The overall efficacy observed after administration of a drug (in the absence of any other procedures or intercurrent events) results from the combination of three factors: the natural history of the disease, the drug's specific effects due to its pharmacological action, and the placebo effect (a psychological effect that is not yet fully elucidated). The efficacy observed after administering a placebo (a substance containing no active ingredients) can only result from two factors: the natural history of the disease, and the placebo effect.

The aim of a placebo-controlled trial is to estimate the drug's specific effects, i.e. those resulting from its pharmacological properties. Note that in many situations, it is unethical to "treat" the patients in the control group with a placebo, in particular when the condition is serious and a satisfactory treatment exists. In a randomised trial, a random process is used to determine which patients will receive the drug and which patients will receive the placebo. The aim of random assignment is to make the groups as similar as possible and ensure that the only difference between them during the trial is the treatment received. In a double-blind trial, neither the patient nor the investigator can tell the drug and the placebo apart.

A so-called "pure" placebo is generally used in trials, i.e. a substance with no pharmacological effects, such as lactose, bread or normal saline. Sometimes however, to make sure that neither the patient nor the healthcare professional can tell whether the patient is receiving the drug or the placebo, it is useful to choose a placebo that shares certain effects or features with the study drug, e.g. a placebo ointment containing the same greasy excipient as the ointment being tested; a real or sham injection, and so on.

©Prescrire 1 November 2020

Source: "Placebo-controlled clinical trials: the use of “pure”, “impure” and “false” placebos" Prescrire International 2020; 29 (220): 275-278. Subscribers only.

Download the full article