A lung ultrasound B-line score to stratify oxygen therapy in transient tachypnea of the neonate: a prospective cohort study.
Transient tachypnea of the neonate (TTN) is the most common cause of respiratory distress in newborns. While often self-limiting, the severity of TTN varies significantly, ranging from mild tachypnea to severe respiratory failure requiring non-invasive ventilation. Early and objective grading of TTN severity remains difficult in routine practice. Clinical scores are subjective, and chest radiography is insensitive to the extent of lung fluid. This study evaluated a semi-quantitative lung ultrasound (LUS) B-line score as a predictor of oxygen requirement in neonates with TTN.
In a prospective cohort at a tertiary neonatal intensive care unit (NICU), we enrolled 267 neonates (gestational age ≥ 33 weeks) with a clinical diagnosis of TTN. Within 6 hours of admission, we recorded a standardized twelve-zone LUS score, respiratory rate, and arterial blood gas values. The primary outcome was the highest level of respiratory support within 72 hours, classified as no oxygen support, low-flow oxygen support, or escalated respiratory support. Analyses included Spearman correlation, receiver-operating-characteristic (ROC) curves, and multivariable ordinal logistic regression.
The LUS score correlated inversely with arterial oxygen (PaO2) (ρ = -0.705, P < 0.001) and positively with partial pressure of arterial carbon dioxide (PaCO2) (ρ = 0.399, P < 0.001). Mean scores rose with increasing support (0.5 ± 2.1, 15.0 ± 6.4, and 27.0 ± 4.8 for room air, low-flow, and high-flow groups; P < 0.001). For predicting any oxygen use, the area under the curve (AUC) was 0.982 (95% CI [0.968-0.997]) with an optimal cutoff of 5.5. A cutoff of 22.5 identified infants needing high-flow support (AUC 0.965). In multivariable analysis, the LUS score was the strongest independent predictor of respiratory support level (adjusted odds ratio per point, 1.70; 95% CI [1.41-2.05]; P < 0.001).
A semi-quantitative LUS score provides a non-invasive and physiologically coherent measure of disease severity in TTN and can help stratify early oxygen and respiratory support requirements at the bedside. These findings suggest that incorporating LUS scoring into early NICU assessment may help standardize decisions about initiation and escalation of respiratory support in infants with TTN.
In a prospective cohort at a tertiary neonatal intensive care unit (NICU), we enrolled 267 neonates (gestational age ≥ 33 weeks) with a clinical diagnosis of TTN. Within 6 hours of admission, we recorded a standardized twelve-zone LUS score, respiratory rate, and arterial blood gas values. The primary outcome was the highest level of respiratory support within 72 hours, classified as no oxygen support, low-flow oxygen support, or escalated respiratory support. Analyses included Spearman correlation, receiver-operating-characteristic (ROC) curves, and multivariable ordinal logistic regression.
The LUS score correlated inversely with arterial oxygen (PaO2) (ρ = -0.705, P < 0.001) and positively with partial pressure of arterial carbon dioxide (PaCO2) (ρ = 0.399, P < 0.001). Mean scores rose with increasing support (0.5 ± 2.1, 15.0 ± 6.4, and 27.0 ± 4.8 for room air, low-flow, and high-flow groups; P < 0.001). For predicting any oxygen use, the area under the curve (AUC) was 0.982 (95% CI [0.968-0.997]) with an optimal cutoff of 5.5. A cutoff of 22.5 identified infants needing high-flow support (AUC 0.965). In multivariable analysis, the LUS score was the strongest independent predictor of respiratory support level (adjusted odds ratio per point, 1.70; 95% CI [1.41-2.05]; P < 0.001).
A semi-quantitative LUS score provides a non-invasive and physiologically coherent measure of disease severity in TTN and can help stratify early oxygen and respiratory support requirements at the bedside. These findings suggest that incorporating LUS scoring into early NICU assessment may help standardize decisions about initiation and escalation of respiratory support in infants with TTN.