Analysis of Risk Factors for Brain Injury in Infants With Small Gestational Age and Construction and Evaluation of Prediction Model.

Abstract

Aims: To evaluate brain injury risk factors in small for gestational age (SGA) infants, develop and validate a risk prediction model.

Methods: Medical records from 326 SGA infants in the Neonatal Intensive Care Unit at Hebei University's Affiliated Hospital from September 2019 to September 2024 were reviewed. The infants were categorised into brain injury (n = 135) and no brain injury (n = 191) groups based on cranial MRI outcomes. Risk factors for brain injuries in SGA infants were identified through both univariate and multivariate logistic regression analyses, followed by constructing a nomogram to visualise these risk determinants. The predictive model's calibration, discrimination, clinical net benefit, and clinical usefulness were verified using calibration and ROC curves, along with decision curve analysis (DCA) and clinical impact curves (CIC).

Results: Significant risk factors included advanced maternal age, gestational hypertension (GH), fetal distress, reduced gestational age, neonatal septicemia, abnormal platelet counts and elevated neutrophil-to-lymphocyte ratio (NLR). The calibration curve and Hosmer-Lemeshow test verified the model's high accuracy. The model demonstrated good discrimination with an ROC curve AUC of 87.4%. Both DCA and CIC evaluations indicated the model's high clinical utility.

Conclusion: The developed multivariate logistic regression model effectively predicts the risk of craniocerebral injuries in SGA infants, serving as a valuable tool for early identification of at-risk neonates.