Cardiac resynchronisation therapy-defibrillators (CRT-D) would seem the obvious choice for patients with heart failure with reduced ejection fraction, given implantable cardioverter defibrillators (ICD) reduce the risk of sudden cardiac death in this population. However, despite decades of discussion, COMPANION (Comparison of Medical Therapy, Pacing, and Defibrillation in Heart Failure) remains the only trial to include both CRT-D and CRT-pacemakers (CRT-P) in which outcomes between the two approaches were similar.1

In clinical trials, ICDs reduce but do not eliminate the risk of sudden death1 as many deaths are due to progressive pump failure resulting in electromechanical dissociation and asystole, rather than ventricular tachyarrhythmias. Even where an ICD could prevent death, the event rate is largely related to the degree of left ventricular remodelling and since ICD trials were conducted in an era where most patients were treated with less optimal neurohumoral blockade and did not receive CRT, it is uncertain whether these results apply to contemporary populations. While conduction delay (QRS ≥130 ms) is associated with less remodelling with medical therapy, CRT frequently improves left ventricular dimensions and haemodynamics with marked reductions in sudden cardiac death as seen in CARE-HF (Cardiac Resynchronization-Heart Failure) in which the risk in the CRT-P arm was halved (HR 0.54, 95% CI 0.35 to 0.84; p=0.005).2 Further considerations are that patients enrolled in ICD trials were younger, less likely to have comorbidities and more likely to have ischaemic cardiomyopathy than patients encountered in clinical practice, while in DANISH (The Danish Study to Assess the Efficacy of ICDs in Patients with Non-ischemic Systolic Heart Failure on Mortality) receipt of an ICD did not affect survival.3

On the other hand, DANISH was underpowered for its primary end-point of all-cause mortality and, despite enrolling a low-risk and well-treated cohort (58% in both groups received CRT), a …