目的 构建风湿病患者发生静脉血栓栓塞症（venous thromboembolism,VTE）风险预测模型并验证，为VTE高危人群的筛查提供参考依据。方法 回顾性收集2021年1月至2024年4月在安徽省某三级甲等医院风湿免疫科就诊的606例患者的病历资料。根据是否发生VTE，将其分为VTE组（n=50）和对照组（n=556）。采用单因素和多因素Logistic回归分析风湿病患者发生VTE的危险因素并构建列线图风险预测模型。采用受试者工作特征（receiver operator characteristic,ROC）曲线、校准曲线和决策曲线来评估模型的预测效能。结果 两组患者在年龄、卧床时间≥72 h、VTE史、是否使用激素治疗等方面的差异均有统计学意义（均P＜0.05)；多因素Logistic回归分析显示，卧床时间≥72 h、VTE史、是否使用激素治疗、D-二聚体定量、总胆固醇浓度是风湿病患者发生VTE的危险因素（均P＜0.05）。构建预测模型的ROC曲线下面积为0.930（95%CI0.886~0.973），Hosmer-Lemeshow检验χ2为8.859（P=0.354），一致性指数为0.930。临床决策曲线显示模型可提供显著额外的临床净收益。结论 构建的VTE风险预测模型有较好的预测效率和临床适用性，可在临床推广。

Objective To construct and validate a risk prediction model for venous thromboembolism(VTE) in patients with rheumatic diseases,and to provide a reference for screening of high-risk groups of VTE.Methods The medical records of 606 patients who were treated in the Department of Rheumatology and Immunology of a tertiary A hospital in Anhui Province from January 2021 to April 2024 were retrospectively collected.According to VTE occurrence,they were divided into VTE group(n=50) and control group(n=556).Univariate and multivariate Logistic regression were used to analyze the risk factors of VTE in patients with rheumatic diseases and construct a nomogram risk prediction model.The receiver operator characteristic(ROC) curve,calibration curve and decision curve were used to evaluate the prediction performance of the model.Results There were significant differences in age,bed rest time ≥72 h,history of VTE and whether to use hormone therapy between the two groups(all P＜0.05).Multivariate Logistic regression analysis showed that bed rest time ≥72 h,history of VTE,whether to use hormone therapy,D-dimer quantification and total cholesterol concentration were risk factors for VTE in patients with rheumatic diseases(all P＜0.05).The area under the ROC curve of the prediction model was 0.930(95%CI= 0.886-0.973),the Hosmer-Lemeshow test χ2 was 8.859(P=0.354),and the consistency index was 0.930.The clinical decision curve shows that the model can provide significant additional clinical net benefits.Conclusions The VTE risk prediction model has good prediction efficiency and clinical applicability,and can be promoted in clinical practice.

R473.58

国家自然科学基金青年项目（72304261）；安徽省教育厅科学研究项目资助（2022AH051259）