Talk:Acupuncture

Vickers's reply to Ernst

Vickers published a reply to Ernst and other critics in the journal Acupuncture in Medicine[1], which is a journal published by BMJ. Here's what he said:

(possible copyright vio removed)

Passing along in case it's useful. TimidGuy (talk) 11:46, 8 November 2013 (UTC)

Thanks for this! (an aside, fwiw -- we don't need to include Vickers' remarks on the laser study; that's a job for a review article.) --Middle 8 (talk)

Seems like Vickers's response could be included in the article. (Note that Vickers's remarks were refereed and published in a journal, whereas Ernst's remarks were simply a quote in a newspaper.) Perhaps we could add a sentence: "Vickers responded by citing evidence that suggests that the effect is not the result of a placebo." TimidGuy (talk) 11:20, 15 November 2013 (UTC)

What evidence did Vickers cite? Vickers response was his own opinion that has not been proven. QuackGuru (talk) 17:15, 15 November 2013 (UTC)

Well, so was Ernst's. --Mallexikon (talk) 04:20, 16 November 2013 (UTC)

Vickers cited a double-blinded study by Irnich that found a statistically significant difference that couldn't be attributed to a placebo effect, thus refuting Ernst's speculation. TimidGuy (talk) 11:58, 16 November 2013 (UTC)

Well, so was Ernst's? You are correct but what is included in the article is the opinion of Ernst that passed V. Vickers response was also his own opinion. But the proposal made by TimidGuy failed verification. You putting words in Vicker's mouth. "Vickers responded by "citing evidence that suggests"... That is original research. Editors do not make up there own conclusions about what Vicker's said. Where in the text did Vicker's say he "suggests" there is evidence.

"Vickers cited a double-blinded study by Irnich" Did Vicker's say this or are editors conducting there own review.

This is closer to the source: "Vickers responded by saying there was no bias in the study because investigators found a way around the problem of therapist blinding.[2]"

This is what Vickers said but Middle 8 said we don't need to include Vickers' remarks. There could of been consensus if Middle 8 did not reject the response by Vickers. But consensus can change. QuackGuru (talk) 19:42, 16 November 2013 (UTC)

"....we don't need to include Vickers' remarks on the laser study" was what I said, because we don't usually discuss single studies, and a letter to the editor isn't a good springboard for doing so (compared to a review article). But aside from that, I am in favor of summarizing Vickers' reply. --Middle 8 (talk) 18:45, 21 November 2013 (UTC)

However we word it, I think the essential point of Vickers is that the results can't be attributed to a placebo effect. Let's check with Middle 8. I didn't get the impression that he thought we should exclude Vickers's response. There's no policy reason for excluding it, and NPOV would suggest it be included. Also, it's unclear that a quote in a newspaper satisfies MEDRS, which disallows popular media as sources. And personally, I don't understand the opposition to Vickers, since the section already cites a number of research reviews that found a reduction in pain. TimidGuy (talk) 11:38, 18 November 2013 (UTC)

Middle8 only said that we wouldn't have to include Vickers' remark about the laser study, I think QG misunderstood him. Of course you can include Vickers' response if you want to. I personally don't think it is necessary, though. --Mallexikon (talk) 01:54, 19 November 2013 (UTC)

Why don't you think it's necessary? As it stands, it gives the impression that treatment is no better than a placebo, but Vickers study showed otherwise, and Vickers effectively refuted Ernst's speculations regarding why it might not be better than a placebo. TimidGuy (talk) 10:56, 20 November 2013 (UTC)

It would help your cause if you rewrote your proposal without putting words in Vicker's mouth. QuackGuru (talk) 16:54, 20 November 2013 (UTC)

Of particular importance is that Vickers is citing a single study, this does not apply to the Meta Analysis overall. Also clinical significance is of primary importance. A back and forth between two individuals should be limited in a WP article. Does Vickers allege that the Meta Analysis is not compromised by the general lack of blinding or that the results indicate clinical significance? These would be important points. Has another source commented on the Meta Analysis? - - MrBill3 (talk) 05:44, 21 November 2013 (UTC)

Hi, to clarify, I certainly believe Vickers' reply to Ernst should be cited in just as much detail as Ernst's comment. That is simple WP:WEIGHT. But in summarizing Vickers' reply, I do not think we should touch on his remarks about the later study. BTW -- let's not lose sight of the fact that Vickers' meta-analysis vastly out-weighs the followup comments of either Ernst or Vickers. So we might end up with 2-4 sentences on the meta-analysis, and then a single sentence on the back-&-forth. And sure, we can use expository text in footnotes, per an earlier edit by Mallexicon. -Middle 8 (talk) 17:10, 21 November 2013 (UTC)

Considerable undue weight to Ernst

At the moment (this version), we have a big undue weight problem: Vickers' meta-analysis is dwarfed by Ernst's remarks. (Vickers' reply to Ernst does not yet appear.) Remember that in the evidence-tier for both evidence-based medicine (EBM) and WP:MEDRS, meta-analysis ranks at the top and expert opinion at the bottom. Vickers' review is mentioned twice, and each time Ernst's reply gets not less weight, but considerably more. Note also WP:WTA in framing Ernst ("however"; "commented"), not to mention stating the title and position (when a wikilink would do) for one party but not the other.

• First instance:
  • Text devoted to meta-analysis: A 2012 meta-analysis found significant differences between true and sham acupuncture, which indicates that acupuncture is more than a placebo when treating chronic pain (even though the differences were modest).[3]
  • Text devoted to expert opinion: However, Edzard Ernst, emeritus professor of complementary medicine at the University of Exeter, disagreed with the study's conclusions and commented that it demonstrated that the effects of acupuncture were principally due to placebo.Footnoted text: In an article in The Guardian, Alok Jha stated:[4] Edzard Ernst, emeritus professor of complementary medicine at the University of Exeter, said the study "impressively and clearly" showed that the effects of acupuncture were mostly due to placebo. "The differences between the results obtained with real and sham acupuncture are small and not clinically relevant. Crucially, they are probably due to residual bias in these studies. Several investigations have shown that the verbal or non-verbal communication between the patient and the therapist is more important than the actual needling. If such factors would be accounted for, the effect of acupuncture on chronic pain might disappear completely".
• Second instance:
  • Text devoted to meta-analysis: A 2012 meta-analysis found acupuncture is effective for the treatment of chronic pain, specifically back and neck pain, osteoarthritis, and chronic headache.(source as above)
  • Text devoted to expert opinion: However, the study is disputed by professor Edzard Ernst, who commented that the meta-analysis demonstrated that the effects of acupuncture were largely due to placebo, since the difference in results from real or sham acupuncture were small and clinically irrelevant. Footnoted text: (cites to same footnote as above)

Enough said. Of course this is easy enough to fix; I'm just out of time right now. Will try later, or if someone else wants to, go for it. --Middle 8 (talk) 18:37, 21 November 2013 (UTC)

Ernst is not the only person to comment on Vicker's meta-analyis. Which of these comments, published in JAMA Internal Medicine, warrant inclusion? An "Invited Commentary" (this seems to give some weight) by Andrew L. Alvins, Needling the status quo: Comment on “Acupuncture for chronic pain” which has the text, "The conclusion that most of acupuncture's observed clinical is mediated by placebo effects..." and "...critics correctly note that a new pharmaceutical agent that fails to show superiority over placebo will not be approved. Why should the bar be lowered for acupuncture?" Placing acupuncture in perspective by Ziegelstein which questions the lack of context in comparison of effect size (and has another point behind the paywall). Acupuncture's elephant in the room by Barrett and London which points out the difference between study and practice as does Complexity of sham acupuncture (Huang et al.) which also notes some issues with the sham controls.

In terms of weight Vicker's reply to Ernst mentioned ONLY ONE study among those in the meta-analysis this does not support the entire analysis as using double blinded studies that overcome the placebo effect. As above Alvins seems to have the same opinion as Ernst. I know in general we use substantially stronger MEDRS but criticism of the meta-analysis published in the journal which published it seems applicable. Note Vickers also had a reply in JAMA Internal Medicine to the above (I don't have access to it right now).

Regarding the balance of meta analysis vs Ernst's commentary I think you have a valid point we should be able to reach some consensus on editing down the material from Ernst. I do think material in the notes section should be allowed but I remain open to input on that (everything else too for that matter).

refs

Best. - - MrBill3 (talk) 07:56, 22 November 2013 (UTC)

Here's the text of Vickers's reply:

(possible copyright vio removed)

With all these published, refereed sources available, it seems unnecessary to use a newspaper as a source. We can probably leave out Ernst (which in fact may not comply with the MEDRS proscription against popular media). TimidGuy (talk) 11:30, 22 November 2013 (UTC)

Of all the criticisms it seems Alvins is the one that has weight and bearing. Ernst may still qualify as expert opinion. - - MrBill3 (talk) 11:53, 22 November 2013 (UTC)

More great finds, thank you! (Boy do I wish I lived closer to a good library.) Agree, we should cite (and properly weight) the most substantive reply or replies, i.e. those that explain their positions, ideally in a journal. To be continued.... regards, Middle 8 (talk) 13:19, 22 November 2013 (UTC)

Still not clear in what way Ernst complies with MEDRS. Non-evidence-based expert opinion is nearly at the bottom of the MEDRS hierarchy. Ernst doesn't offer any evidence. He is simply speculating about what future studies might show. In addition, MEDRS says, "The popular press is generally not a reliable source for scientific and medical information in articles." TimidGuy (talk) 12:11, 25 November 2013 (UTC)

I tend to agree with TimidGuy's argument. Ernst's newspaper statements deserve low weight. He is a highly published, well recognized authority so a mention of his statement might still be warranted. BTW most of the text of the other refs I gave is available on their "preview" pages. - - MrBill3 (talk) 15:40, 26 November 2013 (UTC)

Thanks. Send me an email if you'd like to see pdfs of any of the sources. TimidGuy (talk) 11:25, 27 November 2013 (UTC)

I'll propose some new wording. And in the spirit of compromise, I'm open to including Ernst's statement. But I just wish there were something in MEDRS that supports using it. I can't see that there is. Non-evidenced-based opinion is at the bottom of the hierarchy. TimidGuy (talk) 12:06, 2 December 2013 (UTC)

Thanks for refocusing on this; I'll have a closer look. Agree re deprecating Ernst; I'd prefer to include something else, something with substantive criticism/comment, not just opinion. And not overdo the commentary's weight relative to the source (something I'm pretty sure you already agree with; just saying). --Middle 8 (talk) 13:05, 2 December 2013 (UTC)

So maybe we should go ahead and replace Ernst with the Avins commentary that appeared alongside the Vickers review. TimidGuy (talk) 15:53, 8 December 2013 (UTC)

This is a notable controversy and there is a disagreement with Vickers. There was a previous discussion to include the disagreement. See Talk:Acupuncture#Disagreement with Vickers. NPOV policy seeks some form of balance. When there is an opposing POV, we try to let each get the coverage according to RS, without allowing the article to become either a hit piece or a hagiographic sales brochure. The WP:SECONDARY source used in the article is a critique of the Vickers meta-analysis reported in The Guardian. The Ernst statement is the opinion of an expert which does have WP:WEIGHT. QuackGuru (talk) 18:03, 8 December 2013 (UTC)

The Ernst source simply doesn't comply with MEDRS. It's at the bottom of the hierarchy. Avins is a much better source for the disagreement. It complies with MEDRS and the criticism is more substantial. TimidGuy (talk) 12:07, 9 December 2013 (UTC)

The article is summarising a real world controversy which included comments from Avins. "At least in the case of acupuncture, Vickers et al have provided some robust evidence that acupuncture seems to provide modest benefits over usual care for patients with diverse sources of chronic pain," wrote Avins.[5] Avins is mostly in agreement with Vickers. The conclusion that most of acupuncture's observed clinical is mediated by placebo effects..." is not really explaining the controversy and the article is also supportive of Vickers.[6] QuackGuru (talk) 21:02, 10 December 2013 (UTC)

Ok we obviously have a pundits' controversy here... And I see two questions: 1.) If we want to include this controversy, what rational criteria do we chose in regards to which opinion to include? (Because we obviously can't include all letters to the editors we can find. And while Avins probably ranks very high in terms of MEDRS criteria... not sure where Ernst would be placed) 2.) We don't usually include commentaries about reviews and I'm a little concerned about creating an original precendent... Should we maybe leave this controversy out altogether? --Mallexikon (talk) 03:18, 11 December 2013 (UTC)

Are letters to the editor compliant with MEDRS? No. Some people think that every utterance from Ernst should be the main body of every article on CAM subjects. Thats ridiculous. Either we only accept secondary sources (systematic reviews) on the subject of efficacy or we don't. If we don't, then primary sources are fair game right? Ernst's letter is a fucking letter and does not belong as a source here. Herbxue (talk) 06:05, 12 December 2013 (UTC)

Well, it's not a letter as far as I understood it, it's a quote from him in a newspaper article... thus, the newspaper article would constitute a secondary source. --Mallexikon (talk) 06:13, 12 December 2013 (UTC)

MEDRS says not to use popular media as a source, so The Guardian shouldn't be used. Ernst simply speculates that future double-blinded studies may show that acupuncture is no better than a placebo. It seems silly to include this speculation in the article. On the other hand, while I disagree with Quackguru, who characterizes this review as controversial, I do feel there are clarifications and qualifications that could be added. For example, Vickers says that the results compared to treatment as usually are clinically significant. The commentary (which is not a letter) by Avins questions whether the results are clinically significant. Also, Vickers emphasizes the finding that acupuncture has a statistically significant effect compared to a placebo, but his characterization of the effect as "modest" glosses over the fact that it's not clinically significant. I vote for clearly stating Vickers's findings and then citing Avins to qualify them somewhat. TimidGuy (talk) 11:36, 12 December 2013 (UTC)

Thank you for pointing this out. I believe that this is an issue of undue weight (WP:UNDUE) and dispute that, in the midst of so many great sources used in the article, an editorial like this would be used (WP:MEDRS). I can't think of any articles of GA or B class in WPMed which use this structure: "A metaanalysis found that... However, in an editorial, X said...". I had a look at this article following the discussion on WP:MED and have left a comment below to this regard. Kindly, --LT910001 (talk) 10:49, 16 December 2013 (UTC)

Alright, that kind of answers my questions... So do we have consensus to leave the Vickers/Ernst controversy out altogether? --Mallexikon (talk) 11:29, 16 December 2013 (UTC)

I support leaving out Ernst. I just don't see how it possibly complies with MEDRS. It's not clear to me that Avins isn't compliant, but I'd go along with leaving it out. I would, however, like to reword the presentation of Vickers so it's clear that he found a clinical significant reduction in pain and that there was a statistically significant difference between acupuncture and a placebo. TimidGuy (talk) 12:19, 16 December 2013 (UTC)

I fixed the structure and weight concerns. The text is shortened to satisfy weight. QuackGuru (talk) 16:54, 16 December 2013 (UTC)

Well, no, the weight concern is not fixed. So far, consensus indicates to leave Ernst's comment out altogether, for the reasons given above. --Mallexikon (talk) 05:00, 17 December 2013 (UTC)

A 2013 meta-analysis did not determine acupuncture is cost-effective

"A 2013 meta-analysis determined that acupuncture is a cost-effective treatment strategy in patients with chronic low back pain."

The source: "Acupuncture as a substitute for standard care was not found to be cost-effective unless comorbid depression was included."

• Taylor P, Pezzullo L, Grant SJ, Bensoussan A. (2013). "Cost-effectiveness of Acupuncture for Chronic Nonspecific Low Back Pain". Pain Practice: The Official Journal of World Institute of Pain. doi:10.1111/papr.12116.. PMID 24138020. {{cite journal}}: Check |doi= value (help)CS1 maint: multiple names: authors list (link)

The current text is not accurate. QuackGuru (talk) 21:35, 8 November 2013 (UTC)

Actually, it is accurate. The source distinguishes between acu being cost-effective according to a given WHO benchmark (it is), and acu being cost-effective as a substitute for standard care. Quote:

“ ....According to this threshold, acupuncture as a complement to standard care for relief of chronic LBP is highly cost-effective, costing around $48,562 per DALY avoided. When comorbid depression is alleviated at the same rate as pain, cost is around $18,960 per DALY avoided. Acupuncture as a substitute for standard care was not found to be cost-effective unless comorbid depression was included. According to the WHO cost-effectiveness threshold values, acupuncture is a cost-effective treatment strategy in patients with chronic LBP. ”

Should we leave text as is or elaborate? --Middle 8 (talk) 06:41, 13 November 2013 (UTC)

Current text: "A 2013 meta-analysis found acupuncture as a substitute for standard care was not cost-effective but as a complement to standard care it was cost-effective for chronic low back pain.[169]"

I did elaborate on it but left out the low level details. The old text was not accurate when it was taken out of context. QuackGuru (talk) 17:55, 13 November 2013 (UTC)

A little clearer grammatically, with respect to the "for chronic low back pain" prepositional phrase which modifies both kinds of cost-effectiveness (i.e. as complement and substitute to standard care): "A 2013 meta-analysis found that acupuncture for chronic low back pain is cost-effective as a complement to standard care, but not as a substitute for standard care." --Middle 8 (talk) 19:09, 21 November 2013 (UTC)

That is not how it was summarizes. It first said acupuncture as a substitute for standard care was not cost-effective. QuackGuru (talk) 19:20, 21 November 2013 (UTC)

Wrong, read the quote above, that's the order it uses and is what I followed. Yours was grammatically unclear, as I said. --Middle 8 (talk) 19:35, 21 November 2013 (UTC)

It was a study. Failed MEDRS. QuackGuru (talk) 19:45, 21 November 2013 (UTC)

"The objective of this study was to assess the cost-effectiveness of acupuncture in alleviating chronic LBP either alone or in conjunction with standard care compared with patients receiving routine care, and/or sham."[7] This study does not meet MEDRS. QuackGuru (talk) 19:45, 21 November 2013 (UTC)

The study failed MEDRS. QuackGuru (talk) 19:50, 21 November 2013 (UTC)

Not true; don't be deceived by semantics. The source is a meta-analysis and does meet MEDRS. See the abstract keywords. "Study" is a general word that can be used to refer to meta-analyses, as does the author of the piece quoting Ernsts's comments on Vickers.[8]. --Middle 8 (talk) 19:56, 21 November 2013 (UTC)

"For acupuncture and standard care vs. standard care and sham, a weak positive effect was found for weeks 12 to 16, but this was not significant."[9]

They state "for weeks 12 to 16...". That is what studies do. The source says study not meta-analysis. You have not shown it passes MEDRS. This was OR and a MEDRS violation. QuackGuru (talk) 20:07, 21 November 2013 (UTC)

QG, you're simply wrong, and I hope you self-rv (if applicable). Read the abstract: "The objective of this study was to assess ... To determine effectiveness, we undertook meta-analyses which found ...."

