Mouth Leak is Associated with Sleep Fragmentation During Nasal CPAP Treatment of OSA and May Be Detected by Leak Waveform Analysis.

Abstract

Rationale: Mouth air leak is a major cause of low adherence to nasal CPAP in patients with obstructive sleep apnea (OSA). However, CPAP reports do not distinguish mouth from mask leak. We hypothesized that mouth air leak is terminated abruptly by an arousal from sleep and mouth closing that can be detected by CPAP leak waveform analysis.

Objective: Describe patterns of mouth air leak waveform during polysomnography (PSG) in patients with OSA treated with well-fitted nasal CPAP.

Methods: PSG recordings with a jaw motion sensor to detect mouth opening were performed in OSA patients treated with nasal CPAP with suspected mouth air leak. Careful mask fitting and visual inspection excluded mask leak. Mouth leak episodes were characterized by an increase (≥ 20%) above the intentional leak. Leak episodes were classified as intermittent (< 5 minutes) or continuous (≥ 5 minutes).

Results: Twenty patients (80% men; age: 63±11 years; body mass index: 29.9±6 kg/m2; baseline apnea hypopnea index: 46.9±19 events/h) treated with nasal CPAP and documented mouth leak completed the study. All but one patient experienced an overlap of intermittent and continuous mouth leak. Most mouth leak episodes ended with mouth closure (97.7%) and an arousal (52.7%) or awakening (38.6%). Only 34.9% of the leak episodes were associated with respiratory events. Intermittent mouth air leak was more common in sleep stages N1+ N2 (P<0.01), while continuous leak was more common in sleep stage N3 (P<0.01). Continuous episodes of air leak were associated with a higher amplitude of mouth opening. CPAP report waveform was able to detect only 29.6% of the leak episodes detected by PSG. Only 10 patients (50%) had a high unintentional leak according to the criterion adopted by ResMed and only 2 patients (10%) presented large leak according to the Philips criterion.

Conclusions: Intermittent and continuous mouth leak during nasal CPAP frequently co-exist and contribute to sleep fragmentation. Identification of leak waveform patterns may help detect mouth air leak which in turn is an important cause of poor CPAP adherence.