To assess the influence of initial mask choice on the clinical effectiveness and tolerance of noninvasive positive pressure ventilation (NIPPV) in the management of hypercapnic acute respiratory failure.
A prospective randomized controlled clinical study.
A medical intensive care unit at a university hospital.
Randomization between two NIPPV interfaces.
Initial mask choice was randomized between two standard masks: face (NIPPVf group) and nasal (NIPPVn group). The main end point was mask failure (i.e., mask change and/or intubation). Secondary end points were tolerance of NIPPV, change in respiratory parameters during the first 3 days, and patient outcome. Results were analyzed on an intent to treat basis. A per protocol analysis was also conducted.
Ninety patients with underlying chronic lung disease were included, 46 in the NIPPVf group and 44 in the NIPPVn group. The overall success rate of NIPPV was 83%. Mask failure occurred significantly more often in the NIPPVn group (32/44 vs. 9/46; p < 0.0001), mainly because of the need for mask change (32/44 vs. 0/46; p < 0.0001) because of the occurrence of major buccal air-leaks in 94% of cases. Improvement in respiratory parameters was similar in the two groups. Whereas air-leaks were more frequent in the NIPPVn group (p < 0.05), respiratory comfort was assessed as lower and complications more frequent by the staff in the NIPPVf group from day 2 (p < 0.05).
A face mask should be the first-line strategy in the initial management of hypercapnic acute respiratory failure with NIPPV. However, if NIPPV has to be prolonged, switching to a nasal mask may improve comfort by reducing face mask complications.
From the Department of Medical Intensive Care (CG, AB, FD, FT, GB), Rouen University Hospital, Rouen, France; GRHV Research Group (CG, GB), UPRES EA 3830-IFRMP.23, Institute for Biomedical Research, Rouen University, Rouen, France; Department of Biostatistics (JB, MFH), Rouen University Hospital, Rouen, France; and Inserm U 657 (JB), Bordeaux, France.
Presented, in part, at the American Thoracic Society International Conference, San Diego, May 19–24, 2006.
The authors have not disclosed any potential conflicts of interest.
For information regarding this article, E-mail: Christophe.Girault@chu-rouen.fr