TY - JOUR T1 - AutoCPAP devices accurately identify obstructive sleep apnea patients with residual apnea during treatment JF - European Respiratory Journal JO - Eur Respir J VL - 38 IS - Suppl 55 SP - p4959 AU - Lena Hess AU - Tsogyal D. Latshang AU - Malcolm Kohler AU - Christian M. Lo Cascio AU - Robert Thurnheer AU - Konrad E. Bloch Y1 - 2011/09/01 UR - http://erj.ersjournals.com/content/38/Suppl_55/p4959.abstract N2 - Background: Objective data on the effectiveness of CPAP therapy of patients with obstructive sleep apnea syndrome (OSA) are needed. The accuracy of the apnea/hypopnea index (AHI) recorded by CPAP devices during treatment is unknown. We tested the hypothesis that AHI derived from CPAP devices accurately predict an elevated AHI during a sleep study.Methods: 48 OSA patients, mean±SD age 54±6y, baseline AHI 52.5±21.6/h, underwent in-laboratory sleep studies once or twice during long-term autoCPAP therapy. AHI from CPAP devices (ResMed S8, Philips Respironics REMstar) during sleep studies and during the preceeding 7 nights were recorded.Results: The AHI derived from 60 sleep studies was 5.0±6.2/h. The mean difference (bias) of the corresponding AHI from CPAP devices was 1.8/h, 95% confidence interval -7.4 to +11.0/h. The difference between the minimal and maximal CPAP-derived AHI in the week preceeding the sleep study was 6.1±5.1/h. In 22 of 60 sleep studies, the AHI was >5/h. The accuracy of the AHI from CPAP devices to correctly predict an elevated AHI >5/h was quantified by the area under the receiver operating characteristic curve of 0.87 (95% CI 0.78 to 0.96). The sensitivity and specificity of an AHI >5/h from CPAP devices to identify a polygraphic AHI >5/h were 86% and 75%, respectively; the negative and positive predictive values were 93% and 56%, respectively.Conclusions: The studied CPAP devices accurately identify OSA patients with elevated AHI during treatment. Because of the high night-to-night variability mean AHI derived from autoCPAP devices over several nights might be clinically more relevant than a single night AHI measured in the sleep laboratory. ER -