Why I Accepted a PLOS One Article about Homeopathy for Depression

plos onePLOS One recently published Homeopathy for Depression: A Randomized, Partially Double-Blind, Placebo-Controlled, Four-Armed Study (DEP-HOM). I’m proud to have been the Academic Editor accepting this paper.

I wrote this blog post as an independent blogger to present my personal views about my decision to recommend acceptance, in the context of some larger issues.

The same day, PLOS Medicine published a paper evaluating acupuncture for depression in primary care. I tweeted (@coyneoftherealm) that if PLOS Medicine was going to keep publishing clinical trials of acupuncture without suitable sham acupuncture controls, I might have to resign as an Academic Editor at PLOS one.

Besides blogging at PLOS Mind the Brain, I’m an occasional blogger at Science Based Medicine. The heavily accessed blog site is well known for not mincing words in expressing contempt for complementary and alternative medicine (CAM) approaches, often referred to as SCAM in its blog posts. A couple of my posts (1,2) there were scathing criticisms of a PLOS Medicine article claiming acupuncture had effects equivalent to antidepressants and psychotherapy for depression.

I have not asked them, but I doubt many of my Science Based Medicine colleagues approve my accepting the homeopathy paper, especially if they were unaware of my rationale.

And then there is the inconsistency. Why, if I accepted a homeopathy paper, did I so strongly objected to publishing of an acupuncture paper? My recovery from a momentary lapse of reason?

I will explain, but offer no apologies.

Thanks to the open access afforded by PLOS One, you can get the article here.

A Clinical Trial of Homeopathy for Depression in PLOS One

The article describes an attempt to recruit patients into a four-armed randomized trial. A homeopathic remedy was compared to placebo. A homeopathic interview that involves a lot of history taking to personalize the choice of medication was compared to a more conventional shorter interview. Thus, the trial had a 2 x 2 placebo controlled design, with practitioners and patients blind to whether placebo or control was being administered.

The investigators intended that 224 patients would be randomized. However, despite extensive efforts, they were only able to recruit 44 patients. They abandoned their efforts and wrote up the study.

The investigators acknowledged that there was a lack of scientific rationale for homeopathic medicine.  They reported finding only one previous study in which its efficacy for major depression had been examined. There are actually more studies, but bad quality.

Their interest in conducting a trial was pragmatic.

Many depressed persons in the community are drawn to this treatment because of their belief that it is effective and lacks the side effects conventional medication or the extensive time commitment required by psychotherapy.

Homeopathy has been recommended by both Prince Charles and Mother Teresa. A controversial Swiss Health Technology Assessment concluded that homeopathy was safe and effective and resulted in continued reimbursement for treatment by Swiss insurance companies.

The German government funded this trial, presumably assuming that results of a well-designed clinical trial could settle the issue of efficacy in a way that could be persuasively communicated to the lay public and professional community.

Based on this inability to recruit patients, the investigators concluded

Although our results are inconclusive, given that recruitment into this trial was very difficult and we had to terminate early, we cannot recommend undertaking a further trial addressing this question in a similar setting.

They went on to explain why a further trial was not recommended.

How Is Homeopathy Supposed to Work?

Practitioners claim homeopathic medicine works by stimulating a self-healing mechanism in the body’s defense. This is accomplished by administering a substance that would cause the symptoms, except that is provided in very diluted form.

In the case of this clinical trial, it was diluted to a standard quinquagintamillesimal (Q or LM) potency, diluted 50,000 times.  But it’s important that homeopathic preparations not simply be diluted, they must be violently shaken between dilutions. Dilution just reduces potency, but dilution plus shaking or succussion as it is called, is believed to increase potency.

It is possible that there is not even a single molecule of the original substance left in the final, diluted remedy. So homeopathic remedies may consist of nothing but water. That does not bother homeopaths because they believe that because of dilution and succussion, the original compound leaves an “imprint” in the water that no longer depends on the substance still being physically present.

