首页|急诊老年髋部骨折患者术前下肢深静脉血栓发生风险列线图预测模型构建与评价

急诊老年髋部骨折患者术前下肢深静脉血栓发生风险列线图预测模型构建与评价

Construction and evaluation of nomogram prediction model for lower extremity deep vein thrombosis in elderly patients with emergency hip fracture before operation

扫码查看
目的 探讨急诊老年髋部骨折患者术前发生下肢深静脉血栓(DVT)的危险因素,并构建列线图预测模型.方法 选择2017年5月至2020年1月首都医科大学附属北京积水潭医院收治的495例急诊老年髋部骨折患者为研究对象.依据术前是否发生下肢DVT将患者分为DVT组和非DVT组.通过医院病历系统收集患者的性别、年龄、体质量指数(BMI)、病程、骨折类型、高血压史、高血糖史、吸烟史、饮酒史、既往手术史、肺部/尿路感染情况、卧床时间、收缩压和舒张压.术前采集患者外周肘静脉血5 mL,3 000 r·min-1离心15 min取血清,使用监测仪检测患者空腹血糖(FBG)、总胆固醇(TC)、红细胞(RBC)计数、白细胞(WBC)计数、三酰甘油(TG)、钠离子、钾离子、钙离子、血小板、纤维蛋白原(FIB)以及D-二聚体水平.比较DVT组和非DVT组患者的各项指标,应用logistic回归分析急诊老年髋部骨折患者术前发生下肢DVT的独立影响因素,采用受试者操作特征(ROC)曲线分析各独立影响因素对急诊老年髋部骨折患者术前下肢DVT发生的预测效能;依据获得的独立影响因素构建列线图预测模型,采用R软件中C指数、ROC曲线及校准曲线评估列线图预测模型对急诊老年髋部骨折患者术前发生下肢DVT风险的预测效能.结果 495例急诊老年髋部骨折患者中,术前发生下肢DVT者52例(占比为10.51%).DVT组和非DVT组患者的性别、BMI、病程、骨折类型、高血压、吸烟史、饮酒史、肺部/尿路感染、收缩压、舒张压、FBG、TC、RBC、WBC、TG、钠离子、钾离子、钙离子及血小板水平比较差异均无统计学意义(P>0.05);DVT组患者的年龄、高血糖占比、既往手术史占比、卧床时间以及FIB、D-二聚体水平均显著高于非DVT组(P<0.05).Logistic回归分析显示,年龄大、高血糖、手术治疗、卧床时间长、高水平FIB以及高水平D-二聚体是急诊老年髋部骨折患者术前发生下肢DVT的独立危险因素(P<0.05);ROC曲线结果显示,年龄、高血糖史、手术治疗、卧床时间、FIB以及D-二聚体对急诊老年髋部骨折患者术前发生下肢DVT均有一定预测价值(P<0.05).基于年龄、高血糖史、手术治疗、卧床时间、FIB以及D-二聚体建立列线图预测模型,校准曲线C指数为0.953;建模组和验证组的AUC分别为0.953和0.940.结论 年龄大、高血糖、手术治疗、卧床时间长、高水平FIB、高水平D-二聚体是急诊老年髋部骨折患者术前发生下肢DVT的独立危险因素;基于年龄、高血糖史、是否行手术治疗、卧床时间、FIB以及D-二聚体构建的急诊老年髋部骨折患者术前发生下肢DVT的列线图预测模型具备较好的区分度和良好的预测能效,能直观地预测急诊老年髋部骨折患者术前发生下肢DVT的概率.
Objective To explore the risk factors of lower extremity deep vein thrombosis(DVT)in elderly patients with emergency hip fracture before operation,and to build a nomogram prediction model.Methods A total of 495 elderly patients with emergency hip fracture admitted to Beijing Jishuitan Hospital Affiliated to Capital Medical University from May 2017 to January 2020 were selected as the research subjects.The patients were divided into the DVT group and non-DVT group according to whether lower extremity DVT occurred or not before operation.The patient's gender,age,body mass index(BMI),course of disease,fracture type,hypertension history,hyperglycemia hist ory,smoking history,drinking history,past surgical history,pulmonary/urinary tract infection,bed rest time,systolic blood pressure,and diastolic blood pressure were collected through the hospital's medical record system.Before operation,5 mL of peripheral elbow venous blood was collected and centrifuged for 15 min at 3 000 r·min-1 to obtain the serum.The fasting blood glucose(FBG),total cholesterol(TC),red blood cell(RBC)count,white blood cell(WBC)count,triglyceride(TG),sodium ions,potassium ions,calcium ions,platelets,fibrinogen(FIB),and D-dimer levels were measured using the monitor.The indices of patients in the DVT and non-DVT groups were compared.Logistic regression was used to analyze the independent influencing factors of lower extremity DVT in elderly patients with emergency hip fracture before operation,and receiver operating characteristic(ROC)curve was used to analyze the predictive effect of each independent influencing factor for lower extremity DVT in elderly patients with emergency hip fracture before operation.The independent influencing factors were incorporated into the nomogram prediction model,and the C-index,ROC curve,and calibration curve in R software were used to evaluate the predictive effect of the nomogram prediction model for the preoperative risk of lower extremity DVT in elderly patients with emergency hip fracture.Results Among 495 elderly patients with emergency hip fracture,52 patients(10.51%)had lower extremity DVT before operation.There was no significant difference in gender,BMI,course of disease,fracture type,hypertension,smoking history,drinking history,pulmonary/urinary tract infection,systolic blood pressure,diastolic blood pressure,FBG,TC,RBC,WBC,TG,sodium ions,potassium ions,calcium ions,and platelet between the DVT group and non-DVT group(P>0.05).The age,proportion of hyperglycemia,proportion of past surgical history,bed rest time,and levels of FIB and D-dimer in the DVT group were significantly higher than those in the non-DVT group(P<0.05).Logistic regression analysis showed that older age,hyperglycemia,surgical treatment,long bed rest time,and high FIB and D-dimer levels were independent risk factors for lower extremity DVT in elderly patients with emergency hip fracture before operation(P<0.05).The ROC curve showed that age,hyperglycemia,operation or not,bed rest time,FIB and D-dimer levels had certain predictive values for lower extremity DVT in elderly patients with emergency hip fracture before operation(P<0.05).Based on age,hyperglycemia,operation or not,bed rest time,FIB and D-dimer levels,the nomogram prediction model was established,and the calibration curve C-index value was 0.953.The AUC values of the modeling group and validation group were 0.953 and 0.940,respectively.Conclusion Older age,hyperglycemia,surgical treatment,long bed rest time,high FIB and D-dimer levels are independent risk factors for lower extremity DVT in elderly patients with emergency hip fracture before operation.Based on age,hyperglycemia,operation or not,bed rest time,FIB and D-dimer levels,the nomo-gram model for predicting lower extremity DVT in elderly patients with emergency hip fracture before operation has good discrimination and predictive effect,which can directly predict the probability of lower extremity DVT in elderly patients with emergency hip fracture before operation.

elderly hip fracturelower extremity deep vein thrombosisrisk factorsnomograph prediction model

刘鑫鑫、刘志伟

展开 >

首都医科大学附属北京积水潭医院急诊科,北京 100035

老年髋部骨折 下肢深静脉血栓 危险因素 列线图预测模型

北京市医院管理中心"青苗"计划

QML20210405

2024

新乡医学院学报
新乡医学院

新乡医学院学报

CSTPCD
影响因子:0.999
ISSN:1004-7239
年,卷(期):2024.41(5)
  • 19