普通外科恶性肿瘤患者术后下肢深静脉血栓形成预测模型的构建  

Development of a prediction model for postoperative lower extremity deep venous thrombosis in patients with malignant tumors undergoing general surgery

作　　者：盛昌 贺爱兰[2,3,5] 万凌燕 王伟[1,5] 黄建华[1,5] 杨璞[1,5] 唐红英[2,4,5] SHENG Chang;HE Ailan;WAN Lingyan;WANG Wei;HUANG Jianhua;YANG Pu;TANG Hongying(Department of Vascular Surgery,Xiangya Hospital,Central South University,Changsha 410008,China;Teaching and Research Section of Clinical Nursing,Xiangya Hospital,Central South University,Changsha 410008,China;Department of Clinical Nursing,Xiangya Hospital,Central South University,Changsha 410008,China;Reproductive Medicine Center,Xiangya Hospital,Central South University,Changsha 410008,China;National Clinical Research Center for Geriatric Disorders,Xiangya Hospital,Central South University,Changsha 410008,China)

机构地区：[1]中南大学湘雅医院血管外科,湖南长沙410008 [2]中南大学湘雅医院临床护理学教研室,湖南长沙410008 [3]中南大学湘雅医院护理部,湖南长沙410008 [4]中南大学湘雅医院生殖医学中心,湖南长沙410008 [5]中南大学湘雅医院国家老年疾病临床医学研究中心,湖南长沙410008

出　　处：《中国普通外科杂志》2023年第6期850-858,共9页China Journal of General Surgery

基　　金：中南大学湘雅医院2020年度国家老年疾病临床医学研究中心适宜技术推广基金资助项目(XYYYJSTG-07)。

摘　　要：背景与目的:下肢深静脉血栓形成(LDVT)是普通外科恶性肿瘤患者术后常见的并发症之一,严重影响患者的术后康复。目前的评估工具无法对LDVT患者进行更加细致的风险分层。因此,本研究探讨普通外科恶性肿瘤患者术后发生LDVT的影响因素,并建立可靠的预测工具,从而为LDVT的诊断和防治提供帮助。方法:回顾性分析2021年1月1日—2022年10月31日中南大学湘雅医院普通外科手术治疗恶性肿瘤患者的临床资料,根据良好设计的纳入、排除标准对病例进行严格的质量控制。使用已经较明确的LDVT的影响因素和一些重要临床特征作为分析变量,单变量分析和多变量分析用以评估LDVT的影响因素以及筛选模型的预测因子。应用编程软件制作基于Logistic回归的列线图并通过受试者工作特征曲线(ROC)评估列线图的预测性能,校准曲线用以评估预测模型与数据的拟合程度。使用决策曲线分析(DCA)比较预测模型与其他单一指标临床应用价值的差异。结果:本研究共纳入342例患者,其中LDVT组167例,对照组175例。单变量分析显示,1个月内有手术创伤史、高血压史、吸烟史、饮酒史、放疗史、ICU住院时间、红细胞(RBC)、血红蛋白(Hb)、纤维蛋白原降解产物(FDP)、D-二聚体、凝血时间、手术时间、术中输注RBC、术中输注血浆、手术方式与LDVT的发生有关(均P<0.05)。多变量分析显示,1个月内有手术创伤史、FDP、凝血时间、手术时间、术中输注RBC、术中输注血浆是术后LDVT的独立影响因素(均P<0.05)。将独立影响因素作为预测因子建立模型,预测术后2周LDVT风险列线图的ROC曲线的曲线下面积(AUC)为0.830 (95%CI=0.787~0.874,P<0.001)。校准曲线中的Hosmer-Lemeshow检验统计量为0.973;DCA分析显示了列线图比单一指标有更好的净效益。结论:本研究构建的预测模型具有良好的鉴别能力和临床应用价值,有助于临床医生Background and Aims:Lower extremity deep venous thrombosis(LDVT)is a common postoperative complication in patients with general surgical malignancies,significantly affecting their postoperative recovery.Currently,assessment tools cannot provide a detailed risk stratification for LDVT patients.Therefore,this study aims to explore the influencing factors for LDVT occurrence in patients with general surgical malignancies after surgery and establish a reliable prediction tool to assist in diagnosing and preventing LDVT.Methods:The clinical data of patients undergoing inpatient surgery for malignant tumors in the Department of General Surgery,Xiangya Hospital,Central South University from January 1,2021,to October 31,2022,were retrospectively collected,and the cases were strictly quality-controlled according to well-designed inclusion and exclusion criteria.Established LDVT influencing factors and important clinical features were used as analysis variables.Univariate and multivariate analyses were performed to evaluate the influencing factors for LDVT and screen predictive factors for the model.A receiver operating characteristic(ROC)curve based on Logistic regression was created using programming software to assess the model's predictive performance.A calibration curve was used to evaluate the goodness of fit between the prediction model and the data.Decision curve analysis was employed to compare the clinical application value of the prediction model with other single indicators.Results:A total of 342 patients were included,with 167 cases in the LDVT group and 175 cases in the control group.Univariate analysis revealed that a history of surgical trauma within one month,hypertension,smoking,alcohol consumption,history of radiotherapy,duration of ICU stay,red blood cell(RBC)count,hemoglobin(Hb)level,fibrinogen degradation products(FDP),D-dimer,coagulation time,surgical duration,intraoperative RBC transfusion,intraoperative plasma transfusion,and surgical approach were all related to the occurrence of LDVT(all P<0.05).

关 键 词：静脉血栓形成 下肢 肿瘤 手术后并发症 危险因素 列线图 

分 类 号：R654.3[医药卫生—外科学]