All this is right there in the abstract, the same one that you read and found the word "study" in.

I don't appreciate having to spell this all out to an editor who's been around for years. It is disruptive to IDHT and ignore what the eye can plainly see. If you're having a hard time focusing on stuff, please take a break from editing instead of wasting other editors' time. If you don't have the scientific literacy necessary to know what "study" means, etc., why make things difficult for those who do? WP:COMPETENCE. --Middle 8 (talk) 21:19, 21 November 2013 (UTC)

From the full text of the the article: "Study Design: We sought to establish the effectiveness of the interventions over the comparators described in Table 1 by undertaking a literature review and meta-analysis. The literature review included clinical trials and systematic reviews. Search terms included...". The study is indeed a meta-analysis and thus is consistent with MEDRS. Puhlaa (talk) 23:32, 21 November 2013 (UTC)

This may need expert comment from someone who has access to the full study. The abstract of the study states, "For acupuncture and standard care vs. standard care and sham, a weak positive effect was found for weeks 12 to 16, but this was not significant." Does this mean when an analysis was done of studies that included sham treatment there was no significant effect? Wouldn't that mean the analysis this study used to draw conclusions was not one based on comparison of actual vs sham treatment studies? The abstract states, "we undertook meta-analyses" note the plural. Did the authors of the study cherry pick amongst analyses? What was the methodology of the "meta-analyses" and how did the authors choose which of them to base their results on? In my opinion any citation of this study should include that, "standard treatment with acupuncture vs standard treatment with sham acupuncture showed no significant results." Again I think someone with research expertise needs to examine the full text of this study. To me it is clear that this study is not "A Meta-Analysis" but used multiple analyses to find specific results. - - MrBill3 (talk) 05:36, 22 November 2013 (UTC)

How is that clear to you if you haven't read the paper? --Middle 8 (talk) 12:03, 22 November 2013 (UTC)

Because the abstract spells it out. How could they have used the meta-analysis that showed no significant effect to demonstrate cost effectiveness? Cost effective at having no effect? Also the abstract say, "we undertook meta-analyses" that means multiple and clearly as before the meta-analysis showing no significant effect could not have been used, thus they must have selected out that meta-analysis among the multiple meta-analyses they undertook. All stated in the abstract. I will read the paper at my earliest convenience. - - MrBill3 (talk) 12:19, 22 November 2013 (UTC)

I share your puzzlement at how they found cost-effectiveness without much in the way of effectiveness. I don't think the plural "meta-analyses" likely implies cherry-picking; sounds more like statistics-talk. But sure; TBD (FWIW.... probably not that big a deal as sources go). regards, Middle 8 (talk) 13:31, 22 November 2013 (UTC)

Removed from article & sticking here, pending resolution (re cost effectiveness vis-á-vis efficacy).

“ A 2013 meta-analysis found that acupuncture for chronic low back pain was cost-effective as a complement to standard care, but not as a substitute for standard care. ”

--Middle 8 (talk) 01:31, 23 November 2013 (UTC)

I have added some excerpts from the full-text of the article for your review. It seems to me that this is a MEDRS-compliant source that carried out 3 different meta-analyses and is quite acceptable for inclusion in this article:

• METHODS
• "Study Design: We sought to establish the effectiveness of the interventions over the comparators described in Table 1 by undertaking a literature review and meta-analysis. The literature review included clinical trials and systematic reviews. Search terms included keywords “acupuncture”, “chronic low back pain”, and “nonspecific low back pain”. We searched PubMed from its inception to 30 Jan 2012. The bibliographies of seven systematic reviews were searched for additional references. We included all randomized controlled trials that included acupuncture with manual or electronic stimulation. Clinical trials were included if a comparator of standard or usual care was utilized in the study. Standard care included one or more of the following:... Most of the studies were not relevant to this analysis because of the comparator; for example, studies were excluded that compared acupuncture alone with sham alone or placebo. All studies included in the analysis were randomized controlled trials. We excluded studies without intention to treat analysis."
• "Meta-Analysis Methodology': To estimate the health gain that could be attributed to each intervention, we conducted a meta-analysis to determine the efficacy of the interventions. We undertook three meta-analyses (one for each category in Table 1) to investigate the differential impact of the intervention vs. the comparator using Comprehensive Meta-analysis software (Version 2.2.050. 2009, Biostat Inc., Englewood, NJ, USA... We use effect size as the main measure of efficacy. We calculated standardized mean differences (SMD) using Hedge’s g, because it includes an adjustment to correct for small sample bias and is used in Cochrane Collaboration systematic reviews. Random effects models were applied because of expected heterogeneity. Studies were grouped according to length of follow-up."
• "Cost-Effectiveness Analysis: Cost-effectiveness ratios were determined as the incremental cost of the intervention divided by the incremental benefit and presented as incremental cost ($A)per DALY saved. The incremental cost is defined as the difference between the cost of the intervention (for example, the cost of acupuncture as a complement to standard care) and the cost of the comparator (for example, standard care). Higher incremental cost-effectiveness ratios indicate lower cost-effectiveness (ICER) . The DALY was chosen as the outcome measure of health gain as it captures both morbidity and mortality effects, and the DALY is used in baseline information on health status for Australia..."
• RESULTS
• "Cost-Effectiveness Analyses: The cost-effectiveness analyses are summarized in Table 5. Acupuncture as a complement to standard care is cost-effective, with a mean cost per DALY avoided of $48,562 (90% confidence interval $28,500 to $76,900). Where comorbid depression is alleviated at the same rate as pain, it is even more cost-effective, with a mean cost of $18,960 per DALY avoided (90% confidence interval of $11,100 to $30,000). In general, acupuncture was not found to be cost-effective when used as a replacement for standard care based on the trial conducted by Cherkin (32). However, when comorbid depression is included in the analysis by Haake (33), acupuncture is cost-effective compared with standard care ($62,946-well below three times GDP per capita)."

Puhlaa (talk) 21:39, 23 November 2013 (UTC)

Thank you for those excerpts. I think the article does qualify as MEDRS and supports the statement, thus it can be replaced in the article. Perhaps some caveat or explanation of the fact that there was no difference between sham and non sham should accompany it. The authors did find it important enough to place it in the abstract. - - MrBill3 (talk) 15:45, 26 November 2013 (UTC)

Yes, thanks for those excerpts, and I agree, restoring with explanation makes sense. --Middle 8 (talk) 15:00, 28 November 2013 (UTC)

Rate of serious adverse events

Restored[10] a source from 2004 that estimated a rate of SAE's of 5 in one million. For this area of research, 2004 seems fine, since it's consistent with later reviews like Xu et.al.,2013, which just didn't mention a specific number. Some things change quickly; this doesn't appear to be one. --Middle 8 (talk) 21:35, 21 November 2013 (UTC)

The 2004 is too old. It failed MEDRS. That is your opinion that the source from 2004 is consistent with recent reviews. The recent reviews covering safety did not think it was important to discuss the specific numbers. You were using the older source to argue against the later reviews. You thought the recent reviews were wrong so you used the 2004 source to argue against recent reviews. QuackGuru (talk) 21:46, 21 November 2013 (UTC)

It's not a MEDRS failure for the reasons I stated; it is consistent with recent reviews. There is no basis in fact for your other assertions, nor have you even attempted to provide one. --Middle 8 (talk) 12:06, 22 November 2013 (UTC)

You know there are recent reviews covering safety including the 2013 you mentioned. QuackGuru (talk) 18:00, 22 November 2013 (UTC)

You know that older sources are fine under MEDRS in a lot of situations. A1candidate put it well just below. --Middle 8 (talk) 01:43, 23 November 2013 (UTC)

I haven't got the time to study both articles in-depth, but I think its wrong to remove a source just for being several years old, especially when taking into account the lack of similar reviews published in medical literature. Unless there's an overwhelming increase in the number of acupunctural reviews being published in the previous weeks/months that explicitly contradicts this particular review, I don't see why it should be removed. Scientific consensus usually takes years, if not decades, to be shaped. A 2004 publication year seems to be fine. -A1candidate (talk) 15:37, 22 November 2013 (UTC)

Your comments show that the reference is indeed several years old. There is no compelling reason to ignore MEDRS. QuackGuru (talk) 18:00, 22 November 2013 (UTC)

And still more IDHT from you. I'm restoring the ref. If other editors agree with me, they'll make sure it stays. Don't expect your edits (inclusions or removals) to stick unless you're willing and able to discuss them. --Middle 8 (talk) 01:43, 23 November 2013 (UTC) Follow-up comment: haven't yet put it back in. More good reasons to add the source have emerged below. --Middle 8 (talk) 12:30, 26 November 2013 (UTC)

I agree with Middle 8 - no BS making up criteria on the fly - discuss here first please.Herbxue (talk) 03:09, 23 November 2013 (UTC)

Per WP:MEDRS:

• Look for reviews published in the last five years or so, preferably in the last two or three years. The range of reviews you examine should be wide enough to catch at least one full review cycle, containing newer reviews written and published in the light of older ones and of more-recent primary studies.
• Within this range, assessing them may be difficult. While the most-recent reviews include later research results, do not automatically give more weight to the review that happens to have been published most recently, as this is recentism.
• Prefer recent reviews to older primary sources on the same topic. If recent reviews do not mention an older primary source, the older source is dubious. Conversely, an older primary source that is seminal, replicated, and often-cited in reviews can be mentioned in the main text in a context established by reviews. For example, the article genetics might mention Darwin's 1859 book On the Origin of Species as part of a discussion supported by recent reviews.

According to WP:MEDRS we use recent reviews. Wikipedia does not engage in death pacts with almost ten-year-old sources. We don't have to say anything about the numbers. The recent reviews decided what is important. There is never a requirement that dubious, potentially misleading, but sourced text must appear, simply because an old source mentioned something several years ago when we know the recent reviews did not make the same point. Even if true, there is no reason to use the older source. Some things may change or may not change; but this is not the job of Wikipedians to make that determination. QuackGuru (talk) 06:29, 23 November 2013 (UTC)

MEDRS also says:

• "These instructions are appropriate for actively researched areas with many primary sources and several reviews and may need to be relaxed in areas where little progress is being made or few reviews are being published."

From PubMed, reviews on adverse events in acupuncture, most recent first:

• Xu, 2013-03 (English language, 6 databases, systematic, 25 countries & regions, 2000-2010; sequel to earlier review up to 2000)
• He, 2012-10 (Chinese language, 1956-2010)
• Wheway, 2012-01 (U.K., review of reports to National Patient Safety Agency, 2009-2011)
• Adams, 2011-12 (Pediatric, all languages, 18 databases, systematic, inception-Sept. 2010)
• Ernst, 2011-04 (Serious adverse events, all languages, 11 databases, 2000-ca.2010)
• Zhang, 2010-12 (Chinese language, 3 databases, 1980-2009)
• Capilli, 2010-01 (clinical trials focusing on pain)
• White, 2004-09 (Review of reviews; computerised databases, previous reviews of case reports, population surveys, prospective surveys, textbooks; English language; inception- 1990-2000)) (note: this is the one I propose to restore; includes estimated rate of serious AE's = 5 in 10^6)
• ... and six or seven more, from 1999-2003.

So since White's general review (2004), there were seven reviews, three of which (Wheway, Adama and Capilli) were about subsets of acupuncture care (pediatric, specific databases), and thus not comparable to White. Of the remaining four, two were in the Chinese literature and thus also not comparable to White, since there are fewer adverse effects in the West. That leaves two comparable to White: Xu-2013 and Ernst-2011. That would certainly qualify as a case of "few reviews are being published", i.e., reasons given in MEDRS to relax to five-year window. I think that pretty much settles the issue. Comments? --Middle 8 (talk) 15:30, 23 November 2013 (UTC)

I think you clarified this very nicely. --Mallexikon (talk) 05:03, 24 November 2013 (UTC)

I think the "few reviews are being published" exception is to allow a 5 year plus review when there is not a more recent one, or a more recent one that is comparable available. Why would it be appropriate to use a 10 year old review when there is one that is from this year and one that is from two years ago? That you have shown seven reviews from 2010 to the present pretty much negates the idea that few reviews are being published. But again the relaxing of the 5 year rule IMO is to allow use of a older study when there is not a more current one available.

If the more recent one doesn't have all the data you are looking for I would suggest a statement about Xu 2013 or Ernst 2011 followed by a mention of the data from White 2004. Alternatively you could argue that White 2004 is a superior source as a review of reviews and the most recent source at that level. - - MrBill3 (talk) 07:08, 24 November 2013 (UTC)

I wasn't suggesting ignoring Xu 2013 or Ernst 2011, but rather augmenting them with White 2004's "5 in one million" figure (for rate of serious AE's). (White is otherwise consistent with Xu and Ernst.) I did explain why the other five reviews aren't comparable to White: there are many different places to dig for data, e.g. language and databases, type of AE, population treated, care setting. Given all those variables, reviews are scarce. --Middle 8 (talk) 12:30, 26 November 2013 (UTC)

This proposal to use the fringe journal Acupuncture in Medicine makes no sense. There is information in the article about the numbers. Why include the same information twice in the article? I already explained that the recent reviews cover this. I think I clarified this very nicely. QuackGuru (talk) 18:48, 24 November 2013 (UTC)

Thanks for putting that number in the article, QG, but the source you used -- Adams 2011, the Pediatrics review -- credits it to White 2004; we should cite the latter since it's the original source. The very fact that Adams cites White shows that Acupuncture in Medicine is not a "fringe journal". Its publications are also used by, e.g., the Cochrane Collaboration[11], and Adams and Vickers each cite it multiple times. We should leave the judgement about "fringe-ness" to the literature. (Some of the stuff published by Acupunct. Med. might not pass MEDRS, but that's true of any journal, e.g. case reports). --Middle 8 (talk) 12:30, 26 November 2013 (UTC)

I agree that if Adams is citing a number from White it should be sourced to White. With that change, as the numbers are in the article, have we reached consensus. - - MrBill3 (talk) 15:49, 26 November 2013 (UTC)

See WP:MEDDATE: Look for reviews published in the last five years or so, preferably in the last two or three years.

The fringe journal Acupuncture in Medicine is still a fringe journal. We must use independent sources. There is no reason to use the 2004 source when we have a much newer source for the numbers. We don't cite older sources unless it is a good source like a Cochrane review. QuackGuru (talk) 17:19, 26 November 2013 (UTC)

QG, I think you are too quick with damning sources you don't like, especially with your assessments regarding "fringe". This is a reliable source as pointed out several times above, even though you think anything acupuncture is evil. If you believe the consensus here is unacceptable, please take it to the reliable sources noticeboard. --Mallexikon (talk) 05:00, 27 November 2013 (UTC)

Acupuncture in Medicine is published by BMJ Group, which can hardly be characterized as fringe. TimidGuy (talk) 11:23, 27 November 2013 (UTC)

I pointed out the source is way too old. In 2014, it would be 10 years old. QuackGuru (talk) 19:28, 27 November 2013 (UTC)

If the source (White 2004) is being cited by Adams in 2011 that supports its use as the best review currently available per the MEDRS exceptions discussed above. I agree that blanket condemnation of a journal as a source is inappropriate and / or a matter that should go to the RS Noticeboard. A published article should be evaluated on its merits for a specific article or fact therein. If a fact / number is used, the reference should not be a citation of that number but the original study. The citation of the number in a more recent source is a matter for editorial discussion about including that fact. - - MrBill3 (talk) 11:01, 28 November 2013 (UTC)

@QG: Your concerns about the age and mainstream-ness of the source have already been addressed, QG. No need to go in circles; consensus isn't always unanimity, and by all indications we have a consensus. I'll restore White 2004 per my earlier edit when I get a chance. --Middle 8 (talk) 13:35, 28 November 2013 (UTC)

MrBill3 said "a matter that should go to the RS Noticeboard." I thought you wanted to include the source. Your earlier edit deleted text from a 2011 source that was not about the numbers. There is a big difference between including the source and rewriting the text. You have not given a reason to rewrite the text.

