Abstract
BACKGROUND: Whether portal venous gas (PVG) is a sign of severe neonatal necrotizing enterocolitis (NEC) and predicts poor prognosis remains uncertain. METHODS: Patients from two centres were randomly assigned to a training set or a validation set. A nomogram model for predicting severe NEC was developed on the basis of the independent risk factors selected by least absolute shrinkage and selection operator (LASSO) regression analysis and multivariate logistic regression analysis. The model was evaluated based on the area under the curve (AUC), calibration curve, and decision curve analysis (DCA). RESULTS: A total of 585 patients met the study criteria, and propensity score matching resulted in 141 matched pairs for further analysis. Patients with PVG had a greater risk of surgical intervention or death compared with patients without PVG. A prediction model for severe NEC was established based on PVG, invasive mechanical ventilation (IMV), serum platelet count (PLT) and pH <7.35 at the onset of NEC. The model had a moderate predictive value with an AUC > 0.8. The calibration curve and DCA suggested that the nomogram model had good performance for clinical application. CONCLUSION: A prediction nomogram model based on PVG and other risk factors can help physicians identify severe NEC early and develop reasonable treatment plans. IMPACT: PVG is an important and common imaging manifestation of NEC. Controversy exists regarding whether PVG is an indication for surgical intervention and predicts poor prognosis. Our study suggested that patients with PVG had a greater risk of surgical intervention or death compared with patients without PVG. PVG, IMV, PLT and pH <7.35 at the onset of NEC are independent risk factors for severe NEC. A prediction nomogram model based on PVG and other risk factors may help physicians identify severe NEC early and develop reasonable treatment plans.