| 食管胃结合部腺癌术后严重并发症预测模型的构建及验证 |
| 投稿时间：2025-05-20 修订日期：2025-07-11 点此下载全文 |
| 引用本文：菅启航,王成,李隆乾,秦卓霖,刘重阳侯 明.食管胃结合部腺癌术后严重并发症预测模型的构建及验证[J].医学研究杂志,2025,54(11):147-152 |
| DOI：
10.11969/j.issn.1673-548X.2025.11.026 |
| 摘要点击次数: 33 |
| 全文下载次数: 18 |
|
| 基金项目:兰州大学教学团队培育项目(1zuyxcx-2022-20) |
|
| 中文摘要:目的 探究食管胃结合部腺癌(adenocarcinoma of gastroesophageal junction, AEG)患者在行根治性切除术后发生严重并发症的相关因素,并建立预测模型。方法 回顾性收集2018年9月~2025年1月在兰州大学第二医院接受AEG根治性切除术的576例患者,随机分为训练组和验证组,使用Clavien-Dindo分级系统评估Ⅲ级及以上的并发症,通过多因素Logistic回归分析探讨影响严重并发症的危险因素；构建严重并发症风险预测模型列线图；绘制受试者工作特征(receiver operating characteristic, ROC)曲线来评估模型区分度,通过校准曲线来检验模型的准确性。结果 多因素Logistic回归分析结果表明,泛免疫炎症值(pan immune inflammation value, PIV)＞173.85、术前肌少症、年龄＞61岁、术前合并贫血和手术时间＞342min均是AEG根治性切除术后严重并发症的独立危险因素(P<0.05)。构建的预测模型具有良好的区分能力,训练组和验证组的曲线下面积(area under the curve, AUC)分别为0.840和0.835；校准曲线结果显示,模型的预测概率与实际发生率拟合情况良好(训练组:χ2=4.88,P=0.771；验证组:χ2=7.08,P=0.528)。结论 术前PIV＞173.85、术前肌少症、年龄＞61岁、术前合并贫血、手术时间＞342min可增加AEG患者术后严重并发症的发生率。构建的模型能够有效预测AEG患者术后严重并发症的发生,且模型评价显示出良好的预测效能。 |
| 中文关键词:食管胃结合部腺癌 肌少症 泛免疫炎症值 术后并发症 预测模型 |
| |
| Construction and Validation of Predictive Model for Severe Postoperative Complications After Adenocarcinoma of Gastroesophageal Junction Surgery. |
|
|
| Abstract:Objective To investigate the related factors of severe complications after radical resection of adenocarcinoma of gastroesophageal junction (AEG), and establish a predictive model. Methods A retrospective study included 576 patients who underwent radical AEG resection at the Second Hospital of Lanzhou University from September 2018 to January 2025. Patients were randomly divided into training group and validation group. Complications of grade Ⅲ and above were evaluated using the Clavien-Dindo classification system. Multivariate Logistic regression analysis was used to explore the risk factors influencing severe complications, and Nomogram prediction model for the risk of severe complications was constructed. Model discrimination was assessed using receiver operating characteristic (ROC) curve, and the accuracy was evaluated using calibration curve. Results The results of multivariate Logistic regression analysis revealed that preoperative pan-immune inflammation value (PIV) > 173.85, preoperative sarcopenia, age >61 years, preoperative anemia, and operation time >342min were the independent risk factors for severe complications after radical resection of AEG (P<0.05). The model demonstrated strong discrimination, the area under the curve (AUC) of the training group and the validation group were 0.840 and 0.835, respectively. Calibration curves showed that the predicted probability of the model fited well with the actual incidence rate (training group:χ2=4.88, P=0.771; validation group:χ2=7.08, P=0.528). Conclusion Preoperative PIV>173.85, preoperative sarcopenia, age >61 years, preoperative anemia, and operation time >342 min can increase the incidence of severe postoperative complications in AEG patients. The constructed model can effectively predict the occurrence of severe postoperative complications in AEG patients, and the model evaluation shows good predictive efficacy. |
| keywords:Adenocarcinoma of gastroesophageal junction Sarcopenia Pan immune inflammation value Postoperative complications Predictive model |
| 查看全文 查看/发表评论 下载PDF阅读器 |
|
|
|
