Effect of Noninvasive Ventilation After Unplanned Extubation

BACKGROUND: Our study set out to test the effect of noninvasive ventilation (NIV) performed after unplanned extubation.

METHODS: Retrospective analysis of prospectively collected data in a university-affiliated mixed ICU of 12 beds during a 5-y period (January 2013 to December 2017). Unplanned extubation was defined as the occurrence of an unplanned removal of the endotracheal tube, whether deliberate or accidental. NIV after an unplanned extubation was not protocolized and was decided by the physician in charge on an individual basis.

RESULTS: A total of 121 subjects (median [25th-75th percentile] age, 62.1 [43.3–73.6] y; median [25th-75th percentile] Simplified Acute Physiology Score II, 45 [36–54]) experienced 131 unplanned extubation episodes. Re-intubation was deemed necessary in 35 subjects (28.9%). NIV was used in 24 subjects (19.8%) (prophylactic NIV, n = 10; rescue NIV, n = 14). The re-intubation rates were 25.8%, 10%, and 64.3% in the no NIV, prophylactic, and rescue NIV subgroups, respectively. The median (25th-75th percentile) time to re-intubation was longer for subjects on NIV (9.1 [3.5–49.2] vs 0.46 [0.25–1] h, P = .001). The median (25th-75th percentile) ICU length of stay and duration of mechanical ventilation were longer in the subjects who underwent NIV (14.5 [7–24.5] vs 6 [3–14] d, respectively, P = .004; and 9 [3–22] vs 3 [1–7.3] d, respectively, P = .003).

CONCLUSIONS: NIV after unplanned extubation had uncertain efficacy, especially when provided as rescue management of postextubation respiratory failure.