Independent design and conduct of clinical trials

Randomized clinical trials (RCTs) have justifiably become the cornerstone for the evaluation of virtually any kind of intervention: drugs, surgical or diagnostic procedures, behavioural modifications, health policies and even systems of care. This commentary focuses on two issues related to independence in clinical trials: first, the independence to ask important questions, and second, independence from bias in the conduct, analysis and reporting of trials, which can lead to misleading conclusions. The most frequent type of RCT are those evaluating therapies, and much of this commentary will be focused on issues relevant to such trials, although the principles apply generally.