下肢动脉血管重建术后并发淋巴漏的风险预测列线图模型建立及预防指导

Establishment of nomogram model for risk prediction of lymphatic leakage after arterial revascularization of lower extremities and guidance for its prevention

目的探讨下肢动脉血管重建术后并发淋巴漏的影响因素,并以此构建风险预测列线图模型,同时提出相对应的预防指导建议。方法回顾性分析于2014年1月至2022年6月在徐州市中心医院行下肢动脉血管重建的879例下肢动脉硬化闭塞症(LEASO)病人的临床资料,根据病人术后有无并发淋巴漏将其分为并发组(31例)和未并发组(848例),比较两组病人的一般资料。采用多因素logistic回归分析法分析下肢动脉血管重建术后并发淋巴漏的影响因素,并基于此采用R 3.4.5软件包绘制列线图模型,且绘制受试者工作特征(ROC)曲线对列线图模型的预测效能进行评估,绘制校准度曲线图并采用Bootstrap法检验预测列线图预测模型的一致性。结果单因素分析显示,两组身体质量指数(BMI)[(26.52±2.01)kg/m^(2)比(25.71±2.13)kg/m^(2)]及Fontaine分期[Ⅳ期38.71%比18.28%]、术中出血量[>100 mL 29.03%比11.20%]、术后血红蛋白(Hb)水平[(95.33±15.25)g/L比(104.57±22.38)g/L]、术后白蛋白(Alb)水平[(25.68±5.59)g/L比(28.28±6.17)g/L]、使用高频电刀(83.87%比55.54%)、腹股沟切口(80.65%比56.25%)差异有统计学意义(均P<0.05)。logistic回归分析可知,BMI[OR=2.52,95%CI:(1.18,3.86)]、Fontaine分期[OR=3.22,95%CI:(2.08,4.36)]、术中出血量>100 mL[OR=3.36,95%CI:(2.23,4.48)]、术后Hb水平[OR=2.97,95%CI:(1.75,4.20)]、术后Alb水平[OR=3.10,95%CI:(2.02,4.19)]、使用高频电刀[OR=3.39,95%CI:(2.20,4.59)]、腹股沟切口[OR=3.31,95%CI:(2.00,4.62)]均是影响LEASO病人下肢动脉血管重建术后并发淋巴漏的独立危险因素(P<0.05)。据此构建预测LEASO病人下肢动脉血管重建术后并发淋巴漏的列线图模型,预测值同实测值基本一致,一致性指数为0.836,模型具有良好的精准度和区分度。结论BMI、Fontaine分期、术中出血量>100 mL、术后Hb水平、术后Alb水平、使用高频电刀、腹股沟切口均为影响LEASO病人下肢动脉血管重建术后并发淋巴漏的影响因素,基于上述因素构建的列线图模型效能高。

Objective To explore the influencing factors of lymphatic leakage after lower extremity arterial revascularization,to establish a risk prediction nomograph model,and to propose corresponding prevention guidance suggestions.Methods The clinical data of 879 patients with lower extremity arteriosclerosis obliteran(LEASO)who underwent lower extremity arterial revascularization in Xuzhou Central Hospital from January 2014 to June 2022 were retrospectively analyzed.The patients were assigned into the complication group(31 cases)and the non-complication group(848 cases)according to whether with lymph leakage after surgery,and the general data of the two groups were compared.Multivariate logistic regression analysis was used to analyze the influencing factors of lymphatic leakage after lower extremity arterial revascularization,and R3.4.5 software package was used to draw the nomogram model based on this,and the receiver operating characteristic(ROC)curve was drawn to evaluate the prediction efficiency of the nomogram model,and the calibration curve was drawn,and Bootstrap method was used to test the consistency of the nomogram prediction model.Results Univariate analysis showed that body mass index(BMI)[(26.52±2.01)kg/m^(2) vs.(25.71±2.13)kg/m^(2)],Fontaine stage(Ⅳ38.71%vs.18.28%),intraoperative blood loss>100 mL(29.03%vs.11.20%),postoperative hemoglobin(Hb)level[(95.33±15.25)g/L vs.(104.57±22.38)g/L],postoperative albumin(Alb)level[(25.68±5.59)g/L vs.(28.28±6.17)g/L],use of high frequency electric knife(83.87%vs.55.54%),inguinal incision(80.65%vs.56.25%)of the two groups had statistical significance(P<0.05).Logistic regression analysis showed that BMI[OR=2.52,95%CI:(1.18,3.86)],Fontaine stage[OR=3.22,95%CI:(2.08,4.36)],intraoperative blood loss>100 mL[OR=3.36,95%CI:(2.23,4.48)],postoperative Hb level[OR=2.97,95%CI:(1.75,4.20)],postoperative Alb level[OR=3.10,95%CI:(2.02,4.19)],use of high frequency electric knife[OR=3.39,95%CI:(2.20,4.59),and inguinal incision[OR=3.31,95%CI:(2.00,4.62)]were independent risk factors for lymphatic leakage after lower extremity arterial revascularization in LEASO patients(P<0.05).On this basis,a nomograph model was constructed to predict lymphatic leakage after lower extremity arterial revascularization in LEASO patients.The validation model results showed that the predicted value was basically consistent with the measured value,and the consistency index(C-index value)was 0.836,indicating that the model had good accuracy and discrimination.Conclusion BMI,Fontaine stage,intraoperative blood loss>100 mL,postoperative Hb level,postoperative Alb level,use of high frequency electric knife and inguinal incision are all influencing factors for LEASO patients with lymphatic leakage after arterial revascularization of lower extremities.The nomogram constructed based on the above factors has high efficiency.