Talk:Acupuncture

IPD meta-analysis: weight issue

Meta-analyses are pretty much unsurpassed as MEDRS's, and the "gold standard" for meta-analysis is the kind using individual patient data (IPD).[1][2][3][4][5] For research on acupuncture's efficacy, there is one such study, Vickers et. al. 2012. It was four large IPD meta-analyses in one and there is plenty there to discuss. It's cited once in the article and isn't in the lede. Other meta-analyses are cited multiple times and are in the lede. This is pretty obviously a WP:WEIGHT problem.

This being a controversial area of research, there were multiple criticisms directed at Vickers, and some of them were valid and are included in the article. However, most of them weren't MEDRS, and those that were were on the low end (e.g., letters to the editor), and to that extent, weigh modestly. One wouldn't guess this from the article, though. Ernst, for example, criticized Vickers in a comment to The Guardian [6], yet the article gives that about the same space as Vickers itself. It's possible that Vickers' critics haven't put their criticisms in stronger MEDRS's because their criticisms are weak. As it happens, Steven Novella, a fairly well-known alt-med critic, said that he'd prepared a full paper criticizing Vickers, but that it was rejected for publication.

So the article is all out of whack with respect to Vickers, which is to be expected when tigers are afoot, as they so often are in CAM topics. This needs fixing; thoughts? --Middle 8 (t • c | privacy • COI) 11:02, 26 October 2015 (UTC)

No comments again? All right, I'll work Vickers into the lede and try to make the 2nd para on efficacy less of a cut-and-pasted laundry list (which is what it, and much else, turned into while this page was effectively under the ownership of QuackGuru). --Middle 8 (t • c | privacy • COI) 07:24, 4 November 2015 (UTC)

Could you stick to the topic and refrain from poking people? You might also bear in mind that tigers are much more in keeping with Wikipedia's WP:FRINGE approach—we know people feel better when someone claiming authority pays attention to them, and we know that practitioners use Wikipedia to promote their beliefs, but there is no knowledge regarding how the insertion of needles might be useful. Johnuniq (talk) 10:18, 5 November 2015 (UTC)

I haven't been following edits and discussions lately - but Johnuniq come on - you cannot be paying much attention to the literature if you feel confident making such a sweeping statement. Even if you only accept Ernst's review of reviews as reliable research into efficacy (which would be ridiculous, since there is more to clinical and biological research than just RCT's), you still would see the potential "usefulness" and the need for more research.Herbxue (talk) 20:20, 5 November 2015 (UTC)

@Johnuniq: (a) Single mention of another editor ≠ poking. (b) True re tigers on WP, but still, they come in multiple stripes. (c) Agree with Herbxue re what is known about the field (and I don't think academic centers would use it if they didn't think it was useful).... but yes, let's stick to the topic. --Middle 8 (t • c | privacy • COI) 23:11, 5 November 2015 (UTC)

This section has been set to not archive for months, so in response to the above perhaps I should highlight the critical word in my earlier comment: there is no knowledge regarding how the insertion of needles might be useful. That makes the topic a bit of a head-scratcher from a scientific point of view. Johnuniq (talk) 09:42, 29 November 2015 (UTC)

You're referring to mechanism, I take it. We don't need proven mechanisms to reasonably consider efficacy, just plausible ones. [7][8][9][10] But please let's not stray too far from the question at hand, i.e. summarizing and weighting efficacy reviews... --Middle 8 (t • c | privacy • COI) 22:44, 2 December 2015 (UTC)

Note: I've refactored further discussion of mechanism into its own section, below. --Middle 8 (t • c | privacy • COI) 13:15, 25 December 2015 (UTC)

No surprises here.

