Nocturnal efficiency and tolerance of a demand oxygen delivery system in COPD patients with nocturnal hypoxemia.

OBJECTIVES We compared the efficacy of the standard nasal cannula and the demand oxygen delivery system (DODS) during sleep in patients with COPD. SUBJECTS Twenty patients with moderate or severe COPD were included in the study. METHODS Four consecutive polysomnographic recordings were performed under the following conditions: DODS powered by compressed air (night 1 [N1]); oxygen delivered with a nasal cannula alone (night 2 [N2]); oxygen delivered through a DODS (night 3 [N3]); and oxygen delivered with nasal cannula alone (night 4 [N4]). Oxygen flow rates with and without DODS were adjusted the day before the first night so that the resulting transcutaneous arterial oxygen saturation (SaO2) was > or = 95%. The following parameters were evaluated each night: apnea-hypopnea index, nocturnal SaO2, total oxygen saving, and several neurophysiologic parameters. RESULTS The oxygen saving with the DODS was, on average, 60%. All parameters obtained during N2 and N4 (oxygen alone) were identical. The percentage of total recording time spent at SaO2 > or = 95% was comparable between N2 ([mean +/- SD]; 69+/-32%) and N3 (61+/-31%) (difference is not significant [NS]), as was the time spent at SaO2 between 90% and 95% (N2, 29.8+/-31%; N3, 35.9+/-27%; NS) and < 90% (N2, 0.75+/-2.6%; N3, 2.5+/-8.6%; NS). Although the mean response time was not significantly different between N2 and N3, two patients experienced a substantial increase in response time with an SaO2 < 90% on the DODS. The DODS device did not induce any difference in the percentage of time spent in rapid eye movement (REM) sleep (N2, 12.3+/-8.7%; N3, 16.4+/-7.8%; NS) or non-REM sleep (N2, 87.7+/-8.7%; N3, 83.7+/-7.9%; NS). Non-REM distribution in stage 1-2 sleep and in stage 3-4 sleep was comparable between N2 and N3. Similarly, no difference was observed for the sleep efficiency index (N2, 71+/-15%; N3, 69.6+/-14%; NS). Differences between sleep onset latency times were NS. CONCLUSIONS In a majority of moderate to severe COPD patients, the use of a DODS device does not induce any significant alteration of nocturnal neurophysiologic and ventilatory profiles. However, the presence of nocturnal desaturation in a few patients justifies the need to systematically perform a ventilatory polygraphic recording when prescribing a DODS device.