胃癌术后肺部感染列线图预测模型的构建

A nomogram for predicting postoperative pulmonary infection in gastric cancer patients

目的:研究胃癌术后发生肺部感染的相关危险因素,并构建列线图预测模型。方法 :回顾性分析2010~2020年我院行手术治疗的330例胃癌患者的相关临床资料,筛选可能引起术后肺部感染的因素,采用Logistic回归进行危险因素分析,并构建列线图预测模型。采用校正曲线、一致性指数(c指数)和决策曲线分析(DCA)评价预测准确性、鉴别能力和临床有用性。结果:多因素Logistic回归分析显示年龄、吸烟史、慢性呼吸系统疾病史、肺功能锻炼依从性、腹腔镜手术、术中失血量是胃癌术后发生肺部感染的独立预测因素。依据独立危险因素所构建的列线图,校准曲线提示有较好的一致性,C-index为0.878(95%CI:0.801~0.956),ROC曲线下面积(AUC)为0.878,提示该列线图预测模型具有良好的区分度和校准度。DCA曲线表明该列线图有较好的临床获益。结论:根据胃癌术后发生肺部感染的相关独立危险因素构建的列线图预测模型,有助于对高危人群进行围手术期干预,减少术后肺部感染发生率。

Objective:To identifying risk factors of postoperative pulmonary infection(POI)in gastric cancer(GC)patients as well as generating an effective nomogram for the POI.Methods:Patients with gastric cancer after surgery from 1st January 2010 to 31st December 2020 were retrospectively analysed.Their clinical and pathological data were collected.Multiple logistic regression analysis was conducted to identify risk factors for POI and to generate the nomogram prediction model.Predictive accuracy,discriminatory capability,and clinical usefulness were evaluated by calibration curves,concordance index(C-index),and decision curve analysis(DCA).Results:Multivariate regression analysis revealed that age,smoking,chronic respiratory diseases,laparoscopic surgery,intraoperative blood loss and lung function exercise compliance were independent prognostic factors for POI in GC patients.The nomogram had a Cindex of 0.878(95%CI:0.801~0.956).The calibration curves showed good consistency between actual and nomogram-predicted probabilities.The area under the curve(AUC)was 0.878,while the cutoff value was-2.088 with a sensitivity of 80.6%and a specifificity of 82.9%.DCA showed that the nomogram had better clinical usefulness.Conclusions:The present study provides a nomogram developed with perioperative features to predict the incidence of POI in GC patients.