【摘要】 目的通过分析神经外科患者术后颅内感染的相关危险因素,建立个体化预测颅内感染风险的列线图模型。方法纳入2014年1月-2018年6月于我院神经外科行开颅手术治疗患者386例,收集临床资料,采用多因素Logistic回归模型分析颅内感染的独立危险因素;应用R软件建立预测颅内感染风险的列线图模型,并进行验证。结果 (1)386例神经外科手术患者有25例发生颅内感染,发生率为6.5%。(2)2型糖尿病(OR=1.353,95%CI:1.141～1.603)、幕下手术(OR=2.452,95%CI:1.107～5.435)、手术持续时间≥4 h(OR=1.125,95%CI:1.076～1.177)、脑室外引流(OR=1.125,95%CI:1.076～1.177)及脑脊液漏(OR=2.022,95%CI:1.28～3.193)均是颅内感染的独立危险因素。(3)对列线图模型进行验证,ROC曲线显示该模型预测神经外科患者术后颅内感染风险的曲线下面积为0.849(95%CI:0.740～0.958);校准曲线为斜率接近于1的直线,Hosmer-Lemeshow拟合优度检验(χ~2=9.068,P=0.106)均显示该模型预测神经外科患者术后颅内感染风险具有良好准确度。结论本研究基于2型糖尿病、幕下手术、手术持续时间、引流方式及脑脊液漏这5项神经外科患者术后颅内感染的独立危险因素,构建的预测神经外科患者术后颅内感染风险的列线图模型,具有良好的区分度与准确度,可指导临床个体化防控神经外科患者术后颅内感染,临床应用价值高。更多还原

【Abstract】 Objective To analyze the risk factors for post-operative intracranial infection in patients undergoing neurosurgery,and to establish a lnomogram model to predict the risk of intracranial infection.Methods From January 2014 to June 2018,386 patients undergoing neurosurgery at Liaocheng people’s hospital were enrolled in this study.Their clinical data were analyzed and multivariate regression analysis was used to determine the risk factors of intracranial infection.A nomogram was developed by R software and validated to predict the risk of intracranial infection.Results A total of 28 cases out of 386 patients undergoing neurosurgery had intracranial infection.The incidence rate was 6.5%.History of type 2 diabetes(OR=1.353,95% CI:1.141～1.603),subsurgical surgery(OR=2.452,95% CI:1.107～5.435),operative time ≥ 4 h(OR=1.125,95% CI:1.076～ 1.177),extra-ventricular drainage(OR=1.125,95% CI:1.076～1.177) and cerebrospinal fluid leakage(OR=2.022,95% CI:1.280～3.193) were all independent risk factors of intracranial infection.ROC curve revealed that the model predicting intracranial infection in patients undergoing neurosurgery was with an area under the curve 0.849(95% CI:0.740～0.958).The slope of the calibration plot was almost one and the model passed Hosmer-Lemeshow goodness of fit test(χ~2=9.068,P=0.106) which demonstrated that the model was with good accuracy.Conclusions The nomogram built based on the history of type 2 diabetes,surgical procedure,operation time,drainage and cerebrospinal fluid leakage has good discrimination and accuracy which could be used for predicting individual risk of intracranial infection in patients undergoing neurosurgery,with potentially high clinical application value.更多还原