Diagnostic value of lung ultrasound in evaluating the severity of neonatal respiratory distress syndrome

Background: It is still unclear whether lung ultrasound (LUS) can be used to evaluate the severity of neonatal respiratory distress syndrome (NRDS). Objective: To evaluate the role of LUS in assessing NRDS. Methods: From January 2017 to January 2018, newborns with suspected NRDS were enrolled. The LUS score and lung consolidation areas were determined. The receiver operative curve (ROC) was used to analyze the LUS score and lung consolidation to predict NRDS severity. Results: Neonates with NRDS had higher LUS scores than those with non-NRDS (23.6 +/- 3.6 vs. 16.2 +/- 1.8, P < 0.05). Among neonates with NRDS, the LUS scores increased with NRDS severity (18.0 +/- 2.7 vs. 24.0 +/- 1.7 vs. 27.0 +/- 1.7, all P < 0.05). There were almost no consolidation areas in non-NRDS, while 1.9 +/- 1.7 consolidation areas were observed in the NRDS group (P < 0.05). The number of consolidation areas also increased with NRDS severity (0 vs. 1.5 +/- 0.8 vs. 4.1 +/- 1.3, all P < 0.05). The LUS score for NRDS vs. non-NRDS showed 80.2% sensitivity and 100% specificity using a cut-off of 21.5 (Area under the ROC curve, AUC = 0.938; P < 0.001). The LUS score for severe vs. mild/moderate NRDS showed 73.1% sensitivity and 95.7% specificity using a cut-off of 25.5 (AUC = 0.944; P < 0.001). The LUS score for predicting mechanical ventilation showed 81.3% sensitivity and 88.8% specificity using a cut-off of 25.5 (AUC = 0.912; P < 0.001). The AUCs of consolidation areas were similar to those of LUS score (all P > 0.05). Conclusion: The LUS score and consolidation areas can discriminate NRDS from non-NRDS and the different grades of NRDS, and predict the application of mechanical ventilation.