[ Extubation failure predictors in the release process of invasive mechanical ventilatory support in pediatrics]

Introduction: Invasive mechanical ventilation (IMV) in pediatric intensive care units (PICUs) is a ubiquitous practice; the weaning process is supported by predictors, none of which guarantees success.

Objetive: To describe the weaning process from IMV and to determine the predictive factors of extubation failure (EF).

Patients and methods: Retrospective study, patients with more than 48 h of IMV, period 11/2019 to 10/2023. The following were excluded: deceased, referred to another institution and tracheostomized patients. Demographic characteristics, ventilatory status prior to extubation (IMV and blood gas parameters), variables related to the spontaneous breathing test and post-extubation support [noninvasive ventilation (NIV)] of successful vs failed extubation groups were compared aimed at determining the existence of predictors.

Results: 510 extubation events were recorded in 445 patients, with an EF rate of 12.7%. Admission for respiratory reasons (92 vs 78%), higher respiratory rate (22 vs 20 per minute) and fraction of inspired O2 (0.32 vs 0.4) programmed on the ventilator, and use of post-extubation NIV (51 vs 24%) were significantly associated with EF. In multivariate analysis, this association persisted for higher expiratory tidal volume prior to extubation and the use of NIV post-extubation. EF resulted in more days in IMV (10 vs 7) and PICU (13 vs 9).

Conclusion: Higher exhaled tidal volume prior to extubation and use of NIV post-extubation were predictive factors of EF, this being associated with longer stay in IMV.