Abstract
BACKGROUND: A total lung ultrasound score (tLUS) is a validated tool to describe parenchymal aeration, evaluate neonatal respiratory distress syndrome (RDS) progression and guide early surfactant replacement. tLUS derives from regional scores (rLUS) from predefined ultrasound views. RESEARCH QUESTION: This paper explores the relative contribution of rLUS to tLUS and their predictive power of surfactant need for RDS, individually and with additional variables. STUDY DESIGN AND METHODS: This was a secondary analysis of multicenter, prospective, observational study. Preterm neonates with RDS were stabilized on nCPAP. Within 2 h of life, we calculated a tLUS (range 0-18) by summing 6 rLUS (using a 0-3 scale on midclavicular, anterior and posterior axillary line views) and the oxygen saturation/inspired oxygen fraction ratio (SatO(2)/FiO(2)). The administration of surfactant by a physician masked to the ultrasound results was used as reference test. RESULTS: We enrolled 175 preterm infants. A midclavicular (MC) score ≥ 2 was an early marker of aeration heterogeneity. Prognostic accuracy for surfactant need was high for the left MC score (AUC: 0.86 with sensitivity 0.79 and specificity 0.90) and the right MC score (AUC 0.87 with sensitivity 0.74 and specificity 0.93; optimal Youden cut-off = 2). A combined left + right MC score lead to an AUC: 0.90 (sensitivity 0.82. specificity 0.89; optimal Youden cut-off = 3). A prediction model including gestational age, SatO(2)/FiO(2) and the combined MC score had an AUC 0.95. INTERPRETATION: rLUS are not always uniformly distributed in early RDS. The combined MC score is a simplified rapid and accurate predictor of surfactant replacement (alone or in combination with noninvasive variables) reducing stressful manipulations in first hours of life for preterm neonates.