O2089: Contemporary Operative Management of Necrotizing Enterocolitis in U.S. Children's Hospitals

Background: Various aspects of the surgical and perioperative management of necrotizing enterocolitis (NEC), including expedient resection, selective use of primary intestinal anastomoses, earlier enterostomy takedown and limited postoperative antibiotics, may improve nutritional and neurodevelopmental outcomes. We sought to evaluate the presentations, operative interventions and outcomes of a large population of infants who underwent intestinal resection for NEC and examine potential opportunities for practice improvement.

Methods: The Pediatric Health Information System (PHIS) database was used to identify infants with Bell Stage 3 NEC who underwent an intestinal resection with or without enterostomy for acute disease between July, 2016 and June, 2021. Demographics, operative approach, antibiotic management, duration of parenteral nutrition, timing of enterostomy takedown and survival were evaluated.

Results: 241 infants at 40 children’s hospitals underwent an intestinal resection (small bowel in 144 (60%) and small bowel and colon in 65 (27%)). In 190 (79%) cases this was the first or only operation. In 27 (11.2%) patients a drain was placed prior to a subsequent resection and 5 (18.5%) of these died. The average age at first operation was 36 days and did not vary by gestational age. (Figure) Postoperative antibiotics most frequently included Vancomycin (49.7%), Piperacillin/Tazobactam (46.6%) and Metronidazole (35%) and continued for a median of 13 days. Full feedings were achieved at a mean of 48 days. A primary intestinal anastomosis was used in 25 (13.3%) patients. 143 (75%) stomas were closed after a mean of 103 days and 37% were closed during the initial admission. 40 (16.6%) infants died at a mean of 48 days after the initial operation. Among infants who survived at least 4 days and had a resection within 10 days of their initial operation, those who underwent resection without an enterostomy were relatively similar in terms of demographics and preoperative vasopressor to those managed with a stoma. Both groups experienced similar mortality and length of stay. (Table)

Conclusion: We found that mortality after resection for NEC was at the lower extent of the 15-40% range previously reported and, in contrast to studies of NEC in general, age at operation did not have an inverse relationship to gestational age. The duration of antibiotic therapy exceeded 2 weeks in half of cases, longer than what has been recommended in prior studies. Most infants did not have intestinal continuity restored during their index admission and the average time to stoma closure was several months longer than what has been shown to be equally safe. Despite evidence that a resection without an enterostomy is safe in many cases, most infants underwent enterostomy. Although patients who underwent a primary anastomosis were similar to those who received stomas, outcomes were equivalent, particularly with staged anastomoses after initial hemodynamic instability.