DOI: 10.1161/circ.148.suppl_1.16189 ISSN: 0009-7322

Abstract 16189: Using Vectorcardiography to Predict Prognosis in Patients With STEMI and Heart Failure

Sidney Perkins, Demetri Monovoukas, Zoey Chopra, Corey Powell, Anuush Vejalla, Vishwaratn Asthana
  • Physiology (medical)
  • Cardiology and Cardiovascular Medicine

Introduction: There is an unmet need for a non-invasive and cost-effective surveillance tool for characterizing the severity and progression of heart failure (HF) to guide evidence-based therapies. We hypothesize that changes over time in unique parameters extracted using vectorcardiography (VKG) have the potential to predict clinical outcomes in patients with HF. The pathophysiology of HF involves abnormal remodeling and a loss of cardiomyocytes, which may alter the strength and direction of the net depolarization signal. These subtle changes can be elucidated using VKG and correlated with prognosis. We hypothesize a similar approach can be utilized for patients post-STEMI as the underlying pathophysiology - loss of myocardium - is comparable and could serve as an initial proof-of-concept.

Methods: We identified 209 patients discharged from Michigan Medicine between 2016 and 2021 with a diagnosis of acute STEMI. For each patient, a single 12-lead EKG pre-STEMI and >1-week post-STEMI were collected. A set of 60 VKG parameters were derived by analyzing features of the QRS complex. We used LASSO regression analysis incorporating clinical variables and VKG parameters to create a predictive model for the composite outcome of mortality or HF at 90 days post-STEMI. Model accuracy was determined by randomly partitioning the data into five folds and running the logistic regression model on all folds but one.

Results: Comorbid diabetes (OR = 3.04, p = 0.06), QTc post-STEMI (OR = 2.44, p <0.05), and P wave onset pre-STEMI (OR = 0.41, p < 0.05) were consistently predictive. A logistic regression with just these variables has a raw accuracy of 80.3% and cross-fold validation accuracy of 79.4%. Future work aims to create one of the largest pre-/post-STEMI ECG databases, with over 1,200 patients, which will be used to corroborate the above results and serve as a tool for other investigators to explore.

Conclusions: ECG changes before and after a STEMI event can be used to assess the risk of mortality or HF and help guide management. The composite outcome appears to be dominated by variables that are primarily associated with arrhythmogenic death. Future work will focus on pump failure alone and determine VKG parameter predictiveness in HF specifically.

More from our Archive