Wikipedia talk:Identifying reliable sources (medicine)/Archive 4

From Wikipedia, the free encyclopedia
Jump to: navigation, search
Archive 3 Archive 4 Archive 5

Contents

Use or overuse of UpToDate

A new editor added considerable material to Grave's disease. Leaving entirely aside any POV issue I have re neuropsychiatric symptoms, the rest of the added material is I think basically correct, although written in a style of a text book. My specific concern is that UpToDate has been used in large measure (eg Ref3 used 24 times and further multiple references to UpToDate too) for the added material. Yet UpToDate is a subscription only site, so I am unable to check the details, and "Diagnosis of hyperthyroidism, Douglas S Ross, UpToDate.com, Last literature review version 17.1: januari 2009 | This topic last updated: augustus 25, 2008" does not seem likely as a reference easily located.

Is UpToDate therefore a permitted source for what should be standard information. Is any deficiency as a source (no abtract per PubMed for published articles to at least vaguely WP:V a topic), preclude it from being a major repeated source ? David Ruben Talk 22:40, 1 May 2010 (UTC)

Generally this talk page is not intended to operate as a noticeboard - WP:MED might be a better spot.I think UpToDate should be discouraged as a source because it's tertiary and harder to access than medical journals, which you can at least access through an interlibrary loan if nothing else. It makes it harder to verify the content which means that subtle or major issues slip by easier. You should explain that to the editor and see if the editor understands. If not, I would request that the editor verify the sources used in UpToDate and provide them on the talk page, and then begin replacing UpToDate citations with secondary sources. If the UpToDate article does not provide its sources, it's a good reason to look for a source which does use sources. II | (t - c) 22:49, 1 May 2010 (UTC)
The policy is at WP:PAYWALL, which I have added to ==See also==.
We might consider developing a section on this point. It could address online access as well as cost. WhatamIdoing (talk) 06:50, 2 May 2010 (UTC)

Transcendental Meditation studies and the press

There is a dispute at Talk:Transcendental Meditation#Use of newspapers as sources in research section over the inclusion of quotations from scientists reported in the popular press in response to studies on the use of Transcendental Meditation by schoolchildren. Some editors believe that doing so violates the "Popular press" section of this guideline. My view is that the press prohibition is aimed at descriptions of study outcomes rather than quotations from scientists, and that NPOV requires we report all significant views. Lastly, if it matters, the TM movement publicizes its studies in the popular press and uses them to seek policy changes to allow the teaching of TM in schools. Any input?   Will Beback  talk  20:40, 16 May 2010 (UTC)

To clarify the discussion and add to Will's comments: Two quotes under discussion are:
"Edzard Ernst, professor of complementary medicine at the Peninsula Medical School in Exeter, has said "there is no good evidence that TM has positive effects on children. The data that exist are all deeply flawed."
"Educational psychologist Kairen Cullen, associate fellow of the British Psychological Society, speaking of TM in a pediatric setting has said it "...is a very difficult sample group to access and it would be very hard to provide empirical evidence - any claims would therefore be pretty speculative".
And the discussion does not center around completely excluding the comments/quotes but also suggests including them in another section specifically for reactions/ comments in the press concerning the TM research. Since there are many articles that reference the TM research NPOV might be best served by a section that shows a cross section of those comments.(olive (talk) 21:04, 16 May 2010 (UTC))
This talk page is only for dicussing the guideline, not for content dispute resolution. I would suggest trying one of these noticeboards, Wikipedia:Content noticeboard, Wikipedia:Neutral point of view/Noticeboard, Wikipedia:Reliable sources/Noticeboard.--Literaturegeek | T@1k? 22:43, 19 May 2010 (UTC)
Apologies - I thought this page functioned as a noticeboard for WP:MEDRS-related issues. I'll post a fresh request at WP:RSN.   Will Beback  talk  22:59, 19 May 2010 (UTC)

Letters to the editor

Are letters to the editor, published in peer-reviewed journals, reliable sources for anything but the opinion of the author? Can they be used without attribution? Can they be sources for fringe claims? If referenced (and lengthy), or published in certain journals (presumably with standards for what LttE will be published) could they be considered on par with an article in the same journal? Where do they fall in WP:PSTS?

And the question that needs to be asked - is this already dealt with somewhere else? WLU (t) (c) Wikipedia's rules:simple/complex 17:47, 10 June 2010 (UTC)

Pretty much all source questions (such as PSTS) depend on what you are drawing from the source. I would think they are similar to letters to newspapers, with the difference that the letter writer is more likely to be a subject expert. Colin°Talk 17:58, 10 June 2010 (UTC)
(edit conflict) I think that they are generally a low evidence level primary source. Higher quality sources, if available, should be used instead. However, I don't think there should be a blanket ban on their use; sometimes letters to the editor raise good points and can be useful if used cautiously. Yes they can be sources of fringe theory, WP:UNDUE, WP:RS and current guidance in WP:MEDRS, WP:CONSENSUS and good editorial judgement should be used to decide if a letter to the editor is being used inappropriately.
Perhaps this guideline needs to discuss and emphasise that editors should strive to use the best level of evidence for articles? Currently it just says secondary is better than primary, but some primary sources are good quality and some are not good quality. Even some reviews are poor quality.--Literaturegeek | T@1k? 18:02, 10 June 2010 (UTC)
This is a perennial comment wrt medical sources. Unfortunately, it takes expertise to judge whether a primary research paper has established its discoveries sufficiently to warrant stating them as facts in an encyclopaedia. The nature of Wikipedia, unlike Britannica say, is that we can't guarantee our editors are experts (even if they say they are) or unbiased (even if they think they are). There's absolutely zero evidence that Wikipedians are better at reviewing the primary literature than review authors. Some may think they are. There's plenty evidence that many Wikipedians are awful at reviewing the primary literature (most don't even have access to the full text). Policy WP:WEIGHT demands we consult the secondary literature in order to judge whether a study's findings are worth mentioning. Policy WP:PSTS greatly limits any use of primary sources. So while we may read what we consider to be a bad review from time to time, we have to live with the constraints WP imposes. Colin°Talk 21:30, 10 June 2010 (UTC)
Yes I agree with what you are saying Colin. I see that you are on a wikipedia break (a wee break, you must be Scottish or Northern Irish :-)), I hope that you enjoy your break.--Literaturegeek | T@1k? 17:03, 29 June 2010 (UTC)

What is this guideline about?

Is this guideline about all medical information or medical information about medicine? A Quest For Knowledge (talk)

This guideline is about all medical information.Doc James (talk · contribs · email) 17:42, 29 June 2010 (UTC)
While I think that's a good idea, that interpretation requires that the title should change to something more applicable. In agreement with this edit, perhaps WP:Identifying reliable sources (medical information) would suit? LeadSongDog come howl! 00:12, 10 July 2010 (UTC)
It also applies to things that most people wouldn't consider to be "medical information", e.g., non-human/pre-clinical research. Just about any sentence that begins "A study in mice showed that..." probably isn't "medical information", but it is covered by this guideline. WhatamIdoing (talk) 01:20, 10 July 2010 (UTC)

Best venue for disputes on application of MEDRS?

I started at Wikipedia:Fringe_theories/Noticeboard#Eye_movement_desensitization_and_reprocessing_.28EMDR.29, but agree that's not the best place for discussion. While this guideline doesn't recommend dispute resolution venues, it suggests WP:NOR/N. Suggestions? Should this guideline recommend such venues? --Ronz (talk) 16:32, 9 July 2010 (UTC)

This has come up from time to time. The usual answer is to discuss at the article's talkpage. If necessary, refer the question to WP:RSN, with courtesy notices on project talkpages, e.g. Wikipedia talk:WikiProject Medicine and the article's talkpage. LeadSongDog come howl! 17:21, 9 July 2010 (UTC)
In this case, I don't think RSN really applies. The basic problem of the article is that reviews are not given preference to primary studies, which is a NPOV/OR issue. Part of the problem may be contributions by researchers themselves. --Ronz (talk) 18:15, 9 July 2010 (UTC)

External link?

User:WeijiBaikeBianji's talk page mentions this page, Warning Signs in Experimental Design and Interpretation by Peter Norvig, chief of research at Google. I wonder whether we might like to have it as an "External link" for this Wikipedia page? WhatamIdoing (talk) 04:31, 14 July 2010 (UTC)

Ideally, it would be great, but I think it would tempt new editors into thinking that it's okay to discuss the shortcomings of primary studies in Wikipedia. It's hard enough for scientifically-minded editors to grasp that reports including physical evidence are irrelevant to WP, except in the rare case where no reviews are available. I was wondering if the Greenhalgh references in Further reading should be deleted for the same reason. (Sad, I know.) Maybe someone could write an article for which these references would be relevant. Best wishes, Postpostmod (talk) 12:44, 21 August 2010 (UTC)

Clarification of scope of guideline -- non-medical assertions in medical articles ???

I would like to propose adding something like the following in the lead, to clarify when this policy does and does not apply:

Note that this guideline only applies to biomedical assertions. Even in articles predominantly about medical subjects, assertions which are not biomedical in nature (e.g. historical facts, biographical information, etc.) do not fall under the scope of this policy, and may use sources which adhere to the normal Wikipedia guidelines for reliable sources.

It could probably use some tweaking, but it should definitely be mentioned. For instance, in an article about vaccinations, we should not be requiring WP:MEDRS sources for biographical information about Louis Pasteur, even though the article is primarily about a medical subject. -- Jrtayloriv (talk) 17:56, 23 August 2010 (UTC)

I think this is already clear in the guideline as currently written. It's explicitly intended to guide our presentation of medical information. If anyone is using this guideline to challenge appropriate biographical sources about Louis Pasteur, please let me know and I'll add my voice to those trying to set them straight. MastCell Talk 17:59, 23 August 2010 (UTC)
I still don't feel that it is clear enough, however, I do feel that the recent change in article title somewhat alleviated this issue (i.e. the title isn't saying "medicine-related articles" anymore, it just says "medicine"). Anyhow, maybe I'll bring this up later -- seems like people are disinterested at the moment. -- Jrtayloriv (talk) 21:37, 28 August 2010 (UTC)

Version for science

This page is being adapted at Wikipedia:Identifying reliable sources (science-related articles) for non-medical science-related articles. Interested editors might want to put the other page on their watchlists. WhatamIdoing (talk) 18:51, 24 August 2010 (UTC)

Anti-research spin most unwelcome

I just changed some verbiage that seemed intended, without coming out and saying it, to prohibit people from describing new and interesting scientific results in Wikipedia articles. I feel as if there is a certain sit-down-and-shut-up mentality that has long stalked the medical articles, in which the greatest of all terrors is that a patient might find out that there is a new drug or experimental therapy being used on a disease, perhaps in some other country. There is no WP:NOUSEFULINFORMATION and we have no ethical obligation to ensure that patients can't learn accurate information about hopeful research developments. I am not pleased with finding yet another anti-science guideline turn up, especially one that is supposed to be about identifying reliable sources. Wnt (talk) 19:29, 28 August 2010 (UTC)

We may be looking at different pieces of the puzzle. As a first point of potential common ground, do you think that entire articles ought to be built from primary sources? Assume that dozens of secondary sources exist (so it's notable), but the editor does not want to use them. WhatamIdoing (talk) 21:28, 28 August 2010 (UTC)
My issue with the current draft [1] is not so radical. I'm not suggesting that secondary sources should be omitted, spurned, or deleted, nor that they should not carry authoritative weight. However, when secondary sources are unavailable, it should be possible to create a section or a small article based on primary results, sometimes leavened with newspaper reports and perhaps independent commentary self-published by known experts. Publication of a primary source in a half-decent biomedical journal should be sufficient to demonstrate a concept's notability, and major conclusions made there should be at least as reliable as major news reports about political and social issues that are routinely considered valid for other articles.
I understand that sometimes, mass media reports a provocative but overly hopeful primary result - and often in such cases, the original researchers carry a near-absolute conflict of interest, and the "newsfomercial" sounds like it's been bought and paid for. But even in such an extreme case, my feeling is that Wikipedia should report the popular delusion, but make strong efforts to get a broader point of view from those who disbelieve it. Because people who see such bogus articles are going to come here for more information, and either we can give them the broader truth, or we can send them out to do web searches on their own. Wnt (talk) 04:43, 29 August 2010 (UTC)
Can you give an example of where Wikipedia would be improved by allowing editors to "create a section or a small article based on primary results, sometimes leavened with newspaper reports and perhaps independent commentary self-published by known experts"? Doing so would break nearly all of Wikipedia's policies never mind this guideline. One has to wonder why such information has not (yet) been discussed by wise folk in reliable sources. What makes us Wikipedian's better than them to disseminate such fresh knowledge to the masses?
You are concerned above that we should give patients early access to the latest research news. That isn't Wikipedia's concern as we aren't a newspaper, magazine or medical charity. There are other resources on the Internet who do that and Google does a good job of finding them. WP:WEIGHT (policy) requires us to justify the inclusion of a topic/finding based on its weight within reliable sources. If no reliable sources have (yet) discussed this topic/finding then it has no weight. It is not the case that publication in a "half-decent biomedical journal" makes a topic/finding notable enough for inclusion. There is simply too much primary research to cover. WP:PSTS (policy) requires that interpretation of primary research requires a secondary source. An encyclopaedia describes the world as it is now, not how it might be in the future or how we might wish it to be if only....
Far from being "anti-science", this guideline is firmly grounded in concepts like evidence-based medicine and a desire to prevent advocacy of quack therapies. The process of scientific research and advancement is to present ones research findings in a publication for discussion by ones peers. Until that discussion has taken place, the "science" process is only half-baked and including it here is usually premature.
There are aspects of this guideline one wouldn't expect to find in the publication guidelines of any magazine or book written by known experts. Those aspects are influenced by the novel feature of Wikipedia: it can be written by anyone, anonymously. It is this that is largely responsible for the policy requirements that we can only really write about stuff other people write about (WP:NOR). Colin°Talk 14:53, 29 August 2010 (UTC)
We're not a medical charity, but we are an information charity, including medical information. People should not be singled out to be denied information simply because they have a disease.
To give an example, I just went to PLoS Biology and found this article. This article can and should be referenced from IL-6 or obesity to say something like "In mice, IL-6 activity in the hypothalamus was found to mediate the ability of exercise to reduce food intake in obese, but not lean animals". The figure [2] could be copied and used as an inline figure, provided that it is labeled something like "proposed role of the hypothalamic anti-inflammatory response mediated by exercise" (which is how the authors labeled it). Wnt (talk) 16:18, 29 August 2010 (UTC)
I should further add that the Discussion section of that PLoS article gives a reasonable review of recent developments involving IL-6 and obesity. As such I think that, despite the usual scientific usage, the section can reasonably be called a "secondary source" for the information it provides, though I would prefix "reviewed in..." to the reference tag just to be clear about the usage. It is very important not to allow policy to rule out the use of such sources, because open access policies did not apply to reviews for quite a while, and probably still don't where some private funders are concerned. So for most readers who actually want a source and not an invitation to buy an article for $36, we should be using such reviews in so-called "primary articles" as often as possible. Wnt (talk) 16:25, 29 August 2010 (UTC)
So you think that free (as in beer) primary sources (e.g., original experimental results) are preferable to WP:PAYWALLed secondary sources (e.g., a literature review)?
I suppose that one of the questions is whether you believe that Wikipedia is an encyclopedia (that happens to contain a list of references), or a bibliography (that happens to contain a text description of what one might find in the sources). (If the former, then your concern about presenting a reader with an invitation to pay $36 is unimportant, because our "product" is the sentences, not the sources.) WhatamIdoing (talk) 16:52, 29 August 2010 (UTC)
By some variant of the equipartition theorem, I expect that the value of the text and the references should approach a fixed ratio, perhaps 1:1. Wnt (talk) 21:32, 29 August 2010 (UTC)
To say we are "an information charity" isn't helpful, as WP:INDISCRIMINATE is at pains to point out. We are an encyclopaedia for the general reader. The primary research in that paper involved experiments on mutant mice and rats. There are very few places on Wikipedia where such basic research could be directly cited as a source. You are right that portions of such a paper can be considered a secondary source on the topic and this has been discussed before in the archives of this guideline. They are not the best choice and a very quick PubMed search for "IL-6 obesity" brings up 263 reviews, at least some of which look promising from their title alone.
Issues of open access and paying for articles are red herrings. Remember also this is a guideline not policy so nothing is "ruled out". Colin°Talk 17:22, 29 August 2010 (UTC)
Wnt, it sounds to me that you believe that the list of sources (presumably including references, recommended books, and external links) is half the value of each encyclopedia article. If this were generally accepted -- if the goal were to provide enough information to get the highly motivated reader started on a reading binge -- then we might favor free-beer sources.
However, I do not believe that the community shares your point of view. I suspect that most editors would say that citations exist primarily (or even exclusively) for the purpose of demonstrating to ourselves that we've correctly written the article, not to save our readers the trouble of looking for other things to read. WhatamIdoing (talk) 01:03, 30 August 2010 (UTC)

Possession of a PMID

An anon at Talk:Blanchard, Bailey, and Lawrence theory is currently asserting that mere possession of a PMID means that the publication cannot possibly be a letter to the editor. This is not an uncommon problem (particularly in this group of articles). Would it be useful to mention that "in the database" is not the same thing as "a reliable source"? WhatamIdoing (talk) 00:13, 29 August 2010 (UTC)

Yes, perhaps the "Searching for sources" should note that PubMed database holds all the articles in these journals, including editorials, letters to the editor and biographies, etc. Perhaps we should repeat that not all articles in "peer-reviewed journals" are peer-reviewed or are research papers or formal reviews of such papers. Colin°Talk 14:59, 29 August 2010 (UTC)
  • Err, no. Unnecessary detail. In this case, however, I think it should be noted that it appears the person you are disputing with points to this InformaWorld response paper as the article in question, which you then say is a letter to the editor (CTRL-F "second paper"). It does not seem to be a letter to the editor - usually these are shown in collected form. Also, many full journal articles are titled "Letter", and in fact many journals are called Letters, so this runs into wikilawyering confusion (examples). I should also note that nobody should be removing primary articles, unless they are replaced with review articles, purely on the basis of being primary. If there's no original research/synthesis going on, WP:PSTS clearly allows primary articles and objections should be raised on substantive issues (e.g., research is emerging and undue). II | (t - c) 18:19, 29 August 2010 (UTC)
I assume you meant "text sourced to primary articles" rather than the articles themselves. WP:PSTS clearly limits the use of primary articles as sources and other policy such as WP:WEIGHT is nearly impossible to satisfy if primary sources are the only ones cited. I agree that one shouldn't go removing perfectly decent text merely because the sourcing isn't first rate. Usually, however, that is a warning sign that things are not as they should be. Colin°Talk 20:00, 29 August 2010 (UTC)
There was discussion a while back at WT:NOR as to whether research articles are really primary sources at all. Most of us think of actual data when we talk about primary sources, and that is how PSTS is worded. Research articles have lots of interpretive and evaluative discussion (read PSTS's description of secondary articles and see the similarities). A question: there's an research article on a RCT cited in a Wikipedia article. This research article discusses four previous RCTs on the same issue and summarizes what these articles found. Is that research article a primary source or a secondary source when it is cited to cover these earlier RCTs? II | (t - c) 21:21, 30 August 2010 (UTC)
ImpIn, that example is, indeed, just a letter to the editor, exactly like it says at the top corner of the page (did you perhaps overlook the label, which says "LETTER TO THE EDITOR" in all caps, on the gray bar?). Informa formats pdfs of letters (and all other content) separately so they can sell them separately. The format on the pdf is not the format in the paper journal. WhatamIdoing (talk) 01:09, 30 August 2010 (UTC)
I actually don't have the article, I just hadn't seen letters to the editor presented that way by any publishers. But I'll take your word for it and I apologize for questioning it - in my defense it's unclear why anyone would argue the point if "Letter to the editor" is stamped on the article. However, this also brings in another point - are letters to the editor necessarily unreliable? I don't think so, and often NPOV will require that we present them in order to present both sides of a contentious issue. Letters to the editor which are formally published are obviously better than "comments" like you see on PLoS One - we could perhaps mention that. But getting into the nitty gritty of technical systems in a guideline seems like too much information. II | (t - c) 21:21, 30 August 2010 (UTC)
A letter to the editor (assuming it's not a hoax, of course) is generally accepted as a reliable source (only) for the fact that the author holds the stated opinion. The fact that letter writers hold opinions is not usually WP:DUE.
Letters (and similar things) should not normally be used to 'debunk' or 'criticize' better quality sources. For example, in this dispute, we have:
  • several secondary sources announcing that the dominant view is X;
  • a primary, peer-reivewed source (Moser 2009) saying that the mainstream view is wrong about one detail;
  • a letter to the editor (Lawrence) saying that Moser's survey was hogwash; and
  • a letter from Moser (2010) saying that Lawrence's criticism of him is hogwash.
IMO neither of the two letters deserve space in Wikipedia -- we do not pretend that two sides of a dispute have equal validity -- and the primary source should be included only in those articles that actually mention the disputed detail (e.g., this one, but not that one: after all, if the detail is too trivial to mention in a given article, then primary-source criticism of the detail is also too trivial to mention in that article). WhatamIdoing (talk) 21:59, 30 August 2010 (UTC)

Wikipedia:RSMED listed at Redirects for discussion

Please comment at the redirect discussion. Should it be decided to maintain that redirect, a secondary consideration is whether it is wise to advertise it on this guideline page. It is one thing having a backup for misremembered typos [something that history shows not to be required as there are only a few links using it], but another by having this guideline confusingly known by two different names (MEDRS and RSMED). Colin°Talk 15:45, 6 September 2010 (UTC)

The RfD closed with no consensus. It is hard to delete things on Wikipedia, even things that would cause confusion and aren't actually used/useful: what links here. However, having the unused/unloved redirect is one thing. Advertising it on this guideline page is another. This guideline is not known as Wikipedia:Identifying reliable sources (medicine), or by the old name Wikipedia:Reliable sources (medicine-related articles) (why on earth was it changed, anyway?), it is known universally as MEDRS. I can't see what benefit there is to having talk pages confused by some folk referring to RSMED and some to MEDRS. So I propose we don't advertise the RSMED shortcut, and the link removed from the top of the guideline. If you agree, could someone do the deed. My earlier edit was reverted by the person who added the shortcut. Colin°Talk 18:29, 29 September 2010 (UTC)

testimonial evidence

I disagree with your editor on the aspartame page. The root cause is wether to accept the position adopted by the government or to accept the testimonials of many people that have cured very serious illnesses by ceasing long term use of aspartame. While I have no problem with presenting the established viewpoint I would like to point out that personnel testimonials are acceptable in the courts and should be also included. I know this is against the no original research policy but the fact is that it is impossible for medical studies to predict the outcome of a 20 year exposure while personal experiences do just that. Perhaps the no original research policy should be modified when large numbers of people independently come to the same conclusion. Arydberg (talk) 13:06, 6 October 2010 (UTC)

Thank you for taking the time to share your opinion. This guideline is simply not allowed to overrule the WP:No original research policy, so I'm afraid there's nothing that we can do to permit what you want -- even if most of us wanted to, which I don't believe is the case.
Anecdotal evidence is a poor source of information, because it doesn't address confounding factors. For example, a primary source of artificial sweetener in most people's diets is diet soda. Dropping artificial sweetener also causes the person to drop artificial colors, carbonation, preservatives, and caffeine. It's possible that it is the act of removing one of these from the diet that actually causes any improvement, rather than the removal of the artificial sweetener. Similarly, dropping artificially sweetened over-processed junk food in favor real food like fruits and vegetables might be associated with health improvements -- but by adding good food to the diet, not by removing artificial sweeteners. (After all, people who dump full-sugar junk food for fruits and vegetables also see similar improvements.)
In short, anecdotes simply don't (can't) prove anything (except "Joe believes this is the cause"), and should not be relied upon in any scholarly article. WhatamIdoing (talk) 18:11, 6 October 2010 (UTC)

Thank you for your response but I must respectfully disagree. Your use of the term anecdotal evidence is not accurate. Testimonial is the accurate term. anecdotal relates to stories passed from person to person not to firsthand accounts. ( see the Wikipedia entry on anecdotal evidence).

Given the following about aspartame the sweetener used in diet Coke and diet Pepsi:

1) That Donald Rumsfield was instrumental in the approval of aspartame after it had been refused approval for eight years.

2) That Pub Med shows 28 case histories of serious problems with aspartame, some from the use of aspartame in drugs.

3) That the internet has many many testimonials of serious medical problems like brain seizures and brain tumors attesting to the benefits of giving up aspartame after long term exposure

4) That medical science is completely unable to accurately test the long term exposure dangers of any substance.

5) That 75 percent of all the complaints received by the FDA are about aspartame.

