Whilst non-invasive ventilation (NIV) has an established role in treating acute on chronic respiratory failure, its impact on re-intubation rate following elective extubation and subsequent respiratory failure is unclear. This multicentre, randomised, controlled trial addressed whether use of NIV after discontinuation of mechanical ventilation can decrease mortality (primary outcome) or re-intubation rate (secondary outcome). Patients electively discontinued from mechanical ventilation by agreed criteria were observed for 48 hours. Those developing respiratory failure were randomised to a standardised protocol of NIV (114 patients) or medical treatment (107 patients).

All cause mortality was significantly higher in the NIV group (25% v 14%, p = 0.048), although re-intubation rates were the same in both groups (48%). The number needed to harm was nine patients. On the basis of these results, the trial was stopped early at interim analysis. The increased mortality was confined to those requiring re-intubation, and the time between onset of respiratory failure and re-intubation was significantly higher in the NIV group (12 hours v 2.5 hours, p = 0.02). A retrospective subgroup analysis of patients with COPD showed lower re-intubation rates with NIV, although the study was neither designed nor powered to look at this specific question.

This study shows that the use of NIV for respiratory failure following extubation increases mortality and has no effect on re-intubation rates. This may be mediated by a delay in time to re-intubation for those patients receiving NIV. Further studies using selected patient groups (e.g. COPD) may yet show a role for NIV in this setting.