Implantable cardioverter-defibrillators (ICDs) are generally reliable medical devices that have the potential to add quality years of life for appropriate candidates. Indications for ICDs have emerged from a series of randomized clinical trials, observational data from cohorts of high-risk patients with less common diseases, and expert opinion based on limited data in uncommon disorders. The randomized trials are limited by inadequate stratification designs that resulted from insufficient funding availability. The result was outcomes that led to uneven applications, based in part on post-implant experience of device utilization. In this document, we explore the basis for the features of the evidence available to support ICD use, the role of clinical judgment in circumstances in which data are limited or lacking, and the need for additional research to improve the specificity of indications. Directions for new research initiatives are considered. In addition, a general overview of a clinical research paradigm is presented, in which the research and health care delivery arms of the health care enterprise combine in research design and funding, as the latter bears the impact of the outcomes of the former. Impact estimates during the design of trials, considering reasonable contingencies for outcomes, are suggested as a means of justifying the size, scope, and appropriate costs of studies. If we who are involved in clinical research and health care delivery do not resolve this problem, for both ICDs and other new therapies that appear in the future, society will do it for us.
2009 by the American College of Cardiology Foundation.