LAUSR

RATIONALE Lung ultrasound is useful in critically ill patients with acute respiratory failure. Given its characteristics, it could be also useful in extremely preterm infants with evolving chronic respiratory failure, as we lack accurate imaging tools to monitor them. OBJECTIVES To verify if lung ultrasound can monitor lung aeration and function, and has good reliability to predict bronchopulmonary dysplasia in extremely preterm neonates. METHODS Multicenter, international, prospective, longitudinal, cohort, diagnostic accuracy study consecutively enrolling inborn neonates with gestational age ≤30+6 weeks. Lung ultrasound was performed on the first, seventh, fourteenth and twenty-eighth days of life and lung ultrasound scores were calculated and correlated with simultaneous blood gases and work of breathing score. Gestational age-adjusted lung ultrasound scores were created, verified in multivariate models and subjected to ROC analyses to predict bronchopulmonary dysplasia at 36 weeks post-menstrual age. MAIN RESULTS Mean lung ultrasound scores are different between infants developing (n=72) or not developing (n=75) BPD (p<0.001,at any time-point). Lung ultrasound scores significantly correlate with oxygenation metrics and work of breathing at any time-point (p always<0.0001). Gestational age-adjusted lung ultrasound scores significantly predict bronchopulmonary dysplasia at 7 (area under ROC curve: 0.826-0.833;p<0.0001) and 14 (area under ROC curve: 0.834-0.858;p<0.0001) days of life. Bronchopulmonary dysplasia severity and gestational age-adjusted lung ultrasound scores are significantly correlated at 7 and 14 days (p always<0.0001). CONCLUSIONS Lung ultrasound scores allow to monitor lung aeration and function in extremely preterm infants. Gestational age-adjusted scores significantly predict the occurrence of BPD,starting from the seventh day of life.