Despite the dearth of evidence that placebos are clinically effective, they have been heralded throughout medicine's history as a means to relieve symptoms and contribute to the well being of patients. Centuries of anecdotal evidence and a general belief in the efficacy of placebos as treatments were given "scientific" status in 1955 with Henry K. Beecher's research on placebos. However, Beecher's study and much of the subsequent research on placebo-controlled trials is limited by the fact that the primary comparison has been between the placebo and the trial therapy not between the placebo and no treatment at all. The problem with such a placebo-therapy design is that it cannot adequately distinguish a placebo effect from the natural fluctuations that often occur in the course of a patient's disease.
In "Is the Placebo Powerless? An Analysis of Clinical Trials Comparing Placebos with No Treatment," Asbjørn Hróbjatsson and Peter Gøtzsche circumvent the limitations of the placebo-therapy design by conducting a systematic review of 130 clinical trials in which approximately 7,500 patients with 40 different clinical conditions were randomly assigned to either placebo or no treatment and evaluated in terms of binary outcomes and continuous outcomes, objective and subjective. With the exception of some small subjective effects on the reduction of pain, the authors report that they found very little evidence of placebos having powerful clinical effects. They conclude that outside clinical trials, there is no justification for the use of placebos.
In a companion editorial to the Hróbjatsson and Gøtzsche article, John Baillor argues that their conclusion may be too broad and hasty . For one thing, some patients did report reductions in their experiences of pain with placebos. Second, there are both statistical and methodological doubts over the quality of some of the clinical trails included in Hróbjatsson and Gøtzsche's study. But Baillor mentions problems of his own concerning placebos—clinical and ethical problems. First, the use of placebos (versus no pill taking) could act as a regular reminder of a patient's illness. Rather than alleviating discomfort, placebos could increase patient discomfort. Second, placebos could mask symptoms and lead patients to not seek "real" treatments if they believe they are being treated. In both cases, the autonomy and well being of patients could be undermined. Finally, placebos involve deception on the part of the physician that could deleteriously affect the physician-patient relationship.
Nevertheless Baillor leaves the door open for non-research placebo use. With the proviso that each and every clinical use of placebo demands justification, he concludes that their contribution to pain relief, particularly, "may merit their continued therapeutic use" .
Questions for Discussion
- If we assume that placebos are sometimes clinically effective, can you think of circumstances in which (outside of clinical trials) their benefits for patients justify physician deception?
- If so, on what grounds would you justify the deception?
- Beecher HK. The powerful placebo. JAMA. 1955;159(17):1602-1606.
- Bailar JC. The powerful placebo and the wizard of oz. N Engl J Med. 2001;344(21):1630-1632.