Outpatient assessment of bronchopulmonary dysplasia using point of care lung ultrasound

Background: Lung point-of-care ultrasound (POCUS) is an emerging technology for assessment of neonatal lung diseases. Prior studies have described lung ultrasound scores (LUS) in bronchopulmonary dysplasia (BPD). However, few have assessed LUS in patients with BPD beyond the initial neonatal intensive care unit (NICU) hospitalization. Methods: We performed a cross-sectional study to evaluate outpatient LUS abnormalities and respiratory morbidity in children with and without a history of BPD. We hypothesized that infants with severe BPD would have higher (more abnormal) LUS compared to healthy infants. Eligible children received a single lung ultrasound evaluation during an outpatient clinic visit. The lung ultrasound images were analyzed and scored (0-18). A LUS greater than 0 was considered abnormal. Results: 125 participants aged 0-24 months were enrolled, including 26 healthy full-term children (controls) and 45 preterm children without BPD (Grade 0). 16, 26, and 12 were enrolled with Grade 1, 2 and 3 BPD, respectively. Lung ultrasound scores were significantly higher in children with Grade 2 and Grade 3 BPD (mean LUS of 0.68 and 2.67, respectively), compared to healthy term children (mean LUS of 0.11). However, 91 preterm participants, including some with severe BPD, had normal ultrasounds (LUS = 0). Adding abnormal LUS to a model including established clinical risk factors significantly improved the model's ability to identify which children with BPD were likely to have a history of hospital readmissions (AUC increased from 0.718 to 0.785). However, due to low inter-observer agreement, these findings should be interpreted as exploratory and observer-dependent. Conclusion: Lung POCUS may be a feasible adjunctive assessment tool for children with BPD, and abnormal LUS may be associated with respiratory-related hospital readmissions. However, normal LUS findings were common, including among some children with severe BPD. Longitudinal studies with a larger cohort are needed to further evaluate its use and limitations.