文章摘要
腹部手术患者术后谵妄的危险因素分析及列线图预测模型构建
Risk factor analysis and nomogram prediction model establishment for postoperative delirium in patients undergoing abdominal surgery
  
DOI：10.12089/jca.2026.05.002
中文关键词: 术后谵妄  腹部手术  危险因素  回归分析  预测模型
英文关键词: Postoperative delirium  Abdominal surgery  Risk factor  Regression analysis  Prediction model
基金项目:叙永县人民医院-西南医科大学科技战略合作项目（2023XYXNYD15）
作者单位E-mail
甘云帆 646000,四川省泸州市,西南医科大学附属医院麻醉科  
郭美洁 泸州市叙永县人民医院麻醉科  
王威燎 646000,四川省泸州市,西南医科大学附属医院麻醉科  
杨晓玲 646000,四川省泸州市,西南医科大学附属医院麻醉科  
王茂华 646000,四川省泸州市,西南医科大学附属医院麻醉科  
王晓斌 646000,四川省泸州市,西南医科大学附属医院麻醉科  
伍佳莉 646000,四川省泸州市,西南医科大学附属医院麻醉科 13982759964@163.com 
摘要点击次数: 848
全文下载次数: 207
中文摘要:
      
目的：探讨腹部手术患者术后谵妄（POD）的危险因素，并构建腹部手术患者POD的预测模型。
方法：回顾性收集2023年1月至2024年7月行腹部手术的患者资料作为训练集，另选择2024年8月至2025年3月行腹部手术的患者资料作为验证集。经单因素及多因素Logistic回归分析筛选独立危险因素，构建列线图预测模型。绘制受试者工作特征（ROC）曲线并计算曲线下面积（AUC），评估模型的预测效能。采用校准曲线和DCA曲线验证模型的准确性和临床实用性。
结果：训练集中有128例（33.8%）发生POD，验证集中有25例（30.5%）发生POD。多因素Logistic回归分析显示，年龄偏大（OR=1.111，95%CI 1.067～1.156）、男性（OR=2.690，95%CI 1.447～4.998）、饮酒史（OR=2.095，95%CI 1.120～3.918）、红细胞计数偏低（OR=1.661，95%CI 1.086～2.538）、尿素偏高（OR=1.130，95%CI 1.023～1.248）、血糖偏高（OR=1.132，95%CI 1.014～1.264）、术中低体温（OR=2.888，95%CI 1.602～5.209）以及术后转入ICU（OR=4.095，95%CI 2.001～8.379）为腹部手术患者发生POD的独立危险因素。基于独立危险因素构建的列线图预测模型在训练集中的AUC为0.818（95%CI 0.774～0.863），敏感性为0.875，特异性为0.610；在验证集中的AUC为0.724（95%CI 0.603～0.844），敏感性为0.600，特异性为0.754。校准曲线和DCA曲线显示模型具有具有较好的一致性和临床实用性。
结论：年龄偏大、男性、饮酒史、红细胞计数偏低、尿素偏高、血糖偏高、术中低体温以及术后转入ICU为腹部手术患者发生POD的危险因素。基于独立危险因素构建的列线图预测模型预测性能良好，具有较好的一致性和临床实用性。
英文摘要:
      
Objective: To investigate the risk factors for postoperative delirium (POD) in patients undergoing abdominal surgery and to establish a nomogram prediction model.
Methods: Patients who underwent abdominal surgery from January 2023 to July 2024 were retrospectively collected as the training set, and patients who underwent abdominal surgery from August 2024 to March 2025 were collected as the validation set. Univariate and multivariate logistic regression analyses were applied to identify independent risk factors, and a nomogram prediction model was established. Receiver operating characteristic (ROC) curves were plotted, and the area under the curve (AUC) was calculated to evaluate predictive performance. Calibration curves and decision curve analysis (DCA) were used to validate the accuracy and clinical utility of the model.
Results: A total of 128 patients (33.8%) developed POD in the training set, and 25 patients (30.5%) developed POD in the validation set. Multivariate logistic regression analysis identified advanced age (OR = 1.111, 95% CI 1.067-1.156), male (OR = 2.690, 95% CI 1.447-4.998), history of alcohol consumption (OR = 2.095, 95% CI 1.120-3.918), decreased red blood cell count (OR = 1.661, 95% CI 1.086-2.538), elevated urea level (OR = 1.130, 95% CI 1.023-1.248), hyperglycemia (OR = 1.132, 95% CI 1.014-1.264), intraoperative hypothermia (OR = 2.888, 95% CI 1.602-5.209), and postoperative transfer to ICU (OR = 4.095, 95% CI 2.001-8.379) as risk factors for POD in patients undergoing abdominal surgery. The nomogram prediction model based on independent risk factors achieved an AUC of 0.818 (95% CI 0.774-0.863) in the training set, with a sensitivity of 0.875 and a specificity of 0.610; and an AUC of 0.724 (95% CI 0.603-0.844) in the validation set, with a sensitivity of 0.600 and a specificity of 0.754. The calibration curve and DCA show that the model has good consistency and clinical practicability.
Conclusion: Advanced age, male, history of alcohol consumption, decreased red blood cell count, elevated urea level, hyperglycemia, intraoperative hypothermia, and postoperative transfer to ICU are risk factors for POD in patients undergoing abdominal surgery. The nomogram prediction model constructed based on independent risk factors has good predictive performance, and possesses good consistency and clinical practicability.
查看全文   查看/发表评论  下载PDF阅读器
关闭