Early extubation followed by immediate noninvasive ventilation vs. standard extubation in hypoxemic patients: a randomized clinical trial.

Abstract

Noninvasive ventilation (NIV) may facilitate withdrawal of invasive mechanical ventilation (i-MV) and shorten intensive care unit (ICU) length of stay (LOS) in hypercapnic patients, while data are lacking on hypoxemic patients. We aim to determine whether NIV after early extubation reduces the duration of i-MV and ICU LOS in patients recovering from hypoxemic acute respiratory failure. Highly selected non-hypercapnic hypoxemic patients were randomly assigned to receive NIV after early orstandard extubation. Co-primary end points were duration of i-MV and ICU LOS. Secondary end points were treatment failure, severe events (hemorrhagic, septic, cardiac, renal or neurologic episodes, pneumothorax or pulmonary embolism), ventilator-associated pneumonia (VAP) or tracheobronchitis (VAT), tracheotomy, percent of patients receiving sedation after study enrollment, hospital LOS, and ICU and hospital mortality. We enrolled 130 consecutive patients, 65 treatments and 65 controls. Duration of i-MV was shorter in the treatment group than for controls [4.0 (3.0-7.0) vs. 5.5 (4.0-9.0) days, respectively, p = 0.004], while ICU LOS was not significantly different [8.0 (6.0-12.0) vs. 9.0 (6.5-12.5) days, respectively (p = 0.259)]. Incidence of VAT or VAP (9% vs. 25%, p = 0.019), rate of patients requiring infusion of sedatives after enrollment (57% vs. 85%, p = 0.001), and hospital LOS, 20 (13-32) vs. 27(18-39) days (p = 0.043) were all significantly reduced in the treatment group compared with controls. There were no significant differences in ICU and hospital mortality or in the number of treatment failures, severe events, and tracheostomies. In highly selected hypoxemic patients, early extubation followed by immediate NIV application reduced the days spent on invasive ventilation without affecting ICU LOS.