Usefulness of noninvasive positive-pressure ventilation delivered by helmet versus facial mask in patients with postextubational respiratory distress

초록

Background Extubation failure is common in intensive care unit (ICU). Noninvasive positive-pressure ventilation (NIPPV) for postexubational respiratory distress does not improve rates of intubation in heterogeneous groups of patients. Delivery of NIPPV with helmet would allow for higher pressure than facial mask without substantial air leak. We hypothesize the significant reduction of reintubation rates in the patients with NIPPV delivered by helmet would be expected. Method Between March 1, 2017 and April 30, 2018, Thirty-nine patients with respiratory failure within two hours after extubation and applied by NIPPV were retrospectively collected in a medical ICU in an academic, tertiary care hospital. Results Respiratory failure due to hypercapnia was sixteen patients, hypoxemia was nineteen patients, and postextubational stridor was five patients. Comparing the NIPPV delivered facial mask with helemt, there was no difference in the rate of reintubation (53.8% vs 46.2%; OR, 1; 95% CI, 0.24-3.80) or ICU mortality (4.8% vs 11.1%; OR, 0.59; CI, 033-4.8), duration of NIPPV (2.6 vs 3.7 days, p<0.16), and removal rate of carbon dioxide in blood. According to the cause of respiratory failure, comparing the difference in the rate of reintubation, ICU mortality and removal of carbon dioxide, there was no differences between delivered by facial mask and helmet. Although there was no statistical significance, the severity was higher in the patients with delivered by helmet (sequential organ failure assessment, 2.23 vs 3.44; p<0.09). Conclusion NIPPV delivered by helmet than facial mask significantly reduced the intubation rate among ARDS patients. However, in this study, the effect of NIPPV delivered by helmet was not significant in patients with postextubation respiratory distress. The severity of the patients would be distorted by the small number of patients might influence our outcomes. Thus, we will collect more cases and analyze again each year.