"A 2011 meta-review showed that serious adverse events, including death have been reported, are frequently due to practitioner error, is rare, and diverse.[16]"

I object to deleting this text or replacing it with the dated source. I don't see an argument to replace the 2011 meta-review. QuackGuru (talk) 17:52, 28 November 2013 (UTC)

Again you're repeating yourself and ignoring points made above (e.g., that White is the original source for the 5-on-one-million number, and as such, should be cited). You also mischaracterize my edit, which did not "delete text from a 2011 source"; you may have misread the diff. That can happen, but with you it's been happening a lot. When you keep misreading/misrepresenting policies and edits, it gets disruptive. --Middle 8 (talk) 03:14, 29 November 2013 (UTC)

Your edit replaced the 2011 meta-review and your edit was OR. For example, the "exceedingly" rare was OR. Replacing the meta-review was against MEDRS. Your proposal does not seem to match your edit. You did not give a good reason to replace the the meta-review when it is a newer source. The 2011 meta-review said nothing about the 5-in-one-million number. IMO, your proposal does not make much sense. QuackGuru (talk) 03:34, 29 November 2013 (UTC)

You're misreading that diff. And above (early in this thread) you misread WP:MEDDATE; in fact, the bullet points that you pasted in come right after the "few reviews being published" exception, so I can't see how you could have missed that without being careless or bad-faith. This is disruptive. --Middle 8 (talk) 03:56, 29 November 2013 (UTC)

Where is your explanation for deleting/replacing the 2011 meta-review. I thought your proposal was to include the numbers but you still want to do this? Again, the meta-review is not about the numbers so what reason is there to replace it. Your original proposal was to include the numbers because you thought the newer sources just don't mention a specific number. But the newer sources do. There are two sources that mention the numbers. I included the numbers using one of the newer sources that explicitly states the 5 in 1 million numbers. The source you used does not explicitly state the 5 in 1 million numbers. QuackGuru (talk) 04:29, 29 November 2013 (UTC)

Enough! We have consensus; you've excluded yourself from it with persistent IDHT. And I think we're near the point where we need some local or global user-conduct consideration. Locally, it might simply involve an emerging understanding that you, QG, aren't the most trustworthy editor and therefore your mainspace edits won't stick unless there's explicit consensus; globally, well, the usual. But that may not be necessary unless you continue disrupting the talk page. --Middle 8 (talk) 10:02, 1 December 2013 (UTC)

To be clear, there's nothing personal here, nor am I generalizing unfairly. QG is good at finding sources, and frequently makes reasonable edits. It's just that he also frequently gets sidetracked, makes factual errors (the correction of which he tends to ignore), and IDHT's (intensely). --Middle 8 (talk) 10:44, 1 December 2013 (UTC)

The 2011 meta-review does not discuss the numbers but you did replace the source with an older source. Are you planning to delete/replace the 2011 meta-review again? The newer source that is about the numbers gives a better explanation about the numbers. The newer source does specify the numbers, including children and adults. See Adams 2011. I did explain the older source you wanted to restore does not explicitly state the 5-in-1 million number. QuackGuru (talk) 18:49, 1 December 2013 (UTC)

The White 2004 dated source is confusing because it said "acupuncture is estimated to be 0.05 per 10 000 treatments, and 0.55 per 10 000 individual patients."[12] This shows the older source did not explicitly state it was 5-in-1 million.

"The authors did not specify if this estimate included adults and children,..."[13] according to the newer Adams 2011 review. The newer review does specify the numbers, including children and adults.

Current text: "The incidence of serious adverse events was 5 per one million, which included children and adults.[166]"

The current text in the article is well written using the Adams 2011 review. The older text is indeed confusing and therefore not appropriate for inclusion in this article. QuackGuru (talk) 05:20, 2 December 2013 (UTC)

@QuackGuru: (1) You've just claimed (twice) that it's too confusing to convert 0.05/10,000 to 5/1,000,000. Among science editors. What do you take us for? (2) Re the newer source, you ignore this, from right above: Thanks for putting that number in the article, QG, but the source you used -- Adams 2011, the Pediatrics review -- credits it to White 2004; we should cite the latter since it's the original source. Then you ignore another editor's agreement with that comment. Pretending not to understand math is trolling for sure, and repeated IDHT-ing of comments (in the same thread) is probably also trolling, or else extreme incompetence. I've engaged you this far to allow you a choice, and you've repeatedly chosen disruption. Why should you be trusted? I won't engage your IDHT or "misunderstandings" further. As long as you behave this way, consensus will and should move along without you. --Middle 8 (talk) 07:10, 2 December 2013 (UTC)

@QG: Middle8 has a point. I would usually post this on your user page but you delete everything written there... We all honor your contributions here but I, too, feel kind of frustrated with your style. How about some good ol' WP:AGF from your side? --Mallexikon (talk) 07:28, 2 December 2013 (UTC)

@QuackGuru: I generally find your contentions reasonably well supported. I respect and value your editing and in many cases your opinion. I think you bring a sharp editors pencil to many places it is needed. That said I find your conduct on talk pages disruptive. Your level of IDHT at times defies belief and your follow up comments can be repetitive without any additional information. I am personally offended that you quoted me out of context. I was clear that if you contested White 2004, you should take it to the RS Noticeboard. Your comments and edit summaries all too often show a lack of civility. I encourage you to continue to contribute to WP but urge you to consider behaving in a more genial and constructive manner. Not wanting to get beyond the appropriate boundaries of a talk page I just want to add there are times and states of mind not conducive to editing. I hope all is well with you and look forward to a time when your participation in talk pages matches your acumen in editing. - - MrBill3 (talk) 07:36, 2 December 2013 (UTC)

Belatedly restored White 2004 and text per consensus above. --Middle 8 (talk) 08:10, 27 January 2014 (UTC)

However

The same review concluded that acupuncture can be considered inherently safe when practiced by properly trained practitioners.[168] "The same review also stated: However, there is a need to find effective ways to improve the practice of acupuncture and to monitor and minimize the health risks involved."[14]

The text may not summarise the source accurately. The text does not explain the "however" part. QuackGuru (talk) 19:59, 27 November 2013 (UTC)

Do you have a proposal for explaining the statement? I think one may be appropriate. We should provide a paraphrase and explanation per policy (IMO). - - MrBill3 (talk) 11:06, 28 November 2013 (UTC)

Looks good, QG. --Middle 8 (talk) 15:35, 28 November 2013 (UTC)

"exceedingly" rare is original research

"One might argue that, in view of the popularity of acupuncture, the number of serious adverse effects is minute. We would counter, however, that even one avoidable adverse event is one too many. The key to making progress would be to train all acupuncturists to a high level of competency."[15]

A 2011 meta-review showed that serious adverse events, including death have been reported, are frequently due to practitioner error, is rare, and diverse.[16]

I removed the original research. I recommend we keep this source. A 2004 cumulative review should not be used to replace the 2011 meta-review. See WP:MEDDATE. QuackGuru (talk) 18:09, 28 November 2013 (UTC)

I think you might be pushing the definition of OR too far. 5 in one million can fairly be called "exceedingly rare", IMO, compared to the rates of SAE's in other areas. How else to summarize it without just saying it directly (which is OK, but sometimes grammatically unwieldy)? --Middle 8 (talk) 02:30, 29 November 2013 (UTC)

The 2011 meta-review did not cover say anything about the 5 in one million. I read the full text. Please don't try to delete this 2011 meta-review against MEDRS. You don't have consensus to delete the 2011 meta-review. The current text is now accurate. QuackGuru (talk) 03:21, 29 November 2013 (UTC)

White 2004 -- remember that one? -- is the source for the 5-in-one-million number, and you're misreading the diff. Your "not getting it" is classic IDHT, and the effect (intentional or not) is trollish. Nobody wants to haggle over distracting misunderstandings. This is just too disruptive. --Middle 8 (talk) 04:00, 29 November 2013 (UTC)

"A 2004 cumulative review showed that serious adverse events (SAE's) are frequently due to practitioner error, exceedingly rare, and diverse.[168]" The dated source was added by Middle 8.

"A 2011 meta-review showed that serious adverse events, including death have been reported, are frequently due to practitioner error, is rare, and diverse.[16] Current text.

I object to using the 2004 cumulative review. Obviously no reason was given in this thread to replace the 2011 meta-review. I support the current text using the newer source. I object to the MEDRS violation. QuackGuru (talk) 04:29, 29 November 2013 (UTC)

More repetition and IDHT. Please stop disrupting this page. --Middle 8 (talk) 10:03, 1 December 2013 (UTC)

Where is your explanation for deleting the 2011 meta-review against MEDRS? This source was not about the numbers. Did you make a mistake when you edited the article? QuackGuru (talk) 18:26, 1 December 2013 (UTC)

You misread the diff. --Middle 8 (talk) 05:43, 2 December 2013 (UTC)

Seems like both sources are fair game, which means that "exceedingly rare" is not original research. Quack Guru, please don't insult everyone else's intelligence. Just add good new info if it is sourced well, but don't dick around with the other good, well-sourced material before discussing it here first. Herbxue (talk) 05:51, 2 December 2013 (UTC)

Both sources are not fair game according to WP:MEDDATE. I don't see a good reason to delete the text from the newer 2011 review and replace it with text from a dated 2004 source. It seems like you did not provide verification for the text that failed verification. The part "exceedingly" is indeed unsourced when no editor provided verification. QuackGuru (talk) 05:59, 2 December 2013 (UTC)

There are separate issues here: First,whether the 2004 source is reliable or not. The link you provided is in a section titled "basic advice" not "strict policy" - the goal is reliability. Does the 2011 source meaningfully and convincingly contradict the 2004 source? No, they both basically say acupuncture is generally safe, but you prefer the wording of the 2011 source. That is the second issue: Does the more recent source automatically trump the older source? Not really but of course the newer source should be included. I am not for deleting new sourced material in favor of old, but deleting material that is still valid and not specifically disproved by recent work should be discussed here first.Herbxue (talk) 08:17, 2 December 2013 (UTC)

Agree. --Mallexikon (talk) 08:21, 2 December 2013 (UTC)

Agree, and just to be clear: contrary to QG's repeated description of my edit that added White 2004, I didn't delete the Ernst 2011 review, I simply moved it to a place where it better fit the text. Ernst 2011 doesn't discuss and shouldn't be cited for rate of SAE's; what he does is give a raw number for SAE's -- that is, a numerator but no denominator. (And as mentioned, his # is consistent with White and other reviews.) The first sentence appearing in the diff was originally sourced to White; I know, because I wrote it, and consensus at the time supported it. Today, all but one of us seem to be fine with including White 2004, so I will redo/improve upon this edit. (Re MEDDATE's guidelines -- the bullet points that say to use < 5.y.o. reviews happen to come immediately under the sentence that says we can user older ones when not many are extant.) struck; already made that point; once is enough... don't feed the IDHT --Middle 8 (talk) 10:14, 2 December 2013 (UTC) (edited 13:11, 2 December 2013 (UTC))

Use this primary source?

An editor recently added material on the cost-effectiveness of acupuncture that was sourced to this large RCT conducted in Germany and published in 2008 in the European Journal of Health Economics. [16] There are some errors in the material added, but I don't want to take the time to fix the errors if the material will eventually be deleted due to this being a primary source. Thanks. TimidGuy (talk) 12:05, 5 December 2013 (UTC)

The previous text was: A 2008 German study found acupuncture was cost-effective for chronic osteoarthritis pain.[175] There are other reviews in the Acupuncture#Cost-effectiveness section. Is there a reason to keep this primary source? QuackGuru (talk) 22:19, 26 December 2013 (UTC)

Significance

Just looking at Acupuncture#Pain. I'm finding an unacceptable vagueness in our portrayal of the significance of treatment effects. Can someone who has access to the Saudi review [17] please tell me whether they are using "significant" in their conclusion to refer to statistical significance or clinical significance, and whether they (or anyone else commenting on their review) say anything about the clinical significance of the difference between acupuncture and control? --Anthonyhcole (talk · contribs · email) 05:34, 8 December 2013 (UTC)

Without looking at that source, there's very few trials caring about clinical significance. It's not usually a term appearing in efficiency discussions. --Mallexikon (talk) 05:53, 8 December 2013 (UTC)

In my limited experience, recent good medical studies and reviews make it very clear whether they are using "significant" to mean statistical or clinical. Given the vast difference in meaning, it's important that we convey to the reader which meaning to assign to the term wherever we use it. --Anthonyhcole (talk · contribs · email) 06:12, 8 December 2013 (UTC)

I'll see if I can get the Saudi review. TimidGuy (talk) 15:46, 8 December 2013 (UTC)

Their meta-analyses found that acupuncture was both clinically and statistically significantly better than sham acupuncture, usual care, and wait list controls. I can send it to you, if you like. TimidGuy (talk) 17:18, 8 December 2013 (UTC)

You've got mail. :o) --Anthonyhcole (talk · contribs · email) 18:24, 8 December 2013 (UTC)

I went ahead and added that the reduction was clinically significant. TimidGuy (talk) 11:41, 12 December 2013 (UTC)

New meta-analysis

• PMID 24146995
• Some commentary from Ernst

Alexbrn ^{talk|contribs|COI} 08:22, 15 December 2013 (UTC)

Commentary from individuals

I had a look at this article based on a discussion in WP:MED and found this article's use of quotations very odd. In particular, it seems to selectively quote from individuals in several places, which not only is an issue under WP:UNDUE and WP:MEDRS but also somewhat strange to read in a medical article. Here is a list of the quotations:

• "A report for CSICOP on pseudoscience in China written by Wallace Sampson and Barry Beyerstein said:...". This particularly concerns me: "A few Chinese scientists we met ..." (?!!)
• "Acupuncturist Felix Mann, who is the author of the first comprehensive English language acupuncture textbook Acupuncture: The Ancient Chinese Art of Healing,[73] has stated in lectures that ..."
• "A delegation of the Committee for Skeptical Inquiry reported in 1995:[48]"
• "but it remains controversial among medical researchers and clinicians.[10] In 2006, a BBC documentary Alternative Medicine filmed a patient "
• "Psychologist John Jackson stated that, "Acupuncture, both as a field of study and in practise, contains many of the hallmarks of pseudoscience. It is based on untenable principles and the small amount of evidence there is to support its use in pain relief can also be called into question"
• "Steven Salzberg, director of the Center for Bioinformatics and Computational Biology and professor at the University of Maryland stated that,"
• " Steven Novella, Yale University professor of neurology, and founder and executive editor of the blog "
• "However, Edzard Ernst, emeritus professor of complementary medicine at the University of Exeter, disagreed with the study's conclusions "
• "The consensus statement and conference that produced it were criticized by Wallace Sampson..."
• "Brent Bauer comments on the Mayo Clinic website that it "it may be worth trying acupuncture" if other, more conventional treatments haven't helped in the treatment of low back pain, while noting that there was no difference in effectiveness between sham and real accupuncture.[187]"

Acupuncture may be effective in some domains, however there appears to be enough medical evidence (WP:MEDRS) for the lack of efficacy in many areas that many of these selective quotes could be replaced with sources more in keeping with WP:MEDRS. Kind regards, --LT910001 (talk) 11:04, 16 December 2013 (UTC)

Thanks for your input. I'm not sure if you a aware of the situation here though... We do have enough MEDRS for both effectiveness of acupuncture in some fields and non-effectiveness for other fields already. However, most skeptics still find the article somehow too pro-acupuncture. Almost all the quotes you listed above were included in order to satisfy the skeptics camp and to present more criticism. I don't see how we can remove these quotes without starting another edit war. --Mallexikon (talk) 11:25, 16 December 2013 (UTC)

Have flagged this on WP:MED. I understand this may be a matter of some contention, but I would point out that MEDRS is not a threshold for an article as a whole, but a guideline for the use of every single source in a medical article. Sources from individuals are not reliable medical sources, and including full quotes from such sources is an example of WP:UNDUE. There appears to be enough reliable sources in the article that reliable sources and quotes could be substituted instead of these claims. --LT910001 (talk) 12:57, 17 December 2013 (UTC)

To be picky MEDRS is not quite "a guideline for the use of every single source in a medical article" but (to quite the guidance) for "biomedical information in all types of articles". Articles on "medical" topics may contain material that doesn't need MEDRS sources, and conversely articles on "non-medical" topics may contain health information which does. Alexbrn ^{talk|contribs|COI} 13:09, 17 December 2013 (UTC)

Editorial discretion is involved in articles. For example, discredited studies from dated German acupuncture trials are allowed to remain in an article. Editors know the primary sources fails WP:MEDRS and low levels details fails WP:WEIGHT. It was is not simply background information dumped into the coat rack article. See WP:COATRACK. QuackGuru (talk) 16:51, 17 December 2013 (UTC)

I think LT910001 made his case very clear and there's actually no rational argument against it. I'll start removing the material he cited. --Mallexikon (talk) 07:59, 22 December 2013 (UTC)

Absolutely not, on ALL counts. LT910001 is wrong about the application of MEDRS, and you would do well to not remove any properly sourced material without consensus. Alexbrn is quite correct about MEDRS. -- Brangifer (talk) 08:11, 22 December 2013 (UTC)

You've got to be f****g kidding! You make such an absurd statement of intention, and one minute later you actually dare to do it! Have you no idea what vandalism is? Deletion of properly sourced material amounts to vandalism. Don't do it. EVER. Discuss your intentions first, and wait (a long time) to see what reaction you get. You must have extremely good reasons for deleting properly sourced material, and usually it's better to tweak the content, rather than delete it. Violating NPOV by deleting opposing POV is very wrong. -- Brangifer (talk) 08:19, 22 December 2013 (UTC)

And you do it again, after I had reverted you with a BRD notification. To make sure the record is clear, I have notified you about edit warring/vandalism on your talk page. -- Brangifer (talk) 08:29, 22 December 2013 (UTC)

Comment - Agree with User:LT910001. Commentary fails WP:MEDRS -A1candidate (talk) 08:58, 22 December 2013 (UTC)

They are flat out wrong. You don't understand MEDRS. It does not cover all content, only direct medical claims. Commentary, opinions, and controversies are not covered by MEDRS. -- Brangifer (talk) 09:15, 22 December 2013 (UTC)

@Brangifer: I made 2 different edits based on the rationales raised here and at Wikipedia talk:WikiProject Medicine#Arbitrary section break - rationales that show very clearly that the material is not properly sourced. Now you've both used profanity ("You've got to be f****g kidding!") and accused me of vandalism. WP:IDENTIFYUNCIVIL comes to mind, and I feel that you try to intimidate me. Would you bother to apologize? --Mallexikon (talk) 09:18, 22 December 2013 (UTC)

It would be stretching it to say this is covered by MEDRS; this is really a case of two academics having a slight difference of opinion on the emphasis given in presenting a conclusion, and not directly "biomedical information" as such. As has been pointed out before, Ernst is not disagreeing with Vickers et al. They say "... acupuncture is more than a placebo. However, these differences are relatively modest ..." and Ernst says "... the effects of acupuncture were principally due to placebo ..." (my emphases). As expert opinion it seems perfectly okay to me as a gloss. I don't see any strong case for its removal (and certainly not without consensus: there's too much edit warring going on here). Alexbrn ^{talk|contribs|COI} 09:22, 22 December 2013 (UTC)

Exactly. Discussion is good, edit warring is bad. Mallexikon, I have explained quite clearly on your talk page why I didn't consider your edits vandalism, even though the appearance and effect was the same, and that continuing to make such deletions would be considered vandalism, since you are now warned. Otherwise we're not children. -- Brangifer (talk) 09:37, 22 December 2013 (UTC)

Edit warring is bad; would encourage consensus-building where possible. For editors who have arrived from WP:MED, I have expressed a previous opinion about the article here (Wikipedia_talk:WikiProject_Medicine#Concerns_of_LT910001) that elaborates on my concerns. --LT910001 (talk) 10:05, 22 December 2013 (UTC)

The comments in question seem to be a clear reflection and exposition of the mainstream scientific consensus and are properly sourced. They are not the type of specific "biomedical information" referred to in MEDRS. They should not be removed unless and until there is material that indicates this is not the mainstream scientific consensus. WP policy is clear that the mainstream scientific consensus is due most weight and should be clearly expressed. Even valid MEDRS of details and results that contradict the consensus should be constricted by due weight. Comments and explanations that reflect the mainstream scientific consensus belong in the article and should not be removed.

