Electrocardiographic Features and Outcome Correlations in 124 Hospitalized COVID-19 Patients with Cardiovascular Events

  • Author Footnotes
    # Pavani Nathala, Vidyulata Salunkhe, Harideep Samanapally, Qian Xu contributed equally and are co-First Authors.
    Pavani Nathala
    Footnotes
    # Pavani Nathala, Vidyulata Salunkhe, Harideep Samanapally, Qian Xu contributed equally and are co-First Authors.
    Affiliations
    Division of Infectious Diseases, Center of Excellence for Research in Infectious Diseases (CERID), University of Louisville, Louisville, KY, USA
    Search for articles by this author
  • Author Footnotes
    # Pavani Nathala, Vidyulata Salunkhe, Harideep Samanapally, Qian Xu contributed equally and are co-First Authors.
    Vidyulata Salunkhe
    Footnotes
    # Pavani Nathala, Vidyulata Salunkhe, Harideep Samanapally, Qian Xu contributed equally and are co-First Authors.
    Affiliations
    Division of Infectious Diseases, Center of Excellence for Research in Infectious Diseases (CERID), University of Louisville, Louisville, KY, USA
    Search for articles by this author
  • Author Footnotes
    # Pavani Nathala, Vidyulata Salunkhe, Harideep Samanapally, Qian Xu contributed equally and are co-First Authors.
    Harideep Samanapally
    Footnotes
    # Pavani Nathala, Vidyulata Salunkhe, Harideep Samanapally, Qian Xu contributed equally and are co-First Authors.
    Affiliations
    Division of Infectious Diseases, Center of Excellence for Research in Infectious Diseases (CERID), University of Louisville, Louisville, KY, USA
    Search for articles by this author
  • Author Footnotes
    # Pavani Nathala, Vidyulata Salunkhe, Harideep Samanapally, Qian Xu contributed equally and are co-First Authors.
    Qian Xu
    Footnotes
    # Pavani Nathala, Vidyulata Salunkhe, Harideep Samanapally, Qian Xu contributed equally and are co-First Authors.
    Affiliations
    Department of Bioinformatics and Biostatistics, School of Public Health and Information Sciences, University of Louisville, Louisville, USA.
    Search for articles by this author
  • Stephen Furmanek
    Affiliations
    Division of Infectious Diseases, Center of Excellence for Research in Infectious Diseases (CERID), University of Louisville, Louisville, KY, USA
    Search for articles by this author
  • Omar F. Fahmy
    Affiliations
    Division of Infectious Diseases, Center of Excellence for Research in Infectious Diseases (CERID), University of Louisville, Louisville, KY, USA
    Search for articles by this author
  • Fnu Deepti
    Affiliations
    Division of Infectious Diseases, Center of Excellence for Research in Infectious Diseases (CERID), University of Louisville, Louisville, KY, USA
    Search for articles by this author
  • Alex Glynn
    Affiliations
    Division of Infectious Diseases, Center of Excellence for Research in Infectious Diseases (CERID), University of Louisville, Louisville, KY, USA
    Search for articles by this author
  • Trevor McGuffin
    Affiliations
    School of Nursing, University of Louisville, Louisville, USA
    Search for articles by this author
  • Dylan C. Goldsmith
    Affiliations
    School of Medicine, University of Louisville, Louisville, USA
    Search for articles by this author
  • Jessica Petrey
    Affiliations
    Kornhauser Health Sciences Library, University of Louisville, Louisville, USA.
    Search for articles by this author
  • Tshura Ali
    Affiliations
    Division of Infectious Diseases, Center of Excellence for Research in Infectious Diseases (CERID), University of Louisville, Louisville, KY, USA
    Search for articles by this author
  • Derek Titus
    Affiliations
    Department of Anesthesiology & Perioperative Medicine, University of Louisville, Louisville, KY, USA
    Search for articles by this author
  • Author Footnotes
    $ Siddharth Pahwa, Ruth Carrico, Julio Ramirez, Sean P. Clifford, Lynn Roser, Maiying Kong, and Jiapeng Huang are Senior Authors.
    Ruth Carrico
    Footnotes
    $ Siddharth Pahwa, Ruth Carrico, Julio Ramirez, Sean P. Clifford, Lynn Roser, Maiying Kong, and Jiapeng Huang are Senior Authors.
    Affiliations
    Division of Infectious Diseases, Center of Excellence for Research in Infectious Diseases (CERID), University of Louisville, Louisville, KY, USA
    Search for articles by this author
  • Author Footnotes
    $ Siddharth Pahwa, Ruth Carrico, Julio Ramirez, Sean P. Clifford, Lynn Roser, Maiying Kong, and Jiapeng Huang are Senior Authors.
    Julio Ramirez
    Footnotes
    $ Siddharth Pahwa, Ruth Carrico, Julio Ramirez, Sean P. Clifford, Lynn Roser, Maiying Kong, and Jiapeng Huang are Senior Authors.
    Affiliations
    Division of Infectious Diseases, Center of Excellence for Research in Infectious Diseases (CERID), University of Louisville, Louisville, KY, USA
    Search for articles by this author
  • Demetra Antimisiaris
    Affiliations
    Department of Pharmacology & Toxicology, University of Louisville, Louisville, KY, USA.
    Search for articles by this author
  • Author Footnotes
    $ Siddharth Pahwa, Ruth Carrico, Julio Ramirez, Sean P. Clifford, Lynn Roser, Maiying Kong, and Jiapeng Huang are Senior Authors.
    Sean P. Clifford
    Footnotes
    $ Siddharth Pahwa, Ruth Carrico, Julio Ramirez, Sean P. Clifford, Lynn Roser, Maiying Kong, and Jiapeng Huang are Senior Authors.
    Affiliations
    Department of Anesthesiology & Perioperative Medicine, University of Louisville, Louisville, KY, USA
    Search for articles by this author
  • Author Footnotes
    $ Siddharth Pahwa, Ruth Carrico, Julio Ramirez, Sean P. Clifford, Lynn Roser, Maiying Kong, and Jiapeng Huang are Senior Authors.
    Siddharth Pahwa
    Footnotes
    $ Siddharth Pahwa, Ruth Carrico, Julio Ramirez, Sean P. Clifford, Lynn Roser, Maiying Kong, and Jiapeng Huang are Senior Authors.
    Affiliations
    Department of Cardiovascular & Thoracic Surgery, University of Louisville, Louisville, USA.
    Search for articles by this author
  • Author Footnotes
    $ Siddharth Pahwa, Ruth Carrico, Julio Ramirez, Sean P. Clifford, Lynn Roser, Maiying Kong, and Jiapeng Huang are Senior Authors.
    ,
    Author Footnotes
    ^ Lynn Roser, Maiying Kong and Jiapeng Huang contributed equally to this work.
    Lynn Roser
    Footnotes
    $ Siddharth Pahwa, Ruth Carrico, Julio Ramirez, Sean P. Clifford, Lynn Roser, Maiying Kong, and Jiapeng Huang are Senior Authors.
    ^ Lynn Roser, Maiying Kong and Jiapeng Huang contributed equally to this work.
    Affiliations
    School of Nursing, University of Louisville, Louisville, USA
    Search for articles by this author
  • Author Footnotes
    $ Siddharth Pahwa, Ruth Carrico, Julio Ramirez, Sean P. Clifford, Lynn Roser, Maiying Kong, and Jiapeng Huang are Senior Authors.
    ,
    Author Footnotes
    ^ Lynn Roser, Maiying Kong and Jiapeng Huang contributed equally to this work.
    Maiying Kong
    Footnotes
    $ Siddharth Pahwa, Ruth Carrico, Julio Ramirez, Sean P. Clifford, Lynn Roser, Maiying Kong, and Jiapeng Huang are Senior Authors.
    ^ Lynn Roser, Maiying Kong and Jiapeng Huang contributed equally to this work.
    Affiliations
    Department of Bioinformatics and Biostatistics, School of Public Health and Information Sciences, University of Louisville, Louisville, USA.
    Search for articles by this author
  • Author Footnotes
    $ Siddharth Pahwa, Ruth Carrico, Julio Ramirez, Sean P. Clifford, Lynn Roser, Maiying Kong, and Jiapeng Huang are Senior Authors.
    ,
    Author Footnotes
    ^ Lynn Roser, Maiying Kong and Jiapeng Huang contributed equally to this work.
    Jiapeng Huang
    Correspondence
    Corresponding author: Jiapeng Huang MD, PhD, Department of Anesthesiology & Perioperative Medicine, University of Louisville, 530 South Jackson Street, Louisville, KY 40202, USA. Tel.: 502-852-8157.
    Footnotes
    $ Siddharth Pahwa, Ruth Carrico, Julio Ramirez, Sean P. Clifford, Lynn Roser, Maiying Kong, and Jiapeng Huang are Senior Authors.
    ^ Lynn Roser, Maiying Kong and Jiapeng Huang contributed equally to this work.
    Affiliations
    Division of Infectious Diseases, Center of Excellence for Research in Infectious Diseases (CERID), University of Louisville, Louisville, KY, USA

