Objective The recent Danish Study to Assess the Efficacy of ICDs in Patients with Non-ischemic Systolic Heart Failure on Mortality (DANISH) trial suggested that implantable cardioverter defibrillators (ICDs) do not reduce overall mortality in patients with non-ischaemic cardiomyopathy (NICM), despite reducing sudden cardiac death. We performed an updated meta-analysis to examine the impact of ICD therapy on mortality in NICM patients.

Methods A systematic search for studies that examined the effect of ICDs on outcomes in NICM was performed. Our analysis compared patients randomised to an ICD with those randomised to no ICD, and examined the endpoint of overall mortality.

Results Six primary prevention trials and two secondary prevention trials were identified that met the pre-specified search criteria. Using a fixed-effects model, analysis of primary prevention trials revealed a reduction in overall mortality with ICD therapy (RR 0.76, 95% CI 0.65 to 0.91).

Conclusions Although our updated meta-analysis demonstrates a survival benefit of ICD therapy, the effect is substantively weakened by the inclusion of the DANISH trial—which is both the largest and most recent of the analysed trials—indicating that the residual pooled benefit of ICDs may reflect the risk of sudden death in older trials which included patients treated sub-optimally by contemporary standards. As such, these data must be interpreted cautiously. The results of the DANISH trial emphasise that there is no ‘one size fits all’ indication for primary prevention ICDs in NICM patients, and clinicians must consider age and comorbidity on an individual basis when determining whether a defibrillator is appropriate.