Electrocardiographic predictors of disease severity, mortality, and advanced ventilatory support among hospitalized covid-19 patients: a two year single center retrospective, cohort study

Abstract Funding Acknowledgements Type of funding sources: None. Introduction For detecting myocardial injury in severe and critical COVID-19, the electrocardiogram (ECG) is neither sensitive nor specific; but in a resource-poor environment, it remains relevant. Changes in the ECG can be a potential marker of severe and critical COVID-19 to be used for predicting not only disease severity but also the prognosis for recovery. Methods The admitting and interval ECGs of 1,333 COVID-19 patients were reviewed in a two-year, single-center, retrospective cohort study. Each was evaluated for 29 pre-defined ECG patterns under the categories of rhythm, rate, McGinn-White and RV overload patterns, axis and QRS abnormalities, ischemia/infarct patterns, and AV blocks before univariate and multivariate regression analyses for correlation with disease severity; need for advanced ventilatory support; and in-hospital mortality. Results Of the 29 ECG patterns, 18 showed a significant association with the dependent variables on univariate analysis. Multivariate analysis revealed that atrial fibrillation, HR >100 bpm, low QRS voltage, QTc >500msec, diffuse nonspecific T-wave changes, and "any AMI" ECG patterns correlate with disease severity; need for advanced ventilatory support and in-hospital mortality. S1Q3 and S1Q3T3 increased the odds of critical disease and need for high oxygen requirement by 2.5-3 fold. Fractionated QRS increased odds of advanced ventilatory support. Conclusion The ECG can be useful for predicting the severity and outcome of more than moderate COVID-19. Their use can facilitate rapid triage, predict disease trajectory, and prompt a decision to intensify therapy early in the disease to make a positive impact on clinical outcomes. Clinicodemographic profile of COVID19 ECG patterns vs outcomes