胫骨平台骨折术前下肢深静脉血栓形成列线图预测模型构建与验证  

Prediction model of preoperative deep vein thrombosis of lower limbs by nomogram in patients with tibial plateau fractures

作　　者：唐茁栋 王明友 王洪平 李亭燕 王川[1] 刘绍江[1] Tang Zhuodong;Wang Mingyou;Wang Hongping;Li Tingyan;Wang Chuan;Liu Shaojiang(Department of Orthopedics,Panzhihua Central Hospital,Panzhihua,Sichuan Province 617000,China)

机构地区：[1]攀枝花市中心医院骨科,四川攀枝花617000

出　　处：《创伤外科杂志》2025年第1期43-49,共7页Journal of Traumatic Surgery

基　　金：四川省临床重点专科建设项目经费资助;攀枝花市指导性科技计划项目(2022ZD-S-16)。

摘　　要：目的 探讨胫骨平台骨折患者术前发生下肢深静脉血栓(DVT)的危险因素,构建风险预测模型并验证。方法 回顾性分析2018年1月—2024年1月攀枝花市中心医院骨科收治的胫骨平台骨折患者166例,男性108例,女性58例;年龄22~76岁,平均49.3岁;摔伤69例,道路交通伤64例,高处坠落伤33例。SchatzkerⅠ~Ⅲ型91例,SchatzkerⅣ~Ⅵ型75例。根据术前下肢深静脉彩色多普勒超声结果将患者分为DVT组(50例)和非DVT组(116例)。通过单因素分析和多因素Logistic回归分析确定术前发生DVT的独立危险因素,根据独立危险因素构建列线图预测模型,绘制受试者工作特征(ROC)曲线,校准曲线和临床决策分析(DCA)曲线验证该模型的预测效能。结果 166例患者中术前检出DVT50例,术前DVT发生率30.1%(95%CI:0.231~0.372)。单因素分析结果显示,年龄、致伤因素、受伤至手术时间和术前D-二聚体比较差异有统计学意义(P<0.05),多因素Logistic回归分析结果显示,年龄≥47.5岁和术前D-二聚体≥2422.5ng/mL是胫骨平台骨折患者术前发生DVT的独立危险因素。ROC曲线显示,该模型曲线下面积(AUC)为0.960,95%CI:0.929~0.991,灵敏度为88.0%,特异度为95.6%,校准曲线显示该模型一致性良好,DCA曲线显示在阈值范围0~1中该模型临床净收益均较高。结论 年龄≥47.5岁、术前D-二聚体≥2422.5ng/mL为胫骨平台骨折患者术前发生DVT的独立危险因素。由此构建的模型具有较好的区分度和一致性,应用于胫骨平台骨折术前DVT的预测具有较好的预测效能。Objective To explore the risk factors of preoperative deep vein thrombosis(DVT)of lower extremities in patients with tibial plateau fractures,and to construct and verify a risk prediction model.Methods Clinical data of 166 patients with tibial plateau fractures admitted to Panzhihua Central Hospital from Jan.2018 to Jan.2024 were retrospectively analyzed.There were 108 males and 58 females aged 22-76(mean 49.3)years.Among them,69 were due to ground-level falls,64 to road traffic accidents and 33 to falls from height,with 91 cases of Schatzker typeⅠ-Ⅲand 75 typesⅣ-Ⅵ.According to the results of whether developing lower limb DVT before surgery,confirmed by the color Doppler ultrasound,patients were divided into DVT group(n=50)and non-DVT group(n=116).Independent risk factors for preoperative DVT were determined by univariate analysis and multivariate Logistic regression analysis,and then the nomogram prediction model was constructed,whose predictive efficacy was further verified by the receiver operator characteristic(ROC)curve,calibration curve and decision curve analysis(DCA)curve.Results Among the 166 patients included in the study,50 were diagnosed to have preoperative DVT,with the incidence being 30.1%(95%CI:0.231-0.372).Univariate analysis showed that age,injury causes,time from injury to surgery,and preoperative D-dimer level were significantly different between the two groups(all P<0.05).Multivariate Logistic regression analysis further showed that age≥47.5 years and preoperative D-dimer level≥2422.5 ng/mL were independent risk factors for preoperative DVT in patients with tibial plateau fracture.ROC curve revealed that the area under the ROC curve(AUC)of the model was 0.960(95%CI:0.929-0.991),the sensitivity was 88.0%,and the specificity reached 95.6%.The calibration curve showed that the model had good consistency.DCA curve showed that the net clinical benefit of the model was high in the threshold range of 0-1.Conclusion Age≥47.5 years and preoperative D-dimer≥2422.5 ng/mL are independent

关 键 词：胫骨平台骨折 深静脉血栓 危险因素 列线图 预测模型 

分 类 号：R683.42[医药卫生—骨科学]