Homeopathy in Healthcare: Effectiveness, Appropriateness, Safety, Costs

Against allopathy Alternative medicine

Clinically unproven

Qi Traumeel Acutonics Appeal to nature Applied kinesiology Gillian McKeith

Woo-meisters

American Holistic Medical Association Burzynski clinic Michael Uzick Allen West Jim Bakker Dee Snider v - t - e

Homeopathy in Healthcare: Effectiveness, Appropriateness, Safety, Costs (aka "The Swizz Report") may refer to one of two documents: it is the title of a report to the Swiss healthcare commission by a group of homeopaths and anthroposohists which began from the premise that homeopathy is effective and popular and from this concluded that — surprisingly — homeopathy is popular and effective. The second document is an edited and expanded version translated into English from the original German, and it is this which is better known, not least through assiduous promotion by Dana Ullman.

Misrepresentation of the report by homeopaths is so widespread that the Swiss Federal Office of Public Health (FOPH) have published a letter in Swiss Medical Weekly explicitly stating that the report is not officially endorsed and reiterating the fact that the FOPH recommended against homeopathy and other scams, and that they are currently reimbursed for political, not medical, reasons.

What it is not

Homeopaths claim a few things about the English translation which are false (and where applicable also false in respect of the original document).

• They claim that it is a report by the Swiss government. It is not. It is a report (or more accurately a polemic) addressed to the Swiss government by a group of true believers.
• They claim this is a translation of the original. It is not the original report but an updated version which includes commentary on Shang et. al. (2005), a document which was also an outcome of the same investigation process as this report.
• They claim that it supersedes reviews such as Shang et. al, 2005. It does not. Shang represents scientifically robust work and is an objective assessment of homeopathy’s efficacy (or lack thereof). This document no more supersedes that review than the latest glossy output from Scholars for 9/11 Truth supersedes the NIST investigation into the twin towers collapse.
• They claim that it is neutral because the Swiss Government is neutral. This is a deliberate abuse of language: the Swiss were militarily neutral during the World Wars, but this has no bearing on their scientific neutrality, and in any case it is not a report by the Swiss Government in the first place.
• They claim the Swiss Government chose, on the basis of this document, to reimburse homeopathy under their Federal health scheme. They did not. In fact, the review process recommended against funding homeopathy (and anthroposophy and various other implausible forms of CAM). Agitation by CAM apologists led to a referendum in which the public opted to extend reimbursement for a further five years, at the end of which it would have to demonstrate its legitimacy.

It may be confidently predicted that the outcome of the final investigation will be exactly the same as every other investigation of homeopathy: apologists asserting "it works for me" based on endless anecdotes, and scientists saying that it is implausible, violates laws of nature, is supported by very badly constructed evidence, and is contradicted by much better quality evidence which shows no effect beyond placebo.

This will end the argument exactly as effectively as the scientific consensus on climate change has ended the controversy about anthropogenic global warming, and the scientific consensus for evolution has ended the attempts by creationists to assert parity within the education system for the "competing theories" of evolution and creation.

What it is

The report is, by all accounts, a very shoddy piece of work indeed.

A case study of research misconduct

A good overall summary was published in Swiss Med. Wkly. 2012;142:w13594:

The report was commissioned following a 2009 referendum in which the Swiss electorate decided that homeopathy and other alternative therapies should be covered by private medical insurance. Before implementing this decision, the government wished to establish whether homeopathy actually works. In February 2012 the report was published in English and was immediately proclaimed by proponents of homeopathy to offer conclusive proof that homeopathy is effective. This paper analyses the report and concludes that it is scientifically, logically and ethically flawed. Specifically, it contains no new evidence and misinterprets studies previously exposed as weak; creates a new standard of evidence designed to make homeopathy appear effective; and attempts to discredit randomised controlled trials as the gold standard of evidence. Most importantly, almost all the authors have conflicts of interest, despite their claim that none exist. If anything, the report proves that homeopaths are willing to distort evidence in order to support their beliefs, and its authors appear to have breached Swiss Academies of Arts and Sciences principles governing scientific integrity. —The Swiss report on homeopathy: a case study of research misconduct

This analysis by Shaw zeroes in on probably the founding flaw in the entire work, as revealed particularly in this comment:

If homeopathy is highly likely to be effective but this cannot be consistently proven in clinical trials, the question arises of what conditions are needed for homeopathy to show its effectiveness and realise its potential, and what conditions threaten to obscure this? —Bornhöft, p. 198

This reveals, as Steven Novella notes, that the authors, frustrated by the fact that the best science shows homeopathy to be useless, seek evidence that instead confirms their intended finding.

