个人化预测甲状腺癌患者术后出血风险的列线图模型的建立  

Establishment of Nomogram Model for Personalized Prediction of Postoperative Hemorrhage Risk of Thyroid Cancer Patients

作　　者：郑艳 张茜[1] 孙菲[1] ZHENG Yan;ZHANG Qian;SUN Fei(Department of Clinical Laboratory Center,Beijing Friendship Hospital Affiliated to Capital Medical University,Beijing 100050,China)

机构地区：[1]首都医科大学附属北京友谊医院临床检验中心,北京100050

出　　处：《医学综述》2021年第3期609-613,共5页Medical Recapitulate

摘　　要：目的探讨甲状腺癌患者术后发生出血的危险因素并建立个人化预测甲状腺癌术后出血风险的列线图模型。方法回顾性分析2016年12月至2019年12月于首都医科大学附属北京友谊医院行手术治疗的2 682例甲状腺癌患者的临床资料,分别采用单因素和多因素Logistic回归分析甲状腺癌术后出血的独立危险因素;应用R软件建立预测甲状腺癌术后出血风险的列线图模型;采用Bootstrap法对模型进行内部验证,并绘制受试者工作特征曲线(ROC曲线)探索列线图模型对甲状腺癌术后出血风险的预测效率。结果 2 682例甲状腺癌患者中术后出血38例,术后出血多发生于术后24 h内,以6 h内最为集中[76.32%(29/38)],出血部位以皮瓣及经深筋膜浅层最为常见[63.16%(24/38)]。合并高血压、甲状腺全切术、肿瘤分期≥Ⅲ期、肿瘤大小≥4 cm、喉返神经浸润及淋巴结转移是甲状腺癌术后出血的独立危险因素(OR=12.330,95%CI 3.354~45.323;OR=6.120,95%CI1.714~21.861;OR=4.137,95%CI 1.510~11.337;OR=5.118,95%CI 1.903~13.767;OR=16.477,95%CI2.007~135.281;OR=5.296,95%CI 1.816~15.445)(P <0.01)。列线图模型预测甲状腺癌患者术后出血风险的一致性指数为0.843,区分度良好;校正曲线趋近于标准曲线,预测值与实测值基本一致。ROC曲线显示曲线下面积为0.870(95%CI 0.838~0.902)。结论甲状腺癌患者术后出血以6 h内最为集中,出血部位以皮瓣及经深筋膜浅层最为常见。合并高血压、甲状腺全切术、肿瘤分期≥Ⅲ期、肿瘤大小≥4 cm、喉返神经浸润及淋巴结转移是甲状腺癌术后出血的独立危险因素,据此建立的列线图预测模型具有可靠的预测能力。Objective To investigate the risk factors of postoperative bleeding in patients with thyroid cancer and to establish a personalized nomogram model to predict the risk of postoperative bleeding.Methods A retrospective analysis was performed on the clinical data of 2 682 patients with thyroid cancer undergoing surgery for thyroid cancer in Beijing Friendship Hospital Affiliated to Capital Medical University from Dec.2016 to Dec.2019.Univariate and multivariate Logistic regression analyses were used to analyze the independent risk factors for postoperative bleeding of thyroid cancer.R software was used to establish a nomogram model to predict the risk of postoperative bleeding of thyroid carcinoma.The model was validated internally by Bootstrap method,and the receiver operating characteristic curve( ROC curve) was developed to explore the predictive efficiency of the nomogram model.Results Altogether there were 38 cases of postoperative hemorrhage in the 2 682 patients with thyroid carcinoma,mostly occurred within 24 h after surgery,and the most concentrated within 6 h[76.32%( 29/38) ],and the most common site of hemorrhage was the flap and the superficial layer of deep fascia[63.16%( 24/38) ].Hypertension,thyroid full cut,tumor staging≥Ⅲ stage,tumor size≥4 cm,laryngeal recurrent nerve invasion and lymph node metastasis were the independent risk factors of thyroid carcinoma postoperative bleeding( OR =12.330,95% CI 3.354-45.323;OR = 6.120,95% CI 1.714-21.861;OR = 4.137,95% CI 1.510-11.337;OR = 5.118,95% CI1.903-13.767;OR = 16.477,95% CI 2.007-135.281;OR = 5.296,95% CI 1.816-15.445)( P < 0.01).The consistency index for predicting the risk of postoperative bleeding in patients with thyroid cancer by the nomogram model was 0.843,with good differentiation;the calibration curve was close to the standard curve,and the predicted value was basically consistent with the measured value.The area under the ROC curve was 0.870( 95% CI 0.838-0.902).Conclusion Patients with thyroid cancer have the most concentrated postopera

关 键 词：甲状腺癌 术后出血 危险因素 列线图模型 

分 类 号：R736.1[医药卫生—肿瘤]