During the last two decades, many doctor’s attitudes towards alternative medicine have become more liberal. The general attitude seems to be: “I don’t care how it works, as long as it helps my patients.” At first glance, this argument seems correct—after all, clinicians have a duty to do all they can to alleviate the suffering of their patients. Once we dissect the argument carefully, it turns out to be ill conceived when debating the value of ineffective interventions.

Essentially, the argument links two fundamentally different issues: the mechanism of action, and the clinical effectiveness of a treatment. To make things clearer, it might be best to discuss them not in the abstract, but in the context of a concrete form of alternative medicine: homeopathy. About 200 clinical trials of homeopathy have so far been published. Their results are, of course, not entirely uniform but the totality of this evidence fails to show that the effects of homeopathy differ from those of placebo.

The principles of homeopathy are scientifically implausible, and even most homeopaths admit they have no clear explanation as to “how this treatment works.” In fact, scientists are keen to point out that a mechanism of action of highly dilute homeopathic remedies is not just unknown but unknowable, there is no way to explain how they work without re-writing several laws of nature.

At the same time, nobody can deny that many patients feel better after having consulted a homeopath. Therefore the argument applies in an almost classic way: “I don’t care how homeopathy works, as long as it helps my patients.” In a way, it sanctions the use of a disproven treatment for the sake of pleasing the patient.

Proponents of homeopathy disagree and argue that, as patients benefit from it, the use of homeopathy is adequate. Vis a vis thousands of patients who are convinced that homeopathy has helped them, the case against homeopathy is often not easy to make but it is nevertheless supported by several logical arguments.

Patients 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 etc. Thus even if thousands of patients of swear by homeopathy, this does not constitute proof for its efficacy. The most logical explanation of the apparently different results from experience and clinical trials is that the empathetic encounter with a homeopath is therapeutic, while the homeopathic medicine is not.

This should encourage clinicians to build good therapeutic relationships with their patients rather than to dish out placebos. Good medicine constitutes maximising the context effects and applying the optimal therapy. Encouraging an attitude whereby the latter element is neglected is not good medicine and arguably unethical. I fear that the argument “I don’t care how this treatments works, as long as it helps my patient,” might do just that.

Edzard Ernst is a professor of complementary medicine at the Peninsula Medical School, Exeter.