Having presented my opinion I urge collegial consensus building. - - MrBill3 (talk) 12:55, 22 December 2013 (UTC)

MrBill3, regarding your statement here: "They should not be removed unless and until there is material that indicates this is not the mainstream scientific consensus." - There are tonnes of mainstream academic sources indicating that the material being added is a false claim. I would be happy to show my sources to you, as well as anyone else who is willing to listen. -A1candidate (talk) 15:33, 22 December 2013 (UTC)

My sense from having looked at research reviews is that mainstream consensus is generally that acupuncture reduces pain. See, for example, Berman in the New England Journal of Medicine, which recommends it for low back pain. I don't think there's mainstream consensus that it's the same as a placebo, since some meta-analyses find it the same, others find it statistically significantly better than a placebo, and at least one recent meta-analysis finds it clinically significant compared to a placebo. We should be using these sources, not what someone says in a newspaper. TimidGuy (talk) 15:45, 22 December 2013 (UTC)

Alexbrn is correct regarding the application of MEDRS. A1candidate, if you have sources we are not using, please do present them. Thanks!   — Jess· Δ♥ 16:32, 22 December 2013 (UTC)

Ernst's expert opinion is the mainstream POV. It must be kept to maintain NPOV. If editors want to tackle the real MEDRS violations they should look here. QuackGuru (talk) 19:50, 22 December 2013 (UTC)

@A1candidate: First "the material being added" material is not being added. Second "is a false claim" please be specific what is being claimed in which statement and what is said in what source that falsifies it. Please refrain from invoking tonnes of sources without specifying the source and the content that supports your assertion. Remember as has been explained what is being discussed is the mainstream scientific consensus.

@TimidGuy: Please give citation for Berman that you refer to. Mainstream consensus seems reflected by the included comments. A few analyses wavering between equal to placebo or somewhat better than placebo do not seem to have swayed mainstream consensus. However if as you say NEJM has an article actually recommending acupuncture as treatment that should be included. Expert commentary that represents mainstream consensus belongs in the article if you have multiple sources that seem to indicate mainstream consensus is different from the comments in the article, provide the sources, identify the comments they contradict and propose new content. - - MrBill3 (talk) 00:35, 23 December 2013 (UTC)

There does not appear to be the consensus among users for inclusion of certain content, regardless of whether the content does or does not reflect scientific consensus. On my current read of this situation: --LT910001 (talk) 00:47, 23 December 2013 (UTC)

• In support of inclusion: QuackGuru, MrBill3, BullRangifer, Alexbrn)
• Objecting to inclusion: LT910001, TimidGuy, A1candidate, Mallexikon

As pointed out in my comment to A1candidate. This is not a discussion about inclusion or not. This is a discussion started by LT910001 objecting to content that has been present in the article for quite some time. Consensus was reached when the material was added and has stood until this recent objection. The lengthy bulleted list represents 8 comments, specific edits regarding each comment have not been discussed. So far I have yet to hear a policy based argument that these comments do not represent mainstream scientific consensus either. - - MrBill3 (talk) 01:01, 23 December 2013 (UTC)

Can you point me to any evidence of consensus on the talk page? I have had a very quick look through the archives and can't find the consensus to which you refer. --LT910001 (talk) 01:10, 23 December 2013 (UTC)

MrBill3, the false claim is this:

"Commenting on this meta-analysis, professor Edzard Ernst stated that it demonstrated that the effects of acupuncture were principally due to placebo."

Ernst's commentary is a fringe viewpoint that is not accepted by the mainstream academic community. Are you willing to let me show you why? -A1candidate (talk) 01:24, 23 December 2013 (UTC)

A1candidate, you're forcing me to repeat myself. If you have sources, present them. Please stop edit warring. You don't need to request permission to state your case, and even if you felt you did, you shouldn't be edit warring to your preferred version in the meantime. Either demonstrate what you're claiming, or move on, please. Well sourced content should not be removed from the article, amid opposition on talk, without attempting to reach consensus first.   — Jess· Δ♥ 01:45, 23 December 2013 (UTC)

What he said. - - MrBill3 (talk) 01:52, 23 December 2013 (UTC)

1.) The only rationale I've seen here for including Ernst's comment is that it's an "expert opinion reflecting mainstream consensus". Well it's definitely an expert opinion, but can you prove with a MEDRS that it's also "mainstream consensus"? 2.) Vickers study is a scientifically well done meta-review. Ergo, it's a MEDRS. Ergo, its conclusions already are mainstream consensus (because they're scientifically proven). The question here is, do we want to set a precedent? And have some professor's commentary added to the conclusions of every meta-review, just because those conclusions don't fit some people's world view? --Mallexikon (talk) 02:33, 23 December 2013 (UTC)

I've just added the result of a more up-to-date meta-analyasis of the same data set, from a less conflicted source (i.e. not the Acupuncture Trialists' Collaboration). This too is emphasizing "little evidence" between sham and true acupuncture (and some interesting things besides). As I said above, there is really no great dispute here: all the sources agree in substance and only differ in how they emphasize certain aspects of their findings.

Incidentally, this whole section is very poor content: largely a shopping-list of undigested conclusions from papers. It would be better replaced with a paragraph or two simply saying what needs to be said here: MrBill3 said it nicely: "A few analyses wavering between equal to placebo or somewhat better than placebo". Alexbrn ^{talk|contribs|COI} 02:47, 23 December 2013 (UTC)

@Mallexikon: 1)I think the more recent meta analysis and edit by alexbrn validate Ernst's commentary and place it in appropriate context. MEDRS do no provide mainstream consensus they provide biomedical information. 2) "Ergo, its conclusions already are mainstream consensus" is simply not true. Mainstream scientific consensus is built over time across a broad spectrum of studies, replication, analyses and exploration of mechanisms as well as meetings of professional organizations etc. etc. A single meta analysis does not automatically create mainstream scientific consensus with its results. In this particular case another meta analysis was done and the results were different. - - MrBill3 (talk) 02:58, 23 December 2013 (UTC)

1.) The new study Alexbrn included is not helping us here with the Ernst comment problem. This new meta-analysis might use the same data set as Vickers (I haven't checked whether that's true) but it doesn't look for efficiency difference between verum and sham acupuncture - it takes the efficacy of acupuncture as a fact ("Background - Recent evidence shows that acupuncture is effective for chronic pain...") and only studies in which way the characteristics of verum acupuncture influence the outcome. Please read the whole text and not just the summary (the cherry-picked quote "When comparing acupuncture to sham controls, there was little evidence that the effects of acupuncture on pain were modified by any of the acupuncture characteristics evaluated" can be a little misleading... It gets clearer in their conclusion: "There was little evidence that different characteristics of acupuncture or acupuncturists modified the effect of treatment on pain outcomes.) This study doesn't belong into the "efficiency" subsection, since it's not an efficiency analysis.

2.) Alright, I find your definition of mainstream consensus "built over time across a broad spectrum of studies, replication, analyses and exploration of mechanisms as well as meetings of professional organizations etc. etc." acceptable, however, Ernst's comment doesn't fit it. It's a single man's single comment on a single meta-review, and we shouldn't set this kind of precedent. --Mallexikon (talk) 04:09, 23 December 2013 (UTC)

Mallexikon, thank you for your considered response. I concede that I should take some time and read the new study in its entirety and that I have not done so. IF (and it seems there is some contention) Ernst's comment is expert commentary reflecting mainstream consensus it should remain and is by no means a new precedent. A1candidate is presenting some information to support the contention that Ernst does not reflect mainstream consensus. I think some material to support that it is mainstream consensus is reflected in one of the sources I provided in the prior discussion of Vickers v. Ernst. I will recheck that and await possible additional material from A1candidate. As I said a full reading of the newer study may help clarify the issues for me. - - MrBill3 (talk) 04:42, 23 December 2013 (UTC)

I think Mallexikon is right about the thrust of the newer study not being directed at the sham/real difference, and have clarified accordingly. Nevertheless, its emphasis on this issue seems aligned with Ernst's: "Given the results of the primary research [1] showing small differences between real and sham acupuncture, it is not surprising that the current analysis showed little evidence of substantial differences between alternative approaches to acupuncture". Alexbrn ^{talk|contribs|COI} 04:48, 23 December 2013 (UTC)

I'd like to emphasize once more that this study has the statement "Recent evidence shows that acupuncture is effective for chronic pain" in its Background section. Regarding the discussion about what the mainstream consensus on acupuncture efficiency is, this sentence is actually the closest thing to evidence I've seen here so far. --Mallexikon (talk) 06:29, 23 December 2013 (UTC)

I think the National Health Service is the most mainstream-ish source I've seen. WP:MEDRS "in a nutshell" says to look for "reputable medical journals, widely recognised standard textbooks written by experts in a field, or medical guidelines and position statements from nationally or internationally recognised expert bodies". Isn't the NHS an example of the latter? The NHS says: "There is some scientific evidence that acupuncture is effective for a small number of health conditions." This exquisitely hedged statement is an almost perfect summary of where the literature stands. The lede has been wobbling (example diff, not necessarily illustrative) around language like: "Existing evidence { is consistent with | does not rule out } acupuncture being no more effective than a placebo". Equally, existing evidence is consistent with there being a small effect beyond placebo. It sounds like we need to come up with a non-weaselly way of saying, in the lede, that there's disagreement over whether or not it's a placebo -- like I just did, maybe -- as opposed to playing the semantic glass-half-full-or-empty word games. --Middle 8 (talk) 10:26, 23 December 2013 (UTC)

Middle 8, I like your reasoning. Would you please formulate a wording we could use? We might be able to hammer out a compromise here. -- Brangifer (talk) 16:11, 23 December 2013 (UTC)

Extended discussion

@MrBill3 and Mann_jess - The challenge for me, really, is to decide what type of sources to present to both of you, whether it is the consensus of international health authorities, national health authorities or mainstream medical journals. For starters, let me quote from a mainstream medical textbook for physiology students:

"...Because the Chinese were content with anecdotal evidence for the success of AA (acupuncture analgesia), this phenomenon did not come under close scientific scrutiny until the last several decades, when European and American scientists started studying it. As a result of these efforts, an impressive body of rigorous scientific investigation supports the contention that AA really works (that is, by a physiologic rather than a placebo/psychological effect)..."

— Human Physiology: From Cells to Systems (2013)

My experience of editing this page and my previous debates with many of the same editors have taught me that in most cases, it is futile to even start a discussion for various reasons. Having said that, I sincerely hope that both of you keep a fair and open mind, and I believe that both of you are willing to do so. Otherwise, we'll just be wasting all of our precious time. If you have any honest questions or issues which you would like to raise, I'll be happy to address them. -A1candidate (talk) 02:49, 23 December 2013 (UTC)

This source seems to present at least marginal support for your contention. I see two substantial issues with it. First is that the material is provided in "boxed content" with the heading "Concepts, Challenges and Controversies" and the title "Acupuncture: Is It For Real?" it is not presented as mainstream consensus or accepted biomedical information included in the text. Second it also states, "In the United States, AA [Acupuncture Analgesia] has not been used in mainstream medicine [...]". That seems to contradict your assertion. Also the support it cites is for the "acupuncture endorphin hypothesis" not for qi, meridians or much of what acupuncture is purported to do.

@A1candidate: Thank you for presenting this information and I look forward seeing what other sources you (or others can provide). As for your quandary regarding what information to provide I suggest you start with the strongest, particularly those which explicitly state there is a consensus in the scientific community. You addressed your comment to myself and another editor and you stated your experience with "many of the same editors". Two specific/many of the same? If you are frustrated by prior seemingly futile attempts you should have used the appropriate processes. In posting here and now please proceed with good faith and an assumption of such. In other words skip the complaints about the past, hope for fair and open mindedness and concerns about wasting time and stick to discussion of editing content and the policies that apply. - - MrBill3 (talk) 05:36, 23 December 2013 (UTC)

"Several Cochrane reviews of acupuncture for a wide range of pain conditions have recently been published. All of these reviews were of high quality. Their results suggest that acupuncture is effective for some but not all types of pain." Source: Lee, MS; Ernst, E (2011). "Acupuncture for pain: An overview of Cochrane reviews". Chinese Journal of Integrative Medicine 17 (3): 187–9. [18] --Mallexikon (talk) 07:31, 23 December 2013 (UTC)

@Mallexicon, ahem, that's a fringe journal ... oops... gulp... ZOMFG it's Ernst.... (brain explodes). (/snark) As Alexbrn linked to above, Ernst himself thinks acu is a placebo (doesn't he?). And scientific consensus may well end up there. But sci consensus lags the opinions of whoever turn out to be right, just as WP lags the science. I don't think the article should assume consensus in any direction, although it's amusing to see (as we sometimes do) better MEDRS's for the it's-not-just-a-placebo side than the yes-it-is side. --Middle 8 (talk) 10:45, 23 December 2013 (UTC)

I wouldn't call the text A1Candidate cites a sci-consensus-level source; only statements from scientific academies and comparable bodies rank at that level (see Scientific opinion on climate change and List of scientific societies explicitly rejecting intelligent design for examples of such sources). But is an entirely acceptable MEDRS, and on the high end (i.e. closer to a meta-analysis of blinded RCT's than to a statement of expert opinion, because it's from a mainstream textbook and therefore has substantial editorial oversight). --Middle 8 (talk) 10:45, 23 December 2013 (UTC)

MrBill3 - Whether it's "boxed content" or non-boxed content seems to be a rather trivial issue to me. I don't think we should be arguing over how the authors decided to design their textbooks. Either the source is reliable, or it isn't. As for the part about acupuncture in the United States, the authors used the past form ("has not") instead of the present form ("is not") because they assert that the field of acupuncture has been accepted in recent years. Nobody (neither me nor the textbook authors) is claiming that acupuncture is scientifically proven to be based on meridians or whatsoever. -A1candidate (talk) 18:06, 26 December 2013 (UTC)

Primary source

At the present, our article includes this: "A delegation of the Committee for Skeptical Inquiry reported in 1995:[48] 'We were not shown acupuncture anesthesia for surgery, this apparently having fallen out of favor with scientifically trained surgeons. Dr. Han, for instance, had been emphatic that he and his colleagues see acupuncture only as an analgesic (pain reducer), not an anesthetic (an agent that blocks all conscious sensations).'" This is a primary source, and it's anecdotal too. If nobody objects, I'd delete it. --Mallexikon (talk) 02:56, 23 December 2013 (UTC)

This seems to be a third party source reporting on the mainstream scientific consensus (msc). It does seem very weak as it says, "apparently" and cites the statement of a single "Dr. Han" speaking about "he and his colleagues". I think if this is a representation of msc it should have other sources that support it. I do think it is important to get some clarity and solid sources establishing msc. The argument could be made that the bulleted list is an assembly of data synthesized into an assertion of msc as OR. I'll try reading the source of this particular content and see if it warrants remaining. If you want to remove this content please paste the full content you remove to this discussion including ref. - - MrBill3 (talk) 05:01, 23 December 2013 (UTC)

I know you mean well, but you keep mentioning a red herring, and that is that a skeptical source must be describing mainstream scientific consensus (MSC). Drop that red herring, because it will surely lead you astray and confound the issues. These statements are not coming from alternative medical sources, but are skeptical opinions and commentaries from the same side as mainstream science. As such, statements from skeptical organizations and notable individuals are perfectly acceptable content, even if they were to disagree with the far from fully formed consensus.

There is no unified scientific consensus, other than the situation that acupuncture is still classified as alternative medicine, ergo fringe, and thus not fully accepted. The current state of flux (over 50 years!), with occasional statements of support, then disappointment over another failure of acupuncture to conclusively work, etc., leaves scientific skeptics as the unflinching watchdogs who tend to be very hardline in their demand for clear scientific evidence of effect before acceptance. That's why one finds many notable skeptics who will not accept some statement of seeming acceptance from a mainstream scientist or textbook (the author's opinion), because they know from experience that such statements often lead to a dead end. They want stronger evidence, and their doubts are perfectly useful as content here.

So, please stop demanding that skeptics toe the line of some non-existent mainstream consensus. They won't do it until the evidence is unequivocal, and then they will reclassify acupuncture as no longer alternative, but mainstream. This article, per NPOV, is required to document much about acupuncture, including the whole spectrum of notable opinions from believers, scientists, and skeptics. Efforts to delete such opinions are a form of whitewashing and are not allowed. We need all of it. This article should not paint a picture documenting a unified opinion, because that is not the real world situation. It must describe all the varying conflicting POV. -- Brangifer (talk) 07:38, 23 December 2013 (UTC)

I agree that as an encyclopedia we must include multiple sources of information from multiple points of view, and appreciate your acknowledgment of a lack of acceptance of what "mainstream" means. But skeptics continue to reference "mainstream" opinion the same way that alternative proponents say "you can't measure my woo" - it is special pleading. We either accept all primary sources or none - and we duke it out for consensus on each individual source. After the 1990's, "mainstream" is a meaningless term. We are in full "po-mo" times and there are facts OR opinions. What skeptics want is for Barrett and Ernst to be installed as unquestionable authorities on all alternative med topics, and that is simply not acceptable (and probably not even "mainstream"). Herbxue (talk) 08:00, 23 December 2013 (UTC)

@Brangifer: I think you have presented the issue quite clearly and completely. WP does emphasize that weight should be determined by MSC but your summary of that I believe is entirely accurate to whit: Acupuncture is classified as alternative and despite flux it remains alternative not mainstream. As there is flux the article should include the whole spectrum of notable opinions with weight towards the fact that acupuncture is not an established, validated, accepted treatment. I agree that attempts to remove such commentary are whitewashing. I do not insist that notable opinions reflect MSC I instead assert they are due weight as reflecting MSC (which I agree is not a comprehensive or unified position beyond the aforementioned alternative, non mainstream status of acupuncture). Evidence for efficacy is adequately discussed in the article already and the comments of those who exemplify the lack of acceptance as a standard treatment should remain in the article. Thank you for sharpening the focus of this discussion and pointing out the clear attempt to remove notable comments as whitewashing. Biomedical claims are already discussed in detail with MEDRS, the broader context and ongoing flux require the inclusion of skeptical viewpoints with the weight they are due.

