Primary contributors to gastrostomy tube placement in infants with Congenital Diaphragmatic Hernia

Abstract

Objective: To identify factors associated with gastrostomy tube (GT) placement in infants with congenital diaphragmatic hernia (CDH).

Methods: Retrospective cohort study of 114 surviving infants with CDH at a single tertiary care neonatal intensive care unit from 2010-2019. Prenatal, perinatal and postnatal characteristics were compared between patients who were discharged home with and without a GT. Prenatal imaging was available for 50.9% of the cohort. Logistic regression was used to assess the association between GT placement and pertinent clinical factors. ROC curves were generated, and Youden's J statistic was used to determine optimal predictive cutoffs for continuous variables. Elastic net regularized regression was used to identify variables associated with GT placement in multivariable analysis.

Results: GT was placed in 43.9% of surviving infants with CDH. Prenatal variables predictive of GT placement were percent predicted lung volume (PPLV) <21%, total lung volume (TLV) <30 ml, lung-head ratio (LHR) <1.2 or observed to expected LHR (O/E LHR) <55%. Infants who required a GT were diagnosed earlier prenatally (23.6 ± 3.4 vs. 26.4 ± 5.6 weeks). Patients whose stomach was above the diaphragm on prenatal ultrasound (up) had a higher odds of GT placement compared to those with stomachs below the diaphragm (down) position by a factor of 2.9 (95% CI: 1.25, 7.1); p = 0.0154. Postnatally, infants with GT had lower Apgar scores at 1 and 5 min, longer lengths of stay and higher proportion of flap closures. Infants with a type C or D defect and extracorporeal membrane oxygenation (ECMO) were associated with increased odds of needing a GT. Postnatal association included being NPO for >12 days, need for transpyloric (TP) feeds for >10 days, >14 days to transition to a 30 min bolus feed, presence of gastro-esophageal reflux (GER), chronic lung disease and pulmonary hypertension. In multivariable analysis, duration of NPO, time to TP feeds, transition to 30 min bolus feeds remained significantly associated with GT placement after adjusting for severity of pulmonary hypertension (PH), GER diagnosis and sildenafil treatment.

Conclusion: Identification of risk factors associated with need for long-term feeding access may improve timing of GT placement and prevent prolonged hospitalization related to feeding issues.

Level of evidence rating: Level II (Retrospective Study).

Keywords: Congenital Diaphragmatic Hernia (CDH); Gastroesophageal reflux (GE Reflux); Gastrostomy Tube (g-tube); fetal MRI, antenatal ultrasound.