Then I think there is a reasonable probability that there may be real dangers associated with the use of aspartame and these deserve mention. Arydberg (talk) 14:16, 7 October 2010 (UTC)

There are sufficient recent high-quality reviews for this subject to establish that the scientific debate continues, such as doi:10.1080/10408440701516184 (PMID 17828671), this (PMID 18604921) and even doi:10.1089/fpd.2008.0089 (PMID 18666862). There is absolutely no need or excuse for resorting to low-quality anecdotal testimonials to make this point. LeadSongDog come howl! 17:08, 7 October 2010 (UTC)
It would also create a most unwanted precedent: we would end having such kind of anecdotical evidence in every alternative-completely unproven-even dangerous-treatment in the market article, since there are always some (or many) who see incredible benefits (or worsenings) where there are none. It goes against some of the most core policies of wikipedia, so I hardly see the point even in its discussion. On a side note: precisely what you (or many others) think, if not published as a reliable source- its what we do not want to have in wikipedia since it is OR and quite irrelevant.--Garrondo (talk) 17:48, 7 October 2010 (UTC)

Sorry, I just think large numbers of people with accounts of personal experience should trump experiments with rats and mice. Arydberg (talk) 11:28, 9 October 2010 (UTC)

But if the weight of personal experience is that strong, then this would surely be compelling data for some reliable source to use, and that source could be referenced within WP. Otherwise we slide into WP:OR and soapboxes. For example, we can find many powerful personal testimonies from people who have won the lottery, but does that mean that winning the lottery is the norm, or even a realistic possibility to be recommended for the average person. Peer review and the editors of reliable sources weed out such issues. Surely it's not our job to have these battles on the talk pages of WP? -- TerryE (talk) 20:03, 9 October 2010 (UTC)
An alternative way of looking at things is that if Arydberg is allowed to insert his original research, then opposing original research viewpoints would need to be included, for example if someone wants to say aspartame side effects from long-term use are psychosomatic in origin, individuals are wacko and so forth, also will have to be included. The whole system then degenerates into chaos and a battle. Surely Arydberg you can understand why original research is not allowed? The WP:NOR policy occasionally gets in the way of editing articles and is not perfect but it is many many times better than the chaos that would ensue if original research was allowed.
To change the topic, this talk page is only meant for improving this guideline; this really is the wrong place to be discussing this, it should be discussed at WP:NOR talk page but I can tell you that there is essentially zero chance of the community changing their minds and allowing for original research.--Literaturegeek | T@1k? 20:21, 9 October 2010 (UTC)

Wikipedia talk:Verifiability#WP:PAYWALL and freely accessible sources

Editors may wish to take a look at this discussion, which is about the use of paywalled sources. NW (Talk) 02:59, 21 August 2010 (UTC)


Or this one http://www.theatlantic.com/magazine/archive/2010/11/lies-damned-lies-and-medical-science/8269 —Preceding unsigned comment added by 131.109.225.2 (talk) 17:51, 30 October 2010 (UTC)

Alternative Medicines and Reliable Sources

This stems from my recent atempt to to add Aloe Vera to the Arthritis page and having it removed - a brief discussion with the person who removed it is at User_talk:Jfdwolff. Having read the endless arguments from others trying to post alternative medicines I believe the problem actually lies here in the Wiki policy for Identifying reliable sources.

Many ailments on Wikipedia should have an Alternative Medicine/Other therapies section. Arthritis, for example, could include Aloe Vera, Acupuncture, Aromatherapy, Chinese Medicine, and so on. But the field of alternative medicine is so wide and the strength of evidence varies - some have a reasonable amount in literature and others are extremely spurious - and very few have proper scientific research.

The reason for this is fairly clear - natural substances or processes cannot be patented so they do not attract the funding required for full clinical trials. Effectively this means a natural substance may be better than a prescription drug but no one is prepared to put up the funds to prove it. This is a defect in the modern world of medicine, not just Wikipedia.

What we end up with is contributors on Wikipedia trying to please the guidelines by quoting the best research they can find, but it will not be conclusive. This leads to inconsistancy. My contribution for Aloe Vera on the Arthritis page was deleted although it contained three references while on Rheumatoid arthritis Ayurveda has been accepted even though there is no citation.

It's fair to say that society uses alternative medicine and its use is widespread. The British Medical Journal (Vol. 312, 20/3/1996) found the precentage of people using over the counter or complementary medicine rather than seeking a consultation with a doctor was extremely high in certain cases: headache: 80%; heartburn: 62%; mouth ulcers: 51%; etc. I suggest the figures are even higher now.

As a first strike I am arguing the following:

1) There should be nothing to stop an umbrella statement on many ailment pages, under the Treatment section, to say that 'there are alternative medicines/other therapies in use'. This reflects society.
2) Where there is a large amount of work on the subject in printed books (vanity and self-published excluded) these should be accepted as 'reliable sources'. But how to quantify this?

Polishwanderer (talk) 14:29, 11 October 2010 (UTC) 14:23, 11 October 2010 (UTC)

Why would I use aloe vera on my knee when water is as effective and free? Why should it be included in a reference work when it is no more effective as a placebo than any randomly selected inert substance. Fred Talk 19:10, 31 October 2010 (UTC)
I totally agree that we need to cover commonly used modalities of complimentary and alternative medicine for medical conditions. The difficulty is, as you correctly note, finding out how widespread the use of particular modalities actually is. There is plenty of data about acupuncture in osteoarthritis of the knee, saw palmetto for benign prostatic hypertrophy, echinacea for the common cold, and so on. For conditions like arthritis (without modifier), the data is much less clear.JFW | T@lk 14:47, 11 October 2010 (UTC)
Yes, we should note widespread use, if there is a good reference for that use. And note widespread anecdotal reports, if they exist. However, care should be taken to avoid phrasing that scientific research supports the use, unless that evidence does exist. Fred Talk 19:10, 31 October 2010 (UTC)

Biology and Medicine

I am curious, as far as Wikipedia is concerned, where exactly does medicine end and biology begin? In other words, I would like to know if MEDRS guidelines apply to all of the biological sciences. razorbelle (talk) 05:05, 31 October 2010 (UTC)

The special requirements for sources for medical information are related to the obvious risks of reliance by either physicians and patients on information which relates to medical conditions and treatments. Obviously such reliance is misplaced, see Wikipedia:General disclaimer, but we are obliged to, within the framework of our software and editing policies, do what we can to maintain a higher standard then usually followed for scientific subjects such as biology. So, if the subject is disease or treatment of disease, the special requirements apply, if the subject is simply biology, no. There is, however, a gray area, relating to biochemistry, where the medical standards may be more appropriate. Fred Talk 18:43, 31 October 2010 (UTC)
No matter what the subject, you ought to be using the best sources available to you. If the information on this page helps you identify better sources for biology -- or engineering, or music, or whatever else you're writing about -- then please take whatever good you can from it. Nobody who thinks Wikipedia needs good sources for medicine-related articles is going to advocate for using lousy sources in, say, botany-related articles just because "plants are not people".[3] WhatamIdoing (talk) 21:24, 31 October 2010 (UTC)

Funding of cited research

This study, "How Do US Journalists Cover Treatments, Tests, Products, and Procedures? An Evaluation of 500 Stories" looked at media reports on phamaceutical products. Among other things, it criticised the fact that many media reports cite experts or studies that have a financial tie to the manufacturer of the drug, without disclosing these ties. I think that is a valid criticism that we should take on board as well. If we cite studies and experts that have such ties, we should make the reader aware of these ties, and I propose adding a sentence to that effect to the guideline. --JN466 14:28, 27 October 2010 (UTC)

Can you give me an example of such a source in a live article, with how it is currently presented versus how you think it should be presented?
Also -- this complaint is usually made (by reasonable people) about prescription medications that are currently under patent. But we run into FRINGEy challenges on this basis. For example, there was an editor last year complaining that all the studies about Endoscopic thoracic sympathectomy (article needs serious work) were biased because the study authors were almost all surgeons who had been paid to do the surgeries. WhatamIdoing (talk) 16:16, 27 October 2010 (UTC)
I can give you examples of where I think studies are attributed correctly:
For all I know, this practice may be followed widely already, but it is worth mentioning in the guideline. I am definitely not complaining about our coverage, or saying that studies like this are necessarily biased or unreliable; I just think it is good practice to indicate who performed or funded a study, and especially so where the researchers could be seen as having an interest in promoting a product or treatment. It's part of correct attribution. --JN466 15:01, 29 October 2010 (UTC)
I haven't looked at that specific example (TM), but I suspect that's likely a case where higher quality independent research is available that would trump whatever assertions are made there ... so, MEDRS has it covered already. SandyGeorgia (Talk) 15:05, 29 October 2010 (UTC)
In both the linked cases, the studies are cited and attributed. Surely that's a good thing? --JN466 15:18, 29 October 2010 (UTC)
JN, I actually wanted a bad example/an example of a problem that needs fixed, because if people are already handling this correctly, then telling them to do exactly what they're doing is kind of WP:CREEPy.
So I've got a downside: If we add this, the SPA at the ETS article is probably going to try to spam "In another study by paid, ETS-promoting surgeons" into half of the paragraphs in the article. I also reasonably expect someone at MCS to do something similar for every study showing benefit from any type of psychiatric medication. IMO that's a big downside. Can you show me some upside? Can you show me a problem that this will actually fix? WhatamIdoing (talk) 05:18, 30 October 2010 (UTC)
I'll try and find an example from our articles where a study isn't attributed. To address your concern, a wording like "In another study by paid, ETS-promoting surgeons" is clearly inappropriate, as it adopts a POV in the article's editorial voice, rather than reporting a POV. (It also sounds, frankly, like an unsourced POV.) But I see no harm in disclosing that a study was performed and reported by employees of the manufacturer, for example. Could I ask you to have a look at this commentary? It's authored by about a dozen editors of the most highly regarded medical journals, including the editors of Annals of Internal Medicine, the Journal of the American Medical Association, The Lancet, and the journals of ten other national medical associations. This is the medical publishing mainstream. It stresses the importance of disclosure of potential conflicts of interest, by everyone involved in the publication of a study. If disclosing potential conflicts of interest is best practice in medical publishing, it is a practice which we should enshrine in our guideline too. --JN466 14:11, 30 October 2010 (UTC)
Disagree. See below. Colin°Talk 15:24, 30 October 2010 (UTC)

The problem is that "mak[ing] the reader aware of these ties" is an ad hominem attack and often unencyclopaedic. Any doubts about the soundness of research, bias in a paper, or journal, or endemic problems with certain topics of research (e.g. TM) must come from our sources. We can't introduce them into the article ourselves. Do our sources cast doubt on a particular trial because of its funding? The ENHANCE study mentioned above does appear (I'm no expert, just looked briefly at the article) to be a study that is discussed in itself by reliable sources. Mostly, studies are not notable and our articles should not mention them, just the conclusions. Or better still, document the real-world non-research facts (e.g. rather than cite a research paper to say "drug X has been found to be effective for treating Y", cite a clinical recommendations guideline that says "drug X is recommended as a first-line treatment for Y". Newspapers focus on studies because they are news, they tend to spend little time finding out about the subject, and they just regurgitate the press release, which is effectively a self-published source. The 10-point HealthNewsReview.org Ratings Criteria mentioned in the PLoS paper is excellent. The problem with not considering (or mentioning) the funding-bias of quoted researchers in newspaper stories is only one small aspect of the mistakes newspapers make. Colin°Talk 10:01, 30 October 2010 (UTC)

If sources express doubt about the soundness of research, it may well be appropriate to report what these sources say. If no sources do so, then such doubt has no place in our article. However, this is a separate issue from disclosure of potential conflicts of interest by those who performed a study. Where such disclosures have been made, we should report them if we cite a study. [4] --JN466 14:16, 30 October 2010 (UTC)
The funding bias is only one aspect of a study that might have an impact on its worth. Other aspects include study size, trial design, patient choice, statistical strength of conclusions, degree of benefit, degree of harm. That's before we even consider clinical aspects such as cost and availability. Unless the article is actually about the study (and we have few such notable studies) then it is fairly likely that this level of detail is inappropriate and that disclosure of funding is no more notable than any other aspect of the study.
Even when it is appropriate to discuss studies in article text, we should do so via a secondary source. That secondary source can make comments about inadequacies or strengths. If we use clinical guideline reports for sources, then we can draw on the wisdom of a whole panel of experts who will review all aspects of the studies they use as evidence, including funding bias. Highlighting this one aspect is likely to fail policy. Colin°Talk 15:24, 30 October 2010 (UTC)
I agree that discussing a study should be done via secondary sources commenting on that study. But if the study itself discloses a conflict of interest, we should mention this conflict of interest in our articles if we use the study as a source, in line with best journalistic practice.
This is indeed not just about funding; study design and data access are other points specifically mentioned in the journal editors' document. The document is not covered by copyright; I append it below for your and other editors' reference. --JN466 16:41, 30 October 2010 (UTC)
Colin, you said, "The problem is that "mak[ing] the reader aware of these ties" is an ad hominem attack and often unencyclopaedic." You are talking about information here that study authors are required to provide according to medical science publications policy, that editors of the world's leading medical journals have said should be published to safeguard ethical standards and maintain the public's confidence in science, and that gold-standard media articles should include, according to multiple studies of science reporting. --JN466 17:20, 30 October 2010 (UTC)
JN, the study authors are required to report their sample size, too. They are required to provide a list of references. They are required to provide contact information for at least one author. Should we include everything that they're required to provide?
Or should we perhaps agree that what's appropriate for an original peer-reviewed research paper is not necessarily or automatically appropriate for an encyclopedia article? WhatamIdoing (talk) 22:48, 30 October 2010 (UTC)
These standards apply even to media articles. --JN466 12:05, 31 October 2010 (UTC)

Commentary and guidelines defined by the editors of major medical journals

Quotation from Sponsorship, authorship and accountability

Frank Davidoff, Catherine D. DeAngelis, Jeffrey M. Drazen, M. Gary Nicholls, John Hoey, Liselotte Højgaard, Richard Horton, Sheldon Kotzin, Magne Nylenna, A. Overbeke, Harold C. Sox, Martin Weyden and Michael S. Wilkes The authors are members of the International Committee of Medical Journal Editors. Dr. Davidoff is Editor Emeritus, Annals of Internal Medicine; Dr. DeAngelis is Editor, Journal of the American Medical Association; Dr. Drazen is Editor-in-Chief, New England Journal of Medicine; Dr. Nicholls is Editor, New Zealand Medical Journal; Dr. Hoey is Editor, Canadian Medical Association Journal; Dr. Højgaard is Editor-in-Chief, Ugeskrift for Læger (Journal of the Danish Medical Association); Dr. Horton is Editor, The Lancet; Dr. Kotzin is Executive Editor, MEDLINE/Index Medicus; Dr. Nylenna is Editor-in-Chief, Tidsskrift for Den norske lægeforening (Journal of the Norwegian Medical Association); Dr. Overbeke is Executive Editor, Nederlands Tijdschrift voor Geneeskunde (Dutch Journal of Medicine); Dr. Sox is Editor, Annals of Internal Medicine; Dr. Van Der Weyden is Editor, Medical Journal of Australia; and Dr. Wilkes is Editor, wjm western journal of medicine.

Correspondence to: Dr. John Hoey, CMAJ, 1867 Alta Vista Dr., Ottawa ON K1G 3Y6; fax 613 565-5471; john.hoey@cma.ca

As editors of general medical journals, we recognize that the publication of clinical-research findings in respected peer-reviewed journals is the ultimate basis for most treatment decisions. Public discourse about this published evidence of efficacy and safety rests on the assumption that clinical-trials data have been gathered and are presented in an objective and dispassionate manner. This discourse is vital to the scientific practice of medicine because it shapes treatment decisions made by physicians, and drives public and private health care policy. We are concerned that the current intellectual environment in which some clinical research is conceived, study subjects are recruited, and data are analyzed and reported (or not reported) may threaten this precious objectivity.

Clinical trials are powerful tools; like all powerful tools, they must be used with care. They allow investigators to test biologic hypotheses in living patients, and they have the potential to change the standards of care. The secondary economic impact of such changes can be substantial. Well-done trials, published in high-profile journals, may be used to market drugs and medical devices, potentially resulting in substantial financial gain for the sponsor. But powerful tools must be used carefully. Patients participate in clinical trials largely for altruistic reasons — that is, to advance the standard of care. In the light of that truth, the use of clinical trials primarily for marketing, in our view, makes a mockery of clinical investigation and is a misuse of a powerful tool.

Until recently, academic, independent clinical investigators were key players in design, patient recruitment and data interpretation in clinical trials. The intellectual and working home of these investigators, the academic medical centre, has been at the hub of this enterprise, and many institutions have developed complex infrastructures devoted to the design and conduct of clinical trials.1,2 The academic enterprise has been a critical part of the process that led to the introduction of many new treatments into medical practice and contributed to the quality, intellectual rigor and impact of such clinical trials. But, as economic pressures mount, this may be a thing of the past.

Many clinical trials are performed to facilitate regulatory approval of a device or drug rather than to test a specific novel scientific hypothesis. As trials have become more sophisticated and the margin of untreated disease harder to reach, there has been a great increase in the size of the trials and consequently in the costs of developing new drugs. It is estimated that the average cost of bringing a new drug to market in the United States is about $500 million.3 The pharmaceutical industry has recognized the need to control costs and has discovered that private, nonacademic research groups (i.e., contract research organizations [CROs]) can do the job for less money and with fewer hassles than academic investigators. Over the past few years CROs have received the lion's share of clinical-trial revenues. For example, in 2000 in the United States, CROs received 60% of the research grants from pharmaceutical companies, as compared with only 40% for academic trialists.1

As CROs and academic medical centres compete head to head for the opportunity to enroll patients in clinical trials, corporate sponsors have been able to dictate the terms of participation in the trial, terms that are not always in the best interests of academic investigators, the study participants or the advancement of science generally.4 Investigators may have little or no input into trial design, no access to the raw data, and limited participation in data interpretation. These terms are draconian for self-respecting scientists, but many have accepted them because they know that if they do not, the sponsor will find someone else who will. And, unfortunately, even when an investigator has had substantial input into trial design and data interpretation, the results of the finished trial may be buried rather than published if they are unfavourable to the sponsor's product. Such issues are not theoretical. There have been a number of recent public examples of such problems, and we suspect that many more go unreported.5,6

As editors, we strongly oppose contractual agreements that deny investigators the right to examine the data independently or to submit a manuscript for publication without first obtaining the consent of the sponsor. Such arrangements not only erode the fabric of intellectual inquiry that has fostered so much high-quality clinical research, but also make medical journals party to potential misrepresentation, since the published manuscript may not reveal the extent to which the authors were powerless to control the conduct of a study that bears their names. Because of our concern, we have recently revised and strengthened the section on publication ethics in the "Uniform Requirements for Manuscripts Submitted to Biomedical Journals: Writing and Editing for Biomedical Publication," a document developed by the International Committee of Medical Journal Editors (ICMJE) and widely used by individual journals as the basis for editorial policy. The revised section* follows this commentary. (The entire "Uniform Requirements" document is currently undergoing revision; the revised version should be available at the beginning of 2002.) As part of the reporting requirements, we will routinely require authors to disclose details of their own and the sponsor's role in the study. Many of us will ask the responsible author to sign a statement indicating that he or she accepts full responsibility for the conduct of the trial, had access to the data and controlled the decision to publish.

We believe that a sponsor should have the right to review a manuscript for a defined period (e.g., 30 to 60 days) before publication to allow for the filing of additional patent protection, if required. When the sponsor employs some of the authors, these authors' contributions and perspective should be reflected in the final paper as are those of the other authors, but the sponsor must impose no impediment, direct or indirect, on the publication of the study's full results, including data perceived to be detrimental to the product. Although we most commonly associate this behaviour with pharmaceutical sponsors, research sponsored by governmental or other agencies may also fall victim to this form of censorship, especially if the results of such studies appear to contradict current policy.

Authorship means both accountability and independence. A submitted manuscript is the intellectual property of its authors, not the study sponsor. We will not review or publish articles based on studies that are conducted under conditions that allow the sponsor to have sole control of the data or to withhold publication. We encourage investigators to use the revised ICMJE requirements on publication ethics to guide the negotiation of research contracts. Those contracts should give the researchers a substantial say in trial design, access to the raw data, responsibility for data analysis and interpretation, and the right to publish — the hallmarks of scholarly independence and, ultimately, academic freedom. By enforcing adherence to these revised requirements, we can as editors assure our readers that the authors of an article have had a meaningful and truly independent role in the study that bears their names. The authors can then stand behind the published results, and so can we.


Footnotes

Conflict of interest

Public trust in the peer review process and the credibility of published articles depend in part on how well conflict of interest is handled during writing, peer review and editorial decision making. Conflict of interest exists when an author (or the author's institution), reviewer or editor has financial or personal relationships with other persons or organizations that inappropriately influence (bias) his or her actions. The potential of such relationships to create bias varies from negligible to extremely great; the existence of such relationships does not necessarily represent true conflict of interest, therefore. (Relationships that do not bias judgement are sometimes known as dual commitments, competing interests or competing loyalties). The potential for conflict of interest can exist whether or not an individual believes that the relationship affects his or her scientific judgement. Financial relationships (such as employment, consultancies, stock ownership, honoraria, paid expert testimony) are the most easily identifiable conflicts of interest and the most likely to undermine the credibility of the journal, the authors, and of science itself. Conflicts can occur for other reasons, however, such as personal and family relationships, academic competition and intellectual passion.

All participants in the peer review and publication process must disclose all relationships that could be viewed as presenting a potential conflict of interest. Disclosure of these relationships is particularly important in connection with editorials and review articles, because bias can be more difficult to detect in those publications than in reports of original research. Editors may use information disclosed in conflict of interest and financial interest statements as a basis for editorial decisions. Editors should publish this information if they believe it will be important to readers in judging the manuscript.

Potential conflicts of interest related to individual authors' commitments

When authors submit a manuscript, whether an article or a letter, they are responsible for disclosing all financial and personal relationships between themselves and others that might bias their work. To prevent ambiguity, authors must state explicitly whether potential conflicts do or do not exist. Authors should do so in the manuscript on a conflict of interest notification page that follows the title page, providing additional detail, if necessary, in the accompanying cover letter.

Investigators should disclose potential conflicts to study participants, and should state in the manuscript whether they have done so.

Editors also need to decide when to publish information disclosed by authors about potential conflicts. If doubt exists, it is best to err on the side of publication.

Potential conflicts of interest related to project support

Increasingly, biomedical studies receive funding from commercial firms, private foundations and government. The conditions of this funding have the potential to bias and otherwise discredit the research.

Scientists have an ethical obligation to submit creditable research results for publication. As the persons directly responsible for their work, researchers therefore should not enter into agreements that interfere with their access to the data, their ability to analyze the data independently, to prepare manuscripts and to publish them. Authors should describe the role of the study sponsor(s), if any, in study design; in the collection, analysis and interpretation of data; in the writing of the report; and in the decision to submit the report for publication. If the supporting source had no such involvement, the authors should so state. Biases potentially introduced when sponsors are directly involved in research are analogous to methodological biases of other sorts; some journals therefore choose to include information about the sponsor's involvement in the methods section of the published paper.

If a study is funded by an agency with a proprietary or financial interest in the outcome, editors may ask authors to sign a statement such as, "I had full access to all of the data in this study and I take complete responsibility for the integrity of the data and the accuracy of the data analysis." Editors should be encouraged to review copies of the protocol and/or contracts associated with project-specific studies before accepting such studies for publication. Editors may choose not to consider an article if a sponsor has asserted control over the authors' right to publish.

Conflicts of interest related to commitments of editors, journal staff or reviewers

Editors should avoid selecting external peer reviewers with obvious potential conflicts of interest, for example, those who work in the same department or institution as any of the authors. Authors often provide editors with the names of persons they feel should not be asked to review a manuscript because of potential conflicts of interest, usually professional. When possible, authors should be asked to explain or justify their concerns; that information is important to editors in deciding whether to honour such requests.

Reviewers must disclose to editors any conflicts of interest that could bias their opinions of the manuscript, and they should disqualify themselves from reviewing specific manuscripts if they believe such disqualification would be appropriate. As in the case of authors, silence on the part of reviewers concerning potential conflicts may mean either that such conflicts exist that they have failed to disclose, or that conflicts do not exist. Reviewers must therefore also be asked to state explicitly whether conflicts do or do not exist. Reviewers must not use knowledge of the work, before its publication, to further their own interests.

Editors who make final decisions about manuscripts must have no personal, professional or financial involvement in any of the issues they might judge. Other members of the editorial staff, if they participate in editorial decisions, must provide editors with a current description of their financial interests (as they might relate to editorial judgments) and disqualify themselves from any decisions where they have a conflict of interest. Editorial staff must not use the information gained through working with manuscripts for private gain.

Editors should avoid submitting to their own journal reports of original research to which they have contributed as authors. If they do so, they should recuse themselves from the editorial process, and delegate editorial decisions on those manuscripts to other members of the editorial staff.

Editors should publish regular disclosure statements about potential conflicts of interests related to the commitments of journal staff.


This document is not covered by copyright. It may be copied and distributed without charge for not-for-profit purposes.

Competing interests: None declared for Drs. Davidoff, DeAngelis, Nicholls, Hoey, Højgaard, Horton, Kotzin, Nylenna, Overbeke, Sox, Van Der Weyden or Wilkes. Before July 2000, Dr. Drazen received speaker fees from Merck Frosst and consultancy fees from Merck Frosst, Immunex and Roche Bioscience, and was a member of the scientific advisory boards of Aerocrine, Aradigm, Entelos, Inspire Pharmaceuticals and Lifemasters; he currently has no arrangements with any commercial entity in the health care sector other than his current employer, the Massachusetts Medical Society.


References

  • Henderson L. More AMCs finding growth from reform. Centerwatch 2000;7(6):1,10-3.
  • Kowalczyk L. Harvard, other medical schools aim to give drug firms faster pace for trials. Boston Globe 2000;Jul 28:C4.
  • Mathieu MP. Parexel's pharmaceutical R&D sourcebook, 1998. Waltham (MA): Parexel International Corporation; 1999.
  • Rennie D. Thyroid storm. JAMA 1997;277:1238-43.[Abstract/Free Full Text]
  • Kahn JO, Weng D, Mayer K, Murray H, Lagakos S, for the 806 Investigator Team. Evaluation of HIV-1 immunogen, an immunologic modifier, administered to patients infected with HIV having 300 to 549 x 106/L CD4 cell counts. JAMA 2000;284:2193-202.[Abstract/Free Full Text]
  • Blumenthal D, Campbell EG, Anderson MS, Causino N, Louis KS. Withholding research results in academic life science: evidence from a national survey of faculty. JAMA 1997;277:1224-8.[Abstract/Free Full Text]

The above document is from 2001 and describes the changes in medical publishing guidelines made at the time.

Here is another study from 2007 (Cook DM, Boyd EA, Grossmann C, Bero LA (2007) Reporting Science and Conflicts of Interest in the Lay Press. PLoS ONE 2(12): e1266. doi:10.1371/journal.pone.0001266), finding that in a significant number of cases the media did not report financial ties in their stories even though they were disclosed in scholarly journals; it concludes: "Because science journalists often did not report conflict of interest information, adherence to gold-standard recommendations for science journalism was low. Journalists work under many different constraints, but nonetheless news reports of scientific research were incomplete, potentially eroding public trust in science."