If Your Veterinarian Offers Acupuncture, Find a Different Vet. Guy (Help!) 01:26, 19 December 2015 (UTC)

Silly article. The actual review they cite concludes: "For cutaneous pain and diarrhea, encouraging evidence exists that warrants further investigation in rigorous trials. Single studies reported some positive intergroup differences for spinal cord injury, Cushing's syndrome, lung function, hepatitis, and rumen acidosis. These trials require independent replication. On the basis of the findings of this systematic review, there is no compelling evidence to recommend or reject acupuncture for any condition in domestic animals. Some encouraging data do exist that warrant further investigation in independent rigorous trials." This article is cherry picking and POV-pushing. No surprises here.Herbxue (talk) 21:17, 19 December 2015 (UTC)

Ah yes, the SCAM "get out of jail free" card: more trials are needed. And I support more trials, just as soon as they come up with a credible mechanism and good quality basic science research proving the existence of what they want to test. Anything else would, of course, violate the Declaration of Helsinki. I am still waiting for the proof that qi exists, but it seems oddly elusive :-) Guy (Help!) 22:43, 19 December 2015 (UTC)

The "more studies needed" conclusion is commonplace in medical research of all kinds. This systematic review found encouraging evidence of efficacy - in a population that supposedly does not experience the placebo effect. Herbxue (talk) 01:18, 21 December 2015 (UTC)

I'd be pretty surprised to hear anyone familiar with the scientific texts to criticize the "further investigation is needed" clause in the conclusions of each scientific paper. That's the case we face with the human sciences, be it medicine, economics, or sociology; nothing isn't really carved in stone, and the scientists do hedge their position. Moreover, we do not need - nor do we allow - a second peer-review by anonymous Wikipedia editors. Cheers! Jayaguru-Shishya (talk) 21:04, 2 January 2016 (UTC)

That's irrelevant, because the context is SCAM studies. Any SCAM study with a weak result, especially a negative result that's been spun into a positive one by P-hacking, has this text. SCAM proponents always keep asking until they get the answer they want, and since no competently conducted study ever delivers the answer they want, the conclusion is always that they should keep asking. And, per the Declaration of Helsinki, the vast majority of their trials are unethical, because there is no remotely plausible mechanism by which the study population can benefit form the results of the research (the only way they could would be if fraudulent treatments were dropped, and as we know, no negative result is ever considered valid by SCAM proponents, and many are never even published). Guy (Help!) 09:45, 21 December 2015 (UTC)

Are you suggesting an edit, or just using this as a forum to vent and make stuff up?Herbxue (talk) 20:50, 21 December 2015 (UTC)

I don't see the "summary" you posted above as at all positive. However, Guy, the Declaration doesn't apply to veterinary research. (We could discuss some of the more unethical phrases in the Declaration at a later date, in a different forum.) — Arthur Rubin (talk) 17:59, 22 December 2015 (UTC)

If 'SCAM' isn't some kind of acronym that I am unaware of, then what makes a study's context an a priori 'scam'?

In real science (not sure if medicine counts as such) there are many examples of 'scientific facts' that once had consensus as well as a lack of 'plausible mechanisms' to support alternatives (e.g. the earth standing still and NOT rotating or going around the sun), but that were later determined to be wrong. Having said that, I personally consider acupuncture to be a (originally well-intended) pseudo-science. But I base that view primarily on the outcome of studies (i.e. checking the claims) rather than the largely irrelevant lack of a 'plausible mechanism'.

Try to convince astronomers that Dark Matter is an a priori 'SCAM' because there is no 'plausible mechanism' (yet)... 110.23.118.21 (talk) 09:05, 8 January 2016 (UTC)

SCAM: Supplements, complementary and alternative medicine. Guy (Help!) 00:09, 14 January 2016 (UTC)

I think a translation is required. "further investigation is needed" = "Please continue funding my work" -Roxy the dog™ woof 09:09, 15 January 2016 (UTC)

