P113 A detailed analysis of multicentric sleep staging inter-rater variabilities

Abstract Introduction Correct identification of sleep stages is important in disorder diagnosis; but subject to high inter-scorer variability. Thus, we aimed systematically analyse staging agreements between ten scorers a multicentric environment better understand areas that possess scoring variations. Method Polysomnographic recordings 50 individuals (mean (±SD) AHI 12.0±13.2) were independently scored by technologists from seven centres ( Europe six, Australia one). The epoch-based majority stage was identified all scorings; then the mean scorer agreement with this score calculated. We also investigated how sleep-state transitions affect agreement. Results In total, 48556 epochs analysed. 86.3±2.9% (Randolph’s kappa κ=0.74) overall; and 86.3±9.5% (κ=0.74), 65.6±13.5% (κ=0.39), 86.1±6.2% 88.8±11.2% (κ=0.78), 91.7±4.5% (κ=0.83) for Wake, N1, N2, N3, REM respectively. Across scorers, 20.0% unanimously as most frequent disagreement N2 N3 (19.7% epochs). Only 0.3% N1. There negative association transition frequency (r=-0.61). Discussion High obtained except Surprisingly, considerable disparities observed among distinguishing N3; may reflect both discrepancy associated N2/N3 potentially different perceptions. Greater discrepancies more transitions.