@Herbxue: The skeptics v alternative proponents is a sidetrack. WP policy is the reason mainstream scientific opinion carries weight in WP articles. Your contention " We either accept all primary sources or none" is a false dichotomy again WP policy is what we go by here. Often we do "duke it out for consensus on each individual source". Primary sources may be used but with care. However Ernst commenting on a study is a secondary source as is the CSI delegation reporting on acupuncture in China. Barret and Ernst are qualified experts on alternative medicine and their published comments are precisely the "author's own thinking based on primary sources" in the WP definition of secondary sources. - - MrBill3 (talk) 08:38, 23 December 2013 (UTC)

No, I beg to differ: the CSI delegation report is definitely not a secondary source. I just read it again; this is a report about their own trip to China - as primary as a source can be. In regard to Brangifer's reasoning: sure, acupuncture is not universally accepted (yet?), and we have to keep this article balanced. But that doesn't mean that you get a blank check to include every anecdotal primary source which fits your world view. The CSI report is not tenable. --Mallexikon (talk) 10:20, 23 December 2013 (UTC)

???? You still don't get it. It's not from some unknown, non-expert, person's private little blog. THAT would be a primary source we wouldn't use. This is published in Skeptical Inquirer (SI). It's just as valid a secondary source as if it was published in the New York Times. The way you throw around "primary" shows you seem to still be stuck in your misunderstanding of MEDRS, which doesn't apply to this type of source and content. If you want to discuss "primary source" in the context of MEDRS, then fine, because we don't use "primary scientific research" as sources. The SI article is not primary scientific research, and it's not covered by MEDRS. -- Brangifer (talk) 16:31, 23 December 2013 (UTC)

Well, no, I'm not talking about MEDRS. I'm talking about WP:RS in general. Which says "Primary sources are often difficult to use appropriately. While they can be both reliable and useful in certain situations, they must be used with caution in order to avoid original research. While specific facts may be taken from primary sources, secondary sources that present the same material are preferred. Large blocks of material based purely on primary sources should be avoided." And WP:PRIMARY states "A primary source may only be used on Wikipedia to make straightforward, descriptive statements of facts that can be verified by any educated person with access to the source but without further, specialized knowledge."

If you have any problems seeing the difference between primary and secondary sources, just lemme know, I'll be happy to explain it to you. --Mallexikon (talk) 01:06, 24 December 2013 (UTC)

Let's see....here's the simple version: I write something on my blog. That's the primary source. (We are allowed to use them carefully, and pretty much required to use them in articles about the person, or to get their own statements.) It gets published in a magazine article. That article is now a secondary source. That article then gets quoted in a third article. That's a tertiary source. (We stop counting after that.)

Your reasoning would exclude the use of most sources at Wikipedia. -- Brangifer (talk) 06:09, 24 December 2013 (UTC)

Ok, so that's where your misunderstanding lies... As per WP:PRIMARY: "Primary sources are original materials that are close to an event, and are often accounts written by people who are directly involved. They offer an insider's view of an event, a period of history, a work of art, a political decision, and so on... An account of a traffic accident written by a witness is a primary source of information about the accident; similarly, a scientific paper documenting a new experiment conducted by the author is a primary source on the outcome of that experiment." Since CSICOP wrote an article about their own trip, it's primary - whether it's published or a blog doesn't matter.

A secondary source, as per WP:SECONDARY is defined as the "author's own thinking based on primary sources, generally at least one step removed from an event. It contains an author's interpretation, analysis, or evaluation of the facts, evidence, concepts, and ideas taken from primary sources... A book by a military historian about the Second World War might be a secondary source about the war, but if it includes details of the author's own war experiences, it would be a primary source about those experiences."

Glad we could sort this out :) --Mallexikon (talk) 05:12, 25 December 2013 (UTC)

Well, I consider QW and CSICOP (or whatever it is) to be basically blogs (someone has an opinion, gets some friends to sign on as co-editors, and publishes opinions, since you have friends you get to call yourself a "secondary source"?). But even if it were a serious secondary source, the CSI report is problematic for multiple reasons, the main one being it is racist and makes little to no attempt to correct for the language deficiencies of the TCM partitioners they interviewed. It is a great example of making "marks" out of AGF-ing hosts by intentionally misunderstanding them. We still need consensus to include it and I do not support that source. It is racist. Herbxue (talk) 09:01, 24 December 2013 (UTC)

I hope you all had a wonderful Christmas (if you celebrate it). Herbxue sums up my issue. This source is andecdotal and somewhat racist and I don't see why it needs the quote. As I have pointed out here I don't see other articles with quotes like this: "Yes, we met a native doctor, and he thought it was useful." Bonkers.--LT910001 (talk) 00:24, 27 December 2013 (UTC)

Alright! To summarize, this source is a) primary and b) anecdotal (as in: unfit to tell us anything about the general situation of TCM in modern times). I don't think it's racist, but for the mentioned reasons, I'd delete this source (if nobody objects). --Mallexikon (talk) 05:33, 30 December 2013 (UTC)

I agree it should be deleted. As I understand WP:FRINGE, sources like QW and CSICOP are ideal for tiny fringe topics with little sourcing. But they may be unnecessary and below par in areas where we have better sec sources, let alone MEDRS's, that provide adequate balance. --Middle 8 (talk) 06:52, 30 December 2013 (UTC)

How would the CSICOP report be primary to acupuncture? It is a group developing their own thinking based on interpretation, analysis and evaluation of facts, evidence and concepts from primary sources. The CSICOP delegation are not practitioners or subjects of acupuncture they did not perform a study on acupuncture, they spoke to primary sources, gathered evidence from primary sources, looked at facts in situ of primary sources, discussed concepts and ideas of primary sources with those sources and formed their own thinking based on their analysis of something they were not directly invovled in (at least one step away). If the subject were their trip they would be primary but the subject is their interpretation of the information they found. As for racism and failure to correct for language deficiences I didn't see any examples given. Regarding QW and CSI as RS this is a discussion that has been had multiple times, recognized experts, published in the literature of the fields the comment on, notable established organizations publishing material under editorial oversight with clear editorial policies. Note the guideline on RS discusses blogs with editorial oversight. - - MrBill3 (talk) 12:03, 8 January 2014 (UTC)

As per WP:PRIMARY: "Primary sources are original materials that are close to an event, and are often accounts written by people who are directly involved. They offer an insider's view of an event, a period of history, a work of art, a political decision, and so on... An account of a traffic accident written by a witness is a primary source of information about the accident; similarly, a scientific paper documenting a new experiment conducted by the author is a primary source on the outcome of that experiment." The CSICOP delegation made a trip to China to investigate Chinese medicine, and then wrote a report about it. It doesn't matter whether there was editorial oversight or not.

A secondary source, as per WP:SECONDARY, "contains an author's interpretation, analysis, or evaluation of the facts, evidence, concepts, and ideas taken from primary sources..." But CSICOP didn't evaluate existing literature here - they reported about their own experiences. "A book by a military historian about the Second World War might be a secondary source about the war, but if it includes details of the author's own war experiences, it would be a primary source about those experiences." It's exactly the same situation here. --Mallexikon (talk) 03:11, 9 January 2014 (UTC)

"The CSICOP delegation made a trip to China to investigate Chinese medicine" ← so there is a degree of separation between those doing the investigation and the people and things being investigated; CSICOP are outsiders looking in, not participants. The fact there was a trip abroad here is only incidental: that aspect is not the central topic being addressed. By your argument you could say a systematic review was a primary source because the authors describe how they are "directly involved" in selecting and sifting the research they are going to use. Alexbrn ^{talk|contribs|COI} 07:06, 9 January 2014 (UTC)

One would think that WP's definitions of primary and secondary sources are written clear enough to not have editors work on WP for years without really grasping their meaning... A review of (or a report based on) primary sources constitutes a secondary source. If you base your report on your own experiences, it's a primary source ("An account of a traffic accident written by a witness is a primary source of information about the accident; similarly, a scientific paper documenting a new experiment conducted by the author is a primary source on the outcome of that experiment". And: ""A book by a military historian about the Second World War might be a secondary source about the war, but if it includes details of the author's own war experiences, it would be a primary source about those experiences). Come on, this ain't rocket science. --Mallexikon (talk) 08:32, 9 January 2014 (UTC)

Quite. The Chinese people being interviewed are equivalent to the "witnesses" to the accident (or the "combatants" involved in the war). Their accounts are being digested, written-up and editorially contextualized by the SCICOP writers. If we had access to the verbatim transcripts of the CSICOP interviews they would would be primary sources. Alexbrn ^{talk|contribs|COI} 08:45, 9 January 2014 (UTC)

Illogical. You could just as well argue that a single study is a secondary source, since the researchers "editorially contextualized" the accounts of the participants. The important part is that in order to have a secondary source, you have to base it on a primary source - that means existing literature, not a verbal account you wrote down yourself. --Mallexikon (talk) 09:05, 9 January 2014 (UTC)

Quite often (for non-medical material) a single study is a secondary source. So you're essentially now saying it matters whether the material at the core is spoken or written. Why? Alexbrn ^{talk|contribs|COI} 09:17, 9 January 2014 (UTC)

Man, it's not me making these rules - it's WP. --Mallexikon (talk) 10:08, 9 January 2014 (UTC)

No such rule exists (that I can see) making such a distinction betweeen written and spoken material (which risks, incidentally, being culturally prejudiced: some traditions are predominantly oral). In this case we have the primary material (chinese medical material as related in the interviewees' accounts etc.) and the secondary treatment of it (from SCICOP). If they'd merely published raw transcripts that would be a different matter; but they didn't. Alexbrn ^{talk|contribs|COI} 10:21, 9 January 2014 (UTC)

What part of "a scientific paper documenting a new experiment conducted by the author is a primary source" and "a book by a military historian about the Second World War might be a secondary source about the war, but if it includes details of the author's own war experiences, it would be a primary source about those experiences" did you not understand? This is not a question of "raw transcripts" or "spoken material". If you were there yourself and document what you saw and heard, it's a primary source. If you write a book discussing several primary sources, that book is a secondary source. Those CSICOP people wrote a first-hand report about their trip. This is a primary as a source can be. If you don't want to hear this from me, though, I can't help it. Please take it to the RS noticeboard then, and let them explain it to you. --Mallexikon (talk) 11:14, 9 January 2014 (UTC)

SCICOP were not conducting a "scientific experiment", and as I a have said above it is the TCM practitioners who are analogous to the soldiers: they were giving their interviewers an "in the trenches account" of TCM. You're shifting the true focus of SCICOP's focus by saying the material is "about their trip". While in a superficial sense that is true, in fact what they are writing about is TCM as relayed by the sources (people) they consulted. Alexbrn ^{talk|contribs|COI} 11:26, 9 January 2014 (UTC)

And they did a poor job of it, and in my opinion, were not acting in good faith. This was either a primary account of their trip or poor "gotcha" journalism. Again, I think it reeks of ethnocentrism and I am personally offended by it.Herbxue (talk) 17:47, 9 January 2014 (UTC)

The authors of the SCICOP report were not first hand witnesses, nor participants thus not primary. The analyzed evidence this evidence is not required by WP to be literature or written accounts. Per the example of an automobile accident a reporter writing an article about the accident would be expected to speak to witnesses and participants not rely only on written evidence. No WP policy would exclude an article written based on such interviews. The authors of the SCICOP report conducted just such gathering of evidence, formulated opinions and analysis and published them. Where is there any primary aspect? Again to the example if the reporter were to discuss their efforts to locate witnesses, the discussion of this effort would be primary but not the discussion of statements of witnesses and the interpretation thereof.

If the SCICOP report was poorly conducted or is biased or compromised by ethnocentricity please provide an explanation with clear, specific examples. - - MrBill3 (talk) 11:04, 10 January 2014 (UTC)

Let me just contribute a few quotes from CSICOP's report here... From the chapter in question, i.e., the one they titled "Our visit":

"Although the lab appointments were simple, they were serviceable. The hallways were dimly lit by single, sparsely distributed fluorescent lamps, electricity apparently being expensive. Much of the lab’s equipment had been donated by an admirer of Dr. Han’s, the distinguished opiate researcher, Avram Goldstein, who shipped his furnishings and apparatus to Beijing after retiring from Stanford University. The institute’s walls were decorated with poster presentations from research meetings that described the lab’s discoveries..."

"We next visited several clinics at the China Academy of Traditional Chinese Medicine in Beijing. The main one was a moderate-sized room, approximately 12 by 4 meters, crowded with ten or twelve patients being treated — most with acupuncture, several with acupuncture and moxibustion, and two with cupping, as described below..."

"While touring the TCM complex, we made several observations. The total space in this institution devoted to the practice of TCM was a relatively small portion of its holdings. The rest of the complex of several large buildings was apparently devoted to more mainstream scientific research. We asked what portion of the total medical services delivered in China was TCM, and how people were chosen to receive TCM treatments. We received some surprising answers: Patients generally request TCM treatments themselves, rather than being referred to TCM practitioners by biomedically trained physicians. Most scientifically trained doctors do not practice TCM, nor do they decide on the mode of treatment if they should refer a patient for TCM."

I understand that the fact that they also did interviews makes it tempting to employ this rationale that Alexbrn and Mr. Bill used - i.e. the interviews to be counted as primary sources, and the CSICOP report then allegedly being a secondary source, since it discusses those interviews. You are missing the point, though. Primary/secondary source differentiation is all about: was the one writing it involved or not? And an interviewer is of course involved - the information retrieved is influenced by the way and the nature of the questions that are asked (and by the questions that are not asked). That's why, for example, a single medical study constitutes a primary source, even though the people conducting the study don't take the pills themselves but rather let test subjects take the pills and then interview them about the effects. Or sometimes just interview test subjects without even giving any pills at all.

If CSICOP would not have gone there themselves, but would have shifted through already published interviews (as per WP:RS: "The term "published" is most commonly associated with text materials, either in traditional printed format or online. However, audio, video, and multimedia materials that have been recorded then broadcast, distributed, or archived by a reputable party may also meet the necessary criteria to be considered reliable sources") and then compiled them, their source would constitute a secondary one. Does this make the differentiation clearer to you? --Mallexikon (talk) 04:57, 12 January 2014 (UTC)

Again I cite the example of newspaper articles a clearly acceptable source for WP. Newspaper reporter go to the scene of an event, interview people involved and synthesize original interpretation. This is precisely what CSICOP has done. The scientific study example is a red herring, the researchers set up the research, establish protocols and effect the action. Medical researchers actually administer the treatment studied this is clear direct involvement. CSICOP had no such direct invovlement, they gathered evidence much as a the reporter for a newspaper article would. The examples you gave are clear examples of evidence gathering and do not show any direct involvement in the subject they are reporting on. The evidence gathered for a report which constitutes a reliable source are not required to be previously published. In the history book on a war example the author may visit a battle site and interview soldiers, witnesses etc. that would not make the book primary. What would make the history book primary is if the author wrote about his involvement with the battle. This distinction seems quite clear. - - MrBill3 (talk) 05:14, 12 January 2014 (UTC)

Convoluted argumentation.

1.) A newspaper article very often constitutes a primary source (i.e., if it mainly deals with the reporter's own impression of a scene etc.; this might or might not include interviews he did at that scene).

2.) That doesn't mean this kind of newspaper article is not an acceptable source for WP. We can use primary sources: "Primary sources are often difficult to use appropriately. While they can be both reliable and useful in certain situations, they must be used with caution in order to avoid original research. While specific facts may be taken from primary sources, secondary sources that present the same material are preferred." (per WP:NOTRS).

3.) There are many single studies that do not involve administering treatment to test subject; e.g., finding out how many people in a given country suffer from, let's say, fibromyalgia. A study like that would still constitute a primary source, even though it would only consist of interviews. --Mallexikon (talk) 01:55, 13 January 2014 (UTC)

Thank you for your polite, considered and thoughtful response. Lets go directly to the guidelines/policy to eliminate any contention of convolution.

1.) "Primary sources are original materials that are close to an event, and are often accounts written by people who are directly involved. They offer an insider's view" CSICOP are not "insiders" and are separated from the event/subject as they are gathering evidence about something which they were not participants in nor direct witnesses of. They are not writing about acupuncture practices they witnessed or are close to or insiders in relation to.

2.) "Whether a source is primary or secondary depends on context. A book by a military historian about the Second World War might be a secondary source about the war, but if it includes details of the author's own war experiences, it would be a primary source about those experiences" (emphasis added). CSICOP is not reporting about their own experiences with acupuncture.

3.) Scientific studies: I again argue red herring. In a scientific study, parameters of a specific variable(s) and specific outcome(s) are established by investigators. The results of a scientific study are quantitative (or if subjective are still expressed as results). This is inherently different from an opinion based on analysis. Scientific studies report results of statistical analysis not "an author's own thinking based on primary sources, generally at least one step removed from an event". There is a key differentiation made between "an author's own thinking" and a scientific analysis that separates a scientific study.

4.) On what basis do you contend that the CSICOP report is not "an author's own thinking based on primary sources, generally at least one step removed from an event"? There is a set of primary sources, the authors were one step removed and they provided their own thinking (as distinct from scientific analysis per 3).

As separate issues, I am more than willing to consider contentions of bias, poor or limited analysis, poor or limited sources and ethnocentricity. Please forgive the extensive discussion on a topic of limited value as a learning experience related to WP policy interpretation and consensus editing. I do not wish to waste anyone's time but feel this is a clear "case in point". - - MrBill3 (talk) 08:14, 13 January 2014 (UTC)

No offense taken, and I feel similar about the "case in point".

1.) Beg to differ, but the CSICOP people turned into insiders as soon as they went to China to take a look for themselves. Please read their article again; they actually do write about acupuncture techniques they witnessed (moxa und cupping).

