RESUMO
PURPOSE: Neurocritical care patients receive prolonged invasive mechanical ventilation (IMV), but there is poor specific information in this high-risk population about the liberation strategies of invasive mechanical ventilation. METHODS: ENIO (NCT03400904) is an international, prospective observational study, in 73 intensive care units (ICUs) in 18 countries from 2018 to 2020. Neurocritical care patients with a Glasgow Coma Score (GCS) ≤ 12, receiving IMV ≥ 24 h, undergoing extubation attempt or tracheostomy were included. The primary endpoint was extubation failure by day 5. An extubation success prediction score was created, with 2/3 of patients randomly allocated to the training cohort and 1/3 to the validation cohort. Secondary endpoints were the duration of IMV and in-ICU mortality. RESULTS: 1512 patients were included. Among the 1193 (78.9%) patients who underwent an extubation attempt, 231 (19.4%) failures were recorded. The score for successful extubation prediction retained 20 variables as independent predictors. The area under the curve (AUC) in the training cohort was 0.79 95% confidence interval (CI95) [0.71-0.87] and 0.71 CI95 [0.61-0.81] in the validation cohort. Patients with extubation failure displayed a longer IMV duration (14 [7-21] vs 6 [3-11] days) and a higher in-ICU mortality rate (8.7% vs 2.4%). Three hundred and nineteen (21.1%) patients underwent tracheostomy without extubation attempt. Patients with direct tracheostomy displayed a longer duration of IMV and higher in-ICU mortality than patients with an extubation attempt (success and failure). CONCLUSIONS: In neurocritical care patients, extubation failure is high and is associated with unfavourable outcomes. A score could predict extubation success in multiple settings. However, it will be mandatory to validate our findings in another prospective independent cohort.