    Department of Anesthesiology & Perioperative Medicine, University of Louisville, Louisville, KY, USA

    Department of Pharmacology & Toxicology, University of Louisville, Louisville, KY, USA.

    Department of Cardiovascular & Thoracic Surgery, University of Louisville, Louisville, USA.

    Center for Integrative Environmental Health Sciences, University of Louisville, Louisville, KY, USA
    Search for articles by this author
  • the Center of Excellence for Research in Infectious Diseases (CERID) Coronavirus Study Group on behalf of the COVID-19 CardioVascular Research Group (COVID-CVRG)
  • Author Footnotes
    # Pavani Nathala, Vidyulata Salunkhe, Harideep Samanapally, Qian Xu contributed equally and are co-First Authors.
    $ Siddharth Pahwa, Ruth Carrico, Julio Ramirez, Sean P. Clifford, Lynn Roser, Maiying Kong, and Jiapeng Huang are Senior Authors.
    ^ Lynn Roser, Maiying Kong and Jiapeng Huang contributed equally to this work.
Published:January 13, 2022DOI:https://doi.org/10.1053/j.jvca.2022.01.011

      Abstract

      Objectives

      Electrocardiographic (ECG) changes have been associated with COVID-19 severity. However, the progression of ECG findings in COVID-19 patients has not been studied. The purpose of this study was to describe ECG features at different stages of COVID-19 cardiovascular (CV) events and to examine the effects of specific ECG parameters and cardiac-related biomarkers on clinical outcomes in COVID-19.

      Design

      Retrospective, cohort study

      Setting

      Major tertiary care medical centers and community hospitals in Louisville, KY, USA

      Participants

      124 COVID-19 patients who suffered cardiovascular events during hospitalization.

      Interventions

      None

      Measurements and Main Results

      12 lead ECG parameters, biomarkers of cardiac injuries, and clinical outcomes were analyzed with Spearman's correlation coefficients and Kruskal–Wallis one-way ANOVA. Atrial fibrillation/atrial flutter was more frequent in ECG obtained at the time of CV event when compared to admission ECG (9.5% vs. 26.9%; P = .007). Sinus tachycardia was higher in the last available hospital ECG than CV event ECG (37.5% vs. 20.4%; P = .031). Admission ECG QTc was significantly associated with admission troponin levels (R = 0.52, P < .001). The last available hospital ECG showed non-survivors had longer QRS duration than survivors (114.6 vs. 91.2 ms; P = .026), and higher heart rate was associated with longer ICU length of stay (Spearman ρ = 0.339, P = .032).

      Conclusions

      In hospitalized COVID-19 patients who suffered cardiovascular events, ECGs at various stages of COVID-19 hospitalization showed significantly different features with dissimilar clinical outcome correlations.

      Keywords

      Abbreviations:

      ANOVA (analysis of variance), ARDS (acute respiratory distress syndrome), AV (atrioventricular), BNP (B-type natriuretic peptide), CERID (Center of Excellence for Research in Infectious Diseases), COVID-19 (coronavirus disease 2019), CT (computed tomography), CV (cardiovascular), ECG (electrocardiographic), EHR (electronic health record), ICU (intensive care unit), LOS (length of stay), LV (left ventricular), LVH (left ventricular hypertrophy), NT-pro BNP (N-terminal pro B-type natriuretic peptide), OR (odds ratio), QTc (corrected QT interval), RR (relative risk), RT-PCR (reverse transcriptase-polymerase chain reaction), RVH (right ventricular hypertrophy), SARS-CoV-2 (severe acute respiratory syndrome coronavirus 2), SD (standard deviation)
      Introduction
      The coronavirus disease 2019 (COVID-19) pandemic continues to be the leading cause of mortality and morbidity around the world with over 229 million infections and over 4.7 million deaths as of September 21, 2021.1 In the United States alone, over 40 million patients have been infected, and over 670,000 people have died.2 Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the novel strain of coronavirus that causes COVID-19, is most notably known to cause severe respiratory disease.3 However, COVID-19 is also known to cause multi-organ dysfunction, with the cardiovascular system being a significant target.4-7 Emerging data suggest that cardiac complications occur in as many as 40% of hospitalized patients.6 Those hospitalized with severe COVID-19 who develop cardiovascular complications have higher mortality, longer length of stay (LOS), and worse prognosis than those who do not experience cardiac complications.8,9
      A multitude of heart pathologies have been described in association with SARS-CoV-2 infections. Implicated mechanisms of cardiac involvement include: myocardial ischemia secondary to hypoxic state, COVID-19-associated hypercoagulability and resultant microvascular thrombotic complications, cytokine storm, direct viral inflammatory effect, and drug-induced cardiac damage.10-12 Electrocardiographic (ECG) changes have been correlated with infection severity.7,13-15 Clinical outcome measures, including admission to intensive care14, mechanical ventilation requirement14, acute respiratory distress syndrome (ARDS)7,15-17, acute renal failure requiring renal replacement therapy7,14, and all-cause mortality14,16,17 were higher in patients with abnormal ECGs.
      The Center of Excellence for Research in Infectious Diseases (CERID) recently analyzed 702 adult COVID-19 patients hospitalized between March 2020 and June 2020 in a United States metropolitan area and identified 124 patients who suffered cardiovascular (CV) events.18 Patients with cardiovascular events had a much higher mortality rate at 45.2% than those without cardiovascular events at 8.7%.18 The null hypothesis for this project was there is no significant difference between ECGs at admission, CV events and discharge (death or out of hospital). The purpose of this study was to investigate the ECG features in COVID-19 patients who suffered CV events at three distinct phases of hospitalization: upon admission, during the acute CV event, and on approaching discharge or death.
      Methods
      Study Design and Setting
      The clinical data used in this study was from the Burden of COVID-19 study, an observational retrospective cohort database maintained by Center of Excellence for Research in Infectious Diseases (CERID) at the University of Louisville.18,19,20 The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines for reporting observational studies is followed and the checklist is attached in supplemental material. The primary aim was to describe ECG features at different stages of hospitalization during COVID-19 CV events. The secondary aim was to examine the effects of specific ECG parameters and cardiac-related biomarkers on clinical outcomes in COVID-19.
      Human Subjects Protection
      The Burden of COVID-19 Study, including subsequent research using the database, was approved by the University of Louisville Institutional Review Board (# 20-0257). Information from the patients’ electronic health records (EHR) was entered into a secure database (REDCap) and is compliant with the Health Insurance Portability and Accountability Act. Patient data is secure through the use of standard data security measures and approved by the IRBs of all participating hospitals to protect and safeguard the private healthcare information of all patients.
      Patient Cohort and Study Setting
      Data from 702 adult inpatients hospitalized with COVID-19 from March 9, 2020, to June 20, 2020, at all nine adult hospitals in the Louisville metro area were reviewed. All hospitalized patients with a diagnosis of COVID-19—as defined by a positive reverse transcriptase-polymerase chain reaction (RT-PCR) on the first or the repeat test and/or ground glass opacities on a chest computerized tomography (CT)—met the inclusion criteria for the study. COVID-19 patients seen in the emergency department who were not admitted as a hospital inpatient were excluded from this study.
      Data Collection
      Data were abstracted from the EHRs of hospitalized patients diagnosed with COVID-19. A comprehensive data abstraction instrument was developed by physicians, nurses, epidemiologists, biostatisticians, public health professionals, and research assistants who were members of the CERID study group. Trained researchers (PN, VS, HS) collected information from the patients’ EHRs on COVID-19 test results, past medical history, social history, symptoms of current illness, medications, physical examination, management and therapies, radiologic and laboratory data, clinical course of hospitalization, complications, and outcomes. Only CV events occurring after admission to the hospital were considered in the analysis. CV events included heart failure, cardiogenic shock, acute myocardial infarction, cardiomyopathy, myocarditis, cardiac arrhythmias (including tachycardia, bradycardia, supraventricular tachycardia, atrial tachycardia, and bundle-branch blocks), cerebrovascular events, pulmonary embolism, pulmonary edema, deep vein thrombosis, and cardiac arrest. Clinical diagnoses of CV events were made independently by individual physicians at each site. We planned to use sensitive cardiac injury/function biomarkers (troponin and B-type natriuretic peptide (BNP)) as short-term outcome measures and mortality, hospital length of stay (LOS), and intensive care unit (ICU) LOS as long-term outcome measures.
      ECG data assumed normal calibration of ECG paper at 25 millimeters (mm) per second and 1 millivolt (mV) equivalent to 10 millimeters (mm). Results from ECGs were collected from EHRs through abstraction of confirmed 12-lead ECG results by experienced researchers. First ECG was defined as the earliest available ECG in as patient's EHRs upon admission; CV event–triggered ECG was defined as the ECG within 24 hours of a clinically diagnosed CV event; final ECG was defined as the last recorded ECG before discharge or death.
