Indications for implantable cardioverter defibrillator (ICD) implantation are expanding, but many primary and secondary ICD trials have excluded patients with advanced renal insufficiency. We investigated the effect of renal function on the incidence and time to first appropriate ICD shock.

We analyzed data from all new ICD implantations at a tertiary care center from July 2001 to December 2002.

During a mean follow-up time of 445 ± 285 days, 29 (13%) of 230 patients (age 63 ± 14 years, 79% men, 77% white, 75% coronary artery disease, left ventricular ejection fraction 0.28 ± 0.14) received 41 appropriate shocks. Patients were divided into tertiles according to their serum creatinine level. The 1-year incidence of appropriate ICD shock was 3.8%, 10.8%, and 22.7% in the first, second, and third tertiles, respectively (P = .003). Using the same cut off values of serum creatinine, the 1-year incidence of appropriate ICD therapy (shock and antitachycardia pacing) was 8.8%, 20.8%, and 26.3% (P = .02). After correcting for age, sex, race, left ventricular ejection fraction, indication for ICD implantation, and use of β-blockers in a Cox regression model, serum creatinine was still an independent predictor of the time to first appropriate ICD shock (hazard ratio 6.0 for the third compared with the first tertile, P = .001).

Renal insufficiency is a strong predictor of appropriate ICD shocks. Defibrillator therapy should therefore not be withheld based on the presence of this comorbidity. The mechanisms underlying the relationship between renal function and ventricular arrhythmias deserve further investigation.