Control of OSA during automatic positive airway pressure titration in a clinical case series: predictors and accuracy of device download data.

STUDY
OBJECTIVES: To investigate the factors associated with physiologic control of obstructive sleep apnea (OSA) during automatic positive airway pressure (APAP) titration in a clinical series. To also assess the usefulness of apnea-hypopnea index (AHI) data downloaded from the APAP device (Dev AHI).
DESIGN: Retrospective review of a consecutive series of patients with OSA who underwent APAP titration (Autoset Spirit, ResMed, Bella Vista, New South Wales, Australia ) with simultaneous polysomnographic (PSG) monitoring in the sleep laboratory.
SETTING: Tertiary sleep clinic. PARTICIPANTS: There were 190 consecutive patients with OSA referred for APAP titration. MEASUREMENTS AND
RESULTS: There were 58% of patients who achieved optimal or good control of OSA (titration PSG AHI < 10, or at least 50% reduction in AHI if diagnostic AHI < 15/hr) during APAP titration. The independent predictors of titration PSG AHI were a history of cardiac disease and elevated central apnea and arousal indices during the diagnostic study. Although the median and interquartile range (IQR) AHI from the device (7.0, 3.9-11.6 events/hr) was only slightly less than the PSG AHI (7.8, 3.9-14.4 events/hr, P = 0.04) during titration, case-by-case agreement between the two measures was poor (chi-square < 0.001).
CONCLUSION: In a clinical sample control of OSA during APAP titration is often poor, and close clinical follow-up is particularly needed in patients with a history of cardiac disease or with high arousal or central apnea indices on the diagnostic study. Device AHI does not reliably assess control during APAP titration, and PSG assessment may be required if clinical response to treatment is poor. The findings relate to the ResMed AutoSet device and may not apply to other devices.