      ECG abnormality was defined as any irregularity in rhythm, left ventricular hypertrophy (LVH), right ventricular hypertrophy (RVH), AV block, or ST changes. Rhythm information from ECGs included sinus rhythm, sinus tachycardia, atrial fibrillation, atrial flutter, atrioventricular (AV) blocks (none, first degree block, second degree type I heart block, second degree type II heart block, third degree heart block), left bundle branch block, and right bundle branch block. Collected ECG parameters included P-R segment duration (normal range 120–200 ms), QRS duration (normal range 80–100 ms), corrected QT interval (QTc), presence of pathological Q-wave (>40ms wide, >2mm deep, or >25% of QRS depth), ST segment elevation and depression indicative of ischemia (voltage >0.1 mV above or below baseline). Other measurements for analysis included evidence of left ventricular hypertrophy (LVH) using Sokolov-Lyon criteria (S wave depth in precordial lead V1 plus tallest R wave height in precordial leads V5-V6 >35 mm) and evidence of right ventricular hypertrophy (RVH) (right axis deviation >90 or 110 ° with larger R and smaller S waves in leads V1 and V2).21
      Laboratory data included the first and highest (peak) levels of troponin as well as BNP and N-terminal pro BNP (NT-pro BNP) levels. Measurements of BNP and NT-pro BNP were converted to one standard unit of measurement for comparison because hospitals used either BNP or NT-pro BNP, not both. BNP measurement of 1 picograms per milliliter (pg/mL) is equivalent to 0.289 picomoles per liter (pmol/L) and NT-pro BNP measurement of 1 pg/mL is equivalent to 0.118 pmol/L.22
      Clinical Outcome Variables
      Clinical outcome variables included mortality (date of death and number of days from hospital admission to mortality), hospital LOS, and ICU LOS.
      Data Analysis
      Descriptive statistics were reported; continuous variables were summarized as mean and standard deviation (SD), and categorical variables were summarized as counts and percentages. To examine the association between ECG parameters and troponin/BNP values, Spearman's correlation coefficients were used to evaluate the association between two continuous variables, and Kruskal–Wallis one way ANOVA was conducted to examine the association between a categorical variable and a continuous outcome variable.23 To examine whether ECG parameters were associated with mortality (e.g., patients discharged alive or not), Mann-Whitney U tests were used for continuous ECG parameters and Chi-squared tests or Fisher's exact tests for categorical ECG parameters. Generalized mixed effect models were used to examine whether ECG parameters changed over time. In particular, logistic mixed-effect models were used to examine the time effect for categorical ECG parameters, and linear mixed-effect models were used to examine time effect for continuous ECG parameters.24 In the generalized linear mixed effect models, each subject was considered as a random effect. The intra-individual variations were captured by the random effects, and the inter-individual variations were captured by the individual-level covariates. P-values < .05 were considered statistically significant. Statistical analyses were conducted using R version 4.0.2 software (R Foundation for Statistical Computing). These complex data analysis was carried out by a Ph.D. student in biostatistics (Qian Xu) under the supervision of Dr. Maiying Kong, Professor in Bioinformatics and Biostatistics at the University of Louisville. They were part of the study team and were coauthors of the manuscript.
      Results
      The sample population consisted of 702 hospitalized COVID-19 patients, 124 of whom suffered CV events. Among the 124 CV-event patients, 22 (17.7%) patients had heart failure, 19 (15.3%) had cardiac arrest, 14 (11.3%) had cardiogenic shock, 15 (12.1%) had acute myocardial infarction, 12 (9.7%) had pulmonary edema, 54 (43.5%) had new serious arrhythmia, 19 (15.3%) had acute worsening of long-term arrhythmia, 7 (5.6%) had cerebrovascular accidents, 8 (6.5%) had pulmonary embolism, 2 (1.6%) had myocarditis, and 6 (4.8%) had deep venous thrombosis (DVT). Pre-existing cardiac co-morbidities in hospitalized COVID-19 patients who suffered CV events and who did not suffer CV events were summarized in Supplemental Table 1. The group with CV events showed significantly higher incidences of all preexisting cardiac co-morbidities than the group without CV events except DVT. Among the 124 patients, eighty-six patients required ICU care. Fifteen patients received non-invasive mechanical ventilation, and 69 patients needed invasive mechanical ventilation (IMV).
      Table 1 summarized the ECG parameters in first, CV event–triggered, and final ECGs. Normal sinus rhythm was significantly more common in the first ECG compared to CV event–triggered ECG (59.5% vs. 42.6%; P = .023). Atrial fibrillation or atrial flutter was more frequent in CV event–triggered ECG compared to first ECG (9.5% vs. 26.9%; P = .007). Sinus tachycardia was found to be significantly more frequent in final ECG than in CV event–triggered ECG (37.5% vs. 20.4%; P = .031). Interestingly, we found that QRS duration was significantly longer in CV event–triggered ECG than in final ECG (107.364 ± 30.2 ms vs. 105.59 ± 30.3 ms; P = .047), and the rate of AV block was significantly higher in CV event–triggered ECG than in final ECG (31.5% vs. 17.5%; P = .018).
      Table 1Summarized ECG parameters in the first ECGs, event triggered (ET) ECGs and final ECGs, respectively.
      First ECGET ECGFinal ECGP-value forfirst vs ETP-value forfinal vs ET
      Sample sizesN=74N=108N=40
      Rhythm, n (%)
      Normal Sinus44 (59.5%)46 (42.6%)17 (42.5%)0.023
      p-value<0.05
      0.919
      Atrial Fibrillation/Flutter7 (9.5%)29 (26.9%)5 (12.5%)0.007
      p-value<0.05
      0.167
      Sinus Tachycardia20 (27.0%)22 (20.4%)15 (37.5%)0.2850.031
      p-value<0.05
      Other3 (4.1%)11(10.2%)3 (7.5%)0.1550.056
      Heart Rate, mean (SD) (bpm)94 (22.7)100.31(30.3)95.9 (24)0.0810.530
      P-R interval, mean (SD) (ms)169.3 (73.3)156.9 (35.5)144.9 (38)0.0960.268
      QTc, mean (SD) (ms)457.7 (38.6)469.4 (40.7)463.8 (46.7)0.0800.317
      QRS Duration, mean (SD) (ms)104.0 (26.7)107.4 (30.2)105.6 (30.3)0.8980.047
      p-value<0.05
      Q-Wave, n (%)0 (0%)1 (0.9%)0 (0%)0.9910.996
      LVH, n (%)8 (10.8%)9 (8.3%)3 (7.5%)0.1030.695
      RVH, n (%)1 (1.4%)0 (0%)0 (0%)0.9990.995
      AV Block, n (%)21 (28.4%)34 (31.5%)7 (17.5%)0.4410.018
      p-value<0.05
      ST Changes, n (%)3 (4.1%)10 (9.3%)6 (15%)0.2590.252
      ET – event-triggered, LVH – left ventricular hypertrophy, RVH – right ventricular hypertrophy, AV – atrioventricular
      low asterisk p-value<0.05
      Table 2, Supplemental Table 2 and Supplemental Table 3 showed the associations between the ECG parameters of first, event-triggered, and final ECG and mortality, ICU LOS, and hospital LOS. For first ECG, hospital and ICU LOS were significantly longer for patients with normal sinus rhythm than those with atrial fibrillation or atrial flutter (11.45 vs 7.61 days for hospital LOS and 6.75 vs. 0.62 days for ICU LOS (Table 4), respectively). For final ECG, non-survivors had much longer QRS duration than survivors (114.6 vs. 91.2 ms; P = .026, Supplemental Table 2), and a higher heart rate was significantly associated with longer ICU LOS with a Spearman correlation coefficient of 0.339 (P = .032, Supplemental Table 3).
      Table 2Summary of statistics between ECG parameters at each time point (First, Event Triggered (ET) and Final ECG) and outcome variables in terms of intensive care unit (ICU) length of stay (LOS), and hospital LOS.
      First ECGEvent-triggered ECGFinal ECG
      nHospital LOS:Median (IQR)p-valuenHospital LOS:Median (IQR)p-valuenHospital LOS:Median (IQR)p-value
      All Rhythms:7411.29
      overall median (IQR)