Fradulent misrepresentation

Misrepresentation is endemic in the Swiss report. An obvious example is their evaluation of Efficacy of homeopathic arnica: a systematic review of placebo-controlled clinical trials, Ernst and Pittler, Arch. Surg. 1998 Nov;133(11):1187-90.

Eight clinical studies were subjected to descriptive analysis, the inclusion criteria being: homeopathy and Arnica species. Three showed no difference between A. verum and placebo, among them also two of the three higher quality studies; five showed a trend in favour of homeopathy.

Compare this with the study as published:

RESULTS: Eight trials fulfilled all inclusion criteria. Most related to conditions associated with tissue trauma. Most of these studies were burdened with severe methodological flaws. On balance, they do not suggest that homeopathic arnica is more efficacious than placebo.

CONCLUSION: The claim that homeopathic arnica is efficacious beyond a placebo effect is not supported by rigorous clinical trials

So, the authors cherry-picked through the report and represented it as a collection of data overall somewhat positive to homeopathy, whereas in fact the report concluded the exact opposite.

Why would they do that?

A few people have looked into this:

Edzard Ernst identified a number of structural flaws:

• A substantial undeclared conflict of interest. The authors stated no conflict of interest, but all are practitioners of or otherwise financially vested in homeopathy and have a direct financial interest in state reimbursement of homeopathy.
• "Consistent with their exotic views on research methodology, the authors of the report include all types of study designs in the HTA [Health Technology Assessment]. The authors make it clear that ‘methods that involve no experimental change in the medical intervention (such as case studies) are considered preferable, as they are better able to reflect the healthcare reality’. The findings of such methods are given the term ‘real-world effectiveness’, and it is such results that form the focus of the authors’ attention. More critical thinkers might point out that, by adopting this bizarre strategy, the authors effectively create their own standard. This evaluation of homeopathy thus significantly deviates from the rigour generally expected of HTAs." — in other words, the authors deliberately invert the normal hierarchy of evidence quality, giving the most objective form of evidence the lowest weight and the least objective, most weight. The only obvious reason for this is that this produces the result the authors set out to achieve.
• The authors excluded some systematic reviews. No objective criteria were given for exclusion. The excluded reviews tended to be those which reported negative outcomes.
• Some reviews were quoted as producing results in favour of homeopathy, but the text was quote-mined and the reports in question found on balance that results were at best inconclusive.
• A systematic review of A. montana was stated to include five studies which supported effectiveness, and therefore claimed as supporting evidence, but the conclusion of the systematic review was that: "On balance, they [the RCTs] do not suggest that homeopathic arnica is more efficacious than placebo."
• Ten of 22 reviews were rated by the authors as more positive than a review by Shang et. al, which scored using objective criteria.
• The report included no input whatsoever from anybody even remotely skeptical of homeopathy.

Special pleading

Andy Lewis of Quackometer analysed the report and noted that the original review panel had said:

[I]t is very obvious that all or some of the authors have a positive attitude towards the treatments in question or are convinced about their efficacy. Unquestionably, strict proponents of the usual hierarchy of evidence will regard the presented evaluations as scientifically untenable and unreasonably positive… —HTA review panel

Lewis notes that the authors begin by asserting:

The dilutions' mode of action requires a homeopathic explanation. Homeopathic remedies are backed by 200 years of empirical observation of millions of patients which shows that high potencies are often particularly effective if they are optimally matched to the patient's individual symptom picture. The mode of action of homeopathic remedies cannot be demonstrated with modern scientific methods. Due to a misapplied positivism that sees the reality of nature merely as the sum total of its measurable and quantifiable phenomena, a ‘lack of evidence’ is often seen to mean the same as ‘lack of effectiveness’.

This is a blatant example of appeal to antiquity and is also begging the question, since whether or not homeopathy works was precisely the question the Swiss Government had set out to ask; beginning by asserting that it works is entirely unacceptable. This single paragraph also introduces a third fallacy: special pleading, the claim that traditional (i.e. scientifically objective) forms of analysis cannot be applied to homeopathy. No explanation is given for this other than vague assertions about its being "special". In fact, homeopaths can and have conducted scientifically robust randomised controlled trials, with the predictable result: homeopathic tablets are placebos.

