Non-Invasive Ventilation Versus High-flow Nasal Oxygen in Intensive Care Units

In intensive care units (ICUs), around 20% of patients experience respiratory failure after planned extubation (defined as removal of endotracheal tube after weaning from the ventilator). Nearly 40-50% of them eventually require reintubation with subsequently high mortality rates reaching 30-40%.Consequently, a noninvasive respiratory support strategy aiming at avoiding reintubation deserves consideration. Whereas prophylactic use of non-invasive ventilation (NIV) applied immediately after extubation may prevent respiratory failure, NIV used as rescue therapy to treat established post-extubation respiratory failure could increase the risk of death. The largest clinical trial conducted to date by Esteban and colleagues showed greater mortality with NIV than with standard oxygen. The only factor explaining the deleterious effects of NIV was that the time before reintubation was markedly longer than with standard oxygen, suggesting that NIV may worsen the outcome by delaying reintubation. No further large-scale clinical trial has been performed after this study and thereby, the most recent international clinical practice guidelines suggest that NIV should not be used in this setting. However, recent large-scale clinical trials on extubation have shown that around 40 to 50% of patients with post-extubation respiratory failure are actually treated with NIV. Indeed, NIV as rescue therapy may avoid reintubation in a number of cases, especially in patients with underlying chronic lung disease. In a recent post-hoc analysis focusing on 146 patients with established post-extubation respiratory failure, we found that mortality was 18% with NIV and 29% with high flow nasal oxygen alone (difference, - 11% [95% CI, -25% to 2%]; p=0.12). Mortality rates were even significantly lower with NIV than with high-flow nasal oxygen alone in patients with hypercapnia, suggesting that after all, NIV might not increase the risk of death. To explain the discrepancies with the Esteban study, the time interval between NIV initiation and reintubation was markedly shorter in our study (5 hours in median vs. 12) and predefined criteria for reintubation were established in order to minimize the risk of delayed intubation. Moreover, all participating centers had extensive experience in NIV which was not the case in the previous trial. Whereas high-flow nasal oxygen has never been specifically studied for management of post-extubation respiratory failure, this respiratory support could in this setting constitute an alternative to standard oxygen or NIV. High-flow nasal oxygen is increasingly used after extubation in order to prevent post-extubation respiratory failure, and its beneficial effects have been reported in treatment of acute respiratory failure. To date, the best noninvasive respiratory support strategy in patients with post-extubation respiratory failure remains unknown. However, high-flow nasal oxygen could be included among reference respiratory supports whereas standard oxygen would be considered as a suboptimal strategy after extubation in ICUs.Thereby, we have decided to assess whether NIV alternating with high-flow nasal oxygen as compared to high-flow nasal oxygen alone may decrease mortality of patients in ICUs with post-extubation respiratory failure.

To do that, we propose to conduct a prospective multicenter randomized controlled open-label trial comparing these 2 strategies of noninvasive respiratory support in patients with post-extubation respiratory failure in ICUs. Patients included will be randomized as early as possible after the onset of respiratory failure and will be assigned to one of the 2 following groups, with a 1:1 ratio: high-flow nasal oxygen group treated by high-flow nasal oxygen alone, or NIV group treated by NIV alternating with high-flow nasal oxygen. As NIV may be more effective in hypercapnic patients, randomization will be stratified according to the PaCO2 level measured on blood gas performed at inclusion with the aim at ensuring a balanced allocation of hypercapnic patients (PaCO2 > 45 mmHg) among the 2 groups.