703 - Utility of Lung Point-of-Care Ultrasound in Predicting Duration of Respiratory Support in the Early Neonatal Period: A Pilot Study

Publication Number: 4687.703

Juhi D. Motiani, University of Illinois College of Medicine, Chicago, IL, United States; Stefan Tchernodrinski, University of Illinois College of Medicine, Arlington Heights, IL, United States; Kelly McConnell, University of Illinois College of Medicine, Chicago, IL, United States; Mayra A. Matar, University of Illinois College of Medicine, Chicago, IL, United States; Alan Schwartz, University of Illinois College of Medicine, Chicago, IL, United States; Nauri E. Abreu, Children's Hospital of the University of Illinois, Chicago, IL, United States; SaurabhKumar Patel, Children's Hospital of the University of Illinois, Chicago, IL, United States; Nishant Srinivasan, Children's Hospital of the University of Illinois, Chicago, IL, United States

Background: Point-of-care ultrasound (POCUS) has become an invaluable tool in neonatal intensive care units, commonly used for bedside procedural guidance and hemodynamic monitoring. Its value lies in its immediacy, portability, and avoidance of radiation- often being referred to as "the new stethoscope." Regarding the respiratory system, current literature has focused primarily on diagnostic capabilities of POCUS, rather than on its use for prognosis or guiding clinical interventions. A persistent problem in managing infants on respiratory support is not objectively knowing when to wean. This could lead to re-escalation of support causing further distress to parents. Lung ultrasound (LUS) allows quantification of extravascular lung water via B-line analysis to calculate a lung fluid index (LFI). Although LFI is widely used in adults, its ability to predict respiratory illness course and duration of respiratory support is yet to be established in neonates.

Objective: This study aimed to assess whether LFI through POCUS can serve as an accurate tool in estimating lung fluid burden and predicting duration of respiratory support in neonates in the early neonatal period.

Design/Methods: This pilot, prospective observational cohort study included 10 neonates admitted to the NICU at the University of Illinois Hospital. LUS were performed at standardized intervals: within 6 hours of life, around 24 hours, daily until cessation of respiratory support, and again at 1 week of life or prior to discharge. LFI scores were calculated using a five-point scale based on B-line patterns. Images were scored blindly by two independent reviewers. Clinical data, including gestational age, delivery mode, and respiratory support duration, were collected.

Results: Exam 1 (within 6 hours of life) scores demonstrated a statistically significant correlation with total duration of respiratory support (Pearson r = 0.633, p = 0.049; Spearman ρ = 0.676, p = 0.032), whereas Exam 2 (E2) and Exam 3 (E3) scores did not correlate significantly. All patients were >36 weeks gestation, born via cesarean section, BW >2500g and did not require surfactant administration.

Conclusion(s): These findings demonstrate that LFI especially if done early ( < 6 hours) significantly correlates with duration of respiratory support in the early neonatal period. This information may be useful to clinicians in the decision to wean from respiratory support, reduce radiograph usage and improve counseling of family on duration of illness. Future steps include expanding the cohort and comparing LFI with physician interpretation of chest radiography.

Lung Fluid Index Trend in Serial POCUS Exams
Figure 1: Most patients show a decline in Average Score-H (Average of the highest score per anatomical view)
over time, reflecting improvement in lung fluid burden.

LFI Correlation with Respiratory Support Duration
Figure 2: Here is a side-by-side visualization showing the correlation between lung
fluid index scores at each exam timepoint (E1, E2, E3) and duration of respiratory
support. E1 (left panel) shows a very strong positive linear correlation, confirming that
higher initial scores are associated with longer durations of support. E2 (middle) and
E3 (right) show weak or no clear trend, indicating these later scores have minimal
predictive value for respiratory duration.