Objective: To assess the value of a spontaneous breathing trial (SBT) using a flow-inflating bag in predicting extubation success. Secondary goals were to evaluate the positive and negative predictive accuracy of a 15-min SBT.
Design: Prospective, blinded, clinical study.
Setting: Pediatric intensive care unit (ICU) of a university hospital
Patients: Infants and children intubated for ≥24 hrs.
Interventions: Patients who met defined criteria for extubation underwent a 15-min SBT connected to a flow-inflating bag set to provide 5 cm H2O continuous positive airway pressure.
Measurements and Main Results: Seventy patients underwent the SBT. Respiratory rate, heart rate, blood pressure, and pulse oxygen saturations were recorded at baseline and at 5 and 15 mins into the SBT. The ICU physicians were blinded to the results of the SBT, and all patients were extubated at the end of the trial. Patients were observed for the next 24 hrs, and the need for noninvasive ventilation or reintubation (i.e., extubation failure) was recorded.
Sixty-four patients (91%) passed the SBT with a subsequent extubation failure rate of 7.8% (only 1.6% required reintubation). Six of the 70 (9%) patients enrolled failed the trial, but half were extubated successfully. Successful completion of the SBT has a 95% sensitivity for predicting successful extubation with a positive predictive value of 92% and an odds ratio of 12 (95% confidence interval, 1.3, 53.7). The specificity of the SBT was 37% with a negative predictive value of 50%. Logistic regression analysis revealed a significant association between passing the SBT and extubation success (p = .017).
Conclusions: A 15-min flow-inflating bag SBT represents a practical, reliable bedside test that has 95% sensitivity for predicting extubation success in pediatric ICU patients. A trial failure is associated with but does not accurately predict extubation failure.
From the Department of Pediatrics, Division of Pediatric Critical Care, University of Florida College of Medicine, Gainesville, FL
The authors do not have any financial interest to report.
The authors have not disclosed any potential conflicts of interest.
Address requests for reprints to: Arno Zaritsky, MD, University of Florida, Department of Pediatrics, Division of Pediatric Critical Care, P.O. Box 100296, Gainesville, FL 32610-0296. E-mail: firstname.lastname@example.org.