      (7.42, 18.18)
      1089.02
      overall median (IQR)


      (5.87, 15.24)
      4010.29
      overall median (IQR)


      (6.16, 17.95)
      Normal Sinus4411.45

      (7.13, 18.09)
      0.004
      overall p-value comparing four different rhythms


      p-value<0.05.
      468.63

      (4.79, 14.51)
      0.668
      overall p-value comparing four different rhythms
      178.90

      (5.17, 17.86)
      0.7868
      overall p-value comparing four different rhythms
      Atrial Fibrillation /Flutter77.61

      (5.97, 7.70)
      0.012
      p-value compared to normal sinus rhythm


      p-value<0.05.
      2910.94

      (7.35, 16.45)
      0.336
      p-value compared to normal sinus rhythm
      511.33

      (8.92, 12.02)
      0.7616
      p-value compared to normal sinus rhythm
      Sinus Tachycardia2014.87

      (9.34, 23.51)
      0.159
      p-value compared to normal sinus rhythm
      227.98

      (5.34, 14.66)
      0.964
      p-value compared to normal sinus rhythm
      1510.25

      (7.53, 34.43)
      0.331
      p-value compared to normal sinus rhythm
      Other35.65

      (5.36, 7.42)
      0.088
      p-value compared to normal sinus rhythm
      119.18

      (6.10, 18.91)
      0.407
      p-value compared to normal sinus rhythm
      310.32

      (7.26, 17.19)
      0.999
      p-value compared to normal sinus rhythm
      nICU LOS: Median (IQR)p-valuenICU LOS: Median (IQR)p-valuenICU LOS: Median (IQR)p-value
      All Rhythms:746.10
      p-value compared to normal sinus rhythm


      (1.91, 12.33)
      1084.60
      p-value compared to normal sinus rhythm


      (0, 11.11)
      408.11
      p-value compared to normal sinus rhythm


      (3.54, 13.22)
      Normal Sinus446.75

      (1.82, 12.94)
      0.008
      overall p-value comparing four different rhythms


      p-value<0.05.
      464.74

      (0,11.12)
      0.917
      overall p-value comparing four different rhythms
      174.43

      (0.72, 11.05)
      0.261
      overall p-value comparing four different rhythms
      Atrial Fibrillation /Flutter70.62

      (0, 2.64)
      0.014
      p-value compared to normal sinus rhythm


      p-value<0.05.
      292.80

      (0,9.82)
      0.573
      p-value compared to normal sinus rhythm
      511.14

      (7.08, 11.90)
      0.271
      p-value compared to normal sinus rhythm
      Sinus Tachycardia2010.28

      (5.98, 16.27)
      0.161
      p-value compared to normal sinus rhythm
      225.31

      (0, 11.51)
      0.857
      p-value compared to normal sinus rhythm
      159.13

      (5.92, 32.17)
      0.059
      p-value compared to normal sinus rhythm
      Other32.5

      (1.25, 3.71)
      0.190
      p-value compared to normal sinus rhythm
      114.91

      (0.58, 8.52)
      0.918
      p-value compared to normal sinus rhythm
      310.09

      (5.62, 16.92)
      0.524
      p-value compared to normal sinus rhythm
      a overall median (IQR)
      b overall p-value comparing four different rhythms
      c p-value compared to normal sinus rhythm
      low asterisk p-value<0.05.
      First troponin values were compared between patients with and without an ECG abnormality in the first ECG. There were no significant difference in first troponin level among those with and without an ECG abnormality (Table 3). Supplemental Table 4 showed the correlation coefficients for continuous ECG parameters of the first ECG. QTc was significantly associated with the first troponin levels (R = 0.52, P < .001).
      Table 3Comparisons of the first troponin between the patients with abnormality and the patients without abnormality based on the first ECG abnormality.
      First ECG AbnormalityN(%)First Troponin with Abnormality (ng/ml)Mean±SDFirst Troponin without Abnormality (ng/ml)Mean±SDP-value
      Rhythm: AF/Flutter7(9.5%)0.109±0.10.234±0.80.2374
      Rhythm: Sinus Tachycardia20(27.0%)0.096±0.10.234±0.80.3221
      Rhythm: Others3(4.1%)0.03±00.234±0.80.1855
      LVH8(10.8%)0.082±0.10.187±0.70.6769
      RVH1(1.4%)0.03±NA0.177±0.70.7558
      AV Block21(28.4%)0.164±0.20.179±0.80.0788
      ST Changes3(4.1%)0.033±00.181±0.70.3309
      LVH – left ventricular hypertrophy, RVH – right ventricular hypertrophy, AV – atrioventricular
      Table 4 showed a comparison of peak troponin values between patients with and without ECG abnormality in rhythm, LVH, RVH, AV block, or ST changes. Peak troponin levels of patients with atrial fibrillation or atrial flutter were significantly lower than those with normal sinus rhythm (0.073 ± 0.1 vs. 0.681 ± 1.3; P = .043) (Table 4). There were no significant correlations between peak troponin levels and CV event–triggered ECG parameters (heart rate, P-R interval, QTc and QRS duration, Supplemental Table 5).
      Table 4Comparisons of the peak troponin between the patients with abnormality and those without abnormality based on the event triggered ECG abnormality.
      Event Triggered ECG AbnormalityN(%)Peak Troponin with Abnormality (ng/ml)Mean±SDPeak Troponin without Abnormality (ng/ml)Mean±SDP-value
      Rhythm: AF/Flutter29 (26.9%)0.073±0.10.681±1.30.0430*
      Rhythm: Sinus Tachycardia22(20.4%)0.614±1.40.681±1.30.7121
      Rhythm: Others11(10.2%)0.157±0.20.681±1.30.1654
      LVH9 (8.3%)0.05± NA0.506±1.10.4102
      AV Block34(31.5%)0.365 ± 0.70.556± 1.20.6647
      ST Changes10 (9.3%)0.922 ± 1.90.433 ± 10.8219
      AF – Atrial Fibrillation, LVH – left ventricular hypertrophy, AV – atrioventricular *p-value<0.05, NA-Not available due to only one observation under the conditions.
      The effect of the first troponin, peak troponin, and BNP levels on mortality, ICU LOS, and hospital LOS for all 702 patients are summarized in Table 5. Levels of first troponin (1.256 ± 11 vs. 0.116 ± 0.6 ng/ml; P < .001), peak troponin (0.389 ± 1 vs. 0.311 ± 0.8 ng/ml; P < .001), and BNP (1733.718 ± 4536.2 vs. 565.063 ± 2909 pmol/l; P = .011) were higher in patients who died in the hospital. First troponin levels were significantly correlated with both hospital LOS (R = 0.165; P < .001) and ICU LOS (R = 0.172; P < .001). Peak troponin levels were significantly correlated with ICU LOS (R = 0.301; P = .001). BNP was significant correlated with both hospital LOS (R = 0.204; P < .001) and ICU LOS (R = 0.212; P < .001).
      Table 5Association study between each one of the biomarkers (i.e., first troponin, peak troponin, BNP) and each one of the outcome variables (i.e., mortality, ICU length of stay and hospital length of stay) in 702 patients.
      Deadmean±SDAlivemean±SDp-valueSpearman Correlation(LOS)p-value(LOS)Spearman Correlation(ICU LOS)p-value(ICU LOS)
      First Troponin (ng/ml)1.256±110.116±0.6<0.001
      p-value<0.05
      0.165<0.001
      p-value<0.05
      0.172<0.001
      p-value<0.05
      Peak Troponin (ng/ml)0.389±10.311±0.80.011
      p-value<0.05
      0.170.072
      p-value<0.05
      0.3010.001
      p-value<0.05
      BNP

