胸腺瘤切除术后患者发生重症肌无力的风险列线图预测模型构建与验证  

Construction and validation of a risk nomogram prediction model for myasthenia gravis in patients after thymoma resection

作　　者：黑晓欢[1] 卢丹[1] 王小飞[1] Hei Xiaohuan;Lu Dan;Wang Xiaofei

机构地区：[1]首都医科大学附属北京同仁医院胸外科,北京100730

出　　处：《中华现代护理杂志》2022年第21期2884-2890,共7页Chinese Journal of Modern Nursing

基　　金：北京市医管局扬帆计划(临床技术创新项目)(XMLX201839)。

摘　　要：目的构建胸腺瘤切除术后患者发生重症肌无力(MG)的风险列线图预测模型,并对模型的预测效能进行验证。方法采用便利抽样法,选取2018年3月—2021年2月首都医科大学附属北京同仁医院胸外科收治的477例行胸腺瘤切除术的患者为训练集,并按照相同标准选取2017年3月—2018年2月的62例行胸腺瘤切除术患者设为验证集进行回顾性分析。分析纳入患者的临床资料,采用Logistic回归分析探讨胸腺瘤切除术后发生MG的独立危险因素,并构建风险列线图预测模型。结果经胸部X线、CT检查和相关检测确认477例行胸腺瘤切除术患者术后MG发生率为14.05%(67/477)。发生MG和未发生MG的两组患者在合并免疫疾病、术前病程、手术路径、肿瘤是否完整切除、WHO病理分型、术后肺部感染以及术后放化疗等资料上的比较差异均有统计学意义(P<0.05);Logistic回归分析结果表明,合并免疫疾病、手术路径为开胸、肿瘤未完整切除、WHO病理分型为A+AB型、术后肺部感染以及术后未放化疗是胸腺瘤切除术后发生MG的独立危险因素(P<0.05)。基于6项独立危险因素建立胸腺瘤切除术后发生MG的风险列线图预测模型,结果显示,训练集和验证集的C-index分别为0.837(95%CI:0.807~0.867)和0.817(95%CI:0.784~0.850),两集的校正曲线均与理想曲线拟合反映良好,受试者工作特征曲线下面积分别为0.834(95%CI:0.794~0.874)和0.825(95%CI:0.789~0.861)。结论合并免疫疾病、手术路径为开胸、肿瘤未完整切除、WHO病理分型为A+AB型、术后肺部感染以及术后未进行放化疗是行胸腺瘤切除术后发生MG的独立危险因素;基于上述危险因素建立的风险列线图预测模型可准确评估和量化胸腺瘤切除术后MG发生的风险,具有良好的预测能力。Objective To construct a risk nomogram early prediction model for myasthenia gravis(MG)after thymoma resection and to verify the predictive performance of the model.Methods Using the convenient sampling method,a total of 477 patients undergoing thymoma resection who were admitted to Department of Thoracic Surgery in Beijing Tongren Hospital Affiliated to Capital Medical University from March 2018 to February 2021 were selected as the training set.A total of 62 patients who underwent thymoma resection from March 2017 to February 2018 were set as the validation set for retrospective analysis.The clinical data of the included patients were analyzed.Logistic regression analysis was used to explore the independent risk factors of MG after thymoma resection and a risk nomogram prediction model was constructed.Results The incidence of postoperative MG in 477 patients who underwent thymoma resection was confirmed by chest X-ray,CT and related tests,which was 14.05%(67/477).There were statistical differences between the two groups of patients with MG and those without MG in terms of combination of immune diseases,preoperative course of disease,surgical route,complete tumor resection,WHO pathological classification,postoperative pulmonary infection,postoperative radiotherapy and chemotherapy and other data(P<0.05).Logistic regression analysis showed that combined immune disease,thoracotomy,incomplete tumor resection,WHO pathological classification of A+AB,postoperative pulmonary infection and no postoperative chemoradiotherapy were independent risk factors for MG after thymoma resection(P<0.05).Based on 6 independent risk factors,a risk nomogram prediction model of MG after thymoma resection was established.The results showed that the C-index of the training set and the validation set were 0.837(95%CI:0.807-0.867)and 0.817(95%CI:0.807-0.867),respectively.The calibration curves for both sets showed good fit to the ideal curve,with areas under the receiver operating characteristic curves of 0.834(95%CI:0.794-0.874)and 0.825(

关 键 词：胸腺瘤 重症肌无力 危险因素 列线图 预测模型 

分 类 号：R736.3[医药卫生—肿瘤]