2.) They do report about their own experiences with acupuncture / the whole TCM system (cf their description of premises above; cf their description of acupuncture techniques etc.)

3.) Some convolution in your argumentation here. A single medical study constitutes a primary source. For a secondary source, you'd need a review of single studies. The crucial difference is that the single study doesn't base its analysis on published sources, whereas the review does. In light of these facts, could you rephrase your point?

4.) A secondary source has to base its analysis on published primary sources. The CSICOP people did that in a small part of their report, where they discuss the research work of one of the doctors they interviewed. However, it's only a small part of their report. Overall, the report is about what they've seen and heard. --Mallexikon (talk) 10:08, 13 January 2014 (UTC)

Where CSICOP reports on their own experiences they are a primary source to a certain extent. However as the guidelines repeatedly stress what is important is context. For the portion of the quote, "We were not shown" this is primary however it is in a larger context in the article.

CSICOP has not become "insiders" for TCM by simply going to a location and investigating it themselves. The remain analytical critics, this is abundantly clear in the article. For methods of moxa and cupping currently in practice in China they would be primary and anecdotal.

There is no requirement that secondary sources base their analysis on published sources read the guidelines carefully. However it it clear from the article listing 28 published sources as references they clearly have used published sources. It is also clear that they are providing "interpretation, analysis, or evaluation of the facts, evidence, concepts, and ideas taken from primary sources" through their 14 notes detailing such. Dr. Han's statement was not given in isolation it was given as an example of the statement, "acupuncture anesthesia for surgery, this apparently having fallen out of favor with scientifically trained surgeons." That they were not shown such is again an illustrative example. In WP both examples are anecdotal and may be excised, however; the contention that:

• According to Beyerstein and Sampson of CSICOP, acupuncture as anesthesia for surgery seems to have "fallen out of favor with scientifically trained surgeons" in China.

Is fully supported and properly so by the reference. This is clearly the product of their interpretation and analysis and they have made very clear in context that they sorted through substantial primary source material. - - MrBill3 (talk) 19:52, 26 January 2014 (UTC)

Proposal to remove statement

I propose this statement "Commenting on this meta-analysis, professor Edzard Ernst stated that it demonstrated that the effects of acupuncture were principally due to placebo.[n 2]" from here "A 2012 meta-analysis found significant differences between true and sham acupuncture, which indicates that acupuncture is more than a placebo when treating chronic pain (even though the differences were modest).[99] Commenting on this meta-analysis, professor Edzard Ernst stated that it demonstrated that the effects of acupuncture were principally due to placebo.[n 2] " be deleted from the article. Concerns: --LT910001 (talk) 00:38, 27 December 2013 (UTC)

• Opinion of scientist in newspaper is a weak WP:MEDRS reliable source and should be replaced with a better one. See "The popular press is generally not a reliable source for scientific and medical information in articles", and "Roughly in descending order of quality, lower-quality evidence in medical research comes from individual RCTs; other controlled studies; quasi-experimental studies; non-experimental, observational studies, such as cohort studies and case control studies, followed by cross-sectional studies (surveys), and other correlation studies such as ecological studies; and non-evidence-based expert opinion or clinical experience." (WP:MEDASSESS, same page).
• This is WP:OR: "This includes any analysis or synthesis of published material that serves to advance a position not advanced by the sources. "
• This violates WP:UNDUE: "Do not combine material from multiple sources to reach or imply a conclusion not explicitly stated by any of the sources. If one reliable source says A, and another reliable source says B, do not join A and B together to imply a conclusion C that is not mentioned by either of the sources." Source A (Metaanalysis) says yes, source B (expert opinion in the Guardian) says no. It's improper to link them together like this.

Ernst is just about the most reliable source on this subject regardless of the venue. Proposal rejected. jps (talk) 00:54, 27 December 2013 (UTC)

Thanks for your input, but I would remind you that WP is build around consensus, so we'll need to wait for a few more users to chip in before we jump to that conclusion. Additionally, WP:MEDRS is a guideline for statements making claims about medicine on Wikipedia. Ernst may be very knowledgable, but his statements do not carry the same weight as meta-analyses and systematic reviews. --LT910001 (talk) 01:15, 27 December 2013 (UTC)

The review needs to be explained and Ernst is just the person to do it. jps (talk) 02:09, 27 December 2013 (UTC)

This was discussed before. Ernst is an expert on the topic. His evaluation on the meta-analysis maintains NPOV. QuackGuru (talk) 01:45, 27 December 2013 (UTC)

Um, not so sure. He clearly has a POV. I am not suggesting his studies be excluded (they are notable), but his commentary on someone else's study is not really worth including.Herbxue (talk) 06:27, 27 December 2013 (UTC)

Ernst is not reliable. -A1candidate (talk) 01:56, 27 December 2013 (UTC)

He sure is. He's the only person who has ever had a professorship in alternative medicine. In some ways, he is the only expert on this subject. jps (talk) 02:09, 27 December 2013 (UTC)

Ernst is very reliable, and an excellent expert to use as a source in this area. --Roxy the dog (resonate) 03:07, 27 December 2013 (UTC)

He is certainly notable and must be included, but for fuck's sake he is not an expert on acupuncture or TCM, and is definitely not the "only expert"! He's just the only source you like. One thing I like about him is that he is never so sloppy as to say "acupuncture is just placebo", I'm less pleased that he does indulge in every opportunity to suggest that it might be, in a manner that does not require him to prove it or be responsible for it, only suggest it. Herbxue (talk) 06:22, 27 December 2013 (UTC)

The discussion here is not about Ernst being notable or not (he sure is). The discussion is about whether it is permissible to add his comment here. We have a MEDRS (the Vickers study). And we have a newspaper article citing Ernst. The former obviously trumps the latter. Even if Ernst wrote a letter to the editors this source wouldn't trump a meta-review. We have to leave Ernst's comment out; otherwise, we set a bad precedent. For all the medicine-related articles on WP. --Mallexikon (talk) 07:29, 30 December 2013 (UTC)

Yes, it's massive undue weight; Vickers and individual expert commenters like Ernst occupy opposite ends of the spectrum of good medical sources. In fact they're so far apart that the only way I can see giving a sentence to Ernst is if we have at leas a couple para or more devoted to Vickers -- the spread is easily that big. There are other good sec sources with different findings, and therefore ample balance in the article. The literature obviously shows conflicting results and conclusions, and that's exactly what we should portray. --Middle 8 (talk) 10:54, 30 December 2013 (UTC)

Ernst's the go-to expert on all things acu* - his opinion (attributed as such) is worthy of inclusion I'd have thought. Alexbrn ^{talk|contribs|COI} 11:01, 30 December 2013 (UTC)

If you show me some reliably sourced evidence that Ernst is the internationally accepted "go-to expert on all things acu" - fine. Until then, he's not going to get any extra treatment. WP:MEDRS doesn't say "look for the one expert you can find on a given topic and try to insert his views and opinions as often as possible in an article". WP:MEDASSESS demands that "in general, editors should rely upon high-quality evidence, such as systematic reviews, rather than lower-quality evidence, such as case reports, or non-evidence, such as anecdotes or conventional wisdom. The medical guidelines or position statements produced by nationally or internationally recognised expert bodies often contain an assessment of the evidence as part of the report." No mention of single-expert opinions taken from a newspaper article. --Mallexikon (talk) 03:32, 31 December 2013 (UTC)

For starters, see Edzard Ernst#Work in complementary medicine. QuackGuru (talk) 03:41, 31 December 2013 (UTC)

Yep. Read it. This guy has been doing a lot of research (then again, that's what researchers do, no?). However, no reliably sourced evidence that Ernst is the internationally accepted "go-to expert on all things acu" yet. --Mallexikon (talk) 04:45, 31 December 2013 (UTC)

Well, The Guardian went to him for the opinion we're citing; the UK Parliament go to him when taking evidence on CAM; Exeter University went to him in appointing a CAM chair. He has been called a "leading critic of alternative medicine" in the BMJ and by The Times "a leading expert". Respected academic publishers repeatedly publish his work in books and journals, covering CAM topics in general and acupuncture in particular. Even CAM publications cite him as an authority, sometimes even as "Edzard Ernst himself"! And he's a trained acupuncturist. Really, the claim that Ernst doesn't represent noteworthy "expert opinion" appears rather tendentious. Alexbrn ^{talk|contribs|COI} 08:02, 31 December 2013 (UTC)

Yes, WP:MEDRS guides us what to do "in general"; and in this case in particular my assessment is that Ernst is good to have, especially considering the WP:FRINGE context of the topic as a whole. WP:MEDRS is a useful guideline, but WP:PSCI is mandatory policy. Because acu* is pseudoscience, we need to ensure that we frame all these medical investigations (which, in any case show it to be nothing-or-nearly-nothing) within the wider mainstream scientific context, which is that investigating acupuncture is an absurd endeavour (see PMID 20457720). Alexbrn ^{talk|contribs|COI} 07:20, 31 December 2013 (UTC)

1.) Acupuncture is not characterized as pseudoscience at WP (cf Category:Pseudoscience). Middle8 gave a very good overview on this (complex) situation at Wikipedia talk:WikiProject Medicine#WP:FRINGE/PS answers this question: No

2.) Yes, Ernst got a chair. Yes, he's a scientist published by respected scientific academic publishers. Yes, The Times has called him a "leading expert" (on whatever). But you still haven't produced any reliably sourced evidence that Ernst is the internationally accepted "go-to expert on all things acu". He's just one scientist among many, even though you love him. In comparison to Vickers meta-review, Ernst's commentary doesn't even constitute a MEDRS. Putting the two in direct context gives undue weight to an inferior source. --Mallexikon (talk) 08:36, 31 December 2013 (UTC)

Middle 8 gave an acupuncturist's view after everybody else had moved on. However the consensus on that thread is against it, as it is in RS: acupuncture is pseudoscience (or at the very least, as an altmed practice, fringe) and so falls under the WP:PSCI policy. As to Ernst, requiring some source that uses the exact words "internationally accepted 'go-to expert on all things acu'" is silly. By any reasonable measure, and as stated in RS, he's a "leading expert" on the topic for the reasons I have already given. If these reasons are not good enough for you, then we shall have to disagree; they're good enough for me. So in my view his opinion is worthy of inclusion. Alexbrn ^{talk|contribs|COI} 09:05, 31 December 2013 (UTC)

In addition, I would suggest that except for acupuncturists here, there is consensus that it is a pseudoscience, and we would expect people who practise this pseudoscience to want to defend this dubious 'medical' treatment. --Roxy the dog (resonate) 09:19, 31 December 2013 (UTC)

@Alexbrn - "Acupuncturist's view"? Oh really? By your logic, does the fact that I'm also a scientist give my words greater weight than your non-scientist's view, or do the science and woo cancel each other out? Ad hominem is lazy, and sure enough you offer no rebuttal to my argument. Of course you're right that it has fringe-aspects, but it also has non-fringe ones, and neither has to do with a semantic argument.

@Roxy the dog - there's exactly one acupuncturist here. And he's curious if you're going to clarify your assertion that acupuncture's good for nothing) in light of gold-standard sources that say otherwise. I see above you've perhaps walked it back to "dubious", which is actually about right, all things considered (i.e. given all the stuff it's said, falsely, to be good for). --Middle 8 (talk) 11:45, 31 December 2013 (UTC)

When you boil it down to basics, those "gold-standard" sources say it is good for a bit of placebo induced pain relief. That is all. It is bonkers that we big it up like we do. Also, and this goes to competence, how can a scientist actually be an acupuncturist? The former actually would be expected to totally obliterate any belief in the latter, assuming normal scientific understanding of the world. --Roxy the dog (resonate) 12:40, 31 December 2013 (UTC)

@Roxy - Funny how those same gold-standard sources (Ernst re Cochrane; Vickers '12) say it's good beyond (sham) placebo. For nausea, too ("Acupoint P6 reduces PONV"). I know others like Novella believe the import of those same sources (for pain, anyway) ultimately show placebo, but we have a well-documented range of views, and Cochrane's and Vickers' meta-analyses obviously have more weight than Novella's blog articles. Of course Novella may be right, but WP lags the science, and there's no sci consensus yet.

On your question, this isn't the place to dwell on my biography, but: (1) both the evidence and my understanding of EBM were different when I trained to be an L.Ac.; (2) my views have evolved and I now figure it's all up for grabs, and that it's a question of science not belief; (3) I'm not practicing anymore, so even if it weren't the right thing, I can afford to put the science first; (4) as is obvious from even some of the better sources here, some scientists seem to find merit in it; ask one sometime -- scientific opinion isn't as homogeneous as opinion among non-scientist skeptics, FWIW. For example, the map (qi and meridians) isn't the same as the territory (results). Mechanism isn't that implausible; it's not like it's action at a distance or something. --Middle 8 (talk) 16:06, 31 December 2013 (UTC)

@Middle 8 The presence of interested positions in a consensus-forming process distorts it which is why, at more serious venues than Wikipedia, declarations of interest are often required and conflicted people often withdraw. You are correct that I did not bother to rebut your argument (by which one could argue that homeopathy was not pseudoscience, since it has been subject to serious investigation). I think you'll find that there's more than 1 acupuncturist participating here - of which more soon ... Alexbrn ^{talk|contribs|COI} 12:53, 31 December 2013 (UTC)

@Alexbrn - Simply having a profession isn't considered a conflict on WP, and rightly so. See my comments above to Roxy-t-d, and bother to read them at Talk:GERAC, if you wonder about my interests and views. With respect to my qualifications, editors should be thanking me for having the topic and scientific literacy to actually make this a good article, and for caring enough to follow sources and talk pages. (But no good deed goes unpunished.)

Now I know that you haven't read WP:FRINGE/PS, because the determining factors don't include the presence of serious research (although that can be informative). We had an ArbCom ruling on pseudoscience (which adopted some of my suggestions on demarcation), and incorporated their points into NPOV and then NPOVFAQ and then over FRINGE, precisely to bypass endless arguing, based on dueling criteria. My post at WT:MED was nothing more than a straight recap of policy (or maybe it's technically guideline now, but it's basically NPOV) that nobody else had bothered to read. --Middle 8 (talk) 16:06, 31 December 2013 (UTC)

Belatedly @Roxy the dog: "I would suggest that except for acupuncturists here, there is consensus that it is a pseudoscience" -- demonstrably false, and anyway, we go by NPOV's guidelines on demarcation ([[WP:FRINGE/PS, and WP:RS/AC, per which acu isn't to be unambiguously depicted as a pseudoscience, although RS's saying it is are obviously fine to cite. Also, your use of scare quotes above -- "gold-standard" sources -- suggests you don't understand that Cochrane and other meta-analyses really are. No gold-standard source calls acu a pseudoscience, btw. --Middle 8 (talk) 13:57, 27 January 2014 (UTC)

Of course, Cochrane doesn't address the question of where topics lie on the science/pseudoscience continuum. We need to turn to expert RS that does deal with this. And acupuncture has its own entry in Shermer's encyclopedia of pseudoscience, I see. Alexbrn ^{talk|contribs|COI} 14:04, 27 January 2014 (UTC)

(Add) And in the Encyclopedia of Pseudoscience. I feel a category coming on ... Alexbrn ^{talk|contribs|COI} 14:22, 27 January 2014 (UTC)

Done --Roxy the dog (resonate) 16:27, 27 January 2014 (UTC)

You need a better RS than that. I've pointed you to the germaine policies already and it's been discussed plenty in the archives and elsewhere ... you guys are now into IDHT and approaching WP:POINT, not to mention a lovely ad hominem above from Alexbrn. This has been global consensus since WP:ARB/PS. I summarize generally-accepted logic explained on my user page -- User:Middle_8#Using_Category:pseudoscience -- if you can be arsed to read it. --Middle 8 (talk) 19:06, 27 January 2014 (UTC)

Hang on

But, looking at the article I see the Vickers study and Ernst's reaction to it is duplicated in two sections, and we have a socking great quote from Ernst in the reference. We don't need all that, just a brief summary will do. I've attempted to de-duplicate and slim Ernst down. See what you think ... Alexbrn ^{talk|contribs|COI} 09:54, 31 December 2013 (UTC)

Thank you. :-) A good start. I'd prefer that we give more space to discussing Vickers, and include his reply, for the sake of WP:WEIGHT and because it just lays out the areas od disagreement nicely. Here's my suggestion, including a minor reorg, putting the '12 efficacy study ahead of the '13 followup, for better flow and because it's an efficacy section.

On the disagreement with Ernst: Vickers' belief that the difference between real-acu and no-acu is more clinically relevant than the difference between real-acu and sham-acu is indeed perverse (per Novella et. al.) from the standpoints by which medicine ordinarily evaluates placebos, but it is a significant view in medicine -- e.g. also held by independent German body who evaluated the results of GERAC. (And in the case of GERAC, there was no statistically significant difference between sham and real acu. At least with Vickers, there is, though it roughly splits the difference between no-acu and real-acu, and is small. Vickers has pain relief at about 30%, 42.5% and 50%, respectively, for no acu, sham acu, and real acu. I mentioned this in my edit.) --Middle 8 (talk) 15:16, 31 December 2013 (UTC)

Sorry I mucked up the section headers, but anyway, we can continue below at Talk:Acupuncture#Pain_section. --Middle 8 (talk) 22:23, 31 December 2013 (UTC)

Legal and political status recent edits

In 2006, German researchers published the results of one of the first and largest randomized controlled clinical trials.[194] As a result of the trial's conclusions, some insurance corporations in Germany no longer reimburse acupuncture treatments.[194] The trials also had a negative impact on acupuncture in the international community.[194]

"...some insurance companies in Germany stopped reimbursement for acupuncture treatment]."[19]

• He, W.; Tong, Y.; Zhao, Y.; Zhang, L.; Ben, H.; Qin, Q.; Huang, F.; Rong, P. (2013). "Review of controlled clinical trials on acupuncture versus sham acupuncture in Germany". Journal of traditional Chinese medicine. 33 (3): 403–7. PMID 24024341. {{cite journal}}: Unknown parameter |displayauthors= ignored (|display-authors= suggested) (help)

As you can see from reading the 2013 review, the text is obviously sourced and is accurately written. Recently, I have noticed editors disagree with me every single time I start a new thread related to acupuncture. Let's see what happens this time. Remember, it does constitute a reliable source according to Mallexikon. QuackGuru (talk) 02:13, 27 December 2013 (UTC)

Avoid WP:TE and WP:IDHT and you'll find less disagreement arising. --Middle 8 (talk) 11:10, 30 December 2013 (UTC)

As I said, the text is neutrally written and it is also WP:MEDRS compliant. This review overrules primary sources and outdated sources. I object to deleting this relevant source. QuackGuru (talk) 19:22, 31 December 2013 (UTC)

@QG: I replied some at FTN, where you cross-posted about the issue. [20] My comment is about pushback in general, not this specific proposal, re which I haven't developed an opinion. What can I say that I haven't said before re: my hope that you'd take a new approach to editing? --Middle 8 (talk) 22:19, 31 December 2013 (UTC)

Another null-result from a sham acupuncture controlled study

http://onlinelibrary.wiley.com/doi/10.1002/cncr.28352/abstract

Consider adding it to the article.

jps (talk) 00:03, 28 December 2013 (UTC)

A small primary study; it adds little to the picture painted by the strong secondary material we've got. Alexbrn ^{talk|contribs|COI} 03:55, 28 December 2013 (UTC)

The article states, "research suggests that it is helpful in relieving some symptoms of cancer or the side effects of cancer treatment." And this study directly contradicts that. jps (talk) 04:11, 28 December 2013 (UTC)

@jps - This study actually supports, not contradicts, that statement from our article -A1candidate (talk) 05:24, 28 December 2013 (UTC)

jps, yes, it does contradict, but the source is not MEDRS compliant, so the text should be changed using other means and sources. It's still instructive, so thanks for sharing.

