• Stephen D Pittman, Giora Pillar, Richard B Berry, Atul Malhotra, Mary M MacDonald, and David P White.
• Division of Sleep Medicine, Brigham and Women's Hospital, 75 Francis Street, Boston, MA 02115-5817, USA. spittman@rics.bwh.harvard.edu
• Sleep Breath. 2006 Sep 1; 10 (3): 123-31.

AbstractThis study aimed to assess the accuracy of a wrist-worn device based on peripheral arterial tonometry (Watch_PAT 100) to detect residual episodes of respiratory disturbance during continuous positive airway pressure (CPAP) therapy. Concurrent polysomnography was used as the reference standard to identify sleep disordered breathing (SDB) events. The study was conducted in three sleep laboratories affiliated with tertiary care academic medical centers. Seventy patients using CPAP to treat obstructive sleep apnea for at least 3 months, following an in-laboratory titration to determine the optimal therapeutic positive airway pressure, participated in this study. Symptoms indicating suboptimal therapy were not required for participation, but self-reported adherence to CPAP therapy was necessary for inclusion. Interventions are not applicable in this study. The accuracy of the PAT-derived respiratory disturbance index (PAT RDI scored by automated algorithm) to detect residual SDB on CPAP was assessed against polysomnography (PSG) using Bland-Altman analysis, receiver-operator characteristic (ROC) curves, and likelihood ratios for increasing (LR+) and decreasing (LR-) the probability of moderate-severe SDB in the study population. Respiratory events on the PSG were quantified using standard criteria for research investigations ("Chicago criteria") to yield a PSG RDI.C. Based on the PSG results, 19% of the participants had moderate-severe SDB (PSG RDI.C>15 events per hour) on their prescribed pressure. For PAT RDI >15 events per hour, the area under the ROC curve was 0.95 (SE 0.03, p < 0.0001, 95% CI 0.89 to 1.00), the LR+ was 8.04 (95% CI 3.64-17.7), and the LR- was 0.17 (95% CI 0.05-0.62). The mean difference between the PAT RDI and PSG RDI.C was three (2SD 14.5) events per hour. Therefore, residual moderate-severe SDB on CPAP was not uncommon in a multicenter population self-reporting adherence to CPAP therapy to treat obstructive sleep apnea. The Watch_PAT device accurately identified participants with moderate-severe SDB while using CPAP in the attended setting of a sleep laboratory.

36 keywords...

hide…