Background: Previous systematic reviews of implantable cardioverter defibrillators (ICDs) used for primary prevention of sudden cardiac death (SCD) concluded that ICDs are less effective in women and the elderly.
Purpose: To examine ICD effectiveness for primary prevention of SCD across subgroups by sex, age, New York Heart Association class, left ventricular ejection fraction, heart failure, left bundle branch block, QRS interval, time since myocardial infarction, blood urea nitrogen level, and diabetes.
Data sources: MEDLINE and the Cochrane Central Register of Controlled Trials through 3 September 2013 with no language restriction.
Study selection: Researchers screened articles for studies comparing ICD versus no ICD for primary prevention.
Data extraction: Data were extracted about study design, patients, interventions, mortality and SCD outcomes, subgroup characteristics, and subgroup effects. Quality of subgroup analyses was determined by consensus. Relative odds ratios comparing subgroup effects were calculated, and random-effects model meta-analyses were conducted on these ratios.
Data synthesis: Meta-analysis of 14 studies showed a decrease in deaths and SCDs due to ICD treatment. Ten studies provided subgroup analyses. Nine studies compared ICD versus no ICD, whereas one compared cardiac resynchronization therapy plus a defibrillator versus no ICD. Within-study interaction tests and across-study meta-analyses yielded weak evidence that did not show differences for all-cause mortality in subgroups by sex, age, and QRS interval. The evidence was indeterminate for other evaluated subgroups because of a paucity of data.
Limitation: Many subgroup analyses were underpowered, which may have resulted in false-negative findings.
Conclusion: Weak evidence fails to show differences for all-cause mortality in subgroups of sex, age, and QRS interval. Evidence is indeterminate for all-cause mortality in the other subgroups and for SCD.
Primary funding source: Agency for Healthcare Research and Quality.