目的 探索改良式空肠造瘘管营养治疗在食管癌术中的应用效果及并发症风险预测模型构建。 方法 回顾性分析的 197 例食管癌患者均在南京医科大学附属淮安第一医院/ 淮安市第一人民医院 2020 年 4 月至 2023 年 10 月期间收集,根据治 疗方法分为两组,对照组(n = 98)采用经鼻插入鼻肠管,试验组(n = 99)采用改良式空肠造瘘管营养治疗,比较两组患者术前、 术后 14 d 的营养指标和并发症。 同时需统计患者术后并发症发生情况,根据有无发生分为两组,即发生组(术后发生并发 症,n = 24),未发生组(术后未发生并发症,n = 173),对单因素分析有差异的指标再进行二元 Logistic 回归分析术后并发症发生 的独立危险因素,并验证二元 Logistic 回归模型的评估效能。 结果 试验组干预后总蛋白、白蛋白、血红蛋白水平高于对照 组,术后并发症发生率更低于对照组,差异具有统计学意义(P<0. 05)。 经单因素分析,年龄、手术时间、饮酒史、吻合部位、营 养风险、干预方式会对术后合并并发症造成影响(P<0. 05)。 而经二元 Logistic 回归分析,年龄≥60 岁、手术时间>240 min、有 饮酒史、吻合部位为颈部、有营养风险、经鼻导管是食管癌患者术后并发症发生的独立危险因素(P<0. 05)。 根据危险因素建 立评估模型,所有研究的样本分为训练集和验证集,其中验证集,评价二元 Logistic 回归模型的评估效能,AUC 为 0. 840, 95%CI 为 0. 766~ 0. 913,特异度为 75. 10%,敏感度为 79. 20%;训练集的 AUC 为 0. 828,95% CI 为 0. 711 ~ 0. 944,特异度为 98. 84%,敏感度为 66. 67%,表明该模型具有良好区分度。 结论 改良式空肠造瘘管营养治疗相比于经鼻插入鼻肠管,更能够 改善食管癌患者术后营养状态。 但术后仍受到吻合部位、年龄、饮酒史、营养风险、干预方式、手术时间等因素影响,导致并发 症发生,因此需重点关注上述危险因素,进行相应预防、干预措施,从而降低食管癌术后并发症,改善患者预后。

Objective To explore the application effect of modified jejunostomy nutrition support in esophageal cancer and construct the risk prediction model of complications. Method Retrospective analysis of 197 esophageal cancer patients was collected from the First Hospital of Huai 'an affiliated to Nanjing Medical University / First People's Hospital of Huai 'an City from April 2020 to October 2023. The patients were divided into two groups based on treatment methods the control group of 98 cases received nasogastric tube insertion while the experimental group of 99 cases received modified jejunal stoma for nutritional support. The study compared the nutritional indicators and complication rates during the treatment period at 14 days post-intervention. Additionally the incidence of postoperative complications was recorded and divided into two groups the group with complications n = 24 and the group without complications n = 173 . For indicators showing differences in univariate analysis binary Logistic regression was used to identify independent factors influencing postoperative complications and the evaluation effectiveness of the binary logistic regression model was verified. The above procedures were summarized for the training group. Result The total protein albumin and hemoglobin levels in the intervention group were higher than those in the control group after the trial with a lower incidence of postoperative complications compared to the control group and the difference was statistically significant P<0. 05 . Univariate analysis showed that age surgery duration alcohol consumption history anastomosis site nutritional risk and intervention method had an impact on postoperative complications P<0. 05 . Bivariate Logistic regression analysis revealed that age ≥60 years surgery duration < 240 min alcohol consumption history anastomosis site at the neck nutritional risk and nasogastric tube use were independent risk factors for postoperative complications in esophageal cancer patients P<0. 05 . The samples of all the studies are divided into the training set and the validation set. An evaluation model was established based on these risk factors with the training set showing an AUC of 0. 840 a 95%CI of 0. 766 to 0. 913 a specificity of 75. 10% and a sensitivity of 79. 20% the validation set had an AUC of 0. 828 a 95%CI of 0. 711 to 0. 944 a specificity of 98. 84% and a sensitivity of 66. 67% indicating that the model has good discriminative power. Conclusion Compared to nasogastric tube insertion modified jejunal stoma nutrition support can better improve the postoperative nutritional status of esophageal cancer patients. However postoperative complications are still influenced by factors such as nastomosis site age alcohol consumption history nutritional risk intervention method and surgery duration. Therefore it is essential to focus on these risk factors and implement corresponding preventive and intervention measures to reduce postoperative complications in esophageal cancer patients and improve patient outcomes.