Wikipedia should adhere to that "gold standard" which the media fail to uphold: where conflicts of interest are disclosed in the scholarly journals, we should report them if we use the study as a source. --JN466 16:41, 30 October 2010 (UTC)

The problem here is the "if we use the study as a source". Don't. Just don't. See WP:PSTS, which is policy. If editors follow policy, the facts that they do draw from a primary source (should they find the need to use one) are so limited and uncontroversial that no qualifier is necessary. Colin°Talk 17:19, 30 October 2010 (UTC)
The document quoted above (hatted) makes clear that this applies to review articles as much as it does to individual studies: "Disclosure of these relationships is particularly important in connection with editorials and review articles, because bias can be more difficult to detect in those publications than in reports of original research. Editors may use information disclosed in conflict of interest and financial interest statements as a basis for editorial decisions. Editors should publish this information if they believe it will be important to readers in judging the manuscript." I am entirely in agreement with you that articles should usually cite review articles rather than primary sources, but the WP:MEDRS guideline explicitly permits citation of primary sources. If you think this is wrong, that is a separate topic for discussion; but as long as editors are allowed to cite primary sources, the question of disclosed conflicts of interest in these studies remains relevant. --JN466 17:31, 30 October 2010 (UTC)
We aren't writing a "review article" in a medical journal. I don't think MEDRS is wrong to "permit" citing primary sources, as to do otherwise would be in conflict with policy. But, as I said, the facts that we can draw from such sources don't need a COI qualification as they are very limited indeed. Perhaps you could give an example were you felt it was justified to cite a primary research paper and where a funding disclosure was necessary. Colin°Talk 18:08, 30 October 2010 (UTC)
Colin, FWIW, this concern arose from wider discussions at WP:V talk which spilled onto the Foundation list: [5][6]. I don't normally edit medical articles, but I was concerned about the possibility that we do not follow ethical policies that are considered best practice in reliable sources. --JN466 19:03, 30 October 2010 (UTC)
You know, I don't know anyone on WP that doesn't want to prevent Big Pharma from influencing our articles. But the ideas suggested by some editors on those discussions would actually have the opposite effect. Big Pharma OWN the press. Sure, there's the odd journalist doing investigative work, and there are articles discussing drug controversy, but nearly all press articles on drugs are rehashed press releases written by pharmaceutical companies. By citing such an article, one is effectively citing a self-published source. Colin°Talk 19:46, 30 October 2010 (UTC)
Outside of investigative work, or articles that result from some academic whistleblower contacting the media, I'm sure it's much as you say. --JN466 01:24, 31 October 2010 (UTC)

Example of an article citing a study without disclosing financial ties

WhatamIdoing, here is an example: [7]. This says,

Natalizumab was evaluated in two randomized, double-blind, placebo-controlled trials in people with multiple sclerosis. Both studies enrolled individuals with MS who experienced at least one clinical relapse during the prior year and had a Kurtzke EDSS score between 0 and 5. In these trials natalizumab was shown to reduce relapses in individuals with MS by 68% vs. placebo, a margin far greater than had been seen for other approved MS therapies.[9] Natalizumab also slowed the progression of disability in patients with relapsing MS.[9] In combination with interferon beta-1a (IB1A), relapsing and disability progression were reduced more than IB1A alone.[11] Other benefits of natalizumab use by patients with relapsing MS included reduced visual loss,[12] a significant increase in the proportion of disease-free individuals,[13] significantly improved assessments of health-related quality of life in relapsing individuals,[14][15] reduced cognitive decline of a portion of individuals with MS,[16] reduced hospitalizations and steroid use,[17] and prevention of the formation of new lesions.[9][18] ...

These are not "very limited" facts. The first primary source cited is this; at the bottom, it says "Supported by Biogen Idec and Elan Pharmaceuticals. Data were analyzed by Biogen Idec and Elan Pharmaceuticals", followed by a long list of financial ties each of the study's authors has to the manufacturers. Another source cited in the same paragraph is [8]; this too says, "Supported by: Biogen Idec and Elan Pharmaceuticals." Our article makes no mention of this, but I think it should. It shouldn't make a big deal out of it, but it should mention it as a matter of course, for example like this: "The product's manufacturers funded/performed two randomized, double-blind placebo-controlled trials of Natalizumab in people with multiple sclerosis ..." --JN466 18:54, 30 October 2010 (UTC)

No, the text you quote is in the "Indications" section and has much wrong with it. Such a section should cite clinical guidelines and greatly limit discussion of individual studies (often to zero). Let me be clear: a section describing when a drug should be prescribed has no business citing research studies. Nor has it any business citing newspaper articles either. If there is controversy about a drug being prescribed/promoted/advertised beyond the level justified by the research, or concerns the research may be biased, then such controversy should be noted (in another section) with the weight it deserves. Such controversy is not a "medical fact" and may cite various reliable sources including newspapers. Colin°Talk 19:48, 30 October 2010 (UTC)
That is your opinion, and it sounds like it might be a very sensible opinion, but there is nothing to that effect in this guideline. We say,
"The best evidence comes from meta-analyses of randomised controlled trials (RCTs), and from systematic reviews of bodies of literature of overall good quality and consistency addressing the specific recommendation. Narrative reviews can help establish the context of evidence quality. Roughly in descending order of quality, lower-quality evidence in medical research comes from individual RCTs, other controlled studies, quasi-experimental studies, and non-experimental studies such as comparative, correlation, and case control studies."
Studies like the ones cited here are defined as the second most reliable source for this kind of information in the guideline. If I went to the Natalizumab article to remove these sources, I'd most likely be in for a hard time. --JN466 21:22, 30 October 2010 (UTC)
That quote is in a section on assessing evidence rather than on clinical best practice (which is what Indications is about), it is also guidance on knowing study strengths rather than wholly on what sources to cite. If you read the nutshell and the lead, you find that medical (clinical) guidelines are given top billing along with systematic reviews (such as Cochrane). Neither the nutshell nor the lead encourage primary sources (i.e. directly citing studies, biased or otherwise). Perhaps MEDRS should clarify the sourcing for the Indications section of a medical article. Research is research. Practice is practice. There's a gap between the two that should not be filled by WP editors. Colin°Talk 21:48, 30 October 2010 (UTC)
JN, I don't think you'd get much opposition to re-writing that section. Io io editor hasn't been active for a long while, and the few experienced editors who have that page on their watchlist would probably appreciate it.
But what bothers me is the DUE issue here:
We currently present some details of study design (randomized, double-blind, placebo-controlled, etc.). These details, although IMO unnecessary, tell the (educated) reader something specific about this particular trial. That information is 'actionable', in that the reader can decide that this specific trial has virtues that others don't.
Adding "a study conducted by the drug manufacturer" isn't actionable in the same way. It smears the study by association -- Some drug manufacturers produce bad studies (more commonly, they simply fail to publish unfavorable ones), so all such studies need disclaimers -- but it doesn't tell you anything about the particular study in question.
As for what I'd do in this particular instance: I wouldn't tag it with a sentence that amounts to "yes, the only people on the planet with legal access to the molecule at the time were involved in this study". Instead, I'd chop it dramatically to say just "Natalizumab reduces relapses and slows disease progression in MS patients." WhatamIdoing (talk) 23:08, 30 October 2010 (UTC)
WhatamIdoing, I understand where you're coming from when you say "it smears the study by association", but it is an untenable argument when the ethics standards in scholarly and media publishing say such links should be disclosed. Newspapers and scholarly journals do not "smear" the authors of a study or a review by including this information. They include the information because they consider it indispensable for maintaining "the credibility of the journal, the authors, and of science itself". (The current version of the conflict-of-interest section of the Uniform Requirements for Manuscripts Submitted to Biomedical Journals is here, by the way.)
The Uniform Requirements also state that journal editors should "publish this information if they believe it is important in judging the manuscript." So if this information has in fact been published, as in the example above, it would tend to indicate that the journal's editor did consider it an important factor in judging the manuscript. This means it is actionable in the same way as the information that the study was randomised, placebo-controlled, etc. These are the ethics standards of our most reliable sources – JAMA, The Lancet, etc. – and we, as an encyclopedia, should boldly adopt them. --JN466 01:14, 31 October 2010 (UTC)
The ethics standards in scholarly and media publishing say such links should be disclosed in the original papers, not that they should be re-disclosed every time someone cites an original paper. Have you seen a scholarly journal that makes a habit of doing what you want to do here?
For example, the NEJM piece you cite above has a list of 29 references, and I didn't see any sentence in it that was qualified with a "funded by" statement. What you propose, in effect, is that when this paper says, "Current therapies for multiple sclerosis, including interferon beta and glatiramer acetate, are only moderately effective, reducing the annualized rate of relapse by about one third", that we should say the same thing, only tossing in "according to studies financed by drug manufacturer Serono" (who apparently funded at least part of the PRISMS study that is ref 16 in that source). So if journals aren't repeating these disclosures when they cite other people's papers, why should we go beyond their actual practice to do so? WhatamIdoing (talk) 03:23, 31 October 2010 (UTC)
I'll look into this, but note that the "gold standard" for media reports is to disclose such links. In terms of what we are doing, we are probably closer to being a media source than a scholarly source. --JN466 12:02, 31 October 2010 (UTC)
If we, like the media, were writing an article about the latest research study, then the funding of that study would be important and notable information to present. But we tend to write articles about diseases and how they are prevented or treated. Neither editors nor our readers are in a position to judge whether one aspect of the research (like funding) affects how it should be judged. This is why we base articles on secondary sources, and build our text on the analysis and judgement of experts. Colin°Talk 13:08, 31 October 2010 (UTC)
Here are some medical books that mention study funding, in a non-obtrusive but matter-of-fact way: "This was a randomized double-blind study sponsored by the drug's manufacturer". Similarly: [9][10][11][12][13][14][15][16][17][18]. There is of course a huge amount of literature about the controversy as such: [19][20][21][22][23] --JN466 13:21, 31 October 2010 (UTC)
Similar examples in google scholar: [24]; google news: [25] These are simple, matter-of-course mentions that do not cast aspersions against the authors. --JN466 17:05, 31 October 2010 (UTC)
We've got a mismatch on scale here. So look at your first source: In the entire book, with dozens or perhaps hundreds of references, they appear to name the sponsor for exactly one study -- a study that is described in rather more detail than any encyclopedia article should. I find the same is true for the second book, and I suspect that the same is true for all of them.
Identifying funding sources is clearly the exception, not the rule. Scholarly sources mention funding far less often than they mention study design, sample size, control selection, randomization, etc. (all of which are details that we normally skip entirely). Why should we "always" or "normally" do what the industry "rarely" or "never" does? Why should funding, which the scholarly sources usually skip, be included, when study design, which the scholarly sources frequently identify, is often considered too detailed to bother mentioning? WhatamIdoing (talk) 21:17, 31 October 2010 (UTC)
Concur with many of the comments/concerns already listed. While funding does factor into how to evaluate any article, so many other factors do as well (reputation of journal, reputation of author, blinding, appropriate statistical analysis choice, n sufficiently powered, numbers lost to follow-up, pre-specifying primary/secondary end points, etc - the list goes on). Not sure why we should elevate this particular component above others to point out. Yobol (talk) 03:13, 2 November 2010 (UTC)

Template:Cite journal

Template:Cite journal could be modified to include optional information about funding of the research the journal article is based on. This is rather low key, but seems appropriate if the information is known. Fred Talk 20:29, 30 October 2010 (UTC)

It's not necessarily a bad idea, but the disclosure list for a multicenter study involving a pre-approval drug could take half a page. I'm also concerned about promotional efforts: "This study was funded by <name of charity here>." WhatamIdoing (talk) 23:10, 30 October 2010 (UTC)
An appropriate summary would need to be used in that case. Fred Talk 00:59, 31 October 2010 (UTC)
It's a good idea. It is rather low-key, in that the average reader may not notice it. But it might be enough in some cases, and the presence of the field would help build editor awareness. --JN466 01:33, 31 October 2010 (UTC)
I'm a bit concerned by this. If we do this, we'd give funding sources way more weight than any comparable reference source. I am not aware of any commonly used citation format for medical journals which includes the authors' funding sources. I'm not saying that this information is unimportant - in fact, it's one of the first things I look at when critically evaluating a journal article - but I don't think the answer is to create a novel citation format to address the issue. And that's not even getting into the practical issues identified by WhatamIdoing. MastCell Talk 04:18, 31 October 2010 (UTC)
That is a valid point; we should not perhaps create a new citation format that doesn't have a precedent in RS. --JN466 17:25, 31 October 2010 (UTC)
I'm opposed to this for the same reasons given in the above discussion wrt doing so in article text + the issues raised by WhatamIdoing about the practicalities of it. Colin°Talk 07:37, 31 October 2010 (UTC)
this is a pervasive problem beyond Wikipedia's role. It affects essentially all of biomedical research. The question of sponsorship affects many other research studies as well. Furthermore, it is strongly counterproductive. The worst cases are the ones where funding is NOT declared. If the funding is declared, the reader of the actual article will see it; if the funding is not declared, then those papers will be cited here in such a way as to make them seem more reliable, when exactly the opposite is the case. Additionally, since in earlier years many fewer papers had funding declared, then this will look like it deprecates much of earlier research.
There are in any case equally important distinctions to be made affecting validity of published research; whether there is a truly adequate control and large enough sample can be even more important than sponsorship. so an the reputation of the investigator--for better or worse. We are not able to indicate such things in cite templates.
In some cases, it deserves to be deprecated. But it is not our role to decide that. The only people qualified to do that are those writing with professional credentials for journals with excellent peer review and editorship. Some few people at Wikipedia are at that level, but they know that the opinions will be respected on the basis of their authority only if published in a proper source, not here.
All we can do in disputed fields is cite proper studies that discuss the status of the work, and quote or paraphrase their key conclusions in the article. DGG ( talk ) 17:01, 31 October 2010 (UTC)
I think that concern about inadvertently elevating sources which don't disclose funding could be easily fixed by simply having that template field default to 'undeclared' or 'not declared', 'not published'. Any of those would fully equalize the playing field, and if anything, they would lean in the opposite direction, as 'undeclared' sounds a bit covert. Ocaasi (talk) 02:50, 2 November 2010 (UTC)
As far as I can make out, the Uniform Requirements for Manuscripts Submitted to Biomedical Journals, including its ethics guidelines, have been widely adopted by journals in the field. They have been in force for the past 10 years. I can't imagine there would be much recent reliably published research that would fail to declare conflicts of interest. --JN466 17:19, 31 October 2010 (UTC)
The degree to which those guidelines have been successfully implemented is somewhat dubious. Think Lancet and Andrew Wakefield. Or Claudia Henschke's studies of lung-cancer screening. Or, well, pretty much anything linked to the tobacco industry (see PMID 15791022). MastCell Talk 03:22, 2 November 2010 (UTC)
I would be against using such disclosure: if the source has been published in a reliable journal there has been a peer-review proccess who has decided that independently of the COI of the author the info of the article is reliable. Generally we are not the ones to decide if it invalidates the conclusions of the article, since we use secondary sources, although common sense would have to be used some times. In addition such disclosure could be used by quackers to censor any article they do not like and has a disclosure of some kind (I have already seen that kind of behavior in WP). --Garrondo (talk) 07:21, 2 November 2010 (UTC)
I would be strongly in support of a particular version of this. 1) Funding details in the template only; 2) 'No data available' is the default for studies which don't provide the info; 3) COI info cannot be used to determine reliability or to exclude a source, or even to comment on the source's potential POV, unless a third party reliable source makes such a connection. In other words, make the information available but don't allow editors to synthesize any new uses for it. Ocaasi (talk) 07:41, 2 November 2010 (UTC)
Garrondo, the conflict-of-interest policies of medical journals leave it to the discretion of the journal's editor to judge the relevance of disclosed conflict-of-interest information. The very fact that it is published along with the study following peer review indicates that the journal editor considered it ethically correct and necessary to publish it, after and in addition to the entire peer review process.
I agree with the point Ocaasi makes under 3) above; it is sound and of vital importance. --JN466 17:35, 4 November 2010 (UTC)
MastCell, are you aware that Wakefield published his study in 1998? The requirement for conflict-of-interest disclosure, as per the hatted document above, was introduced in 2001, three years after the Wakefield case. And are you aware that The Lancet officially withdrew Wakefield's paper in 2010, citing ethical problems, including the fact that he had not disclosed his conflict of interest? The change in the Uniform Requirements for Manuscripts Submitted to Biomedical Journals
  1. was made precisely to prevent recurrence of cases like Wakefield's, where published research is compromised by undisclosed conflicts of interest, and
  2. would have prevented the entire Wakefield fiasco if it had been in force in 1998, because Wakefield would have been required to disclose his conflict of interest to The Lancet.
This is an argument for these policies, not against them. (I don't know the background for Henschke, but her argument that CT screening saves lives appears to have been vindicated by study results published today.)
The PubMed document you cite argues that conflict-of-interest disclosure policies are perhaps not yet stringent enough, which if anything argues that we should at least reflect the ones that currently exist, rather than include no COI disclosures at all. --JN466 17:33, 4 November 2010 (UTC)
With Henschke, the question wasn't so much whether CT screening prevents lung cancer deaths (I suspect it does, although the evidence is nowhere near as clear-cut as one might assume by sampling the most recent Times coverage). It had to do with poorly disclosed conflicts of interest (see here).

While we're on the subject, though, the Times coverage is interesting for what it doesn't say. There was apparently a 20% relative risk reduction in death from lung cancer with CT screening, but the absolute risk reduction is nowhere to be found (not on the NCI's website, either). That may be in part because the result is an interim result, not a final one, but still. There were 53,000 participants. Assuming 1:1 randomization, that means there were 26,500 in each arm. So the lung-cancer mortality rate was 354/26500 (1.4%) with CT screening, and 442/26500 (1.7%) with chest-film screening. The absolute risk reduction with CT screening would therefore be 0.3%. Put another way, you would have to screen 333 people with CTs (as opposed to chest X-rays) to prevent one death from lung cancer. 332 of those people would have zero benefit, and possible (though unquantified) risk from increased radiation exposure and biopsies prompted by false-positive CT findings.

Of course, and as always, no one is willing to document the false-positive rate - how many people had unnecessary and invasive biopsies on the basis of incidental lesions picked up by CT scans? That's an important counterbalance, especially since the absolute benefit appears so small. If you randomize 53,000 patients, it's very easy to find statistically significant results. The clinical (real-world) significance of those results is not determined by the p-value, however. But I digress. MastCell Talk 19:32, 4 November 2010 (UTC)

Ocaasi (and others),
How exactly are you going to manage #3?
Consider the editor at ETS, who has done her own 'research' to determine that the surgeons publishing case series are paid to do surgeries (shock!) and was trying to exclude or disparage top-quality sources on that basis. (The studies don't "disclose" this so-called conflict of interest, and the Uniform Requirements does not consider "being employed" to be a reportable conflict of interest.)
So if it's already happening now, why do you think that increasing the prominence of this information (to a point well beyond what any scholarly source does) will do anything except increase this undesirable behavior?
(Keep in mind that nobody reads the directions, so "We'll say they're not allowed to do that!" is not a good enough plan.) WhatamIdoing (talk) 18:36, 4 November 2010 (UTC)
There's nothing I can find in Citing Medicine that discusses any mention of conflicts in citations. While there may be a good argument for doing so, it clearly would be leading rather than following the practices used by the ICMJE journals. The closest Citing Medicine gets to this is at example 75 for journal citations:

75. Other types of notes for journal articles

Harlow BL, Barbieri RL. Influence of education on risk of hysterectomy before age 45 years. Am J Epidemiol. 1999 Oct 15;150(8):843-7. Supported by a Public Health Service grant from the National Institute of Mental Health.

In fact, the contrary seems to be the practice even here. The journal identifies conflicts affecting the author(s) of the citing article, but not those of the articles cited within it. LeadSongDog come howl! 19:34, 4 November 2010 (UTC)

I'm increasingly at a loss of what any of this discussion has to do with Wikipedia. MEDRS says: "Reliable primary sources may occasionally be used with care as an adjunct to the secondary literature". It is the job of the secondary literature to review research and to make conclusions or recommendations, taking into account all factors including any funding bias. Our core facts should come from the secondary literature. Our readers should not have to worry that the single study we cite was funded by Bad Pharma Inc or by Good Hospital, Honesttown, or that the fifth researcher in the team once received £500 from Iffy Drugs plc. They should be reassured that we cite a review in a respected journal by a respected expert, or a clinical guideline by a respectable body of experts. Colin°Talk 19:41, 4 November 2010 (UTC)

Colin, this applies to review articles as much as it does to primary sources. [26]. Review articles also come with published conflict-of-interest information. Both review articles and individual studies are copiously cited in WP. The idea is simply that if that respected journal considers it ethically necessary to mention the funding sources for a particular review article, or individual trial, that we don't withhold that information from the reader, but report it along with all the other details, by just adding those three words: "sponsored by X". --JN466 20:21, 4 November 2010 (UTC)
Yes, but you're persistently missing the point. The journal requires disclosure of the author's COI in the author's paper itself. It does not require disclosure of the dozens of source's COIs when they are mentioned in other papers.
Perhaps this will help:
  • Mary Jones writes for Cochrane. The conclusion is that lab result foo is a seriously flawed marker for identifying disease bar. Mary discloses her COI: "I once worked for a company that did one of the 23 studies used for this analysis."
  • Tom Smith writes a paper for NEJM. Tom writes in his paper, "Don't rely on foo if you're trying to figure out if this patient has bar". Tom also discloses his COI: "I once did some paid consulting for the company selling the bar test." Tom does not disclose Mary's COI; he only discloses his own COI.
Do we agree that this is what actually happens in the real world, in reputable journals that follow current ethical standards? WhatamIdoing (talk) 21:05, 4 November 2010 (UTC)
No, we do not agree on this point. I gave plenty of examples of sources [27][28] that do discuss research in much the same way that we discuss it here, and that do mention who sponsored it. The gold standard for media coverage of medicine is to mention manufacturer and/or commentator involvement in research when discussing that research, per multiple sources provided. Quality media article do it, and those that fail to do so have been severely criticised for this failure. What I do accept is that this information is not mentioned each and every time that a study is described, or mentioned in passing, in a published source. On the other hand, if you look at recent Cochrane reviews, they will certainly mention who funded the studies included in the review (of course they also mention a wealth of other detail that goes way beyond what we would typically mention here), and will include industry funding in their assessment of potential bias (example). --JN466 22:57, 4 November 2010 (UTC)
But you've given examples of news stories on individual studies, not journal reviews on a disease or a therapy. We all agree that when studies are described in detail, which mainly occurs when they are recently published or press-released, funding is a relevant aspect to note. But, by and large, studies are and should not be described in detail on WP unless they are notable enough to deserve their own articles or a significant section in another article. It is the misuse of primary sources, and a good dollop of OR, that leads so many WP articles to be overfull of "a study showed that...". Encyclopaedias do not do that. We say aspirin treats a headache. We say valproate is a first line therapy for epilepsy. The research that showed these things to be true belongs in a history section, if it belongs anywhere. Colin°Talk 23:55, 4 November 2010 (UTC)
I gave several links above. One was to google news, the other one was to google scholar. I also linked to a listing of recent Cochrane reviews, and checked half a dozen of them. Each of them commented on the funding of the studies that were reviewed. I agree with you on primary sources in principle, but the fact is that policies and guidelines—including MEDRS—do not forbid citing primaries, that they are widely cited, and that editors not providing funding information when they cite primary sources exacerbates the problems. Lastly, literature reviews and meta-analyses may also be sponsored (example), and I think it is good practice and aids transparency to point that out where it is so. --JN466 00:12, 5 November 2010 (UTC)
Even the Google Scholar links are to news items or to pieces discussing research and funding issues in general. I wish people wouldn't say things like "policies and guidelines—including MEDRS—do not forbid citing primaries" as it misses the point. Both MEDRS and WP:NOR heavily restrict primary sources. MEDRS says they can only be used "as an adjunct to the secondary literature" and WP:NOR says "Wikipedia articles should be based on reliable, published secondary sources". Here's an analogy. My company's dress code is (for men) smart trousers (no jeans), smart shirt (with a collar), tie optional. So ties aren't forbidden but you can't go to work wearing just a tie. The fact that primary sources are widely cited is an WP:OTHERCRAPEXISTS argument and should be dealt with by replacing text citing better sources rather than worrying about what the primary source citation contains. Colin°Talk 08:28, 5 November 2010 (UTC)
Part of the problem here is that many editors are not aware that in some fields, like medicine, articles in peer-reviewed journals are primary sources. It is different in other fields; an article in a peer-reviewed literary journal analysing the work of a major writer is a secondary, not primary, source, for the article on that writer or their work. --JN466 13:22, 5 November 2010 (UTC)
It is not true that "articles in peer-reviewed journals are primary sources". These journals contain many types of article, such as editorials, research papers, literature reviews, meta-analysis, biography, obituary and letters. Also, within one article, the actual text one is drawing upon might be primary, secondary or tertiary material. Further, whether something is primary or secondary depends on what you are saying, not just the material itself. For example, most research papers contain background sections that are secondary sources for their facts. All publications are primary sources for the fact that authors of the text said what they said when they said it. Your literary example is similar to a literature review paper in a journal, though those tend to cover many studies, not just one. An editorial is a secondary source on the research or news item the editorial is commenting on, but a primary source on the opinions of the editors. All very messy and never as simple as some folk on policy talk pages would like it to be. Colin°Talk 15:44, 5 November 2010 (UTC)
I am familiar with the distinctions that you outline here, Colin, but had dispensed with the complicated detail which you have (correctly) added. However, I would disagree that a monograph in a peer-reviewed journal that presents original research on a literary figure's output is more alike to a [research] literature review than to an individual study presenting original biomedical research about measles. A paper presenting original research is a paper presenting original research, whether the subject is medicine or literature. The same is the case in sociology or anthropology: if an anthropologist travels to the Brazilian rainforest and publishes an anthropological study of an Amazonian tribe based on her original research, her study will be a valid secondary source for our article on that tribe, whereas a paper of original research on measles would be a primary source in our article on measles. No? --JN466 19:46, 5 November 2010 (UTC)
Our articles on Primary source and Secondary source highlight that different disciplines use somewhat different definitions and that the two kinds are relative rather than hard distinctions. What matters for MEDRS is that it is the first publication of new research that counts as primary literature, and any articles (or portions of articles) that draw on those published studies are secondary literature. Both the monograph and literature review examples are based on previous published works, so whatever they say about those works is secondary. Anyway, this is discussion isn't taking MEDRS forward. Colin°Talk 15:35, 6 November 2010 (UTC)
There's a simple approach to this that almost always works. Trust the indexers to make the determination of whether the article is a review. If Pubmed (or other indexing service) calls it a review, then we can treat it as a review. If it is in a journal that only publishes reviews, treat it as a review. Otherwise, seek concensusconsensus on the talkpage before doing so. LeadSongDog come howl! 20:46, 5 November 2010 (UTC)
That's consensus, not concensus. The thing is, there aren't review articles on every drug or treatment in existence. Where there aren't any, individual studies are the only sources editors have, and readers should know if the people conducting them were employed by the manufacturer. --JN466 14:07, 6 November 2010 (UTC)
Drugs don't get prescribed to patients when the only documentation on the drug is some piece of primary research. There will be documentation by a drug authority, often clinical guidelines by a professional or government body, and an entry in some drug formulary that the physician consults when writing his or her prescription. PubMed makes it easy to locate primary research papers online, but they are not the only source of information on drugs. As far as what we tell the readers about the research, it is clear we don't agree and if you want to put that info in an article then do so and debate its relevance on a case-by-case basis if need be (as WhatamIdoing notes below). Colin°Talk 15:35, 6 November 2010 (UTC)

Since we disagree on the basic facts, I don't think that further discussion is going to be worthwhile. I went to half a dozen random articles and found zero disclosures of someone else's COIs; Jayen goes to sources selected for disclosure and finds, out of dozens of sources, one or two disclosures. Jayen then apparently comes to the conclusion that since this is very rarely done by sources, then listing the disclosures for all 216 cited sources at Schizophrenia would improve the article. We aren't going to agree on that point.