I just finished reading Paul Offit's book, Do You Believe in Magic?: The Sense and Nonsense of Alternative Medicine. Very interesting and well-written. Among many things he writes about acupuncture, he tells about Steven Novella's appearance on the Dr. Oz Show:

Dr. Oz: "I just think it's very dismissive of you to say that because we couldn't take this idea [acupuncture] that exists with a different mind-set and squeeze it into the way that we think about it in the West, that it can't possibly be effective."

"Novella knew that acupuncture was by definition a sham, a trick, a deception; yet he never once said, "Acupuncture doesn't work." Rather, he questioned why it worked. "It's the ritual surrounding a positive therapeutic interaction: a comforting, caring [clinician]," he said. "You're relaxing for half an hour or an hour. That's where the effect is. There's no effect to actually sticking a needle through the skin." In other words, the placebo effect." (p. 224)

So, that's where it's at. It's the placebo effect, a very non-specific effect. When a method, like acupuncture, does not rise above the placebo effect, it is a universal truth in science and medicine that the method is itself considered ineffectual. In common parlance we just say that it doesn't work as claimed, and the claims are false. Something else is going on that could just as well be induced by numerous other methods. Since other, truly helpful, methods exist which actually do far more than stimulate the placebo response, they should be preferred, and failing to do so is to practice below the expected standard of care, and that's dead wrong, unethical, quackery, and in some cases illegal.

It really worries me (for their own sake, and especially for their patients' sake) that A1candidate doesn't understand this. This failure means they are practicing deception and substandard practice, which can have serious and even fatal consequences for their patients. As far as editing here is concerned, it also means they don't understand the workings of the scientific method well enough to be competent in understanding and editing the medical and scientific aspects of subjects. -- Brangifer (talk) 06:55, 28 December 2013 (UTC)

Again with the over-the-top drama-Queen rhetoric. Even if most acupuncturists didn't realize that much of their clinical success came from non-specific effects of acupuncture (the "relaxation response"), they still would not being doing any more harm than collecting a paycheck for ineffective service. You suggest that their ignorance may have "fatal consequences" - really? Other than that Canadian massage therapist (a non-acupuncturist) practicing acupuncture without proper training or licensure, show me one bonafide example in the US or Canada of danger from a properly trained and licensed acupuncturist in the west. There is a reason why an acupuncturist's malpractice insurance costs $500 US per year while a typical MD pays $200,000 per year. Its because acupuncture is safe and often effective therapy.Herbxue (talk) 08:25, 28 December 2013 (UTC)

Didn't you read what I wrote?: "Since other, truly helpful, methods exist which actually do far more than stimulate the placebo response, they should be preferred, and failing to do so is to practice below the expected standard of care, and that's dead wrong, unethical, quackery, and in some cases illegal." Acupuncturists "treat" AIDS, cancer, diabetes, and myriad other conditions for which there are proven, effective, methods, methods which they often recommend against with claims that acupuncture will help. Failing to use effective methods can cause great harm. This applies to many more so-called "alternative" methods than acupuncture. Diversion from effective methods is one of the greatest weaknesses and dangers of using so-called "alternative" medicine. If it were used solely for treatment of relatively innocuous, self-limiting, conditions, the only harm would be the robbing of the pocketbook, but that's only part of the problem. -- Brangifer (talk) 17:52, 28 December 2013 (UTC)

I don't know the laws state-by-state, but AFAIK it's against the law in quite a few states in US for acupuncturists to claim to treat such things. And acupuncture schools teach that it's unethical to make such claims. --Middle 8 (talk) 01:43, 29 December 2013 (UTC)

That's good! Unfortunately many places that is not the case, and it doesn't apply just to acupuncturists. -- Brangifer (talk) 06:31, 29 December 2013 (UTC)

Sorry, why is the source not WP:MEDRS compliant? jps (talk) 10:58, 28 December 2013 (UTC)

It's one single piece of primary research, not a review or meta analysis. We rarely allow use of such sources. -- Brangifer (talk) 17:52, 28 December 2013 (UTC)

Have you read this article lately? It's full of references to such studies. jps (talk) 20:53, 28 December 2013 (UTC)

I could only find one such source (a null result for drug detox; we already have two reviews on the subject), and removed it. I didn't count the total number of reviews in the article, but it's on the order of dozens. User jps is perhaps thinking of a different article -- maybe the TCM one? --Middle 8 (talk) 04:19, 29 December 2013 (UTC)

That' probably true. I suspect the TCM article has such problems. -- Brangifer (talk) 06:31, 29 December 2013 (UTC)

I just checked, and the TCM article is also OK in terms of using reviews instead of individual studies. I could have missed a couple, but I saw only Cochrane-type reviews cited for efficacy. I have no idea what jps is referring to. --Middle 8 (talk) 23:31, 29 December 2013 (UTC)

That is problematic. Please tag each instance, and after that's done we can work on deleting them one at a time. That will allow specific objections and specific reversions to be dealt with appropriately. It needs to be done individually because it will likely demand revision of content. -- Brangifer (talk) 21:20, 28 December 2013 (UTC)

While we're on the subject of non-specific effects, a parallel study about homeopathy makes the same point as with the acupuncture study:

• Homeopathy has clinical benefits in rheumatoid arthritis patients that are attributable to the consultation process but not the homeopathic remedy: a randomized controlled clinical trial, Rheumatology, Volume 50, Issue 6, Pp. 1070-1082

I suspect this is the case with many methods where one sees an effect, but where there is no difference between control arms. "Doing something is better than doing nothing, but not better than doing anything else, and is worse than doing something effective." (you can quote me ) -- Brangifer (talk) 18:23, 28 December 2013 (UTC)

Therein lies the problem with this article and the TCM one - the average reader will come away with the impression that both treatment modalities are worthwhile, and might help what ails them, when in reality nothing could be further from the truth. It is really bad that they have been allowed to get this bad. --Roxy the dog (resonate) 21:01, 28 December 2013 (UTC)

That's what happens when scientifically illiterate editors are allowed to get their way unopposed. Giving equal weight to sources that are unequal also causes these problems. We need to find the problem statements and copy them here, then deal with them to fix these systemic problems where bias in favor of ineffective methods and unreliable sources has taken over. -- Brangifer (talk) 21:24, 28 December 2013 (UTC)

The reason the article gives the impression that acupuncture may have an effect -- and the reason some editors argue that the article should reflect that -- is because there are good sources that say so. See Ernst's summary of Cochrane Collaboration reviews, among others. (While he thinks acu is a placebo, that review-of-reviews shows he acknowledges that the literature at this point doesn't say that.) Most of the sources in the article, last time I read it carefully anyway, discussing efficacy are good, secondary ones.

Brangifer, are you really claiming above [21] that it's problematic -- and suggestive of editorial incompetence -- if A1Candidate does not believe acupuncture is a placebo?

I don't think that Roxy the dog and many other editors (based on comments I've seen around WP) understand the literature -- either that, or they're not giving due weight to the good sources that find acu is more than a placebo. (My comments about weighting sources apply to this article, not the TCM one.) --Middle 8 (talk) 02:51, 29 December 2013 (UTC)

I know that you understand this, but some here don't. Something can cause an effect (and the literature - and quote from Novella/Offit - shows that acupuncture does cause effects) and still be a placebo, because placebos do cause effects and draw forth placebo responses. This fact that there is an effect seems to be misunderstood as meaning that the effect of placebo/acupuncture/homeopathy/etc. is meaningfully comparable to a proven medical treatment which has a far greater effect, much further than the placebo effect one would expect from any action taken or exposure to a sympathetic practitioner. Again: "Doing something is better than doing nothing, but not better than doing anything else, and is worse than doing something truly effective." (you can quote me ) -- Brangifer (talk) 06:31, 29 December 2013 (UTC)

Your point would be well taken if we were discussing the primary sources on proposed mechanisms (that an effect of acupuncture, like triggering inflammatory cascade and endorphin release, is not equivalent to a clinically significant result). But we are talking about systematic reviews that compare acupuncture to placebos, so your point is not well taken. Acupuncture is shown to have an effect greater than inert placebo for certain types of pain. The ongoing insistence on using the word placebo in reference to acupuncture is based on an emotional feeling towards acupuncture that people have rather than the evidence. And by the way, ever wonder why MD's give antibiotics for obvious viral URI? Or dermatologists give steroids for everyrthing regardless of diagnosis? When you say that acupuncture does not rise to the standard of accepted biomedicine, what exactly is that standard? It is definitely not the utopian science based rational brilliant medicine you seem to think it is. Herbxue (talk) 09:12, 29 December 2013 (UTC)

@Brangifer - When Ernst's review of Cochrane reviews says that they find acupuncture is effective for pain, it means "effective beyond placebo", and this should be obvious from context. You didn't answer my question about what you meant, but the reason I asked was to give you a chance to walk it back, because it leads to a conclusion that's over the top. Your comment is pretty clearly saying (a) acupuncture is a placebo (i.e., only a placebo), (b) A1candidate doesn't understand this, and therefore (c) A1candidate lacks the scientific literacy necessary to edit WP sci + med articles (and is practicing deception and substandard medical care). Well, if that's true, then any editor/doctor who accepts Cochrane reviews, and believes Vickers' 2013 review etc., is guilty of the same things (or just the lack of WP-competence part if they're not a doctor). That's an absurd conclusion, since Cochrane reviews and sources like Vickers are indicators of mainstream opinion and (for that reason) among the best MEDRS's we have. I must be missing something... --Middle 8 (talk) 09:23, 29 December 2013 (UTC)

Research history and context

Some of the above discussion has made it apparent that this article is lacking some context for the research that has taken place, while there is quite a lot of acupuncture-specific material out there we could be using. Here are some sources as a starting point: can anybody suggest more?

• Edzard Ernst, E (2009). "Complementary and Alternative Medicine: Between Evidence and Absurdity". Perspectives in Biology and Medicine. 52 (2): 289–303. doi:10.1353/pbm.0.0083. PMID 19395826.
• Greasley, P (2010). "Is evaluating complementary and alternative medicine equivalent to evaluating the absurd?". Evaluation & the health professions. 33 (2): 127–39. doi:10.1177/0163278710361923. PMID 20457720.
• Foot, Doug; Ridge, Damien (2012). "Constructing the placebo effect in the placebo wars: What is the way ahead?". Health Sociology Review. 21 (3): 355. doi:10.5172/hesr.2012.21.3.355.
• Jeremy H. Howick (23 February 2011). The Philosophy of Evidence-based Medicine. John Wiley & Sons. pp. 92–94. ISBN 978-1-4443-4266-6.

Note the topic here is not biomedical as such: the content we're missing is on the nature and value of the scientific investigations that have been conducted on acupuncture. Alexbrn ^{talk|contribs|COI} 09:26, 31 December 2013 (UTC)

Please suggest acupuncture-specific excerpts that you think would benefit the article. What is it specifically about the last source, Howick, that you like? The part where it's critical of sham acu, and ergo the design of GERAC (which was the first large study to find sham = verum) and every other subsequent sham-controlled trial? Did you even bother to read it? I already criticized it [22] at Talk:GERAC -- acupuncture partisan that I am [23]. --Middle 8 (talk) 11:01, 31 December 2013 (UTC)

You said it :-) Alexbrn ^{talk|contribs|COI} 11:28, 31 December 2013 (UTC)

(Yes, the diff beside my last comment above should be read, lest the sarcasm be lost....) I said it, but did you read it? You said at Talk:GERAC that "later better sources (Howick) have discounted the trials' ability to emit clinically significant data". [24] Hmmmm. So why do you think Howick (pp.92-94) is good for contextual purposes here? --Middle 8 (talk)

Not sure yet, just gathering potential sources ... at first blush it bears on the statements (we have) that trial design is highly problematic. Alexbrn ^{talk|contribs|COI} 16:46, 31 December 2013 (UTC)

Got it; thanks. Yes, sham acu is non-trivial, and there's no one accepted way to do it. Since most efficacy studies use some form of sham, Howick's wholesale dismissal is very much a minority view.

Here's an interesting article with Ben Goldacre as co-author that could be used for contextual purposes:

• Interpretive Bias In Acupuncture Research? A case study

GERAC is cited as an example -- not as a criticism of study design, but of how it was intrepreted (cf. my comment above, 2nd para). The authors also rebut Howick's objection about sham, i.e., he thinks sham acu can't be a true placebo because it's more active than an inert pill. But as the above source says, the more noxious or intense the stimulus, the stronger the placebo (or nocebo) effect. So two pills are more intense than one pill, dry needling is better than two pills, etc. --Middle 8 (talk) 21:25, 31 December 2013 (UTC)

Howick needs careful consideration. It appears he is a foremost authority on placebo use. Alexbrn ^{talk|contribs|COI} 14:07, 2 January 2014 (UTC)

Pain section

It's got some outdated cruft and could use a cleanup. Most recent Cochrane source on PONV is straightforward enough: P6 acupoint stimulation prevents postoperative nausea and vomiting with few side effects. Will suggest wording later.... too long of an editing marathon, damn, it's like a hangover, I never learn. --Middle 8 (talk) 16:53, 31 December 2013 (UTC) I was talking about nausea with the above struck text, but accidentally typed pain for the section header instead. Sorry --Middle 8 (talk) 20:24, 31 December 2013 (UTC)

It's painful, this section. Who can imagine a general reader getting much out of it? It should really be digested into a couple of paras. Alexbrn ^{talk|contribs|COI} 17:27, 31 December 2013 (UTC)

I strongly disagree with this unorganised hard to read version. QuackGuru (talk) 18:29, 31 December 2013 (UTC)

There was no consensus for the hard to follow recent change. QuackGuru (talk) 19:49, 31 December 2013 (UTC)

Sorry folks, I meant the header to say "nausea", not pain; just a typo when I was tired. Anyway, let's discus the pain section here. I agree with Alexbrn that the section on efficacy for pain needs work too.

As for recent edits: we were talking above about the longstanding issue of Ernst's opinion of Vickers' review: whether to include or how to weight the former. Alexbrn noticed some duplication in the article, cf. Talk:Acupuncture#Hang_on. I commented further and offered an edit. I'm cool with including Ernst's opinion -- he is, after all, unsurpassed as an expert -- but for [[WP:WEIGHT] purposes would like to expand discussion of Vickers. If my edit looks sympathetic to acupuncture, that's because the source's conclusions are. I paraphrased it very closely (just enough to avoid copyvio) and didn't cherry-pick. I think my edit fairly portrays the source, objections to it, and the author's reply.

In my ES I noted that I did a minor "reorg", i.e. put the original '12 efficacy study before the '13 followup about lack of differences across styles of acu. QuackGuru objects because it's "unorganised" and "hard to read". I don't buy this explanation; the prose is quite clear, whatever else one might say about it. It's mostly simple sentences, fer heaven's sake. Have a look: diff here, and you can see how it appears at this version of the article. --Middle 8 (talk) 22:45, 31 December 2013 (UTC)

Editors objected to your change and now you are trying to force changes to the article? QuackGuru (talk) 22:55, 31 December 2013 (UTC)

Correction: one editor -- you -- has objected so far. I'm not trying to force anything, and note that one other editor besides me thought your reversion was meritless. Like I said in my ES, I only ask that objections be plausible ones, and specific. Just repeating "hard to follow" doesn't cut it; I'm sick of the IDHT. It's not too much to ask that the discussion be substantive.