The authors claim:

Future research methods must respect the unique qualities of homeopathy by attaching more weight to single case evaluations, by including practically and expertly applied homeopathic treatment into research and clinical practice in order to identify its real potential and limitations. The recently introduced outcome studies are promising in this respect as they do not focus on specific effect but on the overall practical treatment and patient care in homeopathy.

This, again, is a reversal of the usual hierarchy of evidence. Single case reports are, by common consent, the least reliable form of evidence that is used in medicine. The authors seem to want to use these because they are the most likely to produce favourable results. That is, of course, precisely why medicine weighs such things very lightly: the probability of false results, confirmation bias and even false results due to simple stochastic variation means that very little of value can be drawn from individual cases.

Where the authors went wrong

Section 3 sets out the authors' introduction to the "speciality of homeopathy", which they do with all the objectivity of a comedy Jewish mother.

They begin with a bold statement: "Administering homeopathic remedies and mixtures merely on the basis of a clinical diagnosis goes against the principles of homeopathy." In other words, the vast majority of uses of homeopathy, being over-the-counter purchases, are not — according to the authors — valid homeopathy. Unless, of course, you want to prove how popular homeopathy is, or include any study that is favourable to homeopathy, however conducted.

They describe the "law of similars", neglecting to mention that there is absolutely no independent evidence to support this supposed law, and that, in fact, it is contradicted by rather a lot of other research.

They describe the supposed principle of potentisation or dynamisation, noting that

The fact that ever higher potencies were used in the history of homeopathy was a frequent bone of contention.

Is that bone of contention a thing of the past? Not hardly, though it is true that it was identified as a problem quite early, even before we knew that the nature of matter made it physically impossible (Holmes, 1842).

From a conventional pharmacological point of view, a medicinal effect seems most improbable, as according to Avogadro’s number, no molecule of the medicinal substance will be left in the solution from the 24^th decimal or 12^th centesimal potency. The dilutions' mode of action requires a homeopathic explanation.

So, when the laws of nature tell you that you are wrong, an explanation is required which is outside the laws of nature. This is classic woo territory, of course.

They don't know what they are supposed to be doing

According to the authors,

HTA is short for Health Technology Assessment, an established scientific procedure which, in contrast to the meta-analyses and systematic reviews specified by the Cochrane Collaboration Standards, examines not only the efficacy of a particular intervention, but especially also its ‘real-world effectiveness’, its appropriateness, safety and economy. HTAs are therefore much wider in scope and politically more informative. They include material that is ‘normally’ not taken into consideration, such as observational studies.

HTA is usually defined as: "a multidisciplinary field of policy analysis [that] studies the medical, social, ethical, and economic implications of development, diffusion, and use of health technology." It is not "in contrast to" meta-analyses; assessment of objective merit is a core part of an HTA, and any intervention which cannot demonstrate effect will be shown by a competent HTA to be an ineffective intervention. "Real world effectiveness" is a canard – it does not matter how many "it worked for me" anecdotes you collect; if the intervention is shown by well-designed studies to be a placebo, then an honest HTA will not invert this and claim that "real world effectiveness" — as measured by diligent application of the post-hoc fallacy — somehow overcomes that fact.

The statement that Cochrane does not consider factors such as safety and efficacy is simply false.

The phrase "politically more informative" is special pleading; where an intervention fails objective scientific tests, politics may be abused (as it was in Switzerland, with a referendum decision promoted by the authors' own Yes campaign to allow a further trial of homeopathy despite lack of evidence of effectiveness beyond placebo), but this does not validate the intervention any more than North Carolina can stop sea levels rising by making it illegal to predict that they will rise.

They lie about their own importance

This is a doozie. The authors assert in the Introduction that:

In contrast to this subsidiary result, which was of little relevance for the political decision, the much more comprehensive and differentiated HTAs ascertained that the individual CAM interventions, especially homeopathy, were effective, under Swiss conditions safe and, as far as could be judged from the trial situation, also cost-efficient.

If the "subsidiary result" (the scientific meta-analysis by Shang et. al. which showed that homeopathy is not better than placebo) was of "little relevance for the political decision", how come the political decision weighed that evidence more strongly than the Swiss report? How come the final verdict was not to reimburse?

