Abstract
Sham surgery is a controversial and rarely used component of randomised clinical trials evaluating surgical interventions. The recent use of sham surgery in trials evaluating efficacy of intracerebral fetal tissue grafts in Parkinson's disease has highlighted the ethical concerns associated with sham surgery controls. Macklin, and Dekkers and Boer argue vigorously against use of sham surgery controls. Macklin presents a broad argument against sham surgery controls while Dekkers and Boer present a narrower argument that sham surgery is unnecessary in the specific setting of fetal tissue engraftment for Parkinson's disease. I defend sham surgery controls against both these criticisms. Appropriate clinical trial design, sometimes including sham surgery, is needed to ensure that false positive trial results do not occur and endanger public safety. Results of a completed trial of fetal tissue grafting for Parkinson's disease are used to illustrate the potential benefits of, and problems associated with, sham surgery controls. Sham surgery controls, however, should be employed only when absolutely necessary. I suggest criteria for appropriate use of sham surgery controls.