Are lung ultrasound and chest radiograph equally reliable for neonatal imaging? A scoping review

Chest radiograph (CR) has traditionally been the standard in imaging neonatal respiratory diseases. Lung ultrasound (LUS) has recently emerged as a radiation-free alternative. This scoping review compares the reliability of these two techniques for diagnosing respiratory distress syndrome (RDS), evaluating the need of surfactant replacement, and predicting the development of bronchopulmonary dysplasia (BPD). A literature search was conducted in PubMed, Embase, and Cochrane Library for studies published between 1974 and December 2024. The following search terms were used: “neonate,” “lung ultrasound,” “chest radiography,” “reliability,” “Respiratory Distress Syndrome,” “Bronchopulmonary Dysplasia,” and “surfactant administration.” The search was limited to full articles. Only studies evaluating the inter- and intra-observer agreement were included. Out of 322 evaluated papers, 14 met the inclusion criteria and were included in this review. LUS had four papers (308 neonates) for RDS diagnosis with Cohen’s k or intraclass correlation coefficient (ICC) ≥ 0.9, four studies (480 neonates) for the need of surfactant replacement with Cohen’s k ≥ 0.83, and three papers for BPD prediction (280 infants) showing a Cohen’s k ≥ 0.82. We found no publication on CR reliability for RDS diagnosis, one paper on 56 neonates on its prediction of surfactant need with a Cohen’s k = 0.86, and two studies (146 neonates) on BPD prediction with Cohen’s k ranging from 0.19 to 0.41. Conclusion: We show that, unlike CR, LUS has good quality evidence in favor of its reliability for both diagnosis and prognosis of significant neonatal respiratory diseases and guiding surfactant delivery. (Table presented.)