A nomogram model for early recognition of acute necrotizing encephalopathy: a retrospective study.

BACKGROUND

Acute necrotizing encephalopathy (ANE) is an infection-associated encephalopathy characterized by acute onset and poor prognosis. Currently, there is a lack of effective prediction models for its early identification. To address this gap, we developed and validated a nomogram model for the early identification of ANE in pediatric patients.

METHODS

This retrospective study analyzed children presenting with seizures or consciousness disorders following coronavirus disease 2019 (COVID-19) or influenza infection at Children's Hospital of Soochow University between November 2021 and February 2024. The single-center patient data were stratified into ANE/non-ANE groups, and then randomly divided into training and validation cohorts at a 7:3 ratio for internal validation. Logistic and least absolute shrinkage and selection operator (LASSO) regression were used to identify risk factors for ANE, and a nomogram was subsequently developed. The model's performance was validated using the area under the curve (AUC) and the area under the precision-recall curve (AUPRC), and further validated through calibration curve and decision curve analysis (DCA).

RESULTS

Altered consciousness, aspartate aminotransferase (AST), lactate dehydrogenase (LDH), and blood urea nitrogen (BUN) were independently associated with ANE development. A nomogram model incorporating these four factors demonstrated high predictive performance, with AUC values of 0.961 and 0.978, and AUPRC values of 0.792 and 0.809 in the training and validation cohorts, respectively. Calibration curves, DCA, and clinical impact curves further indicated the model's accuracy in predicting ANE onset.

CONCLUSIONS

Our nomogram model provides a novel tool for the early identification of children at high risk of ANE, assisting clinicians in formulating timely interventions to improve patient prognosis.