Prediction of Bronchopulmonary Dysplasia by Diaphragmatic Ultrasound in Preterm Infants: A Prospective Pilot Study

Background: Diaphragmatic ultrasound showed significant differences between patients with established bronchopulmonary dysplasia (BPD) and healthy controls. The aim of this study was to assess whether diaphragmatic ultrasound could predict the development of BPD in preterm infants < 32 weeks of gestational age. Methods: Diaphragmatic ultrasound was performed on day 3 of life (T0), and at 7 ± 1 (T1), 14 ± 1 (T2), and 21 ± 2 (T3) days of life. Diaphragmatic excursion (DE), diaphragmatic thickness at end of inspiration (DTins) and expiration (DTexp), inspiratory and expiratory peak velocities (I-peak and E-peak) and their ratio to body surface area (BSA), and diaphragmatic thickness fraction (DTF) were measured. Logistic regression and ROC curve analyses were performed to evaluate possible role of these variables as predictive factors for BPD and their accuracy for the prediction of BPD. Results: DE/BSA, DTins/BSA, DTexp/BSA, and DTF did not differ between patients who developed or did not developed BPD. I-peak/BSA and E-peak/BSA were significantly higher in patients who developed BPD at all timings. I-peak/BSA at T2 and E-peak/BSA at T0 and T1 were independent predictive factors for BPD after adjustment for gestational age and respiratory support. At T1 I-peak/BSA > 18 cm/s/m2 and E-peak/BSA > 17.1 cm/s/m2 accurately predict BPD with sensitivity of 92% and specificity of 92%, and sensitivity of 99% and specificity of 79%, respectively. Conclusions: I-peak/BSA and E-peak/BSA are independent risk factors for the development of BPD. They can early and accurately predict the risk for BPD in very preterm infants contributing to targeted treatment of patients at higher risk of BPD.