'Take two placebos and see me in the morning'
The notion of a placebo as an inactive 'dummy' treatment in randomised controlled trials is fading rapidly. A systematic review published in the
Annals of the Rheumatic Diseases included trials of a variety of active treatments (pharmacological and otherwise) for osteoarthritis. The effect size (ES) of the placebo interventions was compared with the ES in the 'no treatment' control group. In this study a treatment ES of 0.2 was described as small, an ES of 0.5 as moderate and an ES of 0.8 as large. In the untreated group the ES for pain was 0.03. The ES for pain was 0.51 in the placebo groups. Placebo achieved similar ES for function (0.49) and stiffness (0.43). ES was smaller for more objective measurements such as walking distance. The ES for placebo in hand OA was high (0.8) and low in hip (0.37). It would appear that placebo is quite a good treatment for OA – especially in the hand – and with none of those nasty side-effects. Intriguingly placebos were often more effective in drug trials. Placebo injections and sham acupuncture were more powerful than placebo tablets. Placebos for new drugs were more powerful than placebos for standard drugs. Expensive placebos were more powerful than cheaper placebos. In an interesting accompanying editorial 1 an attempt is made to define the placebo effect and place this study in context. Other interesting studies with similarly interesting results are quoted – too many to reference here. The placebo effect includes everything else that goes on in the study apart from the active intervention. It is also clearly dependent on what the patient believes about and expects from the intervention and, perhaps most importantly, it depends on the quality of the interaction between doctor or therapist and the patient. It is suggested that our colleagues in complementary medicine understand and exploit these concepts rather well and that we could learn from them with some advantage for our patients. 2