      (pmol/l)
      1733.718±4536.2565.063± 2909<0.001
      p-value<0.05
      0.204<0.001
      p-value<0.05
      0.212<0.001
      p-value<0.05
      ICU- intensive care unit, LOS- length of stay
      low asterisk p-value<0.05
      Associations of first troponin, peak troponin, and BNP with mortality, ICU LOS, and hospital LOS among the 124 patients with CV events are summarized in Supplemental Table 6. However, no significant correlations with any of these three main clinical outcomes were observed.
      Discussion
      Comprehensive analysis of ECGs in 124 patients who suffered CV events demonstrated distinct features of the first ECG (on admission), CV event–triggered ECG (within 24 hours of CV event), and final ECG (last ECG before discharge or death). Atrial fibrillation/flutter was the most common arrhythmia during CV events, and sinus tachycardia was the most common rhythm in the final ECG. QRS duration was longer, and AV block was more commonly present in the CV event–triggered ECGs.
      Two systematic reviews of ECG findings in COVID-19 patients have been published recently.25, 26 Mehraeen et al included 20 articles and found ST-T abnormalities—notably ST elevation—were the most observed ECG findings, but the relationship with myocardial injuries is debatable. Garcia-Zamora et al performed a meta-analysis of 28 studies with 12,499 participants and found that the overall prevalence of cardiac arrhythmias was 10.3%, with the most common arrhythmias documented during hospitalization being supraventricular arrhythmias (6.2%), followed by ventricular arrhythmias (2.5%). The incidence of cardiac arrhythmias was higher among critically ill patients (relative risk [RR], 12.1; 95% confidence interval [CI], 8.5-17.3) and among non-survivors (RR, 3.8; 95% CI, 1.7-8.7). However, none of these studies analyzed ECG parameters based on the timing of ECG or how ECGs change during hospitalization in COVID-19 patients. This information is critical to understand how cardiac injuries evolve in hospitalized COVID-19 patients. Therefore, this study separated ECGs into three important phases—admission, within 24 hours of CV events, and near discharge/death—in order to study distinct ECG features and their progression. Because troponin level reflects and quantifies acute myocardial damage, we decided to correlate the first ECG parameters with the first troponin and the CV event–triggered ECG parameters with peak troponin, indicative of the most severe cardiac injuries during hospitalization. This analysis might provide insights into how cardiovascular injuries occur and develop during hospitalization in COVID-19 patients.
      ECG abnormalities commonly seen in cardiac injury are arrhythmia, ST elevation, PR depression, new-onset bundle branch block, QT prolongation, pseudoinfarct pattern, premature ventricular complexes, bradyarrhythmias, and ventricular tachycardia. Consistent with previous studies25, 26, this study found atrial fibrillation or flutter to be the most common arrhythmia during CV event–triggered ECG (26.9%). However, atrial fibrillation or flutter was much lower on first ECG (9.5%) and final ECG (12.5%). A recent COVID-19 study found atrial fibrillation in 7.0% of the 201 patients at admission, and atrial fibrillation was associated with increased mortality (odds ratio [OR], 12.74; 95% CI, 3.65-44.48; P < .001).27 Other studies have shown that atrial fibrillation or flutter was the most frequently reported serious arrhythmia14,17, was associated with higher troponin values, and carried a higher mortality rate than other rhythms.17 In the present study, AV block was also found be significantly higher in CV event–triggered ECG (31.5%) compared to admission or discharge ECGs. Both atrial arrhythmia and AV block could represent direct viral infection of the conduction system, myocardial ischemia/edema, diastolic dysfunction, left ventricular (LV) dysfunction, pulmonary hypertension from ARDS, or pulmonary embolism. For the first ECG, atrial arrhythmia was found to be correlated with shorter hospital and ICU LOS. However, no correlation was found between CV event–triggered or final ECG and either hospital or ICU LOS. One possible explanation could be early mortality with atrial arrhythmia.
      QRS duration longer than 120 ms was associated with worse clinical outcomes and higher levels of myocardial injury biomarkers in COVID-19.28 Prolonged QRS could reflect active myocardial injury during CV events caused by the direct impact of SARS-CoV-2 on the cells.29 A wide QRS complex has been previously associated with increased mortality risk in non-COVID-19 patients.30, 31, 32 Our findings further confirm the importance of QRS duration in predicting mortality in COVID-19 patients. CV event–triggered ECG demonstrated significantly longer QRS duration than that on the first or final ECGs. Furthermore, this study found that QRS duration on final ECG was significantly longer in non-survivors than survivors among hospitalized COVID-19 patients with CV events.
      Among all the parameters reviewed on the first ECGs, only QTc showed a significant correlation with the first troponin. A recent meta-analysis found that the prevalence of QTc > 500 ms was 12.3% in COVID-19 patients.25 QTc interval prolongation was found to be associated with increased COVID-19 severity and mortality.33, 34, 35, 36, 37 However, the risk of torsades de pointes was not increased in hospitalized COVID-19 patients who showed a marked prolongation of QTc interval.38, 39 In our study, we did not find any correlation between ECG parameters and peak troponin levels among CV event–triggered ECGs. This might suggest that the CV event–triggered ECG parameters are not useful predictors of further outcomes.
      It has been demonstrated that ECG ST-T segment alterations are the most commonly reported ECG anomaly in patients with COVID-19 23,24 and are more frequent among patients with severe COVID-19.13, 40, 41 Barman et al showed that ST-T changes on admission ECG were closely associated with the severity of COVID-19 infection.13 The present study found ST changes on the first ECGs in 4.1% of patients who suffered CV events and on CV event–triggered ECGs in 9.3%. However, there was no significant correlation with either first or peak troponin.
      Sinus tachycardia was reported to be the most common arrhythmia in patients with COVID-1913,14,17,33 and was more frequent in non-survivors. This study found that heart rate on the final ECG was significantly correlated with ICU LOS. This could indicate higher levels of stress due to sepsis, hypovolemia, pulmonary embolism, anxiety, or mechanical ventilation.
