Lung Ultrasound for Prediction of Bronchopulmonary Dysplasia in Extremely Preterm Neonates: A Prospective Diagnostic Cohort Study

Abstract

Background: Bronchopulmonary dysplasia (BPD) remains a common and serious complication of preterm birth. Early prediction of BPD can be useful to alter its course using clinical, quality improvement and research strategies. Our aim is to evaluate the diagnostic and predictive ability of lung ultrasound at three time points in the first two weeks after birth for predicting BPD among infants less than 29 weeks gestational age. Methods: This was a prospective, diagnostic cohort study conducted between July 2019 and December 2020 at two Neonatal Intensive Care Units. Eligible infants were enrolled after parental consent. Lung ultrasound was performed on days 3, 7 and 14 after birth and lung ultrasound scores (LUS) were calculated in blinded fashion. Diagnostic test characteristics and area under receiver operating characteristic (AUROC) curves were calculated and compared with values obtained from clinical predictors. Findings: A total of 152 infants were enrolled with mean (SD) gestational age of 26·5 (1.5) weeks and mean (SD) birth weight of 923 (250) grams. Of them, 87 (57%) infants were diagnosed with BPD. The LUS were significantly higher in infants diagnosed with BPD compared to those without BPD at all scan time points (p 10 at all 3 timepoints had higher sensitivity (0·89, 0·89, and 0·77) with specificity (0·87, 0·90, and 0·92) and corresponding clinically important positive and negative likelihood ratios. The AUROC for LUS at the three time points were 0·96, 0·97 and 0·95 on day 3, 7 and 14 respectively. Compared to the model using clinical characteristics, LUS alone had higher AUROC (p<0·05 for all three time points). Interpretation: In this cohort, LUS in the first 2 weeks after birth had a very high predictive value for the diagnosis of BPD among infants of <29 weeks’ gestation. Trial Registration: Study was registered at http:// www.clinicaltrials.gov (registration # NCT04756297). Funding: No specific funding was received. Declaration of Interests: The authors have no conflicts of interest or potential conflicts of interest relevant to this article to disclose. Ethics Approval Statement: Research Ethics Board (REB) approvals were obtained from both participating institutions. An informed consent was obtained from parents prior to patient’s enrollment.