546 - Association Between Gastrointestinal Adverse Events and Extracorporeal Membrane Oxygenation (ECMO) Utilization in Neonates: A Retrospective Cohort Study

Poster Number: 546
Publication Number: 546.2164

Background: Extracorporeal membrane oxygenation (ECMO) is a lifesaving intervention used in severe cases of cardiopulmonary insufficiencies in neonates. Despite its efficacy, ECMO has been associated with several adverse effects. However, there is not enough evidence that describes gastrointestinal adverse effect arising in neonates under ECMO treatment.

Objective: To examine the association of necrotizing enterocolitis (NEC) and/or gastrointestinal bleeding (GIB) with ECMO in newborn infants.

Design/Methods: we used the National Inpatient Sample (NIS) dataset produced by the healthcare cost and utilization project (HCUP) for the years 2016-2020, retrospective cohort analysis was preformed. We identified NEC using the International Classification of Diseases-Version 10 (ICD10) codes P77.1, P77.2, P77.3, P77.9, and P780. We identified GIB using the ICD10 codes P54.0, P54.1, P54.2, and P54.3. We identified ECMO using ICD10 code Z92.8 and procedure codes 5A1522F, 5A1522G, and 5A1522H. We only included infants of gestational age (GA) > 36 weeks and birth weight (BW) > 2500g. We excluded infants with severe anatomical congenital heart diseases, congenital diaphragmatic hernias, or abdominal wall defects. We calculated odds ratios using chi square test and used logistic regression models to calculate adjusted (aOR) while controlling for confounding variables.

Results:

Results: Weighted sample included 16,784,559 newborn infants. 0.02% diagnosed with NEC, 0.05% with GIB, and about 0.01% received ECMO. Among ECMO infants, NEC occurred in 2.9% in infants treated with ECMO compared to 0.02% in infants who did not receive ECMO, adjusted odds ratios (aOR) after controlling for confounding factors (including maternal hypertension, diabetes mellitus, or chorioamnionitis, placental abruption, breech or malpresentation, cord prolapse or nuchal cord, sex, race and small for gestational age status, central nervous system (CNS) and lung anomalies, chromosomal disorders, pulmonary hemorrhage or hypertension, sepsis or severe asphyxia) were 1.8 (95%CI: 1.1-2.9, p =0.02). GIB occurred in 4.8% in ECMO infants compared to 0.05% in non ECMO infants, aOR 3.8 (95%CI: 2.5-5.6, p < 0.001). Among ECMO infants, NEC was strongly associated with pulmonary hemorrhage, while, GIB was strongly associated with maternal chorioamnionitis.

Conclusion(s): Gastrointestinal adverse events are increased in newborn infants who received ECMO. GI bleeding occurs more frequently than NEC. GI bleeding is more common in ECMO infants who were born to mothers with chorioamnionitis. NEC is common in ECMO infants who developed pulmonary hemorrhage.