2022 年第 8 期 第 17 卷

栓塞风险预测模型对中国感染性心内膜炎患者症状性栓塞事件的预测效能

Efficacy of embolism risk prediction model on predicting symptomatic embolism events in Chinese patients with infective endocarditis

作者：刘妍石秀梅朱光发

英文作者：Liu Yan Shi Xiumei Zhu Guangfa

单位：首都医科大学附属北京安贞医院感染科北京市心肺血管疾病研究所，北京100029

英文单位：Department of Infectious Disease Beijing Anzhen Hospital Capital Medical University Beijing Institute of Heart Lung and Blood Vessel Diseases Beijing 100029 China

关键词：感染性心内膜炎；症状性栓塞；栓塞风险预测计算器；赘生物；金黄色葡萄球菌

英文关键词：Infectiveendocarditis;Symptomaticembolism;EmbolivRiskFrenchCalculator;Vegetation;Staphylococcusaureus

• 摘要：

• 目的 验证栓塞风险预测计算器（ER-Calculator）对中国感染性心内膜炎(IE)患者症状性栓塞（SE）的预测效能，为中国人群IE治疗管理提供依据。方法 本研究为回顾性队列研究，纳入2018年1月至2019年12月入住首都医科大学附属北京安贞医院并诊断为IE的患者。开始抗菌药物治疗2周内出现的SE定为终点事件。应用Logistic回归方法分析栓塞患者与未栓塞患者在既往病史、化验指标及心脏超声等各方面的差异。使用Kaplan-Meier生存曲线（K-M曲线）分析经ER-Calculator确定的高风险人群与低风险人群SE预后的差异。应用受试者工作特征(ROC)曲线验证ER-Calculator在中国IE患者中的预测效能。结果 本研究共纳入171例患者，发生SE的患者共计32例，发生率为18.7%，以中枢神经系统栓塞最为常见。Logistic回归分析结果显示，既往栓塞病史、白细胞计数>15×109/L、C反应蛋白、赘生物长度、赘生物长度>10 mm以及ER-Calculator评分、ER-Calculator评分＞8%为SE事件的危险因素。K-M曲线分析结果显示，高风险人群SE累积发生率显著高于低风险人群，差异有统计学意义（P=0.002）。ROC曲线分析结果显示，ER-Calculator模型预测IE患者开始抗菌药物治疗2周内发生SE风险的特异度为0.849、敏感度为0.625、曲线下面积为0.725，效能良好。结论 ER-Calculator模型对中国IE患者的SE事件具有良好的预测效能。

• Objective To verify the effectiveness of Embolic Risk French Calculator(ER-Calculator) on predicting symptomatic embolism(SE) in Chinese patients with infective endocarditis(IE), so as to provide basis for the treatment and management of IE in Chinese population. Methods This retrospective cohort study included patients who were admitted to Beijing Anzhen Hospital, Capital Medical University and diagnosed with IE from January 2018 to December 2019. The endpoint was SE within 2 weeks of antibiotic initiation. The differences between embolism patients and non-embolism patients in medical history, laboratory examinations and echocardiography were analyzed by Logistic regression analysis. Kaplan-Meier survival curve(K-M curve) was used to analyze the difference in SE prognosis between the high risk population and low risk population determined by ER-Calculator. Receiver operating characteristic(ROC) curve was used to verify the predictive efficacy of ER-Calculator in Chinese IE patients. Results A total of 171 patients were included in this study, including 32 patients with SE, with an incidence of 18.7%, and central nervous system embolism was the most common. Logistic regression analysis showed that history of embolism, white blood cell count>15×109/L, C-reactive protein, vegetative length, vegetative length>10 mm, ER-Calculator score and ER-Calculator score>8% were risk factors for SE. K-M survival curve analysis showed that the cumulative incidence of SE in high-risk population was significantly higher than that in low-risk population(P=0.002). ROC curve analysis showed that the specificity, sensitivity and area under the curve of ER-Calculator model for predicting the risk of SE in patients with IE within 2 weeks after starting antimicrobial treatment were 0.849, 0.625 and 0.725, respectively, and the predictive efficiency was good. Conclusion The ER-Calculator model is a useful predictor of SE among Chinese IE patients.