Evidence for homeopathy?

FIH: The Lancet published a systematic review of homeopathy in 2005.  The authors concluded that homeopathy doesn’t work – a view endorsed by The Lancet’s editorial team.  Now the European Committee for Homeopathy have issued a counter-statement entitled 'New Evidence for Homeopathy' and claiming that the study was flawed.  Who is right, and what is the strength of the evidence either way?

GL: As always with disputes about the interpretation of evidence the answer is that both The Lancet and the European Committee for Homeopathy are right.  The reality is that there are about 200 clinical trials in homeopathy, comparing homeopathic remedies with placebo.  The majority of them are not really generalisable to clinical practice, in that they don’t look at the process of classical homeopathy as it is practised within a primary care environment. 

They involve artificial constructs where homeopathy is almost never used, for instance in the treatment of post operative abdominal problems, or looking at use of homeopathic allergens in the treatment of allergic illness;  this certainly isn’t classical homeopathy. 

The second issue is the conclusions you can draw from these 200 trials depend very much which ‘filters’ you use to look include or exclude evidence.  One extra trial here or there can radically change the conclusions of a systematic review and I think the point that the homeopaths are making is that we don’t really have enough evidence to come to a definitive conclusion. 

Interestingly The Lancet seems to have come to a definitive conclusion on what I would suggest is a fairly fragile evidence base; six studies of conventional medicine and eight studies of homeopathic medicine that were not really directly comparable.

The real issue here is that homeopathy seems implausible; of course the clinical effects may be much more to do with the process of the consultation than the actual prescription of a particular remedy without any actual molecules in it.  Homeopathy is a whole system of medicine which implies lots of untested assumptions; one of them is that very dilute medicines are clinically effective; the other is that the consultation is both special and very effective.  The evidence we have suggests the remedies might just work better than placebo, but we are not sure.  The difficulty for many conventional scientists is the implausibility of the whole process, but then much science, from the Enlightenment onwards, has appeared to be implausible at first glance and later proved, with adequate research, to offer us surprising new insights.

FIH:  There's a continuum in 'unproven' statements between 'unproven, but perfectly likely' (I have three tomatoes in my fridge) and 'unproven, and very unlikely' (I have a giraffe in my fridge). Some would argue that explanations of homeopathy so outrage the laws of chemistry and physics that they fall into the 'giraffe in fridge' category and are therefore not worth testing. What would you say to people who take this view?

GL: I don’t think there’s either a giraffe or an elephant in my fridge!  I do think, however, that there are some interesting observations, both about the effect of homeopathic remedies on the structure of water and indeed the clinical effects of homeopathy, in certain conditions (fibromyalgia and acute flu-like illness). 

Rigorous science requires thoughtful investment and large amounts of data.  While it is self evident that there isn’t a giraffe in the fridge, I don’t think conclusions about homeopathy are self-evident. The lack of evidence means that people take their own personal opinions as a starting point and interpret the material accordingly.

FIH:  What kinds of studies would you like to see to help reach stronger conclusions? - and what kinds of evidence might move you to a stage where you felt reasonably confident that homeopathy either does or doesn't work?

GL: The reality is that we need more pragmatic and economic studies that look at comparative safety and comparative effects. 

The whole debate around complementary medicine is a patient driven debate.  For the first time conventional physicians have had to respond to patient demand and patient need on a world wide basis.  The kinds of questions that patients have are:

•              Could homeopathy help?
•              What will it cost me?
•              Will it be dangerous?
•              Can I use it instead of using a conventional medicine?
•              Can I use it with a conventional medicine?

You can’t use something as a placebo until you know what the active ingredient is.  At the moment we are not sure whether the active ingredient in homeopathy is the medicine or the particular consultation style that homeopaths employ.  What we need right now are pragmatic Health Service based studies that answer the sort of practical questions that patients, and those funding the provision of healthcare, really want to know.

FIH: Homeopathy may be effective for some illnesses but not others. Is it possible to have a theory about where it might be most effective?

GL: Without really understanding why homeopathy may be working (I know that is very confusing, but it is actually the same for a lot of conventional treatment) it is difficult for us to develop a theory about what it might be best, and indeed worst, for.  If it is all down to the consultation and the processes involved, then we really need to find out what the ‘active ingredient’ might be that seems to impress so many intelligent people.  When we have some good data about that, probably through some careful health psychologist investigations, we can then develop a plausible approach and a sensible research strategy.  This will provide a reasonable assessment of where homeopathy may be useful, or indeed useless.

FIH: We've encountered some healthcare practitioners who believe that homeopathy has no physical mechanism, but nevertheless use it on their patients, believing that the process of questions and answers is psychologically helpful. The patient may therefore have a very different perception from the practitioner as to the kind of treatment they are receiving. Is this a fair way to operate?

GL: We know that many doctors prescribe placebo.  A recent US survey of rheumatologists suggested that many were sending patients to buy over the counter anti-inflammatories and painkillers for the placebo effect.  (Still that is rather ‘interesting’ way to describe the concept of placebo and certainly not one with which many healthcare regulators would agree.) 

All treatments have psychological mechanisms; we know that in primary care the way you approach the patient may make a huge difference to the symptoms someone suffers with a sore throat or chest infection.  Psychological mechanisms involved in treating people are a vital component of the therapeutic relationship and remain an essential and very poorly understood element of all medical treatments.

FIH: Do you think there's an argument for homeopathy on the NHS given the current uncertainty?

GL: Interestingly, I am just reviewing the evidence for the various conventional treatments for endometriosis.  Were I to be absolutely fair about it, I would say that no interventions should be made available through the NHS if there is confusion and difficulty surrounding their evidence.  Leading authorities in the area of endometriosis have described our current medical and surgical treatments of this condition, which affects one in ten women in their reproductive years, as something akin to ‘therapeutic anarchy’. 

I think, providing we know what we are dealing with, and providing we ‘do no harm’ then homeopathy would fall into the very substantial gamut of unproven treatments that are currently provided at enormous cost by the National Health Service in the United Kingdom.  At least we know it is safe even if the medicine itself may be ineffective!