Saving Thousands of Lives With Homeopathy

A spoof video posted on YouTube by Myles Power Powerm1985-72936129announced that he was going to save thousands of lives by dumping small quantities of homeopathic remedies into Scottish streams that flowed into the North Sea. He obtained the remedies from a homeopathic first aid kit advertised on Amazon that promised cures for stroke, heart attacks, poisoning and drowning. The remedies would be appropriately diluted and people all over Europe and perhaps eventually the rest the world would obtain the protection of homeopathy against these conditions.homeopathic first aid kit

The problem with this spoof is that the joke had impeccable logic from a homeopathic perspective, except maybe that being rolled around in North Sea storms did not provide sufficient succession.

Many uses of homeopathy are drawn to claims that it is safe and draws upon the body’s natural healing potential. I doubt many users understand the dilution. Dana Ullman notes that Alexa Ray Joel, daughter of Billy Joel and model Christy Brinkley attempted suicide by taking an “overdose” of her homeopathic medicine.

Homeopathy as Evidence-Based Medicine

A Dana Ullman 2010 article in Huffington Post, Homeopathy: A Healthier Way to Treat Depression drew over 500 “likes.”  Ullman bills himself as an evidence-based homeopath.

Another Ullman article, the 2012 The Homeopathic Alternative to Antidepressants, is a spirited defense of the advantages of homeopathy over conventional antidepressants. Ullman is obviously aware of the scientific literature and draws freely, even if selectively interpreting articles that have appeared in New England Journal of Medicine and (ugh) PLOS Medicine to argue that antidepressants are no more effective than a placebo. Ullman also argues that even if antidepressants are effective in relieving the symptoms of depression, their effectiveness comes at the cost of frustrating the body’s natural reactions to depression and so any improvement obtained with cannot be expected to continue after stopping antidepressants.

A United Kingdom National Health Services (NHS) webpage denounces the lack of a scientific basis for homeopathy, cites the authoritative 2010 UK House of Commons Science Technology Committee Report on Homeopathy to argue that

The ideas that underpin homeopathy are not accepted by mainstream science, and are not consistent with long-accepted principles on the way that the physical world works.

As for the succession process, the NHS further quotes the 2010 report

We consider the notion that ultra-dilutions can maintain an imprint of substances previously dissolved in them to be scientifically implausible.

However,  the NHS article then wimped out, indicating NHS does not take a stand against homeopathic medicine and offers web links for referrals.

Why I Liked the PLOS Article

The article had a number of strengths in terms of trial design and a transparent reporting of what actually happened. No confirmatory bias here—or Barnum conclusion that further research is needed.

  • The protocol for the study had been pre-registered and was publically available.
  • The patients and the whole study team remained blinded to the identity of the four treatment groups until the end of the study.
  • The use of both a placebo control for the medication and a more conventional interview for the longer homeopathic interview.

This latter feature allowed for some control of the ritual, attention, and support with which homeopathic medications are delivered. Without the interview, homeopathic practitioners could argue that the medication was administered without appropriate personalization. Yet, knowing that depression is responsive to support inattention delivered with positive expectations, it was imperative to control for these elements of the treatment.

  • The write up of the trial complied with CONSORT in its transparent report of rationale, methods, and results.
  • The frank admission that the investigators failed in their effort to recruit sufficient numbers of patients and this  failure suggests that another attempt might not be warranted.

PLOS One is Not Just Any Journal

The PLOS One website notes

PLOS ONE will rigorously peer-review your submissions and publish all papers that are judged to be technically sound. Judgments about the importance of any particular paper are then made after publication by the readership (who are the most qualified to determine what is of interest to them).

PLOS ONE publication criteria are

  1. The study presents the results of primary scientific research.
  2. Results reported have not been published elsewhere.
  3. Experiments, statistics, and other analyses are performed to a high technical standard and are described in sufficient detail.
  4. Conclusions are presented in an appropriate fashion and are supported by the data.
  5. The article is presented in an intelligible fashion and is written in standard English.
  6. The research meets all applicable standards for the ethics of experimentation and research integrity.
  7. The article adheres to appropriate reporting guidelines and community standards for data availability.

Science-Based Medicine Instead of Evidence-Based Medicine?

Paul Ingraham, an editor at Science Based Medicine asked

Why “Science”-Based Instead of “Evidence”-Based?

