A Risk Model For Predicting Fatal Ventricular Arrhythmia In Patients With Acute Myocardial Infarction

Background Ischemia-related ventricular tachycardia(VT)/ventricular fibrillation(VF)is the leading cause of sudden cardiac arrest and major complication of acute myocardial infarction(AMI).There were lots of previous studies about risk factors VT/VF in patients with AMI.But there is still lacking effective risk predictive model for VT/VF in patients with AMI.Objectives The aim of this study was to establish a potent risk predictive model that could predict the probability of VT/VF in patients with AMI and provide a reference for clinician.Methods AMI patients who experienced VT/VF during hospitalization admitted to the First Affiliated Hospital of Shantou University Medical College from December 2014 to June 2020 included in this study,whereas AMI patients hospitalized in same period with no VT/VF included.Overall,41 demographic characteristics,serum biomarkers,ECG on admission and Procedure-related factors of the study patients were analyzed.Regarding establishment of the predictive model,we made use of the least absolute shrinkage and selection operator regression to choose variables,and then utilized multiple factor logistic regression to build a risk predictive model,as well as a nomogram to visually display this model.To validate the model,we used the area under the receiver operating characteristic(ROC)curve(AUC)to evaluate the discriminating capability,utilized the calibration curve to visually reveal calibration,and used decision curve analysis(DCA)to assess the clinical validity.Results In total,4315 AMI patients were hospitalized in the First Affiliated Hospital of Shantou University Medical College from December 2014 to June 2020.Among them,114 had VT/VF,whereas 209 had no VT/VF in same period.The 323 AMI patients were divided into two groups,at a ratio of 7 to 3 by random sampling,comprised of a training and a validation set respectively.The least absolute shrinkage and selection operator regression was used to select variables,while multivariate logistic regression was used to establish a predictive model shown by a nomogram.The predictive model included white blood cell(WBC)counts,serum potassium(K),serum creatinine(Scr),serum uric acid(UA),Killip class,left ventricular ejection fraction(LVEF),onset to balloon time,and atrial fibrillation(AF)as predictors.The AUCs of the predictive model and that of the internal validation set were 0.848 [95%confidence interval(CI)0.799–0.897] and 0.827(95% CI 0.756–0.899),respectively.The predictive model showed a good degree of fit,and DCA showed the predictive model was clinically effective.Conclusions A potent risk predictive model was established.Timely percutaneous coronary intervention(PCI),lowering of UA levels and appropriate control for K and AF may significantly reduce the risk of VT/VF in AMI patients.