脑肿瘤术后颅内感染危险因素及风险预测模型构建

Risk factors of postoperative intracranial infection in intracranial tumor patients and establishment of risk prediction model

目的构建脑肿瘤患者术后颅内感染的风险预测模型,为干预和防控脑手术后颅内感染提供参考。方法以某医院脑肿瘤术后患者为研究对象,通过多因素logistic回归分析构建颅内感染风险预测模型,评估风险预测模型的拟合优度和效能。结果该医院脑肿瘤术后患者颅内感染发生率为9.01%,感染病原菌主要为革兰阳性菌。多因素logistic回归分析结果显示,手术时长≥4 h、脑室引流时长≥4 d和入ICU等因素为脑肿瘤术后颅内感染的独立危险因素。脑肿瘤术后患者颅内感染风险预测模型拟合度良好,总体准确率为92.96%;其预测效能的ROC曲线下面积为0.844(95%CI值为0.759~0.928),灵敏度为0.75,特异度为0.83。结论脑肿瘤患者术后颅内感染发生率较高,建立的颅内感染风险预测模型总体辨别能力较好,具有良好辅助诊断价值。

Objective To construct a risk prediction model of intracranial infection in patients with brain tumor after surgery,and to provide a basis for intervention and prevention of intracranial infection after brain surgery.Methods The postoperative patients with brain tumor in a hospital were selected as the research objects.A multivariate logistic regression analysis was used to construct a risk prediction model,and the fit and efficacy of the risk prediction model were evaluated.Results The incidence rate of intracranial infection in postoperative patients with brain tumors was 9.01%,and the main pathogens of infection was Gram-positive bacteria.Multivariate logistic regression analysis showed that surgery duration≥4 h,ventricular drainage duration≥4 d,and admission to the ICU were independent risk factors for intracranial infection after brain tumor surgery.The risk prediction model for intracranial infection in patients after brain tumor surgery fitted well,and the overall accuracy of the model was 92.96%.The area under the ROC curve for predicting performance was 0.844(95%CI 0.759-0.928),with a sensitivity of 0.75 and a specificity of 0.83.Conclusion The infection is relatively high in patients with brain tumors in clinical setting.The risk prediction model for intracranial infection established in this study has the high ability to identify patients with high risk of postoperative intracranial infection with good diagnostic value.