前纵隔肿瘤患者术后呼吸和循环系统并发症的风险预测模型

Risk prediction models for postoperative respiratory and circulatory complications in patients with anterior mediastinal tumors

目的构建预测前纵隔肿瘤患者术后呼吸和循环系统并发症发生风险的列线图模型。方法收集2019年1—12月于上海市胸科医院择期行全身麻醉下前纵隔肿瘤切除术且年龄≥18岁的患者499例。根据是否发生术后呼吸和(或)循环系统并发症将患者分为并发症组和非并发症组。采用单因素分析和多因素logistic回归分析并发症发生的影响因素,根据影响因素建立预测前纵隔肿瘤患者术后呼吸和循环系统并发症的列线图模型,绘制ROC曲线,采用AUC评价列线图的预测效能。计算约登指数,确定列线图模型的风险概率界值。结果499例患者中有57例(11.4%)发生呼吸和(或)循环系统并发症。多因素logistic回归分析显示,年龄、主气道受压面积≥50%、限制性通气障碍、心包积液、手术方式、手术时间是前纵隔肿瘤手术后呼吸和循环系统并发症发生的影响因素。列线图模型显示,总分越高,术后并发症的发生风险越高。ROC曲线分析显示,列线图模型的AUC为0.842(95%CI:0.780~0.905,P<0.01)。最大约登指数为0.555,其对应列线图模型的风险概率最佳界值为0.193。结论此列线图模型的构建,能较为准确地预测前纵隔肿瘤患者术后呼吸和循环系统并发症的发生风险。

Objective To construct a nomogram model to predict the risk of postoperative respiratory and circulatory complications in patients with anterior mediastinal tumors.Methods A total of 499 patients aged≥18 years who underwent anterior mediastinal tumor resection under general anesthesia in Shanghai Chest Hospital from January to December 2019 were enrolled in this study.They were divided into complication group and non-complication group according to the occurrence of postoperative respiratory and/or circulatory complications.Univariate analysis and multivariate logistic regression analysis were used to analyze influence factors of the complications.A nomogram model was established to predict probability of the occurrence of postoperative respiratory and circulatory complications in patients with anterior mediastinal tumor based on these influence factors.The area under the receiver operating characteristic(ROC)curve(AUC)was used to evaluate the predictive efficacy of the nomogram.The cut-off value of risk probability of the nomogram model was determined by the Yoden index.Results Respiratory and/or circulatory complications occurred in 57(11.4%)patients.Multivariate logistic regression analysis showed that age,preoperative compression area of the main airway≥50%,restricted ventilation obstruction,pericardial effusion,operation method,and operation time were the influence factors for respiratory and circulatory complications after surgery for anterior mediastinal tumors.The higher total score,the higher risk of postoperative complications,according to the nomogram model.ROC curve analysis showed that the AUC of the nomogram model was 0.842(95%CI:0.780-0.905,P<0.01).The largest Yoden index was 0.555,and the optimal cut-off of risk probability of the corresponding nomogram model was 0.193.Conclusion The nomogram model can efectively predict the risk of postoperative respiratory and circulatory complications in patients with anterior mediastinal tumor.