Lung ultrasound score as a tool to predict severity of bronchopulmonary dysplasia in neonates born ≤25 weeks of gestational age

Abstract

Objective: The primary aim was to evaluate whether the addition of the posterior lung aided in diagnostic accuracy of predicting bronchopulmonary dysplasia (BPD) vs moderate-severe BPD (msBPD); the secondary aim was to explore the diagnostic accuracy of two protocols for BPD vs msBPD.

Study design: This was a single-center prospective observational study. Preterm infants with a gestational age ≤ 25 weeks were included. Two LUS score protocols were evaluated on the 14th day of life (DOL): (A) evaluating the anterolateral (LUS score-al) lung and (B) the anterolateral combined with posterior (LUS score-alp) lung. The LUS score range for the two protocols was 0-32 and 0-48, respectively.

Results: A total of eighty-nine infants were enrolled. Both the LUS score-al and LUS score-alp were higher in neonates developing BPD and msBPD than in the rest of the cohort (LUS score-al 24 (23,26) vs 22 (20,23); LUS score-alp 36 (34,39) vs 28 (25,32)) (LUS score-al 25 (24,26) vs 23 (21,24); LUS score-alp 40 (39,40) vs 34 (28,36)). The LUS score-al on the 14th DOL showed a moderate diagnostic accuracy to predict BPD and msBPD (AUC 95% CI: 0.797 [0.697-0.896]; 0.811[0.713-0.909]), while the LUS score-alp significantly improved diagnostic accuracy of BPD and msBPD (AUC 95% CI: 0.902 [0.834-0.970]; 0.922 [0.848-0.996]). A cutoff of 25 points in the LUS score-al provided a sensitivity, specificity, positive likelihood ratio, and negative likelihood ratio of 76.9%, 79.4%, 3.7, and 0.3 respectively to predict msBPD. Meanwhile, that of 39 points in the LUS score-alp provided a sensitivity, specificity, positive likelihood ratio, and negative likelihood ratio of 81%, 98.4%, 50.5 and 0.19 to predict msBPD, respectively.

Conclusions: The LUS score on the 14th DOL can predict BPD and msBPD with moderate diagnostic accuracy. Apart from that, scanning posterior enhanced diagnostic accuracy.