Reliability of respiratory event detection with continuous positive airway pressure in moderate to severe obstructive sleep apnea — comparison of polysomnography with a device-based analysis

Purpose Monitored polysomnography (PSG) is considered the gold standard technique to diagnose obstructive sleep apnea (OSA) and titrate continuous positive airway pressure (CPAP), the accepted primary treatment method. Currently, the American Academy of Sleep Medicine (AASM) considers automatic PAP therapy initiation at home comparable to laboratory titration and recommends telemonitoring-guided interventions. Advanced CPAP devices evaluate and report the residual apnea–hypopnea index (AHI). However, in order to control the effectiveness of the prescribed therapy outside of a PSG setting, the automatic event detection must provide reliable data. Methods A CPAP titration was performed in the sleep laboratory by PSG in patients with OSA. The residual event indices detected by the tested device (prismaLine, Loewenstein Medical Technology) were compared to the manually scored PSG indices. Results of the device (AHI FLOW ) were compared according to the AASM scoring criteria 1A (AHI 1A , hypopneas with a flow signal reduction of ≥ 30% with ≥ 3% oxygen reduction and/or an arousal) and 1B (AHI 1B , hypopneas with a flow signal decrease by ≥ 30% with a ≥ 4% oxygen desaturation). Results In 50 patients with OSA, the mean PSG AHI 1A was 10.5 ± 13.8/h and the PSG AHI 1B was 7.4 ± 12.6/h compared to a mean device AHI Flow of 8.4 ± 10.0/h. The correlation coefficient regarding PSG AHI 1A and AHI Flow was 0.968. The correlation regarding central hypopneas on the other hand was 0.153. There were few central events to be compared in this patient group. Conclusion The device-based analysis showed a high correlation in the determination of residual obstructive AHI under therapy. The recorded residual respiratory event indices in combination with the data about leakage and adherence of the studied device provide reliable information for the implementation and follow-up of CPAP therapy in a typical group of patients with OSA. Trial Registration Number: ClinicalTrials.gov Identifier: NCT04407949, May 29, 2020, retrospectively registered.