Gastrostomy Tube Placement in Neonates Undergoing Congenital Heart Surgery: A Novel Assessment for Improving Utilization and Timing of Placement.

Abstract

Neonates undergoing congenital heart defect repair require optimized nutritional support in the perioperative period. Utilization of a gastrostomy tube is not infrequent, yet optimal timing for placement is ill-defined. The objective of this study was to identify characteristics of patients whose postoperative course included gastrostomy tube placement to facilitate supplemental tube feeding following neonatal repair of congenital heart defects. A single-institution, retrospective chart review identified 64 consecutive neonates who underwent cardiac operations from 2012 to 2016. Perioperative variables were evaluated for significance in relation to gastrostomy tube placement. A total of 27 (42%) underwent gastrostomy tube placement. Diagnosis of a genetic syndrome was associated with the likelihood of placement of gastrostomy tube (P = .032), as were patients with single ventricle physiology (P = .0013) compared to those felt to be amenable to eventual biventricular repair. Aortic arch reconstruction (P = .029), as well as the need for delayed sternal closure (P = .05), was associated with increased frequency of gastrostomy tube placement. Postoperative outcomes including the number of days intubated (P = .0026) and the presence of significant dysphagia (P = .0034) were associated with gastrostomy placement. Additionally, genetic syndrome (P = .003), aortic arch reconstruction (P = .01), and postoperative intubation duration (P = .0024) correlated with increased length of stay, where increased length of stay was associated with gastrostomy tube placement (P = .0004). Patient characteristics that were associated with a high likelihood of eventual gastrostomy placement were identified in this study. Early recognition of such characteristics in future patients may allow for reduced time to gastrostomy tube placement, which in turn may improve perioperative growth and outcomes.