391 - Association of Time-to-Liberation from Respiratory Support with Gastrostomy Tube Placement in Infants with Bronchopulmonary Dysplasia

Publication Number: 1375.391

Gozdem Kayki, Nationwide Children's Hospital, Columbus, OH, United States; Khushboo Gupta, Nationwide Children's Hospital, Dublin, OH, United States; Sara Conroy, Nationwide Children's Hospital, Columbus, OH, United States; Audrey Miller, Nationwide Children's Hospital, Columbus, OH, United States; Leif D. Nelin, Nationwide Children's Hospital, Columbus, OH, United States; Edward Shepherd, Nationwide Children's Hospital, Columbus, OH, United States; Matthew Kielt, Nationwide Children's Hospital, Columbus, OH, United States

Background: Infants with bronchopulmonary dysplasia (BPD) have delayed achievement of full oral feeding and a higher incidence for gastrostomy tube (GT) feeding as compared to infants without BPD. Prolonged duration of mechanical ventilation and non-invasive positive airway pressure, known risk factors for feeding difficulties, may account for these adverse feeding outcomes by delaying initiation of oral feeding attempts. However, the relationship between the time-to-liberate from respiratory support (LRS) and GT placement remains unknown.

Objective: To estimate the association between the time-to-LRS and GT placement in infants with grade 2/3 BPD.

Design/Methods: This is a retrospective study of infants admitted to the Nationwide Children's Hospital Comprehensive Center for Bronchopulmonary Dysplasia with a diagnosis of grade 2/3 BPD between January 1, 2015 and September 1, 2025. Time-to-LRS was defined as the postmenstrual age (PMA) when infants weaned to nasal cannula < 2L/min or room air. Time-to-LRS was compared between subjects who did and did not undergo GT placement using cumulative incidence plots and the Log-rank test. The association between PMA at LRS and GT placement was estimated using multivariable regression analysis adjusting for gestational age, small-for-gestational age status, any treatment with surfactant, and grade 3 BPD.

Results: Among 638 included infants, 188 (29%) underwent GT placement. Demographic characteristics are presented in Table 1. Birth weight, small-for-gestational age status, treatment with surfactant, and rates of grade 3 BPD differed between groups, as did PMA at LRS (Table 1). PMA at LRS was later in the patients who underwent GT, P< 0.01 by Log-rank test (Figure). In adjusted analyses, the probability of GT placement increased by 3% for each one-week increase in PMA at time-to-LRS (Table 2).

Conclusion(s): In this study, we observed that the probability of GT placement was higher in infants with grade 2/3 BPD who achieved LRS at older PMAs. Future studies are needed to determine therapeutic practices that accelerate time-to-LRS and improve feeding outcomes in infants with grade 2/3 BPD.

Table 1
Demographics and clinical characteristics of cohort

Table 2
Model Results estimating association of post-menstrual age (PMA) at liberation of respiratory support with gastrostomy tube (GT) placement

Figure
Cumulative incidence of post-menstrual age (PMA) at time of liberation of respiratory support (LRS) stratified by gastrostomy tube (GT) placement. Subjects who underwent GT placement achieved LRS at a later PMA as compared to infants who did not undergo GT placement (p < 0.01 by log-rank test)