老年骨性关节炎全膝关节置换术后深静脉血栓形成预测模型的构建及验证

Construction and validation of a predictive model for deep venous thrombosis after total knee arthroplasty in elderly patients with osteoarthritis

目的 探讨老年骨性关节炎(OA)全膝关节置换术(TKA)后深静脉血栓形成(DVT)预测模型的构建及验证。方法 选取2020年11月至2022年11月在重庆全域肿瘤医院行TKA治疗的OA患者194例。根据术后3天影像检查结果,46例归入DVT组,148例归入Non-DVT组。分析OA患者TKA后DVT的危险因素,并构建列线图预测模型及验证,采用X-tile软件将模型分层,探讨其应用价值。结果 DVT组和Non-DVT组年龄、BMI、糖尿病史、TC、D-二聚体、卧床时间、手术时间、术中出血量、LDL-C比较差异有统计学意义(P<0.05)。年龄≥60岁、BMI≥26 kg/m2、糖尿病史、TC≥5.17 mmol/L、LDL-C≥1.04 mmol/L、D-二聚体≥0.5 mg/L是术后发生DVT的影响因素(P<0.05)。训练集和验证集的C-index分别为0.847、0.815,AUC分别为0.853、0.815,两集的校正曲线均与理想曲线拟合反映良好,低风险组患者的DVT发生率明显低于中、高风险组(P<0.05)。结论 预测模型能较好地预测OA老年患者TKA后下肢DVT的发生风险,为早期预防提供较为准确的决策依据。

Objective To investigate the construction and validation of prediction model of deep vein thrombosis(DVT)after total knee arthroplasty(TKA)in elderly patients with osteoarthritis(OA).A total of 194 OA patients who received TKA treatment in our hospitals from November 2020 to November 2022 were selected.According to the imaging results after 3 days of operation,46 patients were classified into a DVT group and 148 patients were classified into a Non-DVT group.Logistic regression analysis was conducted to analyze the risk factors of DVT after TKA in the OA patients,and a nomogram prediction model was constructed and verified.X-tile software was used to stratify the model and explore its application value.There were significant differences in age,BMI,diabetes history,total cholesterol(TC),D-dimer,bed time,operation time,intraoperative blood loss and low-density lipoprotein-cholesterol(LDL-C)between the two groups(P<0.05).Age≥60 years old,BMI≥26 kg/m 2,diabetes history,TC≥5.17 mmol/L,LDL-C≥1.04 mmol/L,D-dimer≥0.5 mg/L were all influencing factors for postoperative DVT(P<0.05).The C-index of the training set and the validation set were 0.847 and 0.815,and the AUC of the two sets were 0.853 and 0.815,respectively.The calibration curves of the two sets were well fitted to the ideal curve.The incidence of DVT in the low-risk group was significantly lower than that in the middle and high-risk groups(P<0.05).The prediction model can better predict the risk of lower limb DVT after TKA in elderly patients with OA.It provides a more accurate decision-making basis for early prevention.