文章摘要
腹腔镜结直肠癌根治术患者术后谵妄的预测模型
Prediction model for postoperative delirium in patients after laparoscopic radical colorectal cancer surgery
  
DOI：10.12089/jca.2025.04.002
中文关键词: 结直肠癌  术后谵妄  危险因素  预测模型  列线图
英文关键词: Colorectal cancer  Postoperative delirium  Risk factor  Prediction model  Nomograms
基金项目:江苏省卫生健康委员会面上项目（M2020084）
作者单位E-mail
唐秀芳 225300,南京医科大学附属泰州人民医院手术室  
王宏刚 225300,南京医科大学附属泰州人民医院麻醉科  
邢海林 225300,南京医科大学附属泰州人民医院肠外科 xinghailin_tz@sina.com 
摘要点击次数: 241
全文下载次数: 33
中文摘要:
      
目的：探讨行腹腔镜结直肠癌根治术患者发生术后谵妄（POD）的独立危险因素，并构建个体化预测模型。
方法：回顾性纳入2019年1月至2024年1月全身麻醉下行腹腔镜结直肠癌根治术的结直肠癌患者385例（训练集），男232例，女153例，年龄60～85岁，BMI 15.0～32.0 kg/m 2，ASA Ⅰ—Ⅲ级，并选取2024年2月至2025年1月130例患者作为验证集。采用意识模糊评估法评估患者术后7 d内POD发生情况，将患者分为两组：POD和非POD组。收集资料包括受教育程度、术前1 d MMSE评分、吸烟、酗酒、肿瘤分期、急诊手术、联合脏器切除、手术时间、围术期输血、术中低血压、术中药物使用、术后转ICU、术后自控镇痛等情况，术前1 d血红蛋白（Hb）、白细胞计数（WBC）、C-反应蛋白（CRP）、蛋白/纤维蛋白原比值（AFR），中性粒细胞/淋巴细胞比值（NLR），单核细胞/淋巴细胞比值（MLR），预后营养指数（PNI）和改良虚弱指数（mFI）。采用多因素Logistic回归分析POD的独立危险因素，采用R 4.4.1软件绘制并评估列线图预测模型。采用受试者工作特征（ROC）曲线、校准曲线、决策曲线分析（DCA）进行难证。
结果：术后7 d内训练集有51例（13.2%）患者发生POD，验证集有19例（14.6%）。与非POD组比较，训练集和验证集中POD组年龄、酗酒、急诊手术和术后转 ICU 比例、 CRP 异常比例、NLR、mFI明显升高，手术时间明显延长， AFR、PNI明显降低（P<0.05）。训练集多因素Logistic分析显示，年龄增大（OR=1.107，95%CI 1.015～1.208，P=0.022）、手术时间延长（OR=1.019，95%CI 1.006～1.033，P=0.005）、NLR升高（OR=25.939，95%CI 9.135～73.635，P<0.001），mFI升高（OR=9.097，95%CI 3.816～21.688，P<0.001）是腹腔镜结直肠癌根治术后POD的独立危险因素，AFR升高（OR=0.606，95%CI 0.414～0.888，P=0.010）、PNI升高（OR=0.896，95%CI 0.820～0.981，P=0.017）是独立保护因素。训练集根据这些因素构建的列线图预测模型AUC为0.902（95%CI 0.866～0.939），验证集评估模型AUC为0.933（95%CI 0.891~0.975）。
结论：基于年龄、手术时间、AFR、NLR、PNI和mFI构建的列线图模型可有效预测结直肠癌患者POD的发生。
英文摘要:
      
Objective: To identify independent risk factors for postoperative delirium (POD) in patients with colorectal cancer and to construct an individualized prediction model.
Methods: A total of 385 colorectal cancer patients between January 2019 and January 2024 (training set), 232 males and 153 females, aged 60-85 years, BMI 15.0-32.0 kg/m 2, ASA physical status Ⅰ-Ⅲ, who underwent laparoscopic radical colorectal cancer surgery under general anesthesia were included, and 130 patients between February 2024 and January 2025 were selected as the validation set. POD was assessed using the confusion assessment method within 7 days after surgery, and patients were divided into two groups: the POD group and the non-POD group. Data were collected including education level, MMSE score on the day before surgery, smoking history, alcohol abuse, tumor stage, emergency surgery, combined organ resection, operative duration, perioperative blood transfusion, intraoperative hypotension, intraoperative medication use, postoperative intensive care unit (ICU), and postoperative patient-controlled analgesia. Preoperative laboratory parameters obtained on the day before surgery included hemoglobin (Hb), white blood cell count (WBC), C-reactive protein (CRP), albumin-to-fibrinogen ratio (AFR), neutrophil-to-lymphocyte ratio (NLR), monocyte-to-lymphocyte ratio (MLR), prognostic nutritional index (PNI), and modified frailty index (mFI). Multivariate Logistic regression analysis was used to identify independent risk factors for POD, and a nomogram prediction model was constructed and evaluated using R 4.4.1 software. Receiver operating characteristic (ROC) curve, calibration curve and decision curve analysis (DCA) were used for validation.
Results: In this study, 51 patients in training set (13.2%) developed POD within 7 days after surgery, while 19 patients in validation set (14.6%). Compared with the non-POD group, the POD group in both training and validation sets exhibited significantly higher age, proportion of alcohol abuse, emergency surgery, postoperative ICU transfer, elevated CRP levels, NLR, and mFI, as well as prolonged operative time, while AFR and PNI were significantly lower (P< 0.05). Multivariate Logistic regression analysis in training set identified the following independent risk factors for POD after laparoscopic radical colorectal cancer surgery: advanced age (OR = 1.107, 95% CI 1.015-1.208, P = 0.022), prolonged operative time (OR = 1.019, 95% CI 1.006-1.033, P = 0.005), elevated NLR (OR = 25.939, 95% CI 9.135-73.635, P < 0.001), and elevated mFI (OR = 9.097, 95% CI 3.816-21.688, P < 0.001). Increased AFR (OR = 0.606, 95% CI 0.414-0.888, P = 0.010) and increased PNI (OR = 0.896, 95% CI 0.820-0.981, P = 0.017) were independent protective factors. The nomogram prediction model constructed based on these factors in training set demonstrated an AUC value of 0.902 (95% CI 0.866-0.939), while the validation set demonstrated an AUC of 0.933 (95%CI 0.891-0.975).
Conclusion: The nomogram model constructed based on age, operative time, AFR, NLR, PNI, and mFI can effectively predict the occurrence of POD in colorectal cancer patients.
查看全文   查看/发表评论  下载PDF阅读器
关闭