0848 The Utility of Pulse Wave Amplitude to Improve Interscorer Reliability

Abstract Introduction While accurate scoring of arousals is crucial for identification of disrupted sleep architecture and hypopnea scoring, achieving acceptable interscorer reliability represents a major challenge. The drop in pulse wave amplitude (PWA) signal is a sensitive marker for arousals and is readily available in most conventional polysomnogram software. The aim of the current study was to examine interscorer agreement in scoring cortical arousals using EEG alone versus utilizing PWA drop signal as surrogate marker. Methods Arousals were scored using the same data on duplicate studies by the sleep laboratory medical director and education coordinator who is registered polysomnographic technologist (RPSGT). The first study was scored according to the American Academy of Sleep Medicine (AASM) arousal rule only. The second study used the drop in PWA as a marker for a possible cortical arousal in conjunction with the AASM arousal rule. A drop in PWA signal of at least 30% that lasted for 3 seconds was needed to identify possible arousals. Interscorer agreement and Cohen’s Kappa were calculated as measures of reliability. Results When scored using conventional EEG arousal criteria alone, there was an overall 90.3% agreement for all epochs with a corresponding Cohen’s Kappa of 0.642 (95% CI: 0.580-0.704), representing substantial agreement. When PWA assistance was utilized, agreement was significantly improved to 96.5% overall with a corresponding Cohen’s Kappa of 0.878 (95% CI: 0.840-0.917), representing almost perfect agreement. Conclusion Pulse wave amplitude may hold promise as a surrogate tool for identifying cortical arousals and improving interscorer reliability. Support (if any)