By Ian Musgrave, Senior lecturer in Pharmacology at University of Adelaide.
This has been a bad year for homeopathy, first there was the Draft Information Paper on Homeopathy from the NHMRC, which concluded there was no reliable evidence for the use of homeopathy in the treatment of the 61 health conditions looked at.
Then a homeopathic remedy manufacturer left the North American market due to law suites over the ineffectiveness of their products, then the Federal Court has found that Homeopathy Plus! was engaged in misleading conduct over its homeopathic “vaccines”.
Quoting from the ACCC website “…[Homeopathy Plus!] engaged in misleading and deceptive conduct and made false and misleading representations to the effect that there was an adequate foundation in medical science for the statement that homeopathic treatments are a safe and effective alternative to the whooping cough vaccine, when in fact no such foundation exists..”
However, this is not a one-way street. Complementary Medicines Australia has claimed, six months after the public consultation process had closed, that the NHMRC process was flawed. In the August publication Food Navigator Asia it was claimed to be “fatally flawed”.
Before we examine these claims, let me remind you that homeopathy is based on two principles “like cures like” and extreme dilution, in most cases to levels so dilute that there is almost no chance of a single remaining molecule of original compound being present in the remedy. Thus caffeine diluted one in a hundred 30 times is used to treat insomnia and Uranium nitrate diluted one in a hundred 30 times is used to treat diabetes.
In the latter case it is fortunate at no uranium will actually be present, as uranium nitrate causes kidney failure. In uranium nitrate-induced kidney failure some glucose turns up in the urine, as the kidneys ability to reabsorb it is damaged. This is completely unlike what happens in diabetes, where high blood glucose overwhelms the kidneys capacity to reabsorb it (in uranium nitrate toxicity blood glucose is not elevated so it is not “like” diabetes at all). Thus the rationale for homeopathic treatment is flawed at many levels.
But back to the draft report of the NHMRC’s review of homeopathy. This represents the largest and most extensive recent review of homeopathy research. The review looked at both systematic reviews of the use of homeopathy in 61 heath conditions and submissions on behalf of interested parties, which contained a mix of systematic reviews and individual randomised controlled trials. All submissions and papers were carefully evaluated against strict criteria recognised internationally for this type of review. The Australasian Cochrane Centre independently reviewed the overview report to ensure that it was valid and high quality.
To remind you, the review found there was no good evidence that homeopathy was effective for any of the 61 medical conditions considered. In some cases, there was clear evidence that homeopathy was ineffective; in others the evidence base was too weak to give a clear result. These findings are in concert with other large reviews of homeopathy. Let’s look at the claimed “flaws”.
There was no adequate explanation of why randomised controlled trials (RCTs) were excluded.
They were not excluded. The main review focused on systematic reviews, which included randomised controlled trials (and other types of high level evidence). This is the best way to compare multiple studies. Randomised controlled trials are considered the highest level of evidence, but the results of a single randomised controlled trial may be misleading for many reasons.
Chance is one, if a therapy has no actual effect,by chance alone you will find some studies that appear to show an effect.
Thus it is far better to compare as many high quality trials as possible to get a clearer picture. Randomised trials were not excluded, but an integral part of the evidence through systematic reviews. Randomised controlled trials submitted by stakeholders that were not already part of systematic reviews were considered as well.
While there are limitations to this approach (specifically the most recent research may be excluded), it is widely used in making clinical decisions and in no way invalidates the findings of the report. One of the biggest limitations is that negative findings tend to be under-reported, so that systematic reviews tend to overestimate the effectiveness of a therapy. That homeopathy cannot pass muster under these conditions is telling.
Three academics invited to comment on the review all broadly agreed there was no high quality evidence recommending homeopathy for any disorder.
The review excluded too many studies.
Of the 1367 publications considered in the main review, only 60 were finally considered. Not because of anything sinister, but because only those met the review criteria. 374 were duplicate citations, 729 were the wrong study type (not peer-reveiwed, not systematic reviews or metaanalyses, or not looking at controlled trials or high level evidence) or were not looking at the conditions considered in the review or did not report the outcomes (etc. etc.) (see the main review for details).
Of the reports submitted by stakeholders, only a few passed the inclusion criteria or were not already included. Pro tip, if the NHMRC asks you for peer-reviewed systematic reviews and randomised controlled trials in humans, don’t submit books on the life of Hahneman and studies of frogs exposed to thyroxine (yes, I went through the papers).
The review did not consider any publication not in English.
While this excludes some studies, most high quality studies are published in the English language press. As well, the practicalities of translating foreign language papers to ensure there are no complicating errors in translation are avoided. Overall, the impact of this decision on the reliability of the report is marginal at best.
The review did not consider animal studies.
These homeopathic preparations are already in use in humans, so the appropriate studies are ones in humans in the first place. As well, the studies in animals suffer the same flaws as those in humans, too many are of poor quality and many are unable to be interpreted or make claims that cannot be supported. For example, one study submitted to the review that claimed to demonstrate that homeopathic treatments kill breast cancer cells actually shows that the ethanol diluent is the lethal factor.
Overall, while there are some limitations to the study, this is a wide ranging, carefully interpreted study. While overall the broad conclusion is that there is no good evidence the homeopathy being effective in the 61 studied conditions, in at least 13 studies there was good evidence that homeopathy was ineffective (asthma for example). The results of this study are in broad agreement with previous studies of homeopathy (see also this, and before you bring up the “Swiss Report” see here and here).
The NHMRC study conclusion that “…the assessment of the evidence from research in humans does not show that homeopathy is effective for treating the range of health conditions considered” cannot be ignored or dismissed.