And summarized a recurring theme going all the way back to the first post at the blog

The idea of emphasizing science in general instead of evidence in particular was first publicly proposed by Yale neurologist Dr. Steven Novella and infamous medical blogger and surgical oncologist Dr. David Gorski in early 2008, along with several other physician co-authors:

EBM is a vital and positive influence on the practice of medicine, but it has its limitations. Most relevant to this blog is the focus on evidence to the exclusion of scientific plausibility. The focus on evidence has its utility, but fails to properly deal with medical modalities that lie outside the scientific paradigm, or for which the scientific plausibility ranges from very little to nonexistent.


It is not that we are opposed to EBM, nor is it that we believe EBM and SBM to be mutually exclusive. On the contrary: EBM is currently a subset of SBM, because EBM by itself is incomplete. We eagerly await the time that EBM considers all the evidence and will have finally earned its name. When that happens, the two terms will be interchangeable..

A comment left on the inaugural SBM blog post

Why is homeopathy implausible? Among other matters, its signature proposition is implausible mainly because never in the whole of human experience or research has dilution of solutions been found to enhance their intrinsic physical, chemical or biological properties (Hormesis is a property of a few biological systems, not the consistent behavior of solutions that homeopathy requires). Thus, dilution doesn’t make our coffee taste stronger and we don’t expect otherwise no matter how much we shake or stir it.

You can find lots of posts at Science Based Medicine concerning homeopathy, including Harriet Hall’s fine discussion of homeopathy first aid kits and Steven Novella’s expression of upset over the Swiss endorsement of homeopathy.

What if

  • …I had rejected the article?

The authors could have gone elsewhere and presented the results with the more confirmatory spin and a call for further research.

Maybe they wouldn’t get published anywhere that would attract attention from anybody but homeopaths. But if so,  maybe the German government would be tempted to finance another trial that was less responsibly conducted and well reported.

  • …The trial had recruited a sufficient number of patients and found a significant effect favoring homeopathic medication when was it administered based on the extensive interview?

I might still have accepted the article, but would not be persuaded of the efficacy of homeopathic medication. I’m enough of a Bayesian to be unshaken by one trial in my belief that a scientifically absurd mechanism could produce effects. I would require the author to acknowledge the lack of a scientific basis for effective homeopathy on depression and propose other mechanisms, perhaps the greater ritual and positive expectations in the longer interview.

One trial does not undo 200 years of claims that are scientifically nonsense.

  • … I had been in a position to participate in the grant review that resulted in funding of the study?

I would say there is not a sufficient scientific basis for homeopathy that would justify the resources required of a well-designed study.

Just because I would accept this article, doesn’t mean that I would approve funding of the study, or be construed in collaborating in the study.

So why was I indignant that PLOS Medicine published a clinical trial comparing acupuncture to antidepressants?

Acupuncture similarly lacks a credible scientific explanation for its effects beyond the rituals in which it is administered. Appeals to ancient Chinese medicine are not scientific.

I would expect a sham treatment having the same ritual with provider and patient blinded would produce the same effect, unless some risk of bias had been introduced.

I think the lack of evidence for the mechanisms proposed by practitioners of acupuncture is sufficient to require that the role of rituals be tested with an appropriate control group, such as sham acupuncture delivered by someone blind to the purpose of hypotheses of the study.

The PLOS Medicine in question did not have an appropriate comparison group controlling for ritual. The authors were allowed to interpret the results with a confirmatory bias.

The article should not have been published in PLOS Medicine because it was scientifically flawed and the authors did not acknowledge the flaws. It greatly embarrasses me that this article got published, and should embarrass the editor who accepted it.

My reaction to publication of the article is to make a determined effort to educate PLOS  editors about the necessity of insisting on appropriate control groups. And about the need to protect the journal from those who would want to exploit its interest in a broader range of of articles to promote fake treatments based on bad science. I am also going to seek some sort of general recommendations from the PLOS management prevent this happening in the future.




