Since the publication of the Institute of Medicine's influential To Err Is Human report in 1999, clinicians and policymakers have embarked on an unprecedented quest to improve patient safety and the quality of health care. While some successes have been achieved, the best methods of improving care remain uncertain, and tension exists between those advocating for rapid dissemination of innovative strategies and those calling for evaluation of such strategies via clinical trials before dissemination. In this commentary, the authors outline arguments supporting and opposing each approach. They conclude that quality improvement interventions should be held to the same standards for determining effectiveness as other medical therapies, and use examples of recent patient safety interventions to illustrate the possible unintended consequences of ineffective initiatives.