To assess the prognostic significance of screening for sleep-disordered breathing in patients with implantable cardioverter-defibrillator (ICD) with regard to appropriate ICD therapy and total mortality.Methods and results
Overnight sleep studies were performed in 204 ICD recipients not known to have sleep apnoea and with no history of daytime sleepiness. Sleep-disordered breathing was diagnosed in the presence of an apnoea–hypopnea index of five or more events per hour. Seventy patients (34%) had no sleep apnoea, 105 patients (51%) had central sleep apnoea, and 29 patients (14%) had obstructive sleep apnoea. During 38 ± 26 months follow-up, 80 patients (39%) received appropriate ICD therapy for ventricular tachycardia (VT) or ventricular fibrillation (VF), and 54 patients (26%) died. On multivariate Cox regression analysis, age, left ventricular (LV) end-diastolic diameter, secondary prevention ICD indication, use of diuretics, and absence of aldosterone antagonist therapy but not sleep apnoea were associated with appropriate ICD therapy for VT or VF. In addition, multivariate Cox analysis identified age and LV ejection fraction but not sleep apnoea as predictors of total mortality.Conclusion
Undiagnosed sleep-disordered breathing is common in ICD recipients. The presence and severity of previously unknown sleep apnoea in ICD recipients, however, does not appear to be an independent predictor of appropriate ICD therapy or morality during follow-up.