PubMed2003
BACKGROUND/PURPOSE: Localized intestinal perforation (LP) is thought to be a distinct entity when compared with perforation associated with necrotizing enterocolitis (NEC). Studies have indicated that LP is more amenable to percutaneous drainage and associated with a better outcome. We sought to determine whether LP and NEC could be distinguished based on clinical parameters alone. METHODS: A retrospective review of 40 neonates with gastrointestinal perforations between January 1990 and May 1998 was performed. All had undergone laparotomy and had histologic specimens available for evaluation. RESULTS: Twenty-one neonates had necrotizing enterocolitis (NEC), and 19 had localized perforation (LP) based on histologic criteria. More neonates with LP were exposed to prenatal indomethacin (37% v 5%; P <.05), received intravenous dexamethasone (42% v 10%; P <.05), had umbilical artery catheters (63% v 14%; P <.05), and had a higher white blood cell (WBC) count (27.1 +/- 23.1 v 14.3 +/- 11.5; P <.05). More neonates with NEC had pneumatosis intestinalis (47% v 11%; P <.05). No significant differences existed in enteral feeding (16% LP v 38% NEC) or overall mortality rate (37% LP v 38% NEC). No statistical differences in the timing of perforation or clinical presentation were found. CONCLUSIONS: NEC and LP are difficult to distinguish based on clinical parameters alone. The authors did find associations between LP and prenatal indomethacin, intravenous dexamethasone, umbilical artery catheters, and a higher WBC count. Mortality rate and clinical outcome were nearly identical in both groups. Pneumatosis intestinalis, thought to be pathognomonic for NEC, was seen on abdominal radiograph in 2 babies with histologically proven LP.