A prediction model for heart failure occurrence after myocardial infarction based on left ventricular ejection fraction and Gensini scores
Weng Zhengyan
Li Kemin
Mei Tingfang
Li Hao
Gao Pengliang
Abstract：Objective To establish a predictive model for occurrence of heart failure(HF)after acute myocardial infarction(AMI)based on left ventricular ejection fraction(LVEF)and Gensini scores.Methods AMI patients(n=180)were chosen and give a retrospective analysis from the First People's Hospital of Bengbu City from Jan.2022 to Jan.2024.All patients were divided,according to HF occurrence,into non-HF group(n=122)and HF group(n=58).The influence factors for HF occurrence after AMI were investigated by using multi-factor Logistic regression analysis.The value of predictive model was analyzed by using ROC curve established based on LVEF and Gensini scores in predicting HF occurrence after AMI.Results The results of multi-factor Logistic regression analysis showed that multi-vessel lesions,Gensini scores>60,heart rate and LVEF all were influence factors of HF occurrence after AMI(P<0.05).The results of ROC curve analysis showed that AUC of multi-vessel lesions,Gensini scores,heart rate and LVEF was,respectively,0.618,0.757,0.693 and 0.742,sensitivity was,respectively,0.293,0.724,0.914 and 1.000,and specificity was,respectively,0.943,0.697,0.443 and 0.500 in predicting HF occurrence after AMI.After included multi-vessel lesions,Gensini scores,heart rate and LVEF into regression equation for obtaining model data,AUC of predictive model was 0.946,sensitivity was 0.845 and specificity was 0.926 in predicting HF occurrence after AMI.Conclusion The influence factors for HF occurrence after AMI include multi-vessel lesions,Gensini scores>60,heart rate and LVEF,and predictive model has a higher value to HF occurrence after AMI established based on LVEF and Gensini scores.
Keywords：Myocardial infarctionHeart failureLeft ventricular ejection fractionGensini scores
Publication Date：2025-03-20
Online Publishing Date：2025-08-15(First online date of this platform, not the publication date of the document)
Pages：5( 313-316,321 )