JzG's removal of overview of SRs

JzG claimed in his revert of my edit in which I added this article that the journal in which it was published is "low level". [11] However, this is not true: the journal in question, Scientific Reports, is published by the prestigious Nature Publishing Group and has an IF of 5.578. [12] It is, however, true that, as JzG also said, the authors of the article (an overview of systematic reviews) are Chinese. But I don't think this is necessarily, in and of itself, a good reason for excluding this article, which meets the MEDRS criterion about being a secondary source exceptionally well, being an overview of systematic reviews. I would like to hear from other editors as to whether they think this overview belongs in this article. Everymorning (talk) 02:22, 14 January 2016 (UTC)

Agree with Everymorning. The ethnicity of the authors is irrelevant, what matters is their scientific background and the quality of the journal. The article is written by researchers of the prestigious Chinese University of Hong Kong, and published by a reputable scientific journal, and should be considered reliable. -Zanhe (talk) 04:42, 14 January 2016 (UTC)

Unsuitable. Did anybody look at the provenance of the authors. I'm surprised it was added. Interested to see other comment. -Roxy the dog™ woof 08:48, 15 January 2016 (UTC)

The source is reliable. JzG is pulling a Donald Trump. Roxy, how much do you know about those specific institutions? Herbxue (talk) 21:52, 15 January 2016 (UTC)

That statement is deeply (and I think deliberately) offensive. The problem of Chinese studies of quackery is documented in reliable independent sources. It is a systemic bias based on well understood cultural norms. It is no more racist to describe this than it was racist for Peter Piot to describe the spread of AIDS among sex workers in African countries. Here's a paper addressing precisely this area: bias in Chinese studies of acupuncture. [13] Guy (Help!) 00:07, 16 January 2016 (UTC)

The only pic I have ever seen of Guy, he was wearing a bicycle helmet. I doubt an experienced editor like him would dare to sport a Trump haircut. -Roxy the dog™ woof 08:35, 16 January 2016 (UTC)

The accusation of racism is not trivial, and is deeply offensive. The issue of systemic bias in Chinese studies has been discussed before. The rhaphiophiles have already been shown this evidence, they just don't like it. Guy (Help!) 09:39, 16 January 2016 (UTC)

The "bias in Chinese studies of acupuncture" you cited above is itself written by mainland Chinese researchers, which shows how ridiculous it is to dismiss papers purely based on the ethnic background of its authors, especially those published in peer-reviewed Western journals. If the Nobel Prize committee finds it suitable to award a Chinese scientist with a strong focus on traditional Chinese medicine (Tu Youyou) the Nobel Prize for Medicine, there's no reason for us to dismiss all Chinese research as unreliable. -Zanhe (talk) 20:53, 16 January 2016 (UTC)

That is rather an odd view. The fact that the authors are Chinese shows that the report is not motivated by anti-Chinese sentiment. I can't think of anyone more appropriate to analyse the systemic bias in Chinese studies. And this is not the only paper to find this - for example, Ernst has also noted that almost all Chinese-authored studies of acupuncture are positive. And the context here is a study by Chinese authors purportedly finding benefit in some random conditions for which no consistent plausible mechanism of action exists. That speaks of P-hacking, if nothing else. Guy (Help!) 22:26, 16 January 2016 (UTC)

The Trump comment was heavy-handed in retrospect, I apologize, it was intended to be a light-hearted jab . If you refer to the discussions about including ethnicity as a way of excluding sources, the consensus was formed that we did not need to protect against using ethnicity to exclude a source because it was so absurd to do so that nobody would ever do it. But here we are again. It is not acceptable to dismiss a source on the basis of ethnicity, period. You have to look at the quality of the journal, and the body of work of the individuals. I highly doubt you have had enough exposure to Chinese medicine researchers in Asia to appropriately assess the quality of their work. Ernst's article citing one (Chinese) review of SOME poorly conducted studies is just not enough to rule out all Chinese journals, much less articles in respected Western journals written by Chinese authors. There may be real concerns with this paper, but I haven't seen a REAL concern raised yet.Herbxue (talk) 17:37, 16 January 2016 (UTC)

