mandarine

I used to be very heavily prejudiced against homeopathy and alternative medicines (acupuncture, aromatherapy, reflexology, you name it), for the following reasons:

• the homeopathic system is founded on pure dogma (law of similars); it is invoked repeatedly, without having been put to the test in the first place.
• the theoretical models and the underlying concepts invoked are disconnected from the rest of science.
• the more serious the clinical trials, the less evidence there was in support of an efficacy above placebo effect. It also seemed rather suspicious that homeopathic theory (one doctor, one patient, one treatment) had arguments to refuse randomized double-blind clinical trials as adequate evidence.

Now that I have been married for ten year to someone who feeds me latin-named sugar grains on a regular basis, I will share with you how I managed to come to terms with homeopathy in a rational way, without having to split my personality or sell my soul. I confess it is a rather jesuitic move. Such moves tend to be quite customary to me, and it makes me wonder whether I would be able to keep any sort of principles in troubled times, but this is another matter altogether.

On dogmatic principles

In my article entitled is medicine a science, I have already written about how traditional western medicine is plagued with dogmatic truths which take repeated assaults with serious evidence from clinical observations and trials to overturn. Dogmas hammered into the heads of medical students are quite durable, and probably much more durable in medical practice than in medical research. I will write later about vaccination as one of these persistent dogmas. But even in research, some principles are taken for granted without further critical thinking. A funny example I heard recently on Nature’s podcast (which I heartily recommend) was about Broca’s area. It is a small area in the brain that was first described by Paul Broca in the XIXth century. It was found damaged in two patients with severe speech disabilities and therefore Broca claimed that it was the seat of language. A neurologist, probably intrigued by discrepancies in neuroscience publications as to where that area was really located, went back to Broca’s samples, and determined that the area that was damaged in Broca’s patients was not where recent MRI-scan-based research had put it. And the reason was: after more than a hundred years being taught that way in medical curricula worldwide, Broca’s area had become a magical dogma that was used as a synonym for ’seat of language’. Any research identifying brain activity related to speech tasks in a certain area in MRI scans would immediately tag this area ‘Broca’s area’, when in fact, nobody was quite sure what Broca had actually described in the first place.
This is an illustration with rather benign consequences, but I am certain that contemporary western medicine is strewn with such landmines and pitfalls. This is why constant doubt is so critical to scientific thinking: my personal superstition is that when you stop doubting truths and probing facts, they gradually decay into myths and errors.

Now if classical medicine is so full of dogma from the XIXth century, I can hardly blame homeopathy for embracing dogmatic views inherited from the XVIIIth century, can I?

On dubious models

Firstly, a definition: a model is a formal description of underlying objects and phenomena that allow to explain observations and predict the results of future experiments. The model of plate tectonics invokes a series of layers in the Earth’s depths, with various convection motions that move continental ‘plates’ around. Nobody has ever been down there to check. The validity of the model is all down to: does the model account for observations (seismic soundings, earthquakes, mountain ranges, mid-ocean ridges), and can it help to predict future events?
The model for homeopathy uses notions like the law of similars and concepts like dynamization, potentization or succussion that do not belong in medical theory anymore and are disconnected from science. Another nice illustration of such disconnection is the model for acupuncture, where the word ‘energy’ is used with a meaning that has nothing to do with the agreed convention in the rest of science. Apparently, practicioners more and more use the word ‘qi’ so that there is no such confusion. As they are not built upon other ‘reliable’ models, these models look suspicious, and for all I know, there is no reason to believe that they are anywhere near correct, at least when we understand that no sort of scientific method was used in the process. Conversely, just because a model is based on conventional scientific concepts does not mean it has a single extra ounce of credit.