In fact it was this report, not Shang, which turned out to be of "little relevance" because it was (rightly) downgraded due to blatant authorial bias. The idea that homeopathy is cost-effective relies on it being effective, and the science shows it is not. The authors assert that you should ignore the science; PEK, the commissioners of the report, disagreed. The authors lost.

They lie about the science

The authors state:

The "negative result" for homeopathy caused a massive furore prior to conclusion of the PEK project and following its subsequent publication (Shang et al. 2005), culminating in the unfortunately titled Lancet editorial "The end of homeopathy" (editorial 2005).

There is no need for the scare quotes. The conclusion of Shang et al. was soundly negative, as was Linde et al. before them. Science has learned how to weed out poorly constructed trials and weigh the evidence by quality first before quantity. The furore existed only because homeopaths created it, and the only unfortunate thing about the Lancet editorial title is that it turned out not to be the end of homeopathy.

As always, homeopaths prefer to use emotion and distortion rather than science.

They even lie about the question they were put

According to the authors,

The PEK programme was prompted by the high demand and widespread use and acceptance of complementary medical treatment, as well as by the political hope for its economic and preventive use.

This is false. CAM proponents like the authors might have "hope for its economic and preventive use", but the PEK programme was prompted by demands by CAM to re-evaluate after CAM modalities were withdrawn due to the strong scientific consensus that most CAM is bordering on fraudulent. The "hope" existed only in the minds of those selling objectively unverifiable remedies, and they were disappointed as the result was not to reimburse unproven CAM.

Some more of the authors' greatest mistakes

These criticisms are the tip of the iceberg. A detailed reading of the document shows many more fundamental flaws.

Special to us, unless...

The authors frame the review in terms of the situation in Switzerland, and are careful to invoke the fallacy that homeopathy cannot be tested by randomised trial because of the way remedies are made specific to the patient.

They make a big play of assessing the quality of evidence, to make sure that it measures up to their special criteria.

In Section 7.3.5 they then assert the popularity of homeopathy by reference to a vast list of studies on CAM generally, which in some cases do not even include the data collection methodology (whether by survey or whatever); they include as users of CAM and by extension homeopathy, any person who has at any time in their lives used any form of complementary or alternative therapy, including nutritional supplements, vitamins, herbal remedies or massage. Using these absurdly broad criteria they reach the astonishing conclusion that homeopathy and CAM generally are popular.

Bias on bias

Section 5.2 is devoted to "The Risk of Bias in Clinical Trials". It begins: Instead of just formally evaluating the studies, we examined their content for bias. Their definition of bias is then made clear: bias is where a result is not to their liking. This does indeed appear to be a very real risk in any half competently conducted clinical trial…

To evaluate bias the authors use criteria which include: Can the outcome be attributed to the intervention according to CBM (cognition-based medicine) criteria?. Cognition-based medicine is an anthroposophic concept which lacks any substantial support. It is proposed by anthroposophists as an alternative to evidence-based medicine. Put briefly, it elevates the patient’s feeling of being cared for and the clinician’s opinion of the effect of the treatment over more prosaic factors such as objectively measurable physiological changes. In individual single cases it is asserted to make “evident” the effect of a given therapy, but it is "unable to establish a generally valid efficacy". It is extremely unlikely that any mainstream medical investigation would consider this approach valid, let alone relevant.

The authors invoke novel judgements of internal and external validity (which, they note apparently without sensing the irony, "is not always used in the same sense"); they say:

Instead of just formally evaluating the studies, we examined their content for bias. Wherever this was possible on the basis of the data available we searched for bias due to methodical deficiencies (internal validity) or whether the studies were not or only partially transferable to our research because they inadequately reflected the real health care situation.

Shang et. al also assessed content for bias, as did Linde et al. Both concluded that reports in CAM-specific journals were more likely to be biased in favour of homeopathy, and positive studies were more likely to have severe methodological errors.

The authors choose instead to redefine bias as not "reflecting the real health care situation" — which, by definition, means agreeing with their uncritical acceptance of "it worked for me" testimonials, which they brand "real-world" results. Scientific evaluation of homeopathy shows that these results are in fact entirely consistent with the well-documented placebo effect, which is also the null hypothesis in this case. Homeopathy has failed to refute the null hypothesis, so the authors choose instead to redefine the null hypothesis as agreement with their pre-existing supposition.

They then introduce the term model validity, which, as they acknowledge:

[I]s not generally known in medical circles. It describes the conformity between the study setting and an ideal procedure ("state of the art" of the intervention under consideration, e.g. classical homeopathy) with regard to indication, intervention, expertise of the therapist etc.

