肝切除术围手术期症状性静脉血栓栓塞症风险评估模型的建立及验证

目的分析肝切除术患者围手术期症状性静脉血栓栓塞症（VTE）与临床特征的关系，建立风险评估模型来预测发生VTE的高危人群。 方法收集2017年1月至2020年6月接受肝切除术206例患者的临床资料。根据围手术期症状性VTE诊断标准分为VTE组（n=45例）和非VTE组（n=161例）。采用SPSS 20.0软件对VTE与临床特征之间的关系进行单因素及多因素Logistic回归分析。以P<0.05为差异有统计学意义。并构建肝切除术患者围手术期症状性VTE的风险评估模型，通过描绘受试者工作特征曲线（ROC）并计算曲线下面积（AUC）来评价模型的评估能力，并以术后实际发生结果为金标准对评估模型进行验证。 结果两组在BMI、VTE史、是否有肝脏恶性肿瘤、手术时间、术后肺部感染、术后间歇充气加压装置（IPC）使用、D-二聚体浓度之间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，VTE史、肝脏恶性肿瘤、手术时间≥3 h、术后肺部感染及D-二聚体≥0.2 mg/L是肝切除术患者围手术期发生症状性VTE的独立危险因素，术后IPC使用是肝切除术患者围手术期发生症状性VTE的保护性因素。VTE风险评估模型的AUC值为0.913（95%CI为0.869~0.974，P=0.004）；当Y的最佳临界值为2.85时对VTE发生风险具有评估价值，灵敏度为86.7%，特异度为89.4%，约登指数为0.761。以术后实际发生VTE为标准对评估模型进行验证，诊断准确度为92.7%（191/206），灵敏度为80.0%（36/45），特异度为96.3%（155/161）。 结论VTE史、肝脏恶性肿瘤、手术时间≥3 h、术后肺部感染及D-二聚体≥0.2 mg/L是肝切除术围手术期发生症状性VTE的独立危险因素。本研究建立的肝切除术患者围手术期症状性VTE风险评估模型对于VTE发生具有较高的预测价值，有良好的适用性。

Establishment and validation of a risk assessment model for perioperative symptomatic venous thromboembolism after hepatectomy

ObjectiveTo analyze the relationship between symptomatic venous thromboembolism（VTE）and clinical features during perioperative period of hepatectomy，and to establish a risk assessment model to predict the high risk population of VTE. MethodsThe clinical data of 206 patients who underwent hepatectomy from January 2017 to June 2020 were collected.According to the diagnostic criteria of perioperative symptomatic VTE，the patients were divided into VTE group（n=45 cases）and non-VTE group（n=161 cases）. SPSS 20.0 software was used for univariate and multivariate Logistic regression analysis of the relationship between VTE and clinical features，and P<0.05 was considered as statistically significant. The risk assessment model for symptomatic VTE in hepatectomy patients was constructed. The evaluation ability of the model was evaluated by describing the receiver operating characteristic curve（ROC）and calculating the area under the curve（AUC）. The evaluation model was verified by the actual postoperative outcome as the gold standard. ResultsThere were statistically significant differences between the two groups in BMI，VTE history，liver malignancy，operation time，postoperative pulmonary infection，use of intermittent pneumatic pressurization device（IPC），and D-dimer concentration（P<0.05）. Multivariate Logistic regression analysis showed that the history of VTE，hepatic malignancy，operation time≥3 h，postoperative pulmonary infection and D-dimer≥0.2 mg/L were independent risk factors for perioperative symptomatic VTE in patients undergoing hepatectomy，and postoperative IPC use was a protective factor for perioperative symptomatic VTE in patients undergoing hepatectomy. The AUC value of the VTE risk assessment model was 0.913（95%CI：0.869~0.974，P=0.004）.When the optimal critical value of Y is 2.85，it had evaluation value for the risk of VTE，with a sensitivity of 86.7%，specificity of 89.4% and Yoden index of 0.761. The diagnostic accuracy，sensitivity and specificity of the evaluation model were 92.7%（191/206），80.0%（36/45）and 96.3%（155/161），respectively，based on the actual occurrence of postoperative VTE. ConclusionHistory of VTE，hepatic malignancy，operation time ≥3 h，postoperative pulmonary infection and D-dimer ≥0.2 mg/L were independent risk factors for perioperative symptomatic VTE.The risk assessment model of perioperative symptomatic VTE in patients with hepatectomy established in this study has high predictive value for the occurrence of VTE and has good applicability.