For more than two decades, I have been in regular contact with practitioners of this or that alternative therapy. Whenever I point out to them that there might be good reasons to doubt the value of their interventions, they tell me: ‘I don’t care how it works; as long as it helps my patients.’ What they mean is clear enough. Every day, they see patients who improve after receiving their treatment and feel it is wrong to doubt their experience. For them, the proof of the pudding is in the eating.
At first glance, this line of argument seems reasonable, logic and correct. It would be foolish, perhaps even unethical, to reject a helpful treatment simply because we fail to understand the exact mechanisms by which it brings about the benefit. This would be neither pragmatic nor kind, and it is not what we generally do in medicine: aspirin, for instance, was widely used and helped many patients long before we understood how it worked. However, once we analyse the way this notion is employed to defend the use of unproven or disproven therapies, we see that it is fallacious.
The argument amalgamates two fundamentally different issues which really should be separated: 1) the mechanism of action of a therapy (‘how it works’) and 2) its clinical effectiveness (‘it helps patients’). The matter gets clearer if we do not discuss it in the abstract, but in relation to a concrete example: Bach flower remedies (BFRs).
Like most homeopathic preparations, BFRs are so dilute that they do not contain any active ingredients to speak of (they differ from homeopathic preparations, however, in that they do not follow the ‘like cures like’ principle). Several clinical trials of BFRs have been published; collectively, their results show that the clinical effects of BFRs do not differ from those of placebo.
The principles backing up BFRs are scientifically implausible, and even BFR practitioners admit that they have no idea how their remedies work. Scientists might add that a mechanism of action of such highly dilute remedies is not just unknown but it is unknowable: there is simply no way to scientifically explain how BFRs work; any rational explanation would necessitate re-writing entire sections of textbooks of physics and chemistry.
The situation is thus quite clear: BFRs are not effective and there is no plausible mechanism of action. Yet many patients feel better after having consulted a BFR-practitioner (or after self-medicating BFRs), and those satisfied customers insist: ‘I don’t care how BFRs work, as long as they help me.’
These people forget, of course, that symptoms can improve for a range of reasons which are related to any specific therapeutic effect: the natural history of the condition, regression towards the mean, placebo effects — to mention just the three most obvious reasons. Only rigorously controlled trials can tell us whether the therapy or other factors caused the observed clinical outcome; our perception alone can often not identify cause and effect correctly, and we thus get easily misled by our own experiences.
The fact that thousands of patients swear by BFRs does not, therefore, constitute proof for their effectiveness. The discrepancy between the impressions generated by our experiences and the results of clinical trials is not difficult to explain: the empathetic encounter with a therapist and/or a placebo effect and/or the natural history of the condition are perceived as symptomatic improvements, while the BFRs have no specific effects at all.
Enthusiasts of BFRs would probably claim that a placebo effect is also a plausible mechanism of action. ‘Surely,’ they might say, ‘this means that BFRs are useful and should be widely employed.’ What they forget is that we do not need a placebo to generate placebo effects.
This sounds paradoxical, but is both true and simple: an effective treatment administered with time, compassion and empathy will, of course, also generate a placebo effect — what is more, it would generate a specific therapeutic effect on top of it. Thus, BFRs — or other placebo therapies — are quite useless in comparison. Seen from this perspective, there is rarely a good justification for using placebos in clinical routine.
My conclusion from all this is straightforward: proponents of alternative medicine often try to justify the use of unproven or disproven treatments by saying: ‘I don’t care how it works, as long as it is helpful.’ This may sound reasonable but, on closer scrutiny, turns out to be fallacious and wrong. Moreover, it is dangerous because it promotes quackery. The use of such therapies in clinical routine cannot be justified and is not in the interest of patients.
Edzard Ernst, emeritus professor at the University of Exeter, is the author of Homeopathy: The Undiluted Facts and the awardee of the John Maddox Prize 2015 for standing up for science. He blogs at edzardernst.com.