*Corresponding Author
Full name: Almudena Alonso-Ojembarrena
Department: Neonatal Intensive Care Unit
Hospital: Puerta del Mar University Hospital
Street Name & Number: Avenida Ana de Viya 11
City, State, Postal code, Country: Cádiz 11010. Spain
Tel: 34956002314; Fax: 34956004801.
E-mail: almudena.alonso.sspa@juntadeandalucia.es
ORCID number: 0000-0002-2413-9758
KEYWORDS: lung/diagnostic imaging, newborn, preterm, ultrasonography.
Abstract
We explain our concerns on using plain LU score to predict extubation failure in preterm infants, as gestational age at birth can be an important confounder.
Dear Editor,
We read with interest the article published in the last issue of Pediatric Pulmonology by El Amrousy D. and coworkers1on lung ultrasound score (LUS) as a predictor of extubation failure (EF) in neonates during the 48 hours following the discontinuation of mechanical ventilation. The authors conclude that LUS above 10 is a reliable EF marker.
Although we commend the authors for translating to the NICU a strategy already tested in adult critical care2, we would like to express a word of caution.
Our concern regards a significant GA difference: 32 weeks (95% confidence interval (CI) 27.5-36.5 weeks) for babies failing extubation versus 35 weeks (95% CI 31.2-38.8 weeks) for infants who were not reintubated (p=0.045).
According to a previous publication 3 and our personal experience, LUS has a marked GA dependence. Compared to late preterm babies, ELBW infants tend to keep higher LUS regardless of duration of mechanical ventilation or development of bronchopulmonary dysplasia. This translates in a variable number of B lines often reaching coalescence long after extubation, regardless of the respiratory support. Preterm infants under 28 weeks without BPD may remain with similar LUS than those evolving to BPD until 4-6 weeks of age (LUS from 5-10), irrespective of the respiratory support they receive4.
Since LUS is a reliable marker of lung aeration5, the persistence of high scores may be due to pulmonary insufficiency of prematurity 6, a 50% to 70% loss of lung volume that makes these infants depend on non-invasive ventilation for a long time after recovering from the initial respiratory distress syndrome.
For these reasons, we believe that a single LUS extubation threshold over a wide GA span is incorrect. GA is a powerful confounder that should have been accounted for in a multivariate analysis.