Potential Benefit of Implantable Cardioverter-Defibrillator Placement During or Early After Acute Hospitalizations in Older Adults with Heart Failure

Background: Patients with heart failure (HF) are at high risk for sudden cardiac death (SCD). Landmark trials demonstrated the efficacy of primary implantable cardioverter-defibrillator (ICD) to prevent SCD and improve overall survival in patients with chronic stable HF. However, hospitalization selects a different population of patients for whom the benefit of preventing SCD may vary. The trajectories of mortality following ICD placement during or early after hospitalization for cardiac or non-cardiac causes and their comparative risks have not been assessed.

Objectives: To compare trajectories of mortality after primary ICD during or within 90 days after acute hospitalization, for either cardiac or non-cardiac diagnoses.

Methods: We identified HF patients > 65 years who received primary ICD, using 20% random sample Medicare data (2008-2018). Implant settings were categorized as: 1) current hospitalization (‘current-H’), 2) ‘recent-H’ within 90 days of hospitalization, or 3) ‘chronic stable HF’. We further stratified by cardiac or non-cardiac hospitalization. Mortality rates were estimated and hazard ratios were calculated using Cox regression for each setting and hospitalization cause at 0-30 days, 30-90 days, 90-365 days, 1-3 years, and 3-5 years following ICD placement.

Results: Of 61,710 beneficiaries receiving ICD (mean age 76, 65% male and 85% White), 19%, 25%, and 56% were implanted in ‘current-H’, ‘recent-H’, and ‘chronic stable HF’ settings. Mortality at 1 year was 21.5% in ‘current-H’ and 15.7% after ‘recent-H’ compared to 6.6% for ‘chronic stable HF’, with highest excess mortality by 30 days after implant (38.2 for ‘current-H’, 21.6 for ‘recent-H’, and 4.4 for ‘chronic stable HF’ per 100 person-years). Compared to ‘chronic stable HF’, adjusted mortality hazard ratios were greatest in time intervals closest in proximity to hospitalization / ICD placement and began to plateau after 90-days survival (0-30 days: ‘current-H’ = 5.52 [4.45 – 6.84], ‘recent-H’ = 3.43 [2.76 – 4.25]; 30-90 days: ‘current-H’ = 3.49 [3.02 – 4.03], ‘recent-H’ = 2.18 [1.89 – 2.52]; 90-365 days: ‘current-H’ = 1.98 [1.83 – 2.13], ‘recent-H’ = 1.61 [1.50 – 1.73]). Early excess mortality was similar for cardiac and non-cardiac causes of hospitalization.

Conclusions: ICD placement during or soon after hospitalization was associated with high mortality, regardless of hospitalization reason, with quickly diminishing risk after 90 days. Our results support a more conservative strategy for primary prevention of SCD in patients with current or recent hospitalization, such as wearable defibrillator and/or deferral of ICD implantation until 90 days following discharge, to approach benefit seen for chronic stable HF.