肺癌胸腔镜术后并发肺部感染的列线图模型构建

Construction of a nomogram model for lung cancer patients complicated by lung infection after thoracoscopic surgery

目的:分析肺癌患者胸腔镜术后肺部感染的危险因素并建立列线图预测模型。方法:回顾性分析2017年01月至2021年06月期间接受胸腔镜肺切除的肺恶性肿瘤患者。根据术后是否发生肺部感染,参与者被分配到肺部感染组和非肺部感染组。通过多因素Logistic回归模型,识别独立的风险因素,并构建了列线图模型。采用重复抽样1 000次的bootstrap检验对列线图模型进行内部验证。采用一致性指数(concordance index,C-index)和校准曲线来表示模型的预测性能和预测准确度。决策曲线分析(decision curve analysis,DCA)评价该列线图模型的临床应用价值。结果:共有872名符合研究标准的患者纳入了这项研究。最终模型中包括手术部位、肿瘤直径、胸腔引流管留置时间和泛免疫炎症值(pan-immune-inflammation value,PIV)。校准曲线表明该模型拟合较好,受试者工作特征(receiver operating characteristic,ROC)曲线下面积(area under curve,AUC)为0.763(95%CI:0.683~0.842);DCA结果表明该模型具有较高的净获益水平。结论:该研究建立了列线图模型,对肺癌患者胸腔镜术后肺部感染的发生有较好的预测能力及良好的临床应用价值。

Objective:To analyze the risk factors for lung infection after thoracoscopic surgery in lung cancer patients and establish a nomogram prediction model.Methods:A retrospective analysis was performed on patients with lung malignancies who underwent thoracoscopic pulmonary resection from January 2017 to June 2021.Participants were assigned to a group with a lung infection or a group without a lung infection based on whether or not they developed a lung infection after surgery.A multivariate Logistic regression model was used to identify independent risk factors,and a nomogram model was constructed.A bootstrap test with repeated sampling 1 000 times was used to validate the nomogram model internally.The concordance index(C-index) and calibration curve were used to represent the predictive performance and accuracy of the model.Decision curve analysis(DCA) was used to evaluate the clinical application value of this model.Results:A total of 872 patients who met the study criteria were included in the study.The final model included surgical site,tumor diameter,thoracic drainage retention time,and pan-immune-inflammation value(PIV).The calibration curve showed that the model fit well.The area under curve(AUC) of receiver operating characteristic(ROC) was 0.763(95% CI:0.683 ~ 0.842).DCA results showed that the model had a higher net benefit level.Conclusion:In this study,the nomogram model was established,which has better prediction ability and good clinical application value for lung cancer patients with thoracoscopic postoperative lung infection.