TY - JOUR T1 - Inter-device variation in dual oximetry for polysomnography JF - European Respiratory Journal JO - Eur Respir J VL - 40 IS - Suppl 56 SP - 1663 AU - Patrick Jamieson AU - Paul Burns Y1 - 2012/09/01 UR - http://erj.ersjournals.com/content/40/Suppl_56/1663.abstract N2 - AimsSome sleep laboratories use two SpO2 channels to better detect artefact. This study aimed to examine the agreement between two SpO2 monitors used concurrently on the same polysomnography (PSG).Methods117 PSGs were audited for this study. The PSGs included channels from an integrated SpO2 monitor (Nonin) and an external device from another manufacturer (Masimo). The mean SpO2 and the desaturation index (DI) for both oximeters were compared. Comparisons were made using Bland Altman analysis, with limits of agreement (LOA) of 1.96 standard deviations, and confidence intervals (CI) of 95%.ResultsBland Altman analysis indicated a bias of +1.7% (CI +1.4%, +2.0%), showing a systematic offset towards Masimo. The LOA were -1.2% (CI -1.7%, -0.7%) to 4.5% (CI 4.0%, 5.0%), indicating a variation of ±2.85%.Bland Altman analysis of the DI indicates a -38.5% (CI: -28.0%, -49.1%) smaller DI for the integrated sensor as compared to the external sensor.ConclusionsWhen used to assess mean SpO2 there is a variation of ±2.9%, this may not be clinically acceptable. Additionally the external sensors produced systematically higher readings than integrated sensor. Therefore the outcome of an assessment of hypoxia is device dependant, if only one of these devices is used.There is poor agreement between devices with regards to the desaturation index. The integrated sensor detected 38.5% less desaturations compared with the external sensor. This systematic bias suggests that an assessment of sleep disordered breathing is device dependant. ER -