老年髋部骨折后静脉血栓栓塞风险预测模型的构建与效率分析。
Construction and efficiency analysis of prediction model for venous thromboembolism risk in the elderly after hip fracture.
机构信息
Department of Orthopaedic Trauma, Xiangya Hospital, Central South University, Changsha 410008, China.
出版信息
Zhong Nan Da Xue Xue Bao Yi Xue Ban. 2021 Feb 28;46(2):142-148. doi: 10.11817/j.issn.1672-7347.2021.190722.
OBJECTIVES
To screen the risk factors for predicting venous thromboembolism (VTE) risk after hip fracture in the elderly, to establish a prediction model based on these factors, and to analyze its prediction efficacy.
METHODS
A total of 52 hip fracture patients over 60 years old with VTE admitted to the Department of Orthopaedic Trauma, Xiangya Hospital, Central South University from March 2017 to April 2019 were selected as a thrombus group, and another 52 hip fracture patients over 60 years old without VTE were selected as a control group. The differences of hospitalization data and examination results between the 2 groups were compared. Logistic regression model was used to explore the influence of risk factors on VTE risk after hip fracture in the elderly and construct the prediction model based on these factors. The receiver operating characteristic curve was used to analyze the predictive effectiveness of model, Hosmer-lemeshow goodness of fit test was used to evaluate the fitting degree of prediction model.
RESULTS
Univariate analysis showed that injury-admission interval, Caprini score, WBC count, platelet count, neutrophil count, neutrophil-to-lymphocyte ratio, platelet-to-lymphocyte ratio, monocyte-to-lymphocyte ratio, systemic immune-inflammatory index (SII), and fibrinogen in the thrombus group were higher than those in the control group (all <0.05). Logistic regression analysis showed that injury-admission interval, Caprini score, and SII were independent predictors of VTE risk after hip fracture in the elderly. The AUC was 0.949 (95% CI 0.901 to 0.996) when the sensitivity and specificity were 82.70% and 96.20%, respectively, which were significantly higher than each single index, and the prediction model had perfect fitting degree (Hosmer-lemeshow =14.078, >0.05).
CONCLUSIONS
SII, Caprini score, and injury-admission interval are independent predictors of VTE after hip fracture in the elderly. The prediction model based on these 3 factors has a good efficacy on the prediction of VTE risk, and could provide important reference for the prevention, management, and treatment of VTE after hip fracture in the elderly.
目的
筛选预测老年髋部骨折后静脉血栓栓塞症（VTE）风险的危险因素，基于这些因素建立预测模型，并分析其预测效能。
方法
选取 2017 年 3 月至 2019 年 4 月中南大学湘雅医院骨科创伤病房收治的 52 例年龄大于 60 岁的髋部骨折合并 VTE 的患者为血栓组，另选取同期 52 例年龄大于 60 岁的髋部骨折未合并 VTE 的患者为对照组。比较两组患者的住院资料及检查结果差异，采用 logistic 回归模型探讨影响老年髋部骨折后 VTE 风险的危险因素，并基于这些因素构建预测模型。采用受试者工作特征曲线分析模型的预测效能，Hosmer-lemeshow 拟合优度检验评估预测模型的拟合程度。
结果
单因素分析显示，血栓组患者的损伤至入院间隔时间、Caprini 评分、白细胞计数、血小板计数、中性粒细胞计数、中性粒细胞与淋巴细胞比值、血小板与淋巴细胞比值、单核细胞与淋巴细胞比值、全身免疫炎症指数（SII）和纤维蛋白原均高于对照组（均<0.05）。logistic 回归分析显示，损伤至入院间隔时间、Caprini 评分和 SII 是老年髋部骨折后 VTE 风险的独立预测因素。当灵敏度和特异度分别为 82.70%和 96.20%时，模型的 AUC 为 0.949（95%CI 0.901~0.996），均明显高于各单项指标，且预测模型拟合度良好（Hosmer-lemeshow=14.078，>0.05）。
结论
SII、Caprini 评分和损伤至入院间隔时间是老年髋部骨折后 VTE 的独立预测因素。基于这 3 个因素的预测模型对 VTE 风险的预测效果较好，可为老年髋部骨折后 VTE 的预防、管理和治疗提供重要参考。
Zotero 插件