神经外科手术部位感染危险因素的logistic回归和分类树模型分析  

Analysis of risk factors for surgical site infections in neurosurgery department patients based on logistic regression and classification tree models

作　　者：巨睿 祝苗 王一洋 JU Rui;ZHU Miao;WANG Yiyang(Heze Municipal Hospital,Heze,Shandong 274000,China)

机构地区：[1]菏泽市立医院感染管理科,山东菏泽274000 [2]菏泽市立医院医院办公室,山东菏泽274000

出　　处：《中华医院感染学杂志》2025年第4期545-549,共5页Chinese Journal of Nosocomiology

基　　金：山东省卫健委卫生政策研究基金资助项目(WZY202337)。

摘　　要：目的 采用logistic回归和分类树模型归纳神经外科患者颅脑术后手术部位感染(SSI)的危险因素,为提出有效的SSI防控措施提供循证依据。方法 回顾性分析菏泽市立医院2020年3月-2022年6月神经外科1 259例手术患者的临床资料。根据手术记录、病程记录、临床经验等对可能引起手术部位感染的因素进行整理汇总,并通过logistic回归及分类树模型两种方法对两组患者的危险因素进行分析。结果 神经外科手术患者1 259例,根据手术部位感染情况分为感染组26例、非感染组1233例,手术部位感染率为2.07%。手术次数、手术风险(NNIS)分级、基础疾病及脑脊液漏患者SSI感染率,差异有统计学意义(P<0.05),logistic回归分析结果显示,NNIS评分(OR=2.247,95%CI:1.158~4.361,P=0.017)、基础疾病(OR=6.834,95%CI:2.586~18.058,P=0.001)是神经外科手术部位感染的危险因素。分类树模型分析,筛选出基础疾病、二次手术及NNIS分级为危险因素,基础疾病为首要因素。结论 NNIS评分、基础疾病、二次手术为发生手术部位感染的危险因素,应密切关注该类患者并采取有效措施。logistic回归和分类树模型均适合于手术部位危险因素的判断,logistic回归模型更适用于寻找危险因素,而分类树模型可明确因素间的交互作用,判断更为直观。OBJECTIVE To summarize the risk factors for surgical site infections(SSI)after craniosurgery in neu-rosurgical patients based on logistic regression and classification tree modeling,and to provide evidence for propo-sing effective SSI prevention and control measures.METHODS The clinical data of 1259 Neurosurgical patients in Heze Municipal Hospital from Mar.2020 to Jun.2022 were retrospectively analyzed.Factors that may cause SSI were collected based on surgical and medical records and clinical experience.The risk factors in the pre-,intra-and postoperative periods were compared between the infected and the non-infected groups and analyzed by logistic re-gression and classification tree modeling.RESULTS Totally 1259 cases of neurosurgical patients were divided into the infected group with 26 cases and the non-infected groups with 1233 cases based on whether SSI occurred,with SSI rate of 2.07%.There were statistically significant differences in the number of surgeries,SSI with NNIS(Na-tional nosocomial infection surveillance)risk index,underlying disease,and cerebrospinal fluid leakage between the two groups(P<0.05).Logistic regression analysis revealed that the NNIS score(OR=2.247,95%CI:1.158 to 4.361,P=0.017)and underlying disease(OR=6.834,95%CI:2.586 to 18.058,P=0.001)were risk fac-tors for neurosurgical site infection.The classification tree model identified underlying diseases,reoperation and NNIS grading as risk factors,and underlying disease was the primary one.CONCLUSIONS NNIS score,underly-ing diseases and reoperation are risk factors for SSI,and patients with such factors should be taken close attention and effective measures.Both logistic regression and classification tree model are suitable for evaluating the risk fac-tors of SSI;however,the logistic regression model is more suitable for identifying risk factors,while the classifi-cation tree model can clarify the interaction between factors and provide more intuitive assessment.

关 键 词：手术部位感染 神经外科手术 危险因素 分类树 

分 类 号：R619.3[医药卫生—外科学]