Association Between Pressure Support During Extubation Readiness Testing and Time to First Extubation in Children with Congenital Heart Disease.

BACKGROUND: Extubation readiness testing (ERT) is often performed in children with congenital heart disease prior to liberation from mechanical ventilation. The ideal ERT method in this popula-tion is unknown. We recently changed our ERT method from variable (10, 8, or 6 cm H2O, depend-ing on endotracheal tube size) to fixed (5 cm H2O) pressure support (PS). Our study assessed the association between this change and time to first extubation and need for re-intubation. METHODS: We studied 2 temporally distinct cohorts, one where ERT was conducted with variable PS and another using PS fixed at 5 cm H2O. Data were prospectively collected as part of a quality improve-ment project. The primary outcome was time to first extubation. Secondary outcomes were need for re-intubation and percentage of successful ERTs. We performed Poisson regression or logistic regres-sion for the association between PS during ERT and time to first extubation or re-intubation, respectively. RESULTS: We included 320 subjects, 186 in the variable PS group and 152 in fixed PS group. In unadjusted analysis, median time to first extubation was longer in the fixed PS group com-pared to the variable PS group (4.1 [2.0-7.1] d vs 3.1 [1.1-5.9] d, P = .02), and there was no differ-ence in re-intubation rate (11% vs 8%, P = .34). Subjects in the fixed PS group were significantly more likely to be mechanically ventilated after cardiac arrest, have a Society of Thoracic Surgeons -European Association for Cardio-Thoracic Surgery (STAT) category of 4 or 5, be extubated on day shift, receive enteral feeds at extubation, have higher respiratory support at extubation, and higher dead-space-to-tidal-volume ratio. After controlling for these variables in multivariable regression, we found no association between the choice of PS and time to first extubation or re-intubation. CONCLUSIONS: The use of a fixed PS of 5 cm H2O instead of variable PS during ERT was not associated with longer time to first extubation or extubation failure.