Nothing in this guideline prohibits editors from mentioning funding if (according to their best editorial judgment) it appears to be a detail worth mentioning. We also don't prohibit them from mentioning other details, such as the sample size or study design -- or the phase of the moon, or the color of the book binding, or anything else -- if, according to their best editorial judgment, a given detail appears to be worth mentioning.

I see no reason to change the guideline to elevate funding above other details, and I think there is no hope of an agreement to do so. WhatamIdoing (talk) 01:13, 5 November 2010 (UTC)

I am not deaf to what you and Colin have been saying, and your critique of my search methods is valid. However, there has been a very substantial controversy about this general topic, and about the way it is handled in both scholarly and media publishing. Why should saying nothing about it be the best option, when out there, in the real world, the issue has attracted so much attention, has led to new policies in scholarly publishing, and to widespread criticism of science reporting?
Now, do you both accept that sources such as Cochrane Library reviews—first-rate sources, according to this guideline—do make a point of clarifying the funding of any studies included in the review? --JN466 13:45, 5 November 2010 (UTC)
Could you list some of the Cochrane papers that do this? Regardless, this is detail on a par with what search criteria the Cochrare paper authors used when they did their literature search. It is all necessary for the paper, but really it is only the conclusions that we are interested in. Colin°Talk 15:44, 5 November 2010 (UTC)
Cochrane reviews usually provide this information in the "Results" section where the studies included in the review are described (click on "· Results" under "The Review" on the left of your screen), and/or in pop-up frames like the "Methodological quality graph of included studies" or the "Characteristics of included studies". Here are some examples:
These reviews are all from the New Systematic Reviews page. --JN466 20:22, 5 November 2010 (UTC)
I'm afraid those reviews are just excellent examples of the sort of secondary source we should cite and how we can be satisfied that the experts have looked into the funding issues (they have documented that they have) and that that is one of the aspects they took into account when drawing their conclusions. So we cite the conclusions and say "there is insufficient evidence that..." or whatever. Colin°Talk 15:35, 6 November 2010 (UTC)

Funding source was significantly related to conclusions

Research showing "Funding source was significantly related to conclusions" Fred Talk 18:54, 1 November 2010 (UTC)

Doesn't surprise me. This is soft-drinks "research". Their possible causes #1 "Industrial sponsors may fund only those studies that they believe will present their products in a favorable light" seems highly likely. Newspapers publish news. To get a story, you need something new to say. One way, with little risk, is to do research where you already know the result. Publish a press release. Newspapers repeat verbatim. Free advertising. Colin°Talk 20:03, 1 November 2010 (UTC)
Yes, a milk processor who pays for research which shows consumption of milk by the elderly reduces the severity of osteoporosis has not necessarily influenced the researcher; they have only documented something predictable and produced a citable scientific journal article, assuming any journal would publish such commonly known information. Fred Talk 00:49, 2 November 2010 (UTC)

Request for assistance

Are there any editors here who would consider looking carefully at this discussion [31], the source, the source's editors, the current thread, and would consider making a comment based on that careful research. I'm happy to withdraw my support of this source but feel more input and closer scrutiny of the source would be helpful.(olive (talk) 21:40, 9 November 2010 (UTC))

Template

Do we want Template:UnreferencedMED?

If we do, then it should be listed on this page and have its grammar fixed. If not, then we should send it to WP:TFD. WhatamIdoing (talk) 03:57, 11 November 2010 (UTC)

Peer review for books

Regarding the above thread by Olive, the consensus at RSN seems to be that this book on integrative cardiology and published by McGraw-Hill Medical[32] and edited by the former president of the American Heart Association isn't a reliable source because it's not peer reviewed. It seems like if that's that standard for books, this guideline should stipulate that. TimidGuy (talk) 12:12, 10 November 2010 (UTC)

It depends on what the book is being used for. Claiming that it is a "medical textbook" is not appropriate. There was also issues of WP:DUE. It is not the best source for looking at research methods. --Doc James (talk · contribs · email) 17:38, 10 November 2010 (UTC)
My impression was that McGraw-Hill exclusively publishes textbooks, but I agree that that was misleading and possibly not accurate. I'd be fine with not calling it a textbook. Since no one has responded, does that mean that there's not a consensus that books and medical textbooks must be peer reviewed in order to be considered a reliable source? TimidGuy (talk) 12:08, 11 November 2010 (UTC)
My understanding is that the editorial oversight academic publishers apply is analogous to peer review. The problem may partly be due to the unfortunate wording in WP:V, "Where available, academic and peer-reviewed publications are usually the most reliable sources". Some editors seem to misconstrue this as meaning that these reliable sources ought to be both academic and peer-reviewed, which is not the meaning intended. There is no mention of peer review in regard to academic publishers in Wikipedia:Identifying_reliable_sources_(medicine)#Books, although there is mention of editorial oversight. --JN466 14:31, 11 November 2010 (UTC)

Policy/Guideline

See RFC at Wikipedia talk:Identifying reliable_sources#RfC on the relationship between the sourcing policies and guidelines.

An editor removed: "In the event of a contradiction between this page and our policies regarding sourcing and attribution, the policy takes priority and this guideline should be updated to reflect it," explaining that it was an 'editing instruction'. I replaced it noting that the hierarchy of policies was significant for properly applying the guideline.

This was removed again as 'litter' (not perhaps the word I would choose for attempting to restore a longstanding and explanatory note). I'm not fan of merely pointing to other instances, but sometimes other stuff exists for the same reason, and WP:IRS also includes that verbatim statement. I think removing it on grounds to clean up the page is not a bad idea, and I like a tighter policy; however, I didn't think the original grounds for removing were sufficient, particularly not if the purpose was to give MEDRS more weight than it actually has.

Content guidelines such as NOTABILITY, CITE, and RS are all widely used, but do not have the force of policy in the way that the core 3 do. Since there is sometimes confusion regarding which aspect of Wikipedia's documentation takes precedence , I think the explanation is not trivial and may be worth mentioning. Thoughts? Ocaasi (talk) 11:53, 11 November 2010 (UTC)

The page is already quite clearly marked as a guideline, so the text was redundant. SandyGeorgia (Talk) 11:56, 11 November 2010 (UTC)
Would you apply the same to WP:RS, or any other content guideline for that matter? Ocaasi (talk) 12:00, 11 November 2010 (UTC)
Do you think it would be helpful if the {policy} and {guideline} header templates made brief mention of their respective roles. For a while when I started editing I didn't know which was which, since they looked similar and were often used in the same contexts. Something short like, Policy has fewer exceptions. Guidelines are subsidiary to policies. (which I ripped from the WP:List of policies and guidelines talk archives). Ocaasi (talk) 12:17, 11 November 2010 (UTC)
If you think the distinction between guideline and policy isn't sufficiently clear, the place to addess that would be here: SandyGeorgia (Talk) 12:22, 11 November 2010 (UTC)
Do you think we should remove the same text from WP:IRS or other content guidelines? Also, what's your an opinion about whether a note such as "Policies lay out standards, while guidelines outline best practices for following those standards" or "Policy has fewer exceptions. Guidelines are subsidiary to policies", or something along those lines, would improve the {content guideline} template? Ocaasi (talk) 12:30, 11 November 2010 (UTC)
I think you should take those questions to the talk page of that template. SandyGeorgia (Talk) 12:31, 11 November 2010 (UTC)
Ok, I'm still curious if you have an opinion about it. Also, the more general question, do you think we should remove the text from WP:IRS as well? I'm trying to establish the broader rationale, so keeping the discussion where it started helps do that. Ocaasi (talk) 12:34, 11 November 2010 (UTC)
I don't mean to be ignoring your questions, I just don't have time today for them, and I think more qualified people can answer at the correct page. SandyGeorgia (Talk) 12:36, 11 November 2010 (UTC)
Gotcha, thx. Ocaasi (talk) 12:37, 11 November 2010 (UTC)
The text wasn't "longstanding"; it was added in July without discussion. All writing should be focused. Repeating policy text (such as that policy > guideline) throughout guideline body text is just poor practice. Should we also repeat that discussions on medical sourcing should be done assuming good faith and without personal attacks, to name just one other policy? Where would it end? If it is important enough that you feel all guideline pages should repeat this point, then (as Sandy says) discussion on the common guideline template is the place to go. Rather than casting aspertions that both editors were trying "to give MEDRS more weight than it actually has", perhaps you should consider why one editor thought it necessary to put MEDRS (and only MEDRS) in its place. Colin°Talk 13:11, 11 November 2010 (UTC)
Seems I will need to find the time to catch up here: maybe next week. SandyGeorgia (Talk) 13:22, 11 November 2010 (UTC)
1) Obviously, no slippery slope is intended or proposed. Policy pointers are helpful where there is likely to be confusion about them. Assuming good faith applies equally everywhere, but the respective roles of NPOV/V/NOR/RS/MEDRS is more varied and frankly, confusing, especially to newer editors.
2) The text was added here but hit has been a part of WP:RS for longer. I don't think it necessarily must be here, but I think the two pages should basically follow the same practice; so, do you think it doesn't belong there or on other content guidelines as well?
3) I added back the text after QuackGuru removed it; he commented to me that my intention in doing so was to downgrade MEDRS. I took that as an implication that he was trying to bolster it--which wasn't a necessary conclusion, and I could have left that speculative mention out of my description above. So can you. I am not trying to put MEDRS in its place, unless you think an accurate statement of its position is somehow biased. As for why MEDRS and not others, well, for one, it's already on RS, and it popped up on my watchlist; that was the extent of the selectivity. Editing Chiropractic for the past few months has also introduced me thoroughly to MEDRS, as opposed to other content guidelines, hence my particular attention to it and the editors who frequently cite it. Still, I think all content guidelines should follow similar logic here, however that plays out. Ocaasi (talk) 14:57, 11 November 2010 (UTC)
Ocassi, you have been asked twice now to take this discussion to the proper page. You haven't done that, but you have canvassed others to come here, which still isn't the proper page. And I've seen the ANI thread now. Please take this discussion where it belongs, as there seems to be a WP:POINT in play here. SandyGeorgia (Talk) 16:13, 11 November 2010 (UTC)
Sandy, policies benefit from having some consistency in their presentation and logic. The fact that RS does include the statement makes this page a natural place to discuss how the various content guidelines handle it (or at least this one). I'm not sure to which AN/I you are referring, but I don't think either Chiropractic or NCAHF are particularly relevant. I was not actively editing MEDRS until a piece of it was removed. Now I have opened that edit for discussion, which is just what should happen. Here is as good a place for it as any, and if there's necessary spillover to be handled at a different page, I'll do that once this discussion has run its course. As for canvassing, SlimVirgin originally added the statement, both here and at RS, so I thought she should chime in. You don't think that was fair? Also, I forgot to mention or you forgot to mention that I canvassed Quackguru as well. Ocaasi (talk) 16:48, 11 November 2010 (UTC)

somewhat arbitrary break

I was asked to comment here. I think it's important to retain that text. Most of the content guidelines that are linked to the sourcing policies have words to that effect. It's to make sure newbies know there's a sourcing policy they can refer to directly, and that this page should reflect it. SlimVirgin talk|contribs 16:17, 11 November 2010 (UTC)

If most of the guidelines have it, then why isn't it just globalized into the template, for consistency, and to avoid page bloat? Do we add the other examples Colin gave as well? Do we repeat every policy on every guideline page? No, we don't repeat text across pages, lest we want to end up like the Mess at MOS. SandyGeorgia (Talk) 16:20, 11 November 2010 (UTC)
I have no idea about templates. I only know that RS had it for ages, and should still have it, and several others do too. It's important for newbies, and it's important to stop policy forks developing. Ocassi said someone here told him that policies and guidelines are equivalent, but that's not correct. The sourcing guidelines have to be consistent with the three core content policies, not the other way round. SlimVirgin talk|contribs 22:10, 11 November 2010 (UTC)
I'm not aware that our other guidelines repeat the policy>guideline text? WP:RS got that added in 27 November 2009, as Ocaasi noted, and the history shows that RS guideline had issues with conflict between guideline and policy. So there was motivation to remind people which makes the edit understandable but it doesn't make it correct. The software folk call this issue DRY. We have a whole section on conflict over at WP:POLICY. I agree with Sandy that if this should be done, it should be done in the template. But, IMO, it isn't needed. Colin°Talk 16:40, 11 November 2010 (UTC)
I don't think it could easily be done in a template, because not all guidelines have a policy that directly relates to them, as this one does. SlimVirgin talk|contribs 22:16, 11 November 2010 (UTC)
I'm agnostic on whether this happens in the content guideline template or in guidelines themselves. I do think that especially for the most commonly cited guidelines (RS, MEDRS, PLAGIARISM, FRINGE, AND CITE) that a brief clarification is useful. Colin, do you think that this kind of thing is not needed here, in a template, or on any content guideline? Ocaasi (talk) 17:17, 11 November 2010 (UTC)
(BTW: those are guidelines, not policies). That's what I'm saying. The added text is just lecturing. Anyone who needs lecturing on the most basic and fundamental points of WP should be pointed at WP:POLICY and the rest of us shouldn't have to suffer pointless reminders. How would you like it if I added "This policy may contain statements that may be confusing. If you find such statements, please open up a discussion on the talk page and debate it endlessly until everyone else grows weary." to the lead of WP:V. Just because something is true doesn't mean folk want or need to read it. Colin°Talk 19:21, 11 November 2010 (UTC)
Colin, it was obvious from the context that I meant guidelines; that's the whole reason we're having this conversation. I don't find the policy confusing; I find the relationship between commonly sited guidelines and their overarching policies to be inadequately explained where it will be most relevant, particularly for new editors. Again, I'm not suggesting we add junk to policies for no reason but for that reason. As for WP:V it has a prominent explanation of the role of the core 3 policies, while WP:NPOV and WP:NOR have the additional note that they are not superseded by other policies or by consensus. That is analogous to what is being proposed here. Apologize if you're weary, but removing content leads to discussions; that's Wiki for everyone, even new editors, even those who wrote the policy. Ocaasi (talk) 19:38, 11 November 2010 (UTC)
I think you've just misunderstood everything I just said. The thing in quotes is an example and was supposed to be mildly humorous. Ok, you think new editors need to be reminded of a bit of WP:POLICY in paragraph three of lots of guidelines. I don't. You are right, I'm not interested in debating the issue endlessly. We have a few people with different opinions and they're pretty stuck on them. Let's now go our separate ways rather than waste time discussing a sentence that adds nothing to WP because it is already part of policy. This isn't an issue of "removing content" though it might have seemed that way to you because you hadn't seen the history. Your edit ended up being a revert of a revert. The original text addition wasn't discussed, it was just added. There's no consensus for it. The only reason I see is one of "ignorant newbies must be told". Too much policy change/debate is made on the basis of someone's gut feeling that "people must know this". Colin°Talk 19:59, 11 November 2010 (UTC)
Alright... Given my mix-up, I couldn't tell if your example was humorous or, um, deprecating. Happens online, unfortunately. Three months is not long enough to be particularly meaningful, but it does seem like our various policies take at least inconsistent approaches to clarifying these kinds of issues. That discussion would be good to have, if you have any interest. Cheers, Ocaasi (talk) 20:06, 11 November 2010 (UTC)
Technically, Colin, Ocaasi is asking editors to be "reminded" of something that isn't in WP:POLICY. I left the relevant quotations from that policy on Ocaasi's talk page before seeing this long discussion. WhatamIdoing (talk) 20:52, 11 November 2010 (UTC)
Wow. You are right. I should have read it more closely. All the more reason to follow DRY: if policy changes, the little snippets of policy repeats scattered throughout the guidelines will be out of date. I think that has put a close on this particular issue. Ocaasi, wrt discussing improvements to policies to make them consistent, I have so little free time for WP that I'm trying to focus it on the few areas I enjoy or that I care deeply about. Colin°Talk 21:25, 11 November 2010 (UTC)
I'm glad whatamidoing pointed it out. I take a slightly different read from the situation, which is that this is still a BFM if a 6-month user with 4000 edits and a 5 year editor with 8 thousands edits can't keep things straight w/r/t policies and guidelines that any help for other editors trying to do so is worth looking into. Either way, policy needs to get simpler and clearer, and maybe merged-er. Otherwise, no need to waste time on what you aren't feeling interested in... Ocaasi (talk) 21:49, 11 November 2010 (UTC)
I didn't understand what was explained here, but policies and guidelines are not equivalent. And WP:POLICY was abandoned some time ago because it was being edit-warred over by newbies, so it's not a good place to take any guidance from. But Wikipedians are clear that policies almost always have to be adhered to (and there are never exceptions to V, NOR, NPOV, or BLP), whereas guidelines incline without necessitating, to steal a phrase from Leibniz. SlimVirgin talk|contribs 22:19, 11 November 2010 (UTC)
SV, you've made me break my usual rule but such disregard for WP:CONSENSUS and policy is shameful. WP:POLICY is marked as a policy page. If you don't like it or consider it "abandoned" then go fix that bit of WP and come back here. Then, and only then, you can try to establish a consensus for your additional text. Until, and only until, you have established such consensus, I suggest you refrain from editing this page. Colin°Talk 22:41, 11 November 2010 (UTC)
Colin, don't start, please, and I have no idea what your usual rule is. It's a standard disclaimer on guidelines closely related to core policies. It's on RS. I think it's also on CITE. The point of the disclaimer is not to allow sourcing forks to develop, for obvious reasons. Policy is as a matter of fact enforced more rigorously than guidelines; and as they have to be descriptive, as well as prescriptive, what matters is what Wikipedians practice, and we do enforce the sourcing policies without exception. We don't enforce the related guidelines in the same way. SlimVirgin talk|contribs 22:51, 11 November 2010 (UTC)
Colin, somewhat ironically, if policy has changed to no longer put Policy ahead of Guidelines or establish a clear hierarchy among discrepancies, conflicts are supposed to be simultaneously and equally resolved among the various pages ("When apparent discrepancies arise between pages, editors at all the affected pages should discuss how they can most accurately represent the community's current position, and correct all of the pages to reflect the community's view.") So either the Policy policy didn't change, in which case the statement-at-issue is still binding, or the Policy policy did change, and now we have to sort it out with no interim guidance or default position. That doesn't resolve the issue of whether it should be on the page, but it does mix things up a bit again. Ocaasi (talk) 23:19, 11 November 2010 (UTC)
Ocaasi, am I obtuse or are you? Can you please explain why this discussion is still going on, when it can clearly be solved in the template? Your persistence in this matter, when there is an easy solution, makes no sense. SandyGeorgia (Talk) 23:24, 11 November 2010 (UTC)
I would like to think the answer is neither of us, and just blame the 86 policies and 235 guidelines for being obtuse. Your solution is very reasonable, but it will have to meet the objections of editors on each of the content guidelines pages, as well as the meta-policy/guideline pages. On this page Colin has raised a specific objection to the inclusion of this material in any form, template or otherwise, and WhatamIdoing has raised an objection to the hierarchy itself. So we can address that here, see if it comes to a resolution, and if necessary continue the discussion in a more general or aptly focused place, or we can do exactly the same thing on another page. I guess I don't see the benefit from moving the discussion at this point. Ocaasi (talk) 23:47, 11 November 2010 (UTC)

break 2

Perhaps someone could explain what the objection to it is. All it says is that the policies take precedence, and that this page should be updated if inconsistency develops. That means nothing can be recommended here that would violate NPOV, NOR, or V, and I think that would have strong consensus wikiwide, because they are the core policies and are strongly supported. What is the objection to pointing that out to new editors? SlimVirgin talk|contribs 22:55, 11 November 2010 (UTC)

  1. It's unnecessary, redundant WP:CREEPiness.
  2. If the guideline reflects the community's view, and the policy page doesn't, then it's the policy page that needs fixing, not the guideline. The Real Policy™ is what the community does and believes, not the words on a page with a {{policy}} template at the top.
  3. The bit about educating newbies is specious, because there aren't any (identified) conflicts between MEDRS and any content policy. Telling them what to do in a circumstance they haven't encountered (and aren't likely to) is silly. WhatamIdoing (talk) 23:12, 11 November 2010 (UTC)
Why is this discussion occurring here when there is already a template that is used across all guidelines, and wording can be tweaked there, in one place, to avoid repeating it across multiple pages? Is there a hidden agenda I'm missing, since the solution is so apparently simple? SandyGeorgia (Talk) 23:14, 11 November 2010 (UTC)
Because not all guidelines have such a direct relationship with certain policies, so the same words on a central template wouldn't necessarily work. SlimVirgin talk|contribs 23:48, 11 November 2010 (UTC)
Frankly, MEDRS is slightly better attended than the content guideline template talk page, this is where the issue originated, and it's important to get input from the people who will see this affect them. I think they are more here than there, although it's a somewhat technical difference. Now, however, we have an actual dispute about Policy, not just about whether or not to cite it, and it has to happen somewhere, and it has already begun in depth here. We can move it, but it seems about as apt to have it on any content guideline page (perhaps with a centralized discussion note) as it would be to have it at WP:Policy or the template page or elsewhere. What's the big deal about continuing it where it started? Ocaasi (talk) 23:25, 11 November 2010 (UTC)
As this effects CITE and IRS too (and probably others that I can't offhand think of), I've posted an RfC here. SlimVirgin talk|contribs 23:48, 11 November 2010 (UTC)
What's the big deal? The waste of time. I don't know if you value yours, but I do value mine. SandyGeorgia (Talk) 23:52, 11 November 2010 (UTC)
Then why the argument about one sentence that's in IRS and CITE anyway? SlimVirgin talk|contribs 23:55, 11 November 2010 (UTC)
The same mess across all of MOS-- duplicate wording that no one keeps up with, sometimes contradictory. Put it in one place-- we have a template. SandyGeorgia (Talk) 00:01, 12 November 2010 (UTC)

Let's summarise since this is now an RFC. The text added by SV was:

"In the event of inconsistencies between this page and the policies, the policies take priority, and this page should be updated to reflect them."

This is similar to text added by SV to WP:RS in November 2009 and WP:CITE in January 2010 (so the "in IRS and CITE" argument isn't helpful to that case). I agree with WhatamIdoing and Sandy. In summary, two aspects of the added text are:

  1. Policy > guideline. This is non-contentious and enshrined in policy (WP:POLICY) but doesn't need repeating in the third paragraph of the lead of this guideline or in the third paragraph of the lead of any guideline. If the community decides we need reminded of this on a regular basis then add it to the template for all guidelines. Perhaps we should be reminded of WP:CONSENSUS on a regular basis too because some editors here choose to ignore it. Maybe I should stick that in the third paragraph of the lead of WP:V?
  2. If policy ≠ guideline then guideline should be fixed. This is contentious and currently not supported by policy at WP:POLICY. The added text repeated this against-policy advice. Surely common sense should be used to work out what needs fixed, and really, whatever mechanism the community decides by which the two come back into agreement should be, you know, be documented in a procedural policy like, em, WP:POLICY. Since both guideline and policy pages can be changed by any editor at any time, I don't really see how one can assume the policy page gets it right 100% of the time. Anyway, this is an argument for another page, not this one.

There should be no agenda here. Nobody should be trying to "put MEDRS in its place" by reminding folk of basic rules, nor should anyone think that the absence of the text allows MEDRS to supersede policy. It just simply doesn't belong here (per DRY) and the issues raised should be discussed and resolved on policy pages elsewhere. Colin°Talk 00:27, 12 November 2010 (UTC)

NOTE: I thought the RFC was directed here but it appears to be at WT:IRS so suggest it continue there... Colin°Talk 00:31, 12 November 2010 (UTC)

Break 3

I think this clarifies the text. QuackGuru (talk) 22:43, 14 November 2010 (UTC)

Resolved: To mention "complementary and alternative medicine" on this page

I propose that we add at least the words "complementary and alternative medicine" to this page. Here are my main reasons:

  1. Readers looking for information about CAM topics deserve articles just as good as readers looking for information about non-CAM topics. CAM is not the ugly stepsister whose articles should be written from the weakest sources technically permissible under WP:V just because it's not "real" medicine.
  2. We regularly get questions about whether this advice should be applied to CAM pages, so we have evidence of confusion (i.e., directly answering the question would not be WP:Instruction creep).
  3. The line between CAM and non-CAM subjects is sometimes hard to find. We don't want editors to choose sources based on their personal belief that Self care, Massage, etc. are/aren't CAM subjects. That way lies edit warring, content forks, bias, and original research.

I'm open to various ways of addressing the issue. On the minimalist side, we could simply inserting the words into the lead; a maximalist position might add a separate section discussing the existence of high-quality sources (a complete surprise to some CAM editors) and the challenges of finding excellent sources for less popular topics.

