LAUSR

The global crisis due to coronavirus disease 2019 (COVID19) has caused more than hundred million infections and over 3 000 000 deaths.1 The disease is caused by the severe acute respiratory coronavirus 2 (SARSCoV2). While lung injury and systemic inflammation are main viral disease pathogenesis, myocardial injury is frequent among patients hospitalized with COVID19 and is associated with a poor prognosis.2 Possible mechanisms accounting for myocardial injury included the inhibition of myocardial angiotensinconverting enzyme 2 (ACE2) expression after SARSCoV infection and systemic cytokine storm.2 Although numerous modalities can be used for assessing myocardial injury in the patients with COVID19, electrocardiography (ECG) is ubiquitous, timesaving, and easytoperform. The examination time of 12 leadsurface ECG is 5 to 10 minutes, which is shorter than that by echocardiography, computed tomography, or magnetic resonance image. Remote ECG monitoring by mobile or portable device is feasible. Therefore, physical contact could be reduced to avoid crossinfection. Characteristic ECG manifestations, suggestive of myocardial involvement, predict disease severity and future death.3,4 ECG is also an important tool to monitor QT intervals as some of therapeutic drugs would prolong QT intervals and increase arrhythmia vulnerability. These make ECG a better tool and active research field to screen cardiac injury or risk stratification in the patients with COVID19. In this issue of the journal, Alsagaff et al conducted a systematic review and metaanalysis to appraise the latest evidence of the correlation between ECG on admission and clinical outcomes, including intensive care unit (ICU) admission, severe illness, and mortality in COVID19 patients.5 This work consisted of seven studies with a total of 2539 patients.5 The results revealed that characteristic ECG features on admission were associated with poor clinical outcomes. These ECG features encompassed the longer QTc interval (weighted means difference, WMD: 6.04), a prolonged QTc interval (more than 460500 ms, relative risks, RR: 1.89), longer QRS duration (WMD: 2.03), faster heart rate (WMD: 5.96), higher incidence of LBBB (RR: 2.55), premature atrial contraction (RR: 1.94), premature ventricular contraction (RR: 1.84), Twave inversion (RR: 1.68), and STdepression (RR: 1.61). In brief, ECGs on admission can be used for risk stratification of COVID19 patients and should be closely monitored in the hospitalized patients.5 The present work reaffirmed that the value of ECG abnormality on admission could be used as an early biomarker of poor clinical outcomes. These ECG features might not only be applied as alerting signs but also a potential tool for risk stratification to guide necessary treatment. To be noted, it is not clear whether these ECG features were the manifestation of the patients' preexisting comorbidities, the consequence of secondary hemodynamic changes (hypoxia, electrolytes imbalance, or acidosis), or direct cardiac involvement from COVID19. The differentiation between these pathologies might lead to different treatment strategies. Although numerous ECG changes have been linked to clinical outcomes in the patients with COVID19 in the present work, these features are also common risky ECG features for adverse cardiovascular events. The development of a specific ECG feature that links to direct cardiac involvement of COVID19 might be important to guide COVID19related cardiac therapy. Meanwhile, these ECG features are very nonspecific. Therefore, a significant number of the patients might be considered risky, subjected to overtreatment, and inadequately increase workload of medical staff. The incorporation of the risky ECG features and other cardiac biomarkers such as brain natriuretic peptide as the scoring system might be a better fit to guide clinical treatment of cardiac involvement. The present review mainly analyzed the data from 12 lead surface ECG. As COVID19 is highly contagious, the translation of these 12 lead ECG findings to realtime monitoring by artificial intelligenceassisted portable or mobile device will lead to future transformation of patient care in COVID19. Nonetheless, there are still several limitations in this study. First, due to retrospective study design, the interstudy heterogeneity was the main drawback, leading to insufficiently matched or adjusted confounders. Second, each study in the metaanalysis stated different cutoff values for a prolonged QTc interval. Moreover, Bazett's formula used in most studies may cause overcorrection of QTc interval at higher heart rates, which might overexaggerate the predictive impact on poor clinical outcomes.5 Only ECGs on admission are reviewed, and however, dynamic change of ECGs during serial