AHA Scientific Sessions 2026 Abstract · Nov 2, 2026
Chronic Kidney Disease Stage Predicts Long-Term Mortality Following ICD Implantation: Evidence from the All of Us Research Program
By Yerukala Sathipati S, Popat A, Sharma P
Background: Implantable cardioverter-defibrillators (ICDs) reduce arrhythmic death, yet long-term all-cause mortality remains substantial. Chronic kidney disease (CKD) is prevalent among ICD recipients, but its graded relationship with post-implant survival — particularly across racially diverse populations — remains incompletely characterized. Research Questions/Hypothesis: We hypothesized that CKD stage independently predicts long-term all-cause mortality after ICD implantation in a graded fashion, consistently across racial groups. Methods: We identified adult ICD recipients in All of Us using validated OMOP-standard procedure concept codes. Pre-implant comorbidities were ascertained via concept-ancestor hierarchies applied to condition occurrence records, capturing all clinical subtypes. CKD was staged by the most recent pre-implant eGFR (KDIGO 2022: G1 ≥90, G2 60-89, G3a 45-59, G3b 30-44, G4 15-29, G5 <15 mL/min/1.73m ² ). All-cause mortality was the primary outcome. Multivariable Cox proportional hazards regression was performed with a minimum 90-day follow-up. A CKD х race interaction term was tested to evaluate whether the CKD-mortality relationship differed by race. Results: Among 1,662 ICD recipients (37% female, 18.9% Black or African American, 13.8% Hispanic, 56.7% White non-Hispanic), 128 deaths (7.7%) occurred over a median follow-up of 2,414 days (IQR 1,572-3,717). In multivariable analysis, CKD was the strongest independent predictor of mortality (HR 1.81, 95% CI 1.41-2.34, p<0.001), followed by prior myocardial infarction (HR 1.32, 95% CI 1.03-1.68, p=0.027) and age at implant (HR 1.01/year, 95% CI 1.003-1.022, p=0.007). A striking dose-response relationship emerged across CKD stages (log-rank p=3.6х10 ⁻ ²⁰ ): mortality increased from 3.1% in G1 to 6.5% in G2, 15.6% in G3a, 18.8% in G4, and 36.8% in G5/ESRD. In continuous analysis, each 10 mL/min/1.73m ² decline in eGFR conferred a 1.13-fold increase in mortality risk (p<0.001). Black patients showed a numerically higher hazard (HR 1.22) that did not reach significance (p=0.175), and no significant CKD х race interaction was detected (p=0.14), suggesting the CKD-mortality burden is similarly distributed across racial groups. Conclusions: In a racially diverse national ICD cohort, CKD stage independently predicted all-cause mortality in a dose-dependent fashion, with ESRD-range disease associated with nearly 12-fold higher absolute mortality compared to normal renal function. These findings support prospective integration of eGFR staging into ICD candidacy discussions and shared decision-making, particularly in patients with advanced CKD where competing non-arrhythmic mortality may attenuate device benefit.