重症脑室出血患者术后血流感染的危险因素分析及预测模型构建

Analysis of Risk Factors and Construction of a Predictive Model for Postoperative Bloodstream Infection in Patients with Severe Intraventricular Hemorrhage

摘要: 目的：分析重症脑室出血(sIVH)患者术后血流感染(BSI)的危险因素，构建风险预测模型，为临床早期识别与干预提供参考。方法：回顾性分析2017年6月~2025年9月于杭州师范大学附属医院接受脑室外引流(EVD)联合有创颅内压监测治疗的137例sIVH患者的临床资料。依据住院期间是否发生血流感染分为感染组(n = 20)和非感染组(n = 117)，采用单因素及多因素Logistic回归分析筛选独立危险因素，构建预测模型并进行内部验证。结果：137例sIVH患者中血流感染发生率为14.6%。多因素Logistic回归分析显示，术前血糖升高(OR = 1.22, 95% CI: 1.01~1.47, P = 0.040)及红细胞计数降低(OR = 0.37, 95% CI: 0.14~0.99, P = 0.049)是血流感染的独立危险因素。为探索不同维度指标的预测价值，分别构建了基于影像学指标的“核心影像模型”(术后2~5 d ONSD联合术前血糖)和基于常规实验室指标的“实验室模型”(红细胞计数联合术前血糖)。ROC曲线分析显示，核心影像模型的AUC为0.73 (95% CI: 0.62~0.83)，校准曲线显示模型校准度良好。结论：术前高血糖及红细胞计数降低是sIVH患者术后血流感染的独立危险因素。基于术后ONSD与术前血糖构建的预测模型对血流感染具有中等强度的风险评估价值，初步显示可为临床早期识别高危患者提供参考，但其效能有待多中心、前瞻性研究进一步验证。

Abstract: Objective: To analyze the risk factors for postoperative bloodstream infection (BSI) in patients with severe intraventricular hemorrhage (sIVH), and to establish a risk prediction model for early clinical identification and intervention. Methods: Clinical data of 137 sIVH patients who underwent external ventricular drainage (EVD) combined with invasive intracranial pressure monitoring at the Affiliated Hospital of Hangzhou Normal University from June 2017 to September 2025 were retrospectively analyzed. Patients were divided into BSI group (n = 20) and non-BSI group (n = 117) based on the occurrence of BSI during hospitalization. Univariate and multivariate logistic regression analyses were performed to identify independent risk factors. A predictive model was constructed and internally validated. Results: The incidence of BSI was 14.6%. Multivariate Logistic regression analysis showed that elevated preoperative blood glucose (OR = 1.22, 95% CI: 1.01~1.47, P = 0.040) and decreased red blood cell count (OR = 0.37, 95% CI: 0.14~0.99, P = 0.049) were independent risk factors for BSI. To explore the predictive value of indicators across different dimensions, two models were developed: a “core imaging model” based on imaging parameters (ONSD at 2~5 days postoperatively combined with preoperative blood glucose levels) and a “lab model” based on conventional laboratory parameters (erythrocyte count combined with preoperative blood glucose levels). ROC curve analysis demonstrated an AUC of 0.73 (95% CI: 0.62~0.83) for the core imaging model, indicating good model calibration. Conclusion: Preoperative hyperglycemia and decreased red blood cell count are independent risk factors for postoperative BSI in sIVH patients. The predictive model based on postoperative ONSD and preoperative blood glucose provides moderate risk assessment value for BSI and may preliminarily assist in early identification of high-risk patients, yet its efficacy warrants further validation through multicenter, prospective studies.