股骨粗隆间骨折患者PFNA术后静脉血栓风险预测模型的构建与验证

Construction and Validation of a Prediction Model for Postoperative Venous Thrombosis after PFNA in Patients with Intertrochanteric Femoral Fractures

摘要: 背景：静脉血栓形成是股骨粗隆间骨折患者围手术期常见并发症之一，严重者可进一步发展为肺栓塞，影响患者术后康复及预后。近端股骨防旋髓内钉(proximal femoral nail antirotation, PFNA)是股骨粗隆间骨折常用内固定方式，但术后静脉血栓仍较常见。目前临床对PFNA术后静脉血栓风险的评估多依赖经验判断，缺乏简便、直观的个体化预测工具。本研究旨在筛选股骨粗隆间骨折患者PFNA术后静脉血栓发生的影响因素，并构建列线图预测模型，为临床早期识别高危患者提供参考。方法：回顾性收集2020年1月至2024年12月青岛大学附属医院收治并接受PFNA内固定治疗的股骨粗隆间骨折患者449例临床资料。按照7:3比例随机分为训练集314例和验证集135例。根据术后是否发生静脉血栓分为血栓组和非血栓组。收集患者一般资料、基础疾病、围手术期资料及实验室检查指标。采用单因素及多因素Logistic回归分析筛选PFNA术后静脉血栓发生的独立影响因素，并据此构建列线图预测模型。采用受试者工作特征曲线及曲线下面积评价模型区分能力，采用校准曲线评价模型预测概率与实际发生概率的一致性，并采用决策曲线分析评价模型临床应用价值。结果：449例患者中术后发生静脉血栓92例，总体发生率为20.49%；其中训练集发生静脉血栓64例，验证集发生静脉血栓28例。单因素Logistic回归分析显示，年龄、糖尿病、脑梗死、受伤至手术时间及卧床时长与PFNA术后静脉血栓发生相关。多因素Logistic回归分析显示，年龄(OR = 1.21, 95%CI: 1.12~1.31, P < 0.001)和卧床时长(OR = 1.49, 95%CI: 1.25~1.78, P < 0.001)是术后静脉血栓发生的独立危险因素；无糖尿病(OR = 0.23, 95%CI: 0.12~0.46, P < 0.001)、无脑梗死(OR = 0.42, 95%CI: 0.21~0.86, P = 0.017)及受伤至手术时间 < 3 d (OR = 0.30, 95%CI: 0.16~0.59, P < 0.001)为保护因素。基于上述变量构建列线图预测模型。ROC曲线显示，模型在训练集中的AUC为0.827 (95%CI: 0.769~0.885)，在验证集中的AUC为0.824 (95%CI: 0.741~0.907)。校准曲线显示，训练集模型预测概率与实际发生概率一致性较好，验证集中低至中等预测风险区间校准效果尚可。决策曲线显示，训练集及验证集中模型均可在一定阈值概率范围内获得临床净获益。结论：年龄、糖尿病、脑梗死、受伤至手术时间及卧床时长是股骨粗隆间骨折患者PFNA术后静脉血栓发生的重要影响因素。基于上述因素构建的列线图预测模型具有较好的区分能力和一定临床应用价值，可为PFNA术后静脉血栓风险的个体化评估提供参考。

Abstract: Background: Venous thrombosis is one of the most common perioperative complications in patients with intertrochanteric femoral fractures. In severe cases, it may progress to pulmonary embolism, adversely affecting postoperative recovery and prognosis. The proximal femoral nail antirotation (PFNA) technique is a widely used internal fixation method for intertrochanteric femoral fractures; however, postoperative venous thrombosis remains relatively common. Currently, assessment of thrombosis risk after PFNA largely relies on clinical experience, and a simple and intuitive individualized prediction tool is lacking. This study aimed to identify the risk factors associated with postoperative venous thrombosis in patients with intertrochanteric femoral fractures treated with PFNA and to develop a nomogram prediction model to facilitate early identification of high-risk patients. Methods: Clinical data from 449 patients with intertrochanteric femoral fractures who underwent PFNA fixation at the Affiliated Hospital of Qingdao University between January 2020 and December 2024 were retrospectively collected. Patients were randomly divided into a training cohort (n = 314) and a validation cohort (n = 135) at a ratio of 7:3. According to the occurrence of postoperative venous thrombosis, patients were classified into a thrombosis group and a non-thrombosis group. Demographic characteristics, comorbidities, perioperative variables, and laboratory indicators were collected. Univariate and multivariate logistic regression analyses were performed to identify independent predictors of postoperative venous thrombosis. A nomogram prediction model was subsequently established based on the identified factors. Model discrimination was evaluated using receiver operating characteristic (ROC) curves and the area under the curve (AUC). Calibration curves were used to assess the agreement between predicted and observed probabilities, while decision curve analysis (DCA) was performed to evaluate the clinical utility of the model. Results: Among the 449 patients, postoperative venous thrombosis occurred in 92 cases, yielding an overall incidence of 20.49%. Of these, 64 cases occurred in the training cohort and 28 in the validation cohort. Univariate logistic regression analysis demonstrated that age, diabetes mellitus, cerebral infarction, time from injury to surgery, and duration of bed rest were associated with postoperative venous thrombosis. Multivariate logistic regression analysis identified age (OR = 1.21, 95%CI: 1.12~1.31, P < 0.001) and duration of bed rest (OR = 1.49, 95%CI: 1.25~1.78, P < 0.001) as independent risk factors for postoperative venous thrombosis. Absence of diabetes mellitus (OR = 0.23, 95%CI: 0.12~0.46, P < 0.001), absence of cerebral infarction (OR = 0.42, 95%CI: 0.21~0.86, P = 0.017), and time from injury to surgery < 3 days (OR = 0.30, 95%CI: 0.16~0.59, P < 0.001) were identified as protective factors. A nomogram prediction model was constructed based on these variables. ROC analysis demonstrated that the model achieved an AUC of 0.827 (95%CI: 0.769~0.885) in the training cohort and 0.824 (95%CI: 0.741~0.907) in the validation cohort. Calibration curves showed good agreement between predicted and observed probabilities in the training cohort, while acceptable calibration performance was observed in the low- to moderate-risk range of the validation cohort. Decision curve analysis indicated that the model provided a positive net clinical benefit across a range of threshold probabilities in both cohorts. Conclusions: Age, diabetes mellitus, cerebral infarction, time from injury to surgery, and duration of bed rest are important factors associated with postoperative venous thrombosis in patients with intertrochanteric femoral fractures treated with PFNA. The nomogram prediction model developed based on these factors demonstrated good discriminative performance and potential clinical utility, providing a practical tool for individualized risk assessment of postoperative venous thrombosis.