So, their measure of validity is the extent to which a homeopath of their preferred type would agree with every part of the process. This allows them to exclude almost any result they like, on a whim.

Quantum woo

See the main article on this topic: Quantum woo

Lewis also notes the authors' poor understanding of the science they purport to discuss. In stating:

Modern physics with its theory of relativity and quantum physics has long overtaken Newtonian mechanics and is paving the way for an understanding of the homeopathic mechanism of action.

The authors combine an assertion with a hope. In fact, it is quantum physics, which posits irreducible minima and the Heisenberg uncertainty principle, which is fundamentally incompatible with homeopathy, not classical physics, which allows for a continuous and deterministic universe. The only light that quantum physics sheds on homeopathy is the cold light of day.

Vitalist nonsense

Fundamentally, the authors' statement that:

The vital force, just as vitality in general, cannot be measured and quantified by science, but it exists as a phenomenon.

reveals the shifting sand on which their case is built. Vitalism is considered refuted. No objective evidence has ever been produced to support the existence of the supposed vital force under any of its names, be it "miasma", "qi", "innate intelligence" or any other, nor has any anatomical structure ever been identified that can be shown to have any part in its regulation or flow.

To quote from section 3.7:

Many acute and chronic conditions, especially in primary health care, can be treated homeopathically as long as the patient's regulatory and self-healing powers ("vital force") can still be adequately stimulated.

It is probably just bad luck that so many people who try homeopathy for chronic conditions have apparently passed this mystical point where the vital force can still be adequately stimulated. What a pity that mainstream medicine has not caught onto this powerful healing process; sadly they are hamstrung by having to deal only with things that can be shown to actually exist.

Begging the question

In discussing the history of homeopathy, the authors state:

Despite his successful treatment of acute diseases, Hahnemann often found that an underlying chronic disease continued to gradually progress. While conventional medicine today regards and treats disease as a local or biochemical disorder, Hahnemann looked beyond the local, momentary manifestations to the chronic disease process of the entire organism. The importance of the chronic disease doctrine lies in the observation of chronic disease processes and in establishing the suitable treatment.

In making this statement they fail to point out relevant historical context: Hahnemann repudiated such treatments as purging and bloodletting, and maintained a clean hospital — his success relative to conventional doctors of the day is generally ascribed by historians of science to the simple fact that they actively killed their patients and he did not. This is not a successful treatment in any medically meaningful way.

They also fail to clarify what Hahnemann's chronic disease doctrine actually meant. His claim was that 7/8 of chronic disease was caused by the psora (itch) miasm, and the balance by the syphilis and gonnorrhoea miasms. The authors present this as a finding of a man they represent as a prototypical medical scientist, but it is complete rubbish.

In stating that [f]rom the homeopathic point of view, the disease itself consists in a regulatory disturbance of the organism. It cannot be observed directly, but only via the sum total of its externally perceptible symptoms and signs, they allude to the fact that homeopathy repudiates the germ theory of disease, one of the many ways in which it is provably wrong.

False false negative

The authors focus constantly on the "danger" of false negative results in trials. This is always a risk, but there is extensive documentation to the effect that false positives are far more common, and far more dangerous. Again, the primary motivation for this concern appears to be providing them with a mechanism to wave away robust evidence that shows homeopathy to be a placebo.

Who are these author people anyway

Peter F. Matthiessen and Gudrun Bornhöft list their affiliation as the PanMedion Foundation, Zurich, whose objectives include: recognition and promotion of complementary medicine. According to the website as translated,

Currently [our] focus is on the upcoming referendum on complementary medicine. The Foundation has launched the "Yes to Complementary Medicine" initiative

So the authors are committed financially, emotionally and politically to the promotion of alternatives to medicine. One may wonder why a supposedly neutral Swiss government commissioned them to report on the validity of homeopathy. The answer, of course, is that the Swiss government did no such thing: they asked them to report on the way homeopathy is practised in Switzerland; the authors, knowing that this would be set up against an objective review of whether that practice delivered any measurable clinical benefit, chose instead to write an extended paean to magic sugar pills which was, in the event, unpersuasive.

Naturally the truth is not palatable to homeopaths, so as usual they misrepresent the report, its purpose, its authorship and the outcome.

What are the odds on that, eh?

External links

• Steve Novella