Page 126 - 南京医科大学学报自然科学版

南京医科大学学报（自然科学版） 第43卷第2期
·268 · Journal of Nanjing Medical University（Natural Sciences） 2023年2月

·公共卫生与预防医学·

食管癌术后颈部吻合口瘘危险因素分析及预测模型的建立

周 瑾 ，马红霞 1*
1，2
南京医科大学公共卫生学院，江苏 南京 211166；南京医科大学附属肿瘤医院感染管理办公室，江苏省肿瘤医院，江苏省
1 2
肿瘤防治研究所，江苏 南京 210009

［摘 要］ 目的：探讨影响食管癌患者术后颈部吻合口瘘的危险因素，构建并验证吻合口瘘发生风险的列线图模型。方法：回
顾性分析2019年1月—2022年5月南京医科大学附属肿瘤医院行食管癌颈部吻合术的362例患者的临床资料，对术后发生颈
部吻合口瘘相关危险因素进行单因素和多因素Logistic回归分析，基于独立危险因素构建列线图模型预测颈部吻合口瘘发生
风险，通过受试者工作特征曲线（receiver operating characteristic curve，ROC）及曲线下面积（area under curve，AUC）、Calibration
校正曲线验证模型的预测效能。结果：362例食管癌患者术后颈部吻合口瘘发生率11.88%（43/362）。多因素Logistic回归分析
显示，糖尿病、慢性支气管炎、胸腹部手术史、新辅助治疗、术后肺部感染是食管癌术后颈部吻合口瘘的独立危险因素（P＜
0.05）。通过整合这5个因素构建预测颈部吻合口瘘风险的列线图模型，该预测模型的AUC为0.844（95%CI 0.771~0.918），Cal⁃
ibration校正曲线显示预测曲线与理想曲线走势基本一致。结论：糖尿病、慢性支气管炎、胸腹部手术史、新辅助治疗、术后肺
部感染是食管癌患者术后颈部吻合口瘘发生的独立危险因素，该列线图模型具有良好的预测效能，可为临床评估患者病情采
取干预措施提供参考。
［关键词］ 食管癌；吻合口瘘；危险因素；列线图；预测模型
［中图分类号］ R735.1 ［文献标志码］ A ［文章编号］ 1007⁃4368（2023）02⁃268⁃08
doi：10.7655/NYDXBNS20230218

Analysis of risk factors and establishment of a prediction model for the cervical
anastomotic leakage after esophagectomy

1，2 1*
ZHOU Jin ，MA Hongxia
1 Department of Public Health，Nanjing Medical University，Nanjing 211166；Department of Infection Management，
2
Nanjing Medical University Affiliated Cancer Hospital，Jiangsu Cancer Hospital，Jiangsu Institute of Cancer
Reaserch，Nanjing 210009，China

［Abstract］ Objective：The current study aims to explore the risk factors for the cervical anastomotic leakage after esophagectomy，

establish a nomogram prediction model，and test its predictive ability. Methods：A retrospective analysis was performed on clinical
data of 362 patients undergoing radical resection of esophageal carcinoma in Nanjing medical university affiliated cancer hospital
between January 2019 and May 2022. The independent risk factors for postoperative cervical anastomotic leakage were analyzed by
univariate analysis and multivariable logistic regression analysis. Based on these factors，a nomogram model was established to predict
the risk of the cervical anastomotic leakage. The predictive performance of the nomogram was verified by receiver operating
characteristic curve（ROC）and calibration curve. Results：The incidence of the postoperative cervical anastomotic leakage was 11.88%
（43/362）. The multivariable logistic regression analysis showed that diabetes，chronic bronchitis，the history of thoracic or abdominal
surgery，neoadjuvant therapy and postoperative pulmonary infection were the independent high risk factors for cervical anastomotic
leakage after esophagectomy（P＜0.05）. A nomogram prediction model was established based on these factors. The area under the
receiver operator characteristic curve achieved 0.844（95% CI 0.771~0.918）.The calibration curve displayed a general consistency
between the prediction and ideal curves. Conclusion：Diabetes，chronic bronchitis，the history of thoracic or abdominal surgery，
neoadjuvant therapy and postoperative pulmonary infection are the independent high risk factors for the cervical anastomotic leakage
after esophagectomy，and clinical prediction model shows good evaluation efficacy and provides a reference for the assessment and
［基金项目］ 南京医科大学姑苏学院引进人才项目（GSRCKY20210201）
∗
通信作者（Corresponding author），E⁃mail：mahongxia@126.com