全髋置换术后下肢深静脉血栓形成的风险预测

Risk prediction of lower extremity deep vein thrombosis after total hip arthroplasty

摘要:

[目的] 探讨全髋关节置换术（total hip arthroplasty, THA） 后下肢深静脉血栓形成（deep vain thrombosis, DVT） 的风险因素。[方法] 回顾分析本院骨科2018 年6 月—2022 年6 月364 例初次THA 患者临床资料。依据术后临床与超声检查，将患者分为DVT 组和非DVT 组，比较两组资料，行二元多因素逻辑回归分析，依据逻辑分析结果构建预测列线图。采用受试者工作特征曲线分析评价列线图的预测效能。[结果] 364 患者中，72 例患者发生DVT，占19.8%；另外292 例患者未发生DVT，占80.2%。DVT 组年龄、BMI、合并糖尿病、血栓史、手术时间、双侧置换比率和卧床时间均显著大于非DVT 组（P＜0.05），但是，前者使用抗凝剂的比率显著低于后者（P＜0.05）。逻辑分析表明，年龄≥70 岁（OR=4.127, 95%CI 1.712~5.392, P=0.012）、BMI≥28 kg/m2 （OR=2.916, 95% CI 1.153~4.124, P=0.063）、糖尿病（OR=2.328, 95% CI 1.412~4.411, P=0.067）、既往血栓病史（OR=5.270, 95%CI 1.091~2.559, P=0.011）、手术时间≥120 min（OR=3.541, 95%CI 1.215~4.089, P=0.027）、双侧置换（OR=2.286,95%CI 1.407~4.531, P=0.034）、术后卧床≥3 d（OR=4.427, 95%CI 1.136~1.973, P=0.023） 是DVT 发生的危险因素；而使用抗凝药物是DVT 发生的保护因素（OR=-2.019, 95%CI -1.723~-5.802, P＜0.001）。逻辑分析构建列线图预测模型并验证，ROC 曲线下面积（AUC） 为0.867（95%CI 0.711~1.000）。[结论] 本研究筛选出高龄、糖尿病、血栓史、手术时间、双侧关节置换、卧床时间是全髋关节置换术后下肢深静脉血栓形成的相关因素，并构建了效果较好的列线图预测模型，可能为防治DVT 提供参考。

关键词:全髋关节置换术;深静脉血栓;危险因素;列线图预测模型

Abstract:

[Objective] To search the risk factors of deep vain thrombosis (DVT) after total hip arthroplasty (THA). [Methods] A retro-spective study was conducted on 364 patients who received initial THA in our hospital from June 2018 to June 2022. According to postopera-tive clinical and ultrasound examination, patients were divided into DVT group and non-DVT group. The data of the two groups were com-pared, and binary multi-factor logistic regression analysis was performed, and the prediction nomogram was constructed according to the re-sults of logical analysis. The predictive efficiency of the nomogram was evaluated by the receiver operating characteristic curve analysis. [Re-sults] Among 364 patients, 72 had DVT, accounting for 19.8%, whereas the other 292 patients did not develop DVT, accounting for 80.2%.The DVT group proved significantly greater than the non-DVT group in terms of age, BMI, diabetes mellitus, history of thrombosis, operationtime, bilateral replacement ratio and bed rest time (P＜0.05), but the former was significantly less than the latter regarding to the ratio of anti-coagulant usage (P＜0.05). As results of logical analysis, the age≥70 years (OR=4.127, 95%CI 1.712~5.392, P=0.012), BMI≥28 years (OR=2.916, 95%CI 1.153~4.124, P=0.063), diabetes mellitus (OR=2.328, 95%CI 1.412~4.411, P=0.067), previous history of thrombosis (OR=5.270, 95%CI 1.091~2.559, P=0.011), operation time ≥120 min (OR=3.541, 95%CI 1.215~4.089, P=0.027), bilateral replacement (OR=2.286, 95%CI 1.407~4.531, P=0.034), and postoperative bed rest ≥3 d (OR=4.427, 95%CI 1.136~1.973, P=0.023) were the risk factor forDVT, whereas the use of anticoagulants was a protective factor for DVT (OR=-2.019, 95%CI -1.723~-5.802, P＜0.001). A nomogram predic-tion model was created based on the logical regression outcome and verified by ROC with area under curve (AUC) of 0.867 (95%CI 0.711~1.000). [Conclusion] In this study, advanced age, diabetes, previous thrombosis, operative time, bilateral joint replacement, and bed resttime were selected as the related factors for deep vein thrombosis after total hip arthroplasty, and a better predictive model was established,which may provide reference for the prevention and treatment of DVT.

Key words:total hip arthroplasty; deep vein thrombosis; risk factors; nomogram prediction model