基于血浆渗透压建立急性ST段抬高型心肌梗死重症患者发生院内死亡风险预测模型  

Development of A Plasma Osmolality Prediction Model for the Risk of In-hospital Death in Critically Ill Patients with Acute ST-segment Elevation Myocardial Infarction

作　　者：贾吉 陶四明 JIA Ji;TAO Siming(Dept.of Cardiovascular Medicine,Affiliated Hospital of Yunnan University,Kunming Yunnan 650021,China)

机构地区：[1]云南大学附属医院心血管内科,云南昆明650021

出　　处：《昆明医科大学学报》2022年第12期58-65,共8页Journal of Kunming Medical University

基　　金：云南省卫生和计划生育委员会医学后备人才培养计划基金资助项目(H-2017019);云南省“高层次人才培养支持计划”入选名医专项基金资助项目(YNWR-MY-2020-024)。

摘　　要：目的 建立并验证血浆渗透压对急性ST段抬高型心肌梗死(acute ST-segment elevation myocardial infarction,STEMI)重症患者院内死亡风险预测模型。方法 通过回顾性分析患者电子病历资料,选取2015年1月至2020年12月于云南大学附属医院心血管内科住院的STEMI重症患者,提取患者的一般信息、实验室检查、合并疾病及用药情况等,筛选STEMI重症患者院内死亡风险危险因素并建立预测模型。结果 利用LASSO回归及多因素Logistic回归筛选出白蛋白(ALB)、白细胞(WBC)、血小板(PLT)、血肌酐(Scr)、是否服用他汀类药物、是否服用血管紧张素转化酶抑制剂(ACEI)或血管紧张素转化酶受体拮抗剂(ARB)类药物、血浆渗透压为STEMI重症患者是否发生院内死亡的独立预测因子(P <0.05),据此7个预测变量绘制列线图,模型检验结果显示其区分度、校准度较好,决策曲线分析(DCA)显示阈值概率5%~95%时,临床使用该模型是可以获益的。结论 构建的包含7个变量预测模型具有较好的区分度和校准度,可作为评估STEMI重症患者院内死亡风险参考工具。Objective To develop and validate a plasma osmolality prediction model for the risk of inhospital death in critically ill patients with acute ST-segment elevation myocardial infarction. Methods By retrospective analysis of patients’ electronic medical record data, patients with severe STEMI admitted to the Department of Cardiovascular Medicine,Affiliated Hospital of Yunnan University from January 2015 to December2020 were selected. General information, laboratory examination, comorbid diseases and medications of the patients were extracted to screen the risk factors of in-hospital death of severe STEMI patients and establish a prediction model. Results LASSO regression and multivariate Logistic regression were used and identified albumin(ALB),leukocyte(WBC),platelet(PLT),serum creatinine(Scr),statins medication,angiotensin-converting enzyme inhibitors(ACEI) medication or angiotensin-converting enzyme receptor antagonists(ARB),and plasma osmosis Pressure as independent predictor of in-hospital death in severe STEMI patients(P < 0.05). A nomogram was drawn based on the 7 predictive variables.Model test results showed good differentiation and calibration,and decision curve analysis (DCA) showed a threshold probability of 5%-95%.Conclusion The prediction model has good discrimination and calibration,and can be used as a reference tool for assessing the risk of in-hospital mortality in critically ill patients with STEMI.

关 键 词：急性ST段抬高型心肌梗死 院内死亡 预测模型 列线图 

分 类 号：R541.4[医药卫生—心血管疾病]