25 thoughts on “Why I Accepted a PLOS One Article about Homeopathy for Depression”

  1. Some minor (but important) corrections to what you say about the Swiss homeopathy ‘HTA’.

    Please see That ‘neutral’ Swiss homeopathy report | Zeno's Blog

    Although we didn’t know it at the time Sven Rudloff and I wrote the blog post, the Swiss Government later stated that it was not an HTA at all (despite what its authors claimed) and that the version of the report that was published in English was published “without any consent of the Swiss government or administration”. In fact, its authors had added to what they had submitted to the Swiss Government commission (the PEK) a few years previously and included a ‘critique’ of the Shang et al. meta-analysis that had also been part of the Swiss Government’s investigation into homeopath and other therapies.

    The report as submitted was somewhat favourable to homeopathy and its flaws were recognized by the PEK, commented on and revised. The full story can be read in our blog post.

    The report has been roundly criticised elsewhere, eg:

    A critique of the Swiss report Homeopathy in Healthcare

    The Swiss report on homeopathy – a case study of research misconduct

    However, as a direct result of the report, homeopathy was removed from the Government health insurance reimbursement scheme. It was only after campaigning by homeopathists that it was temporarily added back in. Homeopaths have until 2015 to provide scientific evidence of the efficacy, appropriateness and cost-effectiveness of homeopathy or it will be permanently removed in 2017.


    1. Fascinating background. I think this situation may have influenced the German government to fund this, who clearly feels under pressure. They too have ferocious CAM and spa/inpatient psychotherapy lobbies.

      The outsider perspective is that CAM won in Switzerland. I know my colleagues have lamented this at the SBM blog.


      1. The German government didn’t fund this. The study was funded by the Carstens foundation, Germany’s largest lobbying organisation for altmed.


  2. As Alan points out, the Swiss thing was not an HTA report, and the “outsider perspective” is assiduously fostered by homeopaths, notably Dana Ullman, who promoted the Bornhoft paper as a “neutral Swiss Government report” and “Swiss Government finds homeopathy safe and effective”.

    In fact, Bornhoft and other authors were a part of the *political* process that led to a *temporary* reinstatement of reimbursement for homeopathy (it will be reviewed again, and will have to provide credible evidence of effectiveness, something homeopathy has never actually managed as you know).

    The overall Swiss PEK conclusion was exactly in line with the House of Commons Science and Technology Committee: http://www.publications.parliament.uk/pa/cm200910/cmselect/cmsctech/45/45.pdf

    The question of whether homeopathy is effective, is answered. Its doctrines are refuted, no remotely plausible mechanism of action has ever been proposed, and the conclusion of systematic reviews shows, in line with the expectation form the first two facts, no evidence that it actually works.

    Further trials are unethical. They can only have one of two results: a negative (which will probably not be published) or a false positive, which will be exploited by believers to further delay the day when their fraudulent claims are finally banned.

    Right now there is a “charity” raising money to send homeopaths to disaster areas, claiming as validation precisely the kinds of false positives that the Bornhoft report elevates above systematic reviews and the absence of credible mechanism. This is the West imposing delusions on desperate and sick people. It is utterly immoral.


    1. More fascinating background. But was does this mean?

      “In fact, Bornhoft and other authors were a part of the *political* process that led to a *temporary* reinstatement of reimbursement for homeopathy (it will be reviewed again, and will have to provide credible evidence of effectiveness, something homeopathy has never actually managed as you know).”?

      What is a “political” process in this context? Unofficial? Compromise of an otherwise legitimate process by illegitimate considerations? We can’t get away from the Swiss continuing to reimburse homeopathy, a very unfortunate situation.

      I am not advocating further trials, but think that it would not be necessarily unethical to conduct trials exploring alternative, science-based explanations like ritual, etc. I object to further trials of acupuncture except where there is an appropriate control for ritual as is obtained with sham acupuncture. Comparisons with sham acupuncture have actually served to undermine claims based on Eastern wisdom. Something similar might hold for homeopathy.

      You can subject an unscientific idea to a scientific test if you apply established methods that explore plausible mechanisms.


      1. James Coyne PhD said:

        We can’t get away from the Swiss continuing to reimburse homeopathy, a very unfortunate situation.