What do you think? WhatamIdoing (talk) 20:16, 12 November 2010 (UTC)

Hi. Sorry for butting in. I followed a link on Colin's talk page. I'd prefer to see something like "this guideline applies to all health-related content." That wording would include CAM, evidence based medicine, and even content that isn't strictly medical, but might impact health; and I fully endorse Colin's observation that wrong advice that impacts health is potentially far more fraught - morally and legally - than a flawed BLP. Anthony (talk) 17:55, 13 November 2010 (UTC)
BTW: I think Anthony is confusing me with SandyGeorgia wrt that observation, but I agree it is important. Colin°Talk 18:26, 14 November 2010 (UTC)
Agree there is not to type of science just one. And just because normal science is unable to support someones POV does not mean that they get to make up how the wish to do the research or decrease the bar. Doc James (talk · contribs · email) 20:34, 13 November 2010 (UTC)
The phrase "complementary and alternative medicine" itself is problematic. While "complementary" medicine can and should be evidence based, "alternative" cannot. Conflating the two in the convenient term "CAM" is not really helpful. We can rationally discuss requirement for high-quality scientific sources for complementary medicine articles. To do so for alternative medicine makes no sense. LeadSongDog come howl! 05:56, 14 November 2010 (UTC)
The conflation happens to be deliberate, created by those who advocate the use of "alternative" methods in a "complementary" fashion. The methods are exactly the same. In practice, as long as an "alternative" method, no matter how ridiculous and absurd, is used in conjunction with mainstream medicine, it's classified as "complementary". The fact that the most famous promoters of "complementary/integrative" medicine claim the methods are more scientifically justified doesn't make it so. Scientific testing of exactly these methods produces the same inconclusive or negative results, regardless of what one calls them. The NCCAM has, without success, used over ten years trying to come up with positive results to justify continued funding (which isn't a scientific manner of working....). In spite of using the best methods available to them, they have gotten negative results in almost all studies conducted over a ten year period at a cost of $2.5 billion.[33][34] -- Brangifer (talk) 20:39, 14 November 2010 (UTC)
Um, LeadSongDog, could you try that again? You've just said—inadvertently, I'm sure—that high-quality scientific sources shouldn't be used in articles like Homeopathy, Chromotherapy, and Gerson diet. I personally think it makes an enormous amount of sense to use the best-quality sources in these articles. "High-quality" and "says that shining a yellow light on someone in a diabetic coma is helpful" are not the same. WhatamIdoing (talk) 22:40, 14 November 2010 (UTC)
Wasn't I clear? If there were reliable sources on such subjects, they wouldn't be alternative, they'd be either "evidence-based" or "disproved". But there are no high quality science based sources. For inert interventions it is essentially impossible to reliably distinguish "no effect beyond placebo" from "miniscule effect beyond placebo" in a finite-sized study, so there in fact may never be such sources. The very absence of these sources is what makes the subject an "alternative medicine". So saying we must have such sources would effectively say we can't even address the subject. Instead, we should use the best sources available with editorial judgement to avoid endorsing what isn't already supported by such high quality science. We could not, for instance, rationally exclude a mention of "water memory" from homeopathy. While purest bunkum, it is a significant part of the modern doctrine discussed in the homeopath's literature.
In contrast, much of "complementary" could as readily be called "supportive". Sound nutrition is often not treated as fully "mainstream" medicine (check the hours allotted in a med school curriculum sometime), but it certainly has an extensive basis in scientific literature. Herbal medicine, ethnobotany, chiropractic, accupuncture, and even talk therapy are clearly grey areas, but few could still honestly dismiss them as purely "alternative". They are all academic disciplines with formal methodologies and an advancing, experimentally testable knowledge base, even though the way in which the experiments are constructed and (poorly) funded is such that precise, statistically significant results may be few and far between. LeadSongDog come howl! 19:11, 15 November 2010 (UTC)
No, there are high-quality, science-based sources about these things. What makes something like chromotherapy be "alternative" is the very fact that despite being disproven, there are a few crackpots who believe that you can reverse a diabetic coma by shining a yellow light on the patient. IMO articles like chromotherapy should be based on high-quality, scientific-based sources, not self-published tracts or purported testimonials on some crackpot's website. WhatamIdoing (talk) 19:42, 15 November 2010 (UTC)
Sorry, I'm not familiar with chromotherapy. But from what you say, that isn't even "alternative", it's "fringe" (and I don't mean the optical kind). If there are good wp:MEDRS sources saying that it doesn't work, we shouldn't be mealy-mouthed about it, we should say so and cite the sources. However, to describe "what" it is, there is a place for thinking more like the wikiprojects for religion or fiction. The article needs to clarify what the mythology is, not just say that it doesn't work in the real world. If we can source descriptions of the mythology in MEDRS of course that is preferable, but it is not essential. We can simply state "Chromotherapy texts by Smith and Jones assert that blah blah blah" without endorsing it, and elsewhere make it clear that the best scientific evidence shows it doesn't work and that its use is equally likely to cause cuddly puppies to erupt in spontaneous projectile rabies. LeadSongDog come howl! 21:07, 15 November 2010 (UTC)
  • I support this. A small section dealing with alternative medicine would be very helpful. do you have a specific suggestion for wording? --Ludwigs2 15:23, 14 November 2010 (UTC)
  • Comment: is our current scope of "appropriate for the medical and health-related aspects of all articles" not inclusive of CAM too? Colin°Talk 18:26, 14 November 2010 (UTC)
  • Yes. If it's not obvious enough, then make it crystal clear: "applies to any and all health-related details in all types of articles, regardless of whether they are considered mainstream or fringe", or something like that. -- Brangifer (talk) 18:32, 14 November 2010 (UTC)
I'm just wary of adding "and, yes, we mean you too" to the guideline. Colin°Talk 18:52, 14 November 2010 (UTC)
LMFAO! -- Brangifer (talk) 19:04, 14 November 2010 (UTC)
Now if it weren't for wp:FORUM we could give Colin mod points for humour. LeadSongDog come howl! 19:11, 15 November 2010 (UTC)

I favor the minimalist approach, if anything. Basically, the point is covered by the third paragraph: (This guideline supplements the general reliability of sources guideline at Identifying reliable sources with specific attention to sources appropriate for the medical and health-related aspects of all types of articles where there are biomedical assertions.) and the third paragraph of WP:MEDASSESS (Speculative proposals and early-stage research should not be cited in ways that suggest wide acceptance.). #Use independent sources is pretty much spot-on for the problems identified here, though it enjoins against overstating any proposal, not just those from alternative medicine practitioners. The idea of adding some advice to the sourcing section might have legs; I seem to recall something along these lines at WP:MED - would it be useful to develop there and link here? - 2/0 (cont.) 23:04, 15 November 2010 (UTC)

Proposal

Taking the minimalist approach, how about changing the third paragraph from "...with specific attention to sources appropriate for the medical and health-related aspects of all types of articles where there are biomedical assertions." to "...with specific attention to sources appropriate for the medical and health-related information in any type of article, including information about alternative medicine."

Ludwigs, perhaps you and I should someday write an essay on sourcing AltMed articles. WhatamIdoing (talk) 22:45, 24 November 2010 (UTC)

This would weaken MEDRS because it is not only for alternative medicine. Even parts of alternative medicine pages where there are no biomedical claims MEDRS does not apply. QuackGuru (talk) 01:03, 25 November 2010 (UTC)
I do not understand how this "weakens" the guideline or says that it only applies to AltMed articles. "Everything, including the kitchen sink" is not usually understood to mean "Everything, by which I mean the kitchen sink and nothing else."
Furthermore, IMO the "biomedical claims" is an inappropriate and GAMEable restriction. We do not want people to say, "I get to use this lousy self-published source, because I say that 'your pain will go away' is a 'spiritual' claim, not a 'biomedical' one." WhatamIdoing (talk) 04:07, 25 November 2010 (UTC)
The proposal says MEDRS is specific for alternative medicine articles which implies it is not for all biomedical assertions. Editors t:hink MEDRS only applies to alternative medicine articles at the V discussion page. Even adding a link to MEDRS with some context at V was deleted for no reason. QuackGuru (talk) 04:38, 25 November 2010 (UTC)
The proposal says nothing of the sort. It says 'any type of article, including information about AltMed'. "Including information about AltMed" means that information about AltMed is a member of the previously specified, larger set, "medical and health-related information in any type of article." "Including AltMed" does not mean "limited exclusively to AltMed".
No one at WT:V thinks the advice on this page applies to Homeopathy but not to Heart disease; if you have come to this conclusion, you may wish to read their comments more carefully. WhatamIdoing (talk) 05:24, 25 November 2010 (UTC)

Areas to improve

  1. "Respect Secondary sources" is restatement of WP:NOR, except it creates a weird new requirement that claims from primary sources have some kind of time limit, which is completely out of whack with our normal sourcing policies.
  2. "Summarize scientific consensus" completely conflicts with WP:RS/AC. We shouldn't be "assembling proof of consensus" from disparate primary sources. Academic consensus needs to be declared by (and generally sourced to) a secondary source.
  3. "Assess evidence quality" I see nothing wrong with reporting early stage research in a "treatment" section of an article, as long as it's presented as early stage research.
  4. "Use up-to-date evidence" This section is completely baffling. Use sources that are recent, but don't use recent sources, because that's recentism. Whatever this is trying to say, it's not saying it very well.
  5. "Use independent sources" this is another section that makes my head hurt. It seems to be a convoluted summary of WP:SELFPUB. It also misrepresents notability. "Notable opinions" doesn't mean anything within Wikipedia policy. Notability is our criteria for having a standalone article on a subject, ideas within an article are not required to be notable.
  6. "Popular press" I know this has been a subject of recent controversy, but it really does conflict with our other policies and guidelines. You can't just dismiss the mainstream press because they might oversimplify things. Reliability is always contextual, based on the claim being made.
  7. In general: There's too many poor summaries of other policies and guidelines here. A lot of this should be trimmed down so that editors can actually see what this guideline adds to the body of Wikipedia guidance. Right now it's really hard to extract what the purpose of this policy is because so much of it is redundant summaries. WP:POL advises us to avoid redundancy, maintain the scope, and to be concise. Right now this guideline meets none of that.

Gigs (talk) 19:21, 15 November 2010 (UTC)

Here's my take:

  1. The "weird new time limit" is perfectly compatible with all of our sourcing policies, since none of them favor a "first-come, first-served" approach. Weak sources (most primary sources) should normally be replaced by a strong sources (such as secondary sources).
  2. "Summarize scientific consensus" says nothing of the sort. The word "secondary" doesn't even appear anywhere in the section; we are not advocating that anyone avoid them. This section relates very closely to WP:GEVAL, which is a particular problem in some health-related articles.
  3. Early stage research has been a particular problem. The two common problems are NOR violations (It worked in isolated cells in a petri dish, so you should eat more ____!) and refspam (Please join this clinical trial!). Additionally, MEDMOS recommends that research be reported in its own section, ==Research directions==.
  4. I'm sorry that you're confused by the balancing act between avoiding outdated information and the health fad of the week. If you run into a specific problem where this applies, I'd be happy to help you sort it out.
  5. A simpler summary might be "Do not believe everything you read, and especially do not blindly trust the health claims made on a multi-level marketing website."
  6. This page both gives people advice on how not to screw up when they're using the popular press, and good reasons to seek out better sources. There simply is no sourcing policy that prefers weak sources over better sources.
  7. I think this page does an excellent job of showing how to apply the community's standard advice to this subject. I think it has done a remarkably good job of maintaining its scope. And while it is long, it is shorter than what some editors would like it to be. WhatamIdoing (talk) 19:53, 15 November 2010 (UTC)
If these issues aren't fixed, it's not going to remain a guideline for long. Your attitude that I need to "sort it out" instead of the guideline getting written in a clear and concise manner is completely unproductive. I can see that this guideline has kind of developed in a back room, out of view of the main community. I doubt it would stand up to wider scrutiny in the form it is in now. If you want to save it, then be a little more accepting of feedback. Gigs (talk) 20:12, 15 November 2010 (UTC)
Your comment about a "back room" guideline couldn't be more wrong. I'm confused why you are so offended by it. Many editors find it immensely helpful. Could you give some examples of articles where editors have been prevented from improving Wikipedia by the guidance given here. With diffs. Please don't make threats to destroy what other people have worked on and find useful. Colin°Talk 21:26, 15 November 2010 (UTC)
There is so much redundant information here, it's hard to see what the actual meat of this guideline is. This guideline has always been controversial, and was marked a guideline with the support of around 15 editors, along with some minor opposition. I'm not offended by it, I just think it's a very poor quality guideline, which does not necessarily reflect a wider consensus because of limited scrutiny. To be honest it reads more like an essay than a guideline. Gigs (talk) 21:38, 15 November 2010 (UTC)
There's always room for improvement, and there are parts that are perhaps too verbose. The "limited scrutiny" charge isn't fair. MEDRS is one of the most cited guidelines on WP and you can be sure that any time someone gets MEDRS mentioned as a reason why their edit wasn't accepted or was revised, they will have come here to read it. It has been around a while is and is pretty stable. Unlike some policy pages that are "copyedited" on a daily basis and whose talk pages are frequented by folk with nothing better to do that argue, this one is helpful even if it is a little rough. That's why I'm upset at language like "If you want to save it", because it seems like an attack to destroy something that might not be perfect but is very much useful, exists to improve Wikipedia, and helps to keep our professional standards high. Colin°Talk 21:51, 15 November 2010 (UTC)
Well keep in mind my response was specifically colored by Whatamidoing's reply where he seemed to dismiss all my feedback and imply that the guideline is perfect the way it is. I don't think any guideline can survive that kind of attitude, especially not one that is starting to draw increased scrutiny for being seemingly out of line with consensus. Gigs (talk) 22:09, 15 November 2010 (UTC)
There is no problem with the current wording. You seem to not understand this page or what you really want to do is rewrite this page to promote fringe views. QuackGuru (talk) 20:47, 15 November 2010 (UTC)
Fully agree with WhatamIdoing: much of the balance in med articles that we have now is thanks to this guideline, and I do not see that the comments above really apply and are a true problem.--Garrondo (talk) 21:03, 15 November 2010 (UTC)
Thank you, Gigs, for your feedback. I'll certainly try to find time to see if the page can be reworded to avoid confusion. Wrt "popular press", I'm unaware of the "recent controversy" (don't find enough time to keep up these days). If you'd be good enough to point me towards it on my talk page, that would help. MEDRS has never been in conflict with WP:V or indeed any policy, though some editors with a POV to push would like to paint it that way. WP:V has pretty much always said "Articles should be based on reliable, third-party, published sources with a reputation for fact-checking and accuracy." What MEDRS highlights is that although the quality press might do a reasonable job with current affairs, fashion, politics, celebrity gossip, the arts, etc, they are piss poor when it comes to medical facts. There's plenty evidence that the press simply regurgitate press releases from medical institutions, quacks and big pharma alike; that they have an inherent bias towards recent news (discoveries, studies) rather than the big picture; that they have absolutely no reputation for fact checking and are as gullible as is possible; and that they are pretty careless when it comes to accuracy. There are links in MEDRS to pages that give such evidence for those who need convincing. Note that I mention this in regard to medical "facts". There are some aspects of medicine the press can be used for, but it is usually pretty obvious those are in areas they are strong in, such as current affairs. Colin°Talk 21:05, 15 November 2010 (UTC)
Thanks for the reply. I think a good copy edit is in order, with an eye on conciseness. I don't think it stands in direct conflict with the spirit of any policy. It's the wording and some of the specific things where things get muddy. I think part of that is the hazard of redundantly summarizing other policies, and part of it might just be the need for some overall copy editing. I totally understand where the mainstream media section is coming from, and I agree that the mainstream press usually does a poor job at science reporting. At the same time, I think the current wording might go too far, depending on the scope of what it's applied to. Mainstream press coverage is important for analysis, the same as any secondary source. Our job is not to only report what insiders are saying about something. Gigs (talk) 21:20, 15 November 2010 (UTC)
These sort of discussions tend not to go anywhere without specific examples. I'll only misinterpret what you mean by "Mainstream press coverage is important for analysis" for example. Remember that many people get hot under the collar about controversies in medicine but 90% of it is boring. Since the press don't tend to do investigative journalism any more, many press articles very much give only the insider POV about something. Is the best drug for angina one that can be prescribed for pennies and has been around for a decade, or a drug newly launched by the drug company to replace their out-of-patent one? The newspaper will repeat the press release that tells you how wonderful the new drug is. An independent report by a body producing clinical guidelines will perhaps tell a different story. When folk take issue with MEDRS's stance on the popular press (for example, to stand up to Big Pharma and all those nasty doctors in their pay), I think they often don't realise how bad it would be if we didn't have these guidelines. Colin°Talk 21:51, 15 November 2010 (UTC)
Right now I don't think the guideline makes a very good case for its own existence. You make a much better case for it here than the guideline does in its own text. Gigs (talk) 22:09, 15 November 2010 (UTC)
Well I'm glad to see we're not mortal enemies but you must forgive WhatamIdoing for being defensive because your language here and at the other discussion was very much on the offensive. I will try to address your issues and I will also see if there is some language in the guideline that is needlessly upsetting folk when it really shouldn't. However, I'm always pretty short of time so I won't promise to be quick about it. For now, I don't see anything that is at odds with policy or prevents people improving Wikipedia. As long as those two things are OK, then we can take our time about the rest? Colin°Talk 22:25, 15 November 2010 (UTC)

Taking a step back

I have made an edit to address what should be the least controversial of my above concerns, that the guideline misrepresented the concept of Wikipedia notability. [35] Notability, in the Wikipedia sense of the word, doesn't have anything to do with article contents, it's the standard we use for whether a topic is suitable for a standalone article. Gigs (talk) 01:17, 16 November 2010 (UTC)

Fringe theories in many cases would not be notable to have their own article but neither to be mentioned in other articles. Changed.--Garrondo (talk) 07:21, 16 November 2010 (UTC)
I agree with the spirit of your change, but I don't think we should use the word notability in a non-wikipedian sense while linking to the notability guideline. How's this? [36] Gigs (talk) 14:54, 16 November 2010 (UTC)
Fine with me.--Garrondo (talk) 14:55, 16 November 2010 (UTC)

Templates

I discovered Template:MEDCN today; should it be merged with Template:MEDRS? WhatamIdoing (talk) 05:32, 27 November 2010 (UTC)

Good find... They seem related to me but not identical. I take {MEDCN} as analagous to {CN}, asking for a cite in the absence of one; however {MEDRS} questions whether a source already in use is qualified to address a medical claim. Unless you want to group the issues together and rewrite the text to mean 'general medicine-related problem here', I'm not sure why or how we'd consolidate them. Template bloat doesn't bother me as much as policy bloat... is there something that can be gained from tidying these up? Ocaasi (talk) 06:51, 27 November 2010 (UTC)
On second thought, if you are taking a broader reading of {MEDCN} to include situations where there is no citation but where a specifically medical one would be required, then you could rewrite the two with "reliable medical citation needed" or something like that. It covers both grounds; its only drawback is that is 'asserts' rather than 'asks' in the unreliable medical source? sense. So maybe, "reliable medical citation needed?" That's broad and diplomatic. For kicks, is it a 'citation' that's needed or a 'source'? Ocaasi (talk) 06:55, 27 November 2010 (UTC)
For kicks? A source that isn't cited might as well be original research (except for that nasty plagiarism thing). If it is cited, then you've got the source, or at least a way to find it. So perhaps "please cite a reliable medical source" instead?LeadSongDog come howl! 18:28, 1 December 2010 (UTC)

MEDRS with some context deleted from V policy

MEDRS is a guideline and has the support of the community but a few editors continue to object to having MEDRS with some context in V policy. QuackGuru (talk) 23:02, 5 December 2010 (UTC)

Strengthening the wording

Have encountered a few issues with disagreements over interpretations of WP:MEDMOS and wondering if we should clarify things. A few examples:

  1. A tertiary textbook is being used to contradict a review article here "A 2007 medical textbook on heart disease said that "TM has been shown to not only improve blood pressure but also the insulin resistance components of metabolic syndrome and cardiac autonomic nervous system tone."John Vogel, Rebecca Costello, and Mitchell Krucoff, Chapter 47 in Braunwald's Heart Disease: A Textbook of Cardiovascular Medicine, Peter Libbie, et al, eds, Saunders Elsevier, 2007, p. 1157"
  2. Here primary research is being used as supposedly no secondary research is available. This is in a topic ares in which pubmed lists nearly 11,000 review articles somehow a primary study of 30 people is notable [37]. There are a total of 90,000 primary studies.
  3. We have some people interpretation of the policy here [38] stating "To the extent that MEDRS is imposing "considerably more stringent" requirements than our official policies that enjoy widespread consensus, MEDRS is wrong and should be updated accordingly." and "MEDRS is clearly outdated and needs an urgent update and review by the community. SG, when I cite a primary source I *only* state what the source states." And here "You have no right to insist that every reliable source be a review article in a journal."

So...

  1. IMO primary studies should never be used when review are available as they are either 1) not notable 2) not WP:DUE 3) not sufficiently independent 4) not encyclopedic (we do not need large blocks of text from primary sources in our articles)
  2. When there is a disagreement between a secondary source and a primary source / medical textbook / tertiary source the secondary source should take precedence. We should not be refuting reviews with text books of questionable validity. Review articles are the primary reliable source along with statement by major organization. This needs to be enforced.

Doc James (talk · contribs · email) 04:08, 30 December 2010 (UTC)

  1. The distinction between primary research papers and any secondary review of the literature is vitally important to help with all policy-related sourcing matters. I'm not so sure the secondary/tertiary distinction is quite such an important one. The book mentioned above seems (from a quick Google) to be a respected book on heart disease[39][40] so I don't think we can classify it as an unreliable source. However, even at 2,000 pages, it is covering a big subject. There are bound to be some facts that reflect the author's idiosyncrasies and limited knowledge. Would you disagree that perhaps 95% of that book is accurate? A general source is likely to be weaker than a specific one and a book for a wide audience weaker than a book for specialists. For example, I didn't find any Neurology textbooks to contain anything of any value for ketogenic diet, but they'd be fine for some of our more general articles. But these are just rules of thumb and if you are looking for a "review > textbook" algorithm in order to win a fight at TM, then I don't think we can or should help.
  2. I sometimes wonder which bit of "Wikipedia articles should be based on reliable, published secondary sources" people have problems understanding? It is policy as well as repeated in this guideline. I think this particular dispute is adequately covered by our policy and guideline. IMO, articles on major medical disorders should not need to cite primary research papers at all, unless that research study is notable enough to warrant detailed coverage in and of itself (such as the history section, or a current research section).
  3. It is unfortunate that someone stated MEDRS was "considerably more stringent" than policy, for this is not true and just feeds the anti-MEDRS trolls. MEDRS is the application of policy to a topic area. If any editor starts questioning MEDRS as a guideline, just cite policy instead. In the above discussion, the editor fell into all sorts of NOR traps when attempting to use a primary research paper as a source.

This guideline needs to serve the needs of uncontentious topics as well as those attacked by POV pushers. It needs to be useful for specialist subjects as well as general ones. I would be opposed, for example, to saying medical articles should be based on reviews in journals rather than on books. I don't think you could write a comprehensive article on many topics with such a restriction. Colin°Talk 11:40, 30 December 2010 (UTC)

WRT books I am only saying that reviews should usually take precedence (as you found with the ketogenic diet). Books usually do not do systematic reviews for each question they cover and a lot of it is just based on the opinion ( typically expert ) of the author.
IMO all health claim should be based on reviews. Social, cultural, and historical content does not need guidelines as stringent.Doc James (talk · contribs · email) 14:03, 30 December 2010 (UTC)
Perhaps we could expand some text to discuss why recent reviews are preferable secondary sources to (some) books, but I understand Colin's points, your frustration, and think the real problem here is how excruciatingly slow and ineffective dispute resolution processes are on Wiki. Generally, the anti-MEDRS editors are tenditious, disruptive editors who don't (or won't) understand WP:DUE, WP:NPOV and WP:V and are trying to advance a POV with synthesis, undue weight and cherrypicking, but our current policies are clear, and this page must only support those core policies, not extend them.

On the general frustration of dealing with tenditious POV warriors considering Wiki's broken dispute resolution processes, I try to focus my limited time now on articles with higher page views, and remember never to wrestle with a pig in the mud, because you'll both get dirty, while the pig enjoys it, and they're generally superior at slinging mud. SandyGeorgia (Talk) 15:14, 30 December 2010 (UTC)

The point about ketogenic diet was that general neurology textbooks weren't much use there. Just as I wouldn't expect a general book on cardiology to be of much use dealing with a decidedly non-mainstream allegedly preventative therapy (TM). But I did use two specialist professional books (one on the history of epilepsy and one on the ketogenic diet). Most reviews aren't systematic either and systematic reviews are (in my limited experience) largely reserved for very specific and easily measurable medical questions. There are a few systematic reviews on the effectiveness of the ketogenic diet but they are old and haven't been updated to include the big RCT. If I were to write an article based only on systematic reviews of the KD, it would be a few sentences long and five years out-of-date.
Agree with Sandy's point: editing and reviewing articles that aren't a battleground is much more enjoyable and productive. Colin°Talk 17:26, 30 December 2010 (UTC)
I once wrote an article based on the one and only book on the topic that was available at the time (late 2006, Intrusive thoughts); some reviews are now available, but even those focus only on specific aspects related to OCD, and the book is still the broadest coverage available (at least the book was written by a well known and respected expert in the field). Books do have some usefulness, but as reviews become available, those sources should trump. If we try to strengthen MEDRS beyond only supporting and clarifying WP:V, we'll only end up in the mud with the pigs :) We have policies and dispute resolution to deal with POV warriors, but the process is just too darn slow and time consuming, zaps good editors, and you get dirty when trying to deal with those-- edit something fun when you tire of the mudslingers :) If you look at Schizophrenia as an example, poor Casliber darn near wore himself out dealing with the POV pushers, but they finally moved along, and only now can we really and finally fix that article. If you spend too much of your time in battleground topics, it may wear you down to the point of leaving, as so many other have; if you wait, sometimes the POV warriors are encouraged to move along or are dealt with by other editors, and then you can get some work done. SandyGeorgia (Talk) 18:09, 30 December 2010 (UTC)
Agree with both of you. However if Wikipedia ever wishes to be taken seriously by academia some of these issues do need to be addressed. I mostly edit non controversial stuff with the odd controversial one throw into the mix. We should not leave controversial topics completely to those who wish to advance fringe POVs or religious ideas using the language of science.Doc James (talk · contribs · email) 18:44, 30 December 2010 (UTC)
Textbooks aren't always tertiary. Quite a lot of their content amounts to a literature reviewss, which means they are properly classified as secondary sources. The distinction that concerns me more is between decent university-level textbooks and self-help garbage. They're both "books", but only one is really suitable for hard-core medical information.
Additionally, I'd almost always rather have a tertiary source than a primary one, so we probably have bigger fish to fry.
I think the odd situation with that textbook should be addressed from the perspective of DUE, rather than trying to discourage books altogether: If one author says X, and all the other sources say not-X, then the overall balance still not-X. We do that in every area, because even the most reliable types of sources come out with occasional outrageous errors.
Also—lousy reviews exist, too. I'd hate to reject accurate, widely accepted information from a top-quality textbook simply because some poorly conducted review was published in a weak journal. Being a "review" does not guarantee that it isn't garbage. I count eight "reviews" at PubMed by the notorious HIV denialist Peter Duesberg: surely we'd take some immunology text over his "reviews". WhatamIdoing (talk) 19:35, 1 January 2011 (UTC)

Fringe journals

I've become more involved as of late in looking carefully at how articles in alternative medicine are source. Predictably, many of these articles source alternative medicine and nutrition journals. While these journals are great for getting the opinions of alternative medicine proponents, they should not be used as evidence for efficacy of an alternative medicine nor should they be used as a criticism for a mainstream study (from, for example, JAMA or NEJoM). Wikipedia right now is ripe with this sort of thing.

