食管癌根治术后肺部感染预测模型的构建  

Construction of a Predictive Model for Lung Infection after Radical Resection of Esophageal Cancer

作　　者：李翠花 钱鑫磊[1] 李莲业 谢占玉[1] LI Cui-hua;QIAN Xin-lei;LI Lian-ye;XIE Zhan-yu(Qinghai Provincial People’s Hospital,Xining,China,810001)

机构地区：[1]青海省人民医院,青海西宁810001

出　　处：《食管疾病》2023年第2期117-121,共5页Journal of Esophageal Diseases

摘　　要：目的构建食管癌根治术后肺部感染的临床预测模型。方法收集整理2015年1月至2021年1月在青海省人民医院胸外科行食管癌根治术患者的临床资料,将所有患者随机分为训练集和验证集。在训练集中,应用单因素Logstic回归初步筛选食管癌术后影响因素,基于初步筛选的结果,应用多因素Logstic回归模型构建食管癌术后肺部感染预测模型,并且构建诺莫图。应用ROC分析在验证集中验证该模型的预测效能。结果本研究共收集126例食管癌术后的患者,其中训练集74例,验证集52例。单因素Logstic回归结果表明,吸烟(P=0.004)、合并糖尿病(P=0.007)、慢性呼吸系统疾病(P=0.014)、术后第一天CRP(P=0.009)和PCT(P=0.032)的升高与食管癌根治术后肺部感染密切相关。通过多因素Logstic回归得到术后肺部感染危险指数计算公式:术后感染危险指数=2.655×吸烟+3.009×合并糖尿病+1.244×合并慢性呼吸系统疾病+0.099×CRP+0.009×PCT-12.398。基于Logstic回归结果构建了诺莫图。验证集中验证结果表明,该模型在外部验证集中仍然具有良好的预测能力(AUC=0.876,P<0.01)。结论本研究成功构建了一个可以预测食管癌根治术后肺部感染的模型,为食管癌术后肺部感染的精准控制提供帮助。Objective To construct a clinical prediction model for postoperative lung infectien after radical resection of esophageal cancer.Methods The clinical data of patients who underwent radical esophagectomy in the department of thoracic surgery of our hospital from January 2015 to January 2021 were collected and sorted out.All patients were randomly divided into training set and verification set.In the training set,univariate logistic regression was used to preliminarily screen the influencing factors of postoperative esophageal cancer.Based on the results of preliminary screening,multivariate logistic regression model was used to construct the prediction model of postoperative pulmonary infection of esophageal cancer,and Nomogram was constructed.ROC analysis is applied to verify the prediction efficiency of the model in the verification set.Results A total of 126 postoperative patients with esophageal cancer were collected in,including 74 cases in the training set and 52 cases in the verification set.Univariate Logstic regression results showed that smoking(P=0.004),diabetes mellitus(P=0.007),chronic respiratory disease(P=0.014),CRP(P=0.009)and PCT(P=0.032)on the first day after operation were closely related to pulmonary infection after radical resection of esophageal cancer.The postoperative infection risk index was 2.655×smokin+3.009×diabetes mellitus+1.244×chronic respiratory disease+0.099×CRP+0.09×PCT-12.398 The validation results of the verification set show that the model still has good prediction ability in the external validation set(AUC=0.876,P<0.01).Conclusion This study successfully constructed a model that can predict the pulmonary infection after radical resection of esophageal cancer,which could provide help for the accurate control of pulmonary infection after radical resection of esophageal cancer.

关 键 词：预测模型 食管癌 肺部感染 

分 类 号：R735.1[医药卫生—肿瘤]