Page 66 - 南京医科大学学报自然科学版

南京医科大学学报（自然科学版） 第43卷第10期
·1392 · Journal of Nanjing Medical University（Natural Sciences） 2023年10月

·临床研究·

神经重症患者术后颅内感染列线图模型的构建

支晓卉 ，徐修鹏 ，岳 震 2*
2
1
南京医科大学第一附属医院康复医学中心，神经外科，江苏 南京 210029
1 2

［摘 要］ 目的：建立预测神经重症患者术后颅内感染风险的列线图模型。方法：回顾性分析2018年1月—2021年1月南京
医科大学第一附属医院神经外科监护病房行开颅手术的200例患者的临床资料。按照7∶3的比例随机分为训练集（n=140）和
验证集（n=60）。采用单因素分析和多因素Logistic回归筛选神经重症患者术后发生颅内感染的危险因素并构建列线图预测模
型。通过绘制受试者工作特征（receiver operating characteristic，ROC）曲线及决策曲线分析（decision curve analysis，DCA）评价
模型的效能及临床净获益。结果：神经重症患者的原发病、脑室外引流时间、腰大池引流时间是术后发生颅内感染的危险因素
（P < 0.05）。绘制列线图模型的 ROC 曲线显示，训练集和验证集的曲线下面积（area under curve，AUC）分别为 0.774（95%CI：
0.695~0.853）、0.831（95%CI：0.725~0.936），DCA曲线显示颅内感染发生的预测可提高临床获益率。结论：基于神经重症术后
颅内感染的危险因素构建了列线图预测模型，有助于早期筛查神经重症术后颅内感染高危患者，利于早期诊治，改善患者预后。
［关键词］ 神经重症；颅内感染；列线图；预测模型
［中图分类号］ R619.3 ［文献标志码］ A ［文章编号］ 1007⁃4368（2023）10⁃1392⁃06
doi：10.7655/NYDXBNS20231009

Construction of nomogram model for postoperative intracranial infection in patients with

severe neurological disease
2
1
ZHI Xiaohui ，XU Xiupeng ，YUE Zhen 2*
2
1 Department of Rehabilitation，Department of Neurosurgery，the First Affiliated Hospital of Nanjing Medical
University，Nanjing 210029，China

［Abstract］ Objective：This study aims to establish a nomogram model for predicting the risk of postoperative intracranial infection
in patients with severe neurological diseases. Methods：A retrospective study was conducted on clinical data from 200 patients who
underwent surgical treatment in our hospital’s neurosurgical care unit between January 2018 and January 2021. The patients were
randomly divided into a training set（n=140）and a validation set（n=60）. Univariate analysis and multivariate logistic regression
analysis were used to screen the risk factors，constructing the prediction model by nomogram. The receiver operating characteristic
（ROC）curve was plotted to assess the predictive efficacy of the nomogram model for intracranial infection in patients with severe
neurological diseases. Additionally，validation of the model and evaluation of its clinical net benefit were performed using decision
curve analysis（DCA）. Results：The protopathy，external ventricular drainage time and lumbar cisterna drainage time were the risk
factors for postoperative intracranial infection（P < 0.05）. ROC curve of the nomogram model showed that the area under curve（AUC）
of the training set and the validation set were 0.774（95%CI：0.695~0.853）and 0.831（95%CI：0.725~0.936），respectively. DCA
curves showed that the prediction of intracranial infection could improve the clinical benefit rate. Conclusion：Our nomogram
prediction model based on risk factors associated with postoperative intracranial infection in patients with severe neurological diseases
offers an effective approach for early identification of high⁃risk individuals，facilitating prompt diagnosis and treatment while improving
the prognosis of patients.
［Key words］ severe neurological disease；intracranial infection；nomogram；prediction model
［J Nanjing Med Univ，2023，43（10）：1392⁃1397］

［基金项目］ 国家自然科学基金（82203767）
∗
通信作者（Corresponding author），E⁃mail：yzicu5336@njmu.edu.cn