It is unfortunate that there is nothing in the MEDRS guideline addressing this issue. I pen now a paragraph for possible inclusion in the Biomedical journals section:

Papers from a journal that are indexed under PubMed or with a DOI are not necessarily all of equal quality. In particular, journals which are written to support the alternative medicine or nutritionism communities must be carefully considered before including in an article. Generally, papers in such journals should not be considered as fully vetted as compared to papers found in prominent medical and scientific journals. These papers can be used to source opinions of members of the alternative medicine and nutritionism communities, but as these communities are academically marginal, studies that appear in such journals are not on par with studies appearing in mainstream journals. Studies appearing in alternative medicine and nutritionism journals should not be used as evidence of rigorous testing, research, or conclusive results. Papers published in such journals which criticize studies in mainstream journals should not be used as more than evidence that the particular authors (or, at the very most, the particular community) objects to the conclusions of the mainstream study.

jps (talk) 22:33, 30 November 2010 (UTC)

Much of your concerns are covered by WP:PARITY. Yobol (talk) 22:42, 30 November 2010 (UTC)
To some extent, but PARITY doesn't demarcate explicitly. I want it stated that Alternative Medicine and Nutritionism journals are not on par with medical and science journals. jps (talk) 22:52, 30 November 2010 (UTC)
I think that, as a matter of good editorial practice, we should already be evaluating sources in part upon their acceptance in the community of experts. This is inherent to the "reliability" of any source which claims academic credibility. Explicitly calling out alt-med and nutrition journals is going to be like waving a red flag in front of some segments of the community. It's sometimes worth doing so, but I don't think we need a special exception in this case.

The guideline already states that we should not cherry-pick or prioritize individual journal articles to create a narrative that runs counter to current mainstream expert thought. If editors are, as you suggest, mining obscure literature to "debunk" or refute the conclusions of more rigorous scholarly sources, then I think policy (and more importantly, best practices) already exist to correct that problem. MastCell Talk 23:14, 30 November 2010 (UTC)

This is covered to some extent in Wikipedia:Identifying reliable sources (medicine)#Use independent sources. QuackGuru (talk) 07:37, 1 December 2010 (UTC)

Adding the above paragraph to this guideline is the surest way to see it turned into an essay. Colin°Talk 12:55, 1 December 2010 (UTC)
While I personally agree with the content of the paragraph, I agree with both MastCell and Colin that it is both unnecessary and likely to do more harm than good for this guideline. Yobol (talk) 15:02, 1 December 2010 (UTC)
Yes, it can be a problem, but I don't think that AltMed is the only offender. It might be useful to add a generic sentence along the lines of "Consider the reputation of the journal; merely being indexed at PubMed is not an endorsement of the journal's quality" to mirror the "Guess what, just because it has a PMID doesn't mean that it's peer-reviewed!" paragraph that we already have.
Beyond that, what I've found most helpful is having the encyclopedia articles about the individual journals contain a section on their reputation. Medical Hypotheses is the best example (and also not AltMed), and perhaps whenever you have a reason to look up a journal, you'll consider adding whatever you find to its article. WhatamIdoing (talk) 20:37, 1 December 2010 (UTC)
Along those lines: The article PubMed could use some love, including perhaps a section that gives an overview of its contents (e.g., includes some journals that aren't peer-reviewed). WhatamIdoing (talk) 20:58, 1 December 2010 (UTC)
I have had what appear to be identical experiences with jps, so I undestand why he put in so much detail, likely all based on avoidin a repeat of problems due to some ambiguity created by brevity. However, Colin's remarks about brevity, above and in the section below, is also good. Perhaps Colin can suggest another very brief "cut to the chase" versoin of jps's proposal, and jps can add to this very brief abstract what he thinks is needed to avoid use of ambiguities inherent in brevity. PPdd (talk) 01:22, 8 February 2011 (UTC)

Please vote - A consensus vote as to whether to consider the journal Homeopathy an RS for physics, science, or medical conclusions

Spam
Resolved

A consensus vote as to whether to consider the journal Homeopathy an RS for phsyics, science, or medical conclusions is happening here[41]. PPdd (talk) 02:08, 3 February 2011 (UTC)

Vote here. -- Brangifer (talk) 05:01, 3 February 2011 (UTC)

Resolved

Just spam. Colin°Talk 19:27, 7 February 2011 (UTC)

Proposed addition on MEDRS "peer reviewed" by alternative medicine panels

Propose addition to MEDRS - Note: Discussion moved to this page from WP:RS by request

This seems to already be covered in this article, but it is apparently not worded well enough for some editors to understand.

An alternative medicine or alternative science journal (or book) that is peer reviewed by its practitioners can not be used as a reliable source for scientific or medical statements. These should be supported by articles (or books) that are peer reviewed by professional scientists or science based medical doctors. In some cases, however, the alternative journal or book can be used as a reliable source to describe beliefs or practices of the alternative medicine or science.

This was voted on here[42]. Some editors found the vote to be preposterously unnecessary. However, the unanimous consensus vote shut down the many interminal talk page discussions about using Homeopahty as RS for a physics conclusion (because it was "peer" reviewed), which occured in volume above the vote at talk:Homeopathy, and being able to reference has already stopped a possible edit war at talk:Anthroposophical medicine here[43]]. I propose adding the wording above to the WP:MEDRS article, as it will be usefull for . Please comment. PPdd (talk) 18:50, 6 February 2011 (UTC)

If you are thinking of adding this to WP:MEDRS, this discussion should take place at WP:MEDRS. Not here. Blueboar (talk) 13:05, 7 February 2011 (UTC)
Collapsed personal comments irrelevant to actual proposal
The above signature by Blueboar was faked by Ppdd, who moved this post here from a different location [44], apparently without notifying Blueboar. Of course it is ridiculous in this location, but that's not Blueboar's fault. Hans Adler 17:19, 7 February 2011 (UTC)
Um... Hans, I would not say that the my signature was "faked"... it was cut and pasted from WT:IRS. However, I did make that comment at WT:IRS (and I signed it there). OK, it might have been nice if PPdd had notified me that my comment was going to be copied over to this board, but I don't really see it as a big problem.... I note that my complete exchange is quoted in the section below. Blueboar (talk) 17:55, 7 February 2011 (UTC)
Telling someone to go to WT:MEDRS in a conversation on WT:MEDRS seems rather strange. Also, this was copied in a way that makes it hard for other editors coming here later to understand which part of the conversation was copied and which was first made here. I would not have reacted in this way if this had been the first instance. Hans Adler 18:33, 7 February 2011 (UTC)
Ppdd, do you have any plans to stop misrepresenting other editors at some point in the near future? That would make it more likely that we will be able to keep you as an editor.
The question whether repeated dilutions work in practice as one would expect them to work in theory is hardly typical for the intended scope of WP:MEDRS. It is a borderline case, and as such must be treated with common sense. To give a similar example unrelated to pseudoscience, we can of course write about counterfeit medicines being sold in Europe [45] without waiting for meta-studies of scientific studies related to this problem. Hans Adler 17:07, 7 February 2011 (UTC)

Anyone want to try that again, without all the hate? Colin°Talk 19:24, 7 February 2011 (UTC)

My proposal stands. I cannot help it that one isolated editor from pseudoscience pages keeps following me around and repeatdly accusing me of bad faith" and "misrepresentation", like here[46] and here[47]. Despite that one editors repeated public accusations of bad faith misrepresentation (false as I pointed out here[48], to which no apology was given) and "canvassing" here[49] (to which no apology was given, even though it is appropriate to notify relevant projects and talk pages that may be affected without an ability to know a discussion was going on).

MEDRS does not govern non-biomedical content

This article's specific application of the RS policy uses inclusive language (what it governs), but it needs a disclaimer because MEDRS does not govern non-biomedical content, so I have just added the following wording to make this point:

  • Note that all the other content is governed by the general reliable sources policy, not this MEDRS policy. ("This" added later.)

While experienced editors should understand this, they occasionally don't. This reveals itself in controversial situations where their very justified attempts to prevent fringe POV pushers from including nonsense leads them to demand MEDRS quality sources for ordinary information. They may even use this tactic against other experienced editors who share their pro-mainstream POV! That's an unfair "raising of the bar" for inclusion of ordinary, non-biomedical, details.

ALL information must still meet our RS standards, but it's a different (general) standard than the MEDRS (specific, higher, but limited) standard. Many controversial articles contain both biomedical information (governed by MEDRS) and information about related controversies and public reception. (NPOV requires the inclusion of such info.) The last two are specifically not governed by MEDRS. I hope that I make the point clearly and that this inclusion will help to prevent misuse of this policy. Please tweak as necessary, just so the point doesn't get lost. I welcome a discussion of this matter here. -- Brangifer (talk) 17:44, 1 January 2011 (UTC)

If I weren't aware of the problem, I might oppose this as redundant instruction creep—but sometimes we have to keep telling them, over and over, until they get it. You can also point these people at the "popular press" section, which specifically enumerates a few subjects that are not 'scientific facts and figures' and therefore should be sourced to more appropriate sources.
And I agree that it's irritating and an abuse of this guideline to have someone pound on the table and demand a peer-reviewed article in a reputable medical journal for (say) the date that a drug was first sold, the sales figures from last year, or the name of the drug manufacturer's current CEO. WhatamIdoing (talk) 18:50, 1 January 2011 (UTC)
Ok, but can we please be careful not to use language like "higher" or "raising of the bar" wrt MEDRS when compared to general guidelines or policy. "Articles should be based on reliable, third-party, published sources with a reputation for fact-checking and accuracy." (WP:V). This guideline is merely the application of that policy to medical facts (as well as a fair amount of WP:NOR and WP:NPOV for good measure). MEDRS is required because there are a great number of people who wish (through ignorance or otherwise) to use sources that fail this policy requirement for medical facts, or sources whose use will inevitably lead to text breaking some other key policy. We don't feel the need to have guidelines on drug company's CEOs (it is fairly easy to source that to the quality business press, say, or an published annual report) but the same standard applies. Colin°Talk 19:17, 1 January 2011 (UTC)
Don't worry, those words were just my opinion. "Raising the bar" was referring to logical fallacies and unfair tactics and needn't be addressed in this policy. -- Brangifer (talk) 21:30, 1 January 2011 (UTC)
I'm just sensitive because such language in a recent talk-page comment elsewhere provoked the usual "MEDRS must be demoted" rant. Also, be careful to get the distinction between policy and guideline correct. Don't want people thinking we regard MEDRS as policy now, do we :-) Colin°Talk 22:22, 1 January 2011 (UTC)
LOL! Dang, when's that going to happen? -- Brangifer (talk) 01:03, 2 January 2011 (UTC)
Colin, i;m sure with you on that; I regard it as important as policy. PPdd (talk) 20:11, 8 February 2011 (UTC)

A subtle (but very important) MEDRS question about quoting auth0ritatively named sources such as alternative science/medicine journals, to mislead about medical conclusions

A question came up at the acupuncture article that deserves a broader treatment here at MEDRS. The subtle question that came up at talk:Acupuncture here[50]. An alternative medicine journal with an authoritative sounding name was quoted in the acupuncture article in a way that misleads readers about "medical" conclusions in a way which meets MEDRS, but is not in the "spirit" of MEDRS -

"The American Academy of Medical Acupuncture (2004) states: 'In the United States, acupuncture has its greatest success and acceptance in the treatment of musculoskeletal pain.' "[51]

This sentence is techincally true, but it leaves an ordinary reader with the impression that a national "medical" academy has concluded acupuncture is successful in general, and has "great" success in treating pain. That is how I read it, until I investigated and found that the AAMA is a POV pushing organization that is not peer reviewed by independent medical practitioners. This seems to be an end run around MEDRS prohibitions about using alternative medicine journals as RS for medical conclusions. Does anyone have a suggested fix of MEDRS to handle this subtle problem of "misleading by quoting medical conclusions by alternative organizations that have assumed titles of authority "? PPdd (talk) 15:05, 7 February 2011 (UTC)

I don't actually find the above sentence misleading. Perhaps I'm getting cynical, but I read that as "greatest", meaning "could be 1% success and acceptance, while all other uses are 0% success and acceptance".
Please don't get hung up on peer review. It's not the only method of ensuring decent work. Most university textbooks aren't peer reviewed, for example, and we'd still want to be able to use them as sources for scientific information. Similarly, it's sometimes appropriate to take statistics from the WHO's website (not peer-reviewed), basic information about diseases from the CDC or NIH websites (not peer-reviewed), and information about drugs from prescriber's inserts (not peer-reviewed). WhatamIdoing (talk) 16:02, 7 February 2011 (UTC)
I agree about not getting hung up on peer review for inclusion of RS for medicine, by doctors. The problem is with POV editors citing in house "peer review" to justify inclusion as indicating in-house beliefs or practices, though not for medical conclusions, then quoting the otherwise NRS (with what appears to a reader new to the field as expert national medical opinion), then leaving the reader with an impression that a recognized national medical authority has done systematic reviews and reached a conclusion.
You probably were not misled by it because you have a familiarity with WP policies and practices, and you did not think AAMA was a national medical board stating a conclusion via systematic review. A standard strategy of alternative medical journals and organizations is to put "National" and "Medical" in their title, and "peer reviewed" in their website or publiscation, then using weasel wording in the article, all done to mislead. PPdd (talk)

Note: the following is a partial copy from WP talk:RS to begin to centralize the discussion. PPdd (talk) 17:11, 7 February 2011 (UTC)

...If you are thinking of adding this to WP:MEDRS, this discussion should take place at WP:MEDRS. Not here. Blueboar (talk) 13:05, 7 February 2011 (UTC)
I put it here (WP:RS) because it is about "alternative science journals" in general, in addition to alternative medicine journals. ~
Fair enough... Well, from a purely RS standpoint, I would agree with the statement.
As a further thought, when you get around to discussing it at MEDRS (and I expect your proposal will require some discussion) I would suggest leaving a note at both WT:FRINGE and WT:NPOV, pointing to that centralized discussion. The issue relates to multiple policies and guidelines, and a full discussion now, ending in a broad consensus now will help prevent potential conflicts later. Blueboar (talk) 14:58, 7 February 2011 (UTC)
I agree on cenralizing the discussion. I also agree that this will be a major decision for modifying WP:RS or WP:MEDRS (or WP:Fringe or WP:NPOV), since it will affect a huge number of edit war-prone articles (and likely stop a huge number of edit wars and interminable and multi-repeated talk page debates. I got accused of "canvassing" last time I posted a vote on a related topic at relevant talk pages, but I fully agree with the need for multiple notifications, and a centralized discussion. PPdd (talk) 17:11, 7 February 2011 (UTC)

Note:The above block was copied from part of a discussion at WP:RS, done to centralize the discussion. PPdd (talk) 17:11, 7 February 2011 (UTC)

  • Note: Notice of this discussion was posted at WP:RS, WP:Fringe, WP:NPOV, WP:WikipediaProjects related to this dicussion, and talk pages of articles which may be affected by this discussion. If you know of an article which might be significantly affected by the outcome of this discussion, please post notice at that articles talk page before consensus may be reached, not after, as the outcome might ultimately be cited on that talk page to affect the article. PPdd (talk) 17:45, 7 February 2011 (UTC)
There are multiple issues here... As I said at WP:IRS, the proposed addition is fine from a purely WP:IRS stand point... if you are going to discuss what adherents of alternative medicine (or science, or history) believe in an article, you have to be able to cite the sources where they publish their beliefs ... and alternative "journals" are often reliable source for such statements of belief. I call this "Situational reliability". IRS has long noted the distinction between using iffy sources for citing statements of opinion and using them to cite statements of fact. The key is to present the information in a way that makes it clear to the reader that what you are discussing is someone's (fringe) opinion, and not accepted fact.
However, situational reliability is only half of the equation... the other part involves WP:NPOV, and especially WP:Undue weight. Whether a specific opinion should be discussed in a specific article (and if so, how much it should be discussed) also depends on the situation. The same opinion may appropriately be given a lot of weight in one article, and appropriately be given no weight at all not in another. I think PPdd's proposal does not address this.
All too often POV warriors on both sides spend hours debating the black and white issue of whether a source is reliable or not... with both sides neglecting the issues of situational reliability and due weight. Can an alternative journal be reliable? Yes... but only in limited situations. Should we mention what alternative journals say? sometimes yes, sometimes no... that also depends on the situation, in this case the context of the article. WHEN you do mention them, however, it is vital to present the information as being an opinion, and not to present it as accepted fact. Blueboar (talk) 18:51, 7 February 2011 (UTC)

Please discuss this elsewhere. Colin°Talk 19:26, 7 February 2011 (UTC)

Proposal for addition needed to MEDRS

Resolved

Alternative medicine/science journals and practioners should not be quoted in the article in a way which may mislead the reader into thinking a mainstrem medical/scientific authority is making the statement. For example, "The American Academy of Medical (Name of alternative medicine), which treats for many maladies, has been found to have its greatest effectiveness in treating cancer", which might mislead a reader into thinking a national medical academy finds the alternative medicine effective, and particularly effective in treating cancer when in fact, it might not be effective at treating anyting. Similarly, "Nobel laureate in physics, John Doe, finds this alternative medicine to possibly have a scientific basis", when the nobel laureate is not in the field applicable to the alternative medicine, and has views on it outside the scientific mainstream. Such "authorities" may sometimes be quoted as stating that "Practitioner of this alternative medicine believe that... (or have these practies..."), if it is a major source for that alternative medicine, but never stating making an outright statement of effectivenes, which might mislead the reader.

Colin, this problem bears on multiple guidelines, of which MEDRS is the most relevant, and is where I propose putting the content (WP:Fringe might also be second for being appropriate, but is less strict than MEDRS). I have noticed various talk pages on which it bears, and talk page articles that will be affected by the outcome, to come here, in order not to have multiple parallel discussions. The issue mighe get more attention if it were not centralized (since watch page notifications may not occur for comments here, but I (and abother editor) felt a centralized discussion to be best. I felt the MEDRS guideline to be the best place to discuss, because, like it or not, readers may use Wiki for medical information, and it is critical that misleading titles and weasel words of alternative medicine "authorities" not appear in a way that might mislead a reader. PPdd (talk) 20:02, 7 February 2011 (UTC)
Don't mislead the reader. Don't appeal to authorities or celebrities who are talking outside their area of expertise. Why do we need 200 more words in MEDRS to say that? Colin°Talk 22:03, 7 February 2011 (UTC)
I don't think this is a good idea. Fringe opinions should be attributed to the person or organization that holds them, even if they've made up a respectable-sounding name for the organization. The fact that one Nobel laureate is an HIV denialist is important to some articles. Celebrity endorsements can be reported as facts. If the reader is so hopelessly stupid as to believe that an actress is a better source for medical advice than his or her own physician, then there's really very little we can do about that.
Furthermore, not everything published by an AltMed practitioner or in an AltMed journal is bad research. I chatted with someone once who was designing a double-blind, randomized clinical trial on whether people with a specific medical condition subjectively felt better with this type of aromatherapy vs that type. The experimental design was fairly robust. If she managed to enroll sufficient people in the trial to get a statistically significant outcome (which I doubt, because it was about ten times the number of people with that condition in her current client list, but I don't know the outcome), then it would have been a perfectly fine bit of research on that topic. I personally expect that the outcome would have shown the two equal, but if not, it would be incredibly arrogant—and, frankly, a serious scientific sin—for me to say, "Well, her data showed that more people said they felt better after doing Foo than after doing Bar, so obviously that means that it's all rubbish." I have seen and even occasionally cited decent work in AltMed journals. Data is data, even if the data is about AltMed.
All of this feels like trying to re-write the guideline to win a specific dispute. It might not be intended that way, but that's how it feels. WhatamIdoing (talk) 22:25, 7 February 2011 (UTC)
Colin's brevity is good, with a slight addition suggested -

"When a quote is not from a medical authority, don't mislead the reader with quotations by authoritatively named bodies. Instead use the content of the quote to make a statement about a field's beliefs and practices, if the quote comes from an authority on its own field's practices and beliefs. Don't appeal to authorities or celebrities who are talking outside their area of expertise."

I am trying to get something like this in a guidline so many editors can stop reading and responding to interminably repeated, massive blocks of discussions, with POV pushers trying to put some new quote along with the name of an entity that has an impressive sounding name ("American Medical...") into alternative medicine articles, so as to create the impression of scientific validation about efficacy, done in order to circumvent not being able to use the content itself under MEDRS. I added wording not in Colin's bevity to avoid future problems so this will be dispositive. PPdd (talk) 01:14, 8 February 2011 (UTC)
I wasn't actually proposing anything. All I captured was the "common sense" argument underlying the proposal. The problem with this proposal and the shortened form is that you are trying to rate all of alt-medicine's publications and practitioners at the WP:NOTRELIABLE level in WP:V. I don't think WP will buy that.
The two problems with the original article text you quoted were, IMO, that it wrongly attributed the opinion of one doctor writing an article in a journal as the opinion of an entire organisation (with an impressive sounding name). The question of whether the opinion of that doctor or even that organisation is worth noting or quoting is one of WP:WEIGHT. Do most reliable sources agree with that opinion? THe other problem is where we quote opinions as a way to get round stating facts that actually can't be supported by evidence, but this handled by WP:NPOV. Colin°Talk 09:03, 8 February 2011 (UTC)
If it was not a proposal, it would still be a good proposal. I checked the source in that one specific example, and you are correct. I fixed that article to cite as source the individual, not the science medicine sounding organization. But it could have been the national science based medical organization (they are not such, but they use a title like that) making the statement. My goal is to have a dispositive guideline to stop nonsence POV arguments in alt med and pseudoscience talk pages, which keept repeating in specific cases the general attempt to convey legitimacy by quoting from an aturhitatively named source, and argued it is NPOV since it is just quoting a major body's statement on its beliefs, and so not misleading. One editor here even said he did not find it misleading, but with my limited knowledge of which is and which is not a legitimate medical source by name (I am just a simple desert mathematician), I was misled at first reading. Maybe you are right that this discussion should be at WP:POV talk, however. PPdd (talk) 20:08, 8 February 2011 (UTC)
My point was that there's nothing MEDRS about the issues here. My "proposal" could be incorporated in some standard WP policy page. If you feel it would be helpful, then perhaps you should see if those words could be incorporated into some general policy or guideline. MEDRS isn't an anti-alt-med guideline. It is just about finding the best sources for medical facts and using them honestly. It isn't really going to solve the wider-world problems such as some minor organisation using a puffed-up name to sound more important than they are, or drug companies burying unfortunate research results, or editors trying to name-drop a notable person to make their argument sound more impressive. For what it is worth, even pro-western-medicine writers make that mistake (see Talk:MMR vaccine controversy#Bill Gates). When dealing with problematic issues, you have many weapons at your disposal. MEDRS is only one of them. Rather than trying to rubbish the journal the source comes from or the practitioner making the claim, it might be easier, say, to look at WP:WEIGHT to see if too much importance is being given to a minority viewpoint. WP:V prefers third-party sources. WP:OR can be used against an editor trying to extrapolate from some study of three mice to the general population. And so on. Colin°Talk 21:52, 8 February 2011 (UTC)
Point taken. I will bring it up at WP:POV, which is where I now think it is best dealt with, and mark this section "resolved". PPdd (talk) 21:25, 9 February 2011 (UTC)
Resolved

Using studies as sources

Some time ago we have a long and circular argument about the suitability of using primary sources for their "secondary content". Every trial report has an introduction and a discussion that is often cited from other content. It was argued that this was acceptable. I have just seen this study (Annals of Internal Medicine) that shows how selective these sections can be in citing previous research. JFW | T@lk 00:13, 4 January 2011 (UTC)

I agree we should stay away from the intros of primary sources. It is often not clear how much of the literature they have looked at. Doc James (talk · contribs · email) 22:59, 4 January 2011 (UTC)
How about this: At the end of Wikipedia:Identifying_reliable_sources_(medicine)#Respect_secondary_sources, we add a sentence that says:
"Original experimental reports frequently include a brief review of prior publications. While these sections are nominally secondary sources, they frequently present a biased subset of the literature,[footnote to study] and higher quality secondary sources should be selected whenever these better sources are reasonably available."
Will something like that do? WhatamIdoing (talk) 06:02, 7 January 2011 (UTC)
There's some overlap with the Biomedical journals section: "Research papers that describe original experiments are primary sources; although they normally contain previous-work sections that are secondary sources, these sections are typically less reliable than reviews". Can we merge the above comments into this? Also, I don't think the study found they were "biased", only that they were grossly incomplete. Also the study was focusing on whether the papers mentioned earlier versions of the same experiment (same question) rather than some of the more general comments about the disease/therapy/whatever made in such sections. How about modifying the existing sentence to say "less reliable and not nearly as comprehensive as dedicated literature reviews", citing the above paper. Colin°Talk 08:17, 7 January 2011 (UTC)
I finally got back to this. See this, and improve as needed. Do we want to footnote this study? WhatamIdoing (talk) 06:46, 11 February 2011 (UTC)
I've tweaked it. I know the "grossly incomplete" was a phrase I used above, but I think it applied (as far as the study was concerned) to their treatment of previous same-question experiements. In terms of other background detail that they cover, the study doesn't say, though one would assume they are going to be less complete because those sections are short. So I've dropped "grossly" and tweaked one other word. Colin°Talk 08:33, 11 February 2011 (UTC)

