Percutaneous Endoscopic Gastrostomy After Cardiothoracic Surgery in Children Less Than 2 Months Old: An Assessment of Long-Term Malnutrition Status and Gastrostomy Outcomes.

Objectives: Infants with critical congenital heart disease undergoing cardiothoracic surgery commonly experience chronic malnutrition and growth failure. We sought to determine whether placement of a percutaneous endoscopic gastrostomy was associated with reduced moderate-severe malnutrition status and to describe percutaneous endoscopic gastrostomy-related clinical and safety outcomes in this population.

Design: Single-center, retrospective cohort study.

Setting: Two hundred fifty-nine-bed, tertiary care, pediatric referral center.

Patients: Children with congenital heart disease less than 2 months old undergoing cardiothoracic surgery from 2007 to 2013 with and without percutaneous endoscopic gastrostomy.

Interventions: None.

Measurements And Main Results: Primary outcomes were weight for age z scores during hospitalization, at 6 months, and 1 year after cardiothoracic surgery. Secondary outcomes were frequency of percutaneous endoscopic gastrostomy revision, percutaneous endoscopic gastrostomy complications, and mortality. Statistical analyses included Wilcoxon rank-sum, Fisher exact, and Student t tests. Two hundred twenty-two subjects met study criteria, and 77 (35%) had percutaneous endoscopic gastrostomy placed at a mean of 45 ± 31 days after cardiothoracic surgery. No differences were noted for demographics, comorbidities, and weight for age z score at birth and at the time of cardiothoracic surgery. The percutaneous endoscopic gastrostomy cohort had greater Society of Thoracic Surgeons-European Association for Cardio-Thoracic Surgery risk category (4 [4-5] vs 4 [2-4]) and length of stay (71 d [49-101 d] vs 26 d [15-42 d]). Mean weight for age z score at the time of percutaneous endoscopic gastrostomy was -2.8 ± 1.3. Frequency of moderate-severe malnutrition (weight for age z score, ≤ -2) was greater in children with percutaneous endoscopic gastrostomy at discharge (78% vs 48%), 6 months (61% vs 16%), and 1 year (41% vs 2%). Index mortality was lower in children with percutaneous endoscopic gastrostomy at 30 days (8% vs 0%) and hospital discharge (19% vs 4%). However, no mortality differences were observed after discharge. Growth velocity after percutaneous endoscopic gastrostomy was greater (44 ± 19 vs 10 ± 9 g/d). Children tolerated percutaneous endoscopic gastrostomy without hemodynamic compromise, minor percutaneous endoscopic gastrostomy complications, and anticipated percutaneous endoscopic gastrostomy revisions. Children without mortality had percutaneous endoscopic gastrostomy removal at a median duration of 253 days (133-545 d). Children with univentricular physiology had improved in-hospital mean growth velocity (6.3 vs 24.4 g/d; p < 0.01) and reduced 1-year rate moderate-severe malnutrition (66.7% vs 36.9%; p < 0.01) after percutaneous endoscopic gastrostomy placement.

Conclusions: Percutaneous endoscopic gastrostomy placement was well tolerated and associated with improved postoperative growth velocity in children with critical congenital heart disease undergoing cardiothoracic surgery less than 2 months old. These findings were also noted in our subanalysis of children with univentricular physiology. Persistent rates of moderate-severe malnutrition were noted at 1-year follow-up. Although potential index mortality benefit was observed, definitive data are still needed.