358 - Serial Assessment of Lung Ultrasound Scores to Assess Response to Furosemide in Preterm Infants

Publication Number: 358.6848

Ianna Blanchard, University of Kentucky College of Medicine, Lexington, KY, United States; Ravi Bhavsar, University of Kentucky College of Medicine, Lexington, KY, United States; Mina Hanna, University of Kentucky - - Lexington, KY, Lexington, KY, United States; Ranjit Torgalkar, University of Kentucky College of Medicine, Lexington, KY, United States; M. Douglas. Cunningham, University of Kentucky College of Medicine, Lexington, KY, United States; Hemanthi Veligaram, University of Kentucky College of Medicine, Richmond, KY, United States; Prasad Bhandary, University of Kentucky, Lexington, KY, United States; Madison Meister, University of Kentucky College of Medicine, Clearfield, KY, United States; Sanjay Dhar, University of Kentucky College of Medicine, Lexington, KY, United States; Rupin Sharma, University of Kentucky College of Medicine, Lexington, KY, United States

Background: Lung ultrasound scores (LUSc) correlate excellently with oxygenation status in preterm infants. Despite limited evidence on utility and concerns for adverse effects, diuretics are commonly used in infants on respiratory support to improve pulmonary mechanics and interstitial edema. Clinical response is qualitatively assessed, subject to observer interpretation causing practice variation and potential for extended use. LUSc may provide an objective assessment of pulmonary air:fluid ratio during diuretic use. We aim to use bedside ultrasound to identify which infants may benefit from furosemide.

Objective: To quantitatively assess response to furosemide in preterm infants weighing < 1500g requiring respiratory support using serial LUSc.
To determine the correlation between LUSc and traditional clinical parameters (fraction of inspired oxygen, respiratory support, urine output, and weight change) of response to furosemide.
To identify baseline clinical variables and LUSc characteristics associated with a positive response to furosemide.

Design/Methods: Infants born < 32 WGA with birth weight < 1500g requiring respiratory support (mechanical ventilation or non-invasive support) and receiving repeated-dose furosemide for evolving BPD were included in this prospective observational study. 
LUSc were measured on days 0, 1, 3, 5, and 7 during a course of furosemide, using a validated scoring tool. Imaging was performed by 4 credentialed neonatologists and reviewed by two blinded experts. Informed consent was obtained prior to enrollment. Statistical testing was performed using repeated measures ANOVA, linear regression, and Spearman’s rank-order correlation test to measure correlation across variables.

Results: 30 infants with median BW 770g and GA 26w4d were enrolled (Table 1). Repeated measures ANOVA was performed to evaluate change in LUSc by furosemide day, revealing a significant decrease in LUSc over time. There was a significant positive correlation between % change in FiO2 and % change in LUSc with time by Spearman’s rank-order correlation (rₛ(40)=1.00, p=.0167)(Figure 1). LUSc on D0 was used to predict % change in LUSc from D0 to D7 of furosemide using simple linear regression. For each point increase in LUSc on D0 there was about a 10% decrease in LUSc by D7(Figure 2). % decrease in FiO2 on D0-3 was significantly higher vs D4-7, while % decrease in LUSc was higher on D4-7 vs D0-3.

Conclusion(s): We demonstrate a statistically significant decrease in mean LUSc, and strong correlation with a reduction in FiO2 when exposed to furosemide. Higher LUSc on D0 predicts higher % decline in FiO2 by D7 of furosemide.

Table 1.
Patient characteristics and demographics

Figure 1.
Spearman’s rank-order correlation, rₛ(40) = 1.00, p = 0.0167

Figure 2.
Linear regression, F (1, 28) = 16.98 , p < 0.005, R2 = 0.38