I had not noticed this section when I started a similar,but not identical, section below[52], in which peer reviewed incidental statements like "We could not do such and such in this study because setting bones without pain relievers is painful", as a "peer reviewed"-ish secondary-like or tertiary-like source for stating "setting bones without pain relievers is painful". Sometimes peer reviewers or editors do not seek to modify what they consider "obvious" or "common knowledge". PPdd (talk) 14:26, 11 February 2011 (UTC)

A situation when primary sources can be considered secondary sources

An issue regarding MEDRS has come up in the specific context of acupuncture, which may need a more generalized MEDRS guidance. In medically peer reviewed primary source articles studying efficacy of acupuncture for something, it was incidentally mentioned that acupuncture can be a “painful treatment”, or that “Japanese acupuncture is ‘far less painful’ than Chinese acupuncture”. The peer reviewers allowed the incidental remark, apparently because it was not disputable, as it is not disputable that setting a bone is painful. So the primary source for the study is actually a secondary source for its peer reviewed incidental comments. PPdd (talk) 23:27, 9 February 2011 (UTC)

MEDRS does not say that primary sources (e.g., original reports on experiments) are unreliable. We say that articles "should be based on" secondary sources, not 'articles must exclusively use' secondary sources. For myself, I promise that when I mean "use exclusively with zero exceptions", I will directly say "use exclusively with zero exceptions". I suspect that other editors who have followed this guideline for a long while have similar commitments to clarity.
Similarly, MEDRS does not say that review articles are the only acceptable type of source. Widely used textbooks, for example, are also excellent sources for many purposes (e.g., anatomy facts, whether something hurts, what something is). MEDRS recommends the use of several different types of sources, each with their own inherent strengths and limitations.
BTW, if you want to get extra eyes on a dispute, there's a note at the top of this talk page that suggests three other, much more heavily watched places to leave a note about it. WhatamIdoing (talk) 23:15, 9 February 2011 (UTC)
Thanks, I had not noticed what you point out at top. I modified the title and my comment per your remarks. The point I was making is that in peer reviewed and accepted incidental comments, the comment actually sums up scientific consensus, and is presumed as a given, rather than assumed as a hypothesis. PPdd (talk) 23:27, 9 February 2011 (UTC)
There needs to be clearer delineations on exactly WHICH cases a quality, peer-reviewed primary source is acceptable. (see acupuncture talk page, for example). What if a "primary source" summarizes the observations of one or more primary sources in its introduction section -- is it a primary or secondary source then? Why or why not? Jimhsu77479 (talk) 00:35, 14 February 2011 (UTC)
There are several essays that might help you with the broader concept. I was the primary author of WP:Party and person.
Beyond that, it's not really feasible to draw bright lines that apply in all cases. What works best for a huge, well-researched topic like Myocardial infarction is not available for rare diseases like ODDD. The best sources for articles that largely provide general information (Disease, Cancer, Heart disease) aren't the same as the best sources for obscure details.
One question to ask yourself is whether a source is the ideal type of source for the specific material (rather than a merely acceptable and perhaps handy type of source for the material at hand). I would probably want a textbook for the definition of disease; I would normally prefer a large meta-analysis or systematic review for statistics about the incidence of a given disease. WhatamIdoing (talk) 19:03, 19 February 2011 (UTC)

Appropriateness of some sources

Although medical guidelines should ideally come from from nationally or internationally reputable expert bodies, the following online sources often complement with information that cannot be found elsewhere, or at least otherwise would take much longer time to find or expensive to buy access to. Therefore, I'd like to know what is the general consensus on using them as sources.

eMedicine

According to their own page: "New eMedicine articles undergo several levels of physician peer review plus an additional review by a PharmD prior to publication. Sources for subsequent revisions of articles include the author, a physician or PharmD member of the editorial board, user-driven feedback, and systematically reviewed medical literature." Mikael Häggström (talk) 10:10, 19 February 2011 (UTC)

I don't know how stringent this is carried out in practice, but my own experience with their articles give the notion that it is as reliable as expert bodies. In my opinion, eMedicine can be used as a source without any major restrictions, but individual claims should be replaced with even more reliable sources when such are available. Still, a substantial advantage of the free online nature of eMedicine is an excellent verifiability, actually giving more reliability to an associated claim than a reference to an inaccessible journal article, which may in fact be totally [not in citation given]. Mikael Häggström (talk) 10:10, 19 February 2011 (UTC)

Merck Manuals

This is produces by a pharmaceutical company, Merck & Co., which can give rise to bias in the information provided. I found no general description on peer review and other quality checks on their articles, but individual articles are apparently written and reviewed by medical professionals. In my opinion, Merck Manuals should not be used for pharmaceutical information, but otherwise, I see no major problem in using it as a reference similarly to with eMedicine. Mikael Häggström (talk) 10:10, 19 February 2011 (UTC)

Much of what I said immediately above applies to both of these. These are decent sources for many purposes, especially basic information. They may not be the ideal source for any given bit of information, but they are acceptable sources—even for pharmaceutical information. WhatamIdoing (talk) 19:31, 19 February 2011 (UTC)

A plea

  • If you have come here to canvas readers for a !vote or discussion on some medical issue not specifically requiring a change to the text of this guideline, we already have a project talk page at WT:MED. Please use it.
  • If you want to discuss the reliability of specific sources, please go to Wikipedia:Reliable sources/Noticeboard.
  • If a discussion has occurred elsewhere you want to continue here, link to the discussion and summarise things here. Do not copy portions of that discussion here as it makes it very confusing where to respond to and people dislike getting their text fiddled with.
  • If you link to a discussion on another talk page using a section-anchor link, please don't change the section name afterwards. We're not mind readers.
  • If you have a proposal for adding or changing text in this guideline, please keep it civil. We are discussing a guideline, not editor behaviour.
  • If you are currently engaged in a heated discussion elsewhere and want MEDRS changed so you can win, please go away and come back in a week or two.
  • If your section header is long enough to be entered for the Booker Prize, please shorten it. A handful of words is plenty.

Colin°Talk 19:19, 7 February 2011 (UTC)

  • I posted here at MEDRS because there seemed to be a general, not specific issue consistent with MEDRS' spirit of not putting things in articles that might mislead WP users about medicine.
  • I posted notice to come here at relevant projects, which is the purpose of projects.
  • I posted at some (not all) talk pages that might be affected by the outcome of the discussion here, and I was accused of canvassing (again) for doing so, but this is the opposite of canvassing, because the project members share my POV, and failing to alert article talk pages where there are anti-POV editors, when the outcome will affect them, would not be fair (equivalent to hiding the discussion by failing to notify).
  • Thank you for pointing out about not changing section headers.
  • I have been civil, and responded only to the content of repeated accusations of bad faith, canvassing, "nonsense" (all of the other editors who participated did not find my words to be "nonsense", the consensus vote closed down interminable WP:make a point discussions) and misrepresentation by a single editor who is following me aroung (and who did not apologize for his false accusations as at the top of here[53], ad here[54], which are not even the article where the general issue came up.
  • There is no heated discussion anywhere that I am aware of.
  • Regarding the length of my section header, I apologize, but I just drank coffee from this mug], so you are likely right on point on that one. PPdd (talk) 20:38, 7 February 2011 (UTC)
  • This is not a project page. It is a Wikipedia guideline. We discuss here how to improve the guideline text.
  • The "An alternative medicine or alternative science journal ..." proposal was not civil. If you can't see what is (still) wrong with it, then I can't help you. Colin°Talk 21:42, 7 February 2011 (UTC)
What are you referring to as cuncivel? Since there are only two references in your quote, I guess use of "alternative medicine? In an article that starts off in the lede first sentence describing the practice as "alternative medicine", what is uncivil about this guideline for such articles? This makes no sense. PPdd (talk) 23:34, 7 February 2011 (UTC)
  • Please don't make it worse by posting links to behavioural issues. This is not an RfC on other users.
  • I'm finding it hard to see any cool discussions at the pages linked to.
Please take a break. Colin°Talk 21:42, 7 February 2011 (UTC)
What is a "cool discussion, and which page linked to are you referring to? PPdd (talk) 23:34, 7 February 2011 (UTC)
PPdd, here is some more free advice: When I stop explaining things to you after comments such as "Hans is misstating facts", then it is not because you made a winning move (which you didn't) but because I simply give up trying after it has become clear that you are just not getting it, and enough other editors have become aware of the problem. I am sure there are plenty of examples in my user talk page archives where you can see how I act when I was actually wrong. As a mathematician I have received years of training for noticing when that is the case. I am not going to apologise for my "false accusations" to you because I did not make any. I did apologise (on my talk page) for an actual false accusation I unwittingly but negligently made against George1918. The misunderstanding was caused by you, and I am not sure that you have apologised to George1918 yet for your talk page manipulation and your over-the-top attacks against the user. Hans Adler 22:06, 7 February 2011 (UTC)
Hans. You too. Stop listing user-behaviour issues that have happened elsewhere. Can both of you just discuss guideline text or else take a break. Colin°Talk 22:17, 7 February 2011 (UTC)
Hans, in addition to insulting me, now you insult yourself. I, too, am a mathematician (retired though - Stanford for 11 very long years), but I would never post that self-insult admission online in a public space... ? A mathematician is like a fashion model; way over the hill at age 35, hence retirement. Unlike a physicist, a mathematician is not connected to reality (except at Caltech, which has a math dept full of closet-physicists)- a retired mathematician is even less connected (less than zero). The rankking goes like this - an engineer is just a stupid experimental physicist, an experimental physicist is just a stupid theoretical physicist, and a theoretical physicist is just a stupid mathematician. But "stupid mathematician" contains a redundancy. So everyone is just a mathematician... except retired mathematicians, which are less. PPdd (talk) 22:32, 7 February 2011 (UTC)
(ec) Sorry, Colin. I felt compelled to respond here by PPdd's accusation that I had accused them wrongly, and then I got carried away. Regarding guideline text, I find it hard to take PPdd's proposal seriously. ("Such 'authorities' may sometimes be quoted as stating that 'Practitioner of this alternative medicine believe that... (or have these practies...'), if it is a major source for that alternative medicine" – "may sometimes"? seriously? after all those other qualifications? and this belongs in MEDRS?) Now if anyone could think of a good explanation of the difference between scientific peer review and the kind of internal peer review that happens inside the more alternative fields of alternative medicine, then that could perhaps be added here, although I am not sure it's needed. Any misconceptions about peer review processes are usually clarified quickly and efficiently at the noticeboards – unless someone chooses to make that impossible by grotesque overreactions. Hans Adler 22:36, 7 February 2011 (UTC)
Actually, if anyone's got a good, verifiable description, the place to put it is in the Peer review article, not here. Please put information about individual journals' reputations directly in the articles about them. (See Medical Hypotheses for an example.) And if you've got any leverage with academics, Wikipedia would seriously benefit from each field publishing a couple of articles about their journals—you know, all that stuff that the grad students learn in the hallway as they desperately seek publication, with the result that they all agree on what constitutes a C-class journal, and which journals favor which kinds of papers, but nobody (apparently) ever puts in print. WhatamIdoing (talk) 00:24, 8 February 2011 (UTC)
Maybe the Wikimedia Foundation would like to found the Journal of Publication Studies. Hans Adler 17:03, 22 February 2011 (UTC)
Such studies, e.g., at MIT and Rutgers, are already being done - see the Notability stub I recently started. PPdd (talk) 17:25, 22 February 2011 (UTC)

Proposal for adding to MEDRS and AFD: re Viagra for AFD

There is a proposal for adding to MEDRS and AFD being discussed here[55]. PPdd (talk) 17:27, 20 February 2011 (UTC)

Altmed scope?

On 4 Dec. 2010, User:WhatamIdoing added a phrase, apparently without discussion, that explicitly extended the scope of this guideline to alternative medicine articles. This extension is being applied by User:PPdd to dismantle articles like Acupuncture points. It seems to me that wikipedia is going to lose all of its info on alternatives if it only accepts WP:MEDRS sources. Is this really what we want? Is it WP:NPOV, or is it just a tool to push the standard western scientific medical POV? Dicklyon (talk) 03:09, 22 February 2011 (UTC)

Ignoring the specifics of any particular content dispute, it seems clear that any medical claims falls under this guideline, including those made by alt-med practitioners. That is not to say that a description of alt-med claims need necessarily be sourced to a MEDRS, but that the evaluation of those claims as to whether they actually work, are safe, etc. should be sourced as described here. MEDRS clearly encompasses all medical claims, alt med or not. Yobol (talk) 03:15, 22 February 2011 (UTC)
  • Dicklyon, you should notify me if there is a discussion you start about me.
  • The complaits I have gotten have been about adding too much to articles like acupuncture, not dismantling them. My explanations regarding acupuncture points are on that article's talk page, which you have not responded to. My only use of MEDRS has been to remove medical claims that are (1) based on primary source studies, as there is an expected random positive outcome if only for the meaning of p-value; and (2) that are "peer reviewed" by alternative medicine practitioners, who are not qualified to peer review for medical conclusions. That is my only use of MEDRS.
  • Alt med journals and other sources may sometimes be usable as RS for sourcing claims about practices and beliefs, but never for medical claims. I have not abused anything to dismantle anything, and you should cite specifics if I might have erred somewhere.
  • On a brighter side, I just made an edit about the use of "tiger's penis" in the first section of traditional chinese medicine, and "Dick lyon", your name sounds surprisingly related. :) PPdd (talk) 03:34, 22 February 2011 (UTC)
  • I just checked WhatamIDoing's edits, and I did not use these in any way to delete anything. I delete all medical claims that are not backed up by MEDRS, whether or not they are in a med article, in an alt med article, or in any article, as long as they are medical claims, they must have MEDRS. That is the clear spirit of MEDRS, whether or not you think alt med articles can have medical claims without MEDRS supporting them. PPdd (talk) 04:08, 22 February 2011 (UTC)
  • Dicklyon you should retract the false accusations you are making against me here and elsewhere like here[56], which you also made without even notifying me that or where you are making them. PPdd (talk) 04:08, 22 February 2011 (UTC)
@Dicklyon. Forgive me if I have misunderstood, but you seem to think WP:MEDRS does not apply to some Wikipedia articles; that you can make claims about the safety, efficacy and strength of evidence for some health interventions without WP:MEDRS-compliant sources, so long as you make them on certain types of articles. Have I got that right?
Personally. I think this guideline doesn't need to single out alt med articles, it just needs to make it plain that any biomedical claim, particularly claims about safety, efficacy and strength of evidence, in any article, must be supported by a WP:MEDRS-compliant source. --Anthonyhcole (talk) 04:54, 22 February 2011 (UTC) Corrected 05:02, 22 February 2011 (UTC)
It shouldn't need to, but the mere presence of this topic on this talk page suggests some may think alt-med is outside the scope of this guideline, which is clearly not the case. Yobol (talk) 04:59, 22 February 2011 (UTC)
Some kind of clarification is necessary, I'm saying it should be along the lines of "this guideline covers medical claims in all Wikipedia articles" rather than "this guideline covers medical claims in alt med articles too." I understand there is nothing in the guideline that remotely implies it only applies to a certain class of article, or doesn't apply to a certain class but, obviously, it needs to be spelled out for some readers, and I'd prefer the former rather than the latter formulation. --Anthonyhcole (talk) 05:13, 22 February 2011 (UTC)
Well, this guideline already says, "This guideline supports the general sourcing policy at Wikipedia:Verifiability, with specific attention to sources appropriate for the medical and health-related aspects in any type of article." The additional portion of "including information about alternative medicine" was the portion in question, which I do not see how it could be objectionable to anyone except those who don't think alt-med falls under the scope of this guideline. Yobol (talk) 05:20, 22 February 2011 (UTC)
Sorry. I wrongly assumed "any type of article" was part of the recent addition of "alt med." --Anthonyhcole (talk) 05:59, 22 February 2011 (UTC)

We got here when I noticed what User:PPdd had done at Acupuncture points, reducing its content by about 90%. His first diff cited this guideline, which I was not otherwise familiar with, as a reason for removal of a section that wasn't much more than a quotation from the altmed journal "Evidence-based Complementary and Alternative Medicine." So I looked at this guideline; at first I didn't see that it mentioned alternative medicince, so it seemed odd to apply it this way to prevent journal outside its scope from ever being used as sources. Then it was pointed out that it does mention altmed, and I noticed that was a recent addition that had not even been discussed. Probably the real problem is that the guideline is being misapplied, not that there are different sets of articles where it is applicable or not. But for infomration about alternative medicine, the alternative journals are where it's likely to be found. Censoring that through the lens of standard medicine seems like it defeats the point of trying to write encyclopedically on the topic. Based on his proposal below, I'm guessing that PPdd thought this section was about "claims about safety, efficacy and strength of evidence." I don't see why one would think that. Dicklyon (talk) 05:44, 22 February 2011 (UTC)

The point is that all medical claims, no matter the article or subject are judged on the same basis. We don't carve out a special exception to our guideline for alt med sources. While alt med journals are more likely to have information (by quantity) about alt med than non alt med sources, they also are also generally of lower quality and often much less reliable (to the point where some are not much better than SPS) when discussing objective measure of their safety/efficacy. Yobol (talk) 05:57, 22 February 2011 (UTC)
Hi Dick,
About that "apparently without discussion", you'll of course want to actually look in this page's archives. There's a search button, which you'll want to use because it was discussed several times, but don't miss Wikipedia_talk:Identifying reliable sources (medicine)/Archive 4#Resolved:_To_mention_.22complementary_and_alternative_medicine.22_on_this_page, which is the discussion most immediately related to the change. The main point was to get the word "alternative" into guideline, so that people who are searching the page will find it, rather than having to guess whether Wikipedia has different rules for "conventional" and "alternative" medicine (and then, how editors classify the specific subject in question, since the line between the two is porous.)
It sounds like PPdd's edits have exceeded the advice in this guideline. This guideline supports the use of scholarly, peer-reviewed journal articles, regardless of whether they are pro- or anti-. The conclusion that the authors come to is utterly unimportant in determining whether the source is the right type of source (which is what MEDRS cares about). There are many excellent types of sources available for AltMed subjects, and even non-peer-reviewed publications are acceptable for descriptions of what something is ("___ is a smelly plant...") or what AltMed folks think about something ("...that is used by herbalists to treat the common cold"). WhatamIdoing (talk) 06:21, 22 February 2011 (UTC)
As long as "alternative medicine" or "complementary medicine" contain the words "medicine" and claim to be a therapy that treats illnesses, then it falls under this guideline. PPdd has tried and failed to get support for the idea that AltMed journals cannot be used for medical claims, or that such journals are absolutely banned by MEDRS. PPdd, if you wish to repeat this again, could you emphasise that it is your own opinion. I agree with WhatamIdoing on this issue. Colin°Talk 08:50, 22 February 2011 (UTC)
Colin, how can a journal not peer reviewed by doctors or scientists be MEDRS for science or medical conclusions? An acupuncturist needs only one year (1850 hours) of training in sticking needles in people to stop the flow of a supernatural force. How can this qualify that person to "peer review" science and medical studies or conclusions? PPdd (talk) 09:17, 22 February 2011 (UTC)
Are you saying that you need a western medical degree or science degree before you can judge (let alone publish) reliable medical or scientific material? How many hours of medical or science training would one need before being able to peer review science? Does belief in supernatural forces disqualify someone from peer reviewing science or medical studies? Colin°Talk 10:58, 22 February 2011 (UTC)

Proposal to add to WP:MEDRS

Add

"Biomedical claims, such as claims about safety, efficacy and strength of evidence for a health intervention, in any Wikipedia article, should normally consider this guideline."

Please comment.

  • SUPPORT PPdd (talk) 05:25, 22 February 2011 (UTC)
Are they my words? Anyway, they're wrong. Either it's any or it's not. "Any, but particularly X" is self-contradictory. How about

"Any biomedical claims, including claims about safety, efficacy and strength of evidence for a health intervention, in any Wikipedia article, must be supported by a WP:MEDRS-compliant source."

--Anthonyhcole (talk) 06:07, 22 February 2011 (UTC)
Changed above. PPdd (talk) 06:20, 22 February 2011 (UTC)
PPdd, at this time, I do not believe that you understand this guideline well enough to be capable of proposing appropriate changes to it. Additionally, your proposed statement is a circular argument, which is poor form for policy writing. I cannot support this proposal. WhatamIdoing (talk) 06:22, 22 February 2011 (UTC)
Please suggest a rewording. It's to cover Dicklyon's implied objection to your edit that he stated, but did not argue for, in the section above, to which this is a subsection. I may not have alot of experience of wiki editing, but I did philosophy of and data analysis at stanford for 11 years, and I have seen alot of abuse of primary source studies. But you are correct in that it takes editing experience to anticipate effects of policies and guidelines. As to "circular", self referentiality is my forte, and my image in the mirror just seconded my having said that, so I am not alone in the opinion. PPdd (talk) 06:33, 22 February 2011 (UTC)

1. There is no such thing as a "MEDRS compliant source". All sources have varying degrees of reliablity wrt the material they contain and the reliability varies depending on what you are taking from the source and how emphatically you are stating the facts, etc. In addition to reliability, there are WP:WEIGHT issues and the need to keep our facts up-to-date. These are all factors on the sources we choose and how we use them. That is why this is a guideline, and not a policy page. It is complicated. Those who wish it were simple are deluding themeselves.

2. Guidelines can never contain the word "must". Go edit the policy pages if you want that level of authority.

Colin°Talk 08:50, 22 February 2011 (UTC)

I modified it per your comments. Please reevaluate to make sure I understood you. PPdd (talk) 09:04, 22 February 2011 (UTC)
It is still circular and unnecessary. Have a read of our nutshell and lead. We cover the scope of this guideline already. Colin°Talk 10:09, 22 February 2011 (UTC)
Colin, I understand "WP:MEDRS-compliant source" to mean a source that supports a claim in accordance with this guideline. I'd be surprised if anyone reading this didn't understand it to mean that. Certainly, there are some things journals peer-reviewed by acupuncturists are good sources for, but I'm surprised to hear you think they can be cited for medical claims. What journal/s and what kind of medical claims? I've missed the discussion involving PPdd that you refer to. Perhaps someone could just link me to it if it's convoluted.
As for the word "must," it's time there was a medical content policy. It's odd that Wikipedia is only worried about harm to reputation and being sued for libel. --Anthonyhcole (talk) 11:37, 22 February 2011 (UTC)
The problem with "WP:MEDRS-compliant source" is similar to the problems people have with the whole primary/secondary source debate. It puts emphasis on the noun as though it can be judged on its own. Context is everything. The other issue is the idea there is some threshold where one can be compliant or not. There's always scope for judgement and sources are on a spectrum. Very little can or needs to be sourced to an independent meta-analysis of randomised controlled trials.
The discussion was just further up in this page. The issue is fundamentalism. A Richard Dawkins-mindset (I'm right, you're deluded) just doesn't work on WP. We cannot rule out sources by characterising their authors as nutters, which is what seems to be happening.
As for a medical content policy, do you mean medical facts should be supported by science? We currently have WP:NPOV, not WP:SPOV. I don't see that changing any time soon. We are fortunate that the majority of experts in the English-speaking world approves of and makes some use of the sciencific process, so NPOV is generally useful. But the idea of western medicine practice being build on the scientific method is a remarkably recent phenomenon and by no means a universal feature. Colin°Talk 13:24, 22 February 2011 (UTC)
I know what you mean by Richard Dawkins mindset. I had the same view when he blustered into Acupuncture. However, he is actually a very fast learner and absorbs and applies policy with alacrity. When he asked Whatamidoing above to be specific about what she believes he doesn't grasp about policy, he was being sincere and I hope she explains her remark. He is quite overwhelmingly prolific and poor WLU at Acupuncture has pretty well just let PPdd run with it because keeping up is too much. Though WLU would be out of the blocks in a second, I'm sure, if he saw it going off the rails.
I'd like to see the med community here be a bit more AGF and helpful with this new user, who has been unfailingly polite and accommodating in all of the interactions I've wittnessed, is clearly energetic, intelligent and happy to work within policy. They don't turn up here that often.
Re medical content policy, I'd like Wikipedia to take medical content at least as seriously as it does BLPs. I'd like to be able to erase unsourced medical claims (at least those that, if wrong, could lead readers to make unhealthy decisions) from articles, with the same impunity as I could an unsourced potentially libelous claim. --Anthonyhcole (talk) 14:25, 22 February 2011 (UTC)
If WP is to have articles on altmed topics, there has to be room for describing what those altmed doctrines and practices are, based on what the practitioners hold to be reliable sources. That does not, of course, mean that WP should endorse those doctrines as facts, but it would be absurd to reject all works by the practitioners in a field. We could compare the religion articles: would we only admit articles on saints to contain statements backed up by assertions in agnostic publications, or would we couch the statements in language that conveys what followers believe without endorsing those beliefs? LeadSongDog come howl! 13:53, 22 February 2011 (UTC)
I don't see that happening here. Describing the POV and claims made by believers is indeed required per NPOV, and any sources can be used to do that. It would often be good to attribute such claims. What MEDRS then requires is that the mainstream POV, backed up by real research, gets more weight as a counter to what are usually unproven, speculative, disproven,and sometimes nonsensical claims. MEDRS is the trump card for actual statements about the scientifically verifiable facts of the matter, and lesser quality sources used by believers can only be used to document the existence of claims and beliefs. -- Brangifer (talk) 16:32, 22 February 2011 (UTC)
Yes, that seems to be one of PPdd's major misunderstandings.
No source is "reliable" for everything. A source's reliability can only be determined in comparison with the specific statement that is made. A Cochrane review on the common cold is a bad source for information about the general theory of relativity. A website published by a nutter is a perfectly fine, MEDRS-compliant source of information about the nutter's view. PPdd is trying to enshrine an inappropriate distinction in this guideline. WhatamIdoing (talk) 16:57, 22 February 2011 (UTC)

Prior consensus on application of MEDRS re biomed conclusions "peer reviewed" by out of field non-experts

The question arose as to whether an alt med journal could be considered MEDRS for making medical related science findings, using its claims to be “peer reviewed”.[57] A homeopathy journal published a primary single trial physical chemistry conclusion, “peer reviewed” by homeopathy believers, not physicists or chemists. The issue has arisen again, regarding acupuncture journals, “peer reviewed” by acupuncturists or self-proclaimed believer in supernatural “qi energy”, as to whether or not they can be considered MEDRS for highly medically esoteric bio-medical conclusions. You will likely be surprised that experienced and rational editors who you might already know and highly respect, are taking the position that the answer is “yes”, and will not allow a FAQ at MEDRS talk, or a statement on MEDRS that the answer is “no”. One editor objected to as being a “nonsense” poll, something akin to “idiotic” or “rhetorical”, or “straw man”, and the fact that I notified so many pages to participate, that he objected that this was similar to canvassing, and many editors did not bother to vote as the answer was “obvious”.[58].

