Predicting extubation failure in preterm infants using lung ultrasound: a diagnostic accuracy study

Area of Science:

• Neonatal Medicine
• Pediatric Pulmonology
• Medical Imaging

Background:

• Premature infants (<32 weeks' gestation) often require reintubation after extubation.
• Predicting extubation failure is crucial for optimizing respiratory support and outcomes.
• Lung ultrasound (LUS) is a non-invasive imaging technique with potential for assessing respiratory status.

Purpose of the Study:

• To evaluate the diagnostic accuracy of pre-extubation lung ultrasound (LUS) in predicting reintubation within 72 hours in preterm infants (<32 weeks' gestation).
• To assess secondary outcomes including reintubation within 7 days, stratified by postnatal age, and the utility of lateral imaging alone.
• To conduct subgroup analyses in extremely preterm infants (<28 weeks' gestation).

Main Methods:

• A prospective diagnostic accuracy study was conducted in two neonatal intensive care units.
• Anterior and lateral LUS was performed before extubation, with scores ranging from 0-24.
• Receiver operating characteristic (ROC) analysis was used to determine accuracy, sensitivity, and specificity, reporting areas under the curves (AUCs).

Main Results:

• The study included 100 preterm infants (mean gestational age 27.4 weeks).
• The overall AUC for LUS predicting reintubation within 72 hours was 0.63.
• Accuracy was higher in extremely preterm infants (AUC 0.70) and those extubated <72 hours (AUC 0.90), but poor in infants >7 days old.

Conclusions:

• Lung ultrasound (LUS) was not a strong predictor of reintubation in the overall preterm infant cohort.
• LUS accuracy increased significantly in extremely preterm infants and those extubated at a younger age.
• Future research should target high-risk infants and explore simplified LUS protocols.