Early noninvasive ventilation prevents extubation failure in at-risk patients

Abstract

For patients at high risk of respiratory failure after extubation, does the early implementation of non-invasive positive pressure ventilation (NPPV) avert respiratory failure and reduce mortality? Randomised, controlled trial with concealed allocation. Participants were recruited from two intensive care units. Patients intubated for at least 48 hours who tolerated a spontaneous breathing trial were eligible if they had at least one of the following risk factors for postextubation respiratory failure: age over 65 years; cardiac failure as the cause of intubation; an Acute Physiology and Chronic Health Evaluation (APACHE)-II score greater than 12 on the day of extubation. Exclusion criteria were contraindications to NPPV and orders against resuscitation. Eight eligible patients refused to participate. The remaining 162 were randomised to intervention (n = 79) or control (n = 83) groups. Immediately after extubation, the intervention group received NPPV for 24 hours, followed by oxygen via a Venturi mask. The control group received oxygen via a Venturi mask after extubation. Both groups were otherwise managed according to the clinical protocols of each unit. Respiratory failure in the first three days, mortality in the intensive care unit, and 90-day survival. Respiratory failure was defined as at least two of the following: respiratory acidosis; hypoxaemia at an inspired oxygen fraction of 0.5 or more; respiratory rate > 35/minute; decreased consciousness, agitation or diaphoresis; and respiratory muscle fatigue or increased work of breathing. A subgroup analysis was performed based on the presence of hypercapnia during the spontaneous breathing trial. Compared with the control group, respiratory failure was less frequent (Relative Risk 0.5, 95% confidence interval (CI) 0.3 to 0.9). This indicates that the number of patients that needed to be treated (NNT) with early NPPV to prevent one case of respiratory failure was 6 (95% CI 3 to 36). Mortality in the intensive care unit was significantly lower in the NPPV group, with the NNT to prevent one death being 8 (95% CI 5 to 30). In the subgroup analysis of those with hypercapnia during the spontaneous breathing trial, this benefit was greater (NNT 6, 95% CI 3 to 61). Survival to 90 days showed no significant difference between the NPPV and control groups overall. Among the subgroup with hypercapnia during the spontaneous breathing trial, however, the difference in survival was significant (NPPV 85% vs control 50%, p = 0.006). NPPV for 24 hours post-extubation in patients at high risk of respiratory failure reduced the risk of early respiratory failure and mortality. [Relative Risk and NNT calculated by CAP Editor from data in paper.]