急性胰腺炎并发全身炎症反应综合征列线图预测模型的构建  

Construction of a predictive model for acute pancreatitis combined with systemic inflammatory response syndrome

作　　者：曹海明 李昌平 CAO Haiming;LI Changping(Department of Gastroenterology,Affiliated Hospital of Southwest Medical University,Luzhou,Sichuan 646000,China)

机构地区：[1]西南医科大学附属医院消化内科,四川泸州646000

出　　处：《安徽医药》2024年第3期576-580,共5页Anhui Medical and Pharmaceutical Journal

摘　　要：目的 分析急性胰腺炎(AP)病人并发全身炎症反应综合征(SIRS)的风险因素,构建的列线图预测模型。方法 回顾性分析西南医科大学附属医院2020年4月至2022年4月收治的370例AP病人的临床资料,用于建模及内部验证,根据病人是否合并SIRS分为非SIRS组(n=273)和SIRS组(n=97),通过logistic回归分析确定并发SIRS的独立危险因素,同时建立列线图可视化预测模型,计算一致性指数(C-index),检验模型准确性;并探讨列线图模型对AP病人发生SIRS的预测效能。结果AP病人合并SIRS发生率为26.2%;白细胞计数、心率、并发胸腔积液为AP病人并发SIRS的危险因素(OR>1,P<0.05);男性、高密度脂蛋白胆固醇(HDL-C)为AP病人并发SIRS的保护因素(0<OR<1,P<0.05)。基于上述5个危险因素构建列线图,计算公式为Logit(P)=-7.32+性别×(-1.15)+胸腔积液×0.76+心率×0.06+白细胞计数×0.16+HDL-C×(-0.59),C-index为0.845,H-L偏差度检测χ2=6.74(P=0.565),对列线图模型进行内部验证,绘制受试者操作特征曲线(ROC曲线)发现,列线图模型预测AP病人合并SIRS发生风险的曲线下面积(AUC)=0.85。结论 构建的列线图预测模型可为AP病人合并SIRS预测提供有效预测手段。Objective To analyze the risk factors of patients with acute pancreatitis complicated with systemic inflammatory response syndrome(SIRS),and to provide guidance for early clinical diagnosis and treatment.Methods The clinical data of 370 patients with AP admitted to the Department of Gastroenterology,Southwest Medical University Hospital from April 2020 to April 2022 were retrospectively analyzed for modeling and internal validation,and patients were divided into non-SIRS(n=273) and SIRS groups(n=97) according to whether they were complicated with SIRS.The independent risk factors for SIRS were determined by multi-factor logistic regression analysis,and the accuracy of the model was tested by establishing a visual prediction model of the line graph and calculating the consistency index(C-index);and the predictive efficiency of the line graph model on the occurrence of SIRS in patients with AP was investigated.Results The incidence of SIRS in patients with AP was 26.2%;white blood cell count,heart rate,and combined pleural effusion were risk factors for SIRS in patients with AP(OR>1,P<0.05);gender and high-density lipoprotein cholesterol were risk factors for SIRS in patients with AP.(0<OR<1,P<0.05);gender and HDL cholesterol were protective factors for patients with AP complicated by SIRS.The nomogram was constructed based on the above 5 risk factors,and the calculation formula was Logit(P)=-7.32+ sex ×(-1.15) + pleural effusion × 0.76 + heart rate × 0.06 + white blood cell count × 0.16 + HDL-C ×(-0.59).The C-index was 0.845,and the H-L deviance test χ2=6.74(P=0.565),and the internal validation of the column line plot model and the plotting of ROC curves revealed that the area under the curve(AUC) of the column line plot model to predict the risk of developing SIRS in patients with AP combined with SIRS=0.85.Conclusion The column line graph prediction model can provide an effective predictor for the prediction of SIRS in patients with AP.

关 键 词：急性胰腺炎 全身炎症反应综合征 列线图 白细胞数 高密度脂蛋白胆固醇 胸腔积液 

分 类 号：R576[医药卫生—消化系统]