摘要
目的探讨颅脑术后并发颅内感染的独立危险因素,并构建列线图预测模型,为临床提供指导。方法回顾性分析2017年5月至2022年4月在中山市人民医院接受颅脑手术的419例患者的临床资料。参考我国2021版的《神经外科中枢神经系统感染诊治中国专家共识》诊断颅内感染,多因素logistic回归分析筛选颅脑术后颅内感染的独立危险因素,并构建列线图预测模型,借助shinyapps.io软件制作网页版动态列线图。结果419例中,53例颅脑术后发生颅内感染,发生率为12.6%。多因素logistic回归分析显示手术持续时间长、术中出血量大、术后脑脊液漏、手术次数≥2次、频繁通过引流管留取脑脊液或注射药物是颅脑术后并发颅内感染的独立危险因素(P<0.05)。基于这5个独立危险因素成功建立颅脑术后颅内感染列线图预测模型,ROC曲线评估显示模型具有较好的区分度[曲线下面积为0.826(95%CI0.768~0.884)],Hosmer-Lemeshow拟合优度检验显示列线图模型的预测曲线与理想曲线基本重合,决策曲线显示列线图模型的临床实用价值很高。借助R软件绘制动态列线图,并通过shinyapps.io软件制作网页版动态列线图,网址:https://monige.shinyapps.io/DynNomapp/。结论手术时间长、术中出血量大、术后脑脊液漏、手术次数≥2次、频繁通过引流管留取脑脊液或注射药物是颅脑术后颅内感染的独立危险因素,基于这些独立危险因素构建的列线图预测模型具有良好的预测性能,开发的网页版动态列线图操作方便、计算结果精确,可应用于临床实践。
Objective To explore the independent risk factors for postoperative intracranial infection following craniotomy and to construct a nomogram prediction model for clinical guidance.Methods The clinical data of 419 patients who underwent craniotomy at Zhongshan City People's Hospita from May 2017 to April 2022 were retrospectively analyzed.Postoperative intracranial infection was diagnosed based on the 2021 Chinese Expert Consensus on Diagnosis and Treatment of Central Nervous System Infections in Neurosurgery.Multivariate logistic regression analysis was used to identify independent risk factors for postoperative intracranial infection,and a nomogram prediction model was constructed.A web-based dynamic nomogram was developed using the shinyapps.io software.Results Among the 419 patients,53 developed postoperative intracranial infection,with an incidence rate of 12.6%.Multivariate logistic regression analysis showed that prolonged surgery duration,large intraoperative blood loss,postoperative cerebrospinal fluid leakage,≥2 surgical operations,and frequent cerebrospinal fluid sampling or drug administration via drainage tubes were independent risk factors for postoperative intracranial infection(P<0.05).Based on these five independent risk factors,a nomogram prediction model for postoperative intracranial infection was successfully established.ROC curve analysis indicated good discrimination of the model[area under the curve:0.826(95%CI 0.768~0.884)].The Hosmer-Lemeshow goodness-of-fit test showed that the predicted curve of the nomogram model largely overlapped with the ideal curve.Decision curve analysis demonstrated high clinical utility of the nomogram model.A dynamic nomogram was developed using R software and transformed into a web-based version via shinyapps.io,accessible at:https://monige.shinyapps.io/DynNomapp/.Conclusion Prolonged surgery duration,large intraoperative blood loss,postoperative cerebrospinal fluid leakage,≥2 surgical operations,and frequent cerebrospinal fluid sampling or drug administration via drainage tubes are independent risk factors for postoperative intracranial infection.The nomogram prediction model based on these risk factors exhibits good predictive performance.The developed web-based dynamic nomogram is user-friendly,provides accurate calculations,and is applicable in clinical practice.
作者
区廷枫
欧结艳
李亮明
林其昌
林伟标
QU Tingfeng;OU Jie Yan;LI Liangming;LIN Qichang;LIN Weibiao(Department of Neurosurgery,Zhongshan City People’s Hospital,Zhongshan 528403,China;Brain Center,Zhongshan City People’s Hospital,Zhongshan 528403,China)
出处
《中国临床神经外科杂志》
2025年第10期607-612,共6页
Chinese Journal of Clinical Neurosurgery
基金
2024年度中山市第一批社会公益与基础专项项目(2024B1061)。
关键词
颅脑手术
颅内感染
危险因素
列线图
网页版动态列线图
craniotomy
postoperative intracranial infection
risk factors
nomogram
web-based dynamic nomogram
