Work

Analysis of risk factors and establishment of a predictive model for brain injury in neonates with surgical necrotizing enterocolitis: a retrospective study

Year: 2025

Type: article

Abstract: BACKGROUND: Neonatal necrotizing enterocolitis (NEC) is prevalent among preterm neonates and is associated with high morbidity and mortality. While surgical intervention remains essential for advanced NEC, postoperative neonates exhibit an elevated risk of brain injury. However, which surgery-related factors exacerbate neonatal brain damage remains insufficiently investigated. This study aimed to identify risk factors for brain injury in neonates with surgical NEC and establish a predictive model to facilitate early identification and intervention, which may ultimately decrease neurodevelopmental impairment rates. METHODS: test for nonparametric variables. Significant predictors (p < 0.05) were incorporated into multivariate logistic regression, with model performance validated by receiver operating characteristic (ROC) analysis. RESULTS: Multivariate analysis revealed six independent risk factors (all p < 0.05): lower gestational age (GA) (OR 0.85 [95% CI: 0.73-0.98]; p = 0.028), higher procalcitonin (PCT) levels at surgery (OR 1.03 [95% CI: 1.01-1.06]; p = 0.017), postoperative sepsis (OR 3.46 [95% CI: 1.36-8.76]; p = 0.009), transmural necrosis with perforation (OR 3.03 [95% CI: 1.18-7.78]; p = 0.021), longer diagnosis-to-surgery intervals (> 24 h) (OR 3.52 [95% CI: 1.23-10.08]; p = 0.019), and retention of the necrotic bowel (OR 4.88 [95% CI: 1.50-15.91]; p = 0.009). The ROC analysis revealed an area under the curve (AUC) of 0.86 for the prediction model. CONCLUSION: The incidence of brain injury in neonates with surgical NEC is independently associated with elevated PCT levels at surgery, the presence of sepsis, and the occurrence of transmural necrosis with perforation. Conversely, increased GA, early surgical recognition and intervention, and complete resection of the necrotic bowel may serve as potential protective factors against brain injury. The predictive model constructed on the basis of these findings demonstrated strong discriminative ability and we propose immediate neuroprotective intervention when model-predicted risk exceeds the 53% threshold.

Source: BMC Pediatrics

DOI: 10.1186/s12887-025-06324-x