The response to the poll was unanimous consensus on what was notifies across many talk pages, with many editors not even bothering to vote, because the answer was so obvious as to not even merit a vote. Since this MEDRS talk page was noticed, as well as a very large number of projects and talk pages that might be affected, and the consensus was unanimous, content should go in to MEDRS as to this interpretation of MEDRS re “peer reviewed journals reviewed by persons outside the field of the published biomed study, or at least a FAQ should appear on this talk page regarding this interpretation of MEDRS. Calling in to question the clear consensus should not happen.

I thought there had already been unanimous consensus on this[59], as it was posted here[60], and here[61] re . My lack of experience in editing has been used as an “argument” against my proposals re MEDRS interpretation, so as to justify not responding to them at all. This is an invalid argument, as it is merely an ad homonym attack. Even if it had merit, while it is true that I lack editorial experience as to the MEDRS guideline implications for Wiki, I at least have the potential for understanding MEDRS, having worked doing philosophy of science and data analysis for 5 years at ucla, 11 years at stanford, 2 at caltech, and most recently at mit; but even if I were a middle school student, proposals should be responded to solely on their merits, and not be based on the proposer. PPdd (talk) 15:51, 22 February 2011 (UTC)

PPdd. You have asked us to state that "For all X, X is not allowed". When some disagree with this, you assume they mean "For all X, X is allowed". There's a whopping great big "it depends" in there. Or as Ben Goldacre says "I think you'll find it's a bit more complicated than that". Colin°Talk 16:02, 22 February 2011 (UTC)
Here is a specific proposal, taking into account Colin's comment -

::Add to FAQ on this talk page, ot to article,

Publications that are not peer reviewed or edited by scientists or MDs, should not normally be used as a source to support biomedical conclusions.

This would put a halt to innumerable edit wars. PPdd (talk) 17:20, 22 February 2011 (UTC)

I really, really doubt it. Skeptics would certainly be happy about it, but you would strip many genuinely useful and appropriate pages of much of their historical and theoretical sourcing. Even though qi, vital energy, prana, homeopathic nonsense and the like have never been proven to exist, they are still part of the "theory" of the disciplines. It is appropriate to note that the theory is not supported, it is not appropriate to use MEDRS to strip out much of the detail. A good article both lays out the nonsense, and goes on to cite sources that state that it's nonsense. MEDRS is a guideline, deliberately. In the case of primary sources, the source simply shoudln't be cited. No need to alter MEDRS.
PPdd, please slow down, calm down, and stop trying to turn wikipedia into RationalWiki. You must accept that there are ideas being bandied about in the peer reviewed literature that don't have a scientific basis, that don't have any real support, but are still debated. It's part of science, it's part of modern life, and it's part of wikipedia. You're trying to push the entire project in a direction that many, many experienced editors, myself included, don't seem to like. Alt med claims are being tested regularly. If they are found to work - we cite this when summarized by a secondary source. If they aren't, we cite this as well, with the same threshold. That's it. We don't debunk, either the claims of mainstream science or the claims of alt med, without a source doing it for us. WLU (t) (c) Wikipedia's rules:simple/complex 18:00, 22 February 2011 (UTC)
In scientific matters, the scientific point of view is the neutral point of view. When science is uncertain, it generally says so using words such as "evidence suggests" or whatever; alt med journals rarely do this, are largely uncritical, and certainly do not publish corrections when theya re wrong. The "crisis in epidemiology" brought about by the finding that rather than protecting against coronary heart disease, combined HRT actually increased the risk, is a great example of how science is embarrassed by its mistakes. Maybe you can point out to me the alt med journal which notes that - to pick an example at random - oscillococcinum is unsupportable because it's based on a bacterium that does not exist, isolated in a disease which we now know for certain is caused by another agent entirely. At best, alt med journals keep quiet about the more obvious absurdities of their field (like the homeopathic remedy "venus stella erans"). At worst, they print it uncritically. If you don't understand why realists find uncritical promotion of alt med a problem, then maybe you have not read this: [62]. A woman suffered a year of agony culminating in a needlessly early death because she relied on alt med instead of taking the advice of numerous qualified medical professionals. Reliable sources for biomedical subjects are reliable sources per the scientific method. The impact rating system should be our guide here. Left-field woo journals that will print the most abject nonsense if it serves their purpose, are not reliable biomedical sources, though they may be reliable sources about the many flavours of woo. It's not about mirroring RationalWiki or Conservapedia or any other project, it's about applying our own internal principles correctly. We're talking here about a science subject, not a branch of the paranormal, so scientific criteria are the only important ones when selecting sources. Which means, of course, that the International Journal of Crystal Healing is not a WP:MEDRS and never will be. Ditto any of the journals devoted to alternative medicine - or rather alternatives to medicine. Guy (Help!) 01:18, 23 February 2011 (UTC)
I thoroughly approve of this position. --Anthonyhcole (talk) 17:16, 23 February 2011 (UTC)
PPdd, I repeat what I pointed out previously. WP:MEDRS applies to articles that strictly discuss a topic from a scientific perspective. This is not necessarily true for the subject of acupuncture points, for which the sole reason why I eventually consented to a (momentary) redirect was the poor quality respectively lack of sourcing in the article. However, this topic can and should be discussed also from the viewpoint of Eastern philosophy, whether it is agreeable with Western science or not. There are two possible solutions: one article highlights the scientific side, which acupuncture obviously is doing, while another describes the traditional viewpoints, with both articles referring to each other; or a single article discusses both viewpoints. In any case, MEDRS will only apply to the Western scientific perspective. Nageh (talk) 18:40, 22 February 2011 (UTC)
There is another issue with the use of WP:PRIMARY sources. While Wikipedia articles must be based on secondary sources this is not an argument to exclude primary sources at all! In fact, primary sources are the definite reference when it comes to a specific issue – what is required though is a secondary source to back them up. Particularly, when there are such secondary sources this is not an argument to delete statements that are referenced with primary sources but no secondary ones. Nageh (talk) 18:44, 22 February 2011 (UTC)
Nageh, you can't split an article in two to discuss the opposing viewpoints. See WP:CONTENTFORKING. MEDRS is not an article guideline, it is a content guideline. I agree that many aspects of acupuncture points can be discussed without reference to the sort of sources MEDRS likes but if the article makes medical claims about the points, then MEDRS applies to those claims. Also, wrt primary and secondary sources, this is an old argument. We don't "back up" sources. We use the best source. If we can cite an excellent secondary source for all the facts we want to state, then we use that and the primary source can go gather dust. This is an encyclopaedia, a tertiary source, not a literature review, and we have no obligations to note the original research paper or honour their authors. Colin°Talk 19:06, 22 February 2011 (UTC)
Regarding a split into two articles with opposing viewpoints... well, it was some quick random idea. Another sort of split would probably be better. Concerning the use of primary sources: please respect that there are different viewpoints, and yes, this has been discussed repeatedly. I'm tired of any discussion right now after the lengthy discussions currently around acupuncture point (which are helpful!), so all I'm saying (and that's what I said) is that a statement referenced with a primary source is not an argument for deletion. Cheers, Nageh (talk) 19:57, 22 February 2011 (UTC)
See WP:BURDEN. "The burden of evidence lies with the editor who adds or restores material. Any material lacking a reliable source directly supporting it may be removed." Couple this with "extraordinary claims require extraordinary evidence" for good measure. As for different viewpoints, yes there are different viewpoints wrt primary sources. There are those based on policy and those based on wishful thinking. Colin°Talk 20:28, 22 February 2011 (UTC)
Thank you for your "constructive" attitude, Colin. :( Nageh (talk) 20:31, 22 February 2011 (UTC)
In the instant case, we're not really talking about extraordinary claims. It's stuff like "Somebody did some research on ___ aspect" or "Practitioners think ____". And PPdd's response wasn't to look for good sources, or to ask for good sources, or to copyedit the content to conform more strictly to WP:PRIMARY, but to simply blank 90% of the article's content. WhatamIdoing (talk) 21:08, 22 February 2011 (UTC)
Is the solution to have articles state something along the lines of "<Pseudoscience> has been found to aid with <whatever ailment(s)> in studies peer-reviewed by other <pseudo scientists>. However, numerous studies published in journals reviewed by scientists have consistently failed to reproduce these results. Rather, these studies have found that <pseudoscience> works no better than a placebo." Obviously, the language needs to be cleaned up, but I suspect you could make this statement about every pseudoscience out there and it would be substantially accurate.JoelWhy (talk) 20:46, 22 February 2011 (UTC)
(Re another of WAID's third person "asumption of bad faith" false ad homonym argument by attack on me, I did look for sources before deleting at acupunture point, despite not having to under WP:BURDEN; I have spent much time at acupuncture with identical claims and sources, and added much content back in after deleting, with RS, just as I did at anthroposophical medicine. Please assume good faith and argue not by incorrect and assumptions of bad faith editing on my part, and false and baseless ad homonym attacks on me, but argue about the issue.) PPdd (talk) 21:36, 22 February 2011 (UTC)

Modification responding to the above comments - Add to FAQ on this talk page, ot to article,

Publications that are not peer reviewed or edited by scientists or MDs, should not normally be used as a source to support biomedical assertions. Such non-medical/science peer reviewed assertions might be included as an assertion of the beliefs of an author about the biomedical conclusions they believe can be drawn about their study or review, subject to WP:UNDUE, but normally their biomedical conclusion should not be stated as an assertion, citing their publication as a source. Assertions about methods or practices, about which the reviewers are specialists, might be sourced by the publication.

On another part of this discussion, no one has responded to the already existing recent unanamous consensus on this exact issue, widely noticed both here, and at most possibly affected talk pages, as noted above. PPdd (talk) 21:36, 22 February 2011 (UTC)

The point of peer review is that your peers are doing the review. If you are an acupuncturist, your peers are other acupuncturists. If you insist that the review be done by non-acupuncturists, which seems to be your goal, it is no longer peer review. It is 'outside expert review'. WhatamIdoing (talk) 21:56, 22 February 2011 (UTC)
I modified it. How does it look now? PPdd (talk) 22:05, 22 February 2011 (UTC)
(Aside - WAID, I just looked at your user page, so change "they" to "she or he". :) PPdd (talk) 22:09, 22 February 2011 (UTC))

Rather than remove the questionable material outright, revise the sentence as to remove the medical claim itself; it should not be necessary to remove material in order to comply with the WP:MEDRS guideline. Wikipedia should simultaneously represent alternative medicine and suppress unsupported medical claims. In other words, allow the article to thoroughly discuss the nonscientific assertions, but ensure that it does not suggest that it actually works. If there is difficulty revising a particular section, please feel free to email me. razorbelle (talk) 22:13, 22 February 2011 (UTC)

Well put, and that is precisely what the proposal is intended to do,
  • Simplified proposal -

    For a biomedical assertion based on a study or review, when the "peer reviewers" are peers of the author, but not scientists or MDs, one should change the assertion into an assertion about an authors' beliefs about the conclusions to be drawn from a study or review.

The rest of the words in the longer proposal were just to incorporate all of the sugestions of the above editors to avoid possible misinterpretations. PPdd (talk) 22:25, 22 February 2011 (UTC)

Further to Razorbelle's suggestion, it would be helpful to come to grips with the common markup tags that are used to indicate inadequately referenced assertions, such as {{MEDRS}}, {{dubious}}, {{MEDCN}}, {{vn}} and {{primary}}. It is usual to allow other editors at least a month before concluding no such source is available unless the assertion is very recently added, patently incorrect and beyond correction, constitutes a wp:COPYVIO, or is subject to removal under wp:BLP. Remember that there is wp:NODEADLINE and trust that other editors also want to achieve a high quality product in the end even though we have diverse approaches. LeadSongDog come howl! 22:44, 22 February 2011 (UTC)
Good point, deserving of its own section... -> PPdd (talk) 23:35, 22 February 2011 (UTC)
Actually, none of this is deserving of being in the guideline. The parts that aren't pure CREEP are just disguised POV pushing. The only possible result is that we'll end up with disputes that are nominally about who, exactly, gets to be called a "scientist" rather than disputes that are nominally about which sources, exactly, are "MEDRS-compliant". What we need is more people actually comparing the material to the source, and deciding whether the source is strong enough to support the material, rather than trying to anoint classes of authors as always being right. WhatamIdoing (talk) 00:20, 23 February 2011 (UTC)
What POV "parts"? Maybe you are looking at all three together, so I strck the first two proposals. The last proposal is only one very brief sentence immediately above. CREEP can be used to make a slippery slope argument to forever freeze any policy or guideline. For any proposal, there will always be a an argument using an "infinite regress of definitions", such as "but then who created god?", or "who counts as a scienctist?". As a practical matter, the unanimous consensus vote on this exact issue, posted for comment at this talk page, was helpful, as it was used to utterly shut up debate on this at homeopathy talk. I don't know how much (bad) experience you have had about sourcing alt med articles, but this short sentence would have been very helpful time and again. PPdd (talk) 01:11, 23 February 2011 (UTC)

I feel that this issue has already been adequately addressed in the body of the guideline. If we spend more time revising articles to make them MEDRS compliant and less time halting others from making them less so, then these details will, in time, become unnecessary. We should take a moment to remember that the goal of this guideline is to ensure that no articles on Wikipedia need our markup tags. razorbelle (talk) 02:29, 23 February 2011 (UTC)

I think the issue has been addressed in the body of the guideline, but others don't, leading to a need for things like this[63], posted for comment on this page, and this same difference of reading of MEDRS has come up over and over again. And there still seems to be a disagreement about the above proposal. "Never start editing alt med articles, its a bottomless waste of time sink", I was warned. I am bothering to spend time on this to ultimately save time and talk page fighting overall, for me and many others. Razorbelle, the reasoning skills of alt med pushers are likely very "different" from yours when it comes to reading MEDRS and the logic of thinking about words. Authoritative sounding guidelines work for people who believe things by virtue of dogmatic authority. PPdd (talk) 02:40, 23 February 2011 (UTC)
This straw poll was neither needed nor helpful (nor is it a ‘prior consensus’). I'm afraid your proposal is actually quite contrary to what you want to achieve. Since you chose homeopathy as an example, let me tell you that everybody on Homeopathy (journal)'s editorial board is an MD and/or scientist and there's no reason to expect their reviewers not to be MD's/scientists, too. Right now this doesn't mean it's a reliable source for anything other than homeopath's viewpoints, but if this proposal goes into MEDRS (it doesn't really look like that right now), I predict people will argue since reviewers and authors there are MDs, ‘Homeopathy’ is reliable according to MEDRS. --Six words (talk) 13:30, 23 February 2011 (UTC)
Six words' "prediction" is the first really good argument against my proposal that I have seen. The underlying problem still needs to be addressed, but I am withdrawing it until I can think up a way with dealing with what Six words just pointed out. I am thinking along the lines of a simpified version of adding "a publication can still be considered not to be reliable even if editors and peer reviewers have science or medicine degrees", but I am not yet prosing this until I think more about it. Six words, the problem came up again and again elsewhere. Do you have a suggestion to deal with it, when you, I, and the others here are not going to be watching every place where this will come up. PPdd (talk) 14:13, 23 February 2011 (UTC)
This is a good argument against. There's also the point I keep raising that some of the standards people want to apply to alt-med, if applied to fairly uncontentious conventional med, could make things difficult if someone wanted to be WP:POINTy about it. I know that some of alt-med requires rewriting the laws of physics in order to justify itself, so we've got the "extraordinary claims" defence there. But all I'm saying is "be careful what you wish for".
PPdd, we now have the situation where the very journal you are trying to ban is actually reviewed/edited by the sort of people (on paper anyway) that you approve of. It does rather sound like you are trying to restrict the sources that "those who disagree with me" can use. Now, while I guess nearly everyone here thinks homeopathy is water, using MEDRS as your only weapon is perhaps not goint to achieve what you want. I noted much further up here that other policies are your friend and possibly more useful. Colin°Talk 15:21, 23 February 2011 (UTC)
I agree with Colin, we have other policies and guidelines that are also relevant when choosing sources. We cannot narrow MEDRS so much that it is impossible to suggest any article in an AltMed journal because this will also make it very difficult to source medical articles about subjects that are ‘mainstream’ but not very common. I don't really see it as a problem that every now and then someone suggests to use an AltMed article and we have to explain why it can't be used in the way suggested. Those suggestions often come from new, inexperienced editors who don't know the relevant guidelines, so no matter what the guideline says, discussions about that will come up again. --Six words (talk) 17:04, 23 February 2011 (UTC)
Colin and Six words both make good points about considering implications for other medical claims. But any medical claim should be sourced by RS, and if the source is not peer reviewed by scientist/MD's, or secdondary or tertiary to such, it should not be in Wiki in the form of a sourced assertion, even if it is a legitimate assertion made by unreviewed scientist/MDs. Six word's example is taken care of by the wording I last suggested, from a logical perspective, since it only restricts the form of some assertions, but does not automatically allow things that it does not preclude. But the reality of readers of MEDRS is that they will not think logically; it has to be spelled out for many. If a legitimate biomed conclusion of a systematic review is published in a poetry journal, and peer reviewed by poets, while it may be legitimate, it should not be sourced by the poetry journal. I am still thinking about new wording that will take care of the potential nonlogical application of the above proposal in MEDRS which Six words pointed out. PPdd (talk) 17:45, 23 February 2011 (UTC)

How long should seroius nonMEDRS assertions stay up?

Should nonMEDRS assertions stay up with tags like {{MEDRS}}, {{dubious}}, {{MEDCN}}, {{vn}} and {{primary}}?

Here is a real life Wiki MEDRS abuse story about this. I was caring for my mother with multiple myloma, so, being a perpetual freshman in a new field every year, I decided to enroll in a local city college and state university, took a sewing class and a textiles class, and decided to put on an all-volunteer fashion show. The show made the cover of Apparel News, and I was approached by a retired sr. state senator to bring the show to a breast cancer fundraiser in the center of Hollywood, at the Hollywood Celebrity Center, which was located in the Scientology complex. An inventor/MD in a booth next to a proselytizing Scientologist was marketing his new very early detection-breast-tumor-finder-machine, the studies for which were published in a "peer reviewed" alternative medicine journal article about it, with the conclusion asserting that it detected the tumor in almost 100% of women with breast cancer, and at an early stage at which mastendectomies were highly successful. As with most marketers, Wikipedia is the first place to go to promote sales and investment for things in appearing in marginal publications. It turns out that the invention not only detected tumors in almost 100% who had them at a very early stage, it also showed that about 50% of women had them when they did not, which was not mentioned in the publication. Furthermore, this invention detected tumors at such an early stage compared to other technologies, that a surgeon considering mastectomies had no other means of detecting the tumor when the machine showed one. The article did not appear to be advert, because it was "based" on a study and review published in a leading alt med "journal", as in the section above, which was "peer reviewed" not by scientist/MDs, but by alt med whatevers (see section above on "peer review" and MEDRS for biomed assertions). A celebrity-MD found out about this, and blanked the entire Wiki article as nonMEDRS, instead of tagging it and waiting a month. This was just before the event. The inventor/doctor then reverted this blanking, and immediately nominated his own advert article for AfD (an inventor/MD/wikilawyer, apparently). The article was then protected for a week. The doctor then sold his invention to a huge number of people and got substantial funding for more nonWiki marketing from investors. Who knows how many women may actually have gotten unnecessary mastendectomies as a reusult of the MEDRS-AfD interaction.

  • I say allow editors to apply severe WP:BURDEN on this if an editor reasonably thinks there is any possibility of harm coming from using Wiki as a source of medical information.
  • Any comments on how long to keep up "serious" bio-med claims with questionable or no sources, tagged with things like {{MEDRS}}, {{dubious}}, {{MEDCN}}, {{vn}} and {{primary}}? PPdd (talk) 23:35, 22 February 2011 (UTC)
Editors must use their best judgment. Spelling out "rules" for this isn't helpful. An assertion that is distinctly dubious and cannot possibly be fixed by copyediting (e.g., "All cancer is caused by a fungus"), especially if it is on a minor point (e.g., "___ cures cancer, as proven by pouring the substance on a single dish of cells in one lab"), would probably be removed immediately. An assertion that seems at least highly plausible to you should be left alone—for decades, if necessary. It would be silly to tell people to drop whatever they're doing because "The human hand normally has five digits" has been unsourced for an arbitrary number of months and will be deleted if not supplied with an inline citation before the WP:DEADLINE.
For the fraction that falls between "immediately" and "never", you must use good judgment. As a rule of thumb, however, unless you are prepared to provide sources indicating that the material is wrong, you should wait at least a week after tagging the material or describing your concerns on the talk page. Most editors are willing to wait more than one month.
As for your question about unnecessary mastectomies, the answer is probably "zero". Surgeons are usually quite picky about getting biopsy results before agreeing to perform mastectomies. Their medical malpractice insurers tend to insist upon it. WhatamIdoing (talk) 00:12, 23 February 2011 (UTC)
In an attempt to be brief, I failed to tell the complete story. (I have since modified it.) The point of the invention is that it is the earliest known detection device(or so the claim went, and even the celeb/MD conceded this point), so it is the only thing that will show the tumor at that early a stage, leaving surgeons with only that machine's viewpoint, and the surgeons subconscious bad faith profit incentive, for the surgeon to go by. You raised another issue as to the seriousness of the claim being subjectively measured by the editor. I just started the seriousness article (it was originally intended as a user page joke, but turned out to be a RSable article, with a history section, too). A wiki editor has to be something like a triage nurse at an emergency hospital re med articles. I am not suggesting any proposal for putting in the MEDRS, but I thought someone might have one, and as you know, the issue just came up at acupoint, which is very low on the seriousness-ometer. PPdd (talk) 00:32, 23 February 2011 (UTC)
The added details don't change the outcome. Boiled down, the conversation would have gone like this:
Woman: But the machine said I had cancer!
Surgeon: Well, apparently that non-FDA-approved machine was wrong.
Woman: But the machine said I had cancer that nobody else can detect!
Surgeon: Right, because you were the victim of a swindle by a quack perpetrating yet another hoax based on junk science in the name of alternative cancer treatment and detection. Your health insurance company won't pay for surgery to remove a non-existent cancer, the hospital won't let me admit you for this surgery, and I could even lose my medical license if I went along with this delusional diagnosis. And then we always send 'cancerous' breasts off to pathology for more details, and when they report back that there was no cancer anywhere in the breast, you're going to sue me. So the final answer is no: I will not do this surgery.
Surgeons do actually run into these situations (and worse), and while the goal is to be gentle and tactful, their job is to say no when there is no possible benefit to the surgery. The "profit motive" of a single dubious surgery does not outweigh the costs. The cost of having his license suspended for even one day would far outweigh the profit from a one-hour surgery, especially since that hour could have been spent on a much less risky surgery. WhatamIdoing (talk) 18:49, 23 February 2011 (UTC)
One would hope. Sounds like you live in the same bubble I do. Under your scenario, with that crowd, a more likely outcome would be no surgery, and the woman would have gotten acupuncture for cancer, never developed it, and acupuncture would, once more, have saved the day. I ignored that whole thing, which was only one of many new age "healing" boooths, and since the fashion show had unravelled. I do know that the inventor/MD got lots of investors from the Scientology and Hollywood crowd present. PPdd (talk) 21:53, 23 February 2011 (UTC)
In the case of altmed articles, where the beliefs and finding of the proponents can't generally be sourced to MEDRS sources, it would often be better to convert the statements into attributed statements of belief or opinion, rather than leave them as unsupportable assertions of fact – in preference to just removing them. Dicklyon (talk) 00:21, 23 February 2011 (UTC)
That's essentially the last wording of the proposal in the above section. PPdd (talk) 00:32, 23 February 2011 (UTC)
PPdd, I might be more inclined to believe that assertion if your behavior did not do the opposite. You have been deleting large swaths of articles rather than fixing them this way, in violation of WP:PRESERVE and in conjunction with assertions that this guideline demands deletion of material that is not currently supplied with a source that happens to meet your uniquely stringent reading of the guideline. WhatamIdoing (talk) 04:03, 23 February 2011 (UTC)
Actually, every article I have worked on initially got shorter, and when I was done it was longer and everything had an inline citation. The only exception is acupuncture point. As I was deleting things to reinsert them back, reworded if needed, or with RS, the article ended up almost blank. Then I noticed that there was nothing that was not in acupuncture, and brought it to AfD, instead of doing all that work for a CONTENTFORK. Especially as I had just worked over identical material at acupuncture, so why do the same thing twice. I figured the only way it would survive AfD would be with material not in acupuncture, since that is the definition of cfork, so I sincerely either expected unanimous "merge" votes, or I expected to see something not at acupuncture, that did not belong at acupuncture. I actually did see one thing that I reworded a little and copied over to acupuncture. PPdd (talk) 21:40, 23 February 2011 (UTC)
  • There is onyl one rule: improve Wikipedia. You can improve Wikipedia by removing special pleading. You can improve Wikipedia by clarifying that a claim is made only by true believers and showing evidence that the mainstream considers it to be twaddle. You can improve Wikipedia by clarifying the attribution of claims, so: "According to homeopath Dana Ullman, Montaigner's work proves the scientific basis for homeopathy, but CBC's Marketplace report Montaigner as saying his work cannot be extended to cover homeopathy". Guy (Help!) 01:25, 23 February 2011 (UTC)