保健医学研究与实践2025，Vol.22Issue(6)：99-106,8.DOI:10.11986/j.issn.1673-873X.2025.06.15
颅脑损伤儿童入住ICU的危险因素分析及风险预测模型构建
Risk factors for pediatric traumatic brain injury admission to the ICU and construction of a risk prediction nomogram
摘要
Abstract
Objective To investigate the risk factors for intensive care unit(ICU)admission among children with traumatic brain injury(TBI)and to construct a nomogram risk prediction model.Methods This retrospective study included 351 ped-iatric TBI patients treated at Northwest Women's and Children's Hospital from January 2019 to August 2023 who met inclu-sion and exclusion criteria.Patients were randomly assigned to a modeling group(n=246)and a validation group(n=105)at a 7∶3 ratio.Clinical data,including TBI severity and pupillary light reflex,were collected.Multivariable logistic regres-sion was used to identify factors influencing ICU admission.A nomogram predicting the risk of ICU admission was devel-oped using R 4.3.1,and its discrimination,calibration,and clinical utility were evaluated by receiver operating characteris-tic(ROC)curve,calibration curve,and decision curve analysis(DCA).Results There were no significant differences in baseline characteristics such as age and sex between the modeling and validation groups(P＞0.05).When patients in the modeling group were grouped by ward versus ICU admission,significant differences were observed between groups in TBI severity,presence of brain herniation,pupillary light reflex status,receipt of surgical treatment,and need for mechanical ventilation(P＜0.05).Multivariable logistic regression identified the following independent risk factors for ICU admission:severe TBI(OR=7.176,95％CI:2.586-19.912),brain herniation(OR=4.559,95％CI:1.715-12.119),sluggish pupil-lary light reflex(OR=4.090,95％CI:1.307-12.798)or absent pupillary light reflex(OR=7.771,95％CI:2.846-21.222),surgical treatment(OR=6.580,95％CI:2.620-16.525),and mechanical ventilation(OR=6.667,95％CI:2.392-18.586;all P＜0.05).The constructed nomogram based on the above factors indicates that when the total score for these five factors reaches 232 points,the probability of ICU admission is 0.7,and when the total score reaches 297 points,the probability is 0.9.The ROC area under the curve(AUC)was 0.818(95％CI:0.740-0.896)for the modeling group and 0.856(95％CI:0.748-0.964)for the validation group.Hosmer-Lemeshow goodness-of-fit test results were x2=6.694,P=0.350 for the modeling group and x2=4.070,P=0.539 for the validation group;calibration slopes for both groups were close to 1.DCA showed that the model provided a high net clinical benefit when the threshold probability ranged from 0.06 to 0.89 in the modeling group and from 0.06 to 0.75 in the validation group.Conclusion TBI severity,brain herniation,pu-pillary light reflex status,surgical treatment,and mechanical ventilation are influencing factors of ICU admission in pediatric TBI patients.The nomogram developed on the basis of these factors demonstrates good predictive performance.关键词
颅脑损伤/重症监护室/危险因素/列线图Key words
Traumatic brain injury/Intensive care unit/Risk factors/Nomogram分类
医药卫生引用本文复制引用
余慧,魏曦,党晴,刘艳,袁丹阳,郭肖燕,孙红格..颅脑损伤儿童入住ICU的危险因素分析及风险预测模型构建[J].保健医学研究与实践,2025,22(6):99-106,8.基金项目
陕西省重点研发计划项目(2022SF-504). （2022SF-504）
