### Learning objectives Four decades ago, Mirowski et al reported on the first three humans treated with an automatic implantable defibrillator.1 These devices have evolved from high-voltage defibrillation-only ‘shock boxes’ to the modern implantable cardioverter defibrillator (ICD), replete with superior longevity, reliability, footprint, diagnostic accuracy and new therapeutic functions. A parallel clinico-scientific process has sought to identify the patients most likely to benefit from an ICD. This article reviews the historical trial evidence as a route to understanding contemporary ICD decision making. Practical considerations and potential future developments are discussed. Although obviously associated issues, secondary prevention ICDs and cardiac resynchronisation therapy (CRT) are beyond the scope of this article. Primary prevention ICD trials initially focused on patients with left ventricular systolic dysfunction (LVSD) and coronary artery disease (CAD) or prior myocardial

### Learning objectives Four decades ago, Mirowski et al reported on the first three humans treated with an automatic implantable defibrillator.1 These devices have evolved from high-voltage defibrillation-only ‘shock boxes’ to the modern implantable cardioverter defibrillator (ICD), replete with superior longevity, reliability, footprint, diagnostic accuracy and new therapeutic functions. A parallel clinico-scientific process has sought to identify the patients most likely to benefit from an ICD. This article reviews the historical trial evidence as a route to understanding contemporary ICD decision making. Practical considerations and potential future developments are discussed. Although obviously associated issues, secondary prevention ICDs and cardiac resynchronisation therapy (CRT) are beyond the scope of this article. Primary prevention ICD trials initially focused on patients with left ventricular systolic dysfunction (LVSD) and coronary artery disease (CAD) or prior myocardial