"The review's lead author responded that the decision in clinical practice is not the difference between sham and real acupuncture, but whether or not to refer for acupuncture;[117] he also noted that the effect of real acupuncture vs. sham is in some cases "very comparable to that for widely accepted treatments", such as NSAIDS for osteoarthritis.[117]"

This is a weight violation with the Vickers response. The 2013 meta-analysis is newer and should be at the top of the section. The text should be in a single paragraph. QuackGuru (talk) 20:40, 1 January 2014 (UTC)

@QuackGuru Please explain why you believe it is a WP:WEIGHT violation and why it should be a single paragraph. As for the order: The reason I put Vickers '12 first is because it's the original paper and '13 is the followup; '13 doesn't supersede the results of '12 but rather expands upon them. And they use the same dataset, so one year's difference in publication is a red herring, as I suspect you know already. --Middle 8 (talk) 01:53, 2 January 2014 (UTC)

It is too much extreme detail that is not from a review. I prefer chronological order rather than cherry picking where you like to go first. QuackGuru (talk) 02:05, 2 January 2014 (UTC)

"A 2012 meta-analysis of acupuncture conducted for the Acupuncture Trialists' Collaboration found "relatively modest" efficiency of acupuncture (in comparison to sham) for the treatment of chronic pain, and on that basis concluded it was a reasonable referral option.[115] The review examined four different types of chronic pain – back and neck pain, osteoarthritis, chronic headache, and shoulder pain – and found similar effects in all of them. Patients' pain reduction was estimated at 30% for no acupuncture, 42.5% for sham acupuncture, and 50% for real acupuncture. The authors concluded that the results "indicate that acupuncture is more than a placebo".[115]

This is not all from the conclusion. This is a weight violation to include this extreme detail from one source. QuackGuru (talk) 02:05, 2 January 2014 (UTC)

I appreciate your trying to improve my edit rather than just reverting it, including keeping the chron order, the explicit conclusion re not-placebo and the reply to Ernst's criticism. But you're overlooking the main reason I added more to Vickers: too much weight to the expert opinion relative to the review itself. We don't have to include every detail I mentioned (nor are we limited to the conclusion section), but the general idea is to avoid putting giving equal weight to the review vs. the commentary on it. I'll try to find middle ground. For example, the four different kinds of chronic pain examined should certainly be mentioned. The response rates might be worth keeping. Why one long para when there's an obvious place to split, i.e. at the '13 followup? Anyway, it's nice to move forward rather than revert warring; let's do more of that. :-) --Middle 8 (talk) 06:20, 2 January 2014 (UTC)

P.S. Re the OR-tag you addded: I have the PDF, and Vickers is responding to the exact point Ernst makes. But he directs that response toward Colquhoun, who unlike Ernst made his point in the sci literature.[25]

• This is Coquhoun, as quoted by Vickers: "acupuncture does not work to any useful extent ... Vickers et al showed that the difference is far too small to be of the slightest clinical interest".
• This is how we (accurately) summarize Ernst in the article: "Edzard Ernst stated that it 'impressively and clearly' demonstrated that the effects of acupuncture were largely due to placebo, since the difference between sham and real acupuncture were small and clinically irrelevant."

So, exact same idea. There are three ways to do this, and remain NPOV: (1) we could leave the text as it is (the simplest way that preserves Ernst), or (2) add the fact that Vickers' response was to Colquhoun who made the same point as Ernst (the technically accurate way that preserves Ernst), or (3) remove Ernst and quote Colquhoun (for which a good case can be made, since Colquhoun's quote is actually in the literature). I wrote it as option #1 since I wanted to keep it simple, and I suspected some editors would prefer Ernst to Colquhoun. I'll change it to #2 and remove the OR tag. But 1 would be the easiest, if we can agree it doesn't rise to OR. --Middle 8 (talk)

Have both Ernst and Colquhoun, briefly summarized; don't have the response-to-the-response, it's too much. Alexbrn ^{talk|contribs|COI} 07:12, 2 January 2014 (UTC)

I put mention of Colquhoun in there but haven't quotes him for WEIGHT reasons (keep reading). Check out this version (the current one as of this comment). 3.15 lines to the Vickers review, 2 lines to Ernst's reply (the ratio being wildly unbalanced so far, relative to source quality), but then 3 more lines to Vickers' reply, which kinda brings the weight back around. It's also fair -- we get the full both sides of the debate. Ernst's point is fair and significant; unrebutted, Vickers looks kind of stupid, but his points are also fair and significant, even if one disagrees. --Middle 8 (talk) 07:33, 2 January 2014 (UTC)

I've boiled this down; we've got what the meta-analysis concluded, and what a couple of named people opined in response. The weight is apparent from where the views originate from, and doesn't come from word count. This is enough material for this one paper, in what is already a bloated section. Alexbrn ^{talk|contribs|COI} 07:49, 2 January 2014 (UTC)

A lot of pruning of 2012 relative to the 2013 review; why is that? --Middle 8 (talk) 09:03, 2 January 2014 (UTC)

It was more overgrown. Alexbrn ^{talk|contribs|COI} 09:14, 2 January 2014 (UTC)

Actually, I'd missed the fact that you'd pruned '13 at all when I made that comment, and both look reasonable. Restored the four different kinds of pain they looked at, in keeping with how we handle other reviews. --Middle 8 (talk) 10:27, 2 January 2014 (UTC)

Low level details?

"The review examined four different types of chronic pain: back and neck pain, osteoarthritis, chronic headache, and shoulder pain.[116]" This is too much detail. QuackGuru (talk) 18:25, 3 January 2014 (UTC)

Nah, mentioning that is consistent with how we handle every other review under the efficacy section. This was a big meta-analysis, in effect four in one paper. And readers deserve to know which kinds of pain it worked for. Wouldn't it be a tragedy if we didn't specify that, and readers then sought treatment for a kind of chronic pain OTHER than the four conditions that got a positive result? Then the acu-quack terrorists would win! Do you really want that on your conscience? Give it a rest; you got your way on a bunch of other stuff after Alexbrn trimmed this down; you need to compromise on some things. --Middle 8 (talk) 20:28, 3 January 2014 (UTC)

A 2012 meta-analysis conducted by the Acupuncture Trialists' Collaboration found "relatively modest" efficiency of acupuncture (in comparison to sham) for the treatment of some kinds of chronic pain, and on that basis concluded it "is more than a placebo" and a reasonable referral option.[116]"

The previous sentence says "for the treatment of some kinds of chronic pain,..." There is no need for the low level details. QuackGuru (talk) 20:49, 3 January 2014 (UTC)

WP:IDHT?

The reply above does not address three points that were made in the preceding comment.

Three reasons to keep one sentence.

1. Consistency; scope: Nah, mentioning that is consistent with how we handle every other review under the efficacy section. This was a big meta-analysis, in effect four in one paper. 2. Accurate medical information: And readers deserve to know which kinds of pain it worked for. (paraphrased: they might be misled into thinking it worked for something different) 3. Consensus: Give it a rest; you got your way on a bunch of other stuff after Alexbrn trimmed this down; you need to compromise on some things. --Middle 8 (talk) 20:28, 3 January 2014 (UTC)

The three points were not addressed. --Middle 8 (talk) 05:25, 4 January 2014 (UTC)

We need to be sure not to imply this is for chronic pain in general either by specifying "kinds of pain" or enumerating those kinds. I prefer shorter, but it's no biggie. Alexbrn ^{talk|contribs|COI} 05:30, 4 January 2014 (UTC)

It's better not to be vague. Only QuackGuru would call this kind of information "low level details". It's WP:TE and trollish, frankly. --Middle 8 (talk) 10:55, 4 January 2014 (UTC)

I also prefer it to be shorter. So I made this change. QuackGuru (talk) 21:49, 4 January 2014 (UTC)

Author citation

Re "for" or "by" the Acupuncture Trialists' Colaboration: Here's what it says at the top of the PDF of the article (emphasis mine):

Acupuncture for Chronic Pain
Individual Patient Data Meta-analysis
Andrew J. Vickers, DPhil; Angel M. Cronin, MS; Alexandra C. Maschino, BS; George Lewith, MD; Hugh MacPherson, PhD; Nadine E. Foster, DPhil; Karen J. Sherman, PhD; Claudia M. Witt, MD; Klaus Linde, MD; for the Acupuncture Trialists’ Collaboration

Guessing PMID 24146995 (cf. this ES) has a typo? FWIW, Vickers' response to responses says it was published "by" the Acu. Trialists' Collab'n. (Vickers AJ, et al. Acupunct Med 2013;31:98–100. doi:10.1136/acupmed-2013-010312) Not as if this especially matters. --Middle 8 (talk) 12:34, 2 January 2014 (UTC)

In the metadata record the Collaboration are included as an author: best to stick with what the secondaries say. Alexbrn ^{talk|contribs|COI} 12:41, 2 January 2014 (UTC)

(Add) I take it you know Vickers was (is?) the chair of the Collaboration? Alexbrn ^{talk|contribs|COI} 12:48, 2 January 2014 (UTC)

Hey, "by" is OK by me. --Middle 8 (talk) 15:44, 3 January 2014 (UTC)

Nausea and vomiting section

(Meant this to be about nausea; was tired and typed "pain" above instead). The section on efficacy for nausea and vomiting has got some outdated cruft and could use a cleanup. Most recent Cochrane source on PONV is straightforward enough: P6 acupoint stimulation prevents postoperative nausea and vomiting with few side effects. Will suggest wording later. --Middle 8 (talk) 20:24, 31 December 2013 (UTC)

Evidence Based Science

Correlates Biomedical

I am concerned that the top paragraph stating that no biomedical correlates have been discovered for aspects of acupuncture is incorrect. See citation below. I suggest amending it or deleting the antiquated assertion. The citation below involves prominent researchers from Georgetown University Medical Center, Harvard Medical School, U of M Dept. Anesthesiology and more. They note: "Phase II/III sham-controlled clinical trials have been successfully completed, and a broad range of basic research studies have identified numerous biochemical and physiological correlates of acupuncture...."

I am concerned that this type of misinformation, or at the very least, highly controversial and peer reviewed controverted data is in the top paragraph. Please comment or deletion is appropriate.

Hindawi Publishing Corporation Evidence-Based Complementary and Alternative Medicine Volume 2011, Article ID 180805, 11 pages doi:10.1155/2011/180805 — Preceding unsigned comment added by Acuhealth (talk • contribs) 18:16, 17 January 2014 (UTC)

See WP:MEDRS for Wikipedia's standards for sourcing biomedical information. That article is not MEDLINE-indexed, which is generally expected of biomedical sources. Zad68 18:22, 17 January 2014 (UTC)

Interesting deletion, but seems to contradict other aspects of the article. There are at least two papers in the references list from a peer-reviewed journal that have been part of this article for some years, yet they manage to stay in. 'Medical Acupuncture' is not MEDLINE-indexed, yet references 78 and 96 have not been deleted! Tzores (talk) 21:29, 19 January 2014 (UTC)

They're not being used to source biomedical claims, are they? Alexbrn ^{talk|contribs|COI} 22:00, 19 January 2014 (UTC)

Edits on "Legal and political status" vol. I

User:QuackGuru was kind enough to add some material about the German acupuncture trials here... You know, the set of large trials that resulted in acupuncture being added to the list of reimbursable services in the German statutory health system.
Interestingly, the text that was added by QG reads: "As a result of the trial's conclusions, some insurance corporations in Germany no longer reimburse acupuncture treatments.[195] The trials also had a negative impact on acupuncture in the international community."
I try to AGF in this case, but unfortunately I know that QG knows that the GERAC resulted in acupuncture being reimbursable in Germany (as we have been working on that article together)... To leave this tiny fact out, and instead only present the side-note material about some health insurances allegedy not reimbursing acupuncture anymore and alleged negative impact on the international community, is a disgustingly obvious attempt to skew the facts. I know we have our differences in perspective here, but there're some rules for chrissake. If you're willing to skew the facts in this way, maybe you should stop for a minute and ask yourself whether your POV got the better of you. --Mallexikon (talk) 05:36, 20 January 2014 (UTC)

See Talk:Acupuncture#Legal_and_political_status_recent_edits. We must use secondary sources. QuackGuru (talk) 05:43, 20 January 2014 (UTC)

Edits on "Legal and political status" vol. II

This source QG uses: "He, W.; Tong, Y.; Zhao, Y.; Zhang, L. et al. (2013). "Review of controlled clinical trials on acupuncture versus sham acupuncture in Germany". Journal of traditional Chinese medicine 33 (3): 403–7" [26] is an analysis done by Chinese doctors, published in a Beijiing-based journal. It does not constitute a reliable source regarding details about the German health care system. Their claim about some health insurances not reimbursing acupuncture anymore has not been echoed by any other source about the GERAC (and there are a lot them), it hasn't been backed up with a source by the article's authors, and they didn't specify on it either (which health insurance are they talking about?). It has to be suspected that their claim is just hearsay. I'd invoke WP:MEDSCI ("Be careful of material published in a journal that lacks peer review or that reports material in a different field" - in this case, a TCM journal that reports details about the German health care system) to delete this source, unless someone objects. --Mallexikon (talk) 05:41, 20 January 2014 (UTC)

Adverse events, redux

See ES here (it's actually five in one million; same logic applies) --Middle 8 (talk) 08:29, 27 January 2014 (UTC)

So acupuncture can have Medically significant outcomes! - probably worth mentioning for that reason. And 1 in 200k is not great odds. Alexbrn ^{talk|contribs|COI} 08:35, 27 January 2014 (UTC)

It's pretty low for rate of SAE's, AFAIK. But we should find a source. And remember, most SAE's happen when standard good practices aren't followed -- that should be explicit. --Middle 8 (talk) 13:29, 27 January 2014 (UTC)

The NHS

Apparently "recognizes" acupuncture "for certain minor conditions" (so we say), sourced to this and this—a fair summary of these documents? Alexbrn ^{talk|contribs|COI} 12:55, 27 January 2014 (UTC)

The sentence you mention is supposed to summarize multiple sources from gov't bodies, and it's not clear that it does so, so we should look at them all. Plus, see just below. --Middle 8 (talk) 13:26, 27 January 2014 (UTC)

This is the page we should be using for NHS: [27], and not really the other. It's actually a good template for what the lede could say, being a reliable source for summarizing different POV's. --Middle 8 (talk) 13:34, 27 January 2014 (UTC)

Effiacy in lede

The latter two para's of the lede are problematic; they need to be updated to reflect the best and most current sources. Also, we have an undue weight problem putting a few individual scientists against NIH, NHS and WHO -- that's just silly. We can unpack it in the article but in the lede, it suggests parity of sources, and as we've seen before, it's top-tier vs. bottom. That said, the older NIH and WHO sources (along with the '05 skeptic sources criticizing them) should be de-weighted compared to the more-recent NHS (which is not criticized by the skeptic sources cited).

I'm not sure chronic pain and PONV are minor conditions; better to just say what they are. It's not as if we have that many types with evidence for efficacy.

Re efficacy, we should have in lede: Cochrane on PONV (acu works as well as drugs), Ernst on Cochrane pain reviews (works for some conditions), Vickers re pain (which is the most recent and statistically rigorous meta-analysis to date), and [28] (Vickers re: little variance in outcome depending on type of Tx). We don't need to keep Ernst '06 (too old), but certainly should keep Madsen (little if any difference between sham and verum for pain) and Ernst '11 (similar to Madsen).

Finally, as mentioned above, the NHS summary is a good summary addressing multiple POV's, and could guide our wording. --Middle 8 (talk) 13:26, 27 January 2014 (UTC)

Adams 2011 text was deleted without consensus and replaced with a dated 2004 source that was not about child acupuncture

There never was any consensus to delete the Adam 2011 source. A 2011 review found that pediatric acupuncture is safe when administered by well-trained, licensed practitioners using sterile needles; however, there was limited research to draw definite conclusions about the overall safety of pediatric acupuncture.[169] The same review found 279 adverse events, of which 25 were serious.[169] The incidence of serious adverse events were few; the incidence of mild adverse events were estimated at 11.8%.[169] The most frequent adverse effects were bleeding and pain.[169] The incidence of serious adverse events was 5 per one million, which included children and adults.[169]

This edit was not an improvement.

A 2004 cumulative review showed that serious adverse events (SAE's) are frequently due to practitioner error, exceedingly rare, and diverse.[165] The rate of SAE's is on the order of 5 in one million, below that of many common medical treatments.[165] The most common SAE's are infection due to unsterile needles and injury—such as puncture of a major organ or nerve damage—due to improper placement of needles.[16][165] Most such reports are from Asia, possibly reflecting the large number of treatments performed there or else a relatively higher number of poorly trained acupuncturists.[16] Infectious diseases reported include both bacterial and viral infections.[164][166] Though very rare in practice, traumatic injury to any site in the body is possible by needling too deeply, including the brain,[166] any nerve, the kidneys, or heart.[167] Many serious adverse events are not intrinsic to acupuncture but rather to bad practices (such as improper needling or unsterile needles), which may be why such complications have not been reported in surveys of adequately-trained acupuncturists.[16]

A 2011 meta-review showed that serious adverse events are frequently due to practitioner error, rare, and diverse.[16] The same review found 95 cases of severe adverse effects, including 5 deaths.[16] The most reported adverse event was pneumothorax.[16] The most common encountered adverse event was bacterial infection.[16] Most such reports are from Asia, possibly reflecting the large number of treatments performed there or else a relatively higher number of poorly trained acupuncturists.[16] Infectious diseases reported include both bacterial and viral infections.[164][166] Though very rare in practice, traumatic injury to any site in the body is possible by needling too deeply, including the brain,[166] any nerve, the kidneys, or heart.[168] Many serious adverse events are not intrinsic to acupuncture but rather to bad practices (such as improper needling or unsterile needles), which may be why such complications have not been reported in surveys of adequately-trained acupuncturists.[16]

A lot of the text is duplication too. QuackGuru (talk) 18:20, 27 January 2014 (UTC)

I'm inclined to think all but the first two paragraphs of this section (which reference 2013 & 2014 systematic reviews) can be cut without losing anything valuable. Who knows, with such slimming, one day this article might be ... readable! Alexbrn ^{talk|contribs|COI} 18:35, 27 January 2014 (UTC)

A 2004 cumulative review showed that serious adverse events (SAE's) are frequently due to practitioner error, exceedingly rare,[original research?] and diverse.[166] This is an older source and "exceedingly" rare in not consistent with the newer source.

A 2011 meta-review showed that serious adverse events are frequently due to practitioner error, rare, and diverse.[16] Sourced text using newer source.

This is duplication. QuackGuru (talk) 18:54, 27 January 2014 (UTC)

@QG - you have to be kidding. Everyone but you joined consensus at Talk:Acupuncture#Rate_of_serious_adverse_events and I simply didn't make the edit till now. Your conduct in that section was unbelievable IDHT and this is just more. --Middle 8 (talk) 19:12, 27 January 2014 (UTC)

You have never explained your reason to delete the Adams 2011 source from the article. You also added a lot of duplication. QuackGuru (talk) 19:16, 27 January 2014 (UTC)

Removal of Adams 2011 was an error; my edit was meant to restor White as the ref to the sentence for which is was originally used. AS pointed out above -- in the section where consensus was reached -- White 2004 is the original source for the 5-in-on-milion number that Adams quotes. Proper form is to cite the original. --Middle 8 (talk) 19:33, 27 January 2014 (UTC)