基于诺莫图构建髋部骨折患者发生下肢深静脉血栓的预测模型  

Predictive modeling of the occurrence of lower extremity deep vein thrombosis in hip fracture patients based on nomograms

作　　者：王银平[1] 杜雪莲[1] 王亮云 黄瑞珍[1] 罗乙舒 熊俊琴 何丽连 WANG Yinping;DU Xuelian;WANG Liangyun;HUANG Ruizhen;LUO Yishu;XIONG Junqin;HE Lilian(Seventh Department of Orthopedics,Foshan Traditional Chinese Medicine Hospital,Foshan 528000,China)

机构地区：[1]佛山市中医院骨七科,广东佛山528000

出　　处：《广东药科大学学报》2024年第2期144-149,共6页Journal of Guangdong Pharmaceutical University

摘　　要：目的探讨髋部骨折患者发生下肢深静脉血栓(DVT)的独立危险因素,并构建和验证列线图模型。方法选择2020年1月至2023年6月佛山市中医院骨科收治的髋部骨折患者227例为研究对象,按照7∶3比例将患者分为建模组(n=159)和验证组(n=68)。收集患者的临床资料,采用LASSO回归和Logistic回归筛选髋部骨折患者发生DVT的独立危险因素,建立风险预测模型和列线图。采用一致性指数(C-index)、校准曲线、受试者工作特征(ROC)曲线进行评价。结果建模组159例髋部骨折中发生DVT 33例,发生率为20.89%。LASSO回归结合多因素Logistic回归分析结果显示,年龄≥70岁、糖尿病、受伤至手术时间≥5 d、低蛋白血症、D-二聚体>0.5是髋部骨折患者发生DVT的独立危险因素。基于上述危险因素绘制列线图模型,将上述5个危险因素构建列线图模型,模型区分度良好(C-index=0.932);Hosmer-Lemeshow检验显示,列线图的拟合优度良好(χ^(2)=57.625,P=0.732)。建模组和验证组发生DVT的ROC曲线下面积分别为0.932、0.821,模型准确度高。结论年龄≥70岁、糖尿病、受伤至手术时间≥5 d、低蛋白血症、D-二聚体>0.5是髋部骨折患者发生DVT的独立危险因素,基于上述因子构建的列线图模型可有效预测髋部骨折患者发生DVT的风险。Objective To explore the independent risk factors for the occurrence of lower extremity deep vein thrombosis(DVT)in patients with hip fracture and to construct and validate a column-line diagram model.Methods 227 hip fracture patients admitted to the Department of Orthopedics of Foshan Hospital of Traditional Chinese Medicine from January 2020 to June 2023 were selected as the study subjects,and the patients were divided into the modeling group(n=159)and the validation group(n=68)according to the ratio of 7∶3.The clinical data of the patients were collected,and LASSO regression and logistic regression were used to screen the independent risk factors for DVT in hip fracture patients,and a risk prediction model and column line diagram were established.Consistency index(C-index),calibration curve,and subject work characteristics(ROC)curve were used for evaluation.Results DVT occurred in 33 of 159 hip fractures in the modeling group,with an incidence rate of 20.89%.LASSO regression combined with multifactorial logistic regression analysis showed that age≥70 years,diabetes mellitus,time from injury to surgery≥5 d,hypoproteinemia,and D-dimer>0.5 were the independent risk factors for DVT in patients with hip fractures.A column-line diagram model was drawn based on the above risk factors,and a column-line diagram model was constructed with the above five risk factors,and the model was well differentiated(C-index=0.932).The Hosmer-Lemeshow test showed a good goodness-of-fit of the column-line diagram(χ^(2)=57.625,P=0.732).The area under the ROC curve for the occurrence of DVT in the modeling group and validation group was 0.932 and 0.821,respectively,and the model was accurate.Conclusion Age≥70 years,diabetes mellitus,time from injury to surgery≥5 d,hypoproteinemia,and D-dimer>0.5 are independent risk factors for DVT in hip fracture patients,and the column-line graph model constructed based on the above factors can effectively predict the risk of DVT in hip fracture patients.

关 键 词：髋部骨折 下肢深静脉血栓 危险因素 列线图 

分 类 号：R683.1[医药卫生—骨科学]