Repeatability of sites of sleep-induced upper airway obstruction. A 2-night study based on recordings of airway pressure and flow.

OBJECTIVES

To evaluate repeatability of overnight continuous airway pressure and flow recordings for assessment of obstructive sites and their distribution in patients with obstructive sleep apnea (OSA).

STUDY DESIGN

Open, prospective study of 30 men with sleep-induced obstructive events (apnea and hypopnea).

METHODS

Repeated ambulatory sleep recordings of airway pressure and flow to assess the localization of sites of obstructive events were compared. Obstructive sites were divided into upper (transpalatal) or lower (subpalatal). The exact distribution of obstructive sites was expressed as percentage upper obstructive events. A cut-off at apnea index (AI) = 5 was used for comparison between frequent and infrequent obstructers.

RESULTS

Eighty-two percent (14/17) of patients with frequent, and 58% of those with infrequent apneic events maintained their predominant site of obstructive events classified as upper (transpalatal) or lower (subpalatal) [72% (21/29) in the entire group]. We found a significant correlation between recordings for the percentage upper apneic (R = 0.54, P = 0.024, n = 17) and hypopneic (R = 0.59, P = 0.012, n = 17) events in patients with frequent, but not in those with infrequent apneic events. Patients who had the vast majority of obstructive events located at a single site were more stable than those with more evenly distributed obstructive events.

CONCLUSION

Repeatability of sites of obstructive events is influenced by the severity of illness and the degree of upper or lower obstructive predominance. The distribution of sites of obstructive events (classified as mainly "upper"/"lower" or as percent upper obstructive events of all) can be identified with relative confidence in patients who have frequent apneic events (AI > or = 5) or a high degree of upper or lower obstructive predominance and especially in those who have a combination of these two criteria.