Logistic Regression Analysis and Prediction Model Construction of Related Influencing Factors of Cervical Anastomotic Leakage After VATS Operation in Elderly Patients with Esophageal Cancer
ZHAO Xiang, GU Yuanyuan, et al
Nantong University Affiliated Hospital, Jaingsu Nantong 226001, China
Abstract:Objective: To investigate the influencing factors of cervical anastomotic leakage after video-assisted thoracoscopic radical resection of lung cancer (VATS) in elderly patients with esophageal cancer, and to establish a logistic regression prediction model. Methods: A total of 178 elderly patients with esophageal cancer who underwent VATS surgery in Nantong University Affiliated Hospital from January 2017 to December 2020 were selected. The incidence of postoperative cervical anastomotic leakage was calculated. A single-factor and multi-factor method was used to describe the postoperative neck of elderly esophageal cancer after VATS. Influencing factors of anastomotic fistula, a Logistic regression prediction model was established, and the specificity and sensitivity of the model for predicting postoperative cervical anastomotic fistula were evaluated. Results: In 178 elderly patients with esophageal cancer who underwent VATS surgery, the incidence of postoperative anastomotic leakage was 14.04%. Logistic regression analysis found that the history of abdominal surgery, hospital stay≥16 d, postoperative lung infection, postoperative ventilator use, and albumin level≥35 g/L on the first day after surgery are VATS in elderly patients with esophageal cancer Factors affecting postoperative cervical anastomotic leakage (P<0.05). Evaluation of the Logistic regression model showed that the model establishment is statistically significant. Wald test showed that the coefficient difference of the regression equation is statistically significant. The Hosmer-Lemeshow goodness-of-fit test showed that the model fits well. Logistic regression model statistical analysis of the data set found that the AUC for predicting postoperative cervical anastomotic leakage was 0.926, the prediction sensitivity was 92.00%, and the specificity was 80.26%. Conclusion: There are many factors related to cervical anastomotic fistula after VATS in elderly patients with esophageal cancer, including history of abdominal surgery, lung infection, length of hospitalization, etc. The Logistic prediction model constructed based on high-risk factors can diagnose cervical anastomotic fistula after VATS in elderly patients with esophageal cancer. It has high predictive value and helps guide clinical treatment.
赵湘, 顾园园, 滕亚莉. 老年食管癌VATS术后颈部吻合口瘘有关影响因素的Logistic回归分析与预测模型构建[J]. 河北医学, 2021, 27(12): 2065-2070.
ZHAO Xiang, GU Yuanyuan, et al. Logistic Regression Analysis and Prediction Model Construction of Related Influencing Factors of Cervical Anastomotic Leakage After VATS Operation in Elderly Patients with Esophageal Cancer. HeBei Med, 2021, 27(12): 2065-2070.