

I think the correct form of words, based on these sources would be:

Reviews find no evidence of significant effects on objective outcomes. There are possible small benefits on subjective outcomes (such as pain or nausea) but these are hard to distinguish from bias. There is no justification for the use of placebos outside of clinical trials.

That seems to me to be the closes to the sources. Both find no clinically relevant effect on objective measures, both play down the relevance of the nonspecific effects. What do others think? Guy (Help!) 21:09, 17 June 2018 (UTC)

I’m not sure that it’s to do with not having English as a first language; rather it’s simply the way scientists tend to write. One of the problems we have here is that there is a difference between writing for a peer-reviewed science journal and writing for a general encyclopaedia, so we have to translate “scientific” English aimed at a specialised audience into plain English aimed at the general public. And, of course, we also have to avoid close paraphrasing to avoid copyvio issues. This will always give people the opportunity to object that what we have written for the encyclopaedia isn’t exactly what the source says, ignoring the fact that we have to paraphrase. The paraphrase you suggest seems perfectly OK to me. Brunton (talk) 22:50, 17 June 2018 (UTC)

Yes and no. Their style is more formal than most English scientists, currently. Gøtzsche speaks excellent English but distinctly British-formal, whereas science writing now is dominated by American-informal. Guy (Help!) 22:54, 17 June 2018 (UTC)

• There is a distinction here that is important to scientists. To a scientist, "we did not find an effect" is very different from "we found that there is no effect". A common way of putting this is that absence of evidence is not evidence of absence. Looie496 (talk) 00:11, 18 June 2018 (UTC)

It is hard to prove a negative, but the burden of proof lies with those seeking to prove an effect. The fact that a substantial and careful analysis found none, is telling. That said, my proposed wording says exactly that. Guy (Help!) 04:45, 18 June 2018 (UTC)

Really? Then, how do you manage to "find there is no effect", except by trying to find one and failing? Both are just other wordings for measuring the effect and finding a value of zero, within error bars. --Hob Gadling (talk) 13:03, 18 June 2018 (UTC)

In the vast majority of cases (at least in medicine, biology, and psychology), failing to find an effect means conducting a statistical test and getting a p value larger than 0.05. Looie496 (talk) 13:57, 18 June 2018 (UTC)

Which is the same, in principle, as "measuring the effect and finding a value of zero, within error bars", as I put it.

You did not answer my question. So, how do you manage to "find there is no effect"? --Hob Gadling (talk) 15:55, 18 June 2018 (UTC)

• We should avoid significant per MEDMOS. We should avoid "important" to avoid WP:CLOP (and the fan of the word admitted[4] they did not understand what it meant anyway - so we should avoid it because it's not clear to some readers). Important here (in the context of clinical effect) means of no import, or not worth bothering with. Good synonyms would be "meaningful", "worthwhile" or "useful" or even "therapeutically useful" in this context I think. (So just WP:ASSERT "Placebos are not a useful means of therapy" to keep it plain and to paraphrase it, as we should). Alexbrn (talk) 04:52, 18 June 2018 (UTC)

OK, please add your preferred version. Guy (Help!) 05:27, 18 June 2018 (UTC)

I'm thinking now about how due any of this history is in the lede, especially since we have a Placebo in history article (itself not problem-free). Perhaps once the body is in better shape it should be more apparent what the structure of the lede should be? Alexbrn (talk) 07:47, 18 June 2018 (UTC)

Not sure. For me, the important thing is to establish the facts: placebo is an inert treatment, it is normally encountered in the context of trials, there was a historical belief that the "placebo effect" was a thing, that is no longer viewed as true. Guy (Help!) 08:14, 18 June 2018 (UTC)

Yes, from the sources those seem the salient points (together with some big ethical considerations). The position that actual disease healing happens seems to be a fringe position and is not supported by RS, so that must be clear. Alexbrn (talk) 08:16, 18 June 2018 (UTC)

Wait..what? the Placebo effect doesn't exist?^{[Citation Needed]} --Guy Macon (talk) 18:53, 18 June 2018 (UTC)