      Research on ECG alterations and hospital LOS is limited. Abrams et al compared the characteristics of patients with COVID-19 who died of arrhythmias to those who died of other causes and there was no statistically significant difference in LOS between the two groups (median 5 vs. 4 days; P = .76).42
      In addition to ECG findings, cardiac involvement can be detected by multiple laboratory markers, including troponin and BNP. Higher BNP values have been associated with abnormal ECG14 and higher in-hospital mortality.17 Troponin levels were higher in non-survivors than survivors33, as well as in patients with severe illness compared to those with non-severe COVID-19.40 For the entire patient cohort (patients with and without CV events), the present study found that the first troponin, peak troponin, and BNP were all significantly associated with mortality and ICU LOS. First troponin and BNP were also significantly associated with hospital LOS. However, we did not find any significant correlations between first troponin, peak troponin, or BNP and mortality, ICU LOS, or hospital LOS among patients with CV events. These results were surprising yet explicable. If cardiac injury has been ongoing, additional troponin and BNP values could not help to further predict clinical outcomes. Another explanation could be patients who suffered CV events shared common risk factors that are more dominant to determine the clinical outcomes. It could also be possible that SARS-CoV-2 virus induces myocardial injury through a different mechanism which is not dependent on the level of cardiac biomarkers. This might imply that clinicians should reconsider ordering additional troponin and BNP tests as they did not effectively predict outcomes in COVID-19 CV event patients.
      There are limitations to our study. The results of this retrospective study may not be generalizable to all persons with COVID-19, as they may have significantly different characteristics to patients hospitalized with COVID-19; hence, the external validity of our study is limited to hospitalized patients. Studies with retrospective design may bias the true incidence and influence of ECG abnormalities as well as their prognostic value as predictors of clinical outcomes. To reduce these confounders and risk of bias in future research, large-scale prospective studies are needed to determine whether ECG abnormalities play an important role in predicting adverse clinical COVID-19 outcomes.
      Conclusions
      The simplicity and availability of the 12-lead ECG makes it a potentially valuable predictive tool in the risk stratification and management of COVID -19 patients. Our study sheds light on the importance of ECG findings in hospitalized patients who suffered CV events; we found ECGs at various stages of COVID-19 hospitalization showed significantly different features with dissimilar clinical outcome correlations.
      Acknowledgments
      The authors acknowledge the excellent efforts of University of Louisville CERID group and Kornhauser Health Sciences Library to assist with this project.
      Author Contributions
      LR, MK, JH had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis, including and especially any adverse effects. PN, VS, HS, QX, SF, OFF, FD, AG, TM, DCG, JP, TA, DT, RC, JR, DA, SPC, SP contributed substantially to the study design, data analysis and interpretation, and the writing of the manuscript.
      Harideep Samanapally, Pavani Nathala, Vidyulata Salunkhe: study design, data collection, manuscript preparation.
      Qian Xu: study design, statistical analysis, manuscript preparation.
      Stephen Furmanek: study design, statistical analysis, manuscript preparation.
      Omar F. Fahmy MD, Fnu Deepti, Trevor McGuffin, Dylan C. Goldsmith, Tshura Ali, Derek Titus: study design and manuscript writing.
      Alex Glynn, Jessica Petrey: literature review, study design, manuscript preparation.
      Ruth Carrico, Julio Ramirez, Demetra Antimisiaris, Sean P. Clifford, Siddharth Pahwa, Lynn Roser, Maiying Kong, Jiapeng Huang: Study design, data analysis, manuscript preparation.
      Data collection was performed by the Center of Excellence for Research in Infectious Diseases (CERID) group at the University of Louisville Division of Infectious Diseases. Statistical analysis and draft writing were performed by Qian Xu under the guidance of Maiying Kong. All authors have read and approved the manuscript.
      Funding
      Research reported in this publication was supported by the National Center For Advancing Translational Sciences of the National Institutes of Health under Award Number U18TR003787, NIH P30 (P30ES030283) grant, and Gilead Sciences COMMIT COVID-19 RFP Program grant (Gilead IN-US-983-6063). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health or Gilead Sciences.
      Declaration of Competing Interests
      The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
      References
      1. Center for Systems Science and Engineering (CSSE) at Johns Hopkins University (JHU). COVID-19 Dashboard. 2021; https://coronavirus.jhu.edu/map.html. Accessed September 21, 2021.
      2. Centers for Disease Control and Prevention. United States COVID-19 Cases, Deaths, and Laboratory Testing (NAATs) by State, Territory, and Jurisdiction. 2021; https://covid.cdc.gov/covid-data-tracker/#cases_casesper100klast7days. Accessed September 11, 2021, 2021.
      3. Wu L, O'Kane AM, Peng H, Bi Y, Motriuk-Smith D, Ren J. SARS-CoV-2 and cardiovascular complications: From molecular mechanisms to pharmaceutical management. Biochem Pharmacol. 2020;178:114114.
      4. Inciardi RM, Lupi L, Zaccone G, et al. Cardiac Involvement in a Patient With Coronavirus Disease 2019 (COVID-19). JAMA Cardiol. 2020;5(7):819-824.
      5. Lala A, Johnson KW, Januzzi JL, et al. Prevalence and Impact of Myocardial Injury in Patients Hospitalized With COVID-19 Infection. J Am Coll Cardiol. 2020;76(5):533-546.
      6. Madjid M, Safavi-Naeini P, Solomon SD, Vardeny O. Potential Effects of Coronaviruses on the Cardiovascular System: A Review. JAMA Cardiol. 2020;5(7):831-840.
      7. Shi S, Qin M, Shen B, et al. Association of Cardiac Injury With Mortality in Hospitalized Patients With COVID-19 in Wuhan, China. JAMA Cardiol. 2020;5(7):802-810.
      8. Karbalai Saleh S, Oraii A, Soleimani A, et al. The association between cardiac injury and outcomes in hospitalized patients with COVID-19. Intern Emerg Med. 2020;15(8):1415-1424.
