Easily identified at-risk patients for extubation failure may benefit from noninvasive ventilation: a prospective before-after study

Abstract

Background: While studies have suggested that prophylactic noninvasive ventilation (NIV) could prevent post-extubation respiratory failure in the intensive care unit, they appear inconsistent with regard to reintubation. We assessed the impact of a prophylactic NIV protocol on reintubation in a large population of at-risk patients.

Methods: Prospective before-after study performed in the medical ICU of a teaching referral hospital. In the control cohort, we determined that patients older than 65 years and those with underlying cardiac or respiratory disease were at high-risk for reintubation. In the interventional cohort, we implemented a protocol using prophylactic NIV in all patients intubated at least 24 h and having one of these risk factors. NIV was immediately applied after planned extubation during at least the first 24 hours. Extubation failure was defined by the need for reintubation within seven days following extubation.

Results: We included 83 patients at high-risk among 132 extubated patients in the control cohort (12-month period) and 150 patients at high-risk among 225 extubated patients in the NIV cohort (18-month period). The reintubation rate was significantly decreased from 28% in the control cohort (23/83) to 15% (23/150) in the NIV cohort (p = 0.02 log-rank test), whereas the non-at-risk patients did not significantly differ in the two periods (10.2% vs. 10.7%, p = 0.93). After multivariate logistic-regression analysis, the use of prophylactic NIV protocol was independently associated with extubation success.

Conclusions: The implementation of prophylactic NIV after extubation may reduce the reintubation rate in a large population of patients with easily identified risk factors for extubation failure.

Fig. 1

Flow-chart of the study. All in all, 168 patients in the control cohort and 310 in the noninvasive ventilation (NIV) cohort experienced planned extubation. Patients who were extubated after less than 24 h of mechanical ventilation (MV) or with a do-not-intubate order were excluded. Among all patients who experienced planned extubation, the proportion of patients at high-risk for extubation failure was similar in the two cohorts: 49 % (83/168) in the control cohort and 48 % (150/310) in the NIV cohort, p = 0.83

Fig. 2

The Kaplan-Meier plots of the cumulative reintubation rates within seven days following extubation. The rate of extubation failure (reintubation at day 7) significantly differed between the four groups (p = 0.0165 log-rank test). Among the patients at high-risk, this rate was significantly lower in the NIV cohort (green solid line) than in the control cohort (blue solid line), decreasing from 28 % (23/83) to 15 % (23/150), p = 0.0225 by log-rank test. The difference remained significantly lower after having excluded the 11 patients in the NIV cohort who did not effectively receive NIV: 17 % (23/139) vs. 28 % (23/89), p = 0.0466 log-rank test. By contrast, the rate of extubation failure did not significantly differ during the two periods in patients at low-risk: 10.2 % (5/49) in the control cohort (blue dotted line) vs. 10.7 % (8/75) in the NIV cohort (green dotted line), p = 0.93