旋后-外旋踝关节骨折术前发生下肢深静脉血栓诱导因素的多因素一般Logistic回归分析模型构建及分析

Construction and evaluation of a risk factor model for preoperative LEDVT in supination-external rotation ankle fractures using multivariable Logistic regression analysis

摘要:

目的 探究旋后-外旋踝关节骨折术前发生下肢深静脉血栓（LEDVT）的诱导因素，构建多因素一般Logistic回归分析模型，并予以评价。方法 选取2022年2月—2024年3月浙江省人民医院毕节医院骨科收治的102例旋后－外旋踝关节骨折患者，依据LEDVT诊断标准及术前发生情况将患者分为发生组11例、未发生组91例。比较两组患者人口学资料、社会学资料及血清学指标，采用多因素一般Logistic回归模型分析患者术后发生LEDVT的影响因素，构建术前LEDVT发生的回归方程，对模型总体、模型拟合优度进行验证。结果 两组患者性别构成、体质量指数、致伤原因、吸烟史、饮酒史、术前血红蛋白水平、高血压、糖尿病比较，差异均无统计学意义（P >0.05）。两组患者年龄、发病至入院时间、术前D-二聚体水平、术前白蛋白（ALB）水平、术前纤维蛋白原（FIB）水平、术前血小板与淋巴细胞比值（PLR）水平、血管及神经损伤比较，差异均有统计学意义（P <0.05）。多因素一般Logistic回归分析结果显示：年龄≥ 65岁［O^R=2.063（95% CI：1.402，3.035）］、发病至入院时间≥ 24 h［O^R=1.964（95% CI：1.296，2.976）］、术前D-二聚体≥0.3 mg/L［O^R=2.147（95% CI：1.352，3.409）］、术前ALB< 35 g/L［O^R=2.184（95% CI：1.320，3.614）］、术前FIB≥ 6.0 g/L［O^R=2.230（95% CI：1.205，4.127）］、术前PLR≥ 300×10⁹/L［O^R=2.214（95% CI：1.277，3.841）］、合并血管及神经损伤［O^R=2.517（95% CI：1.073，5.904）］均是术前LEDVT发生的危险因素（P <0.05）。回归方程为Logit（P）＝2.125+0.724×年龄≥ 65岁+0.675×发病至入院时间≥ 24 h+0.764×术前D-二聚体≥0.3 mg/L+0.781×术前ALB水平< 35 g/L+0.802×术前FIB水平≥ 6.0 g/L+0.795×术前PLR水平≥ 300×10⁹/L+0.923×合并血管及神经损伤，回归方程有统计学意义（P <0.05）。受试者工作特征曲线分析结果表明，回归方程的敏感性为90.9%（95% CI：0.587，0.998），特异性为92.3%（95% CI：0.848，0.969），曲线下面积为0.955（95% CI：0.899，1.000）。结论 存在上述诱导因素的旋后-外旋踝关节骨折患者术前发生LEDVT的风险性较大，骨科医护人员应引起重视，可针对以上诱导因素于患者术前采取针对性预防措施，最大限度降低患者LEDVT发生风险。

Abstract:

Objective To explore the risk factors for preoperative lower extremity deep vein thrombosis (LEDVT) in supination-external rotation ankle fractures, to construct a risk factor model using multivariable Logistic regression analysis, and to evaluate the model.Methods A total of 102 patients with supination-external rotation ankle fractures admitted to the Department of Orthopedics at Zhejiang Provincial People's Hospital Bijie Hospital from February 2022 to March 2024 were selected. According to the diagnostic criteria of LEDVT and the preoperative occurrence, the patients were divided into the occurrence group (11 cases) and the non-occurrence group (91 cases). Demographic data, sociological data and serological indicators were compared between the two groups. The factors affecting the occurrence of preoperative LEDVT were analyzed via multivariable Logistic regression, based on which the regression equation for preoperative occurrence of LEDVT was constructed. The overall performance and the goodness of fit of the model were validated.Results There was no difference in terms of the sex composition, body mass index, causes of injury, history of smoking, history of alcohol consumption, preoperative levels of hemoglobin, history of hypertension or history of diabetes mellitus between the two groups (P > 0.05). The age, time from onset to admission, preoperative levels of D-dimer, albumin (ALB), and fibrinogen (FIB), preoperative platelet-to-lymphocyte ratio (PLR), and vascular and nerve injuries were different between the two groups of patients (P < 0.05). Multivariable Logistic regression analysis indicated that age ≥ 65 years [O^R = 2.063 (95% CI: 1.402, 3.035) ], time from onset to admission ≥ 24 hours [O^R = 1.964 (95% CI: 1.296, 2.976) ], preoperative D-dimer levels ≥ 0.3 mg/L [O^R = 2.147 (95% CI: 1.352, 3.409) ], preoperative ALB levels < 35 g/L [O^R = 2.184 (95% CI: 1.320, 3.614) ], preoperative FIB levels ≥ 6.0 g/L [O^R = 2.230 (95% CI: 1.205, 4.127) ], preoperative PLR ≥ 300 × 10⁹/L [O^R = 2.214 (95% CI: 1.277, 3.841) ], and combined vascular and nerve injuries [O^R = 2.517 (95% CI: 1.073, 5.904) ] were risk factors for preoperative LEDVT in patients with supination-external rotation ankle fractures (P < 0.05). The regression equation was constructed as Logit(P) = 2.125 + 0.724 × (age ≥ 65 years) + 0.675 × (time from onset to admission ≥ 24 h) + 0.764 × (preoperative D-dimer levels ≥ 0.3 mg/L) + 0.781 × (preoperative ALB levels < 35 g/L) + 0.802 × (preoperative FIB levels ≥ 6.0 g/L) + 0.795 × (preoperative PLR ≥ 300×10⁹/L) + 0.923 × (combined vascular and nerve injuries), which was statistically significant (P < 0.05). The receiver operating characteristic curve analysis showed that the regression equation had a sensitivity of 90.9% (95% CI: 0.587, 0.998), a specificity of 92.3% (95% CI: 0.848, 0.969), and an area under the curve of 0.955 (95% CI: 0.899, 1.000).Conclusions Patients with supination-external rotation ankle fractures with the above risk factors are at a higher risk of developing preoperative LEDVT. Orthopedic nursing staff should pay close attention to these risk factors and implement targeted preventive measures before surgery to minimize the risk of LEDVT.