Risk Factors Analysis and Construction of a Predictive Scoring System for Poor Long-term Neurological Prognosis of Neonatal Hypoxic Ischemic Encephalopathy Based on Three-year Follow-up Data

LU Yanhui; LIU Zhenkui; LI Shiyang; WANG Yanfei; LI Chunmei; YUAN Erwei; XING Jing; GUO Weiping

doi:10.3639/j.issn.2095-3097.2025.02.017

Abstract：Objective To investigate the risk factors for poor long-term neurological outcomes in neonates with hypoxic-ischemic encephalopathy(HIE)and to construct a predictive scoring system.Methods A total of 277 neonates with HIE treated in the First Affiliated Hospital of Hebei North University between January 2018 and April 2020 were included in this study and followed up for three years.Based on long-term neurological outcomes,they were divided into the good prognosis group(n=221)and poor prognosis group(n=56).Logistic regression analysis was used to evaluate baseline data,laboratory parameters,and imaging findings between the two groups to develop a logistic predictive model.The variance inflation factor(VIF)of the model was calculated,and the receiver operating characteristic(ROC)curve was plotted to assess the model's discriminatory ability.The Hosmer-Lemeshow goodness-of-fit test was used to evaluate the model's calibration.Decision tree analysis was employed to analyze the model's net benefit.A predictive scoring scale for poor long-term neurological outcomes in neonates with HIE was constructed based on the Framingham risk assessment model and subsequently validated for its performance.Results Logistic regression analysis identified that the time to treatment initiation(OR=1.108,95％CI:1.015-1.211),neutrophil-to-lymphocyte ratio(NLR;OR=1.065,95％CI:1.020-1.112),neuron-specific enolase(NSE;OR=1.060,95％CI:1.005-1.117),diffuse-weighted imaging(DWI;OR=1.013,95％CI:1.003-1.023),resistive index(RI;OR=1.269,95％CI=1.017-1.583),placental abnormalities(OR=2.781,95％CI:1.643-4.729),hyperlactatemia(OR=2.563,95％CI:1.487-4.425),and the absence of hypothermia therapy(OR=3.980,95％CI:2.356-6.743)were independent risk factors for poor long-term neurological outcomes in neonates with HIE.The 5-minute Apgar score(OR=0.864,95％CI:0.750-0.995)was identified as a protective factor.The predictive model showed a sensitivity of 95.67％,a specificity of 89.29％,an accuracy of 94.43％,and an area under the curve(AUC)of 0.910(95％CI:0.722-0.987).The Hosmer-Leme showed goodness-of-fit test indicated good calibration(x2=9.420,P=0.371).A predictive scoring scale was developed with a total score of 39 points,categorized into three risk levels:low-risk ≤14 points,medium-risk 14＜score ≤17 points,and high-risk＞17 points.External validation of the scoring scale showed a sensitivity of 81.82％,specificity of 87.18％,accuracy of 86.00％,and AUC of 0.813(95％CI:0.732-0.886).Conclusion Time to treatment initiation,NLR,NSE,DWI,RI,placental abnormalities,hyperlactatemia,and the absence of hypothermia therapy were identified as independent risk factors for poor long-term neurological outcomes in neonates with HIE,while the 5-minute Apgar score was a protective factor.The constructed scoring scale demonstrated a certain predictive effect on long-term neurological outcomes.