摘要
目的 探究急性心肌梗死患者经皮冠状动脉介入治疗(percutaneous coronary intervention,PCI)术后发生对比剂肾病(contrast induced nephronpathy,CIN)的风险,并构建列线图模型。方法 回顾性选取2021年5月至2023年3月收治于六盘水首钢水钢医院的144例急性心肌梗死患者作为研究对象,按照2∶1的比例随机分为建模组(103例)和验证组(41例),将建模组按照术后是否发生CIN分为CIN组(38例)和无CIN组(65例),采用LASSO回归模型和10倍交叉验证法获得术后CIN的最佳风险预测因素,并使用多因素logistic回归分析影响患者术后CIN的独立危险因素。构建列线图模型并对模型进行评价。结果 术后CIN发生率为36.89%(38/103)。年龄、糖尿病、左心室射血分数(left ventricular ejection fraction,LVEF)、对比剂用量、血肌酐(serum creatinine,Scr)是患者术后发生CIN的独立危险因素(P <0.05)。利用建模组和验证组构建的受试者工作特征曲线(ROC)和校准曲线均显示列线图模型预测效能良好。危险分层系统将患者分为4个亚组:极低风险组(总分<65分)、低风险组(65分≤总分<150分)、中风险组(150分≤总分<200分)和高风险组(总分≥200分),不同风险评分患者CIN发生率差异有统计学意义(χ^(2)=17.495,P=0.001)。结论 通过年龄、糖尿病、LVEF、对比剂用量、Scr等构建的列线图模型对术后CIN具有良好的预测价值。
Objective To explore the risk of contrast-induced nephropathy(CIN) after percutaneous coronary intervention in patients with acute myocardial infarction,and to construct a corresponding model.Methods 144 patients with acute myocardial infarction who had admitted to Liupanshui shenggang shuigang Hospital from May 2021 to March 2023 were retrospectively selected as research objects,and randomly divided into a modeling group(103 patients) and a verification group(41 patients) according to a 2:1 ratio.The patients in the modeling group were divided into a CIN group(38 patients) and a CIN group(65 patients) according to whether CIN occurred after surgery.The LASSO regression model and 10-fold cross-validation method were used to obtain the best risk predictors of CIN after the procedure,and multiple logistic regression was used to analyze the independent risk factors affecting CIN.The nomogram model was constructed and evaluated.Results The postoperative incidence of CIN was 36.89%(38/103).Age,diabetes mellitus,LVEF,dosage of contrast agent and Scr were independent risk factors for CIN(P<0.05).The receiver operating characteristic curve(ROC) and calibration curve constructed by the data sets from the modeling group and validation group showed better predictive efficiency of the line graph model.The risk stratification system divided patients into four subgroups:very low risk group(total score <65),low risk group(total score <150),medium risk group(total score <200) and high risk group(total score 200).There were significant differences in the incidence of CIN among the patients with different risk scores (χ^(2) = 17.495,P = 0.001).Conclusions The line graph model constructed by age,diabetes,LVEF,dosage of contrast agent,Scr and other indicators has better predictive value for CIN after percutaneous coronary intervention.
作者
杨绍汪
YANG Shaowang(Liupanshui Shougang Shuigang Hospital,Liupanshui 553000,China)
出处
《实用医学杂志》
CAS
北大核心
2023年第15期1925-1931,共7页
The Journal of Practical Medicine
基金
贵州省科技计划项目(编号:20205614)。
关键词
经皮冠状动脉介入治疗
对比剂肾病
急性心肌梗死
列线图
percutaneous coronary intervention
contrast-induced nephronpathy
acute myocardial in-farction
nomograms
