I recently found my friends posting the following article about a meta-analysis ‘proving’ that homeopathy doesn’t work. This was followed by comments like ‘I love science’ and ‘surprise, surprise’, and equating a degree in homeopathy with a degree in baloney. But what has this meta-analysis really proven and where is the real baloney?
The conclusions of the meta-analysis were that:
‘“No good-quality, well-designed studies with enough participants for a meaningful result reported either that homeopathy caused greater health improvements than a substance with no effect on the health condition (placebo), or that homeopathy caused health improvements equal to those of another treatment,” the report’s summary states.
Read more at http://www.iflscience.com/health-and-medicine/meta-study-confirms-homeopathy-doesnt-work#JbvdfGJSQxZ17fMD.99
HP versus Pharma: Two Very Different Approaches to Medicine
One key issue here may be what the allopathic scientific community considers to be ‘good quality, well designed studies’. The standard one-size-fits-all methodology used for testing pharmaceutical drugs would inherently be inadequate for a study on any homeopathic drug except for commercial preparations sold for specific ailments, which are known to be the least effective homeopathic treatments.
This is because homeopathy is an entirely different system of medicine, with an approach that is opposite to that of chemical pharmaceutical treatments. Confused? I will try to explain.
In standard clinical trials with pharmaceutical agents, a specific chemical drug is tested in a double-blind, randomized, placebo-controlled study. This is necessary to determine the specific effectiveness of that product for any given ailment, while eliminating bias in the recipients and those administering the drug. It must be tested against a placebo because the simple act of believing you are taking a remedy can have an effect on your physical condition, so that aspect has to be factored in for consideration. The person administering the drug and studying its effects must not know who has taken it and who has taken a placebo, or they may be subject to biased reporting. It is understood that each drug is being studied for a specific desired effect which, if it is effective, it should have on a substantial proportion of recipients. Everyone gets the same drug for the same illness. It’s a one-size-fits-all approach.
Homeopathy is fundamentally different in that treatment takes into consideration not only a specific, targeted ailment, but also the person as a whole: their constitution, their life and medical history, physical traits, personality, temperament, habits and tendencies and anything else that can be used to build as complete a picture of the individual as possible. There are thousands of remedies to choose from, made from everything in nature from minerals, to plant materials, to animal sources. Many remedies can be used to address a specific ailment such as a skin problem or anxiety, but the homeopath chooses the one that is best matched to the person as a whole. For example, a person whose anxiety comes on with dizziness would receive a different remedy than a person whose attacks come on with sweating. A person with a social, sanguine disposition would receive a different remedy than a shy, withdrawn individual.
There are homeopathic remedies known as ‘specifics’, which are used to treat specific ailments such as cocculus for nausea or aconite for fever, but even in the cases where these are used, dosing strength (potency) and frequency are tailored to the individual based on their constitution. A person with a frail constitution would receive a lower potency than a person with a robust one. Some pharmaceutical medications interfere with the effectiveness of HP remedies, and some people do not respond to HP or take longer to respond. Once you understand how this system works, it is easy to see how it would be impossible to judge the value of homeopathy as a whole with a standard one-size-fits all approach for any given homeopathic remedy or administration procedure.
Now let’s look at the new meta-analysis, which you can see here:
The above meta-analysis reviewed a number of meta-analyses and a number of individual conditions for which these had reviewed the effectiveness of HP.
Evidence Shows Homeopathy for Ottitis Media As Effective as Standard Treatment and Better than Placebo- But That’s Still Not Proof Enough
If you look at the data on HP treatment of otitis media (ear infections), you will find that the results were that:
In all studies with comparison to standard treatment with antibiotics, there was found to be no difference in treatment outcomes for pain, duration of illness, and improvement in hearing loss. In other words, HP was as effective as standard treatment. In studies against placebo, evidence was found in favor of homeopathic treatment. Evidence was also found in favor of HP versus standard treatment when it came to a couple of specific outcome measures.
Still, it was concluded that there is not enough evidence to recommend HP treatment. This is not the same as having proven that homeopathy does not work.
Let’s look at one of the contibuting meta-analyses that did not find evidence in favor of HP. Altunc et al (2007) which examined HP treatment of ‘childhood and adolescent ailments’ including ADHD (section 4.2.4 in the document), and concluded that “the evidence from rigorous clinical trials of any type of therapeutic or preventive intervention testing homeopathy for childhood and adolescence ailments is not convincing enough for recommendations in any condition”. (Altunc et al (2007)
What did they actually analyze? They looked at data from 16 studies on nine different ailments and noted that ‘with the exception of ADHD and diarrhoea (three primary studies each), no condition was assessed in more than two double-blind Level II studies.’ In other words, they took together data from studies on nine different conditions, on the majority of which no more than one or two studies had been done, lumped them all together, and concluded that there was not enough convincing evidence that homeopathic practice was effective. It seems to me that, from the outset, the design of this study was bound to fail to produce conclusive results of any kind.
