Lung ultrasound predicts surfactant need in neonates on different respiratory support: an international diagnostic accuracy study

Background An early lung ultrasound score (eaLUS) predicts surfactant administration in neonates of any gestational age with respiratory distress syndrome stabilized on CPAP. Whether this is also true for more unwell infants, non invasively and invasively ventilated, is not known. Aim To assess whether an eaLUS may predict surfactant replacement in infants on different respiratory support modes (CPAP, nasal ventilation and mechanical ventilation). We also investigated eaLUS as a potential marker of unfavorable respiratory outcomes. Methods Prospective, international, multicenter diagnostic study conducted from March 2024 to March 2025 in seven, level III Italian and Spanish neonatal intensive care units. Participants: neonates of any gestational age presenting with respiratory distress, enrolled within 2 h of birth and before the administration of surfactant. Intervention: eaLUS calculated within the first 2 h of life masked to attending physician evaluating both the type of respiratory support and the need for surfactant administration. Main outcome: prognostic accuracy of eaLUS to predict surfactant administration overall and in three subgroups based on early respiratory stabilization strategy (CPAP, NIV and MV). As secondary outcomes we investigated the accuracy of oxygen inspiratory fraction (FiO(2)) > 0.3 by mode of respiratory support to predict surfactant need. Additionally, logistic regression analysis was used to investigate the link between eaLUS with prolonged mechanical ventilation and death or BPD as secondary outcomes. Results We enrolled 229 infants stabilized on CPAP (n = 117), nasal ventilation (n = 58) and mechanical ventilation (n = 54). eaLUS had good prognostic accuracy in the total population (AUC, 0.86; 95% CI, 0.81-0.91). Despite the wide gap of mean airway pressure (MAP) across the population, eaLUS and MAP were only moderately correlated (rho = 0.39; p < 0.001). Optimal thresholds for surfactant replacement were slightly different (LUS >= 8 for CPAP; LUS >= 9 for nasal ventilation; LUS >= 10 for mechanical ventilation). FiO2 > 0.3 was a less accurate predictor of surfactant replacement. eaLUS was significantly associated to prolonged mechanical ventilation beyond both 72 and 168 h. Finally, eaLUS emerged as an independent predictor of BPD or death. Conclusions An early LUS accurately predicts surfactant replacement in neonates non invasively and invasively ventilated and it is a noninvasive imaging prognostic marker of disease severity and adverse respiratory outcomes in this population.