A study that contradicts the consensus view, by aggregating studies with known risk of bias. This is exactly the sort of source we should not include. Remember the great sub-prime mortgage crisis? It happened because people aggregated crappy mortgages and pretended that the aggregate was somehow not crappy. Guy (Help!) 22:28, 16 January 2016 (UTC)

Just so I am clear, here we have a systematic review (secondary source) that is published in a high impact, western medical science journal that is indexed in Medline and published by the Nature Publishing Group, but it is being rejected by an administrator (Guy) because the authors are Chinese (added:) and authorship Nationality is a potential source of bias? Is this a correct interpretation of the issue here? Can we really reject a source based purely on the Nationality of the authors? 2001:56A:75B7:9B00:81A8:1EE6:5502:9C83 (talk) 23:49, 16 January 2016 (UTC)

That is a complete misreading of the situation. If you want to make a comment, try responding to the post at 00:07, 16 January 2016 above. Johnuniq (talk) 01:14, 17 January 2016 (UTC)

Johnuniq, the only concern raised by Guy is with regard to author Nationality as a potential source of bias. I believe that the source under question must be considered on more than one potential source of bias, in this case, the Nationality of the authors. Where is the consideration of this being a secondary source? or that it is published in a very high impact journal? or that the journal is a Western medical journal (not a Chinese publication)? or that the journal is published by a highly credible scientific publisher? As such, it seems that the basis for Guy's rejection is based purely on the fact that the Nationality of the authors as a potential source of bias. However, the balance of variables to consider seem to weigh heavily in favor of including the source. PS. Johnuniq: I have added a qualifier to my comment above to ensure that it is understood that my concern is with the emphasis of a single variable on inclusion decision, rather than mistakenly give the suggestion that it is a racial issue.108.181.201.237 (talk) 02:48, 17 January 2016 (UTC)

Did you examine the paper linked in the post I mentioned? That's the part which needs a response. There is no need to speculate about what Guy thinks because he wrote very clearly and provided a link to demonstrate his point. Johnuniq (talk) 04:04, 17 January 2016 (UTC)

Yes, I read the paper, but it is not at all a convincing argument for exclusion of this secondary source. Have you read the results of the 2011 study posted by Guy? It found that most systematic reviews of acupuncture that are published in Chinese journals did not use the most up-to-date Cochrane risk of bias tool available, but it found that ~75% did do some kind of 'inferior' risk of bias assessment (ie: using older Cochrane tools, or the Jadad scale were most common); reporting was also often incomplete. I could actually see this as a reason to consider excluding a source that was published in a Chinese journal, especially if it was itself assessed in the review posted by Guy and found to be flawed. However, in this case, it is a very far-fetched reach to use the 2011 article posted by Guy that critically assessed reviews published in Chinese journals to suggest that a 2015, peer-reviewed secondary source, published in a medline-indexed, high-impact, Western medical journal from a very reputable publisher should not be included in the wikipedia article.2001:56A:75B7:9B00:5923:32E2:CCE:307D (talk) 03:13, 19 January 2016 (UTC)

Until WP:MEDRS is changed, we shouldn't exclude sources on the basis of the country of origin of the authors. As you know, this has been discussed at length and there's no consensus for making this change to the guideline. This source appears to meet the criteria set by Wikipedia's guidelines. TimidGuy (talk) 12:08, 19 January 2016 (UTC)

• Another systematic review/meta-analysis has just been published in the same high-quality journal and might also be worth adding to the article here. Inclusion might as well be discussed along with the other source as they are both relevant secondary sources from the same MEDRS-compliant publication. See: [14] — Preceding unsigned comment added by 75.152.109.249 (talk • contribs)

