Enigmatic placebo

The double-blind, randomised, placebo controlled trial (RCT) is accepted by the pharmaceutical industry and regulatory agencies as an objective scientific method that produces unbiased results. The validity of the RCT is not simply based on theoretical arguments but rather on years of investigation and observation.

We know, for example, that anticipation of relief can lead to an actual easing of pain following exposure to placebo as the brain responds by generating its own pain-inhibiting opioids. An excellent, free to view, article on ‘placebo effects’ in clinical trials can be found in the BMC Medical Research Methodology journal.

In RCT studies patients are told that they may receive a placebo treatment and that this dummy treatment, which looks like the genuine medicine, contains no active ingredients. It is assumed that any observed clinical or pharmacodynamic effect seen after dosing with placebo reflects the variation in background effect that might be seen in untreated patients. However, a recent article published in PLoS ONE suggests we turn current thinking on its head. The work calls for patients to be provided with information on health changes that might accompany their exposure to placebo, with possible tailoring of information depending on the condition being investigated. The implication is that without this possibility being explained to the patient their consent is not truly informed and this practice shows a lack of respect for patients.

The work discusses the deliberation of every subject taking part in clinical trials – am I on the active drug? It is clear that rightly or wrongly, any perceived knowledge (or opinion) challenges the robustness of subsequent causal inferences. We can’t stop study subjects speculating and the present work fails to offer a solution to the human condition. It is perhaps testament to the rigour of science that no aspect is considered beyond question. However, in this case, it seems, more data are needed before we stop observing and start calling for change.