        The Swiss were maneuvered into a very difficult position. Their laws state that only treatments that have evidence of “efficacy, appropriateness and cost-effectiveness” can be reimbursed. The referendum that was held came out in favour of reimbursing homeopathy (after much campaigning by homeopathists, of course) – the results of referendums in Switzerland are binding on the Swiss Government. So, this put them in a difficult position and, as a compromise, they gave homeopathy temporary reimbursement status, pending one last attempt to provide that good evidence.

        I am not advocating further trials, but think that it would not be necessarily unethical to conduct trials exploring alternative, science-based explanations like ritual, etc.

        If you mean separating out the specific from the non-specific effects of homeopathy (ie the effect of the sugar pills vs the effect of the consultation), then that’s already been done and done fairly well.

        Conducted by a well-known homeopath, the robust study showed that the perceived improvement in rheumatoid arthritis was due only to the consultation. [1]

        1 Brien, Sarah, Laurie Lachance, Phil Prescott, Clare McDermott, and George Lewith. 2011. “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 (Oxford, England) 50 (6) (June): 1070–1082. doi:10.1093/rheumatology/keq234.


  3. The most important thing for skeptics right now is to get rid of the ridiculous made-up term “succussion” and replace it with the much more widely recognised and similar term “twerking”.

    Homeopathic remedies become more potent through dilution and twerking.

    Come to think of it, that’s probably the mechanism they’ve been searching for these past 200 years: prior to the advent of YouTube and the VMA awards science had no idea of the sheer power of twerking.


  4. Thanks for this. I agree with your reasoning and appreciate the value of homeopathy trials as a study of the role of ritual, expectation, manipulation, and so on, upon self-reported symptoms. Care does need to be taken to account for the effects of these aspects of an intervention upon self-reported symptoms. I also think that an equally cautious approach to designing trials and interpreting evidence needs to be taken for interventions which researchers consider to be more plausible.

    One concern I have with SBM is that it can serve to justify allowing one’s current assumptions to distort one’s interpretation of the evidence. That interpretations of EBM can allow clearly implausible treatments to be considered effective should be taken as a chance to learn about problems with the way in which evidence is currently interpreted for all trials. It seems unlikely that the problems which can distort findings for quack interventions do not also effect trials of ‘mainstream’ treatments.

    We should be just as vigilant to the dangers of response bias and the need for appropriate controls in trials of cognitive and behavioural interventions as acupuncture, and for pharmaceuticals as homeopathy. I suspect that a number of ‘mainstream’ interventions are no more effective than homeopathy, and are more dangerous for the fact that they seem plausible and respectable, as well as having greater potential for side-effects.


  5. Also, as far as I can see, this trial was *not* funded by the German government. The Veronica and Carl Carstens Foundation is the largest altmed lobbying organisation in Germany.


      1. By the way, the literature review they thank the Foundation for is notorious in German SBM circles, and has been dissected at length for reasons like the one you pointed out.


      2. OK, but the Declaration of Helsinki mandates that human trials should be based on the scientific evidence, should be likely to benefit the population represented by participants, and says that any intervention should be tested against the best current intervention.

        There does not seem to me to be any ethical way to conduct a human trial on homeopathy. The science says there’s no reason to expect it to work and no way it can work, the totality of available studies shows no compelling evidence of effect beyond placebo. In order to recruit even a single patient, the rules on informed consent would require that patients are told that there is no credible evidence that the intervention works (I have never seen a consent document for a homeopathy trial that says this).

        As editors and publishers, PLoS have a duty under the Declaration not to accept for publication, studies which fall below acceptable ethical standards. What was the informed consent regime around this trial?


      3. Perhaps, but like all studies sent out for review at PLOS One, this one cleared human subjects/Institutional Review Board. I am aware of NIH studies of homeopathy routinely clearing in the states, as well.


      4. Sure. Burzynski’s trials cleared IRB too. Turns out, his IRB has – ahem – issues. IRB signoff on an unethical protocol is not a form of alchemy that renders the protocol ethical.


  6. First of all, thanks to James Coyne for taking the time to explain his decision in such detail, and for being aware that this is a topic of wider interest.