It might be useful if reliable sources could actually be produced to state on a good empirical basis that Ernst, also Yuyi Wang, Liqiong Wang, Qianyun Chai, and Jianping Liu are wrong about the unreliability of trials of acupuncture from within the Chinese academic system? Other sources that support great caution with results from certain academic jurisdictions: Controlled Clinical Trials. 1998 Apr;19(2):159-66. Do certain countries produce only positive results? A systematic review of controlled trials. Vickers A, Goyal N, Harland R, Rees R. (China, Japan, Taiwan, Russia are mentioned.) BMJ. 1999 Jul 17; 319(7203): 160–161. Review of randomised controlled trials of traditional Chinese medicine. Jin-Ling Tang, Si-Yan Zhan, and Edzard Ernst Chinese sources only. Can anyone give us a reliable source to contradict these? Richard Keatinge (talk) 16:16, 28 January 2016 (UTC)

First, not sure about Ernst's blog (not MEDRS, so not relevant here), but Yuyi Wang, Liqiong Wang, Qianyun Chai, and Jianping Liu did not say trials of acupuncture published in Chinese journals are unreliable, they said that they commonly did not use the most up to date cochrane risk of bias tool (they used other risk of bias tools) thus there is the potential for bias in such articles in chinese journals. This does not invalidate them, it just adds a potential source of bias. Second, perhaps someone can explain why we care what Yuyi Wang, Liqiong Wang, Qianyun Chai, and Jianping Liu have to say about articles published in Chinese journals when we are discussing reviews published in a mainstream, western medical journal with high impact factor and solid publisher.2001:56A:75B7:9B00:F4B4:1A16:8F56:249C (talk) 16:26, 28 January 2016 (UTC)

Ernst is a highly notable commentator who conveniently sums up an issue. And we are discussing work originating from (and often, but not always, published in) specific academic jurisdictions. I hope this helps. Richard Keatinge (talk) 16:36, 28 January 2016 (UTC)

Does not help with the discussion at all Richard Keating. Ernst's blog is a self-published source and carries no weight in a discussion of MEDRS sources, plus it is irrelevant because he is summing up an issue with Chinese journals and we are not discussing Chinese journals. The source produced by Guy, and discussed in Ernst's blog, states in the discussion: "We only selected SRs and MAs published in Chinese journals and therefore our results are only applicable to those journals." Here we are discussing inclusion of 2 systematic reviews published in a Western medical journal. No editor has provided any convincing rational for why we would exclude 2 secondary sources from a Western medical journal on the basis that an older review suggests there is the potential for bias in articles published Chinese journals.75.152.109.249 (talk) 17:15, 28 January 2016 (UTC)

To repeat, we have RS saying clearly that certain types of work from certain jurisdictions is not really science because those jurisdictions only publish positive results. Rehashing the positive results really isn't helpful. It would be helpful to find RS that give good reason to believe that the papers presented are in fact reliable. Until then we should use the relevant types of work, from the relevant jurisdictions, very carefully indeed and only within a clear frame of their limitations. Using them, without warnings, to support a statement in Wikipedia's voice is not good. Richard Keatinge (talk) 18:08, 28 January 2016 (UTC)

Here, we are not discussing any work from those jurisdictions you speak of (Chinese journals), so there is no logic to your statement. We are discussing work from a western journal, not a Chinese journal. Where is the source that says articles about acupuncture published in Western medical journals are not really science? That is what would be required to suggest that the reviews being discussed here are not worthy of inclusion.75.152.109.249 (talk) 18:21, 28 January 2016 (UTC)

The problem is with the entire corpus of papers on which your source relies. Publishing meta-analyses etcetera of work known to be unreliable is only as good as the primary work. Garbage in, garbage out. Richard Keatinge (talk) 18:38, 28 January 2016 (UTC)

The problem is with systematic reviews published in Chinese journals and the problem is a risk of bias, not 'lack of real science'. The systematic review under question here is not in a Chinese journal and only had 13% of the studies it examined come from Chinese publications (a small minority). Moreover, the review under discussion did a risk of bias assessment for all of the articles that it included (as per appropriate systematic review procedure) and was deemed appropriate for publication in a well-respected western medical journal. There is no rational grounds for excluding this review.75.152.109.249 (talk) 18:48, 28 January 2016 (UTC)