Does it matter that much that a model invokes esoteric concepts and constructs that have no external justifications? I bet I could describe the whole of plate tectonics by invoking chtonian demons who push continents around, with weaker demons on top and stronger ones below, with a certain law governing the amount of wrath (pressure) versus fury (temperature) as the depth increases, as well as a relationship between fury and the amount of fighting (shear stress) depending on demon personalities (viscosity), etc. By trial-and error, I could reach a poetic (and appropriately frightening) theory of volcanism and earthquakes.

What I mean to say here is that there is no way to probe a model in itself: nobody knows whether electrons or photons are what we think they are, or if they are even ‘real’ (I won’t go into details about what ‘reality’ might mean). All we know is that the concepts of electrons and photons as well as the associated equations are mightily powerful at describing observations and predicting the outcome of experiments or the behaviour of designs.

The fact that a model sounds dubious should not count, although William of Ockham could argue that when a simpler model has the same performance, it should prevail over a complex one (as in Copernicus’ heliocentric model versus the epicycles). Possibly, some alternative practices like homeopathy or acupuncture could be onto something, only theirs models are simply not mature enough and need some pruning; but it does not mean we should throw the baby out with the bath water. As I have already mentioned, the tools offered by evidence-based-medicine now allow to evaluate things without prejudice, even when the underlying theoretical model look weak.

On placebo effect

Unfortunately with homeopathy, the key issue is that there is no hard evidence in favor of any efficacy above placebo effect. As a matter of fact, there is a very efficient argument to justify there can in all probability be no effect other than placebo:
- molecules of active substance in dilutions are so scarce that they are negligible in number in comparison to all other traces. And beyond dilutions of 15CH, they are so scarce that there is very probably not a single molecule of the intended substance in the bottle.
- homeopathy advocates argue that there is a trace effect, and that somehow the substrate has kept an imprint of the active molecules, which continues to confer some efficacy by a yet-unknown phenomenon.
- very well, but then we should take into account all other molecules that are present in the substrate, either traces of whatever can find its way through the manufacturing process (dust from the air, plastic from the container, impurities in the sugar or the alcohol), or imprints from all other molecules that have had a contact with the substrate before. It is obvious that the quantity of ‘alien’ real molecules and ‘alien’ imprints overwhelms whatever imprint was intended in the substrate.

And yet it works.

I mean, lots of people rely on homeopathic treatment and are quite happy with it. In fact, we can reverse the meaning of the key sentence: there is hard evidence that homeopathy is effective, and because we can prove that there is no biochemical action, this effect we call placebo. To my knowledge, there is no reason to give a derogatory meaning to the word ‘placebo’. What is placebo in the first place? Does it have something to do with your homeopath spending twice as much time with you in comparison to the average doctor? Does it have something to do with not fearing side-effects from your treatment? Or does it have something to do with natural selection of the gullible?

Just because it is placebo does not mean people should banish homeopathy: on the contrary. Placebo is much better than doing nothing, and it can do no harm, which is a fabulous advantage over many treatments.

Hahnemann was able to attract followers because he appeared to be a healer compared to those who were cutting veins or using poisonous purgatives to balance humors. More of his patients may have survived and recovered not because he healed them but because he didn’t infect them or kill them by draining out needed blood or weaken them with strong poisons. (from Skepdic)

Instead of despising homeopathy, scientists should focus a lot more attention on it; try to look under the hood and understand more about placebo effect. I am sure that the subject of placebo hides a promised land of discoveries about how the human body and mind are connected; how people, words, care, trust and will can combine into fighting off symptoms or disease. Such research would certainly reach way beyond mere medical matters and treatment of illness, way beyond the narrow question of whether homeopathy works.

Epilogue

As for myself, I prefer to avoid going to a doctor’s altogether (unless it looks serious). I am uncomfortable with a conventional doctor because he is full of self-assured dogma and does not take time to talk with me, and I am uncomfortable with the homeopath, even if he does take time to speak with me, because I know that either there is something wrong with his rational thinking, or he is deliberately fooling me for my own good in order to maximize the placebo effect.

This entry is filed under non-sciences, health, viewpoints. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.