To the Editor We read with great interest the meta-analysis by Al-Khatib et al1 in which they concluded that prophylactic implantable cardioverter defibrillators (ICDs) reduce all-cause mortality among patients with nonischemic cardiomyopathy (NICM), based on data pooled from 4 trials. These authors ascribed the greatest weight (40%) to the Sudden Cardiac Death in Heart Failure Trial (SCD-HeFT) for the calculation of the pooled risk ratio (RR). Furthermore, sensitivity analyses that removed 1 study suggested that their positive result favoring ICDs was driven by data from SCD-HeFT only; the removal of SCD-HeFT rendered the pooled RR nonsignificant (RR, 0.87; 95% CI, 0.72-1.05; P = .14), but the removal of any of the 3 other trials did not change the summary results.2 However, the mortality rate for the NICM population in SCD-HeFT was calculated using a subgroup population; such analysis is observational in nature and hypothesis generating.2