Periprocedural Risk and Survival Associated With Implantable Cardioverter-Defibrillator Placement in Older Patients With Advanced Heart Failure

Importance: Little is known about the utilization rates and outcomes of implantable cardioverter-defibrillator (ICD) and cardiac resynchronization therapy defibrillator (CRT-D) placement among patients with advanced heart failure (HF). Objective: To examine utilization rates, patient characteristics, and outcomes of ICD and CRT-D placements among patients with advanced HF. Design, Setting, and Participants: This cohort study was a post hoc analysis of 81 492 Medicare fee-for-service beneficiaries enrolled in the National Cardiovascular Data Registry ICD Registry between January 2010 and December 2014. Inclusion criteria were patients who had received an HF diagnosis, had a left ventricular ejection fraction of 35% or lower, and showed evidence of advanced HF, which was defined as New York Heart Association (NYHA) class IV symptoms, inotrope use within the last 60 days, left ventricular assist device in situ, or orthotopic heart transplant listing. The comparator group included patients with NYHA class II and no HF hospitalization within the last 12 months, no left ventricular assist device, no orthotopic heart transplant listing, and no current or recent inotrope use. All eligible patients underwent first-time ICD or CRT-D placement for primary prevention of sudden cardiac death. Data were analyzed from January 2010 to December 2014. Main Outcomes and Measures: All-cause mortality and periprocedural complications. Results: Of 81 492 Medicare patients, 3343 had advanced HF (4.1%) and 19 424 were in the comparator group (23.8%). Among the advanced HF population, the mean (SD) age of patients was 74 (9) years, and patients were predominantly white individuals (81.5%) and men (71.1%). The all-cause mortality rate at 30 days was 3.1% (95% CI, 2.6%-3.8%) in the advanced HF group vs 0.5% (0.4%-0.6%) in the comparator group (P < .001). In the advanced HF population, the aggregate in-hospital periprocedural complication rate was 3.74% (95% CI, 3.12%-4.44%) vs 1.10% (95% CI, 0.95%-1.25%) in the comparator group (P < .001). Most adverse events in this group were in-hospital fatality (1.82%; 95% CI, 1.40%-2.34%) and resuscitated cardiac arrest (1.05%; 95% CI, 0.73%-1.45%). Patients with NYHA class IV (hazard ratio, 1.40; 95% CI, 1.02-1.93; P = .04), ischemic heart disease (hazard ratio, 1.24; 95% CI, 1.04-1.48; P = .02), or diabetes (hazard ratio, 1.17; 95% CI, 1.04-1.33; P = .01) had a higher risk of death. Conclusions and Relevance: Among patients undergoing ICD or CRT-D placement for primary prevention of sudden cardiac death, only a small proportion had advanced HF. These patients experienced clinically important periprocedural complication rates associated with in-hospital death and cardiac arrest relative to patients with nonadvanced HF.