RT Journal Article
SR Electronic
T1 Validation of a Cheyne-Stokes respiration (CSR) detection algorithm for a CPAP device
JF European Respiratory Journal
JO Eur Respir J
FD European Respiratory Society
SP P2007
VO 44
IS Suppl 58
A1 Dinesh Ramanan
A1 Peter Bateman
A1 Holger Woehrle
A1 Glenn Richards
A1 Jeff Armitstead
YR 2014
UL http://erj.ersjournals.com/content/44/Suppl_58/P2007.abstract
AB Introduction:CSR is a periodic breathing pattern characterized by a waxing-and-waning of ventilation, primarily during sleep. CSR is very common in heart failure (HF) and can also occur during continuous positive airway pressure (CPAP) therapy initiation for obstructive sleep apnoea (OSA) or during ongoing therapy. Persistent CSR on CPAP therapy can be associated with persisting symptoms as well as a worse prognosis in patients with HF. CSR pattern recognition has been established analyzing a qualitative nasal flow signal using a diagnostic device (Weinreich,G et al. Sleep. 2009;32:553-7). We investigated if CSR detection could also be used while monitoring CPAP therapy through the airflow measured by the PAP device.Objective:To assess the accuracy of an algorithm for CSR detection, running on a CPAP device to provide a clinically relevant measure of CSR.Methods:Phase 1: Algorithm efficacy was determined by processing 69 diagnostic nasal flow records (27 OSA, 20 CSR, 11 CSR+OSA & 11 normals) using software simulation. Phase 2: 25 nasal flow records (selected from 150 serially recorded auto-cpap studies) with a Apnea-Hypopnea index ≥ 5. These records were replayed by a lung simulator to a PAP device running the new algorithm. In both phases, results were compared to scoring by an expert blinded to the algorithm results.Results:Phase 1 achieved sensitivity, specificity, accuracy (with 95% CI) of 80%(63%,90%), 90%(76%,96%), 86%(73%90%) for detecting any CSR period. Phase 2 achieved 100%(51%,100%), 81%(60%,93%), 84%(70%,98%).Conclusion:Our study shows that a novel algorithm for detecting CSR breathing on a PAP device is highly accurate and can be a useful tool for monitoring therapy.