Risk Prediction Models And Risk Factors For Early Heart Failure After Acute Myocardial Infarction

Objective: Heart failure is the most common fatal complication of myocardial infarction.The incidence of heart failure after myocardial infarction is still high,and concurrent heart failure increases the mortality and the number of readmissions.There are no large-scale clinical studies on the risk factors for heart failure in the early(30days)after acute myocardial infarction,and there is currently no systematic risk assessment system in China.This article collects and measures relevant clinical indicators of patients with acute myocardial infarction,studies and analyzes potential predictors of heart failure in the early(30 days)after acute myocardial infarction,explores the value of new predictive factors for early heart failure,and attempts to establish a simple risk assessment model has a certain predictive value for the occurrence of early heart failure after acute myocardial infarction.Methods: A total of 318 patients who were admitted to the Second Hospital of Lanzhou University for acute myocardial infarction from December 2017 to October2019 were included in this study.The data collected includes general demographic data,clinical data,laboratory data,past medical history,and 14 possible new predictors.All patients included in the study were followed up for at least 30 days on an outpatient or telephone basis.The follow-up content was mainly heart failure-related symptoms that occurred within 30 days after myocardial infarction.The patients were divided into heart failure group(89 cases)and non-heart failure group(229 cases)according to the follow-up results.We used univariate Logistic regression analysis for the screening of various data to analyze the potential predictors related to early heart failure after acute myocardial infarction.ROC curve is used to calculate the optimal cutoff value for the continuous variables that are selected.We use the Lasso-logistic regression model to further filter the variables to obtain the variables with the best predictive characteristics and incorporate them into the multi-factor logistic regression model.A nomogram of early heart failure after acute myocardial infarction was constructed based on the Logisticmultivariate regression model.The prediction model was evaluated by calculating the consistency index(C-index),the area under the receiver’s working characteristic curve(AUC),and drawing the calibration curve of the prediction model.Results:1.Comparison of data between the heart failure group and the non-heart failure group: There were statistically significant differences between the two groups in terms of male,age,smoking history,lung rale within 24 hours of admission,time from symptoms appear to arriving at the hospital,BNP,LVEF,atrial fibrillation,multi-vessel stenosis,preoperative disease,TIMI classification levels before treatment,Galectin-3,IMA,IL-10,sST2(P <0.05).2.Risk factors related to the occurrence of early heart failure after acute myocardial infarction: male(OR: 0.289,95%CI: 0.157~0.532,P<0.001);age(OR:1.115,95%CI: 1.083~1.147,P<0.001);lung rale within 24 hours of admission(OR:6.282,95%CI: 2.625~15.034,P<0.001);time from symptoms appear to arriving at the hospital(OR: 1.067,95%CI: 1.037~ 1.098,P<0.001),BNP(OR: 1.021,95%CI:1.015~1.027,P<0.001);LVEF(OR: 0.751,95%CI: 0.695~0.811,P<0.001),atrial fibrillation;(OR: 32.096,95%CI: 4.109 ~ 250.702,P=0.001)multi-vessel stenosis(OR: 4.523,95%CI: 1.982~10.322,P<0.001);Galectin-3(OR: 1.143,95%CI:1.058~1.235,P=0.001)、sST2(OR: 1.032,95%CI: 1.003~1.062,P=0.032).all risk factors P < 0.05.3.Using Lasso-logistic regression results,the optimal variables included in the model were screened: gender,age,lung rale within 24 hours of admission,time from symptoms appear to arriving at the hospital,BNP,LVEF,atrial fibrillation,multi-vessel stenosis.Based on multivariate logistic regression,a Nomgram of risk factors related to early heart failure after acute myocardial infarction was constructed,and gender,age,lung rale within 24 hours of admission,time from symptoms appear to arriving at the hospital,BNP,LVEF,atrial fibrillation,multi-vessel stenosis included in the model,and name it "B-FAILURE" risk prediction model.4.The consistency index(C-index)of the established prediction model was calculated to be 0.911,and the area under the receiver’s working characteristic curve(AUC)is 0.911,indicating that the model has a high degree of discrimination and has a higher predictive power for the occurrence of early heart failure after acute myocardial infarction.And the Calibration curve showed the prediction probability of the model is consistent with the observation probability.Conclusions:1.The results of this study indicate that the independent risk factors related to early heart failure after acute myocardial infarction include:gender,age,lung rale within 24 hours of admission,time from symptoms appear to arriving at the hospital,BNP,LVEF,atrial fibrillation,multi-vessel stenosis,Galectin-3,sST2.2.This study established a simple risk prediction model to predict the occurrence of heart failure in the early(30 days)after acute myocardial infarction.The area under the working characteristic curve(AUC)of the subjects was calculated to be 0.911,and the consistency index(C-index)is 0.911,indicating that the model has a high degree of discrimination and has a high ability to predict the occurrence of early heart failure after acute myocardial infarction.