279 - Modelling of Routine Echocardiographic and Clinical Data for Early Prediction of BPD or Death in Very Preterm Infants

Sunday, April 30, 2023

3:30 PM – 6:00 PM ET

Poster Number: 279
Publication Number: 279.341

Roberto Chioma, The Irish Centre for Maternal and Child Health Research, Cork, Cork, Ireland; Daragh Finn, University College Cork, Cork, Cork, Ireland; Brian Walsh, University College Cork, Cork, Cork, Ireland; Deirdre A. O Sullivan, Children's Health Ireland at Crumlin, Dublin, Dublin, Ireland; Vicki Livingstone, UCC, Cork, Cork, Ireland; David Healy, University College Cork, Cork, Cork, Ireland; Eugene Dempsey, university college cork, Cork, Cork, Ireland

Presenting Author(s)

Roberto Chioma, MD (he/him/his)

Trainee
The Irish Centre for Maternal and Child Health Research
Cork, Cork, Ireland

Background:

Bronchopulmonary dysplasia (BPD) is a common complication of prematurity. Progressive respiratory deterioration in infants at high risk of BPD is correlated with PDA exposure. Early identification of these infants is important for early therapeutic decisions.

Objective: This study aimed to design an early predictive model for BPD or death in preterm infants.

Design/Methods:

A retrospective study of infants admitted to a tertiary NICU between January 2017 and November 2021 was conducted. Inclusion criteria were gestational age (GA) ≤ 29 weeks and/or birthweight (BW) < 1500 grams and echocardiography performed between 36 hours and 7 days of life. Exclusion criteria consisted of pulmonary hypertension and major congenital anomalies. The primary outcome was a composite of BPD and death.

Results:

The study included 101 infants. Median(IQR) GA was 27,6[24,6–29,6] weeks and BW was 900[675–1305] grams (Table 1). Multivariable logistic regression identified three independently associated variables with BPD/death (BW, Respiratory Severity Score, and flow pattern across the PDA) (Table 2 and 3). The multi-variable model had an area under the curve of 0,98 (95% CI 0,96-1, p < .001) for predicting BPD/death (Figure 1). Babies developing the primary outcome had a lower overall score than those who did not (-4,1 [-6,1 < ![if !msEquation] >< ![if !vml] >< ![endif] >< ![endif] > -0,6] vs 4,3 [3,2 – 6,5], P < .001) (Figure 2). A cut-off of 0 had positive and negative predictive values of 95% of 91%, and sensitivity and specificity of 95% and 90%, respectively.

Conclusion(s):

Our model allows early prediction of BPD/death. Its possible role in clinical decision-making requires further research.