ICD-detected respiratory disturbance index: accuracy for sleep apnea detection and prognostic value. (25th November 2020)

• Record Type: Journal Article
• Authors: Boriani, G; D'Onofrio, A; Pisano, E; Pieragnoli, P; Locatelli, A; Capucci, A; Talarico, A; Sinagra, G; Rapacciuolo, A; Piacenti, M; Indolfi, C; Checchinato, C; Ricci, R.P
• Is Part Of: European heart journal. Volume 41:(2020)Supplement 2
• Publication Date: 2020-11-25
• Subjects: Ventricular Arrhythmias and SCD - Diagnostic Methods; Cardiology -- Periodicals; Heart -- Diseases -- Periodicals; 616.12005
• DOI: 10.1093/ehjci/ehaa946.0747
• Languages: English
• ISSNs: 0195-668X

Abstract

Abstract: Purpose: In patients affected by heart failure an association exists between sleep apnea (SA) measured by polysomnography and adverse outcome. Impedance-based implantable cardioverter defibrillator (ICD) algorithms have been designed to compute the Respiratory Disturbance Index (RDI) to identify severe SA. The purpose of the DASAP-HF study was to evaluate the accuracy of RDI for the prediction of severe SA, and investigate the prognostic value of device-detected RDI values. Methods: Patients with left ventricular ejection fraction ≤35% implanted with an ICD were enrolled and followed for 24 months. One month after implantation, patients underwent a polysomnographic study (PS) for assessing the apnea-hypopnea index (AHI). The average RDI value was calculated over a 1-week period preceding the sleep study and compared with the assessment of severe SA at PS (AHI ≥30 episodes/h). The endpoint was all-cause death after 24 months. Results: 224 out of 265 enrolled patients had usable RDI values. Patients characteristics: 79% male, 67±10 years, BMI 27±7kg/m 2, ejection fraction 29±5%, 54% ischemic cardiomyopathy, 50% CRT-D. The mean AHI value at PS was 21±15 episodes/h. The mean RDI value recorded during the week preceding PS was 30±16 episodes/h. RDI values accurately identified severe SA diagnosed at PS (AUC 0.77; 95% CI 0.70–0.83; P=0.001). Based on the ROC curve analysis, RDI ≥29 episodes/h and AHI ≥17 episodes/h maximized sensitivity and specificity for the prediction of death. Both indexes were independently associated with all-cause death but, after correction for the other independent significant prognostic variables, RDI≥29episodes/h yielded stronger prediction (HR: 12.22, 95% CI:1.64–91.37, p=0.015) as compared to AHI ≥17episodes/h (HR: 4.14, 95% CI:1.17–14.66, p=0.028). Moreover, severe SA diagnosed at PS (AHI ≥30episodes/h) was not associated with death (HR: 1.20, 95% CI:0.3817–3.8266, p=0.761). Conclusions: In heart failure patients indicated to ICD, severe SA was confirmed to be associated with survival. The ICD-measured RDI accurately identified severe SA detected at PS, and was associated with the risk of death at long-term. Funding Acknowledgement: Type of funding source: Private company. Main funding source(s): Boston Scientific