There was one study included each on warts, conjunctivitis, otitis media, post-operative pain-agitation syndrome, two each on asthma, recurrent URTI (upper respiratory tract infections) and adenoid vegetation, and three on asthma and ADHD.
Can you imagine if a meta-analysis examined this number of studies on this number of various conditions, treated with different pharmaceutical agents, and concluded that there is not enough evidence to convince them that pharmaceutical drugs have any effect? It would be laughable.
Let’s look at the data they included on studies with ADHD, which was one of the two condition for which three studies were considered (although I would hardly call an examination of three studies a meta-analysis). These include Frei et al, 2005, Freitas et al, 1995, and Jacobs et al, 2005. Two out of three of these studies showed intergroup differences in favor of the effectiveness of HP over placebo.
Jacobs et al, 2005: This study found no intergroup differences and included ‘43 children with conﬁrmed ADHD diagnosis (computerised Diagnostic Interview Schedule for Children) with mean age of 9 years. 9 participants were already taking
stimulant medication but still displaying symptoms (n=5 active, n=4 placebo)’. Description is from Cochrane Review.
Stimulants are well-known to interfere with the action of homeopathic treatment for ADHD. Even if the child is not presently on stimulants, having previously been treated with them can affect how well they will respond and how long it will take to get a response. Including some kids who were on Ritalin during the trial would be a bit like including subjects who are on Suboxone in a trial on opiate painkillers, the effects of which this medication is known to nullify.
Strauss et all, 2000, involved: ’20 children with previously diagnosed ADHD (no conﬁrmation) aged between 7-10 years. Half of the participants (n=10) were already taking Ritalin.18 boys, 2 girls.’ Again, half the children were taking a drug known to interfere with the effect of homeopathic ADHD treatment. Moreover, the HP remedy tested was a non-individualized, low-potency formula preparation sold OTC in pharmacies to ‘improve concentration, memory, and alertness’ and, which was given for only 18 weeks. Nonetheless, this study found intergroup difference in favor of the effectiveness of homeopathic treatment for ADHD.
While we’re on the subject of studies being set-up to fail to show accurate treatment results, Frei et al, 2005, the third ADHD study in the Altunc et al review, was discussed in this related study, published in 2007. This was a retrospective analysis of the effect of screening for responsiveness to HP ADHD treatment before randomization. After screening,
‘They then entered the parallel group, radomized, double-blind, placebo-controlled cross-over trial. The double-blind part of the study consisted of two groups of children who received either verum for 6 weeks followed by placebo for 6 weeks, or placebo for 6 weeks followed by verum for 6 weeks.’ (Frei et al, 2007)
The study showed that responders had a 38.5% drop in Connor’s Global Index scores for ADHD, while non-responders had only an 11% drop, and concluded that:
‘Because of the necessity of identifying an optimal medication before response to treatment can be expected, randomisation at the start of treatment in an RCT of homeopathy in ADHD children has a high risk of failure to demonstrate a speciﬁc treatment effect, if the observation time is shorter than 12 months.’ (Frei et al, 2007)
This study is very interesting and shows how standard study design may be a set-up for failure to show accurate results in a RCT. They also noted that subjects who had been pre-treated with stimulants took longer to respond.
Frei et al, 2005: CGI scores decreased from a median of 19 before treatment, to 8 post-crossover (a 58% drop!).
‘During crossover trial CGI parent–ratings of a child was significantly lower under verum (average 1.67 points)
than under placebo (p=0.0479). Long-term CGI improvement reached 12 points (63%, p < 0.0001).
Interpretation: The trial suggests scientific evidence of the effectiveness of homeopathy in ADHD-treatment, particularly in the areas of behavioural and cognitive functions’
It’s possible that the relatively slight difference during the cross-over phase was due to the persistence of the action of the remedies taken in previous weeks. Unlike pharmaceutical drugs, homeopathic remedies seek to address underlying issues and tend to produce more lasting, even permanent results over time.
These meta-analyses make the mistake of lumping together one, two or three studies each on a wide variety of conditions, or of examining a several studies on one condition but with widely different randomization and study approaches, remedy potencies, and other variables like treatment with contra-indicated pharmaceutical drugs, and they conclude that there is not enough evidence that this system of medicine as a whole works. Is that any surprise?
Meanwhile, several pharmaceutical drugs that have been rigorously tested and shown on meta-analysis to be ineffective and even dangerous are regularly recommended by the medical industry as ‘safe and effective’. These notably include statins, flu shots, and anti-depressants.
As far as I can see, all these meta-analyses have proven that the pharmaceutical industry maintains an iron grip on the medical industry and continues to undermine our health freedom. I think we have found where the real baloney is.