0600 Automated Oximetric Versus Standard Sleep Polygraphy Scoring

Abstract

Abstract Introduction Agreement between automated standard respiratory event scoring and a novel, validated, and patented oximetry-based algorithm was assessed. Methods The standard apnea-hypopnea index (AHI) was derived by adding apneas (flow drop >= 90% for 10 to 30 s) and hypopneas (flow drop >= 30% for 10 to 60 s with oxygen saturation (SpO2) dropping >=3%). The novel oxygen index (ODI4) was derived by scoring events where SpO2 dropped in each of three successive samples and cumulatively by >= 4%. Agreement was assessed by Bland-Altman analysis Results AHI versus ODI4 and Bland-Altman plots showed a high prevalence of AHI > ODI4 when AHI was< 30/h. Negative difference outliers were frequent when mean index difference was > 30/h. There was a bias of 2.83/h in the difference between AHI and ODI4 with upper and lower confidence limits of 22.0/h and -16.3/h. Conclusion Standard respiratory event scoring overestimates respiratory disturbance compared to a novel oximetric index. Standard automated scoring frequently over scores events when basal flow amplitude is low. Support None.