Abstract Introduction Drug-induced sleep endoscopy (DISE) represents the gold standard airway evaluation for sleep surgeons, with the FDA requiring DISE prior to hypoglossal nerve stimulation therapy in the United States. However, its ability to recapitulate sleep breathing pathology is unproven, especially in the United States where target-controlled infusion (TCI) is not available. Methods Consecutive patients who underwent both an in-lab polysomnography (PSG) and DISE with a propofol infusion via the Noxturnal US v7.1.1 platform (NoxMedical) were included. Two reviewers analyzed a baseline period of DISE and the first 10 minutes of consecutive non-REM supine sleep beginning with N2. DISE and PSG were analyzed for types of events (obstructive, central, and mixed), duration of baseline, and mean oxygen saturation. Hypopneas were defined as periods at least ten seconds long with a thirty-percent reduction in flow and a three-percent desaturation; apneas were defined as periods at least ten seconds long with a ninety-percent reduction in flow. For each comparison and proportion of each event type (number of events of event type/total number of events), paired t-tests were performed. Results Patients (n=20) were generally older (54.9 +/- 15.4 years), male (75%), and overweight at the time of their DISE (27.5 +/- 4.3 kg/m2) and their PSG (27.7 +/- 4.4 kg/m2). An average of 3.1 minutes and 10.1 minutes were analyzed during DISE and PSG, respectively. DISE had a significantly higher AHI than PSG (72.9 vs 48.0 events/hour, p=0.003) and significantly lower mean oxygen saturation (85.8% vs 92.8%, p 0.001). However, proportions of each event type were comparable between DISE and PSG – obstructive apneas (0.25 vs 0.30, p=0.51), obstructive hypopneas (0.67 vs 0.58, p=0.26), central apneas (0.02 vs 0.01, p=0.80), central hypopneas (0.01 vs 0.02, p=0.85), and mixed apneas (0.05 vs 0.04, p=0.62). Conclusion Our study findings confirm that propofol, compared to natural sleep, worsens upper airway obstruction, as determined by AHI and mean oxygen saturation. However, surprisingly, proportion of each event type was maintained between settings. These data support the use of DISE, even without TCI, to visualize upper airway obstruction as part of the treatment selection algorithm. Support (if any)
Aggarwal et al. (Fri,) studied this question.