Objective: To evaluate the effects of non-invasive positive pressure ventilation (NIPPV) on acute respiratory failure and to identify the factors associated with its failure or success. Patients and methods: This open and prospective study analyzed 60 episodes of NIPPV use in 53 patients with respiratory failure. Results: In 37 occasions (62%), the patients were successfully ventilated with NIPPV (Success Group, SG) whereas in 23 episodes (38%) intubation was necessary (Failure Group, FG). FG patients showed a higher Apache II score (30.4 ± 9 versus 22.2 ± 8, p = 0.001). The use of NIPPV reduced respiratory rate after 2 hs. This reduction was less evident in FG (from 33 ± 9 to 30 ± 8 bpm, p = 0.094) than in SG (from 39 ± 11 to 28 ± 9 bpm, p < 0.001). PaO2 increased from 62 ± 22 to 101 ± 65 mmHg, (p < 0.001) with no differences between groups. Among patients with hypercapnia, PaCO2 level decreased in SG (from 76 ± 20 to 68 ± 21mmHg, p = 0.032) but not in FG (from 89 ± 23 to 93 ± 40mmHg, p = 0.54). Arterial pH increased in SG (from 7.25 ± 0.10 to 7.34 ± 0.11, p = 0.007) and not in FG (from 7.24 ± 0.07 to 7.21 ± 0.12, p = 0.48). NIPPV was used for longer in SG (3.4 ± 2.5 versus 2.3 ± 2 days, p = 0.003). Higher IPAP levels were used in SG (13.2 ± 3 versus 11 ± 4cmH2O, p = 0.02). The 10 patients (17%) who died, all belonged to FG. Skin lesions at the site of mask contact was the most common complication of NIPPV (5, 8%). NIPPV was efficient in treating acute respiratory failure in 2/3 of patients. The factors associated with failure were: greater severity of the illness, smaller reduction in the respiratory rate, no improvement of respiratory acidosis in 2 h and use of lower IPAP levels. The high mortality rate observed in FG (10, 43%) justifies efforts to optimize NIPPV implementation and to recognize its failure, as early as possible, in order to avoid delays in endotracheal intubation.

Keywords: Non-invasive positive pressure ventilation. Respiratory failure. Apache II. Hypercapnia.