      9. Kotecha T, Knight DS, Razvi Y, et al. Patterns of myocardial injury in recovered troponin-positive COVID-19 patients assessed by cardiovascular magnetic resonance. Eur Heart J. 2021;42(19):1866-1878.
      10. Bavishi C, Bonow RO, Trivedi V, Abbott JD, Messerli FH, Bhatt DL. Special Article - Acute myocardial injury in patients hospitalized with COVID-19 infection: A review. Prog Cardiovasc Dis. 2020;63(5):682-689.
      11. Li L, Zhou Q, Xu J. Changes of Laboratory Cardiac Markers and Mechanisms of Cardiac Injury in Coronavirus Disease 2019. Biomed Res Int. 2020;2020:7413673.
      12. Lippi G, Lavie CJ, Sanchis-Gomar F. Cardiac troponin I in patients with coronavirus disease 2019 (COVID-19): Evidence from a meta-analysis. Prog Cardiovasc Dis. 2020;63(3):390-391.
      13. Barman HA, Atici A, Alici G, et al. The effect of the severity COVID-19 infection on electrocardiography. Am J Emerg Med. 2021;46:317-322.
      14. Bergamaschi L, D'Angelo EC, Paolisso P, et al. The value of ECG changes in risk stratification of COVID-19 patients. Ann Noninvasive Electrocardiol. 2021;26(3):e12815.
      15. Giustino G, Croft LB, Stefanini GG, et al. Characterization of Myocardial Injury in Patients With COVID-19. J Am Coll Cardiol. 2020;76(18):2043-2055.
      16. McCullough SA, Goyal P, Krishnan U, Choi JJ, Safford MM, Okin PM. Electrocardiographic Findings in Coronavirus Disease-19: Insights on Mortality and Underlying Myocardial Processes. J Card Fail. 2020;26(7):626-632.
      17. Poterucha TJ, Elias P, Jain SS, et al. Admission Cardiac Diagnostic Testing with Electrocardiography and Troponin Measurement Prognosticates Increased 30-Day Mortality in COVID-19. J Am Heart Assoc. 2021;10(1):e018476.
      18. Xu Q, Samanapally H, Nathala P, et al. Outcomes and Risk Factors for Cardiovascular Events in Hospitalized COVID-19 Patients. J Cardiothorac Vasc Anesth. 2021.
      19. Carrico R, Ali TsS, Hill M, et al. Healthcare Workers Hospitalized with COVID-19: Outcomes from the Burden of COVID-19 study at the University of Louisville Center of Excellence for Research in Infectious Diseases (CERID). University of Louisville Journal of Respiratory Infections. 2020;4(1):Article 75.
      20. Carrico R, Balcom D, Johnson W, et al. Implementation of the Louisville COVID-19 Surveillance Protocol: Experiences from the University of Louisville Center of Excellence for Research in Infectious Diseases (CERID). University of Louisville Journal of Respiratory Infections. 2020;4(1):Article 10.
      21. Surawicz B, Knilans T. Chou's Electrocardiography in Clinical Practice: Adult and Pediatric. 6th ed . Philadelphia, PA: Saunders, 2008.
      22. Weber M, Hamm C. Role of B-type natriuretic peptide (BNP) and NT-proBNP in clinical routine. Heart. 2006 Jun;92(6):843-9.)
      23. Altman DG. Practical statistics for medical research. Boca Raton, Fla: Chapman & Hall/CRC; 1999.
      24. Fitzmaurice GM, Laird NM, Ware JH. Applied Longitudinal Analysis. Hoboken: Wiley; 2012.
      25. Garcia-Zamora S, Lee S, Haseeb S, et al. Arrhythmias and electrocardiographic findings in Coronavirus disease 2019: A systematic review and meta-analysis. Pacing Clin Electrophysiol. 2021;44(6):1062-1074.
      26. Mehraeen E, Seyed Alinaghi SA, Nowroozi A, et al. A systematic review of ECG findings in patients with COVID-19. Indian Heart J. 2020;72(6):500-507.
      27. Denegri A, Pezzuto G, D'Arienzo M, et al. Clinical and electrocardiographic characteristics at admission of COVID-19/SARS-CoV2 pneumonia infection. Intern Emerg Med. 2021;16(6):1451-1456.
      28. Sonsoz MR, Oncul A, Cevik E, et al. Wide QRS Complex and Lateral ST-T Segment Abnormality Are Associated With Worse Clinical Outcomes in COVID-19 Patients. Am J Med Sci. 2021;361(5):591-597.
      29. Centurión OA. Wide QRS Complex and Left Ventricular Lateral Repolarization Abnormality: The Importance of ECG Markers on Outcome Prediction in Patients with COVID-19. Am J Med Sci. 2021;362(1):1-2.
      30. Ukena C, Mahfoud F, Kindermann I, Kandolf R, Kindermann M, Böhm M. Prognostic electrocardiographic parameters in patients with suspected myocarditis. Eur J Heart Fail. 2011;13(4):398-405.
      31. Kalra PR, Sharma R, Shamim W, et al. Clinical characteristics and survival of patients with chronic heart failure and prolonged QRS duration. Int J Cardiol. 2002;86(2-3):225-231.
      32. Horwich T, Lee SJ, Saxon L. Usefulness of QRS prolongation in predicting risk of inducible monomorphic ventricular tachycardia in patients referred for electrophysiologic studies. Am J Cardiol. 2003;92(7):804-809.
      33. Kunal S, Sharma SM, Sharma SK, et al. Cardiovascular complications and its impact on outcomes in COVID-19. Indian Heart J. 2020;72(6):593-598.
      34. Alareedh M, Nafakhi H, Shaghee F, Nafakhi A. Electrocardiographic markers of increased risk of sudden cardiac death in patients with COVID-19 pneumonia. Ann Noninvasive Electrocardiol. 2021;26(3):e12824.
      35. Ding J, Liu W, Guan H, et al. Corrected QT interval in hospitalized patients with coronavirus disease 2019: Focus on drugs therapy. Medicine (Baltimore). 2021;100(28):e26538.
      36. Bessière F, Roccia H, Delinière A, et al. Assessment of QT Intervals in a Case Series of Patients With Coronavirus Disease 2019 (COVID-19) Infection Treated With Hydroxychloroquine Alone or in Combination With Azithromycin in an Intensive Care Unit. JAMA Cardiol. 2020;5(9):1067-1069.
      37. Farré N, Mojón D, Llagostera M, et al. Prolonged QT Interval in SARS-CoV-2 Infection: Prevalence and Prognosis. J Clin Med. 2020;9(9).
      38. Mercuro NJ, Yen CF, Shim DJ, et al. Risk of QT Interval Prolongation Associated With Use of Hydroxychloroquine With or Without Concomitant Azithromycin Among Hospitalized Patients Testing Positive for Coronavirus Disease 2019 (COVID-19). JAMA Cardiol. 2020;5(9):1036-1041.
      39. Jankelson L, Karam G, Becker ML, Chinitz LA, Tsai MC. QT prolongation, torsades de pointes, and sudden death with short courses of chloroquine or hydroxychloroquine as used in COVID-19: A systematic review. Heart Rhythm. 2020;17(9):1472-1479.
      40. Deng Q, Hu B, Zhang Y, et al. Suspected myocardial injury in patients with COVID-19: Evidence from front-line clinical observation in Wuhan, China. Int J Cardiol. 2020;311:116-121.
      41. Antwi-Amoabeng D, Beutler BD, Singh S, et al. Association between electrocardiographic features and mortality in COVID-19 patients. Ann Noninvasive Electrocardiol. 2021;26(4):e12833.
      42. Abrams MP, Wan EY, Waase MP, et al. Clinical and cardiac characteristics of COVID- 19 mortalities in a diverse New York City Cohort. J Cardiovasc Electrophysiol. 2020;31(12):3086-3096.

      Appendix. Supplementary materials