Implementing placebo

Pittrof and Rubinstein make the important point that the “placebo effect” is actually an effect—people often do get better on placebo.¹ However, most, if not all, of the evidence for placebo effects comes from studies where patients expected to have a reasonable chance (generally 0.5) of receiving the active treatment. There is therefore no evidence base for prescribing placebos in a standard clinical setting, which is what the authors seem to advocate.

We do not know whether placebos will have an effect if patients are aware of what they are receiving. The most likely explanation for the placebo effect, particularly in mental disorders, is that it works as a proto-psychotherapy, using the patients’ conviction that they are being helped and mobilising their own positive resources. We know very little about the brain mechanisms of the placebo response, but the available evidence suggests that, like psychotherapy, it partly operates through the same pathways as the relevant active drugs.² I would therefore expect the patient’s belief that some aspect of their brain chemistry is actually being changed to be a crucial part of the placebo effect.

There may still be ways in which patients can benefit from placebos, even if they cannot be prescribed like ordinary drugs. Firstly, in a setting that re-creates the original trial, including randomisation to placebo or active treatment, but then the physician would not have control over who receives the “reduced benefits for much reduced risks.”¹ Secondly, if physicians were to deceive patients, telling them that they are receiving an active treatment when they are not, but this would face both ethical and practical challenges. Thirdly, by administering placebo-like substances in the context of a quasi-medical model such as homoeopathy, but this will work only if the patient (and probably the doctor as well) believes in this model.