    If all editors (of all journals and fields) would show that level of both, openness and self-reflectiveness, science would be in much better shape.

    However, I do have one (maybe slightly paranoid) fear , seeing that the study was funded by the Carstens Stiftung (NOT by the German government – it does make a difference!).

    One of the main arguments of the Carstens Stiftung and other homeopathy supporters in Germany (and I guess, elsewhere too) is some variation of the following:

    “RCTs don’t work for homeopathy, it’s too individual an approach , etc…”.
    Well, the paper you accepted might give them further ammunition along these lines :
    They may say now “look, we’ve tried soooo hard, but RCTs really don’t work for homeopathy.. let’s do Hahnemann’s ‘Arzneimittleprüfung’ instead and trust the users’ experience”
    [which, of course, is both bogus “proof” ].

    That said, I’d agree that’s a rather ‘political’ point and for my part I do understand that your task at hand was to judge the science reported in the paper, not what the homeopathy lobby would make of it.


  7. You really need to stop with the incessant “I’m going to resign as PLoS ONE editor” threats and whininess. Is that the only intervention you feel like you have at your disposal, and/or do you really see yourself as so important that your resignation would be such a major event? “My appeal got rejected, so I’m going to resign”. “If PLoS keeps publishing XYZ I’m going to resign”. “If PLoS doesn’t implement indexed comments I’m going to resign”. etc etc. These detract from the otherwise intelligent commenting that you provide.


    1. As the PLOS board has repeatedly admitted, it is a huge operation and change is difficult, as they say, like turning around the Titanic. I have quite a lot of validation from other Academic Editors for my concerns, have even successfully encouraged some to rejoin who had quit.

      There has been some risk that what has been quite a democratic revolution would revert to an imperial publisher-author relationship. I feel the need to keep that from happening. I speak for myself, but almost always discover I am speaking for lots of others.

      Stay tuned.


  8. My initial response to this blog title was stunned disbelief. How could anyone, with recognized credentials, get hoodwinked into accepting for publication nonsense related to homeopathy (I am a trained chemist and the notion that “Dilution just reduces potency” & “dilution plus shaking or succussion as it is called, is believed to increase potency” is simply laughable.)

    However, I read the full post and have to admire the logic of why this paper was accepted. The fact that the paper presents details on a well defined trial methodology is commendable. The fact that the trial didn’t go through is understandable, but how the trial results would have been reported is interesting speculation.

    While open access is by far the best approach to ensure fair access to taxpayer funded research, the trade-off is dealing with material on the periphery presented as sound science, with its concomitant negative impact on public discourse.


  9. On the presumption that you believe in Avogadro’s number, it’s obvious that the control pills are identical with the homeopathic pills. That being the case, positive trials of homeopathy should be published only as illustrations of the hazards of statistical methods that ignore the false discovery rate (e.g. http://rsos.royalsocietypublishing.org/content/1/3/140216 ).

    From that point of view, the study is quite interesting, because of its implications for real science. I was quite disappointed that PLOS One published this without making the obvious statistical point.


    1. In hindsight, I’m inclined to agree with you that it was an oversight to not require that point be made. You could certainly rectify the situation by going to the article in PLOS and leaving a comment. I recommend that you do so.

      My perspective in recommending publishing the article, was that in Germany and elsewhere, someone might be tempted to conduct a pragmatic trial demonstrating that homeopathy was ineffective. I wanted two points to be made. The first was that there would be ethical problems recruiting patients to a trial that would be unlikely to accrue sufficient patients for analysis. The second point is that homeopathy in Germany and elsewhere is administered with an elaborate ritual of talking to the patients and seemingly arriving at a tailored treatment. There could be a powerful placebo in that ritual that is absent in non-homeopathic routine care. I think the article makes a good point of this. I don’t think anybody would be inclined to advocate homeopathy based on this paper. But those who might be inclined for whatever reason to conduct a trial would be pushed to reconsider the feasibility as well as the necessity of an appropriate placebo control. It’s not about the inertness of the homeopathic remedy, it’s about the elaborate miss of the ritual with which it is delivered.

      Thanks for making an interesting point that should have been made in the review of this paper.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s